Expression of the hepatic specific V1 messenger ribonucleic acid of the human growth hormone receptor gene is regulated by hepatic nuclear factor (HNF)-4alpha2 and HNF-4alpha8

Abstract

Human (h) GH plays an essential role in growth and metabolism, and its effectiveness is modulated by the availability of its specific receptor [hGH receptor (hGHR)] on target cells. The hGHR gene has a complex 5'-regulatory region containing multiple first exons. Seven are clustered within two small regions: V2,V3,V9 (module A) and V1,V4,V7,V8 (module B). Module A-derived mRNAs are ubiquitously expressed whereas those from module B are only found in postnatal liver, suggesting developmental- and liver-specific regulation of module B hGHR gene expression. To characterize the elements regulating module B activity, we studied a 1.8-kb promoter of the highest expressing exon in liver, V1. This promoter was repressed in transfection assays; however, either 5'- or 3'-deletions relieved this, suggesting the presence of multiple negative regulatory elements. Six putative hepatic nuclear factor 4 (HNF-4) response elements were identified. We determined that HNF-4alpha is developmentally regulated in the human liver: HNF-4alpha2 and HNF-4alpha8 are expressed in fetal hepatocytes but only HNF-4alpha2 is expressed in postnatal liver. Transient transfection assays demonstrated that HNF-4alpha2 and HNF-4alpha8 have a similar dual effect on V1 transcription: activation via site 1 in the proximal promoter and repression through site 6, approximately 1.7 kb upstream. EMSA/electrophoretic mobility supershift assays and chromatin immunoprecipitation analyses confirmed these two sites are bound by HNF-4alpha. Based on these data, we speculate there are multiple regions working together to repress the expression of V1 hGHR transcripts in tissues other than the normal postnatal liver, and that HNF-4alpha is a good candidate for regulating V1 hGHR expression in the human hepatocyte.

Fig. 1.

Schematic of the hGHR Gene The hGHR gene is located on the short arm of chromosome 5. Exons 2–10 code for the protein. Thirteen noncoding exons have been reported within the 150 kb upstream of exon 2 in the 5′-UTR (4 5 7 8 ). Seven of the noncoding exons are clustered in two small regions defined as module A (∼1.6 kb) and module B (∼2 kb). V_A, V_B, V_C, V_D, and V3a/b/_E are found between the two modules. V5 is located adjacent to the first coding exon, exon 2.

Fig. 2.

Transcriptional Activity of hGHR Modules A and B Promoter Constructs Module A (panel A) and module B (panel B) constructs were transfected into HEK293, CV1, HepG2, and Huh7 cells by the CaPO₄ method. Cells were harvested 48 h after transfection and assayed for luciferase and β-galactosidase (internal control) activity. The relative luciferase activity is presented as fold change over the promoterless vector. Data are presented as mean ± se; n = 3–10 experiments.

Fig. 3.

Transcriptional Activity of V1 hGHR Promoter Constructs 3′-Deletion (panel A) and 5′-deletion (panel B) V1 constructs were transfected into HEK293, CV1, HepG2, and Huh7 cells by the CaPO₄ method. Cells were harvested 48 h after transfection and assayed for luciferase and β-galactosidase (transfection control) activities. The relative luciferase activity is presented as fold change over activity obtained using the promoterless vector. Data are presented as mean ± se from n = 3–11 experiments. The significance of the observed differences (compared with V1P1) was determined by Bonferroni’s statistical test following an ANOVA analysis: *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Fig. 4.

Putative Regulatory Regions of the hGHR V1 Promoter A, NRRs and PRRs of the V1 promoter as identified by deletion promoter studies. Specific restriction enzyme sites used in the cloning of the promoter constructs are indicated. B, Putative HNF-4 binding sites identified by MatInspector and Signal Scan software programs in the 1.8-kb hGHR V1 promoter (AF322015) (4 ) and in (C) ovine (o1A), bovine (b1A), and mouse (mL1) homologous regions (26 28 30 ). Oval, HNF-4 response element.

Fig. 5.

