George Lyman Duff Lecture Angiotensinogen as a Therapeutic Target for Cardiovascular and Metabolic Diseases

Abstract

Angiotensinogen (AGT) is the unique precursor for the generation of all the peptides of the renin angiotensin system (RAS), but it has received relatively scant attention compared to many other RAS components. Focus on AGT has increased recently, particularly with the evolution of drugs to target the synthesis of the protein. AGT is a non-inhibitory serpin that has several conserved domains in addition to the angiotensin II sequences at the N terminus. Increased study is needed on the structure-function relationship to resolve many unknowns regarding AGT metabolism. Constitutive whole-body genetic deletion of Agt in mice leads to multiple developmental defects creating a challenge to use these mice for mechanistic studies. This has been overcome by creating Agt floxed mice to enable development of cell-specific deficiencies that have provided considerable insight into a range of cardiovascular and associated diseases. This has been augmented by the recent development of pharmacological approaches targeting hepatocytes in humans to promote protracted inhibition of AGT synthesis. Genetic deletion or pharmacological inhibition of Agt has been demonstrated to be beneficial in a spectrum of diseases experimentally, including hypertension, atherosclerosis, aortic and superior mesenteric artery aneurysms, myocardial dysfunction, and hepatic steatosis. This review summarizes the findings of recent studies utilizing AGT manipulation as a therapeutic approach.

Introduction

The renin angiotensin system (RAS) has long been recognized as critical to the regulation of blood pressure as well as water and sodium hemostasis. The RAS has now been implicated in a wide array of cardiovascular and associated diseases. The unique substrate of the RAS is angiotensinogen (AGT). In humans, it is synthesized as a 485 amino acid protein (452 amino acids after removal of the signal peptide), with almost all the attention on the protein having been focused on the bioactive peptides that are produced from the 10 N-terminal amino acids – namely, angiotensin I (AngI).¹ Although many tissues have the potential to synthesize and secrete AGT, the literature is consistent with hepatocytes being the predominant source.^2-4 However, it is unclear where AGT is cleaved to facilitate the generation of angiotensin peptides. For example, several studies are consistent with hepatocyte-derived AGT accumulating in distant tissues such as the proximal convoluted tubules of the kidney.^{2, 5, 6} Overall, there are still many aspects of AGT biology that need further investigation to determine its role in disease processes and its suitability as a therapeutic target.

Structure and Function Relations in AGT

AGT is a non-inhibitory member of the serpin family. AGT’s structure has now been resolved with a high level of specificity for the mouse, rat, and human.⁷ The protein has several highly conserved sequences in regions that are well known to interact with renin, as well as in regions that are distal to the site of renin interaction.^{3, 8} Most research on AGT’s structure has focused on the N terminus, particularly the 10 residues forming AngI that are released following renin cleavage. Based on computational structural modeling, it has been speculated that a disulfide bond between Cys18 and Cys138 (Cys137 in mouse) is redox modulated, making the reduced form less amenable to renin cleavage.⁷ However, the repopulation of AGT-depleted mice with AGT that was unable to form the disulfide bond had no discernable effects on AngII-mediated effects, compared to wild-type AGT.⁴ It was also speculated that the residues immediately proximal to the renin cleavage site of AGT play a crucial role in its cleavage by renin.⁹ However, mouse AGT, in which the two amino acids proximal to the renin cleavage site were mutated to human residues (L1V and Y12I), in mice with deletion of endogenous AGT did not impact angiotensin II (AngII) production or AngII-mediated effects.¹⁰

In addition to the sites on AGT’s proximal face, sites in the beta-sheet region and the loop region on the distal face can interact with renin (Figure 1). It has been proposed that these two regions determine AGT distribution in the kidney.⁵ For example, there has been a consistent demonstration that deletion of hepatocyte-derived AGT results in reduced AGT protein accumulation in the proximal tubules in both mice and monkeys.^{2, 5, 6} Furthermore, protein accumulation is decreased markedly in the absence of low-density lipoprotein-related protein 2 (LRP2, also known as megalin).^{2, 5, 11} However, selective mutations for highly conserved sequences of the distal face of AGT failed to demonstrate any reductions in AGT interactions with LRP2.¹² There are strong structural similarities between LRP2 and LRP1. Indeed, AGT is bound and endocytosed by LRP1 in the heart.¹³ Currently, there have been no data on whether the conserved sequences of AGT are responsible for its interaction with LRP2.

