Selective effects of protein 4.1N deficiency on neuroendocrine and reproductive systems

Protein 4.1N, a member of the protein 4.1 family, is highly expressed in the brain. But its function remains to be fully defined. Using 4.1N−/− mice, we explored the function of 4.1N in vivo. We show that 4.1N−/− mice were born at a significantly reduced Mendelian ratio and exhibited high mortality between 3 to 5 weeks of age. Live 4.1N−/− mice were smaller than 4.1N+/+ mice. Notably, while there were no significant differences in organ/body weight ratio for most of the organs, the testis/body and ovary/body ratio were dramatically decreased in 4.1N−/− mice, demonstrating selective effects of 4.1N deficiency on the development of the reproductive systems. Histopathology of the reproductive organs showed atrophy of both testis and ovary. Specifically, in the testis there is a lack of spermatogenesis, lack of leydig cells and lack of mature sperm. Similarly, in the ovary there is a lack of follicular development and lack of corpora lutea formation, as well as lack of secretory changes in the endometrium. Examination of pituitary glands revealed that the secretory granules were significantly decreased in pituitary glands of 4.1N−/− compared to 4.1N+/+. Moreover, while GnRH was expressed in both neuronal cell body and axons in the hypothalamus of 4.1N+/+ mice, it was only expressed in the cell body but not the axons of 4.1N-/- mice. Our findings uncover a novel role for 4.1N in the axis of hypothalamus-pituitary gland-reproductive system.

Protein 4.1N 1 is a neuronally enriched member of the protein 4.1 family which also include protein 4.1R 2 , 4.1G 3 and 4.1B 4 homologs. Two highly conserved domains, the 4.1-ezrin-radixin-moesin (FERM) domain and the spectrin/actin-binding domain (SABD) are responsible for interactions of 4.1 family of proteins with the plasma membrane proteins and with the spectrin-actin cytoskeleton, respectively 5 . The third conserved domain in the 4.1 family is the carboxy-terminal domain (CTD). It has been shown that all members of the protein 4.1 family undergo extensive tissue specific alternate splicing generating multiple protein 4.1 isoforms. In the case of 4.1N, the 135 kDa 4.1N is the predominant isoform identified in the brain, whereas a smaller 100 kDa isoform is enriched in peripheral tissues, such as the enteric neurons of the intestinal tract and adrenal gland 6 .
In vitro cell biological studies have identified multiple cellular functions for protein 4.1N. Anti-proliferative actions of NGF in PC12 cells has been shown to be due to binding of 4.1N to the nuclear mitotic apparatus protein (NuMA) following its translocation to the nucleus and preventing the mitotic actions of NuMA 7 . Furthermore, the binding of 4.1N to PIKE prevents its interactions with nuclear PI3K, which can influence the regulation of PI3K by NGF 8 . Interaction of protein 4.1N with dopamine receptor is required for the localization or stabilization of dopamine receptors at the neuronal plasma membrane 9 . Among dopamine receptors, protein 4.1N interacts specifically with D2 and D3 dopamine receptors. Moreover, the protein 4.1N-binding region of inositol 1,4,5-trisphosphate receptor type 1 (IP 3 R1) is necessary and sufficient for the localization of IP 3 R1 at the basolateral membrane domain in polarized MDCK cells 10 . Both the CTT14aa and CTM1 sequences of IP 3 R1 can bind to 4.1N 11 , suggesting that the spectrin-actin binding domain of 4.1N can serve as a linker between IP 3 R1 and actin filaments. This actin filament-dependent regulation of IP 3 R1 diffusion may be important for the spatiotemporal regulation of intracellular Ca 2+ signaling which in turn mediates neurite formation [12][13][14] . Interestingly, 4.1N does not interact with the IP 3 R in an epithelial cell line WIF-B, implying cell specific interaction between protein 4.1N and the IP 3 R and tissue-specific mechanism in shaping the pattern of Ca 2+ waves in various cell types 15 . In primary hippocampal cultures, mouse 4.1N is enriched at the discrete sites of synaptic contact, co-localizing www.nature.com/scientificreports/ with the postsynaptic density protein of PSD95 and GluR1, suggesting a potential functional role for 4.1N as a component of the cytoskeletal architecture of excitatory synapses 16 . 4.1N also directly interacts with the GluR1 subunit of the AMPA receptor and co-localizes with AMPA receptors at excitatory synapses 17 .
