Abstract
Seipin, encoded by BSCL2 gene, is a protein whose physiological functions remain unclear. Mutations of BSCL2 cause the most-severe form of congenital generalized lipodystrophy (CGL). BSCL2 mRNA is highly expressed in the brain and testis in addition to the adipose tissue in human, suggesting physiological roles of seipin in non-adipose tissues. Since we found BSCL2 mRNA expression pattern among organs in rat is similar to human while it is not highly expressed in mouse brain, we generated a Bscl2/seipin knockout (SKO) rat using the method with ENU (N-ethyl-N-nitrosourea) mutagenesis. SKO rats showed total lack of white adipose tissues including mechanical fat such as bone marrow and retro-orbital fats, while physiologically functional brown adipose tissue was preserved. Besides the lipodystrophic phenotypes, SKO rats showed impairment of spatial working memory with brain weight reduction and infertility with azoospermia. We confirmed reduction of brain volume and number of sperm in human patients with BSCL2 mutation. This is the first report demonstrating that seipin is necessary for normal brain development and spermatogenesis in addition to white adipose tissue development.
Introduction
Seipin is a protein encoded by BSCL2 gene whose mutation causes the most severe variety of congenital generalized lipodystrophy (CGL), also known as Berardinelli-Seip congenital lipodystrophy (BSCL) (1). BSCL is a disease characterized by a near total lack of adipose tissue from birth (2). Patients with BSCL frequently develop severe insulin resistance, hypertriglyceridemia and fatty liver (3). BSCL due to AGPAT2 (BSCL1), BSCL2 (BSCL2), CAV1 (BSCL3) and PTRF (BSCL4) mutations have been reported so far (1,4–6). BSCL1 and BSCL 2 are the most common varieties and have been reported in patients of various ethnicities (3). However, most of the patients of African origin have AGPAT2 mutation and those from Lebanon have BSCL2 mutation. BSCL2 mutation is also the major cause of BSCL in Japan (7).
1-acylglycerol-3-phosphate O-acyltransferase (AGPAT) is a critical enzymes involved in the biosynthesis of triglyceride and phospholipids from glycerol-3-phosphate. Of known AGPAT isoforms, AGPAT2 is highly expressed in the adipose tissue and its deficiency causes lipodystrophy (8). Caveolin 1 encoded by CAV1 is an integral component of caveolae, which are specialized microdomains seen in abundance on adipocyte membranes (9). Caveolin 1 binds fatty acids and translocates them to lipid droplets. Polymerase 1 and transcript release factor (PTRF) is involved in biogenesis of caveolae and regulates expression of caveolins 1and 3 (6). On the other hand, molecular functions of seipin, a protein encoded by BSCL2 remain unclear although seipin seems to play a role in lipid droplet formation and be involved in adipocyte differentiation. While BSCL2 mutations cause the most severe cases of BSCL, yet seipin remains the most mysterious lipodystrophic protein in terms of function.
Seipin is a 398 (short-form) or 462 (long-form) amino acid protein that has no similarity with other known proteins or consensus motif. Seipin has two distinct hydrophobic amino acid stretches and is speculated to have two transmembrane domains (1). It also has been shown that seipin localizes to endoplasmic reticulum in various cell lines (10–12). BSCL2 mRNA is highly expressed in the brain and the testis other than adipose tissue in human (1). BSCL2 patients exhibit much higher rate of mild mental retardation than do other BSCL patients (13). These evidences strongly suggest physiological roles of seipin in non-adipose tissues.
In the past few years, three independent models of Bscl2 knockout mice have been reported (14–16). All these three knockout mice exhibited severe generalized lipodystrophy, demonstrating clearly that seipin deficiency itself leads to generalized lipodystrophy in vivo. However, in contrast to human BSCL2 patients, all the three knockout mice had low plasma triglyceride levels. In addition, Bscl2 knockout mice showed a decrease of energy expenditure that is generally increased in human BSCL2 patients (16,17). There are also some differences between mouse and human on physiological roles of seipin in non-adipose tissues. Although depression that is not a major symptom in human BSCL2 patients was reported in male Bscl2 knockout mice (18), no phenotypes as to mental retardation that is frequently observed in human BSCL2 patients have been reported in Bscl2 knockout mice. While BSCL2 mRNA is highly expressed in the brain and testis in addition to the adipose tissue in human, BSCL2 mRNA is not highly expressed in mouse brain (11). Furthermore, teratozoospermia was reported in Bscl2 knockout mice (19) but was not observed in our male BSCL2 patients. Instead of that, oligospermia was observed in our BSCL2 patients. Thus, a new animal model of human BSCL2 is required for further understanding of physiological roles of seipin.
In this study, we chose rat as a species for the generation of a new animal model of human BSCL2 after confirming the similarity of BSCL2 mRNA expression pattern between rat and human. We generated a Bscl2/seipin knockout (SKO) rat using with the N-ethyl-N-nitrosourea (ENU) mutagenesis method (20). SKO rat has a homozygous nonsense mutation (L20X) in BSCL2 gene, which is upstream of the first transmembrane domain. SKO rats showed impairment of spatial working memory with reduction of whole brain weight and infertility with azoospermia in addition to phenotypes of lipodystrophy including hypertriglyceridemia and increase of energy expenditure. Therefore, we also analyzed brain volume and semen in human BSCL2 patients. This is the first report demonstrating that seipin is necessary for normal brain development and spermatogenesis in addition to white adipose tissue development.
