Regulation of human growth hormone receptor expression by microRNAs

Abstract

Human GH binds to its receptor (GHR) on target cells and activates multiple intracellular pathways, leading to changes in gene expression, differentiation, and metabolism. GHR deficiency is associated with growth and metabolic disorders whereas increased GHR expression has been reported in certain cancers, suggesting that the GHR gene requires tight controls. Several regulatory mechanisms have been found within its 5'-untranslated region (UTR) promoter and coding regions. However, the 3'-UTR has not been previously examined. MicroRNAs (miRNAs) are small (19-22 nucleotides) noncoding RNAs that downregulate gene expression mainly through targeting the 3'-UTR of mRNAs and enhancing their degradation or inhibiting translation. In the present study, we investigated whether miRNAs regulate GHR expression. To define putative miRNA binding sites in the GHR 3'-UTR, we used multiple in silico prediction tools, analyzed conservation across species and the presence of parallel sites in GH/IGF axis-related genes, and searched for reports linking miRNAs to GHR-related physiological or pathophysiological activities. To test prioritized sites, we cotransfected a wild-type GHR 3'-UTR luciferase reporter vector as well as miRNA binding site mutants into HEK293 cells with miRNA mimics. Furthermore, we tested whether the miRNAs altered endogenous GHR mRNA and protein levels in HEK293 cells and in 2 cancer cell lines (MCF7 and LNCaP). Our experiments have identified miRNA (miR)-129-5p, miR-142-3p, miR-202, and miR-16 as potent inhibitors of human GHR expression in normal (HEK293) and cancer (MCF7 and LNCaP) cells. This study paves the way for the development of miRNA inhibitors as therapeutic agents in GH/GHR-related pathophysiologies, including cancer.

Figure 1.

Putative miRNA binding sites in the human GHR 3′-UTR. Schematic representation of the 3′-UTR sequence from the human GHR mRNA indicating the potential miRNA binding sites within the sequence tested during the present study. A, B, and C denote multiple binding sites for the same miR.

Figure 2.

Effect of miRNAs on the GHR 3′-UTR luciferase reporter vector. Effects of 1nM to 100nM miR-129–5p, miR-142–3p, and miR-202 and 10nM to 100nM miR-16 on the pmiR-Luc-GHR 3′-UTR reporter vector were examined in HEK293 cells. Data are presented as mean ± SE (n = 5). All results were normalized to the empty Luc reporter vector and to negative mimic control treatment at each mimic dose followed by the calculation of luciferase to β-galactosidase ratios. Results for the first 3 miRNAs were highly significant by both ANOVA and Tukey's group comparison tests, whereas the miR-16 data were significant for ANOVA (P < .05) but not Tukey's group comparisons. This is likely due to the relatively high level of endogenous miR-16 in the HEK293 cells. *, P ≤ .05; ***, P ≤ .001.

Figure 3.

Reporter assays of wild-type (wt) vs mutant (mut) human GHR 3′-UTR vectors. Panel A, Effects of 50nM miRNA mimics on wild-type vs mutant GHR 3′-UTR reporter vectors. Results are presented as mean ± SE (n = 4 for all but miR-202/AB n = 5). Panel B, Schematic representation of hybridization between miR-129–5p and its 3 putative GHR 3′-UTR binding sites (A, B and C); the seed sites are bold, noncomplementary nucleotides are italicized, lines represent Watson-Crick complementarity, and dots represent U to G wobble. Panel C, Reporter assays of wild-type GHR 3′-UTR and the predicted miR-129–5p single, double, and triple mutants. Data are presented as mean ± SE (n = 4). Panel D, Schematic representation of hybridization between miR-202 and its 2 putative binding sites in the GHR 3′-UTR (A and B); the seed sites are bold, noncomplementary nucleotides are italicized, lines represent Watson-Crick complementarity, and dots represent U to G wobble. Panel E, Normalized data from luciferase assays of single and double mutants of miR-202 binding sites compared with the wild-type vector. Results are presented as mean ± SE (n = 5). **, P ≤ .01; ***, P ≤ .001; ns, not significant.

Figure 4.

Effect of miRNAs on endogenous GHR mRNA and protein in HEK293 cells. A, Expression of human total GHR mRNA in HEK293 cells transfected with 50nM miR-129–5p, miR-142–3p, miR-202, and miR-16 mimics. The data are presented as mean ± SE of 3 individual experiments for all but miR-142–3p (n = 5); duplicates were assayed within each experiment. Statistically significant differences between the negative control and miR-treated cells were found for miR-129–5p, miR-142–3p, miR-202 (***, P ≤ .001) and miR-16 (*, P ≤ .05). B, Composite results for n = 4 individual experiments (mean ± SE) showing statistically significant differences between negative control and miR-treated cells: ***, P ≤ .001 for all treatments. C, Representative Western blot of GHR and calnexin protein levels in HEK293 cells after transfections with 50nM miR-129–5p, miR-142–3p, miR-202, and miR-16 mimics.

Figure 5.

Effect of miRNAs on endogenous GHR mRNA and protein expression in MCF7 cells. A, Expression of human total GHR mRNA in MCF7 cells transfected with 50nM of each mimic. The data are presented as mean ± SE of 4 individual experiments for all but miR-202 (n = 5); duplicates were assayed within each experiment. Statistically significant differences between the control and miR-treated cells were found for miR-129–5p, miR-202 (***, P ≤ .001), and miR-142–3p (**, P ≤ .01), whereas miR-16 had no effect (not significant [ns]). B, GHR protein levels were determined by Western blot and normalized to calnexin. Normalized values are shown relative to the negative mimics control treatments (100%). Composite results for n = 4 individual experiments (mean ± SE) showing statistically significant differences between negative control and miR-treated cells: *, P ≤ .05 for all treatments. C, Representative Western blot of GHR and calnexin protein levels in MCF7 cells after transfections with 50nM miR-129–5p, miR-142–3p, miR-202, and miR-16 mimics.

Figure 6.

Effect of miRNAs on endogenous GHR mRNA and protein expression in LNCaP cells. A, Expression of human total GHR mRNA in LNCaP cells transfected with 50nM of each mimic. The data are presented as mean ± SE of 3 individual experiments for all but miR-129–5p (n = 5) and miR-202 (n = 4); duplicates were assayed within each experiment. Statistically significant differences between the control and miR-treated cells were found for miR-129–5p, miR-202, and miR-16 (***, P ≤ .001) but not for miR-142–3p (not significant [ns]). B, GHR protein levels were determined by Western blot and normalized to calnexin. Normalized values are shown relative to the negative mimics control treatments (100%). Composite results for n = 4 individual experiments (mean ± SE) showing statistically significant differences between negative control and miR-treated cells for miR-129–5p, miR-202 (***, P ≤ .001), miR-142–3p, and miR-16 (**, P ≤ .01) for all treatments. C, Representative Western blot of GHR and calnexin protein levels in LNCaP cells after transfections with 50nM miR-129–5p, miR-142–3p, miR-202, and miR-16 mimics.