In contrast to our understanding of the multiple functions of 4.1N in various cell types based on in vitro cell biological studies, function of 4.1N in vivo is yet to be defined. To address this issue, we generated 4.1N −/− mice and performed phenotypic characterization of these mice to begin to decipher the function of 4.1N in vivo. The 4.1N −/− mice were born at lower birth rate and showed slow growth with marked hypoplasia of reproductive system. These findings in conjunction with documentation of focal expression pattern of 4.1N in pituitary and hypothalamus suggest a specialized function for 4.1N protein in neuroendocrine system.

Results
Targeted disruption of the 4.1N gene. To study the roles of 4.1N in vivo, we generated 4.1N knockout mice using Embryonic stem (ES) cells generated by the International Gene Trap Consortium. A gene trap cassette with a reported insertion site between exon 1A and exon 2 was confirmed by genotype analysis using primer pairs (Fig. 1A) that could distinguish the wild type allele from the knockout allele (Fig. 1B). Western blot analysis showed that the homozygous gene trap allele reduced expression levels of 4.1N protein to undetectable levels in various tissues tested (Fig. 1C,D and Suppl. Fig S3), confirming the generation of a 4.1N null phenotype. Heterozygous mice (4.1N +/− ) expressing ~ 50% of the normal level of 4.1N were fertile and phenotypically indistinguishable from wild type mice (supplementary Fig. S4,S5). www.nature.com/scientificreports/ Expression of 4.1N in brain and other tissues by Western blot analysis. The protein 4.1N undergoes tissue specific alternate splicing generating multiple isoforms. We compared the expression of 4.1N in different parts of brain and other tissues by Western blot analysis. Western blot analysis of hypothalamus tissue with 4.1N-specific polyclonal antibody revealed a doublet, with the predominant band migrating at 135 kDa and a less prominent lower molecular weight band. A similar pattern was seen for olfactory and cerebrum, whereas the lower molecular band was more prominent than the 135 kDa band in cerebellum (Fig. 1D). A very faint band of 165 kDa was also detected following longer exposures and may represent a larger splice form of 4.1N of low abundance. A prominent 95 kDa band was detected in the kidney and pancreas, in association with a very faint lower MW band seen in the kidney, pancreas and testis (Fig. 1C). The 95 kDa isoform likely corresponds to an alternatively spliced isoform that excludes intervening sequence between the SABD and CTD 6 . Neither the 95 kDa or 135 kDa 4.1N isoforms were found in heart, liver, skeletal muscle and testis. mice revealed that they were passive and when gently lifted by their tail, their movements were powerless relative to those of their wild-type counterparts. When tested for their ability to walk on a rod or on top of a container, some of the 4.1N −/− mice did not move at all while others were able to proceed for a short distance and tired quickly which was in marked contrast to wild type mice that were able to accomplish these tasks readily. The observed neurobehavioral defects in 4.1N −/− mice suggest a potential role for 4.1N in brain function.

4.1N
Slow growth and decreased weight of reproductive organs in 4.1N −/− mice. As the mice aged, we noted a marked age-dependent decrease in size and body weight of 4.1N −/− mice compared with wild-type littermates (Fig. 2). Body weight of 4.1N −/− mice was reduced by 3 to 4 fold compared to their wild-type littermates for both males ( Fig. 2A) and females (Fig. 2B) at 3-5 weeks of age (P < 0.001), and reduced by 2 to 3 fold at 6-9 weeks of age (P < 0.001, Fig. 2). We then monitored the weight of different organs in the age-matched wildtype and knockout mice (6 to 9 weeks old mice). The size and weight of the male and female reproductive organ were noticeably decreased in 4.1N −/− mice compared to that of 4.1N +/+ (Fig. 2). A detailed tissue weight assessment showed that the ratio of weight of the testis and ovary to the body weight for 4.1N −/− male and female mice was dramatically decreased compared with the 4.1N +/+ control mice (Table1). The 4.1N −/− adult testis weight was 0.13 ± 0.018% of body weight, in contrast to 0.41 ± 0.035% of body weight (p < 0.0001) in 4.1N +/+ mice. For the female, the 4.1N −/− adult ovary weight was 0.0073 ± 0.00085% of body weight, while it was 0.014 ± 0.0013% (p = 0.0005) in 4.1N +/+ mice. We also noted that the 4.1N −/− mice displayed a mild decrease in liver weight compared with the control wild-type littermates (p < 0.05). No significant difference for the ratio of organ to body weight was noted for other organs including heart, lung, spleen and kidney in 4.1N −/− mice.
