Autophagy is required for ectoplasmic specialization assembly in ...

AUTOPHAGY 2016, VOL. 12, NO. 5, 814–832 http://dx.doi.org/10.1080/15548627.2016.1159377

BASIC RESEARCH PAPER

Autophagy is required for ectoplasmic specialization assembly in sertoli cells Chao Liua,b, Hongna Wanga,b, Yongliang Shanga,b, Weixiao Liua, Zhenhua Songa,b, Haichao Zhaoa,b, Lina Wanga,b, Pengfei Jiac, Fengyi Gaoa, Zhiliang Xua,b, Lin Yangc, Fei Gaoa,b, and Wei Lia,b a State Key Laboratory of Stem Cell and Reproductive Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China; bUniversity of Chinese Academy of Sciences, Beijing, China; cState Key Laboratory of Molecular Developmental Biology and National Center for Plant Gene Research (Beijing), Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing, China

ABSTRACT

ARTICLE HISTORY

The ectoplasmic specialization (ES) is essential for Sertoli-germ cell communication to support all phases of germ cell development and maturity. Its formation and remodeling requires rapid reorganization of the cytoskeleton. However, the molecular mechanism underlying the regulation of ES assembly is still largely unknown. Here, we show that Sertoli cell-specific disruption of autophagy influenced male mouse fertility due to the resulting disorganized seminiferous tubules and spermatozoa with malformed heads. In autophagy-deficient mouse testes, cytoskeleton structures were disordered and ES assembly was disrupted. The disorganization of the cytoskeleton structures might be caused by the accumulation of a negative cytoskeleton organization regulator, PDLIM1, and these defects could be partially rescued by Pdlim1 knockdown in autophagy-deficient Sertoli cells. Altogether, our works reveal that the degradation of PDLIM1 by autophagy in Sertoli cells is important for the proper assembly of the ES, and these findings define a novel role for autophagy in Sertoli cell-germ cell communication.

Received 29 June 2015 Revised 14 February 2016 Accepted 23 February 2016

Introduction Spermatogenesis is the process of male gamete production with successive cellular renovation and differentiation, consisting of spermatogonial mitosis, spermatocytic meiosis and spermiogenesis.1,2 During these precisely timed and highly organized events, germ cells traverse the seminiferous epithelium from the basal to the apical (adluminal) compartment by adhering to the somatic Sertoli cells.2,3 Sertoli cells play key roles in the control of spermatogenesis. Their functions include providing structural support and nourishment to developing germ cells, mediating the self-renewal and differentiation of spermatogonial stem cells (SSCs), phagocytosing degenerating germ cells, protecting the autoreactive immune response of germ cells, releasing spermatids at spermiation and hormonal regulation.3,4 At all stages of spermatogenesis, Sertoli cells and germ cells have developed a set of communication devices that are involved in attachment, displacement, degeneration, and cellcell transfer of signaling molecules and cellular materials.5 This Sertoli-germ cell communication is necessary for successful spermatogenesis. In the seminiferous epithelium, functional cell interconnections are maintained through both Sertoli-Sertoli cell and Sertoli-germ cell junctions.6,7 Besides the tight junctions and gap junctions that are also present in other epithelia, the seminiferous epithelium exhibits testis-unique anchoring junctions, including the ectoplasmic specialization (ES) and the desmosome-like junction.6,8 The ES is an actin microfilament-rich

KEYWORDS

Atg5; Atg7; autophagy; cytoskeleton organization; ectoplasmic specialization; PDLIM1

anchoring device using F-actin for attachment9 at the SertoliSertoli cell interface known as the basal ES (bES), which together with the tight junctions, gap junctions and the desmosome-like junction create the blood–testis barrier (BTB).9-12 The apical compartment of the Sertoli cell–elongating spermatid interface, called apical ES (aES), interacts with the acrosome region of the elongating spermatid and mechanically grasps the head of spermatids to undergo rapid elongation and maturation.6,7,9,11,13 The main roles of the apical ES include shaping the spermatid head, facilitating spermatid cell movement, orienting the elongated spermatid and releasing mature spermatozoa during spermiation.10,13 In the Sertoli cell-elongating spermatid interface, the apical ES consists of a narrow layer of hexagonally packed, parallel actin bundles sandwiched between the Sertoli cell plasma membrane and an endoplasmic reticulum (ER) cistern on the Sertoli cell side of the plasma membrane.14 Many cytoskeleton and cytoskeleton-related proteins are localized at the apical ES site, including Actin, TUBB, VIM, ACTR3, PLS1, KEAP1, EPS8, PALLD and ACTN1/4 (Table 1). These proteins play important roles in apical ES regulation and organization.6,9,14 The continuous transport of developing germ cells across the seminiferous epithelium requires the ES to undergo dynamic restructuring in the epithelial cycle.2,12,15,16 At the early stage of the epithelial cycle, the disassembly and reassembly of the basal ES-BTB is necessary for the transit of preleptotene spermatocytes across the BTB to enter the apical

CONTACT Wei Li [email protected] State Key Laboratory of Stem Cell and Reproductive Biology, Institute of Zoology, Chinese Academy of Sciences, Beijing, China Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/kaup. Supplemental data for this article can be accessed on the publisher’s website. © 2016 Taylor & Francis

AUTOPHAGY

Table 1. Genes and gene products mentioned in the text. Symbols ACTN1 ACTN4 ACTR3 ATG3 ATG5 ATG7 ATG12 ATG16L1 CDH2 CTNNB1 CTTN EPS8 GCNA1 H2AFX JAM3 KEAP1 MAP1LC3A MAP1LC3B NUDC OFD1 PALLD PDLIM1 PLS1 SHBG SQSTM1 STMN1 SYCP3 TUBB VCP WT1 ZP3R ZBTB16

Full name

Alias

actinin, alpha 1 actinin, alpha 4 ARP3 actin-related protein 3 Autophagy-related 3 Autophagy-related 5 Autophagy-related 7 Autophagy-related 12 Autophagy-related 16-like 1 cadherin 2 catenin (cadherin associated protein), beta 1 Cortactin epidermal growth factor receptor pathway substrate 8 germ cell nuclear antigen 1 H2A histone family, member X junction adhesion molecule 3 kelch-like ECH-associated protein 1 microtubule-associated protein 1 light chain 3 alpha microtubule-associated protein 1 light chain 3 beta nuclear distribution gene C homolog (Aspergillus) oral-facial-digital syndrome 1 gene homolog (Human) palladin, cytoskeletal associated protein PDZ and LIM domain 1 (Elfin) plastin 1 (I-isoform) sex hormone binding globulin sequestosome 1 stathmin1 synaptonemal complex protein 3 tubulin, beta valosin containing protein Wilms tumor 1 homolog zona pellucida 3 receptor zinc finger and BTB domain containing 16

a-actinin1 a-actinin4 ARP3

N-cadherin b-catenin

gH2AX Jam-C LC3A LC3B Palladin Fimbrin ABP p62 SCP3 b-tubulin p97

815

completion to form an autophagosome. Once fused with a lysosome, the inner proteins are eventually degraded in autolysosomes.20 Autophagy participates in cell growth, adaptation to stress conditions, anti-aging mechanisms, intracellular quality control, renovation during development and differentiation and the biogenesis of organelles such as cilia and acrosome.21,23-25 Autophagy is also active in Sertoli cells,26-29 where it is involved in the clearance of SHBG/ABP that is selectively regulated by testosterone.26 However, the role of autophagy in Sertoli cells is still largely unknown. In this study, we found that the disruption of autophagy by the Sertoli cell-specific knockout of Atg5 or Atg7 caused male mouse subfertility due to the disorganized seminiferous tubules and spermatozoa with malformed heads. The well-organized cytoskeleton structure was disturbed in both autophagy-deficient testis and Sertoli cells. A negative regulator of cytoskeleton organization, PDLIM1, was degraded through the autophagy pathway and accumulated in autophagy-deficient Sertoli cells. PDLIM1 accumulation resulted in the cytoskeletal disorganization in autophagy-deficient Sertoli cells and led to the disruption of both apical and basal ES and influenced Sertoli-germ cell communication. Thus, our work reveals a novel role for autophagy in Sertoli-germ cell communication by regulating the cytoskeleton through degrading PDLIM1 to maintain the proper organization of the ES.

sp56 PLZF

Results compartment and prepare for meiosis.2,16,17 After spermatocytes enter the apical compartment, the apical ES replaces the desmosome-like junctions and gap junctions in the Sertoligerm cell interface and persists until the spermatids line up at the luminal edge of the seminiferous tubule. The apical ES disassembly facilitates the release of spermatozoa during spermiation.6,17,18 Both basal and apical ES undergo recycling at the late stages of the epithelial cycle so that the ES proteins can be endocytosed, transcytosed and recycled to assemble new apical and basal ES.9,19 The extensive restructuring of basal and apical ES is crucial to spermatogenesis, and recent studies have shown many molecules and signaling pathways participate in ES restructuring.6,9,11,14,16 However, the precise mechanism remains elusive. Autophagy is a membrane-trafficking process delivering cytoplasmic material such as long-lived proteins and organelles to the lysosome for degradation. The canonical macroautophagy is initiated from an isolated membrane and followed by the formation of a double-membrane autophagosome that then moves along cytoskeletal structures and fuses with lysosomes for degradation. More than 40 autophagy-related (ATG) proteins have been characterized in autophagy.20-22 In the initiation stage of autophagosome formation, a ubiquitin-activating E1-like enzyme, ATG7, activates ATG12 and then facilitates ATG12-ATG5-ATG16L1 complex formation. This complex can function as an E3 ubiquitin ligase for the second ubiquitinlike conjugation system, LC3, which is activated by ATG7 and ATG3 (E2-like enzyme), sequentially, conjugating to the lipidcontaining membrane. As a scaffold candidate, the LC3-lipidcontaining membrane drives membrane expansion and vesicle

Sertoli cell-specific knockout of Atg5 or Atg7 influences male fertility in mice To detect the functional role of autophagy in Sertoli cells, we specifically knocked out Atg5 or Atg7 in Sertoli cells by crossing mice with a floxed Atg5 or Atg7 allele to AMH-Cre mice that express Cre recombinase only in the Sertoli cells of male mice.30-32 These Sertoli cell-specific atg5 and atg7 knockout mice were named AMH-atg5¡/¡and AMH-atg7¡/¡, respectively. Primary AMH-atg5¡/¡and control Sertoli cells were isolated from 19-d-old mice. The total protein was extracted and immunoblotting analysis was performed to detect the Atg5 knockout efficiency. As shown in Figure 1A, the ATG5 protein level was dramatically reduced in the knockout Sertoli cells compared with the Atg5Flox/Flox cells. Consistent with a role for ATG5 in autophagy,33 the membrane-associated form was LC3B-II reduced and the autophagic substrate SQSTM1/p62 accumulated in atg5-deficient Sertoli cells (Fig. 1A). Similar results were seen in AMH-atg7¡/¡ mice (Fig. 1B). Therefore, the autophagic flux is disrupted in both atg5 and atg7 knockout Sertoli cells. The fertility of atg5-deficient male mice and the control group were then assessed by mating 10 to 13 males from each strain with Atg5Flox/Flox females over a 2-mo period. As shown in Figure 1C, no females became pregnant after mating with AMH-atg5¡/¡ male mice, compared with an 84.84 § 1.38% rate of pregnancy after mating with Atg5Flox/Floxmice (Fig. 1C). The AMH-atg5¡/¡male mice were therefore completely infertile. The same experiment was performed with AMH-atg7¡/¡ mice and 42.70 § 2.10% females became pregnant compared