Results
BSCL2 mRNA expression profiles in mouse and rat
It was reported that BSCL2 mRNA is highly expressed in the brain and the testis in human (1). To choose an animal species that is appropriate for generating the experimental model of human BSCL2, we examined Bscl2 mRNA expressions in various tissues in mouse and rat and compared these expression profiles with that in human. In mouse, high expression of Bscl2 mRNA was observed in the testis but not in the brain (Fig. 1A). On the other hand, Bscl2 mRNA was highly expressed in both the brain and the testis in rat like in human (Fig. 1B). Thus, we decided to generate a seipin knockout animal on rat background as a human BSCL2 model.
Development of BSCL2 animal model. (A and B) Bscl2 mRNA expressions in various tissues in 20-week-old male mice (A) and rats (B) were checked with quantitative RT–PCR. Bscl2 mRNA expression levels were normalized by 18S. Values are means ± SEM (n = 10). (C) Normal sequence and heterozygous T to A mutation at nucleotide 239 (red arrows) of Bscl2 gene in WT and G1 male offspring heterozygous mutant rats, respectively. (D) Schematic diagram of seipin that consists of two transmembrane domains. The red arrow indicates the amino acid position of the L20X mutation. (E) Growth curve of body weight in male SKO rats (filled square) and their WT littermates (open square). Values are means ± SEM (n = 10 per group). *P < 0.05, **P < 0.01 (ANOVA). (F) Weights of various tissues in 20-week-old male SKO rats (closed bars) and their WT littermates (open bars). The fold change is displayed as relative to WT rats. Values are mean ± SEM (n = 10 per group). *P < 0.05, **P < 0.01, NS, not significant (Student's t-test).
Generation of seipin knockout rat
By using ENU mutagenesis followed by MuT-POWER screening of the KURMA samples (20), we generated a seipin knockout rat with a homozygous nonsense mutation (Bscl2sko/sko) in Bscl2, the seipin gene. Bscl2sko mutation was T to A transition at nucleotide 239 in the third exon of Bscl2 gene, resulted in a substitution of leucine at codon 20 by the stop codon (L20X), which is upstream of the first transmembrane domain (Fig. 1C and D). Male and female heterozygous SKO rats were intercrossed to obtain homozygous SKO animals. There were 19 homozygous WT, 50 heterozygous SKO and 26 homozygous SKO rats. This ratio did not differ significantly from the expected 1:2:1 Mendelian ratio of genotypes (number of delivery = 10, mean number of pups per delivery = 9.5; c2 = 1.29, P = 0.69). The sex ratios also did not differ significantly from the expected ratio (male; n = 48, female; n = 45; c2 = 0.58, P = 0.45).
The body weight in SKO rats was significantly lower than that in their WT littermates from at least 3 weeks after birth (Fig. 1E). There was no difference of body weight between heterozygous SKO rats and their WT littermates (data not shown). Although there was no significant difference in the amount of food intake and respiratory exchange ratio, the oxygen consumption was significantly higher in SKO rats compared with their WT littermates (Supplementary Material, Fig. S1A–C). When weights of various tissues were compared between SKO and WT rats, the weight in most tissues was significantly increased in SKO rats (Fig. 1F), which might be the organomegaly caused by hyperinsulinemia. In contrast, in addition to the WAT and BAT, weights in the brain and testis, where high expression of Bscl2 mRNA was observed, were significantly reduced in SKO rats.
SKO rat develops generalized lipodystrophy
Dissection and computer tomography of SKO rat revealed the lack of WAT throughout the body (Fig 2A and B). Body composition analysis demonstrated that fat mass in both subcutaneous and intra-abdominal areas was markedly reduced while lean mass was obviously increased in SKO rats (Supplementary Material, Fig. S2A). Enlargement of skeletal muscle was confirmed by magnetic resonance imaging (MRI) (Supplementary Material, Fig. S2B). In fact, the size of epididymal WAT that is one of intra-abdominal adipose tissues and subcutaneous oil red O staining positive area were extremely reduced in SKO rats although small lipid droplets were detected in both regions (Supplementary Material, Fig. S2C and D). Consistent with this, plasma leptin concentration was markedly decreased in SKO rats compared with WT rats (Fig. 2C). BSCL2 patients lack not only the ‘metabolically active’ adipose tissue such as subcutaneous and intra-abdominal ones but also the ‘mechanical’ adipose tissue located in the bone marrow and retro-orbital areas (7,21). No adipose tissue was detected in the bone marrow and retro-orbital areas in SKO rats (Supplementary Material, Fig. S2E and F).
SKO rat develops generalized lipodystrophy. (A) Gross appearance of dorsal view before and after removal of skin in a 20-week-old male SKO rat (right) and its WT littermate (left). (B) Computer tomography image at a slice 15 cm distal from nose in a SKO rat (bottom) and its WT littermate (top). (C) Plasma leptin concentrations in SKO rats (closed bars) and their WT littermates (open bars). (D) Plasma glucose (left panel) and plasma insulin concentrations (right panel) during IPGTT in SKO rats (filled squares or closed bars) and their WT littermates (open square or open bars). (E) Plasma triglyceride concentrations in SKO rats (closed bars) and their WT littermates (open bars). (C–E) Values are means ± SEM (n = 10 per group). *P < 0.05, **P < 0.01, NS, not significant (Student's t-test). (F and G) Macroscopic (F) and histological images (G) of the liver in a SKO rat (right) and its WT littermate (left). For histological examination, hematoxylin and eosin staining was used. Original magnification of ×200 is shown.
