The potential function of KIF17 in large yellow croaker (Larimichthys crocea) spermatid remodeling: molecular characterization and expression pattern during spermiogenesis

KIF17, which belongs to the kinesin-2 protein family, plays an indispensable role in mammalian spermiogenesis. However, the role of KIF17 in fish spermatid remodeling during spermiogenesis remains poorly understood. Therefore, we aimed to study the role of KIF17 in spermatid remodeling during Larimichthys crocea (L. crocea) spermiogenesis. The kif17 cDNA sequence, 3247 bp in length, was cloned from L. crocea testis, which consisted of a 347-bp 5′-untranslated region (UTR), 413-bp 3′ -UTR, and 2487-bp open reading frame. Bioinformatic analyses revealed that KIF17 obtained from L. crocea (Lc-KIF17) exhibited a high sequence identity compared with those from other teleosts and possessed the structural features of other kinesin-2 proteins. Based on structural similarity, we speculate that the role of Lc-KIF17 may be similar to that of KIF17 in other animals. Lc-kif17 mRNA was diffusely expressed in L. crocea tissues and was highly expressed in the testis, especially at stage IV testicular development. Immunofluorescence analysis revealed that Lc-KIF17 signals colocalized with β-tubulin signals and migrated from the perinuclear cytoplasm to the side of the nucleus where the tail forms during spermiogenesis. These findings revealed that KIF17 may be involved in L. crocea spermiogenesis. In particular, KIF17 may participate in spermatid remodeling by interacting with perinuclear microtubules during L. crocea spermiogenesis. Collectively, this study contributes to an improved understanding of the mechanism underlying L. crocea spermiogenesis and provides a basis for further research on L. crocea reproduction and development.


Introduction
Spermiogenesis is a sequential spermatid remodeling, in which round spermatids go through substantial structural and functional changes to differentiate into spermatozoa (Hermo et al. 2010). Such changes, which include the formation of the acrosome, nucleus shaping, disposal of excess cytoplasm, and generation of the sperm tail, contribute to the production of viable sperm and transmission of genetic information to the succeeding generation (Ma et al. 2017). Some of these changes in spermiogenesis rely on the regulation of microtubules, as proven in modified mouse models, which have shown that microtubules are essential for male fertility (O'Donnell and O'Bryan 2014).
Microtubules, an important type of cytoskeleton, consist of α-tubulin and β-tubulin heterodimers and act as "railways" for protein transport during spermiogenesis (Goodson and Jonasson 2018). Cargo protein transport via microtubules during spermatid remodeling is essential for proper acrosome and sperm tail formation and nuclear shaping. Kinesins are microtubule-associated molecular motor proteins that transport cargo proteins along microtubules using energy derived from ATP hydrolysis (Miki et al. 2005). They are divided into 15 distinct families based on sequence similarity (Lawrence et al. 2004). All kinesin family members share a conserved motor head domain (confers ATPase and microtubule-binding activities) that is linked by a coiledcoil stalk (responsible for dimerization) to a greatly diverged cargo-binding tail domain (associated with multiple classes of cargo proteins) (Hirokawa and Takemura 2005). KIF17 is a kinesin-2 family member that mainly exists as a homodimer (KIF17) or a heterotrimer (KIF3A, KIF3B/KIF3C, and KAP3). The structural characteristics of KIF17 are consistent with those of other kinesin family members (Setou 2000). KIF17 plays an indispensable role in spermiogenesis (Macho 2002;Kimmins et al. 2004;Hogeveen and Sassone-Corsi 2006). Previous studies have demonstrated that KIF17 colocalizes with its cargo protein in the manchette and the principal piece of the sperm tail and participates in mouse spermiogenesis (Saade et al. 2007). There have been studies on the function of KIF17 in mammalian spermiogenesis; however, very few studies have focused on the role of KIF17 in fish spermiogenesis, especially with regarding marine fish.
