Species-specific kinetics of mouse renin contribute to maintenance of normal blood pressure in transgenic mice with overexpressed human angiotensinogen

Two Amino Acids Proximate to the Renin Cleavage Site of Human Angiotensinogen Do Not Affect Blood Pressure and Atherosclerosis in Mice-Brief Report

OBJECTIVE

Renin cleavage of angiotensinogen has species specificity. As the residues at positions 11 and 12 are different between human angiotensinogen and mouse angiotensinogen, we determined whether these 2 residues in angiotensinogen affect renin cleavage and angiotensin II-mediated blood pressure regulation and atherosclerosis using an adenoassociated viral approach for manipulating angiotensinogen in vivo. Approach and Results: Hepatocyte-specific angiotensinogen deficient (hepAGT^-/-) mice in an LDL receptor-deficient background were infected with adenoassociated virals containing a null insert, human angiotensinogen, or mouse angiotensinogen expressing the same residues of the human protein at positions 11 and 12 (mouse angiotensinogen [L11V;Y12I]). Expression of human angiotensinogen in hepAGT^-/- mice led to high plasma human angiotensinogen concentrations without changes in plasma endogenous mouse angiotensinogen, plasma renin concentrations, blood pressure, or atherosclerosis. This is consistent with human angiotensinogen not being cleaved by mouse renin. To determine whether the residues at positions 11 and 12 in human angiotensinogen lead to the inability of mouse renin to cleave human angiotensinogen, hepAGT^-/- mice were injected with adenoassociated viral vector encoding mouse angiotensinogen (L11V;Y12I). Expression of mouse angiotensinogen (L11V;Y12I) in hepAGT^-/- mice resulted in increased plasma mouse angiotensinogen concentrations, reduced renin concentrations, and increased renal AngII concentrations that were comparable to their concentrations in hepAGT^+/+ mice. This mouse angiotensinogen variant increased blood pressure and atherosclerosis in hepAGT^-/- mice to the magnitude of hepAGT^+/+ mice.

CONCLUSIONS

Replacement of L11 and Y12 to V11 and I12, respectively, in mouse angiotensinogen does not affect renin cleavage, blood pressure, and atherosclerosis in LDL receptor-deficient mice.

A nuclear receptor, hepatocyte nuclear factor 4, differently contributes to the human and mouse angiotensinogen promoter activities

Angiotensinogen (AGT), mainly produced in the liver, is the precursor of angiotensin II, an important regulator of blood pressure and electrolyte homeostasis. We previously showed, in hepatoma-derived HepG2 cells that a hepatocyte nuclear factor 4 (HNF4) potentiated human AGT (hAGT) promoter activity and identified its binding sites (termed regions C and J) in the hAGT promoter region. We also showed in transgenic mouse (TgM) that the hAGT is abundantly expressed in the kidney where the level of endogenous mouse AGT (mAGT) expression is low. To elucidate molecular mechanisms of the AGT gene activation in the kidney, we first investigated the HNF4 and AGT expression in the mouse kidney. Northern blot, in situ hybridization and immunohistochemical analyses revealed that the hAGT and HNF4 were both expressed in the proximal tubular (PT) cells of the kidney. We then transfected the hAGT reporter constructs into immortalized mouse PT (mProx) cells and found that regions C and J contributed additively to the HNF4-potentiated hAGT promoter activity. Curiously, no obvious HNF4 binding motif was found in the corresponding region of the mAGT promoter and co-transfected HNF4 failed to activate this promoter in neither HepG2 nor mProx cells. These results suggest that the high-level hAGT expression in the TgM kidney is, at least in part, due to a presence of high-affinity HNF4 binding sites in its promoter.

Response to genetic manipulations of liver angiotensinogen in the physiological range

Genetic variation in the human angiotensinogen gene (AGT) influences plasma AGT concentration and susceptibility to essential hypertension by a mechanism that remains to be clarified. When one or two additional copies of the gene were inserted by gene titration (by homologous recombination with gap-repair at the AGT locus), both plasma AGT and arterial pressure were elevated in the physiological range in the mouse. The causal dependency between plasma AGT and blood pressure and the relative contribution of the various tissues that express AGT to these two phenotypic parameters remained to be determined. To address these issues, we generated a transgenic mouse with overexpression of the mouse AGT gene restricted to the liver. The transgene was examined in two contrasted genetic backgrounds, the sodium-sensitive C57BL/6J and the sodium-resistant A/J. Transgenic and control male animals underwent continuous cardiovascular monitoring by telemetry for 14 days while under a standard sodium diet (0.2%). Moderate but significant increases in plasma AGT (40%, p = 0.01) and systolic blood pressure (4-6 mmHg, p ranging from 0.01 to <0.001) were observed in the sodium-sensitive background, but not in the sodium-resistant animals. Statistical analysis of a large number of consecutive, repeated measurements of blood pressure afforded power to detect small effects in the physiological range by use of advanced mixed models of analysis of variances and covariances. Although plasma renin activity was increased in the sodium-sensitive background, it did not reach statistical significance. These observations underline a potential contribution of systemic AGT to the mechanism of AGT-mediated hypertension, but the significance of sodium sensitivity in the genetic background suggests participation of the kidney in expression of the elevated blood pressure phenotype, a matter that will warrant further studies. They also highlight the challenge of identifying the contribution of individual genes in complex inheritance, as their effects are modulated by other genetic and environmental determinants.