Detection of HNF-4α Protein in Human Liver and Cell Lines by Western Blot A, Western blot: cell lines (HEK293 cells or overexpressing HNF-4α2, -α8, or both as well as HepG2 and Huh7) and human fetal hepatocytes (HFH) and human adult liver (HAL) tissues were examined for levels of HNF-4α protein. Nuclear lysates (10–50 μg) were resolved on 12% SDS-PAGE gels and immunoblotted with anti-HNF-4α that recognizes HNF-4α1, -2, -4, -5, -7, and -8. Calnexin was used as a loading control. B, Extended exposure of Huh7, human adult liver (HAL), and human fetal hepatocyte (HFH) lanes to show HNF-4α2 as the major isoform expressed in HAL, whereas HFH and Huh7 express both HNF-4α2 and -α8 isoforms in equal amounts. C, Cumulative Western blot data expressed as mean ± sd, n = 2–11. HNF-4α was not detected (ND) in HEK293 cells. The significance of the observed differences (compared with HAL) was determined by Bonferroni’s statistical test after an ANOVA analysis: *, P < 0.05; **, P < 0.01; M, Markers.

Fig. 6.

Effect of HNF-4α2 and HNF-4α8 on the Transcriptional Activity of V1 Promoter Constructs HNF-4α2 (100 ng), HNF-4α8 (100 ng), and HNF-4α2+α8 (50 ng + 50 ng) expression vectors were cotransfected with (A) 3′- and (B) 5′-deleted V1 constructs in HEK293 cells using Polyfect. Data are presented as mean ± se; n = 5–13. C, HNF-4α2 (50 ng) and varying amounts of HNF-4α8 (50–200 ng) were cotransfected with V1P4, V_XP_B, and V1P5 constructs into HEK293 cells. Cells were harvested 48 h later and assayed as before. Data are presented as mean ± se; n = 3–10 experiments. The significance of the differences observed was determined by ANOVA, followed by Bonferroni’s group comparison test. *, P < 0.05; **, P < 0.01; ***, P < 0.001. Oval, HNF-4 response element.

Fig. 7.

Mutational Analysis of HNF-4 Sites 1 (Most 3′) and 6 (Most 5′) in the Proximal Promoter of V1 Constructs with either the HNF-4 binding sites 1 or 6 mutated (supplemental Table 3) were cotransfected with HNF-α2 or -α8; nonmutated constructs were transfected as controls. Data are presented as mean ± se, n = 3–10 experiments. The significance of the differences observed was determined by ANOVA followed by Bonferroni’s statistical test: *, P < 0.05; **, P < 0.01; ***, P < 0.001. Oval, HNF-4 response element; oval with diagonal line, ½ mutated HNF-4 response element; oval with cross, mutated HNF-4 response element.

Fig. 8.

EMSA and EMSSA Analyses of HNF-4α Proteins Binding to Putative HNF-4 Sites 1, 5, and 6 in the V1 Promoter Double-stranded oligonucleotides representing the (A) 1 and (B) 6 HNF-4 sites (supplemental Table 3) were radioactively labeled and incubated with various nuclear extracts. Mutant oligonucleotides (supplemental Table 3) and a specific HNF-4α antibody that recognizes HNF-4α2+8 were used to determine the specificity of binding. Nuclear extracts were from HEK293 cells (control or overexpressing HNF-4α2 + α8). Arrows indicate NS, nonspecific; S, shift; SS, supershift bands. NE, Nuclear extract.

Fig. 9.

In Vivo ChIP Analysis of Proteins Binding to the Two Putative HNF-4 Sites in the V1 Promoter Huh 7 and HepG2 (data not shown) cells were cross-linked, and the lysates were immunoprecipitated with HNF-4α antibody or goat IgG (con IgG). After the reversal of the cross-links, PCRs were performed across HNF-4 binding sites 1, 5, and 6 and a region 2 kb upstream of HNF-4 site no. 6 as a control (control) (supplemental Table 4). Products were resolved on a 2% agarose gel. con, Control.

Fig. 10.

Alignment of Human V1, Ovine o1A, Bovine b1A, and Mouse mL1 GHR Sequences Human V1 sequence (AF322015; 500 bp) (4 ) aligned with the ovine (26), bovine (U15731) (28 ) and mouse (NT_039747) (30 ) equivalents. TSS (bold and italicized), TATA boxes (bold), and putative HNF-4 sites (bold) are indicated. The human V1 has two TATA/TSS complexes; the downstream TATA/TSS is conserved in the ovine, bovine, and mouse. *, Conserved nucleotides; arrow, upstream TSS; +1, conserved transcriptional start site.