To determine the importance of the third a-helix in bovine growth hormone (bGH) relative to growthrelated biological activities, the following experimental approach was used: (i) mutagenesis ofhelix III of bGH to generate an idealized amphiphilic helix; (ii) in vitro expression analyses of the mutated bGH gene in cultured mouse L cells; (iii) mouse liver membrane binding studies ofwild-type and mutated bGH; and (iv) expression ofthe mutated gene in the transgenic mouse. An altered bGH gene (pBGHlOA6-M8) was generated that encodes the following changes: glutamate-117 to leucine, glycine-119 to arginine, and alanine-122 to aspartate. The plasmid pBGHlOA6-M8 was shown to be expressed in, and its protein product secreted by, mouse L cells. The altered hormone possessed the same binding affinity to mouse liver membrane preparations as wild-type bGH. Transgenic mice containing the mutated bGH gene, however, showed a significant growthsuppressed phenotype. The degree of suppression was directly related to serum levels of the altered bGH molecule. >90% amino acid sequence identity between bGH and pGH, it is likely that bGH has a similar three-dimensional structure. By aligning these a-helical structures into a two-dimensional Edmundson wheel projection (24), it is clear that an amphiphilic a-helical segment exists between amino acid residues 109 and 126 in the third a-helical region (ref. 25; Fig. 1). However, amino acids 117, 119, and 122 are positioned so that an idealized amphiphilic a-helix is not generated.In this study, we have used the following approach to determine whether this amphiphilic a-helix is important in the growth-related properties ofbGH: (i) mutagenesis ofthe third a-helix of bGH so as to generate an idealized amphiphilic a-helix (Fig. 1); (ii) expression analyses of the wild-type and mutated bGH genes in cultured mouse L cells; (iii) mouse liver membrane binding studies comparing wild-type and mutated bGH; and (iv) production of transgenic mice containing wild-type and mutated bGH genes.

Bovine GH (bGH) analogs with single amino acid substitutions at positions 117 (bGH-E117L), 119 (bGH-G119R), and 122 (bGH-A122D) were generated. These analogs bind to mouse liver membrane preparations with affinities similar to native bGH. However, transgenic mice which express the analogs demonstrate different phenotypes ranging from dwarfism to gigantism. For example, expression of bGH or bGH-E117L result in large transgenic mice. In contrast, transgenic mice with a growth phenotype similar to nontransgenic animals result from expression of bGH-A122D. Surprisingly, transgenic mice with relatively high serum levels of bGH-G119R possessed a dwarf phenotype. Together these results suggest that Gly 119 and Ala 122 are involved in growth-promoting activity of GH.