Identification of cis-regulatory sequences in the human angiotensinogen gene by transgene coplacement and site-specific recombination

The function of putative regulatory sequences identified in cell transfection experiments can be elucidated only through in vivo experimentation. However, studies of gene regulation in transgenic mice (TgM) are often compromised by the position effects, in which independent transgene insertions differ in expression depending on their location in the genome. In order to overcome such a dilemma, a method called transgene coplacement has been developed in Drosophila melanogaster. In this method, any two sequences can be positioned at exactly the same genomic site by making use of Cre/loxP recombination. Here we applied this method to mouse genetics to characterize the function of direct repeat (DR) sequences in the promoter of the human angiotensinogen (hAGT) gene, the precursor of the vasoactive octapeptide angiotensin II. We modified a hAGT bacterial artificial chromosome to use Cre/loxP recombination in utero to generate TgM lines bearing a wild-type or a mutant promoter-driven hAGT locus integrated at a single chromosomal position. The expression analyses revealed that DR sequences contribute 50 or >95% to hAGT transcription in the liver and kidneys, respectively, whereas same sequences are not required in the heart and brain. This is the first in vivo dissection of DNA cis elements that are demonstrably indispensable for regulating both the level and cell type specificity of hAGT gene transcription.

Differential roles of renin and angiotensinogen in the feto-maternal interface in the development of complications of pregnancy

We previously identified a transgenic mouse model that developed pregnancy-associated hypertension (PAH) and intrauterine growth restriction (IUGR) by mating females expressing human angiotensinogen (hANG) with males expressing human renin (hRN). These phenotypic defects were not observed in the opposite type of mating combination, despite the feto-placental overexpression of hRN and hANG detected in both types of crossbreeding. Detailed analysis of transgene localization in the labyrinth and its permeability to the maternal circulation revealed that hRN produced in trophoblast giant cells was secreted into the maternal circulation, whereas hANG, produced in chorionic trophoblasts and trophoblastic epithelium, was undetectable in the maternal plasma, probably due to their distinct spatial and temporal expression in labyrinth. These results demonstrated that PAH and IUGR could be mediated by feto-placental hRN through its permeability to the maternal circulation, not by feto-placental hANG production. Furthermore, overexpression of maternally derived hANG in decidua and spiral arteries of pregnant females with PAH and IUGR raises the possibility of local activation of the renin-angiotensin system and its pathophysiological effects on placental hypoperfusion in complications of pregnancy. This study provides in vivo evidence that the cell-specific expression of RN and ANG in the feto-maternal interface impacts their differential roles in pregnancy.

Demonstration of an early and a late phase of ischemic preconditioning in mice

It is unknown whether ischemic preconditioning (PC; either early or late) occurs in the mouse. The goal of this study was to answer this question and to develop a reliable and physiologically relevant murine model of both early and late ischemic PC. A total of 201 mice were used. In nonpreconditioned open-chest animals subjected to 30 min of coronary occlusion followed by 24 h of reperfusion, infarct size (tetrazolium staining) averaged 52% of the region at risk. When the 30-min occlusion was performed 10 min after a PC protocol consisting of six cycles of 4-min occlusion and 4-min reperfusion, infarct size was reduced by 75%, indicating an early PC effect. When the 30-min occlusion was performed 24 h after the same PC protocol, infarct size was reduced by 48%, indicating a late PC effect. In mice in which the 30-min occlusion was followed by 4 h of reperfusion, infarct size was similar to that observed after 24 h of reperfusion, indicating that a 4-h reperfusion interval is sufficient to detect the final extent of cell death in this model. Fundamental physiological variables (body temperature, arterial oxygenation, acid-base balance, heart rate, and arterial pressure) were measured and found to be within normal limits. Taken together, these results demonstrate that, in the mouse, a robust infarct-sparing effect occurs during both the early and the late phases of ischemic PC, although the early phase is more powerful. This murine model is physiologically relevant, provides reliable measurements, and should be useful for elucidating the cellular mechanisms of ischemic PC in genetically engineered animals.