Malformation of the reproductive system in 4.1N −/− mice. Next, histological examination showed hypoplasia of testis in 4.1N −/− male mice with significant changes in mature sperm and in the number of sperm. The seminiferous tubules were reduced in diameter. Spermatogenic cells and spermatocytes were present, but spermatids were not observed. The number of mature spermatozoa was dramatically decreased (Fig. 3C), with few if any detected in seminiferous tubules. No sperm were present in most of the lumens in tubules of the 4.1N −/− testis. The interstitium was sparse, only a few or no leydig cells could be found in 4.1N −/− testis (Fig. 3A). Moreover, no sperm were found in 4.1N −/− epididymis (Fig. 3B). These findings suggest that spermatogenesis was arrested in the spermatocyte stage in the 4.1N −/− mice. Hypoplasia was also noted in the ovary and uterus of female 4.1N −/− mice. The follicles of the ovary were largely undeveloped, and their growth arrested in the pre-antral phase of follicle development in 4.1N −/− mice between ages of 7-23 weeks. No corpora lutea was found, and the interstitium was severely atrophic in 4.1N −/− ovary (Fig. 4A). These findings suggest abnormalities in the maintenance and growth of the follicles. A high degree of hypoplasia was also observed in the endometrium. No estrous cycle changes such as mitotic activity, hyperemic and secretory changes were found in the endometrium (Fig. 4B). In contrast (Fig. 4C), in 4.1N +/+ mice of the same age, corpus luteum or corpus albicans of the ovary was evident, and follicles including graafian follicles at different developmental stages were observed. These findings imply that both male and

Localization of 4.1N protein in mouse pituitary and hypothalamus.
Based on the phenotype of 4.1N −/− mice, we also explored the detailed expression pattern of 4.1N in mouse pituitary and hypothalamus by immunohistochemistry and immunogold electron microscopy which has not been previously reported. Immunohistochemistry using the mouse 4.1N antibody revealed several patterns of staining in the hypothalamus area. There was intense staining of cell bodies, dendrites and axons except nucleus and the ependymal cells of the third ventricle. Both diffuse and punctate patterns of staining were found under high magnification ( Supplementary  Fig. S2).
In the pituitary, 4.1N is expressed predominantly in the neurohypophysis and pars intermedia (Fig. 9A). In the neurohypophysis, a densely speckled pattern was observed (Fig. 9B). In the pars intermedia, discrete foci along the axons was noted (Fig. 9C). In pars distalis (adenohypophysis), which contains the secretary cells, punctate labeling that outlined cell bodies was noted in some of the cells (Fig. 9D). Immunogold electron microscopy demonstrated the presence of 4.1N immunoreactivity in the secretory granules in pituitary (Fig. 9E). No gold particles were found in the control in which the primary antibody was replaced with nonimmune serum (Fig. 9F). These findings together with the phenotype of 4.1N −/− mice suggest that 4.1N plays an important role in hormone secretion and transmission along the axonal route.

Discussion
The findings from the present study have enabled us to document that deficiency of protein 4.1N is associated with abnormal development of the neuroendocrine system leading to phenotypic changes including hypoplasia in the reproductive system and also slow growth. Moreover, the detailed characterization of 4.1N expression in pituitary and hypothalamus enabled the clue of a role for 4.1N in regulating neuroendocrine function.