Figure 1. Structure of angiotensinogen (AGT).

Conserved structural features were analyzed based on the comparisons of AGT protein sequences from human, rat, mouse, and zebrafish. Two highly conserved regions, β-sheet and loop, were identified in the des(AngI)AGT domain. This figure is modified from Figure 6A published in Arteriosclerosis, Thrombosis, and Vascular Biology.³

Overall, there are many unanswered questions regarding the impact of structural aspects of AGT on the biological activities of the protein.

Measurement of Plasma Concentrations of Intact AGT and des(AngI)AGT

Plasma concentrations of intact AGT have been traditionally measured using indirect approaches that are based on generation of AngI in the presence of excess renin.¹⁴ AGT can now be measured by ELISA to provide absolute concentrations of protein mass. Initial use of the ELISA kit developed by IBL-America for rodents provided plasma concentrations in the range of 1 to 2 μg/ml,¹⁵ while subsequent studies resulted in slightly higher concentrations in the range of 2 to 4 μg/ml.³ The ELISA developed by IBL-America to measure AGT in human plasma has demonstrated that the protein is present in much higher concentrations that range from approximately 15 to 70 μg/ml,¹⁶ with higher concentrations in females.¹⁷ Plasma concentrations of AGT are comparable in humans and non-human primates, which may provide a basis for performing studies in the latter species for enhanced translation.⁶ There is also an ELISA kit developed by abcam for rodent measurements. The plasma AGT concentrations determined using this kit have been considerably higher compared to values derived from the IBL-America kit, with measurements in the range of 8 to 20 μg/ml when measured on the same strain of mice.¹⁸ While the absolute concentrations of AGT appear to differ between the two available kits, interventions that alter plasma AGT concentrations have demonstrated consistent relative changes.^{3, 18} These studies illustrate the need to maintain consistent conditions for the measurements of plasma AGT concentrations.

The ELISA kits made by IBL and abcam measure total AGT concentrations, therefore, they cannot distinguish between intact AGT and the renin-cleaved product, des(AngI)AGT. Recent ELISA development has provided an opportunity to determine the concentrations of intact AGT alone. In mouse studies, the relative concentrations of intact AGT are between 10 to 20% of total AGT.^{19, 20} Therefore, there is a minimal concentration of AGT in mouse plasma that can be cleaved by renin. In the few studies in humans and non-human primates, it has been demonstrated consistently that approximately 50% of AGT in plasma is present as intact AGT.²⁰ In aggregate, interpretation of plasma-derived AGT data needs to take into account the absolute concentration, the mode of measurement, and the presence of the intact relative to cleaved forms of the protein.

AGT in Cardiovascular and Metabolic Diseases

Hypertension

The RAS has long been known to be a critical regulator of blood pressure. Many single nucleotide polymorphisms (SNPs) have been identified in the human AGT gene, with the M235T polymorphism being studied the most extensively. An initial study reported an association between the M235T polymorphism and essential hypertension, coupled with a correlation of T235 to higher plasma AGT concentrations.²¹ However, findings regarding the link between M235T and hypertension have exhibited inconsistency in subsequent studies.^22-29 It is also worth noting that these studies measured plasma AGT concentrations by assessing the generation of AngI in the presence of excess human renin. In contrast, a recent study encompassing 5,786 participants from the MESA (Multi-Ethnic Study of Atherosclerosis) cohort, measured plasma AGT concentrations using the ELISA kit developed by IBL-America.¹⁷ The results of this study revealed a positive association between plasma AGT concentrations and the prevalence of hypertension in both male and female subjects.