Large yellow croaker (Larimichthys crocea), a marine fish species belonging to the family Sciaenidae, has a high commercial value in China (Chen et al. 2010). L. crocea undergoes spermatid remodeling, which involves nuclear reshaping, tail formation, and excess cytoplasm removal, during spermiogenesis. It has been used as a model for studying spermiogenesis in fish (Luo et al. 2019;Mu et al. 2019;Zhang et al. 2017). This study aimed to investigate the functions of KIF17 in spermatid remodeling during L. crocea spermiogenesis to contribute to improved understanding of the mechanism underlying fish spermiogenesis and to provide a basis for further research on fish reproduction and development.

Tissue sampling
Tissues (heart, liver, intestine, kidney, brain, muscle, and gill of stage IV L. crocea) and testes (at developmental stages II-V) were collected from six healthy male L. crocea subjects supplied by Gangwan Aquatic Fingerlings Limited Company (Xiangshan, Zhejiang, China).

RNA extraction and cDNA preparation
TRIzol Reagent RNA (Invitrogen, San Diego, CA, USA) was used to extract total RNA from each L. crocea tissue sample. The quantitation and purity of the extracted RNA were determined by agarose gel electrophoresis and nucleic acid protein analyzer. Prime-Script® RT Reagent Kit (Takara, Dalian, China) was used in the quantitative real-time polymerase chain reaction (qRT-PCR) analysis for First-Strand cDNA synthesis. RT-PCR for First-Strand cDNA synthesis was conducted as follows: 37 °C for 15 min and 85 °C for 5 s. Smart Rapid Amplification of cDNA Ends (RACE) cDNA Amplification Kit (Clontech, Mountain View, CA, USA) was used to synthesize the first-strand cDNA using 5′-RACE and 3′-RACE. All operations were performed according to a previous study ).
Cloning and sequence analyses of kif17 cDNA A full-length kif17 cDNA clone from L. crocea (Lc-kif17) was amplified using degenerate PCR and the RACE approach according to a previous study  30 s, and 72 °C for 1 min; and 72 °C for 10 min for the final extension. The 3′/5′ RACE PCR was conducted as follows: 94 RC for 5 min; 8 cycles of 94 C for 30 s, 65 C/68 6C for 30 s (decreased by 0.5 (C/cycle), and 72 aC for 1.5 min; 27 cycles of 94 C for 30 s, 65 C/66 6C for 30 s, and 72 aC for 1.5 min; followed by a final extension at 72 °C for 10 min. The PCR products were sequenced through GENEWIZ (Suzhou, China). The deduced amino acid sequence of the KIF17 protein from L. crocea (Lc-KIF17) was obtained using the Sequence Manipulation Suite (http:// www. bio-soft. net/ sms/ index. html). The molecular mass (Mw) and theoretical isoelectric point (pI) of Lc-KIF17 were predicted using the Compute pI/Mw tool (http:// cn. expasy. org/ tools/ pi_ tool. html). The functional domains and tertiary structures of KIF17 proteins were predicted using the Simple Modular Architecture Research Tool (SMART) (http:// smart. emblheide lberg. de/) and Iterative Threading Assembly Refinement (I-TASSER) (http:// zhang lab. ccmb. med. umich. edu/I-TASSER/), respectively. Vector NTI10 software was used to perform multiple protein sequence alignment analyses. Phylogenetic analysis of KIF17 was conducted using MEGA 5.0 software (Wang et al. 2019).

Quantitative analysis of Lc-kif17 mRNA
The distribution of kif17 mRNA in the different tissues of L. crocea and at different stages of testicular development was performed by qRT-PCR as described previously . The specific primer Lc-kif17-RT-F/R (Table 1) was synthesized by GENEWIZ (Suzhou, China) and was used to amplify the target gene. The β-actin (Actb) gene was chosen as an internal control. Statistical analyses performed with SPSS 16.0 were using one-way analysis of variance (ANOVA) followed by the Duncan's multiple comparison, where P < 0.05 indicated a significant difference.
Fluorescence in situ hybridization (FISH) L. crocea testes at stage IV of testicular development were extracted, and frozen testis sections embedded in optimal cutting temperature compound were produced as described previously ). The RNA fluorescence probes of Lc-kif17 for fluorescence in situ hybridization (FISH) were designed using Primer 5.0 software (Table 1) and were conjugated to fluorescein using fluorescein isothiocyanate (FITC) synthesized by GENEWIZ (Suzhou, China). The specificity of the RNA fluorescence probes was evaluated using the Basic Local Alignment Search   (7) Alexa Fluor 488-labeled donkey anti-rabbit IgG (H + L) (diluted 1:500 for IF staining). All antibodies were obtained from Beyotime (China).

