Mediterranean G6PD variant rats are protected from Angiotensin II-induced hypertension and kidney damage, but not from inflammation and arterial stiffness

Abstract

Rationale:

Epidemiological studies suggest that individuals in the Mediterranean region with deficiency of glucose-6-phosphate dehydrogenase (G6PD) are less susceptible to cardiovascular diseases. However, our knowledge regarding the effects of G6PD deficiency on pathogenesis of vascular diseases caused by factors, like angiotensin II (Ang-II), which stimulate synthesis of inflammatory cytokines and vascular inflammation, is lacking. Furthermore, to-date the effect of G6PD deficiency on vascular health has been controversial and difficult to experimentally prove due to a lack of good animal model.

Objective:

To determine the effect of Ang-II-induced hypertension (HTN) and stiffness in a rat model of the Mediterranean G6PD (G6PD^S188F) variant and in wild-type (WT) rats.

Methods and Results:

Our findings revealed that infusion of Ang-II (490 ng/kg/min) elicited less HTN and medial hypertrophy of carotid artery in G6PD^S188F than in WT rats. Additionally, Ang-II induced less glomerular and tubular damage in the kidneys – a consequence of elevated pressure – in G6PD^S188F than WT rats. However, Ang-II-induced arterial stiffness increased in G6PD^S188F and WT rats, and there were no differences between the groups. Mechanistically, we found aorta of G6PD^S188F as compared to WT rats produced less sustained contraction and less inositol-1,2,3-phosphate (IP3) and superoxide in response to Ang-II. Furthermore, aorta of G6PD^S188F as compared to WT rats expressed lower levels of phosphorylated extracellular-signal regulated kinase (ERK). Interestingly, the aorta of G6PD^S188F as compared to WT rats infused with Ang-II transcribed more (50-fold) myosin heavy chain-11 (MYH11) gene, which encodes contractile protein of smooth muscle cell (SMC), and less (2.3-fold) actin-binding Rho-activating gene, which encodes a protein that enhances SMC proliferation. A corresponding increase in MYH11 and Leiomodin-1 (LMOD1) staining was observed in arteries of Ang-II treated G6PD^S188F rats. However, G6PD deficiency did not affect the accumulation of CD45⁺ cells and transcription of genes encoding interleukin-6 and collagen-1a1 by Ang-II.

Conclusions:

The G6PD^S188F loss-of-function variant found in humans protected rats from Ang-II-induced HTN and kidney damage, but not from vascular inflammation and arterial stiffness.

Although major progress has been made in vascular medicine (physiology and biology), the incidence of vascular diseases (hypertension, arterial stiffness, coronary artery disease, peripheral vascular disease, stroke, and iatrogenic vascular lesions) remain unabated and the leading causes of death in the United States and worldwide.^{1, 2} The etiology of vascular diseases is multifactorial. It is now evident that metabolic reprogramming (increased glycolysis, pentose phosphate pathway, and lactate production) contributes to switching of vascular smooth muscle cells (VSMCs) from a differentiated (contractile) to a dedifferentiated (pro-inflammatory and -proliferative) phenotype.^{3, 4} Dedifferentiated VSMCs participate in the development of atherosclerosis and pulmonary arterial hypertension.^5–7

Glucose-6-phosphate dehydrogenase (G6PD) is the rate-limiting enzyme in the pentose phosphate pathway. G6PD deficiency is the most common enzymopathy in humans with over 400 million people suffering from different degrees of enzyme deficiency worldwide.⁸ G6PD deficiency is commonly associated with hemolytic anemia.⁸ Epidemiological evidence suggests that individuals in the Mediterranean region with a loss-of-function G6PD variant harboring a non-synonymous single nucleotide polymorphism (SNP; S188F; Type A-; G6PD^S188F; severely deficient) are less susceptible to vascular diseases.^{9, 10} Similarly, Africans and African Americans with a non-synonymous SNP (N126D; Type A+; G6PD^N126D; mildly deficient) have lower rates of coronary artery disease.¹¹ However, there are conflicting reports and theories arguing for and against the role of G6PD deficiency in the pathogenesis of vascular diseases.^9–18 Some studies suggest that G6PD-deficient individuals have a 39.6% greater chance of developing cardiovascular diseases than normal individuals.¹⁷ These differences in literature are mainly because some retrospective epidemiological studies are confounded with other non-vascular diseases and not well controlled. Moreover, epistasis and difference in epigenetics of individuals also may contribute to the observed mixed effects of G6PD on cardiovascular pathology.

Lack of good animal models has also hindered our progress to prove or disprove the epidemiological studies. To experimentally determine whether the Mediterranean SNP is protective in the vascular system we recently generated a rat with the Mediterranean SNP (G6PD^S188F) that faithfully mimics the human deficiency phenotype.¹⁹ Using this rat model, we found that G6PD^S188F variant protects rats from large artery stiffness caused by a high fat diet (a model of obesity/metabolic syndrome) or nitric oxide synthase inhibition with L-N^G-nitroarginine methyl ester.¹⁹ Nonetheless, very little is known about the effects of G6PD deficiency on severity of hypertension and pathogenesis of vascular diseases caused by other factors that elicit synthesis of inflammatory cytokines and vascular inflammation. Therefore, the goal of this study was to address the controversy of mixed morbific effects of G6PD^S188F variant or G6PD deficiency using a well-controlled and established rat model of angiotensin II (Ang-II)-induced vascular disease.