To directly focus on the role of AGT, an initial study manipulated the copy number of the Agt gene in mice.³⁰ This study demonstrated that plasma AGT concentrations were a direct indicator of the number of Agt genes. Furthermore, there was a direct relationship between the Agt gene copy number and systolic blood pressure.³⁰ Genetic deletion of Agt profoundly reduced systolic blood pressure, although these mice also had severe developmental abnormalities.^30-34 Interestingly, a recent study demonstrated that effects of Agt deletion on blood pressure are strain-dependent.³⁵ While AGT has the ability to generate several vasoactive peptides, it is generally assumed that AngII is the primary mediator of blood pressure regulation. The dependency of AngII activity on AT1a receptors in the kidney was demonstrated by a combination of whole body deletion of AT1a receptors as well as transplantation of kidney deficient in AT1a receptors.³⁶ Subsequent studies using mice with Cre driven by a Pepck or Sglt2 promoter bred to mice with Agtr1a floxed alleles demonstrated that AT1a receptor in proximal tubules regulated AngII-mediated blood pressure.^{37, 38} The proximal convoluted tubules express Agt mRNA predominantly in the S3 segment of the proximal tubules and presumably have the ability to synthesize AGT.^{2, 11} However, the functional significance of AGT synthesis in this region has not been defined.³⁹ Indeed, genetic deletion of renal Agt has no effect on AGT protein in S1 and S2 of renal proximal tubules.² Hepatocyte-specific deletion of Agt markedly reduced accumulation of the AGT protein in kidney proximal convoluted tubules in both mice and monkeys in multiple studies.^{2, 3, 6} AGT accumulates in the proximal convoluted tubule through an interaction with LRP2.^{2, 5, 11} The functional significance of hepatocyte-derived AGT was demonstrated when mice with hepatocyte-specific deletion of Agt were found to have a marked reduction in systolic blood pressure.³ The inference that the proximal convoluted tubules is the site of AngII production is based on the profound reduction of renal AngII in mice in which LRP2 abundance was decreased by antisense oligonucleotides (ASO) or genetic deletion of Lrp2.^{5, 11} In both circumstances, no changes were detected in plasma AngII concentrations.

While hepatocyte-derived AGT has a major effect on renal accumulation of AGT, and appears to be important for the regulation of blood pressure, there are significant knowledge gaps on the mechanisms of how AGT acts in the kidney to regulate systolic blood pressure. For example, while LRP2 is required for the accumulation of both renin and AGT in proximal convoluted tubules, immunostaining indicated that AGT and renin accumulate in distinct intracellular loci in S1 and S2 proximal convoluted tubules.^{5, 11} Therefore, it is not clear how their cellular accumulation can lead to renin cleavage of AGT to release AngI. Also, while ACE is expressed in proximal convoluted tubules, it is predominantly located in the S3 segment, necessitating the need for AngI to be transferred to this location for the conversion to AngII. The site of AngII action also needs further investigation since there are conflicting studies. Mice in which AT1a receptors were deleted using Ndrg1 driven Cre had no difference in systolic blood pressure compared to their wild type littermates,⁴⁰ whereas mice developed using the Sglt2 driven Cre had a reduction of systolic blood pressure that was approximately half of that reduction observed in mice with global deletion of A1a receptor.³⁸

Atherosclerotic Vascular Diseases

There has been a substantial body of work on the role of the RAS in experimental atherosclerosis.⁴¹ Several modes of attenuating effects of the RAS, including inhibitors of AT1 receptors, ACE, and renin, have been effective in reducing atherosclerotic lesions size across a range of animal models.^42-44 Initial studies in Agt genetic manipulation were performed in mice that overexpress the human protein. Because of the species specificity to the actions of renin, a compound transgenic mouse was generated that has been designated the “Tsukuba hypertensive mouse”.⁴⁵ The mouse was created on a C57BL/6J background and fed a high-fat diet that increased plasma cholesterol concentrations. The augmented atherosclerosis in the mice overexpressing AGT was attributed to hypertension. However, while hypertension is associated with increased atherosclerosis in mice, it does not appear that the increase in blood pressure, per se, is the cause of the increased atherosclerosis.⁴¹