Western blotting
The specificity of the mouse anti-KIF17 antibody and tissue distribution of Lc-KIF17 protein were determined by western blotting, as described by Wang et al. (2019) and Liang et al. (2020). Briefly, total protein was extracted using radioimmunoprecipitation assay (RIPA) lysis buffer (Solarbio, Shanghai, China), diluted with 5 × sodium dodecyl sulfate (SDS) sample buffer, and boiled for 10 min. The protein samples were loaded on 8% polyacrylamide gels, subjected to SDS-polyacrylamide electrophoresis (SDS-PAGE), and transferred onto polyvinylidene fluoride (PVDF) membranes (Bio-Rad, Hercules, CA, USA). The PVDF membranes were blocked with 5% skimmed milk in phosphate-buffered saline (PBS) with Tween 20 (PBS-Tween 20) for 2 h and incubated with anti-KIF17 antibody at 4 °C overnight. Anti-rabbit GAPDH antibody was used as a control. The membranes were washed with 0.1% PBS-Tween 20 to remove the excess primary antibody and were incubated with the secondary antibody for 1 h at 37 °C. The membranes were then washed with 0.1% PBS-Tween 20. Finally, the signals were visualized through chemiluminescence imaging (Tanon 5200; Shanghai, China).
Immunofluorescence staining IF experiments were performed as described by . After drying at room temperature for 15 min, the frozen testis sections of male L. crocea were permeabilized with 0.3% PBS with Triton X-100 (PBS-Triton X-100) for 20 min, followed by blocking with 5% bovine serum albumin in 0.1% PBS-Triton X-100 for 1.5 h and incubation with primary antibody overnight at 4 °C. The frozen sections were washed with 0.1% PBS-Triton X-100 to remove the excess primary antibody and were incubated with the secondary antibody for 1 h. Excess secondary antibody was removed using 0.1% PBS-Triton X-100, and diamino-2-phenylindole (DAPI; Beyotime, China) was used to stain the nuclei for 5 min. Finally, the testis sections were mounted with Antifade Mounting Medium (Beyotime, China) and viewed using a confocal laser-scanning microscope (LSM710/780; Carl Zeiss, Oberkochen, Germany).
Lc-kif17 encoded a protein composed of 828 amino acids, with a calculated molecular weight of 92.3 kDa and pI of 7.06. The functional domains of KIF17 proteins were predicted to be mainly composed of a conserved motor head domain at amino acid (aa) positions 3 to 343 (3-343 aa), a stalk domain (344-654 aa) containing three coiled-coil regions (395-443 aa, 546-577 aa, and 614-654 aa), and a cargo-binding tail domain (655-844 aa) (Fig. 2a). Using Batch CD-Search, the motor domain (3-343 aa) was determined to be conserved in KIF17 proteins. This domain consists of ATP-hydrolysis sites and microtubule-binding sites and is responsible for the movement of proteins using energy produced from ATP hydrolysis (Hirokawa and Takemura 2005). Using I-TASSER, a 3D structural model of Lc-KIF17 was built, showing similarities with the structures of other kinesin-2 family members (Fig. 2b). A model of the putative homodimeric Lc-KIF17 is shown in Fig. 2c.
The deduced motor domain sequence of Lc-KIF17 displayed a high identity with those of KIF17 orthologs from other teleosts, such as Danio rerio (84.5%), Anabas testudineus (93.8%), Astatotilapia calliptera (93.5%), Oncorhynchus tshawytscha (88%), Oryzias latipes (82.7%), Mastacembelus armatus (92.4%), and Larimichthys polyactis (99.1%) (Fig. 3). Phylogenetic analysis revealed that KIF17 was closely clustered with L. polyactis KIF17 and belonged to a clade with other kinesin-2 family members from other teleosts (Fig. 4). Tissue expression of Lc-KIF17 at mRNA and protein levels Lc-kif17 mRNA expression in different tissues of L. crocea was detected by qRT-PCR. The results showed that Lc-kif17 mRNA was expressed in the heart, liver, intestines, kidneys, brain, testes, muscles, and gills, with the brain exhibiting the highest expression level, followed by the testis (Fig. 5a). The specificity of the mouse anti-KIF17 polyclonal antibody was evaluated by western blotting, which showed that this polyclonal antibody can be used against Lc-KIF17 (Fig. 5c). Similar to the mRNA expression pattern, the protein expression pattern showed that Lc-KIF17 was highly expressed in the testes (Fig. 5d).