Ang-II is produced by angiotensin converting enzyme in the liver, kidneys, and endothelial cells. Ang-II regulates fluid balance, Na⁺-K⁺ balance, and vascular tone, to maintain normal blood pressure.²⁰ However, elevated endogenous Ang-II production increases blood pressure, evokes vascular inflammation and large artery stiffness, and causes kidney damage and heart failure.^20–23 Even more, infusion of Ang-II promotes contraction and growth of VSMCs resulting in hypertension and hypertrophy, respectively.²⁰ Severe hypertension and inflammation damages kidney leading to organ failure. Ang-II exerts its pathological effects in cells of blood vessels and other organs by activating G-coupled protein receptor (AT₁R) and downstream signaling pathways, transcription factors, and expression of protein coding and non-coding genes.^{20, 24} Because Ang-II regulates vascular contraction and inflammation by stimulating multiple intracellular signaling pathways, we investigated the effects of Ang-II-induced hypertension and stiffness in a G6PD^S188F variant rat model. Interestingly, we discovered Ang-II stimulated G6PD activity in the wild-type vascular tissue and the loss-of-function G6PD^S188F variant found in humans protected rats from Ang-II-induced hypertension and kidney damage, but not from vascular inflammation and arterial stiffness. Additionally, we demonstrate G6PD deficiency reduced Ang-II-elicited arterial hypertrophy, superoxide production, and phosphorylated ERK.

Materials and Methods

All data supporting the findings of this study are described below and in Online Data Supplement files. Detailed methods are available in the Online Data Supplement.

Animal models and experimental protocols:

All animal experiments were approved by the New York Medical College Animal Care and Use Committee and all procedures conformed to the guidelines from the NIH Guide for the Care and Use of Laboratory Animals. All rats were anesthetized with inhalation of Isoflurane (isoflurane, USP; 1-chloro-2,2,2-trifluoroethyl difluoromethyl ether; induced at 3% and maintained at 1.5%) and placed on a heated table. Echocardiography, hemodynamic measurements, and radiotelemetry were performed as described in the Online Data Supplement.

Statistical Analysis:

Graphs and statistical analyses were prepared with GraphPad Prism 9.2 (GraphPad Software, Inc, La Jolla, CA). Values are presented as the mean ± SEM of the number of samples (n) from different animals. One-way and Two-way ANOVA with post-hoc Sidak’s multiple comparison test was used to compare multiple groups, and unpaired Student’s t-test or Mann-Whitney test was used to compare two groups. The kidney damage was analyzed using a non-parametric Kruskal-Wallis test. Values of p < 0.05 were considered significant.

Results:

Ang-II stimulates G6PD activity in rat aorta:

Metabolism is fundamental to cell survival, function, growth, and death. It is known that carbohydrate and lipid metabolism dominates in VSMCs.²⁵ Since G6PD activity is stimulated by increases in intracellular hydrogen peroxide/oxidative stress and Ang-II increases reactive oxygen species,^{20, 26} we hypothesized Ang-II stimulates G6PD activity. As predicted, application of Ang-II (1 μM) to isolated aorta from wild type rats for 30 min (acute) and 24 hr (prolong) significantly increased G6PD activity (Fig. 1A).

Figure 1: Ang-II increased G6PD activity in the rat aorta.

(A) Application of Ang-II (10 nM) to isolated aorta in tissue bath for 30 mins. (acute) and 24 hrs. (prolong). N=5 in each group and N=5 in acute and N=6 in prolong group. (B) G6PD activity in isolated aorta from wild type (WT) and G6PD^S188F rats. G6PD activity is less in aorta of G6PD^S188F (N=5) than WT (N=5) rats. *Indicates p < 0.05 determined by Mann-Whitney test.

Ang-II increases blood pressure in wild type but not in G6PD^S188F rats:

Increased G6PD activity may either protect or worsen vascular diseases caused by elevated Ang-II. Therefore, our next objective was to determine the consequence of increased G6PD activity on the Ang-II-elicited vascular diseases in wild type (WT) rats and in recently CRISPR-Cas9 engineered rat model expressing G6PD^S188F, a loss-of-function variant found in individuals of the Mediterranean region.¹⁹ Consistent with our prior work that showed decreased G6PD activity in liver of G6PD^S188F rats,¹⁹ the Mediterranean mutation conferred less G6PD activity in aorta as compared to littermate control rats (Fig. 1B). Chronic (14 days) subcutaneous infusion of Ang-II (490 ng/kg/min) elicited higher systolic and diastolic blood pressure in anesthetized (Fig. 2A and B) and in conscious (Fig. 2C and 1D), WT rats than in G6PD^S188F variant rats.

Figure 2: Ang-II increased blood pressure in wild-type rats but not in G6PD^S188F rats.

(A, B) Ang-II (490 ng/kg/min) infusion for 14 days in rats increased systolic (SBP) and diastolic (DBP) blood pressure in wild type (WT, N=11) but not in G6PD^S188F (N=8) as compared to their respective control WT (N=15) and G6PD^S188F (N=15). Blood pressure was measured by inserting Millar Catheter (model SPR-671, tip size of 1.4F, Millar Instruments, US) in left carotid artery of anesthetized rats (with isoflurane 2%) on day 14 after Ang-II infusion. (C, D) Radiotelemetry recording and summary results of blood pressure measurement showing mean blood pressure (MBP) increased less in G6PD^S188F (N=3) than WT (N=3) rats by Ang-II infusion. A dotted line indicates baseline MBP in WT and G6PD^S188F rats before infusion of Ang-II. ***Indicate p < 0.0001 in panel A, **indicate p < 0.005 in panel B and panel D, and *indicate p < 0.05 in panel D determined by two-way ANOVA followed by Sidak’s multiple comparisons test.