Initial studies on the role of AGT in experimental atherosclerosis were performed in genetically manipulated mice with a hypomorphic allele for Agt that resulted in an ~90% reduction in plasma AGT concentrations. These LDL receptor deficient mice had a striking reduction in experimental atherosclerosis with almost complete ablation of lesions, despite the mice having persistent hypercholesterolemia. A similar profound reduction in atherosclerosis also occurred in mice in which AGT was reduced in adult mice by injection of ASO targeting Agt. The liver appears to be the source of AGT that influences atherosclerosis, based on these two studies.^{3, 46} While these effects of AGT inhibition on atherosclerosis are profound, the striking effects in mice with hepatocyte-specific deletion does not have a clear mechanistic basis and are fertile ground for future studies. Previous studies in humans, such as HOPE⁴⁷ and ONTARGET,⁴⁸ have consistently demonstrated that either ACE inhibition or AT1 receptor blockade profoundly reduces morbidity and mortality associated with atherosclerotic cardiovascular diseases. We anticipate that the results observed with AGT inhibition in mouse models will be translated into therapies that reduce atherosclerotic diseases in humans.

Aneurysmal Vascular Diseases

There is an extensive literature consistently demonstrating that AngII infusion into mice promotes a spectrum of aortic pathologies in both the thoracic and abdominal regions.^{49, 50} Conversely, there has been numerous publications that inhibition of AngII activation of AT1 receptors attenuates the aortic pathology in mouse models in which aneurysm formation is provoked by a wide range of manipulations.⁵¹ However, some publications have questioned whether the benefit observed in models of thoracic aortic disease with AT1 receptor blockers is attributable to inhibition of AngII engagement. One publication has proposed that the benefit of losartan administration in fibrillin-1 hypomorphic mice is attributed to stretch-based activation of AT1a receptors, independent of ligand engagement.⁵² Another study attributed the positive benefits of losartan-induced reduction in aortic dimensions of fibrillin-1 haploinsufficient (C1041G/+ mutation) mice to increased nitric oxide release, rather than inhibition of AT1a receptor activation.⁵³ However, a subsequent study using globally deleted AT1a receptors in fibrillin-1 C1041G/+ mice demonstrated profound reductions in aortic dimensions for the 1 year interval of investigation.⁵⁴ This effect was only noted in male mice, since the strong sexual dimorphic effect of fibrillin-1 haploinsufficiency only led to profound aortic expansion over this interval in males, but not females. Consistent with these findings, ASO targeting AGT inhibited aortic root and ascending aortic dilation compared to control ASO in fibrillin-1 C1041G/+ mice. Both findings are consistent with AT1a receptor stimulation being AngII ligand dependent in Marfan-associated aortopathy. The AGT ASO study also suggests a benefit of decreased synthesis of AGT for people with Marfan syndrome, as described in a recently issued patent.⁵⁵ Further investigation of this approach would be valuable, given the limited pharmacological armamentarium of validated medical approaches for Marfan syndrome.

Aortic dissection is a devastating condition that is not always preceded by aortic aneurysm. Administration of BAPN, a lysyl oxidase inhibitor, leads to a high incidence of fatal aortic dissections when administered to 3-4-week-old mice.⁵⁶ Although Agt ASO and inhibition of AT1 receptors mitigated development of aortic aneurysms, neither prevented BAPN-induced aortic dissections.¹⁸ Therefore, inhibition of AGT is not a panacea for the treatment of all aortopathies.

LRP1 is a multifunctional protein with a critical role in the clearance of extracellular proteases.⁵⁷ Smooth muscle cell specific deletion of Lrp1 leads to aortopathies that are augmented by AngII infusion.^{58, 59} In addition, smooth muscle cell-specific deletion of Lrp1 results in development of aneurysms in the superior mesenteric artery. While the mechanism underlying this pathology has not been elucidated, a contributing factor could be the enhancement of many ligands of LRP1 that are augmented by AngII, including PAI-1, CTGF (connective tissue growth factor), and TGF (transforming growth factor)-β.⁶⁰ Consistent with this premise, administration of Agt ASO as well as blockade of AT1 receptors reduced the generation of aneurysms in the superior mesenteric artery in smooth muscle cell-specific LRP1 deficient mice.⁶¹

Overall, there are limited studies in mice evaluating the impact of AGT synthesis on non-atherosclerotic vascular diseases. However, in these limited studies, genetic or pharmacological inhibition of AGT resulted in profound reductions in the development of aortic aneurysm, while not affecting the incidence of aortic dissection in mice administered BAPN. For models in which AGT deletion protected against aortic aneurysm, the mechanistic basis remains unclear.