Expression of the Lc-kif17 mRNA and the Lc-KIF17 protein during stages II-V of testicular development We further evaluated the expression of Lc-kif17 mRNA during the different stages (II-V) of the development of L. crocea testis. The results showed that the Lc-kif17 mRNA increased gradually with testicular development, with stage IV showing the highest expression level (Fig. 5b). Similarly, the expression level of the Lc-KIF17 protein was highest at stage IV (Fig. 5e, h).
Spatiotemporal expression of the Lc-kif17 mRNA during L. crocea spermiogenesis To examine the spatiotemporal expression of the Lc-kif17 mRNA during the spermiogenesis of L. crocea, we designed probes (Table 1) for the Lc-kif17 mRNA using Primer 5.0 and detected the mRNA subcellular localization and expression level via FISH (Fig. 6). Lc-kif17 mRNA signals were distributed in the perinuclear cytoplasm at the early stage of spermiogenesis (Fig. 6, a1-a4) and were enhanced and distributed around the nucleus at the middle stage (Fig. 6, b1-b4). At the late stage of spermiogenesis, Lc-kif17 mRNA signals were detected on the side of the nucleus where the tail forms (Fig. 6, c1-c4). In mature sperm, the mRNA signals were mainly distributed in the sperm midpiece (Fig. 6, d1-d4).
Immunofluorescence localization of Lc-KIF17 and tubulin during L. crocea spermiogenesis After the evaluating the spatiotemporal expression of the Lc-kif17 mRNA during L. crocea spermiogenesis, we further studied the spatiotemporal distribution of Lc-KIF17 and tubulin during spermiogenesis via IF (Fig. 7). The results showed that Lc-KIF17 colocalized with tubulin during spermiogenesis. Similar to Lc-kif17 mRNA signals, Lc-KIF17 protein signals were randomly distributed in the perinuclear cytoplasm at the early stage of spermiogenesis (Fig. 7, a1-a4) and were enhanced and distributed around the nucleus at the middle stage (Fig. 7, b1-b4). At the late stage of spermiogenesis, Lc-KIF17 protein signals were detected on the side of the nucleus where the tail forms (Fig. 7, c1-c4). In mature sperm, the protein signals were mainly distributed in the sperm midpiece (Fig. 7, d1-d4).
In the present study, kif17 cDNA was cloned from L. crocea, and the amino acid sequence of Lc-KIF17 was similar to those of other kinesin-2 family members. Characteristic features of members of the kinesin-2 family include a motor head domain, a coiledcoil stalk domain, and a cargo-binding tail domain (Setou 2000). The motor domain confers ATPase and microtubule-binding activities, while the cargobinding domain interacts with specific cargo proteins. The stalk region, which differs in KIF3A, KIF3B, and KIF3C, contains three coiled-coil domains related to the autoinhibition caused by a folded conformation that suppresses the enzymatic activity of the motor domain in the absence of cargo proteins (Hammond et al. 2010). Multiple alignment analysis revealed that the deduced motor domain sequence of Lc-KIF17 displayed a high identity withthat of KIF17 orthologs from other teleosts. Phylogenetic analysis showed that the deduced amino acid sequence of KIF17 was highly similar to that of other kinesin-2 family members from other teleosts. These results strongly support that Lc-KIF17 involves highly conserved sequences, suggesting that its function may also be conserved.
