The majority of mutations that cause isolated GH deficiency type II (IGHD II) affect splicing of GH-1 transcripts and produce a dominant-negative GH isoform lacking exon-3 resulting in a 17.5-kDa isoform, which further leads to disruption of the GH secretory pathway. A clinical variability in the severity of the IGHD II phenotype depending on the GH-1 gene alteration has been reported, and in vitro and transgenic animal data suggest that the onset and severity of the phenotype relates to the proportion of 17.5-kDa produced. The removal of GH in IGHD creates a positive feedback loop driving more GH expression, which may itself increase 17.5-kDa isoform productions from alternate splice sites in the mutated GH-1 allele. In this study we aimed to test this idea by comparing the impact of stimulated expression by glucocorticoids on the production of different GH isoforms from wild-type (wt) and mutant GH-1 genes, relying on the glucocorticoid regulatory element (GRE) within intron 1 in the GH-1 gene. AtT-20 cells were transfected with wt-GH or mutated GH-1 variants (5'IVS-3 + 2bp T->C; 5'IVS-3 + 6bp T->C; ISEm1: IVS-3 + 28 G->A) known to cause clinical IGHD II of varying severity. Cells were stimulated with 1 µM and 10 µM dexamethasone (DEX) for 24 h, after which the relative amounts of GH-1 splice variants were determined by semi-quantitative and quantitative (TaqMan®) reverse transcriptase-PCR (RT-PCR). In the absence of DEX, only around 1% wt-GH-1 transcripts were the 17.5-kDa isoform, whereas the three mutant GH-1 variants produced 29, 39, and 78% of the 17.5-kDa isoform. DEX stimulated total GH-1 gene transcription from all constructs. Notably however, DEX increased the amount of 17.5-kDa GH isoform relative to the 22-kDa and the 20-kDa isoforms produced from the mutated GH-1 variants, but not from wt-GH-1. This DEX-induced enhancement of 17.5-kDa GH isoform production, up to 100% in the most severe case, was completely blocked by the addition of RU486. In other studies we measured cell proliferation rates, annexin V staining and DNA fragmentation in cells transfected with the same GH-1 constructs. The results showed that that the 5'IVS-3 + 2bp GH-1 gene mutation had a more severe impact on those measures than the splice site mutations within 5'IVS-3 + 6bp or ISE +28, in line with the clinical severity observed with these mutations.

Our findings that the proportion of 17.5-kDa produced from mutant GH-1 alleles increases with increased drive for gene expression may help to explain the variable onset progression, and severity observed in IGHD II.