Ang-II induces less renal injury in G6PD^S188F rats:

One of the consequences of high blood pressure is damage to the kidney.^{20, 22} Therefore, we performed histological analysis of the kidneys collected from control and Ang-II treated WT and G6PD^S188F rats. As expected, Ang-II infusion resulted in significant glomerular injury and extensive tubular dilation in the renal cortex of WT rats (Fig. 3A). In contrast, G6PD^S188F rats presented with a smaller degree of glomerular injury and tubular dilation after treatment with Ang-II, suggesting some renal protection against Ang-II induced kidney damage in this genotype. No differences in kidney interstitial fibrosis were found between genotypes after Ang-II infusion. Interestingly, Ang-II infusion led to similarly elevated plasma creatinine in both genotypes (Fig. 3B). Blood urea nitrogen (BUN) levels were significantly elevated in WT rats treated with Ang-II, but these levels only tended to increase in G6PD^S188F rats after Ang-II infusion (Fig. 3C).

Figure 3: Fourteen days of Ang-II infusion leads to reduced kidney histological injury in G6PD^S188F rats compared to wild type rats.

(A) Representative images of the glomerular injury present in Masson’s trichrome-stained kidney sections at 400X magnification (scale bar = 20μm) and quantification of glomerular injury and tubular damage. Twenty microscopy fields/animal were evaluated for each injury parameter. (B, C) Plasma creatinine and blood nitrogen urea (BUN) levels are significantly elevated in Ang-II treated WT rats, while only plasma creatine levels are significantly elevated in G6PD^S188F rats. *** Indicate p < 0.0001 and ** indicate p < 0.005 for kidney injury determined by non-parametric Kruskal-Wallis analysis.

Ang-II induces less constriction of cerebral arteries and carotid artery hypertrophy but not large artery stiffness and heart failure in G6PD^S188F rats:

Ang-II is a powerful vasoconstrictor, and it constricts cerebral arteries amongst other vascular beds. Therefore, we determined carotid artery peak velocity, a clinically established indicator of blood flow through cerebral vascular bed, in WT and G6PD^S188F rats before and after Ang-II infusion. Lower than normal peak velocity indicates constriction of cerebral arteries and higher than normal peak velocity is indicative of carotid artery disease.²⁷ Ang-II infusion decreased carotid artery peak velocity in WT and G6PD^S188F rats (Fig. 4A). However, reduction of peak velocity in G6PD^S188F rats was lesser than WT rats. This indicates Ang-II-induced less constriction of cerebral arteries in G6PD^S188F rats. In chronically infused Ang-II rats, since vasoconstriction is ensued by medial hypertrophy, we performed morphometric analysis on carotid arteries stained with Verhoeff-van Gieson stain for elastin laminae and calculated medial thickness-to-lumen diameter ratio, which increased less in G6PD^S188F as compared with WT rats treated with Ang-II (media-to-lumen ratio; Fig. 4B). Besides constricting arteries, Ang-II evokes arterial remodeling and stiffness. Progressive remodeling and increased stiffness or elastance (loss of Windkessel function) of large (conduit) arteries are critical risk factors for heart failure.^{28, 29} Therefore, we determined aorta-femoral pulse wave velocity (PWV), an index of large artery stiffness, before and after infusion of Ang-II, and found PWV increased equally in G6PD^S188F rats as well as in WT rats treated with Ang-II (Fig. 4C). Large artery stiffness is caused by many factors, but broken elastin laminae and increased deposition of collagen in media and adventitia surrounding arterial wall are two major factors of arterial stiffness pathology. Therefore, we performed histology by staining aorta and carotid artery for elastin (Verhoeff-van Gieson staining) and collagen (Masson’s Trichrome staining) and found intact elastin laminae and collagen deposition in the vessel wall of Ang-II treated WT and G6PD^S188F rats (Fig. 4D and E). Furthermore, since hypertension and large artery stiffness increase after load on the heart resulting in heart failure, we determined heart function in untreated controls and Ang-II treated WT and G6PD^S188F rats by inserting PV catheter in LV and recording hemodynamics. Our results suggest, Ang-II decreased stroke work (SW), cardiac output (CO), end-systolic volume (Ves), and end-diastolic volume (Ved), in WT and G6PD^S188F rats (Fig. 4F), and there was no difference in impaired heart function between WT and G6PD^S188F rats. Further, echocardiography results shown in Table 1 indicate that Ang-II increased LV free wall thickness and intra-ventricle septal wall thickness, and decreased LV volume, in systole and diastole in both genotype rats.

Figure 4: Carotid artery peak velocity and medial hypertrophy but not pulse wave velocity is reduced in Ang-II treated G6PD^S188F rats than wild type rats.