816

C. LIU ET AL.

Figure 1. Sertoli cell-specific knockout of Atg5 or Atg7 influences male fertility in mice. (A) The ATG5 protein level was dramatically reduced and the autophagic flux was disrupted in the Sertoli cells of AMH-atg5¡/¡ mice. Immunoblotting analysis of ATG5, SQSTM1 and LC3B was performed in both Atg5Flox/Flox and AMH-atg5¡/¡Sertoli cells. GAPDH served as a loading control. (B) The ATG7 protein level was dramatically reduced and the autophagic flux was disrupted in the Sertoli cells of AMH-atg7¡/¡ mice. Immunoblotting analysis of ATG7, SQSTM1 and LC3B was performed both in Atg7Flox/Flox and AMH-atg7¡/¡Sertoli cells. ACTIN served as a loading control. (C) The AMHatg5¡/¡ mice were infertile. In the fertility test, 84.84 § 1.38% of the plugged females were pregnant after crossing with Atg5Flox/Flox males (white column), whereas none of the plugged females were pregnant after crossing with AMH-atg5¡/¡males. (D) The fertility of AMH-atg7¡/¡ mice was decreased. In the fertility test, 92.30 § 3.40% of the plugged females were pregnant after crossing with Atg7Flox/Flox males (white column), whereas only 42.70 § 2.10% of the plugged females were pregnant after crossing with AMH-atg7¡/¡males (black column). (E) The histology of the seminiferous tubules from Atg5Flox/Flox and AMH-atg5¡/¡ mice using H&E staining. Asterisks indicate large vacuoles in seminiferous tubules. (F) The histology of the seminiferous tubules of Atg7Flox/Flox and AMH-atg7¡/¡ mice using H&E staining. Asterisks indicate large vacuoles in seminiferous tubules. See also Figure S1 and S2.

with 92.30 § 3.40% from the control group, indicating decreased fertility in Sertoli cell-specific atg7 knockout male mice (Fig. 1D). Thus, we conclude that autophagic activities in Sertoli cells play important roles in male fertility. The disruption of autophagy in Sertoli cells perturbed the structure of the basal ES To explore how autophagy in Sertoli cells influences male fertility, we first examined the histology of testes from AMH-atg5¡/¡and Atg5Flox/Floxmice by hematoxylin and eosin (H&E) staining. In atg5-deficient mouse testes, 2 types of seminiferous tubules were detected. One type showed normal tubule structure in histology, while the other was disorganized with large vacuoles compared with the normal seminiferous tubules in the control groups (Fig. 1E). The H&E staining of AMH-atg7¡/¡testes showed similar histological results (Fig. 1F). Usually, large vacuoles in the seminiferous epithelium come from cell death, so we asked whether the disruption of autophagy caused the Sertoli cell death. We assessed the Sertoli cell death by TUNEL assay on autophagy-deficient testis sections and found that the TUNEL-positive signals were not

colocalized with the Sertoli cell marker, WT134 (Fig. S1A and S1C), and the total Sertoli cell numbers per testis of autophagy-deficient mice were similar to their control groups (Fig. S1B and S1D). These results revealed that the disruption of autophagy in Sertoli cells has no obvious influence on the Sertoli cell survival but leads to the germ cell loss in the seminiferous epithelium. The blood–testis barrier which is constituted between adjacent Sertoli cells by the basal ES and other junctions is important for successful spermatogenesis.11,12,14 Its disruption often leads to germ cell loss and male infertility;12 we therefore speculated that the BTB structure might be influenced in atg5- or atg7-deficient mice. To test this hypothesis, transmission electron microscopy (TEM) analysis was performed. In the Atg7Flox/Flox mice, the BTB structure was intact between 2 adjacent Sertoli cells, and the integrated basal ES was identified by the actin filament bundles (arrowheads) sandwiched between cisternae of the endoplasmic reticulum (ER) and apposing plasma membranes of 2 Sertoli cells (Fig. S2). However, in atg7-deficient mice, the basal ES structure was disrupted with large vacuoles and the BTB structure was also perturbed (Fig. S2). The BTB disruption might lead to the germ cell loss in seminiferous tubules of the Sertoli cell-specific atg5 and atg7 knockout mice. These results indicate

AUTOPHAGY

that autophagy might be involved in the assembly of the ordered structure of the basal ES and the maintenance of normal BTB function. The disruption of autophagy in Sertoli cells produces spermatozoa with malformed heads and low motility The above-mentioned mechanism accounts for the decrease in the total number of spermatozoa in the cauda epididymis of the AMH-atg5¡/¡mice (5.37 § 2.84 £ 106 versus 19.93 § 3.69 £ 106 in Atg5Flox/Floxmice) (Fig. 2A), but cannot fully explain the infertility or subfertility of the Sertoli cell-specific autophagy-deficient mice because there were still some spermatozoa produced in the cauda epididymis (Fig. S3A). Detailed histological analysis showed that many spermatozoa with abnormal nuclei were found in the cauda epididymis of atg5-deficient mice, but not in the control group (Fig. S3A). Further morphological evaluation revealed that many spermatozoa from AMHatg5¡/¡ mice had irregularly shaped heads (Fig. 2C and 2E), and the single-sperm immunofluorescence experiment using the acrosome-specific marker, ZP3R/sp56,35 showed that acrosome and nuclear morphology were malformed (Fig. 2C). Similar sperm head abnormality and spermatozoa number decrease (17.74 § 0.53 £ 106 vs. 21.70 § 0.25 £ 106 in Atg7Flox/Flox mice) could also be detected in AMH-atg7¡/¡mice by cauda epididymis H&E staining and ZP3R/sp56 single-sperm immunofluorescence analysis (Fig. 2B, 2D and 2F, Fig. S3B). The TEM analysis of the cauda epididymis in the atg7-deficient mice also showed sperm nuclei malformation (Fig. S3C). These results suggest that the disruption of autophagy in Sertoli cells lead to both the spermatozoa number decrease and the production of spermatozoa with malformed nuclei. Vigorous sperm motility is absolutely necessary for normal fertilization in mammals.36 We measured the sperm motility in AMH-atg5¡/¡ and Atg5Flox/Floxmice by computer-assisted semen analysis and found that the sperm motility of AMHatg5¡/¡mice (15.00 § 1.83%) was severely decreased compared to that of Atg5Flox/Floxmice (88.00 § 1.83%) (Fig. 2G). The percentage of progressive sperms (0.50 § 0.71% versus 24.00 § 6.58% in Atg5Flox/Floxmice), average path velocity (VAP) (34.18 § 12.82 mm/s vs. 115.48 § 15.75 mm/s in Atg5Flox/Floxmice), straight line velocity (VSL) (18.18 § 4.82 mm/s versus 78.90 § 14.65 mm/s in Atg5Flox/Floxmice) and curvilinear velocity (VCL) (64.48 § 14.22 mm/s vs. 191.93 § 25.16 mm/s in Atg5Flox/Floxmice) of spermatozoa in atg5-deficient mice were also severely decreased (Fig. 2I, 2K, 2M and 2O). The sperm motility and related parameters also showed significant decrease in atg7deficient mice (Figs. 2H, 2J, 2L, 2N and 2P). Thus, the disruption of autophagy in Sertoli cells produces spermatozoa with low motility. The subfertility of Sertoli cell-specific autophagydeficient mice might due to both the disorganized seminiferous tubules caused by a disrupted basal ES and spermatozoa with malformed heads. The apical ES is disrupted in sertoli cell-specific autophagy-deficient mice The Sertoli cells play a key role in the control of selfrenewal and differentiation of SSCs, spermatocytic meiosis and

817

spermiogenesis.1-4 To explore how autophagy affects sperm head morphology and motility in Sertoli cells, we detected which stage of spermatogenesis was influenced in the Sertoli cell-specific autophagy-deficient mice. ZBTB16/PLZF is a transcription factor involved in SSC self-renewal,37 and its immunohistochemical analysis performed in AMH-atg7¡/¡mice testes showed no significant differences compared with Atg7Flox/Flox mice (Fig. S4). The immunohistochemical analysis of the meiosis related markers GCNA1, H2AFX/gH2AX and SYCP3 also showed no obvious differences between AMHatg7¡/¡ mice and the control group (Fig. S4). Thus, autophagy in Sertoli cells might have no influence on the selfrenewal of SSCs and meiosis process. During spermiogenesis, various cytoskeletal structures participate in shaping the sperm head, including microtubules and microfilament.19,38,39 We therefore analyzed the cytoskeleton in AMH-atg5¡/¡ and Atg5Flox/Floxmouse testes by the immunofluorescent analysis of TUBB/b-tubulin and phalloidin, which labels F-actin (Fig. 3A and 3B). In Atg5Flox/Flox testes, TUBB was oriented in linear arrays parallel to the long axis from the base to the apex of the Sertoli cells, forming a longitudinally oriented cage-like structure around Sertoli cell nuclei (indicated by immunofluorescence with WT1) (Fig. 3A), which was consistent with previous descriptions.40 However, in the AMHatg5¡/¡testes, both the linear structure along the axis and the cage-like structure around the nucleus were abolished (Fig. 3A). In the shaping of the sperm head, a highly ordered stack of F-actin-containing hoops surround the apical region of the elongating spermatid nucleus.38 The phalloidin indicated that the F-actin structure, which circled the elongating spermatid nucleus, was perturbed and disorganized in atg5-deficient testes compared with the control group (Fig. 3B). The AMHatg7¡/¡mouse testes also showed similar abnormal TUBB and F-actin structures (Fig. 3C and 3D). These results suggested that the disruption of autophagy in Sertoli cells influence the cytoskeleton structures during the process of the sperm head shaping. Because the F-actin hoops surrounding the elongating spermatid nucleus are important components of the apical ES in Sertoli cells, and the apical ES also plays a key role in shaping the sperm head,7,11,38 a TEM experiment was performed to detect the apical ES structure. In AMH-atg7¡/¡testes, the apical ES structure was disrupted, with large vacuoles. The actin bundles surrounding the sperm head disappeared compared to Atg7Flox/Flox mice (Fig. 3E). Similarly, the apical ES structure was also perturbed with large vacuoles and actin bundle loss in atg5-deficient mice (Fig. S5). To further confirm the apical ES assembly defect in autophagy-deficient Sertoli cells, we isolated spermatids attached with some Sertoli cell regions from seminiferous tubules, they were stained with phalloidin to directly show the Sertoli cell-spermatid interaction mediated by the apical ES.41 Similar to the TEM experiments, the F-actin fine structure detected by super-resolution microscopy was disrupted in autophagy-deficient mice (Fig. 4A and 4C), and the percentages of abnormal apical ES were significantly higher than control groups (Fig. 4B and 4D).Taken together, these results suggest that the disruption of autophagy in Sertoli cells causes a defect in apical ES assembly, influences the Sertoli cellgerm cell communication and sperm head shaping, and results

