Imprinted gene dosage is critical for the transition to independent life

Abstract

Neonatal survival in mammals is crucially dependent upon maintenance of body temperature. Neonatal body temperature is largely maintained by thermogenesis in brown adipose tissue (BAT). BAT develops perinatally in mice requiring integration of adipogenic and thermoregulatory gene pathways. We describe a regulatory mutation in the imprinted gene cluster on mouse chromosome 12 resulting in early postnatal lethality. Maternal inheritance of this mutation impairs the ability of young mice to maintain body temperature. While mechanisms of perinatal BAT development are well understood, our work highlights a second phase of BAT recruitment necessary to support small animals newly independent of the nest. We show that the imprinted delta-like homolog 1/preadipocyte factor (Dlk1/Pref1) and iodothyronine deiodinase type 3 (Dio3) functions converge on the development of brown fat at the transition to independent life. This shows that appropriate dosage control at imprinted loci can act as a critical determinant in postnatal survival during phases of physiological adaptation.

Graphical abstract

Figure 1

Disrupted Chromosome 12 Imprinting in TG^MAT Animals (A) Top: The maternally inherited chromosome is unmethylated at control regions and expresses a series of noncoding RNAs: Gtl2, Rtl1AS, snoRNAs, and Meg9/Mirg containing multiple microRNAs that initiate from the shared Gtl2 promoter. Dlk1, Rtl1, and Dio3 are silenced on this chromosome. The paternally inherited chromosome is hypermethylated, the noncoding RNAs are silenced, and protein-encoding genes are expressed. Bottom, summary of gene expression following maternal transmission of the Gtl2LacZ insertion. Protein-encoding gene silencing was partially alleviated, whereas noncoding RNA expression was reduced. The paternally inherited chromosome was not genetically modified. (B and C) Gene expression in the e16 embryo (B) and placenta (C), obtained by northern blotting (Dlk1), qRT-PCR (Gtl2, and Meg9/Mirg), and ribonuclease protection assay (RPA) (Rtl1, Rtl1AS, and Dio3). Data was normalized to WT = 1, n ≥ 6 conceptuses from at least 3 litters, ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001 by Mann-Whitney U test performed on prenormalized data. (D) D3 activity at e16. TG^MAT gray bars, WT black bars. All error bars represent SEM.

Figure 2

Impaired Postnatal Survival in TG^MAT Neonates Is Biphasic and Associated with Failure to Thrive (A) Percentage of animals surviving to weaning in 11 litters generated by maternal transmission of the Gtl2LacZ transgene, ^∗∗p < 0.01, Mann-Whitney test. (B) Day of death was ascertained for 69 TG^MAT animals and is presented as frequency of animals dead during a 5-day interval. (C) Appearance of WT and mutant littermates at P6 and P21. (D) TG^MAT animals (all surviving for 12 weeks, males; n = 8) and 30 WT littermates (n = 30) were weighed weekly for 12 weeks. ^∗p < 0.05, p < 0.01 by one-way ANOVA with Bonferroni's multiple comparison post tests; error bars represent SEM. (E) Growth rates derived from the data in D, calculated using the equation: (weight at T₂ − weight at T₁)/ weight at T₁.

Figure 3

Juvenile TG^MAT Animals Are Hypothyroid (A) Expression of thyroid hormone responsive genes in the e16 embryo and placenta assessed by qRT-PCR as described above. (B) Serum total T3 at P1 (n = 6/genotype). (C) Measurement of serum thyroid hormones of TG^MAT animals with WT littermates at P11 (n = 6/genotype), P14 (n ≥ 4/genotype) and P21 (n ≥ 10/genotype). Because sexual dimorphism was not observed in either body weight or serum parameters at any stage, the data was combined. (D) cDNA was generated from individual pituitaries at P11 (n = 6/genotype), and at P21 (n ≥ 11/genotype), and gene expression quantified by qRT-PCR normalized to Hprt and represented as relative to WT = 1. (E) qRT-PCR analysis of preweaning liver gene expression. cDNA was generated from individual livers at P11 and at P21 (n = 9/stage/genotype) and data was analyzed as described in (B). Spot14 could not be detected at P11. All comparisons are by Mann-Whitney U test; ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001. Error bars represent SEM.