(A) Carotid artery peak velocity measurement by echocardiography decreased in wild type (WT, N=10) and G6PD^S188F (N=10) by Ang-II infusion. However, reduction was less in G6PD^S188F than WT rats. A difference in peak velocity measured on 14^th day after Ang-II (post) minus before (Pre) Ang-II infusion is shown. (B) Carotid artery medial hypertrophy measured as a medial thickness to lumen diameter ratio after 14 days of Ang-II infusion showed less ratio in arteries of G6PD^S188F than WT rats. (C) Pulse wave velocity (PWV) measured from carotid to femoral artery by echocardiography increased in WT (N=10) and G6PD^S188F (N=10) by Ang-II infusion. A difference in PWV measured on 14^th day after Ang-II (post) minus before (Pre) Ang-II infusion is shown. (D and E) No differences in Verhoeff-Van Gieson staining internal elastin laminae (dark brown in media) and Masson’s Trichrome staining collagen (dark blue in adventitia and media) are seen in the aorta and carotid artery isolated from WT and G6PD^S188F rats treated for 14 days with Ang-II. A representative micrograph or five experiments in each group is shown. (F) Heart function is impaired in WT and G6PD^S188F rats treated for 14 days with Ang-II. In WT and G6PD^S188F rats, infusion of Ang-II decreased stroke work (SW), cardiac output, end-systolic (Ves) and end-diastolic (Ved) volume, by ~29–50% of control rats. *Indicate p < 0.05 determined by unpaired Student’s t test in panel A and B. *, **, ***, ****Indicate p < 0.05, p < 0.01, p < 0.001 and p < 0.0001, respectively, determined by two-way ANOVA followed by Sidak’s multiple comparisons test in panel F.

Table 1:

AngII increased left ventricle wall thickness in wild-type (WT) and G6PDS^188F rats.

	WT	WT+AngII	G6PDS188F	G6PDS188F+AngII
IVSd (mm)	1.67±0.06	2.42±0.15**	1.64±0.07	2.38±0.16**
IVSs (mm)	2.89±0.07	3.86±0.14***	2.90±0.06	3.67±0.15**
LVPWd (mm)	1.45±0.05	1.91±0.09**	1.47±0.08	2.04±0.16*
LVPWs (mm)	2.34±0.13	2.96±0.10**	2.37±0.09	2.78±0.06**
LVIDd (mm)	8.95±0.17	7.77±0.11***	9.05±0.21	7.22±0.21***#
LVIDs (mm)	5.89±0.12	4.80±0.89****	5.79±0.22	4.43±0.23**
LVEDV (μL)	442.7±18.8	323.2±9.9***	453.9±23.4	274.7±17.8***#
LVESV (μL)	173.2±8.1	107.7±4.6***	167.4±14.6	90.46±11.2**

IVSd and IVSs: intra-ventricle septal wall in diastole and systole; LVPWd and LVPWs: Left ventricle posterior wall in diastole and systole; LVIDd and LVIDs: Left ventricle internal diameter in diastole and systole; LVEDV and LVESV: Left ventricle end-diastolic and -systolic volume.

^*P<0.05

^**P<0.005

^***P<0.0005

^****P<0.0001 vs WT and G6PD^S188F control

^#P<0.05 vs WT+AngII

Ang-II induces less sustained contraction and IP3 production in G6PD^S188F rats:

High blood pressure is a consequence of constriction of blood vessels/contraction of VSMCs. Therefore, to determine whether less contraction of VSMCs contributed to lower blood pressure observed in G6PD^S188F rats than WT rats treated with Ang-II, we performed force generation studies. Application of Ang-II (1 μM) to aortic rings of WT and G6PD^S188F rats stimulated contraction of smooth muscle. While the peak (tonic) contraction of aorta from WT and G6PD^S188F rats was not different (Fig. 5A), sustained contraction of aorta from G6PD^S188F rats as compared with WT rats was significantly less (Fig. 5B). Next, as Ang-II contracts VSMCs by activating G protein coupled receptor and increasing intracellular IP3 and Ca²⁺ levels,²⁰ we determined IP3 levels in aorta of WT and G6PD^S188F rats infused without or with Ang-II. IP3 levels were significantly less in aorta of G6PD^S188F rats as compared with WT rats (Fig. 5C and D). Furthermore, to confirm the reduced IP3 observed in aorta of G6PD^S188F rats is due to loss of G6PD activity, we measured IP3 in isolated aorta of WT rats incubated with Ang-II (1 μM) in the absence and presence of dehydroepiandrosterone (DHEA; 100 μM) and 6-aminonicotinamide (6AN; 1 mM), both inhibitors of G6PD activity. Ang-II increased IP3 and this was prevented in the presence of DHEA and 6AN (Fig. 5E). Increased IP3 stimulates IP3R on sarcoplasmic reticulum leading to the release of Ca²⁺ from IP3 stores. Elevated cytosolic Ca²⁺ facilitates myosin-actin filament interaction and contracts VSMCs.

Figure 5.

Sustained contraction and IP3 levels decreased, and phosphorylation IP3R increased, in aorta of G6PD^S188F rats and by application of G6PD inhibitors to isolated aorta. (A) Peak contraction of isolated wild type (WT; N=8) and G6PD^S188F (N=5) rat aorta normalized to KCl (100 mM; %100 KCl) increased by the application of Ang-II (1 μM). (B) Sustained contraction of isolated G6PD^S188F (N=5) rat aorta decreased more than wild type (WT; N=8). Sustained contraction was normalized to peak contraction (%) of Ang-II (1 μM). (C and D) IP3 levels measured in the aorta isolated from control and Ang-II (490 ng/kg/min) treated (for 14 days) WT (N=5) and G6PD^S188F (N=5) are shown. IP3 levels were less in the aorta of control and Ang-II treated G6PD^S188F rats as compared with WT rats. (E) IP3 levels were measured in isolated aorta treated without (control; N=10) or with Ang-II (1 μM; N=4) and G6PD inhibitors, DHEA: dehydroepiandrosterone (100 μM; N=4) and 6AN: 6-aminonicotinamide (1 mM; N=5) in tissue bath at 37°C for 30 mins. Application of DHEA and 6-AN decreased IP3 increased by Ang-II. *, **Indicates p < 0.05 and p < 0.01, respectively, determined by unpaired Student’s t test in panels C and D, and by two-way ANOVA followed by Sidak’s multiple comparisons test in panels B, and by one-way ANOVA followed by Sidak’s multiple comparisons test in panels E, and F.