Although protein 4.1N is enriched in the nervous system 16 , the distribution in neuroendocrine tissue is not clear. We found that high levels of 4.1N were localized to the neurohypophysis and pars intermedia in pituitary Table 1. Changes of organs weight in 4.1N-/-mice. Data shown are mean ± SD. *P < 0.05; ***P < 0.001; as determined by Student's t-test.  www.nature.com/scientificreports/ and in axons of hypothalamus with speckled or discrete foci pattern. Immunogold electron microscopy confirmed that 4.1N protein mainly exists in secretory granules in cytoplasm. The results suggest that 4.1N participates in hormone secretion or transport in these areas. In 4.1N −/− mice, a remarkable extent of hypoplasia was observed in the reproductive system. The weight of the male and female reproductive organ was noticeably decreased in 4.1N −/− mice, whereas no significant difference for the ratio of organ to body weight was observed for other organs except for a mild decrease in liver. Importantly, histological changes showed underdevelopment of sperm, graafian follicle, uterus and defective of leydig, corpus luteum or corpus albicans. The pathological findings for the testis and ovary as well as growth state led us to explore the upstream changes in 4.1N −/− mice. It is well established that hormone synthesis, secretion and transport are primarily regulated by the hypothalamic-pituitary-gonadal (HPG) axis. Within the HPG axis, reproductive maturation and function is coordinated by the release of GnRH from a neuronal population in hypothalamus which also consists of other clusters such as GHRH secreting neurons [19][20][21] . The release of these hypothalamic hormones influences the secretion of anterior pituitary hormones such as FSH, LH and GH that subsequently regulate the function of target organs including testis and ovary as well as in growth [22][23][24][25][26] . However, molecular and mechanistic regulation of hormone transport is yet to be fully defined. The defective sexual development and infertility in the 4.1N −/− mice is likely the result of defects in hormone transport and secretion from the hypothalamohypophysial system.
Our documentation of decreased GnRH and GHRH expression in pituitary pars intermedia in 4.1N −/− mice suggests that the absence of these hormone signals in the pituitary might be attributed to impaired secretion or migration of the releasing hormone granules from the hypothalamic to the pituitary via the hypophysealhypothalamic portal system (HHPS). Interestingly, immunohistochemical study confirmed that GnRH are, indeed, absent in the axons of hypothalamus neurons in 4.1N −/− mice. The GnRH immunoreactive signal was present both in neuronal cell body and axons with punctate pattern in the median eminence of the hypothalamus in 4.1N +/+ mice. In the 4.1N −/− mice, the expression of GnRH was not detected in the axons, although some reactivity was observed in the cell body of GnRH neurons, indicating that GnRH fails to migrate through axons www.nature.com/scientificreports/ in 4.1N −/− mice. Furthermore, the decrease of GnRH in pituitary can partly explain why 4.1N −/− mice have lower FSH and LH levels and developmental defects in the reproductive system. These findings suggest that 4.1N is required, directly or indirectly, for the hormone transport and secretion in neuroendocrine system. It is interesting to note that the extent of changes in FSH and LH levels in 4.1N −/− male and female mice is different. In contrast to the significant decrease in FSH level in 4.1N −/− male mice, the decrease in FSH levels in 4.1N −/− female mice was not statistically significant. This may in part be a reflection of the fact that serum FSH levels are generally lower in female rodents than in males 27 , and there is considerable individual variability in the serum FSH levels masking the difference between the mutant and wild-type mice. The individual variability and stability could account for the finding that the serum GH level showed no statistically significant difference between groups. However, the body weight of 4.1N −/− mice was reduced even in older mice and the expression of GHRH and GH as well as the number of secretory granules were decreased in 4.1N −/− pituitary, suggesting that the slow growth phenotype of 4.1N −/− mice is likely to be of complex etiolgoy.
HPG axis, whose abnormal hormone secretion or transport is associated with clinical feature involving delayed or absent puberty and reproductive dysfunction. Disorder in the development of the GnRH neuroendocrine system leads to absent GnRH secretion or migration, resulting in heterogeneous reproductive disorders such as congenital hypogonadotropic hypogonadism in humans with infertility or decreased fertility 28 . The molecular identities of the complexes responsible for transporting axonal hormone are not well understood. A subclass of semaphorins such as Sema4B and Sema4C characterized by a PDZ-binding motif at their carboxy-terminus that mediates their interaction with the post-synaptic density protein PSD-95/ SAP90, SFAP75/Norbin and SEMCAP-1/GIPC have been implicated in clustering of semaphorins, axon guidance and neurite outgrowth [29][30][31] .