Myocardial Dysfunction

Myocardial dysfunction is a common occurrence in patients with septic shock. A protective role of the RAS in septic shock has been suggested based on AngII’s role in maintaining normal blood pressure.⁶² However, inhibition of the RAS has also been linked to cardioprotective effects, exemplifying the complexity of the disease.^{63, 64} In a recent study,¹³ the source of AGT and its effect on cardiac function in acute sepsis has been determined using the cecal ligation mouse model. Plasma AGT concentrations were increased in mice with sepsis-induced myocardial dysfunction, accompanied by increased AGT protein in the liver and heart. In mice with hepatocyte-specific Agt deletion, cardiac contractile activity was preserved, while deletion of Agt in cardiomyocytes had no beneficial effects on cardiac dysfunction. Liver-derived AGT accumulated in the heart during sepsis, predominantly in fibroblasts that represent a relatively small portion of total cells in the heart. AGT accumulation in cardiac fibroblasts were attributable to its interaction with LRP1 that resulted in an inhibition of sarcoplasmic reticulum Ca²⁺ ATPase. These observations invoke the central role for hepatocyte-derived AGT in pathological effects of cardiac dysfunction. This publication also provided evidence for both AngII-dependent and independent effects of AGT in promoting myocardial dysfunction.

Obesity and Liver Steatosis

Administration of diets with augmented saturated fats rapidly results in body weight gain and lipid engorgement of the liver in mice. In an initial study, Agt hypomorphic mice or their wild type littermates in LDL receptor −/− background were fed a Western diet containing 42% calories from fat.³ The pronounced increase in body weight, liver weight, and liver triglyceride accumulation were attenuated markedly in AGT hypomorphic mice, compared to their wild type littermates. Diminished body weight gain was attributed primarily to lower fat mass. Despite no significant changes in plasma total cholesterol concentrations, whole body reduction of AGT led to attenuated lipid engorgement in the liver. Subsequent studies using hepatocyte-specific deletion of AGT replicated the effects of the whole body hypomorphism.⁶⁵

There were two indications that reductions in obesity and liver steatosis occurred via an Ang-II independent mechanism. One is that administration of either aliskiren (a renin inhibitor)³ or losartan (an AT1 receptor blocker),^{46, 65} at doses that cause equivalent reductions of blood pressure in hepatocyte-specific Agt deleted mice, had no discernable reductions in body weight gain, liver weight, or liver triglyceride concentrations. Inhibition of renin would have an equivalent effect of AGT deletion on diminishing synthesis of all angiotensin peptides while retaining AGT in plasma, while inhibition of AT1 receptors blocks AngII-AT1 receptor interaction and their downstream signaling.^{46, 65} A second line of evidence for AngII-independent effects was demonstrated in hepatocyte-specific AGT deleted mice in which plasma concentrations of AGT were restored by infection with adeno-associated viruses (AAVs) expressing either Agt or des(AngI)Agt. Expression of both forms of AGT resulted in equivalent restoration of body weight gain and liver steatosis.³ Since des(AngI)AGT lacks the N-terminal 10 amino acids of the full length AGT protein, its expression would not promote any AngII production. While there are now several publications demonstrating AngII-independent effects of AGT on obesity and liver steatosis, the mechanistic basis for these effects is undefined.

Subsequent studies have confirmed the effect of AGT on body weight gain and liver steatosis in normocholesterolemic mice fed fat-enriched diets.⁶⁵ Hepatocyte-specific deletion of AGT decreased sterol response element binding protein 1c and the downstream enzymes acetyl-coA carboxylase and fatty acid synthetase. This effect of AGT on fat synthesis was attributed to upstream effects of the AGT protein sequentially stimulating phospho-protein kinase B (Akt) and mammalian target of rapamycin (mTOR). The mechanism by which AGT affects liver lipid metabolism in an AngII-independent manner needs to be defined. In contrast to the effects reported in mice fed a Western diet, hepatocyte-specific deletion of AGT had no effect in mice fed a diet-enriched in 60% calories from fat.⁶⁶ These differential response may provide insights in future studies on the mechanism of how AGT regulates lipid metabolism under specific dietary conditions.