Expression of genes encoding MYOCD and MYH11 increases, and ABRA decreases, in aorta of G6PD^S188F rats treated with Ang-II:

Expression of VSMC restricted proteins is a determinant of smooth muscle cell phenotype and function.³⁰ Therefore, we performed qPCR to determine expression of VSMC-related genes in aorta of WT and G6PD^S188F rats infused with Ang-II. Consistent with our previous study, G6PD^S188F as compared with WT rats expressing higher myocardin (Myocd) and myosin heavy chain 11 (Myh11),¹⁹ we found higher expression of Myocd (>3-fold) and Myh11 (>40-fold) genes in G6PD^S188F rats as compared with WT rats infused with Ang-II (Fig. 6A and B). Myocd and Myh11 genes are markers of a differentiated SMC encode protein that participate in contractile function of smooth muscle. Interestingly, expression of actin-binding rho activating (Abra), which encodes protein that stimulates serum response factor-dependent transcription pathway and proliferation of VSMC,³¹ increased (>5-fold) in aorta of WT rats but not G6PD^S188F rats treated with Ang-II (Fig. 6C). Immunofluorescence imaging of aorta revealed qualitatively more MYH11 in aorta (media) of G6PD^S188F as compared with WT rats infused with Ang-II (Fig. 6D).

Figure 6.

Expression of genes encoding smooth muscle cell enriched proteins are altered in aorta of wild type vs. G6PD^S188F rats by Ang-II. (A, B and C) Expression of genes encoding Myocd and Myh11 smooth muscle cell enriched proteins increased in aorta of G6PD^S188F rats as compared with wild type rats (WT) treated with Ang-II. Conversely, expression Abra, a smooth muscle cell enriched gene that encodes a protein that promotes cell proliferation, increased in aorta of WT but not G6PD^S188F rats treated with Ang-II. Gene expression was determined by real-time PCR and are expressed as relative to Tuba1a in WT and G6PD^S188F control (N=5) and Ang-II treated (N=5) rats. (G) A representative micrograph of five immunohistochemistry experiments showing more fluorescence MYH11 (red) signals. Nucleus is stained with DAPI (blue), and “A” indicates adventitial side in each micrograph. *, **, ****Indicate p < 0.05; p < 0.01; p < 0.0001, respectively, determined by two-way ANOVA followed by Sidak’s multiple comparisons test in panels A to F.

Ang-II increases expression of genes encoding IL6, CCL5, and COL1a1 proteins in aorta of wild type and G6PD^S188F rats:

Increased inflammation contributes to arterial hypertrophy and stiffness.^{23, 32} As Ang-II is known to activate transcription of Il-6 in vascular tissue,³³ we determined expression of genes encoding inflammatory protein in aorta of rats infused with and without Ang-II. Interestingly, Il-6 and Ccl5 genes increased (>1000-fold) in aorta of WT and G6PD^S188F rats infused with Ang-II (Fig. 7A and B). After we found Ang-II activated transcription of genes encoding pro-inflammatory cytokines, which potentially inflames smooth muscle and activates fibroblasts, we presumed this resulted in the deposition of collagen observed in adventitial and medial layer of aorta and carotid artery from WT and G6PD^S188F rats infused with Ang-II (Fig. 4E). In this regard, expression of Col1a1 gene, a smooth muscle matrix protein that makes arteries stiffer, markedly increased in aorta of WT and G6PD^S188F rats infused with Ang-II (Fig. 7C). Concurrently, CD45⁺ leukocytes, immunogenic and pro-inflammatory cells, accumulated in adventitial and medial layer of aorta of WT and G6PD^S188F rats infused with Ang-II (Fig. 7D). Interestingly, aorta counter stained with LMOD1, another VSMC differentiation marker,³⁴ increased in medial layer of aorta from G6PD^S188F rats as compared with WT rats infused with Ang-II (Fig. 7D). Neither expression of Il-6 nor accumulation of CD45⁺ cells was different in aorta of G6PD^S188F rats from that of WT rats infused with Ang-II.

Figure 7.

Expression of genes encoding inflammatory cytokines are increased in aorta of wild type and G6PD^S188F rats by Ang-II. (A, B, and C) Expression of genes encoding IL-6, CCL5, and COL1a1increased in aorta of WT and G6PD^S188F rats treated with Ang-II. Expression of Il-6 increased even more in G6PD^S188F than WT rats. Gene expression was determined by real-time PCR and are expressed as relative to Tuba1a in WT and G6PD^S188F control (N=5) and Ang-II treated (N=5) rats. A representative micrograph of five immunohistochemistry experiments showing: (D) increased CD45 (shown by arrow heads and by a box) and counter stained with LMOD1 (red), in aorta of G6PD^S188F and WT rats treated with Ang-II. Nucleus is stained with DAPI (blue), and “A” indicates adventitial side in each micrograph. *, **, ****Indicate p < 0.05; p < 0.01; p < 0.0001, respectively, determined by two-way ANOVA followed by Sidak’s multiple comparisons test in panels A to F.