Figure 4

BAT Recruitment Dynamics Are Altered in TG^MAT Animals (A) BAT weight at P0 to P21 (see Table S1 for statistics). (B) The rate of change of BAT weight over 21 days, derived from the data in (A). (C) Proliferation of BAT at P6 to P21 measured using number of cells staining positive for Ki67, n = 4–5 individuals/stage/genotype. (D) Tissue TAG between P6 and P21, expressed as mg triolein units/g tissue, n = 6–16 samples/stage/genotype, ^∗p < 0.05 Mann-Whitney U test. (E) qRT-PCR was performed on BAT at P6 (WT n = 9; TG^MAT n = 11), P11 (WT and TG^MAT n = 6), P14 (WT and TG^MAT n = 6) and P21 (WT n = 8; TG^MAT n = 12). All developmental stages were performed on the same plate and normalized to a single standard curve, then normalized to WT P6 = 1. Prenormalization values were compared using a Mann-Whitney U test, ^∗p < 0.05. (F) Quantification of Dlk1 mRNA expression over the first three weeks of life by RNase protection assay (autoradiograph, left panel, and quantification graph, right panel, n = 4 animals/stage, below). (G) Anti-DLK1 immunohistochemistry in P11 BAT (top panels = 200× magnification; bottom panels = 400×). (H) qRT-PCR performed on BAT at P6–P21 as in (E) but including P18 (WT n = 7; TG^MAT n = 8). Data was normalized to WT = 1. Prenormalization values were compared using a Mann-Whitney U test; ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001. All error bars represent SEM.

Figure 5

Developmental Induction of β3 Adrenoreceptors Is Not Observed in TG^MAT Juveniles (A) Morphology of WT and TG^MAT BAT from P6–P21. H&E staining at 200 × magnification. Note large white circular fat deposits in the mutant cells at P14 and P21. (B) qRT-PCR performed on BAT at P6–P21 (n > 4/stage/genotype), all performed on the same plate and normalized to a single standard curve, then normalized to WT P6 = 1. Prenormalization values were compared by Mann-Whitney U test; ^∗p < 0.05, ^∗∗p < 0.01. All error bars represent SEM.

Figure 6

Thermogenic Properties of BAT in the Preweaning Period (A) qRT-PCR performed on BAT at P6–P21 as in Figure 5H. Data was normalized to WT = 1. Prenormalization values were compared by Mann-Whitney U test; ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001. (B) UCP1 expression in BAT at P6–P21. UCP1-positive cells are stained in brown and cells were counterstained with hematoxylin (blue). (C) Oxygen consumption measured in BAT explants in the presence of NE (1 μM) and succinate (10 mM), normalized to tissue dry weight. P11 WT n = 5, TG^MAT n = 6, P18 WT n = 6, TG^MAT n = 10; ^∗p < 0.05 Mann-Whitney U test. All error bars represent SEM.

Figure 7

Cold Challenge and Rescue at Thermoneutrality (A) Rectal temperature of individual animals at P18 following removal from the nest for 2 hr. WT n = 12, TG^MAT n = 8 (3 litters); ^∗p < 0.05 Mann-Whitney U test. (B) Top, schematic of experiments conducted to investigate the effect of ambient temperature on mortality. In intervention 1 (left graph), mothers were placed at 30°C 2 days prior to parturition and maintained at 30°C until P6 or at room temperature (22°C) during this period. Pups were counted daily and % survival was calculated. Of 37 animals born at 30°C, 22 survived; of 111 animals born at 22°C, 76 survived. In intervention 2 (right graph, top), pups were raised at room temperature until P11, then placed at 30°C or maintained at 22°C until P28. Of the 76 animals incubated at room temperature between P11 and P28, 34 survived, whereas of the 33 animals incubated at 30°C in this interval, 30 survived. During both interventions, the nest was kept intact and the mother was not removed. Bottom right, body weights of animals raised at room temperature or at 30°C between postnatal day 11 and 24. Error bars represent SEM.