In contrast to WAT, interscapular brown adipose tissue (BAT) remained certain (Supplementary Material, Fig. S3A), although its weight was decreased in SKO rats (Fig. 1F). To examine whether the BAT in SKO rats is functional, we conducted cold exposure experiment. Twenty-four hours exposure of 4°C did not change the body temperature in both SKO and WT rats (Supplementary Material, Fig. S3B). At this time, interscapular BAT weight was similarly decreased in SKO and WT rats (Supplementary Material, Fig. S3C). Histological analysis revealed the reduction of lipid droplet number after 24 h cold exposure especially in SKO rats (Supplementary Material, Fig. S3D). In addition, increment of Ucp1 mRNA expression by cold exposure was not just observed in both SKO and WT rats but was greater in SKO rats than in WT rats (Supplementary Material, Fig. S3E). These results indicate that the BAT in SKO rats was, in terms of thermogenesis, physiologically functional.
IPGTT showed impaired glucose tolerance with hyperinsulinemia, indicating insulin resistance, in SKO rats (Fig. 2D). Under ad lib feeding, plasma triglyceride concentration was markedly elevated in SKO rats while plasma non-esterified fatty acid (NEFA) concentration was unchanged (Fig. 2E and Supplementary Material, Fig. S4A). Plasma total cholesterol (T-Chol) was also elevated in SKO rats (Supplementary Material, Fig. S4B). We studied the effect of fasting on SKO rats. During a 24 h fasting, both SKO and WT rats lost body weight (Supplementary Material, Fig. S5A). Glucose concentration dropped slightly in WT rats, but plummeted in SKO rats to WT levels (Supplementary Material, Fig. S5B). Insulin concentration also dropped in both SKO and WT rats although its level was still higher in SKO rats than in WT rats (Supplementary Material, Fig. S5C). Under these conditions, NEFA concentration appropriately increased in WT rats as a normal respond to fasting, meanwhile it did not increase but dropped in SKO rats, indicating that SKO rats had no sufficient lipid stores to respond to fasting (Supplementary Material, Fig. S5D). Since circulating NEFA is metabolized to ketone bodies, such as β-hydroxybutyrate, by the liver, we checked β-hydroxybutyrate concentrations. Consistent with the results of NEFA, β-hydroxybutyrate concentration vastly increased in WT rats but did not in SKO rats (Supplementary Material, Fig. S5E).
The liver in SKO rats was remarkably enlarged and was lighter in color, suggesting severe fatty liver (Fig. 2F). Histological examination showed large number of lipid droplets of various sizes in SKO rats (Fig. 2G). Consistent with these results, liver weight and liver TG content were also remarkably increased in SKO rats (Supplementary Material, Fig. S4C and D).
These results demonstrate that SKO rats develop generalized lipodystrophy and its related phenotypes, which are strikingly similar to those of human BSCL2.
SKO rat develops impairment of spatial working memory
We evaluated the spatial working memory by Y-maze test for the assessment of mental development in SKO rats. Y-maze score was significantly lower in SKO rats than that in WT rats (Fig. 3A) while spontaneous movement was unchanged. To determine whether this impairment of spatial working memory is seipin knockout specific, we performed the same experiment in leptin-deficient Lepmkyo/Lepmkyo rats and A-ZIP/F-1 mice, a mouse model of generalized lipodystrophy due to adipocyte specific expression of dominant negative protein for b-zip protein (22). Both Lepmkyo/Lepmkyo rats and A-ZIP/F-1 mice showed no significant change in Y-maze score when compared with their WT littermates, respectively, indicating that the impairment of spatial working memory in SKO rats is due to neither leptin deficiency nor lipodystrophy (Fig. 3B and C).
The role of seipin in the brain development. (A–C) Y-maze score as an index of spatial working memory in SKO rats (A), Lepmkyo/Lepmkyo rats (B) and A-ZIP/F-1 mice (C). Values are mean ± SEM (n = 10 per group). *P < 0.05, NS, not significant (Student's t-test). (D and E) Bscl2 expression pattern in the whole brain (D) and sections (E) of hippocampus, cerebellum, frontal cortex and occipital cortex assessed by in situ hybridization in a WT rat. Images used with sense probe (top) and antisense probe (bottom) are shown. Original magnification of ×16 is shown for the whole brain. ×200 is for hippocampus and cerebellum and ×100 is for frontal and occipital cortexes. (F) Microscopic images of hippocampus, cerebellum, frontal cortex and occipital cortex in WT (top) and SKO rats (bottom). Nissl staining was used. Original magnification of ×40 is shown. (G–I) The number of Nissl stained cells in four regions (CA1, CA2, CA3, DG) of hippocampus (G), five layers (I, II+III, IX, V, VI) of frontal cortex (H) and three layers (molecular cell layer, granular cell layer and Purkinje cell layer) of cerebellum (I) in SKO rats (closed bars) and their WT littermates (open bars). Values are means ± SEM (n = 9 per group). NS, not significant (Student's t-test).
The role of seipin in the brain development
Gene expression pattern of Bscl2 in the whole brain was analyzed by in situ hybridization in WT rats. Bscl2 gene was ubiquitously expressed throughout the brain including hippocampus, cerebellum and, frontal and occipital cortex in WT rats (Fig. 3D and E). Next, we histologically examined of the brain in SKO rats. In comparison with WT rats, any morphological changes were not detected throughout the brain in SKO rats (Fig. 3F). Since the whole brain weight was slightly but significantly decreased in SKO rats (Fig. 1F), we examined the neuron density in hippocampus, cerebellum and frontal cortex. In any layers or regions of these sections, no significant change of neuron density was detected (Fig. 3G–I). The fact that the brain volume was decreased while neuron density was unchanged means that the total neuron number in the brain was decreased.