818

C. LIU ET AL.

Figure 2. The disruption of autophagy in Sertoli cells produces spermatozoa with malformed heads and low motility. (A-B) The total number of spermatozoa from the cauda epididymis was decreased in AMH-atg5¡/¡and AMH-atg7¡/¡mice. Atg5Flox/Flox (white column), 19.93 § 3.69 £ 106; AMH-atg5¡/¡ (black column), 5.37 § 2.84 £ 106 in (A). Atg7Flox/Flox (white column), 21.70 § 0.25 £ 106; AMH-atg7¡/¡ (black column), 17.74 § 0.53 £ 106in (B). (C-D) The sperm nuclear morphology and acrosomes are malformed in AMH-atg5¡/¡and AMH-atg7¡/¡mice. The single-sperm immunofluorescence analysis for the acrosome-specific marker ZP3R (green) was performed using Atg5Flox/Flox(left panel) and AMH-atg5¡/¡(right panel) spermatozoa in (C), and Atg7Flox/Flox(left panel) and AMH-atg7¡/ ¡ (right panel) spermatozoa in (D). Nuclei were stained with PI (red). (E-F) The abnormal sperm rate was increased in AMH-atg5¡/¡and AMH-atg7¡/¡mice. In AMH-atg5¡/¡ mice (black column), 46.13 § 0.93% of sperm had malformed heads, whereas only 3.44 § 0.34% of Atg5Flox/Flox mice (white column) had malformed heads (E). In AMH-atg7¡/¡ mice (black column), 29.97 § 1.69% of sperm had malformed heads, whereas only 1.97 § 0.28% of Atg7Flox/Flox mice (white column) did (F). (G-H) The motile sperm rate was decreased in AMH-atg5¡/¡and AMH-atg7¡/¡mice. Atg5Flox/Flox (white column, 88.00 § 1.83%), AMH-atg5¡/ ¡ (black column, 15.00 § 1.83%) (G). Atg7Flox/Flox(white column, 88.33 § 3.84%), AMH-atg7¡/¡(black column, 44.67 § 2.40%) (H). (I-J) The progressive sperm rate was decreased in AMH-atg5¡/¡and AMH-atg7¡/¡mice. Atg5Flox/Flox (white column, 24.00 § 6.58%), AMH-atg5¡/¡(black column, 0.50 § 0.71%) (I). Atg7Flox/ Flox (white column, 23.67 § 1.76%), AMH-atg7¡/¡(black column, 9.33 § 2.85%) (J). (K-L) The average path velocity (VAP) of the spermatozoa was decreased in AMH-atg5¡/¡and AMH-atg7¡/¡mice. Atg5Flox/Flox (white column, 115.48 § 15.75 mm/s), AMH-atg5¡/¡(black column, 34.18 § 12.82 mm/s) (K). Atg7Flox/Flox (white column, 93.00 § 8.20 mm/s), AMH-atg7¡/¡(black column, 57.97 § 6.42 mm/s) (L). (M-N) The straight-line velocity (VSL) of the spermatozoa was decreased in AMH-atg5¡/¡and AMH-atg7¡/¡mice. Atg5Flox/Flox (white column, 78.90 § 14.65 mm/s), AMH-atg5¡/¡(black column, 18.18 § 4.82 mm/s) (M). Atg7Flox/Flox (white column, 64.07 § 4.89 mm/s), AMH-atg7¡/¡(black column, 38.50 § 2.57 mm/s) (N). (O-P) The curvilinear velocity (VCL) of the spermatozoa was decreased in AMH-atg5¡/¡and AMH-atg7¡/¡mice. Atg5Flox/Flox (white column, 191.93 § 25.16 mm/s), AMH-atg5¡/¡(black column, 64.48 § 14.22 mm/s) (O). Atg7Flox/Flox (white column, 156.87§ 9.44 mm/s), AMH-atg7¡/¡(black column, 104.33 § 13 mm/s) (P). See also Figure S3.

AUTOPHAGY

819

Figure 3. The apical ES is disrupted in Sertoli cell-specific autophagy-deficient mice. (A and C) The TUBB structure was disordered in AMH-atg5¡/¡and AMH-atg7¡/¡ testes. Immunofluorescence analysis using antibodies against TUBB (red) and WT1 (green) was performed in the seminiferous tubules of Atg5Flox/Floxmice (upper panels) and AMH-atg5¡/¡ mice (lower panels) in (A). That of Atg7Flox/Floxmice (upper panels) and AMH-atg7¡/¡ mice (lower panels) are shown in (C). Nuclei were stained with DAPI (blue). (B and D) The F-actin-containing hoops surrounding the elongating spermatid nucleus were perturbed and disorganized in AMH-atg5¡/¡and AMH-atg7¡/¡ testes. Immunofluorescence analysis using phalloidin (green, labeled by FITC) was performed in the seminiferous tubules of Atg5Flox/Floxmice (left panels) and AMH-atg5¡/¡ mice (right panels) in (B). That of Atg7Flox/Floxmice (left panels) and AMH-atg7¡/¡ mice (right panels) are shown in (D). Nuclei were stained with DAPI (blue). (E) Ultrastructural analysis indicated that the apical ES structure was disrupted in AMH-atg7¡/¡ seminiferous tubules. TEM of Atg7Flox/Floxmice and AMH-atg7¡/¡ mice was performed. In Atg7Flox/Floxmice (left panel), the apical ES was characterized by the presence of actin bundles (arrowheads) sandwiched between the endoplasmic reticulum (ER) and the Sertoli cell plasma membrane. However, in AMH-atg7¡/¡ mice (right panel), the structure of the apical ES was disrupted with large vacuoles (asterisks), and the actin bundles surrounding the elongating sperm head disappeared. Ac, acrosome; Nu, nucleus. See also Figure S5.

820

C. LIU ET AL.

Figure 4. Apical ES assembly defect in autophagy-deficient mice. (Aand C) Apical ES assembly was perturbed in AMH-atg5¡/¡and AMH-atg7¡/¡spermatids attached with some Sertoli cell regions. (A) Super-resolution microscopy analysis of the F-actin structure by immunofluorescence of phalloidin (green, labeled by FITC) in the spermatids attached with Sertoli cell regions of Atg5Flox/Floxmice (upper panels) and AMH-atg5¡/¡ mice (lower panels). (C) Super-resolution microscopy analysis of the apical ES of Atg7Flox/Floxmice (upper panels) and AMH-atg7¡/¡ mice (lower panels). Nuclei were stained with DAPI (blue). Arrows indicated TBCs. (B and D) Disordered F-actin structures of apical ES significantly increased in the autophagy-deficient mice. (B) 42.87 § 0.49% of apical ES with perturbed F-actin structures in AMH-atg5¡/¡ mice (black column), whereas only 1.48 § 0.23% disordered apical ES in the Atg5Flox/Flox mice (white column). (D) 36.14 § 0.98% of disordered apical ES in the AMH-atg7¡/¡ mice (black column), while only 1.35 § 0.29% in the Atg7Flox/Flox mice (white column).

in the infertility or subfertility of Sertoli cell-specific autophagydeficient mice. Disordered tubulobulbar-complex distribution in the autophagy-deficient Sertoli cells In addition to the ectoplasmic specialization, tubulobulbar complexes (TBCs) are also cytoskeleton-related structures in Sertoli cells, they are composed of fine filaments of actin surrounded tubular portion and double-plasma membrane

bulbous portion flanked by smooth endoplasmic reticulum. TBCs are located on both Sertoli–Sertoli cells and Sertoli cell– spermatids interface and that play important roles in the ectoplasmic specialization restructuring, excess spermatid cytoplasm removing, spermatid acrosome shaping and spermiation.42,43 To test whether TBC structures were also affected by the disruption of autophagy in Sertoli cells, we detected TBCs in AMH-atg5¡/¡ and Atg5Flox/Floxmouse testes by immunofluorescence analysis with a TBC component, CTTN/cortactin.44 In contrast to the well organized TBCs in

AUTOPHAGY

the control group (Fig. 5A, left panel), although CTTN could be detected in atg5-deficient mouse testes, and they were still colocalized with F-actin at the sperm head, TBCs distributions were abnormal and they were irregularly lay on the sperm head (Fig. 5A). To further confirm these results, we isolated spermatids attached with Sertoli cell regions from either atg5-deficient or the control seminiferous tubules, then performed the immunofluorescence analysis of CTTN. We found that the distribution of TBCs was indeed disorganized, and they were coupled with malformed sperm head (Fig. 5B). Furthermore, TEM analysis of the autophagy-deficient mouse testis indicated that the tubular and bulbous portions of TBCs were similar to control groups (Fig. S6B). The AMH-atg7¡/¡ mouse testes also showed similar abnormal distributions of TBCs (Fig. 5C and 5D, Fig. S6C). Thus, we conclude that the disruption of autophagy in Sertoli cells mainly affect TBCs distribution but not their assembly. The sertoli cell cytoskeleton structure is disordered after autophagy disruption Different devices or molecules could regulate the cytoskeletal dynamics in the seminiferous epithelium, including Sertoligerm cell communication, cytokines, calcium, proteases and protease inhibitors.5,11,19 We next asked whether the disorganized cytoskeleton structures in testis were directly or indirectly

821

caused by autophagy disruption. To test this, primary AMHatg5¡/¡and control Sertoli cells were isolated, and the cytoskeleton structures were detected in vitro by immunofluorescence of phalloidin. As shown in Figure 6A and 6B, the F-actin structures were disordered with severely accumulated foci in AMHatg5¡/¡ Sertoli cells compared with Atg5Flox/Flox Sertoli cells. The atg7-deficient Sertoli cells also showed F-actin disorganization and disrupted classical net structures (Fig. 6C and 6D). Thus, the disruption of autophagy in Sertoli cells directly affects their cytoskeletal structure organization. Overexpression of PDLIM1 affects the structure organization of F-actin in Sertoli cells Many studies regarding the relationship between the cytoskeleton and autophagy focused on autophagosome formation and transportation.45,46 Few studies have investigated how autophagy modulates cytoskeletal organization. Recently, Zhou et al. have conducted a quantitative mass spectrometry comparison between WT and atg7¡/¡ MEFs. They find that F-actin fibers are disorganized in atg7¡/¡ MEFs. In addition, 66 downregulated proteins and 48 upregulated proteins are identified.47 The upregulated proteins might be potential negative regulators, and their accumulation might result in the disorganization of the F-actin fibers. Among the upregulated proteins, some cytoskeleton associated proteins and gene expression related

Figure 5. TBCs distribution rather than assembly is affected in the autophagy-deficient Sertoli cells. (A and C) The distribution of TBCs was perturbed in AMH-atg5¡/¡and AMH-atg7¡/¡mouse testes. (A) Immunofluorescence analysis of phalloidin (red, labeled by TRITC) and CTTN (green) was performed in the testes of Atg5Flox/Floxmice (left panels) and AMH-atg5¡/¡ mice (right panels). (C) The same immunofluorescence analysis as in (A) but conducted here with Atg7Flox/Floxmice (left panels) and AMH-atg7¡/ ¡ mice (right panels). Nuclei were stained with DAPI (blue). (B and D) Abnormal TBCs distribution in AMH-atg5¡/¡and AMH-atg7¡/¡spermatids attached with some Sertoli cell regions. (B) Immunofluorescence analysis of phalloidin (red, labeled by TRITC) and CTTN (green) was performed in spermatids attached with some Sertoli cell regions of Atg5Flox/Flox(upper panels) and AMH-atg5¡/¡mice (lower panels). (D) The same immunofluorescence analysis as in (B) but conducted here with Atg7Flox/Flox(upper panels) and AMH-atg7¡/¡mice (lower panels). Nuclei were stained with DAPI (blue). See also Figure S6.