Aorta of G6PD^S188F produces less superoxide and Ang-II increases extra-mitochondrial superoxide production in the aorta of wild type rats:

Ang-II activated AT₁R signaling stimulates NAD(P)H oxidases and generation of superoxide which inactivates nitric oxide signaling in vascular tissue and the heart resulting in severe vasoconstriction and impaired heart function.²⁰ We and others have shown G6PD-derived NADPH fuels NAD(P)H oxidases to produce superoxide.³⁵ However, other than G6PD-mediated superoxide production, nothing else is known about the mechanism of G6PD-derived NADPH driven superoxide generation. Hence, we performed immunoprecipitation experiments to determine whether G6PD and NADPH oxidase form a complex in the aorta. Interestingly, we found G6PD co-immunoprecipitated with p47^phox (Fig. 8A top panel) and gp91^phox (Fig. 8A bottom panel), but not with p67^phox (Fig. 8A middle panel), subunits of NADPH oxidases in aorta of WT rats. Silencing of G6PD reduced DHE staining in cells of medial layer of the aorta (Fig. 8B and C). Similarly, we detected less superoxide levels in aorta of G6PD^S188F as compared with WT rats (Fig. 8D). Next, to determine whether Ang-II increases extra-mitochondrial or mitochondrial superoxide, we measured DHE and MitoSox levels by HPLC as published previously and found DHE, but not MitoSox, levels increased in aorta applied with Ang-II (Fig. 8E).

Figure 8.

G6PD silencing and G6PD^S188F variant decreased superoxide, phosphorylated-ERK. (A) Immunoprecipitation (IP) with IgG and anti-G6PD antibody (G6PD) followed immunoblotting (IB) indicate G6PD co-immunoprecipitated with NADPH oxidase subunit p47^phox and gp91^phox but not p67^phox. Input is whole cell homogenate. (B and C) A representative micrograph and summary results of immunoflourescence of dihydroethidium (DHE) staining detecting superoxide signals in sections of un-transfected control (UT) and transfected with scrambled or non-targeting (NT) or G6PD-specific (siG6PD) siRNA aorta show decreased DHE fluorescence in sections of aorta transfected siG6PD. N=5 in each group. (D) Lucigenin (5 μM) chemiluminescence signals detecting superoxide levels decreased in aorta of G6PD^S188F (N=6) as compared with wild type (WT, N=6) rats. (E and F) Application of Ang-II to isolated aorta in tissue bath at 37°C for 60 mins increased DHE (top panel) but not MitoSox (bottom panel) as determined by HPLC. (F) A representative immunoblots, and summary results of densitometry, show increased pERK-to-tERK in aorta of wild type (WT) but not G6PD^S188F rats treated with Ang-II as compared to respective untreated controls N=5 in each group. *, **Indicate p < 0.05 and p < 0.01, respectively, determined by unpaired Student’s t test in panels C, D, and E, and by Mann-Whitney test in panels F.

Ang-II increases phosphorylated-ERK in aorta of wild type but not G6PD^S188F rats:

Increased reactive oxygen species mediate Ang-II-stimulated ERK signaling in endothelial cells and VSMCs.^{20, 36} Therefore, we determined phosphorylation of ERK in aorta of WT and G6PD^S188F rats infused with and without Ang-II. Our results indicate higher pERK-to-tERK ratio in aorta of WT but not G6PD^S188F rat treated with Ang-II (Fig. 8F).

Ang-II increases aldosterone in plasma of G6PD^S188F rats:

Ang-II stimulates aldosterone production, which induces G6PD deficiency-like phenotype in endothelial cells.³⁷ Elevated aldosterone impairs endothelial function, increases blood pressure, and elicits cell proliferation.³⁷ Therefore, we measured aldosterone levels in the blood of WT and G6PD^S188F rats treated with Ang-II. Blood aldosterone levels increased in the blood of G6PD^S188F (control: 148±40 and Ang-II: 2078±290 pg/ml; P<0.001; n=5 in control and n=6 in Ang-II) and WT (control: 155±57 and Ang-II: 1748±501 pg/ml; P<0.01; n=5 in control and n=5 in Ang-II) rats.

Discussion

In the current study, the salient findings are: 1) Ang-II increased G6PD activity, 2) rats expressing loss-of-function G6PD^S188F variant, found in individuals of Mediterranean region, developed less Ang-II-induced hypertension, arterial hypertrophy, and kidney damage, and 3) the Mediterranean G6PD^S188F variant rats are not protected from Ang-II-induced inflammation and large artery stiffness that can lead to heart failure.

G6PD is a major source of NADPH that maintains redox homeostasis in the cell. Since NADPH is a co-factor for glutathione and thioredoxin reductases,²⁶ G6PD is considered as an antioxidant. The central dogma has been that the reduction of G6PD activity increases H₂O₂ – reactive oxygen species or oxidative stress – in the cell. However, recent studies suggest that G6PD-derived NADPH fuels NADPH oxidases to generate superoxide in smooth muscle,³⁵ heart,³⁸ liver,^{39, 40} and adipocytes.^{41, 42} Furthermore, increased expression and activity of G6PD has been associated with increased oxidative stress in hypertension,^{19, 43, 44} heart failure,³⁸ and diabetes.⁴² However, whether the G6PD deficiency is detrimental or beneficial for cardiovascular health remains a controversial topic.⁴⁵