Manipulation of AGT Synthesis by Genetic and Pharmacological Approaches

A classic experimental mode of defining protein function in vivo has been to genetically delete the gene of interest. Genetic deletion of Agt was one of the early examples of this approach.^{30, 34} These pioneering studies revealed that global Agt deficient mice were viable at birth. However, there was a notably high attrition rate, with less than 50% mice surviving beyond the weaning. Those that survived had pronounced kidney defects with severe hydronephrosis and dilated cardiomyopathy.^{32, 33} The difficulties in breeding Agt deficient mice and the multiple defects for the few mice survived after weaning have resulted in a small number of publications using these mice. Subsequently, mice were developed that had a hypomorphic allele.³ These mice were developed to be floxed, but the retention of the Neo cassette in intron 2 of the gene construct presumably led to decreased translation and development of mice that had greater that 90% reduction of plasma AGT concentrations.³ Despite this profound reduction in AGT, the small amount of AGT that was synthesized enabled the birth of mice that were routinely viable into adulthood, while demonstrating overt phenotypes of AngII reduction, such as reduced systolic blood pressure.³

Agt floxed mice were first developed in 2012.² At the time, it was known that many tissues had the ability to synthesize AGT, but the functional significance of tissue-specific synthesis was unknown. Matsusaka and colleagues² demonstrated that both liver and kidney synthesize AGT, but the predominant source of AGT in both plasma and accumulating in kidney is liver. This observation was subsequently confirmed in another floxed mouse model.³ Further studies have shown that liver-specific deletion of AGT resulted in reductions in pathologies that are commonly attributed to AngII, such as elevated blood pressure as well as atherosclerosis formation in hypercholesterolemic mice.^{3, 4} Unexpectedly, these studies demonstrated profound effects of liver-derived AGT deletion on development of Western diet-induced body weight gain and liver steatosis in hypercholesterolemic mice.^{3, 46, 65} These effects are not mimicked by pharmacological inhibition of AT1 receptors, ACE, or renin, and hence suggest that liver-derived AGT exerts effects that are independent of reduction in AngII production.^{3, 46, 65}

Demonstration of liver-derived AGT’s ability to contribute to disease has facilitated interest in inhibiting AGT synthesis using ASO or small interfering RNA (siRNA). While ASO and siRNA therapies predominately target the liver as an inherent characteristic of the technology, liver-specificity can be enhanced by the addition of a triantennary N-acetylgalactosamine (GalNAc) moiety that enables hepatocytes to exclusively endocytose the drug.⁶⁷

Strategies to inhibit liver-derived AGT are in late-stage clinical development with a focus on resistant hypertension in humans. The first study to show this used a GalNAc-conjugated ASO in a study of safety, tolerability, and efficacy in healthy volunteers with either mild or uncontrolled hypertension. It was administered as a monotherapy or as an additional therapy in individuals who were also prescribed an ACE inhibitor or angiotensin receptor blocker (ARB).⁶⁸ The compound, IONIS-AGT-L_Rx, reduced plasma total AGT concentrations within 8 days of administration and sustained this effect for the duration of the study with weekly administration. Following termination of drug administration, plasma AGT concentrations remained significantly lower compared to control for over 6 weeks. This study was not powered to detect reductions in blood pressure, and although reductions were noted, they did not achieve statistical significance. Recently, an siRNA approach to lowering AGT in a phase I study of escalating doses in hypertensive patients was efficacious.⁶⁹ A single dose of GalNAc-conjugated Agt siRNA produced sustained reductions of plasma AGT concentrations for the entire 24 weeks of observation. At the highest dose of 800 mg, reductions in plasma AGT concentrations were greater than 90%. At doses greater than 200 mg, decreases in systolic blood pressure of >10 mmHg and diastolic blood pressure of >5 mmHg were observed. These two studies provide proof that reduction of AGT synthesis in liver can profoundly affect plasma AGT concentrations in a protracted manner. These initial studies also provide promising indications that the approach will have therapeutic benefits.