822

C. LIU ET AL.

proteins have multiple relationships with those downregulated proteins, according to interaction network analysis. One possibility of the cytoskeletal disorganization in atg7¡/¡ MEFs is that autophagy mediates the degradation of some negative regulators to promote the proper cytoskeletal organization, another one is that autophagy regulates the protein level of some cytoskeleton-associated transcription factors to influence the transcription of the cytoskeleton or cytoskeleton related proteins. If any of them were real negative regulator of the cytoskeletal organization, the overexpression of this gene might mimic the autophagy deficient cell to some extent. To identify potential negative cytoskeleton organization regulators influenced by autophagy, we analyzed the upregulated either cytoskeleton associated or gene expression related genes in atg7¡/¡ MEFs, and found that 6 of them were highly suspected to be a

negative regulator of the cytoskeleton (Table S1 and S2). They were cloned into the pRK vector with a FLAG tag and overexpressed in Sertoli cells. Through this small-scale functional screening, we found that the overexpression of PDLIM1, NUDC and STMN1 resulted in disordered F-actin networks (Fig. 7). NUDC and STMN1 overexpression resulted in the disassembly of the F-actin net structure in Sertoli cells. In the PDLIM1 overexpressing Sertoli cells, the F-actin structure was disordered with severely accumulated foci, similar to the autophagy-deficient Sertoli cells (Fig. 7, Fig. 6A and 6C). As PDLIM1 is a member of the PDZ and LIM protein family that could associate with F-actin by interacting with ACTN1and ACTN4,48 we focused on PDLIM1 as a candidate negative regulator that participated in the process of autophagy-modulated cytoskeleton organization.

Figure 6. Disordered F-actin structures in autophagy-deficient Sertoli cells. (A and C) The structure of F-actin was disordered with severely accumulated foci in AMHatg5¡/¡and AMH-atg7¡/¡Sertoli cells. Immunofluorescence analysis using phalloidin (green, labeled by FITC) was performed in Sertoli cells of Atg5Flox/Floxmice (upper panels) and AMH-atg5¡/¡ mice (lower panels) (A). That of Atg7Flox/Floxmice (upper panels) and AMH-atg7¡/¡ mice (lower panels) is shown in (C). Nuclei were stained with DAPI (blue). (B and D) The rate of abnormal Sertoli cells with severely accumulated F-actin foci was increased in AMH-atg5¡/¡and AMH-atg7¡/¡mice. In AMH-atg5¡/¡ mice (black column), 47.49 § 1.40% of Sertoli cells with severely accumulated F-actin foci, whereas 12.73 § 1.59% of Atg5Flox/Flox mice (white column) had disordered structures (B). In AMH-atg7¡/¡ mice (black column), 48.24 § 0.22% of disordered F-actin structures, whereas 15.99 § 2.22% of Atg7Flox/Flox mice (white column) did (D).

AUTOPHAGY

PDLIM1 can be degraded through the autophagy pathway and accumulates in autophagy-deficient Sertoli cells Autophagy is a lysosomal degradation pathway that eliminates long-lived proteins or defective organelles from the cell. The disruption of the autophagy pathway would result

Figure 7. Overexpression of PDLIM1 affects the fine structure organization of Factin in Sertoli cells. A small-scale functional screening was performed by the transfection of FLAG-tagged MAP6/MTAP6, NUDC, PDLIM1, PPP4R2, STMN1 and YBX1 in pRK vectors into Sertoli cells. An empty vector was transfected as a control group. Subsequent immunofluorescence analysis using FLAG (red) and phalloidin (green, labeled by FITC, indicating F-actin) was performed, and nuclei were stained with DAPI (blue).

823

in protein accumulation.21,25 PDLIM1 overexpression could affect F-actin structure similarly to the autophagy-deficient Sertoli cells, so we speculated whether autophagy modulated the cytoskeleton structure by regulating the PDLIM1 protein level. To test this, we examined the PDLIM1 protein level by immunofluorescence and immunoblotting in AMHatg5¡/¡ and Atg5Flox/Flox Sertoli cells. Both the immunofluorescence and immunoblotting results indicated that PDLIM1 accumulated in AMH-atg5¡/¡ Sertoli cells compared to Atg5Flox/Flox Sertoli cells (Fig. 8A and 8B). To prove that ATG5 was crucial for PDLIM1 degradation, cycloheximide (CHX) chase assays were performed in AMH-atg5¡/¡ and Atg5Flox/Flox Sertoli cells. The result showed that ATG5 depletion strongly delayed PDLIM1 degradation compared with the control group (Fig. 8C and 8D). The atg7-deficient Sertoli cells also showed PDLIM1 accumulation and that ATG7 was crucial for PDLIM1 degradation (Fig. 8E to H). To further prove that PDLIM1 was degraded through the autophagy-lysosome pathway, 2 acidic lysosome inhibitors, ammonium chloride (NH4Cl) and chloroquine (CQ),23 were used to inhibit the autophagy-lysosome pathway. After treatment with lysosome inhibitors, PDLIM1 protein significantly accumulated (Fig. S7). Immunofluorescence analysis also showed that PDLIM1 colocalized with LC3A/B after starvation in Sertoli cells (Fig. S8). To further demonstrate whether PDLIM1 really interacts with LC3A/B or not, we carried out coimmunoprecipitation after transient transfection of pRK-FLAG-Pdlim1 into HEK293T cells. As shown in Figure 8I and 8J, PDLIM1 indeed physically interacted with LC3B. Above all, all these results suggested that PDLIM1 could be degraded via an autophagy-lysosome pathway in Sertoli cells. To test whether the accumulation of PDLIM1 was the major cause of the cytoskeleton disorganization in autophagy-deficient Sertoli cells, we knocked down Pdlim1 in autophagy-deficient and the control Sertoli cells by RNA interference, respectively. After the transfection of Pdlim1specific siRNA duplexes, the PDLIM1 protein level was reduced both in autophagy-deficient Sertoli cells and their control groups (Fig. 9B and 9E). Next, we detected the cytoskeleton structure of Sertoli cells by immunofluorescence of phalloidin, and found that the percentages of Sertoli cells with severely accumulated F-actin foci were decreased significantly after knockdown of Pdlim1 in autophagy-deficient Sertoli cells (Fig. 9A, 9C, 9D and 9F). Thus, PDLIM1 might be the major substrate of the autophagy to regulate cytoskeleton organization in Sertoli cells. Next, we detected the protein level of PDLIM1 in autophagy-deficient mouse testes by immunofluorescence analysis using the anti-PDLIM1 antibody. The PDLIM1 signal accumulated in the apical ES region of both AMHatg5¡/¡ and AMH-atg7¡/¡mouse testes compared with their separate control groups (Fig. 10). These results indicate that in the seminiferous epithelium, the disruption of autophagy in Sertoli cells causes the abnormal accumulation of a negative cytoskeleton organization regulator, PDLIM1, in the apical ES region, which leads to the disorganization of the cytoskeletal structure and disruption of the ES assembly, finally influencing spermatogenesis.

824

C. LIU ET AL.

Discussion Spermatids are highly differentiated cells that exhibit complicated cellular changes, such as nucleus elongation and condensation, cytoplasmic redistribution, acrosome formation, disassembly of the microtubule-based manchette and the formation of the flagella. Abnormal sperm head shapes represent a form of teratozoospermia that can cause male infertility.2,19,39,49 The cytoskeleton is a highly organized dynamic structure that plays a key role in the shaping and differentiating

of spermatids.50 Two cytoskeletal structures in the seminiferous epithelium, the Sertoli cell apical ES and the spermatid-containing acrosome-acroplaxome-manchette complex, may play cooperative roles in the shaping of the sperm head.38,39,51 The Sertoli cell apical ES encircles and mechanically grasps the acrosome of the elongating spermatid head with a narrow layer of hexagonally packed, parallel actin bundles.11,13,14,51 The apical ES is unique, with hybrid-like characteristics. It is composed of several proteins that are generally found in the anchoring junction, focal contacts and tight junctions. The high

Figure 8. PDLIM1 accumulates in autophagy-deficient Sertoli cells and can be degraded through the autophagy pathway. (A) PDLIM1 was accumulated in AMH-atg5¡/¡ Sertoli cells. Immunofluorescent analysis using PDLIM1 in Atg5Flox/Flox(upper panels) and AMH-atg5¡/¡ (lower panels) Sertoli cells. (B) The PDLIM1 protein level increased in AMH-atg5¡/¡ Sertoli cells. Immunoblotting analysis of PDLIM1 and ATG5 was performed in both Atg5Flox/Flox and AMH-atg5¡/¡Sertoli cells. GAPDH served as a loading control. (C) ATG5 was crucial for the degradation of PDLIM1. A cycloheximide chase (CHX) assay of PDLIM1 was performed in Atg5Flox/Flox and AMH-atg5¡/¡Sertoli cells. Samples were taken at 0, 2, and 6 h after the addition of CHX for 2 h. The PDLIM1 protein level was detected by immunoblotting, and VCP/p97 served as a loading control. (D) Quantification of the relative PDLIM1 levels in (C). The amounts of PDLIM1 were quantified using the Odyssey software. (E-H) PDLIM1 is accumulated in AMHatg7¡/¡ Sertoli cells, and ATG7 was crucial for the degradation of PDLIM1. Similar immunofluorescence (E) and immunoblotting (F) analysis of PDLIM1 was performed in Atg7Flox/Flox and AMH-atg7¡/¡Sertoli cells. PDLIM1 was measured and detected as in (G). GAPDH served as a loading control. The relative PDLIM1 levels were quantified as in (H). (I and J) PDLIM1 interacts with LC3B. pRK-FLAG-Pdlim1 was transfected into HEK293T cells. Twenty-four h after transfection, cells were collected for immunoprecipitation (IP) with anti-LC3A/B (I) or anti-FLAG (J) and analyzed with FLAG or LC3B antibodies, respectively. See also Figure S7 and S8.

