Paternally expressed gene 3 (Pw1/Peg3) promotes sexual dimorphism in metabolism and behavior

Abstract

The paternally expressed gene 3 (Pw1/Peg3) is a mammalian-specific parentally imprinted gene expressed in stem/progenitor cells of the brain and endocrine tissues. Here, we compared phenotypic characteristics in Pw1/Peg3 deficient male and female mice. Our findings indicate that Pw1/Peg3 is a key player for the determination of sexual dimorphism in metabolism and behavior. Mice carrying a paternally inherited Pw1/Peg3 mutant allele manifested postnatal deficits in GH/IGF dependent growth before weaning, sex steroid dependent masculinization during puberty, and insulin dependent fat accumulation in adulthood. As a result, Pw1/Peg3 deficient mice develop a sex-dependent global shift of body metabolism towards accelerated adiposity, diabetic-like insulin resistance, and fatty liver. Furthermore, Pw1/Peg3 deficient males displayed reduced social dominance and competitiveness concomitant with alterations in the vasopressinergic architecture in the brain. This study demonstrates that Pw1/Peg3 provides an epigenetic context that promotes male-specific characteristics through sex steroid pathways during postnatal development.

Fig 1. Reduced masculinization of metabolisms in Pw1^+/pat- males.

(A) Postnatal growth of Pw1^+/pat- compared with Pw1^+/+ littermates showing Pw1^+/pat- animals are smaller in both sexes throughout the growing phase (p<0.001), and the onset of sexual dimorphism in body weight is delayed in Pw1^+/pat- at four weeks of age. Two-way ANOVA test revealed significant interaction between genotype and sex after 6 weeks of age. (B) Random-fed glucose levels at 2 month-old and its positive correlation with body weight (r = 0.690, p<0.0001). Each symbol represents independent measurement. (C) Sex dimorphic food consumption at 2 month-old. Two way ANOVA test with multiple comparisons demonstrated significant interaction between sex and genotype. (D) Sexual dimorphisms in body composition in young adults (n = 8–14 each group). Male-specific increase in lean mass and decrease in fat mass in Pw1^+/+ males at 2–3 month of age was less prominent in Pw1^+/pat- males, while the females Pw1^+/pat- are proportionally smaller in lean and fat mass. * in black represents comparison in males and * in red represents comparison in females. (E) An inverse correlation between lean mass and fat mass at 10 month of age was only found in the Pw1^+/+ males (r = 0.757, p<0.001) and not in the Pw1^+/p- males (r = 0.076, p = 0.771) nor in the females (r = 0.006, p = 0.981). Comparison between genotypes and sexes was performed using two-way ANOVA with Tukey’s multiple comparisons. Correlation was determined with simple linear regression analysis. *P<0.05, **P<0.01, and ***P<0.001. NS: non-significant. Values are mean ± SEM. Each symbol represents individual animals in (B) and (E).

Fig 2. Deregulated GH/IGF axis in Pw1^+/pat- youngs and insulin homeostasis in Pw1^+/pat- adult males.

(A) Circulating IGF-1 levels at 3 weeks of age and its positive correlation with body weight. Correlation was determined with simple linear regression analysis. (B) Blunted sexual dimorphism of circulating IGF-1 dynamics in Pw1^+/pat- mice as compared to that of Pw1^+/+. N = 4–8 each group from 5 litters. Two-way ANOVA showed that the IGF-1 levels at 3 weeks old are significantly lower in Pw1^+/pat- mice, with no difference between sexes. **P<0.01 in black represent comparison between genotypes, whereas ***P<0.001 in red represent comparison between sex. (C) mRNA expression of growth hormone (Gh) in the pituitary gland, and GH receptor (Ghr) and insulin-like growth factor (Igf1) in the liver at 3 weeks of age. Values were normalized with Tbp and presented relative to the Pw1^+/+ male littermates (n = 6–12 each group). (D) Random-fed blood insulin levels at 3 months and 6 months of age in Pw1^+/pat- and Pw1^+/+ littermates (n = 4–6). (E) Representative insulin tolerance test (ITT) on Pw1^+/pat- and Pw1^+/+ males from a single litter (n = 3 each genotype) at 3 months and 6 months of age. Similar results were obtained from two other litters. (F) Representative oral glucose tolerance test (OGTT) with insulin secretion measurement on the same set of mice as in ITT. (G) Liver size and mRNA expression of lipogenic genes in the liver at 3 months and 6 months of age showing an age-dependent development of hepatic steatosis in Pw1^+/pat- mice as compared to Pw1^+/+ littermates. Srebp1, sterol regulatory element binding protein 1; Acaca, acetyl-CoA carboxylase alpha; Fasn, fatty acid synthese; Scd1, Stearoyl-CoA desaturase; Pparg1 & 2, peroxisome proliferator activated receptor gamma 1 & 2. (H) Fat deposition revealed by Oil Red-O staining in 8-month-old livers (n = 4–6 for each group). Lipid droplets were quantified in number and in size using particle analysis tool in Image-J software. Values are mean ± SEM; *, P < 0.05; **, P < 0.01; ***, P < 0.001 by two-way ANOVA with Tukey’s multiple comparisons.