In line with this pattern, 4.1N co-localizes with the postsynaptic density protein of PSD95 and GluR1 in primary hippocampal cultures and interacts with the GluR1 subunit of the AMPA receptor at excitatory synapses, 1N maybe linked to the axonal cytoskeleton which is required to ensure the migration of releasing hormone. It is likely that protein 4.1N plays this role through direct or indirect interactions with other cytoskeletal proteins such as αII-and βII-spectrin which have been shown to contribute to the organization of neurofilaments 35,36 . However, precise interactions in specific axons need to be confirmed in future studies. In summary, 4.1N −/− mice exhibit unique phenotypes that include low birth rate, slow growth and hypoplasia of reproductive organs due to defects in the neuroendocrine system. Protein 4.1N could be an important component of axons, not only at the hypothalamus, but also in the pituitary region. The presence of 4.1N may be necessary for the transmission of releasing hormone at hypothalamic-pituitary-axon route. The 4.1N knockout mice are likely be a valuable tool for exploring 4.1N function in certain neuroendocrine disorders.

Methods
Generation of 4.1N knockout mice. 4.1N knockout mice were generated using the similar method for generation of 4.1G knockout mice as described previously 37 . Briefly, embryonic stem (ES) cells, containing a gene trap cassette inserted downstream of exon 1A and upstream of exon 2 in the 4.1N gene, were obtained from the International Gene Trap Consortium (ES cell line FHCRC-GT-S12-6H1). ES cells were microinjected into blastocysts and implanted into recipient female mice to permit development of the embryos into chimeras at the University of California, San Francisco, transgenic facility, and subsequent breeding was performed to select founders with germ line transmission of the gene trap allele. The mice were initially bred and screened at the Lawrence Berkeley National Laboratory (LBNL) animal facility and then maintained at the animal facility of New York Blood Center under specific-pathogen-free conditions. All animal experiments and protocols were reviewed and approved by Institutional Animal Care and Use Committees from both the New York Blood Center and Lawrence Berkeley National Laboratory, and conducted in accordance with the New York Blood Center and Lawrence Berkeley National Laboratory guidelines and regulations. Mice were backcrossed more than 9 generations into a C57BL/6 background, and, in all experiments, comparisons were made between littermate wild-type (4.1N +/+ ) and homozygous mice (4.1N −/− ). The genotyping was routinely performed by PCR as described below. In total, we examined forty-four 4.1N −/− (21 males and 23 females) and the same number of wild-type and heterozygous mice. The mice were mainly divided into two groups: 3-5 weeks old and 6-9 weeks old. A few mice older than 9 weeks were also analyzed. Autopsy and histopathological examination. Mice were anesthetized with pentobarbital and perfused through the heart with PBS (pH7.4) containing 4% paraformaldehyde (PFA). After sacrificing the mice, various tissues or organs were rapidly dissected and fixed in 10% phosphate-buffered formalin solution. The testis were fixed in Bouin's solution. The tissues were subsequently embedded in paraffin, and serial sections (4-μm thick) were cut and mounted on slides. The sections were stained with hematoxylin-eosin (HE). Immunohistochemistry was performed using various antibodies. The stained sections were examined microscopically. The number of testis and ovary analyzed were from 8 male and 6 female, separately. The other tissues were from 6 male and 6 female. Histopathology was assessed independently by two pathologists who were blinded to the different groups. Microscopic semiquantitative scoring was calculated on a scale of 0 to 4 as follows: 0 (no sperm, leydig cells, follicles or corpus luteum), 1 (minimal, < 10%), 2 (mild, 10 to 30%), 3 (moderate, 31 to 65%), and 4 (severe, > 65%).  Western blot. Western blot was performed as described previously 6 . Briefly, tissues from adult 4.1N +/+ and 4.1N −/− were homogenized by sonication in ice-cold lysis buffer (0.32 M sucrose, 0.01 M HEPES, pH 7.4, 2 mM EDTA, 1 mM DTT, protease inhibitor cocktail) in 1.5 ml tubes. After centrifugation at 900×g for 5 min, the supernatant was collected and the protein concentration measured by Bradford method using bovine serum albumin as standard. 30 μg of proteins were run on 8% Tris-glycine gels and subsequently transferred onto nitrocellulose membrane (BIO-RAD, Hercules, CA). Membranes were blocked for 1 h at room temperature in blocking buffer (PBS + 0.1% Tween-20 + 4% nonfat dry milk + 1% BSA) followed by overnight incubation at 4 °C with primary 4.1N antibody (1: 2000 dilution) and rabbit polyclonal anti-GAPDH antibody (1: 5000 dilution) in blocking buffer. After washes with 0.1% PBST, membranes were incubated for 1 h at room temperature with rabbit HRP-conjugated secondary antibody (Jackson ImmunoResearch Laboratories, Inc, West Grove, PA) diluted 1:5000 in blocking buffer. Following extensive washing, membranes were processed using chemiluminescent reagent (Thermo Scientific, Rockford, IL).