AUTOPHAGY

825

Figure 9. PDLIM1 might be the major substrate of autophagy to regulate cytoskeleton organization in Sertoli cells. (A and D) The disordered F-actin structures in autophagydeficient Sertoli cells could be rescued by Pdlim1 knockdown. (A) Immunofluorescence analysis of phalloidin (green, labeled by FITC) and PDLIM1 (red) was performed in Atg5Flox/Flox (left panels) and AMH-atg5¡/¡ (right panels) Sertoli cells after transfection with different siRNAs. (D) The same analysis as in (A) but conducted here with Atg7Flox/Flox(left panels) and AMH-atg7¡/¡Sertoli cells (right panels). Nuclei were stained with DAPI (blue). Ctrl siRNA indicates Sertoli cells transfected with nontargeting control siRNA. siPdlim1-a and siPdlim1-b indicate Sertoli cells transfected with siRNA duplexes targeting Pdlim1 in 2 different sites. (B and E) The Pdlim1 knockdown efficiency in Sertoli cells. (B) Immunoblotting analysis of PDLIM1 was performed in both Atg5Flox/Flox and AMH-atg5¡/¡Sertoli cells in (A). (E) Immunoblotting analysis of PDLIM1 in Atg7Flox/Floxand AMHatg7¡/¡Sertoli cells in (D). GAPDH served as a loading control. (C and F) Quantification of the disorganized F-actin structures. (C) The percentage of Sertoli cells with severely accumulated foci was decreased after knockdown of Pdlim1 in autophagy-deficient Sertoli cells. In AMH-atg5¡/¡Sertoli cells, 38.16 § 1.85% of Sertoli cells transfected with nontargeting control siRNA duplexes (red columns) had perturbed F-actin structures, whereas 17.81 § 2.30% (siPdlim1-a, yellow columns) and 19.34 § 1.17% (siPdlim1-b, green columns) of Sertoli cells transfected with siRNA duplexes targeting Pdlim1 had disordered structures.(ctrl siRNA 9.10 § 0.91%, siPdlim1-a 11.14 § 1.29%, siPdlim1-b 9.31 § 0.84% in Atg5Flox/Flox Sertoli cells) (F) In AMH-atg7¡/¡Sertoli cells, 39.89 § 1.76% of Sertoli cells had perturbed F-actin structures in the control group (red columns), whereas 17.19 § 2.04% (siPdlim1-a, yellow columns) and 18.50 § 1.07% (siPdlim1-b, green columns) of Sertoli cells had disordered F-actin structures after knocking down Pdlim1 by specific targeting with siRNA duplexes. (ctrl siRNA 10.49 § 0.63%, siPdlim1-a 10.55 § 0.92%, siPdlim1-b 11.07 § 1.61% in Atg7Flox/Flox Sertoli cells).

826

C. LIU ET AL.

Figure 10. PDLIM1 is accumulated in the apical ES region of the autophagy-deficient mice. (A and B) Immunofluorescence analysis of PDLIM1 (green) was performed in Atg5Flox/Flox(left panels) and AMH-atg5¡/¡ (right panels) testes (A), Atg7Flox/Flox(left panels) and AMH-atg7¡/¡ (right panels) testes (B). Nuclei were stained with DAPI (blue). (C and D) Immunofluorescence analysis using phalloidin (red, labeled by TRITC) and PDLIM1 (green) was performed in the spermatids attached with some Sertoli cell regions of Atg5Flox/Flox(left panels) and AMH-atg5¡/¡ (right panels) mice (C), Atg7Flox/Flox(left panels) and AMH-atg7¡/¡ (right panels) mice (D). Nuclei were stained with DAPI (blue).

heterogeneity is thought to permit the apical ES to accomplish its multifunctional role in supporting spermiogenesis. The CDH2-CTNNB1 complexes function in apical ES stabilization, while JAM3 functions in spermatid polarization and positioning.7 The apical ES has a key role in spermatogenesis, and

structurally abnormal or absent apical ES has been associated with inappropriate release of spermatids or spermatids abnormalities.52,53 Some early studies indicated that the apical ES participates in sperm head shaping.13 However, the ES-associated actin bundles are noncontractile, which is in accordance

AUTOPHAGY

with the unipolar arrangement of the hexagonally packed actin filaments characteristic of the ES and failure to produce contractions in actin at the ES.54,55 They therefore could not present the exogenous constriction force, and so the precise function remains controversial.13 Autophagy is active in Sertoli cells,26-29 and early morphology studies found that the lysosomes of the seminiferous epithelium show cyclical variations. Chemes speculates that germ cell residual body disposal might be initiated by autophagy and completed by Sertoli cell phagocytosis.56 The strong reproductive toxicity drug microcystin-LR (MC-LR) has toxic effects on Sertoli cells by inducing autophagy and apoptosis.28 The carcinogen Lindane disrupts autophagy at the maturation step in Sertoli cells, and the treatment in Swiss mice induces shrunken and distorted seminiferous tubules.57,58 Recently, autophagy has been found to be involved in the clearance of SHBG/ABP that is selectively regulated by testosterone.26 Since both Sertoli cells and autophagy have extensive and multiple functions, autophagy may participate in the manifold regulation of Sertoli cells. Here, we show that autophagy influenced the apical ES structure by regulating a negative cytoskeleton organization regulator, PDLIM1, and participating in sperm-head shaping. From a small-scale functional screening, PDLIM1 was identified as a negative regulator of cytoskeleton organization. It was degraded through the autophagy pathway, and

827

once autophagic flux was disrupted, it accumulated in Sertoli cells (Fig. 8). And PDLIM1 might be the major substrate of the autophagy to regulate cytoskeleton organization because the knockdown of Pdlim1 in autophagy-deficient Sertoli cells could significantly restore proper organization of the cytoskeletal structure (Fig. 9). PDLIM1 is a member of the PDZ and LIM protein family. It contains an N-terminal PDZ domain and a C-terminal LIM domain.59 PDLIM1 could function as a cytoskeletal organization scaffold and an adaptor for the recruitment of several factors. For example, PDLIM1 interacted with F-actin-based stress fibers via an association with nonmuscle ACTN1 and ACTN4 in nonmuscle tissues,48 and PDLIM1 formed a triple complex with ACTN1 and PALLD in stress fibers.60 In adult rat testes, PALLD colocalizes with EPS8 which is an actin filamentbundling and barbed end-capping protein and ACTR3, which is abranched actin polymerization protein, suggesting that it might regulate actin filament bundle dynamics at the ectoplasmic specialization in the epithelial cycle of spermatogenesis.9 PDLIM1 may also participate in the restructuring of the ectoplasmic specialization by interacting with PALLD. Therefore, the PDLIM1 accumulation might lead to the formation of stress fiber-like structures and influence their localization or dynamics, finally resulting in the disorganization of the ectoplasmic specialization in the seminiferous

Figure 11. A proposed model for the role of autophagy in Sertoli cell-germ cell communication. In normal Sertoli cells (left), LC3 is activated by ATG7 and ATG5 and conjugates to the phagophore membrane. The phagophore could engulf the excessive PDLIM1 by forming an autophagosome and then be delivered to the lysosome/autolysosome for degradation. F-actin-containing apical ES structures and the related microtubule-based structures were assembled properly. In the absence of ATG7 or ATG5 (right), autophagy cannot be initiated, and PDLIM1 accumulated inside the Sertoli cell cytosol. The accumulated PDLIM1 disrupted the F-actin hoops of the apical ES and related microtubule-based elements in Sertoli cells, resulting in the disruption of Sertoli cell-germ cell communication.

828

C. LIU ET AL.

epithelium. As illustrated in Figure 11, in the normal condition (left), the autophagy pathway regulates the PDLIM1 protein level by activating LC3 through ATG7 and ATG5 to promote the phagophore engulfment of excessive PDLIM1 and its delivery to the autolysosome for degradation. In this way, the F-actin-containing apical ES structures and the related microtubule-based structures could be properly assembled. However, in the absence of ATG7 or ATG5 (Fig. 11, right panel), autophagy cannot be initiated and PDLIM1 accumulates inside the Sertoli cell cytosol. The accumulated PDLIM1 disrupted the F-actin hoops of the apical ES and related microtubule-based structures in the seminiferous epithelium, finally resulting in the disruption of Sertoli cell-germ cell communication. Although autophagy plays very important roles in apical ES formation, it does not affect the assembly of TBCs’ tubular and bulbous structures (Fig. 5 and Fig. S6). As 2 actin-based structures of the Sertoli cell, the discrepancy of the influence of autophagy on apical ES and TBCs assembly might come from their different arrangements of F-actin. F-actin is packed in hexagonal arrays in the ES, while it appears as a branched network surrounding the tubular portion of TBCs.43,55 As PDLIM1 could work as a scaffold to recruit other molecules to bind the F-actin bundles, and the binding of PDLIM1 to ACTN1 and PALLD may offer more actin binding site in a smaller volume, facilitating the F-actin interconnectivity,9,60,61 so the accumulation of PDLIM1 may facilitate the branched F-actin network formation in TBCs rather than affecting their assembly. The disordered distribution of TBCs might be caused by the abnormal structure of apical ES or malformed sperm head, and these possibilities still need further experimental data to be distinguished in the future. Although a lot of studies regarding the relationship between cytoskeleton and autophagy focus on the functional role of the cytoskeleton in autophagosome formation and transportation,45,46 few studies investigate how autophagy modulates cytoskeletal organization. The ciliopathy protein OFD1 is degraded by autophagy to promote primary ciliogenesis.24 Chaperone-assisted selective autophagy could be induced after strenuous resistance exercise to degrade mechanically damaged cytoskeleton proteins in affected muscles.62 The autophagy-dependent regulation of cytoskeletal stress fibers has been suggested over 2 decades ago by Ridley and Hall (1992).63 They show that once Swiss 3T3 fibroblasts are starved by serum deprivation for a few hours, the preexisting stress fibers and focal adhesions disappear, suggesting that they might be degraded by the autophagy-lysosome pathway. In the present study, we provided new evidence to show that the stress fiber related protein PDLIM1 is degraded through the autophagy pathway in Sertoli cells to promote the proper assembly of the ES, thereby providing a novel mechanism underlying the physiological role of autophagy in building a Sertoli cell-specific structure to facilitate Sertoli-germ cell communication.