Due to lack of animal models that mimic human enzymopathies, it has been difficult to experimentally prove, in a controlled manner, whether G6PD deficiency is or is not a determinant of vascular pathologies. In the present study, we demonstrated G6PD^S188F variant rats – a novel model that mimics G6PD enzymopathy found in humans¹⁹ – as compared with WT rats, developed less Ang-II-induced hypertension, carotid artery medial hypertrophy, and kidney damage. However, G6PD^S188F variant failed to confer protection against Ang-II-induced large artery stiffness, LV hypertrophy, and heart failure. This implies that Ang-II-stimulated G6PD activity perhaps selectively contributed to the pathogenesis of hypertension and vascular remodeling but not large artery stiffness. These findings were somewhat unexpected because we recently found G6PD^S188F variant rats develop less arterial stiffness in high fat diet feeding model.¹⁹ We attribute these differences of arterial stiffness to stimulation of different signaling pathways by Ang-II vs high fat diet in different cell-types of blood vessel wall. Recent experimental and population-based studies indicate that large artery stiffness precedes development of hypertension.⁴⁶ However, our findings suggest Ang-II-induced stiffness may not be directly connected to hypertension in G6PD^S188F rats.

Ang-II induces vascular pathologies through multiple different signaling pathways that promote contraction of VSMCs and inflammation of cells in different organs including blood vessels.²⁰ Ang-II stimulates mobilization of hematopoietic stem cells as well as leukocytes and fibroblasts from the bone marrow to the adventitia, and accumulation of inflammatory cells in adventitia triggers synthesis and deposition of collagen in the adventitial and medial layer making the artery stiffer.³² On those lines, we also observed Ang-II stimulated accumulation of CD45⁺ leukocytes in adventitia of WT and G6PD^S188F rats. Interestingly, CD45⁺ cell accumulation in perivascular region of G6PD^S188F rats was not different from WT rats. Also, Ang-II evokes gene transcription and stimulates expression of pro-inflammatory cytokines in endothelial cells and VSMCs.²⁰ Ang-II is reported to stimulate expression of IL-6 and other cytokines through NF-κB, ERK/MAPK, and long non-coding RNA, signaling pathways.^{20, 24, 47} Ang-II infusion increased Il-6, Ccl5, and Col1a1 more than 100-fold in aorta of WT and G6PD^S188F rats. Further, in the aorta of G6PD^S188F variant rats, Ang-II increased Il-6 expression even more than WT rats. As p-ERK levels, but not Il-6 expression, decreased in aorta of G6PD^S188F vs WT rats, this indicates that G6PD may not be involved in activation of Il-6 expression elicited by long non-coding RNA (Giver) or by other ERK-independent pathways. Furthermore, it is likely that SMCs, which were presumably in differentiated state as indicated by increased Myocd and Myh11 gene expression, did not produce Il-6 but instead Ang-II augmented Il-6 production from endothelial cells/adventitial fibroblasts/infiltrated leukocytes in aorta of G6PD^S188F variant rats. In this regard, others have shown that G6PD deficiency promotes lipopolysaccharide-induced expression of pro-inflammatory cytokines in leukocytes and endothelial cells.^{15, 48} Additionally, increased aldosterone in WT (by 11-fold) and G6PD^S188F (by 14-fold) rats infused with Ang-II potentially activated aldosterone-dependent signaling pathways that may have promoted SMC calcification and stiffness. Therefore, we suggest G6PD deficiency does not mitigate the Ang-II induced inflammation and inflammation-driven arterial stiffness.

Next, we sought to determine the potential mechanisms through which G6PD^S188F variant reduced hypertension and arterial hypertrophy. Binding of Ang-II to AT₁R, a G-coupled protein receptor, increases IP3, a second messenger, that increases intracellular Ca²⁺.²⁰ Elevated Ca²⁺ in the cytosol contracts VSMC and contributes to the development of hypertension. Therefore, less IP3 in the aorta of G6PD^S188F variant than WT rats treated without or with Ang-II, and decreased IP3 by application of DHEA and 6AN to aorta of WT rats, indicate G6PD deficiency interfered and interrupted IP3 signaling stimulated by Ang-II. Moreover, vascular tone (Fig. 3B) and IP3 (Fig. 3C–D) results taken together suggest G6PD^S188F accelerated Ang-II-induced tachyphylaxis of IP3 receptors. Since inhibition and silencing of G6PD increases protein kinase G (PKG) activity that inactivates phospholipase Cβ3 and reduces IP3,^{49, 50} we propose activation of PKG possibly decreased IP3 in aorta of G6PD^S188F rats. Decreased IP3 potentially reduced the release of Ca²⁺ into the cytoplasm and relaxed smooth muscle resulting in vasodilation. Concomitantly, aorta isolated from G6PD^S188F variant rats as compared with WT rats generated less sustained contraction evoked by Ang-II, supporting the notion that dampened IP3 signaling in smooth muscle of G6PD^S188F variant rats, at least partly, contributed to reduce Ang-II-induced vasoconstriction. Likewise, carotid artery peak velocity decreased more in WT than G6PD^S188F rats after Ang-II infusion. Lower than normal carotid artery peak velocity indicates increased constriction or narrowing of cerebral arteries.²⁷ Therefore, lesser reduction in carotid artery peak velocity of G6PD^S188F, as compared to WT, rats suggest less severe constriction/resistance of cerebral circulation in G6PD^S188F rats in response to Ang-II. Since G6PD^S188F mutation hyperpolarizes the membrane potential and reduces Ca²⁺ influx-induced contraction of aorta and carotid artery, this implies Ca²⁺ influx and decreased IP3-induced Ca²⁺ release collectively blocked Ang-II-induced vasoconstriction and hypertension.

