Insulin and insulin-like growth factor 1 receptors are required for normal expression of imprinted genes

Abstract

In addition to signaling through the classical tyrosine kinase pathway, recent studies indicate that insulin receptors (IRs) and insulin-like growth factor 1 (IGF1) receptors (IGF1Rs) can emit signals in the unoccupied state through some yet-to-be-defined noncanonical pathways. Here we show that cells lacking both IRs and IGF1Rs exhibit a major decrease in expression of multiple imprinted genes and microRNAs, which is partially mimicked by inactivation of IR alone in mouse embryonic fibroblasts or in vivo in brown fat in mice. This down-regulation is accompanied by changes in DNA methylation of differentially methylated regions related to these loci. Different from a loss of imprinting pattern, loss of IR and IGF1R causes down-regulated expression of both maternally and paternally expressed imprinted genes and microRNAs, including neighboring reciprocally imprinted genes. Thus, the unoccupied IR and IGF1R generate previously unidentified signals that control expression of imprinted genes and miRNAs through transcriptional mechanisms that are distinct from classical imprinting control.

Keywords: developmental control; diabetes; transcriptional control.

Fig. 1.

miRNA expression profiling in WT and IR/IGF1R DKO cells. Large-scale miRNA expression profiling was performed by qPCR in confluent WT and DKO cells under normal (in the presence of serum: +FBS) or apoptotic (in the absence of serum for 6 h: −FBS) conditions. (A) Heat map representing the miRNA expression profile in WT and DKO cells in normal or apoptotic conditions. Red represents high expression and green low expression. (B) List of miRNAs significantly down-regulated by more than twofold in DKO cells compared with WT cells in the presence of serum. miRNAs in blue share the same chromosomal location in a cluster in region 12qF1; miRNAs in green share the same chromosomal location in a cluster in region 2qA1.

Fig. 2.

Expression of miRNAs and mRNAs in imprinted Dlk1–Dio3 locus on chromosome 12 in WT and DKO cells. (A) Graphical representation of mouse imprinted Dlk1–Dio3 locus on chromosome 12qF1. Genes are shown as colored: white, biallelic expression; hatched, maternal expression; dark gray, paternal expression. An IG-DMR located between Dlk1 and Gtl2 genes is represented by circles (filled, hypermethylated; open, hypomethylated). The genomic region is not drawn to scale. (B) Expression of miRNAs and mRNAs was measured by real-time PCR in confluent WT and DKO cells in the presence (+FBS) or absence (−FBS) of serum for 6 h. Results represent the average ± SEM of six independent experiments.

Fig. 3.

Expression of other imprinted genes in WT and DKO cells. (A) Graphical representation of mouse imprinted Sfmbt2 gene in chromosomal region 2qA1. Dark gray boxes represent exons of the paternally expressed Sfmbt2 gene; dark gray triangles represent the miRNA cluster in an intron of Sfmbt2. (B and C) Expressions of Sfmbt2 and miRNAs from the cluster on chromosome 2qA1 (B) and of other imprinted genes (C) were measured by real-time PCR in confluent WT and DKO cells in the presence (+FBS) or absence (−FBS) of serum for 6 h. Results represent the average ± SEM of six independent experiments.

Fig. 4.

Imprinted gene expression in IR-KO MEFs and FIRKO brown fat. (A) miRNA expression profiling was performed by qPCR in control, heterozygous KO (IR-KO MEF^+/−), and homozygous KO (IR-KO MEF^−/−). Expression of imprinted miRNAs from cluster in intron of Smbt2 gene in chromosomal region 2qA1 is shown. Results are expressed as fold change compared with expression in control cells normalized to 1. Logarithmic scale. (B) Expression of imprinted genes Dio3, IGF2R, and Grb10 in confluent control and IR-KO MEF^−/− was measured by real-time PCR. Results represent the average ± SEM of three independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001 (compared with control cells). (C) Expression of imprinted mRNAs was measured by real-time PCR in brown adipose tissue of 6-mo-old male control and FIRKO mice. Results represent the average ± SEM of five mice. *P < 0.05 (compared with control mice).

Fig. 5.

DNA methylation in WT and DKO cells. (A) Schematic representation of the Dlk1–Dio3 locus and methylation analysis of the IG-DMR by pyrosequencing after bisulfite treatment of DNA from confluent WT and DKO cells. (B) Methylation analysis of the Gtl2 promoter: Southern blotting was performed on DNA from confluent WT and DKO cells digested with methylation-sensitive restriction enzymes and probed for the Gtl2 promoter. Controls using DNA from maternal and paternal uniparental disomy for chromosome 12 (UPD12) were added. Pyrosequencing was performed on DNA from WT and DKO cells after bisulfite treatment. (C–E) Schematic representation of imprinted loci and methylation analysis by pyrosequencing after bisulfite treatment of DNA from confluent WT and DKO cells. (F) Proliferating WT and DKO cells were treated with 0.2 μM 5-Azacytidine (5Aza) for 6 d, with or without 1 μM trichostatin A (TSA) for the last 18 h. Dlk1 and H19 expression were measured by real-time PCR in confluent cells. Results represent the average ± SEM of three independent experiments. *P < 0.05 (compared with control cells). Schematic representations were obtained from ref. .