Materials and methods Animals The Atg5Flox/Flox mouse strain (RBRC02975)30 and Atg7Flox/Flox mouse strain (RBRC02759)31 were purchased from the RIKEN Bio Resource Center with permission from Dr. Noboru

Misushima (The University of Tokyo) and Dr. Masaaki Komatsu (Tokyo Metropolitan Institute of Medical Science). The Atg5Flox/ Flox AMH-Cre and Atg7Flox/FloxAMH-Cre mice were bred from Atg5Flox/Flox, Atg7Flox/Floxmice and AMH-Cre mice.32 All animal studies were carried out in accordance with the protocols approved by the Institutional Animal Care and Use Committee at the Institute of Zoology, Chinese Academy of Sciences (CAS). Antibodies The rabbit anti-ATG5 antibody (A0731) for western blot, mouse anti-ATG7 monoclonal antibody (SAB4200304) for western blot, rabbit anti-LC3B polyclonal antibody (L7543) for immunoblotting and rabbit anti-CTTN (cortactin; KE-20) antibody (C7112) were purchased from Sigma-Aldrich. The rabbit anti-SQSTM1 polyclonal antibody (5114) was purchased from Cell Signaling Technology. Mouse anti-ZP3R (55101) antibody was purchased from QED Bioscience. The rabbit anti-SYCP3 polyclonal antibody (ab15093) and anti-GCNA1 (ab82527) antibodies were purchased from Abcam. The rabbit anti-WT1 antibody was from Epitomics (S0268). The rabbit anti-PDLIM1 antibody (11674-1AP) was purchased from Proteintech Group. The mouse antiLC3A/B (M152-3) was purchased from MBL. The mouse antiZBTB16 (sc-28319) was purchased from Santa Cruz Biotechnology. The mouse anti-H2AFX (05-636) was purchased from Millipore. The anti-FLAG (M20008), anti-Pan-Actin (M20010L) and anti-TUBB (M20005L) antibodies were purchased from Abmart. The GAPDH (ab1019t) antibody was purchased from Bo Ao Rui Jing. The VCP (valosin containing protein) polyclonal antibodies were generated in rabbit by using the corresponding recombinant proteins as antigen. The goat anti-rabbit FITC (ZF-0311), goat anti-mouse FITC (ZF-0312), and goat anti-mouse TRITC (ZF0313)-conjugated secondary antibodies, as well as the goat antimouse (ZB-2305) horseradish peroxidase (HRP), goat anti-rabbit HRP (ZB-2301)-conjugated secondary antibodies were purchased from Zhong Shan Jin Qiao. The FITC-phalloidin (40735ES75) and TRITC-phalloidin (40734ES75) were purchased from YEASEN. The Alexa Fluor 680-conjugated goat anti-mouse (A21057), and Alexa Fluor 680-conjugated goat anti-rabbit (A21109) for immunoblotting were purchased from Invitrogen. Assessment of the fertility of autophagy-deficient mice The fertility assessment experiments were performed as previously described.23 Each male mouse (8 or 9 wk) was caged with 2 CD1 females (7 or 8 wk), and their vaginal plugs were checked every morning. The number of pups in each cage was counted within a week of birth. Each male underwent 4 cycles of the above breeding assay. Primary sertoli cell cultures, transfection and knockdown Mice at 18 to 22 d of age were used for primary Sertoli cell isolation. The isolation method was as previously described, with minor modifications.64 Testes were decapsulated under the dissection microscope, and the seminiferous tubules were pooled and washed with phosphate-buffered saline (PBS; Gibco, C14190500BT) 3 times. The tubules were incubated with 2 mg/ml collagenase I (Sigma, C5138) and 0.5 mg/ml DNase I (Sigma,

AUTOPHAGY

AMPD1) in F12-DMEM (HyClone, SH30023.01B) for 30 min at 37 C, then washed twice with F12-DMEM (HyClone, SH30023.01B) and further digested with 2 mg/ml collagenase I (Sigma, C5138), 0.5 mg/ml DNase I (Sigma, AMPD1) and 1 mg/ ml hyaluronidase type III (Sigma, H3506) for 10 min at 37 C. The tubules were allowed to settle and were then washed twice with DMEM before being digested with 2 mg/ml collagenase I, 0.5 mg/ ml DNase I, 2 mg/ml hyaluronidase, and 1 mg/ml trypsin (Gibco, 25200072) for 30 min at 37 C. This final digestion step resulted in a cell suspension containing primarily Sertoli cells and type A spermatogonia. The dispersed cells were then washed twice with DMEM and placed into culture dishes in F12-DMEM containing 15% fetal calf serum (Gibco, 10270) and incubated at 34 C and 5% CO2. Spermatogonia were unable to attach to the dish and were removed during the medium change after 48 h. For performing the functional screening, Sertoli cells were transfected with the relevant plasmid DNA (~4 mg of DNA per 105 Sertoli cells) for 24 h using 10 ml of lipofectamine 2000 (Invitrogen, 11668019). For knockdown experiments, siRNA duplexes (100 pM siRNA per 105 Sertoli cells with 7.5 ml lipofectamine 2000 reagent [Invitrogen, 11668019]) targeting Pdlim1were used, along with nontargeting control siRNA duplexes, to transfect Sertoli cells for 24 h. The sequences of siRNA duplexes were as follows: siPdlim1-a, 50 -GCUGCUAUAGCGAAUUUAUTT-30 (GenePharma,PdlimMus-290-2015-7-1); siPdlim1-b, 50 -CCAGCAGCAUGACACACUUTT-30 (GenePharma, Pdlim-Mus-348-2015-7-1) Nontargeting siRNA duplex (Negative control; GenePharma) served as the negative control. Cells were harvested at 2 days after transfection of plasmid DNA or siRNA duplexes for fluorescence microscopy and preparation of lysates. Tissue collection and histological analysis Testes from at least 3 mice for each genotype were dissected immediately after euthanasia, fixed in 4% (mass/vol) paraformaldehyde (PFA; Solarbio, P1110) for up to 24 h, stored in 70% (vol/vol) ethanol, and embedded in paraffin. The 5mm sections were prepared and mounted on glass slides. After deparaffinization, slides were stained with H&E for histological analysis. Epididymal sperm count and sperm motility assays The cauda epididymis was dissected from 8- or 9-wk-old mice. Sperms were squeezed out from the cauda epididymis and incubated for 30 min at 37 C under 5% CO2. The incubated sperm medium was then diluted 1:500 and transferred to a hemocytometer for counting. The sperm motility assays were performed as previously described.65 The 10ml aliquots from the sperm sample were taken before dilution and placed into 20 mm deep glass cell chambers (Leja Products BV, NieuwVennep, The Netherlands). The chambers were imaged using an Olympus BX51 microscope (Olympus, Tokyo, Japan) through a 20£ phase objective and maintained at 37 C on a heated platform. The viewing areas on each chamber were captured using a CCD camera. The samples were analyzed via computer-assisted semen analysis using the Mini-tube Sperm Vision Digital Semen Evaluation system. Different sperm motility parameters were analyzed, including the total motility, average path velocity (VAP), straight-line velocity (VSL), and

829

the curvilinear velocity (VCL). Unfixed sperm were spread on glass slides for morphological observation or immunostaining. Immunoblotting Sertoli cell extracts were prepared in cold RIPA-like buffer (50 mM Tris HCl, pH 7.6, 150 mM NaCl, 2 mM EDTA, 1% NP-40 [Sigma, NP40S], 1% sodium deoxycholate [Amresco, D0613], 0.1% sodium dodecyl sulfate [Amresco, 0227]) supplemented with 1 mM phenylmethylsulfonyl fluoride and a protein inhibitor cocktail (Roche Diagnostics, 04693132001) for 30 min on ice. The homogenates were centrifuged at 13,523 £ g for 15 min, and the protein concentrations were determined by the Bio-Rad protein assay. The protein lysates (~25mg) were electrophoresed under reducing conditions in SDS-PAGE gels and transferred onto nitrocellulose membranes. After incubating in primary antibody, the immunoblotting was performed using a fluorescent dye-labeled secondary antibody (Invitrogen), and the blots were scanned using an Odyssey infrared imager (9120, LI-COR Biosciences, Lincoln, NE). Transmission electron microscopy The adult mouse testes were dissected and fixed with 2.5% (vol/ vol) glutaraldehyde in 0.2 M cacodylate buffer overnight. After washing in 0.2 M PBS, the tissues were cut into small pieces of approximately 1 mm3 and immersed in 1% OsO4 in 0.2 M cacodylate buffer for 2 h at 4 C. Then, the samples were dehydrated through a graded ethanol series and embedded in resin (Low Viscosity Embedding Media Spurr’s Kit, EMS, 14300). Ultrathin sections were cut on an ultramicrotome, stained with uranyl acetate and lead citrate, and observed using a JEM-1400 transmission electron microscope (JEOL, Tokyo, Japan). Immunofluorescence and TUNEL assay Testes were dissected from mutant and control mice immediately after euthanasia, fixed in 4% PFA at room temperature for up to 24 h, stored in 70% (vol/vol) ethanol, and embedded in paraffin. The 5-mm sections were prepared and mounted on glass slides. After deparaffinization, sections were boiled for 15 min in sodium citrate buffer for antigen retrieval. For phalloidin and PDLIM1, testes were fixed in 4% PFA at room temperature for 4 h and dehydrated in 30% sucrose. Then, the tissue was embedded in optimum cutting temperature compound (OCT; Tissue-Tek, 4583) and cut into 6-mm sections using a microtome-cryostat (CM1950, Leica, Wetzlar, Germany).The immunofluorescence analysis of testis fragmented material was performed as previously described, with minor modifications.41 Testes were fixed in 4% PFA at room temperature for 2 hand decapsulated in PBS. The seminiferous tubules were minced into small pieces using scalpels and then were gently aspirated, first through an 18-gauge needle and then a 21-gauge needle to fragment the seminiferous epithelium. After sediment for 5 min, the supernatant fraction was collected and concentrated at 800 £ g for 5 min. The fragments were resuspended in a small amount of PBS, and drops of the material were placed on 3-aminopropyl-triethoxysilane (APES; Zhong Shan Jin Qiao, ZL1-9002) precoated slides for 10 min. After

830

C. LIU ET AL.

removing excess fluid, the slides were air dried. After washing with PBS 3 times and blocking with 5% bovine serum albumin (Amresco, Solon, OH, AP0027), the primary antibody was added to the sections and incubated at 4 C overnight, followed by incubation with the secondary antibody. The nuclei were stained with 40 ,6-diamidino-2-phenylindole (DAPI). The images were taken immediately using a LSM 780 microscope (Zeiss, Oberkochen, Germany) or TCS SP8 microscope (Leica, Wetzlar, Germany). The super-resolution images were detected with a TCS SP8 STED 3X microscope system (Leica, Wetzlar, Germany).For cellular immunofluorescence, the Sertoli cells were plated on cover glass slips, and 24 h later, the cells were rinsed 3 times with PBS, fixed at room temperature, and stained as detailed above. Apoptotic cells in testis were detected using the terminal deoxynucleotidyl transferase dUTP nick end-labeling (TUNEL) assay (In Situ Cell Death Detection Kit; Roche, 11684795910) according to the manufacturer’s instructions.66 Briefly, sections of testes were deparaffinized and boiled for 15 min in sodium citrate buffer for antigen retrieval. After treated with proteinase K for 15 min at room temperature and rinsed twice with PBS, the TUNEL reaction mixture was added and incubated in a humidified atmosphere for 60 min at 37 C in the dark, following by immunofluorescence staining as detailed above. Immunohistochemistry The paraffin sections were fixed with 4% PFA and rinsed 3 times in PBS. Then, the sections were boiled for 15 min in sodium citrate buffer for antigen retrieval. Next, 3% H2O2 was added to the sections to eliminate the endogenous peroxidase activity. After blocking with 5% bovine serum albumin, each section was incubated with the primary antibody at 4 C overnight, followed by staining with the HRP-conjugated secondary antibody. The negative controls were prepared without the primary antibody. Finally, the sections were stained with 3,30 -diaminobenzidine (DAB; Zhong Shan Jin Qiao, ZL1-9018), and the nuclei were stained with hematoxylin. The images were acquired using a Nikon 80i inverted microscope equipped with a CCD camera (Nikon, Tokyo, Japan ). Cycloheximidechase assay Sertoli cells were plated one day before the experiment, and cycloheximide (CHX; Sigma, R750107) was added to the culture at 100 mg/ml to block new protein synthesis. Samples were taken at 0, 2 and 6 after the addition of CHX for 2 h. Protein levels were determined by immunoblotting. Immunoprecipitation Transfected cells were lysed in TAP lysis buffer (50 mM HEPES-KOH, pH7.5, 100 mM KCl, 2 mM EDTA, 10%glycerol, 0.1% NP-40 10 mM NaF, 0.25 mM Na3VO4, 50 mM b-glycerolphosphate) plus protease inhibitors (Roche, 04693132001) for 30 min on ice, and centrifuged at 13,523 £ g for 15 min. For immunoprecipitation, cell lysates were incubated with primary antibody overnight at 4 C, and then incubated with

protein A-Sepharose (GE, 17-1279-03) for 2 h at 4 C. Thereafter, the precipitants were washed 2 times with IP buffer (20 mM Tris, pH 7.4, 2 mM EGTA, 1% NP-40), and the immune complexes were eluted with sample buffer containing 1% SDS for 10 min at 55 C and analyzed by immunoblotting. Statistical analysis All data are presented as the mean § SEM. The statistical significance of the differences between the mean values for the different genotypes was measured by the Student t test with a paired, 2-tailed distribution. The data were considered significant when the P value was less than 0.05 () or 0.01 ().