Intellectual quotient test and brain volume in human BSCL2 patient
Intellectual quotient test was examined in two male and four female BSCL2 patients. The age of these six patients was ranging from 18 to 36 years (Supplementary Material, Table. S1). Four patients had R275X, one patient had E189X and the remaining one had Y187C homozygous mutation in BSCL2 gene, respectively. Five of six patients showed mild but obvious reduction in both verbal and performance intelligence quotient scores. Brains of the same six patients were examined by MRI. No apparent morphological change was reported with these MRI examinations in any patients. For the analysis of brain volume, three male and six female healthy subjects whose age was ranging from 21 to 34 years (mean, 25.4 years) with normal body mass index between 18.5 and 25.0 kg/m2 were also examined by MRI. With the small number of subjects, we found no significant difference but the tendency of reduction in whole brain volume in the patients when compared with healthy subjects (Fig. 4A). Next, we checked the brain parenchymal area of coronal sections at three different levels. In two of the three sections, significant reduction of brain parenchymal area was observed in the patients (Fig. 4B). Finally, we checked the brain parenchymal /subarachnoid area ratio to assess the presence of brain atrophy. No significant difference of this ratio was observed at any levels between the patients and healthy subjects, indicating that the reduction of brain volume in BSCL2 patients was not due to brain atrophy (Fig. 4C).
MRI analysis of brain volume in human BSCL2 patients. (A) Whole brain volume in BSCL2 patients (closed bar) and healthy subjects (open bar). The fold change was displayed as relative to healthy subjects. (B) Brain parenchyma area of coronal sections at three different levels in BSCL2 patients (closed bar) and healthy subjects (open bar). (C) Brain parenchyma/subarachnoid area ratio at three different levels in BSCL2 patients (closed bar) and healthy subjects (open bar). Three different levels: (a) the level just above the bilateral ventricles, (b) the level in the center of the lateral ventricles, (c) the level through interventricular foramen. Values are mean ± SEM (Healthy subjects; n = 9, BSCL2 patients; n = 6). *P < 0.05, **P < 0.01, NS, not significant (Student's t-test).
Male SKO rat shows infertility with azoospermia
Although serum levels of hormones including lutenizing hormone (LH), follicle stimulation hormone (FSH) and testosterone, in male SKO rats were all within normal range and showed no significant difference from those in male WT rats (Supplementary Material, Fig. S7), male SKO rats showed remarkably small testis in weight when compared with WT rats (Fig. 5A and D) and were infertile while female SKO rats were fertile. There was no significant change in testis weight in both Lepmkyo/Lepmkyo rats and A-ZIP/F-1 mice when compared with that in their WT littermates, respectively (Fig. 5B–E). Testis histology showed markedly shrunk spermatic duct and lack of mature sperm cells in SKO rats while there was no change in both Lepmkyo/Lepmkyo rats and A-ZIP/F-1 mice when compared with their WT littermates (Fig. 5D and E). These results indicate that azoospermia in SKO rats is due to neither leptin deficiency nor lipodystrophy.
The role of seipin in the spermatogenesis. (A–C) Testis weight in SKO rats (A), Lepmkyo/Lepmkyo rats (B) and A-ZIP/F-1 mice (C). Values are mean ± SEM (n = 10 per group). **P < 0.01, NS, not significant (Student's t-test). (D and E) Macroscopic (top) and histological images (bottom) of the testis in WT (left), SKO (middle) and Lepmkyo/Lepmkyo (right) rats (D), and WT (left) and A-ZIP/F-1 (right) mice (E). For histological examination, hematoxylin and eosin staining was used. Original magnification of ×40 is shown. (F) Bscl2 mRNA expressions checked with quantitative RT–PCR in the testis in 12-week-old WT, Sl/Sld and W/Wv mice. The fold change is displayed as relative to WT mice. n = 6 per group. (G) Bscl2 mRNA expressions checked with quantitative RT–PCR in the testis during the period of postnatal development in WT rats. Values are mean ± SEM (n = 6 per group). (H) Bscl2 expression pattern in the testis assessed by in situ hybridization in a WT rat. Images used with sense probe (top) and antisense probe (bottom) are shown. Original magnification of ×200 is shown.
The role of seipin in the spermatogenesis
Gene expression of Bscl2 in testis was analyzed with two infertile mouse models. One was Sl/Sld mutant mouse that has Sertoli cell dysfunction and the other was W/Wv mutant mouse that has lack of functional C-Kit (23,24). Bscl2 gene expression was detected in WT mice but not in Sl/Sld and W/Wv mice, indicating that Bscl2 gene is expressed in neither Sertoli nor Spermatogonia cells (Fig. 5F). Gene expression of Bscl2 during the period of postnatal development in WT rat testis was examined and the significant increment was detected from 5 to 6 weeks (Fig. 5G). Most rats reach sexual maturity at 6 or 7 weeks old. Finally, gene expression pattern of Bscl2 in testis was analyzed by in situ hybridization in WT rats (Fig. 5H). Bscl2 gene was expressed in spermatocytes and mature sperm cells but not in spermatogonia cells. These results indicate that seipin might have an important role especially in the late phase of spermatogenesis.
Semen examination and gonadal hormone concentrations in human BSCL2 patients
Semen and gonadal hormone concentrations were examined in two male patients with BSCL2. On semen examination, one patient showed obvious oligospermia and the other patient showed normal but minimum of the normal range value in all parameters (Table 1). On the other hand, gonadal hormone concentrations including LH, FSH, testosterone and prolactin were within normal range in both patients (Supplementary Material, Table. S2). Although further study is needed, these results suggest the potential involvement of seipin in spermatogenesis in human.