Fig 3. Altered social behavior and brain architecture in Pw1^+/pat- males.

(A) Interlitter social rank by tube test in Pw1^+/+ and Pw1^+/pat- males from mixed genotypes at 10 months of age (n = 11 vs n = 10, from 5 litters) and at 3 months of age (n = 7 vs n = 10, from 4 litters). (B) Assessment of social dominance in the stranger encounter tube test. Animals used were listed in S3B Fig. The winning rate was calculated from 17–18 matches against unfamiliar opponents. (C) Pw1 reporter expression (β-gal⁺) is observed in the vasopressinergic neurons of PVN (top), whereas β-gal signals are strongly co-localizing with ERα receptor in SON (bottom) in the hypothalamus (x400). (D) Representative images of AVP expressing neurons in the PVN of hypothamamus in the Pw1^+/+ and Pw1^+/pat- male brains (x40). Coronal sections at 120μm intervals through PVN from anterior to posterior axis were immunostained with an anti-AVP antibody. AVP positive area size (dotted line) and cell count were quantified from five sequential sections. E. Digital quantification of AVP⁺ area size and the total cell count in PVN and their positive correlation in the Pw1^+/+ male brain. Columns, mean; bars, SEM; *, P < 0.05, Mann-Whitney U test.

Fig 4. Reduced testosterone production in the Pw1^+/pat- males.

(A) Plasma testosterone measurements at peripubertal age in Pw1^+/+ and Pw1^+/pat- males (n = 7 and 5, respectively, from 3 litters). Two-way repeated measures ANOVA with Sidak’s multiple comparisons test revealed a significant difference between genotypes at 9 weeks of age (*P<0.05). (B) Plasma testosterone levels in Pw1^+/+ and Pw1^+/pat- non-breeder littermates (n = 47 and n = 42) at 2.5–3 months of age. Each symbol represents independent animals. Bars represent mean ± SEM; **, P<0.01, by Mann Whitney U test. (C) mRNA expression of steroidogenic genes in the 2.5–3 month-old testes of Pw1^+/+ and Pw1^+/pat- non-breeder littermates (n = 15 versus n = 11). StAR, Steroidogenic acute regulatory protein; Cyp17a1, cytochrome P-450 17a; Cyp11a1, cholesterol side-chain cleavage enzyme; Hsd3b1, 3-b-hydroxysteroid dehydrogenase; Lhr, luteinizing hormone receptor; Cyp19a1, aromatase enzyme; Hsd17b3, 17-b-hydroxysteroid dehydrogenase. (D) Pw1 reporter expression in the 2.5-month-old testis of Pw1^IRESnLacZ mice, showing X-gal staining (x100) and β-gal immunofluorescence (x400) in the interstitial compartment and in the epithelium of seminiferous tubules. The β-gal immunofluorescence overlapped with that of androgen receptor (AR) in the Leydig and Sertoli cells (arrows). L, interstitial Leydig cells; S, seminiferous tubule.

Fig 5. Hormonal cascades are deregulated in Pw1^+/pat- mice.

Pw1 is abundantly expressed in the gonads and target organs and colocalizes with steroid hormone receptors in hormone secreting cells. GHRH, growth hormone-releasing hormone; GH, growth hormone; IGF-1, insulin-like growth factor 1; GnRH, gonadotropin-releasing hormone; LH, luteinizing hormone; FSH, follicle-stimulating hormone; AVP, arginine vasopressin; +, stimulation; -, inhibition by Pw1.