Abbreviations aES APES bES BTB CHX CQ DAB DAPI ER ES H&E HRP IF IHC IP MEF MS NH4Cl OCT PBS PFA PI SSCs, STRING TBC TEM TUNEL VAP VCL VSL

apical ectoplasmic specialization 3-aminopropyl-triethoxysilane basal ectoplasmic specialization blood–testis barrier cycloheximide chloroquine 3, 30 -diaminobenzidine 40 , 6-diamidino-2- phenylindole endoplasmic reticulum ectoplasmic specialization hematoxylin and eosin staining horseradish peroxidase immunofluorescent staining immunohistochemistry immunoprecipitation mouse embryonic fibroblast mass spectrometry ammonium chloride optimum cutting temperature compound phosphate-buffered saline paraformaldehyde propidium iodide spermatogonial stem cells Search Tool for the Retrieval of Interacting Genes/ Proteins tubulobulbar complex transmission electron microscopy terminal deoxynucleotidyl transferase dUTP nick end-labeling average path velocity curvilinear velocity straight-line velocity

Disclosure of potential conflicts of interest The authors declare that they have no conflict of interest.

Acknowledgments We thank Noboru Misushima and Masaaki Komatsufor providing the Atg5 and Atg7 floxed mice. We thank Hong Zhang for critical reading of the manuscript.

AUTOPHAGY

Funding This work was supported by the National Natural Science Foundation of China (Grant No. 31471277, 31171374), the Major Basic Research Program (Grant No. 2012CB944404) and the Knowledge Innovation Program of Chinese Academy of Science (Grant No. KSCX2-YW-N-071).

References [1] Roosen-Runge EC. The process of spermatogenesis in mammals. Biol Rev Camb Philos Soc 1962; 37:343-77; PMID:14493721; http:// dx.doi.org/10.1111/j.1469-185X.1962.tb01616.x [2] Hess RA, Renato de Franca L. Spermatogenesis and cycle of the seminiferous epithelium. Advan Exp Med Biol 2008; 636:1-15; PMID:19856159; http://dx.doi.org/10.1007/978-0-387-09597-4_1 [3] Griswold MD. The central role of Sertoli cells in spermatogenesis. Semin Cell Dev Biol 1998; 9:411-6; PMID:9813187; http://dx.doi.org/ 10.1006/scdb.1998.0203 [4] Hai Y, Hou J, Liu Y, Yang H, Li Z, He Z. The roles and regulation of Sertoli cells in fate determinations of spermatogonial stem cells and spermatogenesis. Semin Cell Dev Biol 2014; 29:66-75; PMID:24718316; http://dx.doi.org/10.1016/j.semcdb.2014.04.007 [5] Jegou B. The Sertoli-germ cell communication network in mammals. Int Rev Cytol 1993; 147:25-96; PMID:8225836; http://dx.doi.org/ 10.1016/S0074-7696(08)60766-4 [6] Berruti G, Paiardi C. The dynamic of the apical ectoplasmic specialization between spermatids and Sertoli cells: the case of the small GTPase Rap1. Bio Med Res Int 2014; 2014:635979; PMID:24719879; http://dx.doi.org/10.1155/2014/635979 [7] Kopera IA, Bilinska B, Cheng CY, Mruk DD. Sertoli-germ cell junctions in the testis: a review of recent data. Philos Trans R Soc Lond B Biol Sci 2010; 365:1593-605; PMID:20403872; http://dx.doi.org/ 10.1098/rstb.2009.0251 [8] Lie PP, Cheng CY, Mruk DD. The biology of the desmosome-like junction a versatile anchoring junction and signal transducer in the seminiferous epithelium. Int Rev Cell Mol Biol 2011; 286:223-69; PMID:21199783; http://dx.doi.org/10.1016/B978-0-12-3858597.00005-7 [9] Qian X, Mruk DD, Cheng YH, Tang EI, Han D, Lee WM, Wong EW, Cheng CY. Actin binding proteins, spermatid transport and spermiation. Semin Cell Dev Biol 2014; 30:75-85; PMID:24735648; http://dx. doi.org/10.1016/j.semcdb.2014.04.018 [10] Toyama Y, Maekawa M, Yuasa S. Ectoplasmic specializations in the Sertoli cell: new vistas based on genetic defects and testicular toxicology. Anat Sci Int 2003; 78:1-16; PMID:12680465; http://dx.doi.org/ 10.1046/j.0022-7722.2003.00034.x [11] Wong EW, Mruk DD, Cheng CY. Biology and regulation of ectoplasmic specialization, an atypical adherens junction type, in the testis. Biochimica et biophysica acta 2008; 1778:692-708; PMID:18068662; http://dx.doi.org/10.1016/j.bbamem.2007.11.006 [12] Cheng CY, Mruk DD. The blood-testis barrier and its implications for male contraception. Pharmacol Rev 2012; 64:16-64; PMID:22039149; http://dx.doi.org/10.1124/pr.110.002790 [13] Mruk DD, Cheng CY. Cell-cell interactions at the ectoplasmic specialization in the testis. Trends in endocrinology and metabolism: TEM 2004; 15:439-47; PMID:15519891; http://dx.doi.org/10.1016/ S1043-2760(04)00219-X [14] Lee NP, Cheng CY. Ectoplasmic specialization, a testis-specific cellcell actin-based adherens junction type: is this a potential target for male contraceptive development? Human reproduction update 2004; 10:349-69; PMID:15192055; http://dx.doi.org/10.1093/humupd/ dmh026 [15] Cheng CY, Mruk DD. Cell junction dynamics in the testis: Sertoligerm cell interactions and male contraceptive development. Physiological reviews 2002; 82:825-74; PMID:12270945; http://dx.doi.org/ 10.1152/physrev.00009.2002 [16] Li SY, Mruk DD, Cheng CY. Focal adhesion kinase is a regulator of Factin dynamics: New insights from studies in the testis. Spermatogenesis 2013; 3:e25385; PMID:24381802; http://dx.doi.org/10.4161/spmg.25385

831

[17] Mruk DD, Silvestrini B, Cheng CY. Anchoring junctions as drug targets: role in contraceptive development. Pharmacological reviews 2008; 60:146-80; PMID:18483144; http://dx.doi.org/10.1124/pr.107.07105 [18] O’Donnell L, Nicholls PK, O’Bryan MK, McLachlan RI, Stanton PG. Spermiation: The process of sperm release. Spermatogenesis 2011; 1:14-35; PMID:21866274; http://dx.doi.org/10.4161/spmg.1.1.14525 [19] Sperry AO. The dynamic cytoskeleton of the developing male germ cell. Biol Cell 2012; 104:297-305; PMID:22276751; http://dx.doi.org/ 10.1111/boc.201100102 [20] Mizushima N, Yoshimori T, Ohsumi Y. The role of Atg proteins in autophagosome formation. Ann Rev Cell Dev Biol 2011; 27:107-32; PMID:21801009; http://dx.doi.org/10.1146/annurev-cellbio-092910154005 [21] Mizushima N, Levine B. Autophagy in mammalian development and differentiation. Nat Cell Biol 2010; 12:823-30; PMID:20811354; http://dx.doi.org/10.1038/ncb0910-823 [22] Yang Z, Klionsky DJ. An overview of the molecular mechanism of autophagy. Curr Top Microbiol Immunol 2009; 335:1-32; PMID:19802558 [23] Wang H, Wan H, Li X, Liu W, Chen Q, Wang Y, Yang L, Tang H, Zhang X, Duan E, et al. Atg7 is required for acrosome biogenesis during spermatogenesis in mice. Cell Res 2014; 24:852-69; PMID:24853953; http://dx.doi.org/10.1038/cr.2014.70 [24] Tang Z, Lin MG, Stowe TR, Chen S, Zhu M, Stearns T, Franco B, Zhong Q. Autophagy promotes primary ciliogenesis by removing OFD1 from centriolar satellites. Nature 2013; 502:254-7; PMID:24089205; http://dx.doi.org/10.1038/nature12606 [25] Mizushima N, Komatsu M. Autophagy: renovation of cells and tissues. Cell 2011; 147:728-41; PMID:22078875; http://dx.doi.org/ 10.1016/j.cell.2011.10.026 [26] Ma Y, Yang HZ, Xu LM, Huang YR, Dai HL, Kang XN. Testosterone regulates the autophagic clearance of androgen binding protein in rat Sertoli cells. Scientific Reports 2015; 5:8894; PMID:25745956; http://dx.doi.org/10.1038/srep08894 [27] Yefimova MG, Messaddeq N, Harnois T, Meunier AC, Clarhaut J, Noblanc A, Weickert JL, Cantereau A, Philippe M, Bourmeyster N, et al. A chimerical phagocytosis model reveals the recruitment by Sertoli cells of autophagy for the degradation of ingested illegitimate substrates. Autophagy 2013; 9:653-66; PMID:23439251; http://dx. doi.org/10.4161/auto.23839 [28] Chen Y, Zhou Y, Wang X, Qian W, Han X. Microcystin-LR induces autophagy and apoptosis in rat Sertoli cells in vitro. Toxicon 2013; 76:8493; PMID:24047964; http://dx.doi.org/10.1016/j.toxicon.2013.09.005 [29] Eid N, Ito Y, Otsuki Y. Enhanced mitophagy in Sertoli cells of ethanol-treated rats: morphological evidence and clinical relevance. J Mol Histol 2012; 43:71-80; PMID:22076330; http://dx.doi.org/ 10.1007/s10735-011-9372-0 [30] Hara T, Nakamura K, Matsui M, Yamamoto A, Nakahara Y, SuzukiMigishima R, Yokoyama M, Mishima K, Saito I, Okano H, et al. Suppression of basal autophagy in neural cells causes neurodegenerative disease in mice. Nature 2006; 441:885-9; PMID:16625204; http://dx. doi.org/10.1038/nature04724 [31] Komatsu M, Waguri S, Ueno T, Iwata J, Murata S, Tanida I, Ezaki J, Mizushima N, Ohsumi Y, Uchiyama Y, et al. Impairment of starvation-induced and constitutive autophagy in Atg7-deficient mice. J Cell Biol 2005; 169:425-34; PMID:15866887; http://dx.doi.org/ 10.1083/jcb.200412022 [32] Lecureuil C, Fontaine I, Crepieux P, Guillou F. Sertoli and granulosa cell-specific Cre recombinase activity in transgenic mice. Genesis 2002; 33:114-8; PMID:12124943; http://dx.doi.org/10.1002/gene. 10100 [33] Klionsky DJ, Abdelmohsen K, Abe A, Abedin MJ, Abeliovich H, Acevedo Arozena A, Adachi H, Adams CM, Adams PD, Adeli K, et al. Guidelines for the use and interpretation of assays for monitoring autophagy (3rd edition). Autophagy 2016; 12:1-222; PMID:26799652; http://dx.doi.org/10.1080/15548627.2015.1100356 [34] Mundlos S, Pelletier J, Darveau A, Bachmann M, Winterpacht A, Zabel B. Nuclear localization of the protein encoded by the Wilms’ tumor gene WT1 in embryonic and adult tissues. Development 1993; 119:1329-41; PMID:8306891