Semen examination in human BSCL2 patients
| Variables | Patient 1 | Patient 2 | Cut-off value in WHO manual (2010) |
|---|---|---|---|
| Age | 21 | 25 | |
| Volume (ml) | 3.6 | 1.5 | 1.5 |
| Concentration (106/ml) | 20 | 10 | 15 |
| Total sperm number (106/ejaculate) | 72 | 15 | 39 |
| Motility (% motile) | 75 | 60 | 40 |
| Variables | Patient 1 | Patient 2 | Cut-off value in WHO manual (2010) |
|---|---|---|---|
| Age | 21 | 25 | |
| Volume (ml) | 3.6 | 1.5 | 1.5 |
| Concentration (106/ml) | 20 | 10 | 15 |
| Total sperm number (106/ejaculate) | 72 | 15 | 39 |
| Motility (% motile) | 75 | 60 | 40 |
Semen examination in human BSCL2 patients
| Variables | Patient 1 | Patient 2 | Cut-off value in WHO manual (2010) |
|---|---|---|---|
| Age | 21 | 25 | |
| Volume (ml) | 3.6 | 1.5 | 1.5 |
| Concentration (106/ml) | 20 | 10 | 15 |
| Total sperm number (106/ejaculate) | 72 | 15 | 39 |
| Motility (% motile) | 75 | 60 | 40 |
| Variables | Patient 1 | Patient 2 | Cut-off value in WHO manual (2010) |
|---|---|---|---|
| Age | 21 | 25 | |
| Volume (ml) | 3.6 | 1.5 | 1.5 |
| Concentration (106/ml) | 20 | 10 | 15 |
| Total sperm number (106/ejaculate) | 72 | 15 | 39 |
| Motility (% motile) | 75 | 60 | 40 |
Discussion
Using gene-driven ENU mutagenesis, we generated seipin deficient SKO rat. The mutation of Bscl2 gene in SKO rats (Bscl2sko) is located at nucleotide 239 in the third exon of Bscl2, generating a stop codon at amino acid 20 (L20X). Seipin has two distinct hydrophobic amino acid stretches and is speculated to have two transmembrane domains (1). L20X mutation is located upstream of these two transmembrane domains. In patients with lipodystrophy, nearly 30 BSCL2 mutations have been reported so far (25). Among these BSCL2 mutations, R275X mutation is the closest to the C-terminal (7). No relationship between the location of mutation site and severity of lipodystrophy is found, indicating that the region from the position at amino acid 275 to the C-terminal is important for the physiological function of seipin. SKO rat that has a nonsense mutation located upstream of the two transmembrane domains is considered to have no functional seipin.
The mean mutation frequency with ENU mutagenesis of our protocol was one mutation per 3.7 million base pairs (20). Although the chance for the occurrence of an unexpected mutation with a phenotypic effect is relatively small, this possibility also should be taken account for the experimental design and interpretation of the results. To eliminate mutations that might have been generated by ENU in chromosomal regions other than the Bscl2 locus, we performed backcross more than six generations against F344/NSlc inbred background, and we always compared phenotypes between littermates to minimize the effect of possible unexpected mutation.
The body weight in SKO rats was significantly lower than that in their WT littermates (Fig. 2A). Body composition analysis with computer tomography revealed that fat mass in both subcutaneous and intra-abdominal areas was markedly reduced while lean mass was obviously increased in SKO rats (Supplementary Material, Fig. S2A). In human patients with BSCL, organomegaly such as hepatosplenomegaly, cardiomegaly and muscular hypertrophy is generally observed as a consequence of hyperinsulinemia (17). We therefore compared weights in various tissues between SKO and WT rats and found that weights of the brain and the testis were significantly decreased in SKO rats in addition to the white and brown adipose tissues although weights of many other tissues were increased in SKO rats (Fig. 1F). Taking together with the fact that Bscl2 mRNA is highly expressed in the brain and the testis (Fig. 1B), it was strongly suggested that seipin has an important role in the development of these two organs.
Patients with BSCL2 mutation have the most severe variety of BSCL (13,25,26). They lack not only the ‘metabolically active’ adipose tissue such as subcutaneous and intra-abdominal ones but also the ‘mechanical’ adipose tissue located in the retro-orbital, bone marrow and so on (7,21). SKO rats lost more than 95% of fat mass in both subcutaneous and intra-abdominal areas (Supplementary Material, Fig. S2A), while more than 20% of fat mass in WT control mice was preserved in seipin knockout mice (14,15). We also confirmed no detectable adipose tissue in the bone marrow and retro-orbital areas in SKO rats by histological examination (Supplementary Material, Fig. S2E and F). On the other hand, even in BSCL2 patients, examination of subcutaneous biopsies found small adipocytes with low but detectable lipid content (27). In SKO rats, small adipocytes with small oil red O stained lipid droplets were also observed (Supplementary Material, Fig. S2D). SKO rat is a perfect model of human BSCL2 in terms of lipodystrophy and might be a useful model for the analysis of physiological roles of seipin in adipogenesis. We have already started the investigation with primary cultures of fibroblasts from SKO rats.
Lipodystrophy-associated phenotypes of SKO rats are also strikingly similar to those of human BSCL2 patients. Patients with severe lipodystrophy like BSCL2 generally exhibit insulin resistant but non-ketotic diabetes, hypertriglyceridemia, fatty liver, organomegaly and increased metabolic rate (17,26). While SKO rats had hyperglycemia with hyperinsulinemia (Fig. 3D), they did not have increased b-hydroxybutyrate level at baseline and fasting failed to increase it (Supplementary Material, Fig. S5E). Unlike Bscl2 knockout mice (14–16), triglyceride level was markedly elevated in SKO rats (Fig. 3E). While it was reported that the expression of LDL-receptor in the liver was increased in Bscl2 knockout mice (16), LDL-receptor mRNA expression was unchanged in SKO rats (data not shown). This might be the reason for the difference of triglyceride metabolism between mice and rats. Most organs except for adipose tissues, brain and testis were enlarged in SKO rats (Fig. 2B). Moreover, energy expenditure was increased in SKO rats as the oxygen consumption indicated (Supplementary Material, Fig. S1C). In contrast, Bscl2 knockout mice showed a decrease in energy expenditure (16). The generation of SKO rat provided us a powerful tool for studying the pathophysiology of human BSCL2.