Generation
Immunohistochemistry. Immunohistochemical analysis was performed as described previously 6 . Briefly, 4.0 μm thick sections were de-paraffinized in xylene and dehydrated using graded ethanol solutions. Endogenous peroxidase activity was quenched with peroxidase blocking reagent for 6 min. The tissue sections were pretreated with target retrieval solutions of high pH (for 4.1N immunostaining) or with low pH (for GH-growth hormone, GnRH-gonadotropin releasing hormone and GHRH-growth releasing hormone immunostaining) and heated in microwave for 15 min. Following three washes with PBS, the sections were incubated with specific antibodies, at a dilution of 1:100 for anti-4.1NHP, 1:50 for anti-GH(R&D systems, BAF1067), 1:50 for anti-GnRH (Thermo, PA1-121) 39   Electron microscopy and immunogold labelling. Pituitary Holland). The area ratio of secretory granules in cytoplasm were calculated with QWin image analysis software (Germany). For immunogold labeling, tissue specimens, ~ 1 mm 3 in size, were collected and fixed in the mixture of 4% paraformaldehyde and 0.1% glutaraldehyde in 0.1 mol/L cacodylate buffer, pH 7.4, for 1 h at 4 °C. After washing with 0.1 M cacodylate buffer and treatment with 50 mM ammonium chloride to quench residual aldehydes, the fixed tissue was dehydrated and embedded in LR-White Resin (Electron Microscopy Sciences, Hatfield, PA). Ultrathin sections of embedded tissue were cut with a diamond knife and mounted on to parlodion coated nickel grids. Non-specific binding was blocked by applying nonimmune serum to the sections, followed by incubation overnight at 4 °C with 4.1N primary antibody diluted 1:10 in 0.05 mol/L Tris-buffered saline (TBS), supplemented with 0.1 mol/L bovine serum albumin (BSA). After thorough rinsing with TBS containing 0.1% BSA and 0.5% Tween 20, the sections were incubated with 15-nm gold-conjugated goat-anti-rabbit IgG (Amersham Biosciences) diluted 1:20 in TBS for 2 h at room temperature. Following thorough rinsing with TBS, the sections were treated with 4% aqueous uranyl acetate to enhance contrast and imaged using the transmission electron microscope. For control, either the primary antibody was omitted or nonimmune serum was used.
Hormone assays. Mice serum were collected and sent to Laboratory of Comparative Pathology, Memorial Sloan Kettering Cancer Center and Weill Medical College at Cornell University for hormone assay. FSH and LH levels were measured by the radio immunoassay and GH levels by ELISA enzyme immunoassay. www.nature.com/scientificreports/ Statistical analysis. All numerical data were presented as mean ± SD. Statistical analysis were performed by Student's t-test and one-way ANOVA, values of P < 0.05 were considered to be significant.

Data availability
The data generated and/or analysed during the current study are available from the corresponding authors on reasonable request.
Received: 5 January 2020; Accepted: 14 September 2020 Figure 9. The distribution and subcellular localization of 4.1N in mouse pituitary by immunohistochemistry and immunogold electron microscopy. The expression of 4.1N was found in the pituitary (A, 10 × . Scale bar, 200 µm), especially in the neurohypophysis (B, 100 ×. Scale bar, 20 µm) and pars intermedia (C, 100 ×. Scale bar, 20 µm), which revealed densely speckled and discrete foci pattern. In the pars distalis, punctate labeling that outlined cell bodies in part of cells was observed (D, 100 ×. Scale bar, 20 µm). Immunogold electron microscopy shows immunogold labeling for 4.1N in the secretory granules of axons of neurohypophysis and pars intermedia and some cells in adenohypophysis (arrow, and see inset), Scale bar, 500 nm. (E). No gold particles were found in a control in which the primary antibody was replaced with nonimmune serum (F).