To study the physiological effect of the overexpression of myocardial Gsalpha (protein levels increased by approximately threefold in transgenic mice), we examined the responsiveness to sympathomimetic amines by echocardiography (9 MHz) in five transgenic mice and five control mice (both 10.3 +/- 0.2 months old). Myocardial contractility in transgenic mice, as assessed by left ventricular (LV) fractional shortening (LVFS) and LV ejection fraction (LVEF) was not different from that of control mice at baseline (LVFS, 40 +/- 3% versus 36 +/- 2%; LVEF, 78 +/- 3% versus 74 +/- 3%). LVFS and LVEF values in transgenic mice during isoproterenol (ISO, 0.02 micrograms/kg per minute) infusion were higher than the values in control mice (LVFS, 68 +/- 4% versus 48 +/- 3%; LVEF, 96 +/- 1% versus 86 +/- 3%; P < .05). Norepinephrine (NE, 0.2 micrograms/kg per minute) infusion also increased LVFS and LVEF in transgenic mice more than in control mice (LVFS, 59 +/- 4% versus 47 +/- 3%; LVEF, 93 +/- 2% versus 85 +/- 3%; P < .05). Heart rates of transgenic mice were higher than those of control mice during ISO and NE infusion. In three transgenic mice with heart rates held constant, LV dP/dt rose by 33 +/- 2% with ISO (0.02 micrograms/kg per minute) and by only 13 +/- 2% in three wild-type control mice (P < .01). NE (0.1 micrograms/kg per minute) also induced a greater effect on LV dP/dt in the three transgenic mice with heart rates held constant compared with three wild-type control mice (65 +/ 8% versus 28 +/- 4%, P < .05). Pathological and histological analyses of older transgenic mouse hearts (16.0 +/- 0.8 months old) revealed hypertrophy, degeneration, atrophy of cells, and replacement fibrosis reflected by significant increases in collagen volume in the subendocardium (5.2 +/- 1.4% versus 1.2 +/- 0.3%, P < .05) and in the cross-sectional area of myocytes (298 +/- 29 versus 187 +/- 12 micron2, P < .05) compared with control mouse hearts. These results suggest that Gsalpha overexpression enhances the efficacy of the beta-adrenergic receptor-Gs-adenylyl cyclase signaling pathway. This in turn leads to augmented inotropic and chronotropic responses to endogenous sympathetic stimulation. This action over the life of the animal results in myocardial damage characterized by cellular degeneration, necrosis, and replacement fibrosis, with the remaining cells undergoing compensatory hypertrophy. As a model, this transgenic mouse offers new insights into the mechanisms of cardiomyopathy and heart failure and provides a new tool for their study.

The goal of this study was to determine whether chronic endogenous sympathetic stimulation resulting from the overexpression of cardiac stimulatory G protein alpha subunit (Gs alpha) in transgenic mice (15.3 +/- 0.1 mo old) resulted in a clinical picture of cardiomyopathy. The left ventricular ejection fraction, measured by echocardiography, was reduced in older mice with Gs alpha overexpression (50.4 +/- 5.4%) compared with age-matched control mice (70.9 +/- 1.6%; P < 0.05). When ejection fractions were compared at similar heart rates, the Gs alpha mice exhibited a greater left ventricular end-diastolic dimension than control mice (4.3 +/- 0.2 vs. 3.7 +/- 0.1 mm; P < 0.05). Baseline heart rates were elevated in conscious Gs alpha mice (722 +/- 27 beats/min; n = 5) compared with control mice (656 +/- 28 beats/min; n = 5). Moreover, electrocardiographic monitoring demonstrated a high incidence of arrhythmias. Increased mortality compared with control mice (31.6 vs. 3.0%; P < 0.01) was also observed. Thus older mice with Gs alpha overexpression exhibit many of the features of dilated cardiomyopathy. This study supports the concept that chronic sympathetic stimulation over an extended period of time, i.e., over the life of an animal, is deleterious and actually may result in cardiomyopathy.

Alterations in fi-adrenergic receptor-G.-adenylyl cyclase coupling underlie the reduced catecholamine responsiveness that is a hallmark of human and animal models of heart failure. To study the effect of altered expression of Gs., we overexpressed the short isoform of Gs,, in the hearts of transgenic mice, using a rat a-myosin heavy chain promoter. G.. mRNA levels were increased selectively in the hearts of transgenic mice, with a level 38 times the control. Despite this marked increase in mRNA, Western blotting identified only a 2.8-fold increase in the content of the Gs,, short isoform, whereas G, activity was increased by 88%. The discrepancy between Gs. mRNA and Gs,, protein levels suggests that the membrane content of Gsa is posttranscriptionally regulated. The steady-state adenylyl cyclase catalytic activity was not altered under either basal or stimulated conditions (GTP + isoproterenol, GTPyS, NaF, or forskolin). However, progress curve studies did show a significant decrease in the lag period necessary for GppNHp to stimulate adenylyl cyclase activity. Furthermore, the relative number of /3-adrenergic receptors binding agonist with high affinity was significantly increased. Our data demonstrate that a relatively small increase in the amount of the coupling protein Gs. can modify the rate of catalyst activation and the formation of agonist high affinity receptors. (J.