BSCL2 patients exhibit much higher rates of mild mental retardation than do patients with other BSCL genotypes (13). Although these cognitive defects were speculated to relate to the high expression of Bscl2 mRNA in the brain, it was also possible that the lack of adipose tissue contributes to the mental retardation. Furthermore, it had been reported that leptin secreted from adipocytes plays a role in the regulation of cognitive function (28,29). BSCL2 patients present with the most severe lipodystrophy and their leptin levels are extremely low. In this study, we evaluated the spatial working memory by Y-maze test not only in SKO rats but also leptin deficient Lepmkyo/Lepmkyo rats and A-ZIP/F-1 mice, a mouse model of generalized lipodystrophy and clearly revealed that the impairment of spatial working memory observed in SKO rats is due to neither leptin deficiency nor lipodystrophy (Fig. 3A–C). The spatial working memory of the rodents is responsible for recording information about the environment and the spatial orientation (30). In behavioral science, Y-maze is used to investigate how rodents function with spatial working memory from the early 20th century (31). This is the first report demonstrating that seipin deficiency itself leads to cognitive defect in vivo.
To identify the original region responsible for the impairment of spatial working memory in SKO rats, we examined the distribution of Bscl2 mRNA expression in the whole brain by in situ hybridization in WT rats. Bscl2 mRNA was ubiquitously expressed throughout the rostral–caudal extent of the rat brain including hippocampus, cerebellum, frontal and occipital cortex (Fig. 3D and E). Then, we compared the brain histologically between WT and SKO rats but we could not find any morphological changes throughout the brain (Fig. 3F). Next, we examined the neuron density in hippocampus, cerebellum and frontal cortex, there was no significant difference between WT and SKO rats in any layers or regions of these areas (Fig. 3G–I). These results and the fact that the whole brain volume was significantly reduced in SKO rats indicate that the total neuron number of the brain was decreased in SKO rats. No evidence of brain atrophy suggests that the cause of the reduction of neuron number is a suppression of neuron increment rather than a stimulation of neuron decrease such as apoptosis. Furthermore, no evidence of morphological abnormality suggests that the suppression of neuron increment might be due to disorder of neuron proliferation rather than disorder of neuron differentiation. The causal relationship between brain volume reduction and impairment of cognitive function in SKO rats is unclear since the present study did not deny the possibility that seipin has a role in the neuronal function. In this study, we also found the reduction of brain volume in human BSCL2 patients although we need further study with more large number of patients to confirm it (Fig. 4A–C). Ventricular dilatation in CGL patients whose etiology was unknown was reported by pneumoencephalography (17). However, the ratio of subarachnoid area including cerebral ventricles to parenchymal area assessed by MRI in our patients showed no significant difference with that in healthy controls (Fig. 4C), demonstrating that our patients had no ventricular dilatation. In addition, while malformation of the hypothalamus in a CGL patient was also reported (32), all MRI images were assessed by radiologists and no morphological abnormality was pointed out in the present study. This is the first report indicating the pathophysiological role of seipin in the brain development in both rats and humans. Any apparent phenotypes of motor neuron disease that is associated with gain-of-toxic-function mutation in the N-glycosylation site of seipin (33) were not observed in both SKO rats and our BSCL2 patients.
SKO rats showed infertility with azoospermia. Testis weight was remarkably reduced and testis histology showed markedly shrunk spermatic duct and lack of mature sperm cells in SKO rats (Fig. 5D). These testis phenotypes were not observed in Lepmkyo/Lepmkyo rats and A-ZIP/F-1 mice, clearly indicating that the azoospermia observed in SKO rats is due to neither leptin deficiency nor lipodystrophy. Serum concentrations of gonadal hormones including LH, FSH and testosterone in SKO rats were all within normal range and were not significantly different from those in WT rats (Supplementary Material, Fig. S7). Bscl2 mRNA is highly expressed in the testis. These facts indicate that the responsive region for the azoospermia in SKO rats is the testis itself. Analyses using with Sl/Sld and W/Wv mice demonstrated that Bscl2 mRNA is expressed in neither Sertoli nor Spermatogonia cells. In addition, the analysis during the period of postnatal development in WT rats demonstrated that the significant increment of Bscl2 mRNA expression was detected only from 5 to 6 weeks. WT rat produces mature sperm cells from the age of 6 or 7 weeks old. Consistent with these results, analysis of WT rat testis by in situ hybridization showed that Bscl2 mRNA was not expressed in spermatogonia cells. These results indicate that seipin has an important role in the late phase of spermatogenesis in rats.
At least one male BSCL2 patient has been reported with multiple healthy children (17,25). However, semen examination in this patient was not reported. Recently, one other male BSCL2 patient was reported to have teratozoospermia (19). There was also no information on the number of sperm in this patient. In this study, we performed semen examination in two BSCL2 male patients and found that one patient had oligospermia according to the WHO criteria and the other had low sperm concentration but above the cut-off value of the criteria. Although further study is required, these results suggest the potential involvement of seipin in spermatogenesis in human.
