26Isolated growth hormone deficiency type II (IGHD2) is mainly caused by 27 heterozygous splice-site mutations in intron 3 of the GH1 gene. A dominant negative 28 effect of the mutant growth hormone (GH) lacking exon 3 on wild-type GH secretion has 29 been proposed; however, the molecular mechanisms involved are elusive. To uncover 30 the molecular systems underlying GH deficiency in IGHD2, we established IGHD2 31 model mice, which carry both wild-type and mutant copies of the human GH1 gene, 32replacing each of the endogenous mouse Gh loci. Our IGHD2 model mice exhibited 33 growth retardation associated with intact cellular architecture and mildly activated ER 34 stress in the pituitary gland, caused by decreases in the growth hormone releasing 35 hormone receptor (Ghrhr) and Gh gene promoter activities. Decreases in Ghrhr and Gh 36 promoter activities were likely caused by reduced levels of nuclear CREB3L2, which 37 was demonstrated to stimulate the activity of the Ghrhr and Gh promoters. This is the 38 first in vivo study revealing a novel molecular mechanism of GH deficiency in IGHD2, 39 representing a new paradigm, differing from widely accepted models. 40 41 Key Words: dominant negative effect/ endoplasmic reticulum stress/ growth hormone/ 42 growth hormone releasing hormone receptor/ isolated growth hormone deficiency type 43 II 44 45 22-kDa wild-type GH protein; however, the mutant GH1 allele transcript generates a 52 17.5-kDa exon 3 deletion-mutant GH (Δ3 GH), as a result of in-frame skipping of exon 3. 53The fact that patients harboring a deletion in one GH1 allele exhibit normal stature 54indicates that a single wild-type GH1 allele is sufficient to produce normal levels of 55 wild-type GH secretion (Akinci et al, 1992); however, patients with IGHD2 have low 56 serum concentrations of wild-type GH, despite having a wild-type GH1 allele. Thus, it 57 has been suggested that Δ3 GH exerts a dominant negative effect on wild-type GH 58 secretion; however, the precise molecular mechanisms involved have remained elusive 59 for more than 20 years. 60 3 Several in vitro studies have demonstrated that Δ3 GH is not secreted 61 extracellularly (Graves et al, 2001; Iliev et al, 2005; Kannenberg et al, 2007; Mullis et al, 62 2002; Salemi et al, 2006), suggesting that the dominant negative effect of Δ3 GH is 63 exerted within somatotropic cells of the pituitary, where GH is generated. Generally, 64 mutant proteins exert dominant negative effects on secretory pathways of wild-type 65 factors at the protein level (Deladoey et al, 2001; Ito et al, 1999; Jacobson et al, 1997), 66 which has led many researchers to focus on wild-type and Δ3 GH protein interactions, 67 such as heterodimer formation; however, no study has yet demonstrated definitive 68 evidence of heterodimers comprising wild-type and Δ3 GH proteins. At present, it is 69 widely accepted that Δ3 GH itself is not harmful to somatotroph (Graves et al, 2001), 70 and that wild-type GH contributes to the degradation of Δ3 GH via protein interactions, 71 leading ...