In addition, aorta of G6PD^S188F, as compared with WT, rats expressed more Myocd and Myh11 genes as well as MYH11 and LMOD1 proteins (markers of VSMC differentiation).^{34, 51} Higher expression of gene encoding MYOCD, a transcription co-activator of SMC-restricted genes, and MYH11, a selective SMC protein, suggest G6PD^S188F variant pushed SMCs more towards to the differentiated phenotype in aorta of rats infused with Ang-II. Consequentially, reducing the dedifferentiated SMCs, which contributes to the medial thickening/hypertrophic remodeling of the vessel wall, presumably prevented narrowing of aorta and carotid artery (including other resistance arteries) reduced resistance to blood flow in G6PD^S188F rats. Similarly, G6PD-deficient mice, which were generated by chemical mutagenesis with fewer characteristics of human SNPs, are less susceptible to Ang II-induced hypertension and SMC hypertrophy and atherosclerosis.^{43, 52} Therefore, it is reasonable to suggest that more differentiated SMCs in vessel wall of G6PD^S188F rats and G6PD-deficient mice reduced medial hypertrophy potentially decreased vascular resistance and improved blood flow.

Activation of AT₁R signaling stimulates NADPH oxidases and increases generation of superoxide levels, which inactivates nitric oxide and augments VSMC contraction.^{20, 53} We suggest that increased extra-mitochondrial superoxide potentially contributed to the vasopressor effects of Ang-II. In this study, we demonstrate that G6PD directly interacts with p47^phox and gp91^phox, and that silencing G6PD expression decreased superoxide in the arteries of WT rats. Along those lines, aorta of G6PD^S188F, as compared with WT, rats generated less superoxide. In addition, since elevated reactive oxygen species increase p-ERK,³⁶ we suggest decreased superoxide in the aorta of G6PD^S188F rats prevented ERK activation. Stimulation of ERK pathway increases cell proliferation and vasoconstriction.²⁰ Besides activating ERK pathway, Ang-II rapidly stimulates phosphorylation of HDAC5 and activates myocyte enhancer factor-2 (MEF-2) dependent gene transcription.⁵⁴ In this regard, we recently reported silencing G6PD expression and inhibiting G6PD activity: 1) decreases HDAC5 expression and HDAC activity, 2) increases H3K9ac expression in whole cell homogenate, and 3) enriches H3K27ac (mark of open chromatin) on Myocd promoter leading to increases in Myocd and Myh11 transcription.⁵⁵ Consistently, our findings indicated that aorta of G6PD^S188F, but not WT, rats treated with Ang-II expressed higher expression of Myocd and Myh11 transcripts and MYH11 protein. Furthermore, since G6PD^S188F variant prevented Ang-II-stimulated expression of a MEF-2 targeted gene encoding actin-binding rho-activating (ABRA) protein³¹ – that stimulates the serum response factor-dependent transcription pathway and augments proliferation of VSMC – we propose down regulated HDAC5/MEF-2/Abra and decreased p-ERK, at least partly, reduced cell proliferation and medial hypertrophy in arteries of Ang-II treated G6PD^S188F rats.

In addition to reducing hypertension, G6PD^S188F rats had less renal injury than WT rats. Glomerular and tubular damage was less pronounced in kidneys of G6PD^S188F than WT rats in response to Ang-II. Although BUN and creatinine levels in blood of G6PD^S188F rats were not statistically different than WT rats infused with Ang-II, BUN and creatinine were 50% and 20% of WT, respectively. These results suggest kidney function of G6PD^S188F rats may be better than WT rats. Ang-II-dependent aldosterone increase promotes sodium retention and causes hypokalemia that increases blood pressure. Interestingly, our findings indicate Ang-II increased aldosterone but not blood pressure in G6PD^S188F rats. These results potentially suggest G6PD deficiency blocked both Ang-II- and aldosterone-dependent signaling pathways that causes kidney dysfunction and elevates blood pressure. Since high-pressure elicits juxtamedullary injury in kidneys of rats treated with Ang-II,²² our results suggest lower hypertension consequentially reduced renal injury in G6PD^S188F rats.

G6PD deficiency abrogates αB-crystallin mutation (R120GCryAB)-associated cardiomyopathy but does not confer protection from transverse aortic constriction- and ischemia/reperfusion-induced heart failure.^{56, 57} Similarly, G6PD^S188F variant did not improve heart function impaired by infusion of Ang-II. Since increased arterial stiffness is a direct cause or a risk factor for impaired heart function/failure,^{23, 29} we propose unmitigated arterial stiffness, but not hypertension-induced overload, potentially decreased heart function in G6PD^S188F variant rats treated with Ang-II.

In summary, there are two potential mechanisms associated with G6PD deficiency: 1) a protective mechanism wherein decreased IP3 and ERK signaling prevented or reduced Ang-II-induced hypertension and hypertrophy, respectively, resulting in less pressure-induced kidney damage; and 2) a detrimental mechanism that augmented expression of pro-inflammatory cytokines and matrix proteins, contributing to Ang-II stimulated inflammation-driven large artery stiffness and impaired heart function. Thus, we propose that G6PD^S188F selectively protects rats from some vascular pathology, but not all. These differences could be a potential basis for confounding results in the literature regarding effects of G6PD deficiency, including Mediterranean variants, on human cardiovascular health. Therefore, more controlled prospective epidemiological studies are warranted to ferret out the beneficial effects of G6PD mutations prevailing in multiple ethnic groups worldwide and use the outcome to develop future better pharmacotherapies to treat vascular diseases.