In conclusion, through the generation and analysis of SKO rat, a rat model of BSCL2, we found that seipin deficiency leads to impairment of cognitive function with brain weight reduction and infertility with azoospermia in addition to generalized lipodystrophy, those have not been reported in seipin KO mice. We also confirmed reduction of brain volume and number of sperm in human patients with BSCL2 mutation although further study is needed to clarify human phenotypes. This is the first report demonstrating that seipin is necessary for normal brain development and spermatogenesis in addition to white adipose tissue development.
Materials and Methods
Animals
Rats with a Bscl2 mutation were obtained by ENU mutagenesis of F344/NSlc rats, followed by MuT-POWER (Mu Transposition POoling method With sequencER) screening on the genomic DNA of 4608 G1 male offspring in KURMA (Kyoto University Rat Mutant Archive). ENU mutagenesis procedures, screening protocols (20) and intracytoplasmic sperm injection procedure were previously described (34) The forward primer and the reverse primer used for identifying mutation of Bscl2 were 5′-GATGTTGCTTTGTCCTGCTA-3′ and 5′-TTTCTCGGTTTTTCACCAC-3′, respectively. More than six backcross generations were performed against the F344/NSlc inbred background. Genotyping for Bscl2skomutation was performed by real-time PCR system using TaqMan Sample-to-SNP kit (Applied Biosystems, Carlsbad, CA, USA) with a specific primer pair (Forward primer sequences are 5′-TGTGGGCCCAGGAAGTG-3′ and reverse primer sequences are 5′-CCCCAAACTGCAGCATCAG-3′) and TaqMan MGB probes (WT probe sequences are 5′-CCTGCCTACACATGG-3′ and mutant probe sequences are 5′-CCTGCCAACACATGG-3′). Genomic DNA was extracted from whole blood. The cycling conditions were 20 s at 95°C followed by 40 cycles of 3 s at 95°C and 20 s at 60°C. F344/NSlc rats and C57Bl/6J mice were purchased (Japan SLC, Hamamatsu, Japan). Lepmkyo/ Lepmkyo rats on F344/NSlc background were generated previously (35). A-ZIP/F-1 mice were provided from Diabetes Branch, National Institute of Diabetes and Digestive and Kidney Diseases (Bethesda, MD, USA) (22). Sl/Sld mice and W/Wv mice were purchased from Japan SLC. Rats and mice were maintained on a 14 h light/10 h dark cycle (lights on 7:00 AM, lights off 9:00 PM) and fed ad libitum standard pellet diet (MF; Oriental Yeast, Tokyo, Japan).
All animal care and experiments conformed to the Guidelines for Animal Experiments at Kyoto University and were approved by the Animal Research Committee of Kyoto University.
Real-time quantitative RT–PCR for Bscl2 mRNA expression
Each tissue was frozen in liquid nitrogen and stored at −80°C until use for RNA isolation. RNA was prepared using Trizol (Invitrogen, Carlsbad, CA, USA) reagent following the supplier's protocol. The Quality and the concentrations of the extracted RNA were checked using the Nano-Drop 2000 (Thermo Scientific, Yokohama, Japan). Single-stranded cDNA was synthesized from 1 µg of total RNA using SuperScript III First-Strand Synthesis System for RT–PCR, according to the manufacturer's instructions (Invitrogen). Quantitative RT–PCR was performed with SYBR Green (Applied Biosystems) by Applied Biosystems StepOnePlus™ RT–PCR System using gene specific primer. The housekeeping rat or mouse mitochondrial subunit 18S rRNA genes were used for control and quantitative RT–PCR was performed with TaqMan (Applied Biosystems). The sequences of primers (Sigma-Genosys, Tokyo, Japan) used in the present study are as follows: rat Bscl2 forward; 5′-CCCACAAGTGATTTGAGTTGGGGA-3′, rat Bscl2 reverse; 5′-GTGGCTGACGGTCGGCATGT-3′, mouse Bscl2 forward; 5′-GGTTCTCGCCGCTTACGTTGCG-3′, mouse Bscl2 reverse; 5′-GGTCTCCCAGCTCGCGTCACT-3′, rat 18s forward; 5′-GCAATTATTCCCCATGAACGA-3′, rat 18s reverse; 5′-CAAAGGGCAGGGACTTAATCAAC-3′, probe; 5′-AATTCCCAGTAAGTGCGGGTCATAAGCTTG-3′, mouse 18s forward; 5′-CGCGCAAATTACCCACTCCCGA-3′, mouse 18s reverse; 5′-CGGCTACCACATCCAAGGA-3′, probe; 5′-CCAATTACAGGGCCTCGAAA-3′.
Computed tomography
Computed tomography (CT) image of 20-week-old male rats was obtained under anesthesia by La Theta LCT-100 (Aloka, Tokyo, Japan).
Biochemical assays
Blood was obtained from the tail vein under ad libitum feeding at the age of 20 weeks if not otherwise specified. Plasma leptin concentrations were measured by an enzyme-linked immunosorbent assay (ELISA) kit for rat leptin (Millipore, St Charles, MO, USA). Plasma glucose concentrations were measured by a glucose assay kit (Wako Pure Chemical Industries, Osaka, Japan). Plasma insulin concentrations were measured by an insulin-ELISA kit (Morinaga Institute of Biological Science, Yokohama, Japan). Plasma triglyceride concentrations were measured by an enzymatic kit (Triglyceride E-test Wako; Wako Pure Chemical Industries).
Glucose tolerance test
Intraperitoneal glucose tolerance test (IPGTT) was performed after overnight fasting in 20-week-old male rats. Rats received 2.0 mg/g glucose by intraperitoneal injection. Blood was sampled from the tail vein before and 15, 30, 60, 90, 120 min after the glucose load.