We then detected the distribution of Lc-KIF17 at mRNA and protein levels in different tissues of L. crocea and at different stages of testicular development. We found that the Lc-kif17 mRNA expression pattern was similar to that of Lc-KIF17, with the brain showing the highest expression level, followed by the testes. These expression patterns were consistent with those reported in pervious studies (Setou 2000). According to previous studies, the high KIF17 expression level in the brain is likely related to the dendritic transport of neuronal proteins (Chu et al. 2006;Kayadjanian et al. 2007). In the present study, we focused on the high expression level of Lc-KIF17 in the testes, as such an observation may be related to spermiogenesis (Ma et al. 2017). Moreover, the significantly higher expression level of Lc-KIF17 at stage IV than at the other stages of testicular development may help delineate the potential functions of Lc-KIF17 in L. crocea spermiogenesis. Zhang et al. (2016) discovered that stage IV is the most active stage of L. crocea spermiogenesis, during which a large number of spermatids are produced.
We also studied the spatiotemporal expression of Lc-kif17 mRNA during spermiogenesis via FISH, and the results showed that Lc-kif17 mRNA signals were continuously expressed throughout different stages of spermiogenesis in which round spermatids go through substantial structural and functional changes to differentiate into spermatozoa. Hence, we speculate that based on the high expression levels of Lc-KIF17 at mRNA and protein levels in the testes, especially at stage IV of testicular development, Lc-KIF17 may actively participate in spermiogenesis. Microtubules are important parts of the cytoskeleton that act as "railways" for cargo protein transport during spermiogenesis (Goodson and Jonasson 2018). Microtubules consist of α-tubulin and β-tubulin heterodimers arranged in a head-to-tail manner, with the α-tubulin exposed at the minus (slow-growing) end and the β-tubulin exposed at the plus (fast-growing) end. During spermatid remodeling, the manchette, a typical microtubular-based cargo delivery platform for intramanchette transport (IMT), transiently assembles around and intimately associates with the spermatid nucleus. IMT is believed to be indispensable to nuclear reshaping (Kierszenbaum 2002;Mochida et al. 1998;Soley 1997) and protein delivery to the base of the sperm tail (Kierszenbaum and Tres 2004;Rivkin et al. 1997;Tres and Kierszenbaum 1996). Manchette irregularities in spermatid remodeling induced by specific mutations in or chemical treatment of the β-tubulin gene results in nuclear abnormalities and multiple coiling of the sperm tail (Meistrich et al. 1990;Meistrich 1993).
The capabilities of microtubules, which are closely associated with the use of kinesins to recruit multisubunit cargo complexes along microtubule tracks during spermatid remodeling, have been studied in mammals. In particular, KIFC5A-like motor proteins have been associated with the manchette and may contribute to spermatid remodeling (Navolanic and Sperry 2000), KIF3A participates in nuclear elongation and tail formation during spermiogenesis by maintaining the structure of the manchette (Lehti et al. 2013), and KIF17 interacts with its cargo protein in the manchette (Saade et al. 2007). In the present study, IF results showed that Lc-KIF17 signals migrated from the perinuclear cytoplasm to the side of the nucleus where the sperm tail forms and eventually accumulated in the sperm midpiece during the development of spermatids into mature sperms. In addition, the dynamic localization of β-tubulin signals in L. crocea showed the overlapping of β-tubulin signals and Lc-KIF17 signals during spermiogenesis, suggesting that a perinuclear microtubular structure involving Lc-KIF17 may be involved in spermatid remodeling in L. crocea. Based on these results, the predicted functional models of Lc-KIF17 are shown in Fig. 8.

Conclusions
In this study, the cDNA of the Lc-kif17 gene was cloned from L. crocea. The Lc-KIF17 protein exhibited the typical features of kinesin-2 proteins. Both Lc-kif17 mRNA and Lc-KIF17 protein were highly expressed in L. crocea testis, especially at stage IV of testicular development, and were continuously detected during spermiogenesis, suggesting that KIF17 may play a crucial role and have potential functions in spermiogenesis. Lc-KIF17 signals were found to colocalize with β-tubulin signals and migrate from the perinuclear cytoplasm to the side of the nucleus where the tail forms during spermiogenesis, revealing the likely involvement of KIF17 in L. crocea spermiogenesis. In particular, KIF17 may participate in spermatid remodeling by interacting with perinuclear microtubules. This study contributes to improved understanding of the mechanism underlying L. crocea spermiogenesis and provides evidence for the association of Lc-KIF17 with perinuclear microtubules and the function of Lc-KIF17 in spermatid remodeling.