Overall, while it is clear that inhibition of AGT synthesis in liver can be achieved in humans, the question remains as to whether there is an advantage over currently available drugs that have well demonstrated safety profiles and are economically advantageous. These new drugs may not provide a distinct mechanistic approach to inhibiting the RAS. Instead, the enhanced therapeutic benefit may stem from the consistent inhibition sustained over an extended interval. This includes increased patient adherence to prescribed therapies. The drug can be directed exclusively to hepatocytes when conjugated to GalNAc, which may also provide advantages, such as minimization of side effects.

Summary and Perspectives

Recent advances in the ability to manipulate AGT synthesis have led to many new insights. As the unique substrate of the RAS, it is not surprising that many effects of AGT deletion are attributable to depletion of AngII. The complexity to the RAS has evolved with the discovery of an array of angiotensin peptides and receptors that have antagonistic effects (Figure 2). For example, inhibition of AGT reduces synthesis of Ang1-7 and inhibits activation of AT2 receptors, both of which could antagonize AngII activation of AT1 receptors. Although inhibition of AGT results in reductions in all of the downstream bioactive angiotensin peptides, the primary effects are consistent with those seen when either the production of AngII or the stimulation of AT1 receptors by AngII is diminished. Consistent findings support the notion that inhibition of AGT elicits mechanistically similar responses to ACE inhibition or AT1 receptor blockade in AngII-mediated cardiovascular diseases. However, the emerging ASO and siRNA therapies targeting liver Agt are likely to offer greater therapeutic efficacy because of their protracted and consistent effects, which may potentially enhance patient adherence. Inhibiting the AngII-independent effects exerted by AGT also holds the potential for additional benefits (Figure 3). It is noteworthy that inhibition of Agt by either ASO or siRNA specifically targeting hepatocytes may result in fewer adverse effects, compared to systemic inhibition of either ACE or AT1 receptors. Considering the long-lasting effects of Agt deletion, particularly with siRNA, caution should be exercised in situations where maintaining sufficient blood pressure and tissue perfusion is crucial, such as during shock and conditions leading to severe hypotension. Fortunately, a recent study has indicated that the swift administration of AngII or norepinephrine can rapidly reverse low blood pressure in rats administered with siRNA targeting liver Agt.⁷⁰

Figure 2. Angiotensinogen (AGT) serves as the only substrate for all angiotensin peptides.

The mature protein of human angiotensinogen consists of 452 amino acids (aa), undergoing cleavage by renin to produce angiotensin I (AngI) containing 10 aa, and des(AngI)AGT with 442 aa. Further cleavage of AngI generates AngII or other angiotensin peptides, each interacting with their respective receptors. Numbers in parenthesis denote residue numbering, starting from the N-terminus of human AGT protein. This figure is modified from Figure 1 published in Hypertension Research.⁸

Figure 3. Angiotensinogen (AGT) contributes to multiple cardiovascular and metabolic diseases.

AGT, primarily produced in liver, contributes to hypertension, atherosclerosis, and aortic aneurysm through angiotensin II (AngII)-mediated mechanisms. The des(AngI)AGT portion of AGT contributes to a fat-enriched diet-induced obesity and liver steatosis, independent of AngII production. AGT also contributes to the disease process of septic cardiomyopathy through both AngII-dependent and AngII-independent mechanisms.

Studies of cell-specific AGT deletion in mice strongly indicate that liver is the primary source of AGT. Furthermore, liver-derived AGT impacts many physiologic processes, even at distant locations from the liver. It is unclear how liver-derived AGT can influence biological processes in a distant organ. For example, it is unclear whether AGT is transported to the site of its effect on a disease process. For AngII-dependent effects of AGT, is also undefined whether the octapeptide is synthesized locally or at a distant locus that is transported to the tissue site of disease. Another area of AGT biology that needs further study is the AngII-independent effects. These effects have been demonstrated on angiogenesis, body weight gain, and liver steatosis.^{3, 46, 65, 71} For all these areas, there are compelling therapeutic-relevant reasons for further investigation.