832

C. LIU ET AL.

[35] Kim KS, Cha MC, Gerton GL. Mouse sperm protein sp56 is a component of the acrosomal matrix. Biol Reprod 2001; 64:36-43; PMID:11133656; http://dx.doi.org/10.1095/biolreprod64.1.36 [36] Goodson SG, Zhang Z, Tsuruta JK, Wang W, O’Brien DA. Classification of mouse sperm motility patterns using an automated multiclass support vector machines model. Biol Reprod 2011; 84:1207-15; PMID:21349820; http://dx.doi.org/10.1095/biolreprod.110.088989 [37] Costoya JA, Hobbs RM, Barna M, Cattoretti G, Manova K, Sukhwani M, Orwig KE, Wolgemuth DJ, Pandolfi PP. Essential role of Plzf in maintenance of spermatogonial stem cells. Nat Genet 2004; 36:653-9; PMID:15156143; http://dx.doi.org/10.1038/ng1367 [38] Kierszenbaum AL, Tres LL. The acrosome-acroplaxome-manchette complex and the shaping of the spermatid head. Arch Histol Cytol 2004; 67:271-84; PMID:15700535; http://dx.doi.org/10.1679/aohc.67.271 [39] Kierszenbaum AL, Rivkin E, Tres LL. Molecular biology of sperm head shaping. Soc Reprod Fertility Suppl 2007; 65:33-43; PMID:17644953 [40] O’Donnell L, O’Bryan MK. Microtubules and spermatogenesis. Semin Cell Dev Biol 2014; 30:45-54; PMID:24440897; http://dx.doi. org/10.1016/j.semcdb.2014.01.003 [41] Vogl AW, Grove BD, Lew GJ. Distribution of Actin in Sertoli-Cell Ectoplasmic Specializations and Associated Spermatids in the Ground-Squirrel Testis. Anatomical Record 1986; 215:331-41; PMID:3526980; http://dx.doi.org/10.1002/ar.1092150402 [42] Vogl AW, Du M, Wang XY, Young JS. Novel clathrin/actin-based endocytic machinery associated with junction turnover in the seminiferous epithelium. Semin Cell Dev Biol 2014; 30:55-64; PMID:24280271; http://dx.doi.org/10.1016/j.semcdb.2013.11.002 [43] Upadhyay RD, Kumar AV, Ganeshan M, Balasinor NH. Tubulobulbar complex: cytoskeletal remodeling to release spermatozoa. Reprod Biol Endocrinol 2012; 10:27; http://dx.doi.org/10.1186/1477-7827-10-27 [44] Young JS, Guttman JA, Vaid KS, Vogl AW. Tubulobulbar complexes are intercellular podosome-like structures that internalize intact intercellular junctions during epithelial remodeling events in the rat testis. Biol Reprod 2009; 80:162-74; PMID:18799754; http://dx.doi. org/10.1095/biolreprod.108.070623 [45] Monastyrska I, Rieter E, Klionsky DJ, Reggiori F. Multiple roles of the cytoskeleton in autophagy. Biol Rev Camb Philos Soc 2009; 84:431-48; PMID:19659885; http://dx.doi.org/10.1111/j.1469-185X.2009.00082.x [46] Mackeh R, Perdiz D, Lorin S, Codogno P, Pous C. Autophagy and microtubules - new story, old players. J Cell Sci 2013; 126:1071-80; PMID:23620510; http://dx.doi.org/10.1242/jcs.115626 [47] Zhuo C, Ji Y, Chen Z, Kitazato K, Xiang Y, Zhong M, Wang Q, Pei Y, Ju H, Wang Y. - MEFs reveals a close relationship between F-actin and autophagy. Biochem Biophys Res Commun 2013; 437:482-8; PMID:23850690; http://dx.doi.org/10.1016/j.bbrc.2013.06.111 [48] Vallenius T, Luukko K, Makela TP. CLP-36 PDZ-LIM protein associates with nonmuscle alpha-actinin-1 and alpha-actinin-4. J Biol Chem 2000; 275:11100-5; PMID:10753915; http://dx.doi.org/ 10.1074/jbc.275.15.11100 [49] Szczygiel M, Kurpisz M. Teratozoospermia and its effect on male fertility potential. Andrologia 1999; 31:63-75; PMID:10097795; http:// dx.doi.org/10.1111/j.1439-0272.1999.tb02848.x [50] Breitbart H, Cohen G, Rubinstein S. Role of actin cytoskeleton in mammalian sperm capacitation and the acrosome reaction. Reproduction 2005; 129:263-8; PMID:15749953; http://dx.doi.org/10.1530/rep.1.00269 [51] Sun X, Kovacs T, Hu YJ, Yang WX. The role of actin and myosin during spermatogenesis. Mol Biol Rep 2011; 38:3993-4001; PMID:21107714; http://dx.doi.org/10.1007/s11033-010-0517-0 [52] O’Donnell L, Stanton PG, Bartles JR, Robertson DM. Sertoli cell ectoplasmic specializations in the seminiferous epithelium of the testosterone-suppressed adult rat. Biol Reprod 2000; 63:99-108; PMID: 10859247; http://dx.doi.org/10.1095/biolreprod63.1.99

[53] Wolski KM, Mruk DD, Cameron DF. The Sertoli-spermatid junctional complex adhesion strength is affected in vitro by adjudin. J Androl 2006; 27:790-4; PMID:16809272; http://dx.doi.org/10.2164/ jandrol.106.000422 [54] Vogl AW, Pfeiffer DC, Mulholland D, Kimel G, Guttman J. Unique and multifunctional adhesion junctions in the testis: ectoplasmic specializations. Arch Histol Cytol 2000; 63:1-15; PMID:10770585; http:// dx.doi.org/10.1679/aohc.63.1 [55] Lie PP, Mruk DD, Lee WM, Cheng CY. Cytoskeletal dynamics and spermatogenesis. Philos Trans R Soc Lond B Biol Sci 2010; 365:158192; PMID:20403871; http://dx.doi.org/10.1098/rstb.2009.0261 [56] Chemes H. The phagocytic function of Sertoli cells: a morphological, biochemical, and endocrinological study of lysosomes and acid phosphatase localization in the rat testis. Endocrinology 1986; 119:1673-81; PMID:3757907; http://dx.doi.org/10.1210/endo-119-41673 [57] Corcelle E, Nebout M, Bekri S, Gauthier N, Hofman P, Poujeol P, Fenichel P, Mograbi B. Disruption of autophagy at the maturation step by the carcinogen lindane is associated with the sustained mitogen-activated protein kinase/extracellular signal-regulated kinase activity. Cancer Res 2006; 66:6861-70; PMID:16818664; http://dx. doi.org/10.1158/0008-5472.CAN-05-3557 [58] Nagda G, Bhatt DK. Alleviation of lindane induced toxicity in testis of Swiss mice (Mus musculus) by combined treatment with vitamin C, vitamin E and alpha-lipoic acid. Indian J Exp Biol 2011; 49:191-9; PMID:21452598 [59] Bauer K, Kratzer M, Otte M, de Quintana KL, Hagmann J, Arnold GJ, Eckerskorn C, Lottspeich F, Siess W. Human CLP36, a PDZdomain and LIM-domain protein, binds to alpha-actinin-1 and associates with actin filaments and stress fibers in activated platelets and endothelial cells. Blood 2000; 96:4236-45; PMID:11110697 [60] Hasegawa T, Ohno K, Funahashi S, Miyazaki K, Nagano A, Sato K. CLP36 interacts with palladin in dorsal root ganglion neurons. Neurosci Lett 2010; 476:53-7; PMID:20381583; http://dx.doi.org/ 10.1016/j.neulet.2010.03.081 [61] Grooman B, Fujiwara I, Otey C, Upadhyaya A. Morphology and viscoelasticity of actin networks formed with the mutually interacting crosslinkers: palladin and alpha-actinin. PloS one 2012; 7: e42773; PMID:22916157; http://dx.doi.org/10.1371/journal.pone. 0042773 [62] Ulbricht A, Gehlert S, Leciejewski B, Schiffer T, Bloch W, Hohfeld J. Induction and adaptation of chaperone-assisted selective autophagy CASA in response to resistance exercise in human skeletal muscle. Autophagy 2015; 11:538-46; PMID:25714469; http://dx.doi.org/ 10.1080/15548627.2015.1017186 [63] Ridley AJ, Hall A. The small GTP-binding protein rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors. Cell 1992; 70:389-99; PMID:1643657; http://dx.doi. org/10.1016/0092-8674(92)90163-7 [64] van der Wee KS, Johnson EW, Dirami G, Dym TM, Hofmann MC. Immunomagnetic isolation and long-term culture of mouse type A spermatogonia. J Androl 2001; 22:696-704; PMID:11451367 [65] Jimenez T, Sanchez G, Wertheimer E, Blanco G. Activity of the Na, K-ATPase alpha 4 isoform is important for membrane potential, intracellular Ca2C, and pH to maintain motility in rat spermatozoa. Reproduction 2010; 139:835-45; PMID:20179187; http://dx.doi.org/ 10.1530/REP-09-0495 [66] Song ZH, Yu HY, Wang P, Mao GK, Liu WX, Li MN, Wang HN, Shang YL, Liu C, Xu ZL, et al. Germ cell-specific Atg7 knockout results in primary ovarian insufficiency in female mice. Cell Death Dis 2015; 6:e1589; PMID:25590799; http://dx.doi.org/10.1038/ cddis.2014.559