Histology
Rat livers and rat and mouse testes were fixed in 10% neutrally buffered formalin and subsequently embedded in paraffin. Histological sections of 5 mm thickness were stained with hematoxylin and eosin. Rat brains were sampled after perfusion with 4% paraformaldehyde under anesthesia by intraperitoneal injection of sodium pentobarbital (DS Pharma Biomedical, Suita, Japan), fixed in 4% paraformaldehyde and subsequently embedded in paraffin. Sagital sections of 4 mm thickness were Nissl stained with 0.1% Cresyl violet solution.
Y-maze test
Spatial working memory was assessed by the Y-maze test in 20-week-old male SKO rats, Lepmkyo/ Lepmkyo rats and A-ZIP/F-1 mice as described previously (36). Briefly, the Y-maze test was conducted during the dark period (9:00 PM to 11:00 PM) in a dimly illuminated room. The maze consists of three arms (for rats; 425 mm long, 225 mm high and 145 mm wide, for mice; 300 mm long, 150 mm high and 60 mm wide, labeled A, B or C) diverging at a 120° from the central point. Each rat was placed at the end of the start arm and allowed to move freely through the maze during an 8-min session. The sequence of arm entries was manually recorded. An actual alternation was defined as entries into all the three arms on consecutive occasions. The maximum alternation was subsequently calculated by measuring the total number of arm entries minus 2 and the percentage of alternation was calculated as (actual alternations/maximum alternations) ×100. The total number of arms entered during the sessions, which reflect locomotor activities, was also recorded.
In situ hybridization of Bscl2 mRNA
In situ hybridization was performed as described previously (37). Briefly, to prepare cRNA probes, cDNA fragments encoding rat Bscl2 (NM_001012171.1 sequence166–1092) was amplified. Digoxigenin-labeled sense and antisense probes were synthesized with RNA polymerases. To obtain paraffin-embedded blocks and sections of rat brain and testis, rats were dissected after its perfusion. Rat brains and testes were fixed in 4% paraformaldehyde and embedded in paraffin. Sections of 6 mm thickness were deparaffinized, fixed, hydrogen chloride treated. Hybridization was performed with antisense or sense probe (100 ng/ml) at 60°C for 16 h in hybridization solution (Genostaff, Tokyo, Japan). Hybrids were detected with anti-DIG alkaline phosphate-conjugated antibody (Roche Diagnostics GmbH, Basal, Switzerland) and coloring reactions were performed with NBT/BCIP solution (Sigma-Aldrich Japan, Tokyo, Japan). The sections were counterstained with Kernechtrot stain solution (Muto Pure Chemicals, Tokyo, Japan).
Analysis of cell number in the brain sections
Four regions (CA1, CA2, CA3, DG) of hippocampus and five layers (I, II+III, IX, V, VI) of frontal cortex were photographed at high magnification (×400), respectively (Supplementary Material, Fig. S6A and C). The number of Nissl stained cells in a photographed visual field (0.275 × 0.4125 mm) was counted. The molecular cell layer and granular cell layer of cerebellum were also photographed at high magnification (×400), but the number of stained cells was counted in a smaller visual field (0.13 × 0.4125 mm) (Supplementary Material, Fig. S6B). Purkinje cell layer of cerebellum is a monolayer and the number of stained cells in this layer was just counted in a photograph at high magnification (×400) (Supplementary Material, Fig. S6B).
Measurement of brain size in BSCL2 patients
Brain size of BSCL2 patients was assessed by MRI technique. Whole brain images were acquired by a 3-Tesla Trio MRI scanner (Siemens, Erlangen, Germany) in axial orientation using the following parameters: repetition time, 3000 ms; echo time, 30 ms; flip angle, 90°; voxel size, 3 × 3 × 3 mm; field of view, 192 × 192 mm; matrix size, 64 × 64; and number of slices, 48 (38). Whole brain volume was calculated with the Virtual Place software (AZE, Tokyo, Japan). Brain parenchymal and subarachnoid areas of three different coronal sections, those are just above the bilateral ventricles, in center of the lateral ventricles and through interventricular foramen, were calculated by Image J software (NIH, Bethesda, MD, USA). Study protocols were approved by the Ethical Committee of Kyoto University Graduate School of Medicine. After detailed explanation of the study design, written informed consent was obtained from all subjects before study initiation.
Semen examination in BSCL2 patients
Semen examination was performed in two young adult male BSCL2 patients. One was 24 years old with E189X homozygous mutation and the other was 28 years old with R275X homozygous mutation in BSCL2 gene. Semen samples were collected by masturbation that followed 3 days of sexual abstinence. Ejaculate volume and concentration of spermatozoa and percentage of motile spermatozoa in semen were analyzed at room temperature immediately after complete liquefaction (2 h). Motility was determined by Cellsoft Automated Semen Analyzer (Cryo Resources, New York, NY, USA).
Statistical analysis
Data are expressed as means ± SEM. Comparison between or among groups was assessed by Student's t test or ANOVA with Fisher's protected least significant difference test. χ2 test was used for analysis of Mendelian ratios of genotype and sex. P < 0.05 was considered statistically significant.
Funding
This work was supported by research grants from the Japanese Ministry of Education, Culture, Sports, Science and Technology, the Japanese Ministry of Health, Labor and Welfare, Uehara Memorial Foundation, The European Community's Seventh Framework Programme (FP7/2007–2013) under grant agreement no. HEALTH-F4-2010-241504 (EURATRANS) and Industrial Technology Research Grant Program from New Energy and the Industrial Technology Development Organization (NEDO) of Japan.
Acknowledgements
We thank Takashi Shinohara for discussion and Keiko Hayashi for technical assistance. The authors also acknowledge Yoko Koyama for secretarial assistance.
Conflict of Interest statement. None declared.
References