Table 1.

Genetic deletion of Agt in hepatocytes or inhibition of Agt by ASO in mouse models with cardiovascular or metabolic disorders

Disease	Mouse Strain	Sex	Manipulations	Phenotype	Ref
Hypertension	hepAgt+/+ vs hepAgt−/− in C57BL/6N	Male	None	SBP ↓	2
	hepAgt+/+ vs hepAgt−/− in C57BL/6 (substrain unknown)	Male	Low or high-fat diet	SBP ↓	66
	hepAgt+/+ vs hepAgt−/− in Ldlr−/−	Male & Female	WD	SBP ↓	3, 4, 10, 12
	Ldlr −/−	Male	Agt ASO & WD	SBP ↓	3
	Ldlr −/−	Male	GalNAc Agt ASO & WD	SBP ↓	46
Atherosclerosis	hepAgt+/+ vs hepAgt−/− in Ldlr−/−	Male & Female	WD	atherosclerosis ↓	3, 4, 10, 12
	Ldlr −/−	Male	Agt ASO & WD	atherosclerosis ↓	3
	Ldlr −/−	Male	GalNAc Agt ASO & WD	atherosclerosis ↓	46
TAA	Fbn1 C1041G/+	Male	Agt ASO	TAA↓	54
Aortic dissection	C57BL/6J	Male	GalNAc Agt ASO & BAPN	aortic dissection ↔	18
	hepAgt+/+ vs hepAgt−/− in C57BL/6 (substrain unknown)	Male	BAPN	aortic dissection ↔
SMA aneurysm	SMC-specific Lrp1+/+ vs Lrp1−/−	Male	Agt ASO	SMA aneurysm ↓	61
Septic cardiomyopathy	hepAgt+/+ vs hepAgt−/− in C57BL/6 (substrain unknown)	Male	Cecal ligation and puncture	cardiac function ↓	13
	C57BL/6J	Male	Cecal ligation & puncture Control or Agt ASO	cardiac function ↓
Obesity	hepAgt+/+ vs hepAgt−/− in C57BL/6 (substrain unknown)	Male	High-fat diet	body weight gain ↔	66
	hepAgt+/+ vs hepAgt−/− in C57BL/6J or Ldlr−/−	Male	WD	body weight gain ↓	3, 65
Liver steatosis	hepAgt+/+ vs hepAgt−/− in C57BL/6J or Ldlr−/−	Male	WD	liver steatosis ↓	3, 65
	Ldlr −/−	Male	GalNAc Agt ASO & WD	liver steatosis↓	46

↓: the observed phenotype decreased by either hepatocyte-specific genetic deletion of Agt or pharmacological inhibition by Agt ASO or GalNAc Agt ASO

↔: the observed phenotype was not different between the two groups

Agt: angiotensinogen; ASO: antisense oligonucleotides; GalNAc: triantennary N-acetylgalactosamine-conjugated; MFS: Marfan syndrome, SMA: superior mesenteric artery; TAA: thoracic aortic aneurysm

WD: Western diet: 42% kcal from fat; high-fat diet: 60% kcal from fat

Highlights.

1. Angiotensinogen (AGT) serves as the exclusive substrate in the renin-angiotensin system.

2. Deleting AGT, through suppressing its synthesis in hepatocytes, improves cardiovascular functions attributed to inhibition of angiotensin II (AngII) production.

3. Inhibiting AGT synthesis in hepatocytes exerts beneficial effects on obesity and liver steatosis induced by a fat-enriched diet feeding in mouse models that are independent of AngII-mediated mechanisms.

4. The emerging therapies using antisense oligonucleotides (ASO) and small interfering RNA (siRNA) to target liver Agt hold promise for enhancing compliance and reducing adverse effects, compared to systemic approaches of inhibiting angiotensin-converting enzyme or AngII type 1 receptors.

Abbreviations

AGT

Angiotensinogen

Ang

Angiotensin

ASO

Antisense oligonucleotides

LRP1

Low-density lipoprotein related protein 1

LRP2

Low-density lipoprotein related protein 2

RAS

Renin angiotensin system