Treprostinil reduces blood pressure and aortic inflammation in hypertension

Abstract

Prostaglandin I₂ (PGI₂) signaling is vasoprotective. Further, PGI₂ signaling exhibits immunomodulatory properties that largely promote an anti-inflammatory state. Inflammation and immune activity are associated with the development of hypertension. However, it remains unknown whether exogeneous PGI₂ can restrain the immune and inflammatory responses in the context of hypertension. A phenome-wide association study evaluated single-nucleotide polymorphisms (SNPs) in the receptor for PGI₂, IP (gene name PTGIR), and the associated odds of developing cardiovascular pathologies. C57Bl/6J mice underwent one of two models to evaluate the effect of an exogenous PGI₂ analog on aortic inflammation and hypertension. Mice were either administered the PGI₂ analog, treprostinil (TPL), at the initiation of angiotensin II (Ang II) infusion to determine the effect of TPL on the development of inflammation and hypertension, or mice were administered TPL after 2 weeks of Ang II to determine the ability of TPL to reduce blood pressure and inflammation in established hypertension. Humans who were heterozygous for a SNP in IP had a significantly greater odds ratio for several vascular pathologies compared to controls. Mice that received TPL at the onset of Ang II infusion were protected from developing hypertension and exhibited reduced aortic inflammation. Mice that received TPL two weeks after initiation of Ang II infusion exhibited a significant reduction in their blood pressure, decreased aortic inflammation, diminished aortic fibrosis, and fewer splenic Th1 cells compared to vehicle treatment. Exogeneous PGI₂ signaling protects against the development and/or maintenance of Ang II-induced hypertension, possibly through inhibition of Th1 cell function.

Introduction:

Hypertension is an important risk factor for and cause of cardiovascular disease morbidity and mortality. Importantly, the prevalence of hypertension is increasing as it currently effects 45% of US adults and greater than 70% of the elderly (1, 2). These numbers, while specific to the US, are increasing worldwide, making hypertension a global health concern (3).

An important contributor to the development and maintenance of hypertension is inflammation. Specifically, studies over the last 20–30 years have demonstrated that antigen presenting cells, such as dendritic cells, become activated in hypertension. These activated dendritic cells promote the polarization of T cells to pro-inflammatory Th1, characterized by expression of the transcription factor Tbet, and Th17, characterized by the expression of the transcription factor RORyt, phenotypes (4–6). These Th1 and Th17 cells release pro-inflammatory cytokines such as IFN-γ and IL-17A, respectively, which promote vascular dysfunction and renal injury(7, 8). T regulatory (Treg) cells, are protective in hypertension and are characterized by their expression of the transcription factor Foxp3; however, Treg numbers are often reduced or are functionally impaired in hypertension, enabling disease progression (9–12).

Prostaglandin I₂ or PGI₂ is a metabolite of arachidonic acid produced through the cyclooxygenase pathway(13–15). PGI₂ signals through the G-protein coupled receptor termed IP, or gene name PTGIR (14). PGI₂ is a potent vasodilator that has known vasoprotective effects (16, 17). PGI₂ has also been shown to exhibit immunomodulatory effects. Specifically, PGI₂ signaling reduced cytokine production from CD4+ Th1 and Th2 cells and also impaired pro-inflammatory dendritic cell function (18, 19). Further, our group recently demonstrated that PGI₂ signaling promoted Treg functions in models of allergy and asthma and also demonstrated that exogenous administration of PGI₂ analogs in patients with pulmonary arterial hypertension promoted the development of Tregs (20, 21). Several PGI₂ analogs exist and are FDA approved for the treatment of pulmonary arterial hypertension. One PGI₂ analog is treprostinil and was one of the therapeutics given to patients in our report which examined Treg development after PGI₂ analog administration therapeutically for pulmonary arterial hypertension (20).

Studies have demonstrated that mice deficient in IP (IP knockout (KO)), develop hypertension, cardiac hypertrophy, and severe cardiac fibrosis in response to salt feeding (22). Further, intravenous injection of PGI₂ into anesthetized rats with hypertension acutely reduced blood pressure and acutely increased plasma renin activity (23). However, to date, no one has evaluated the role and effect of PGI₂ analogs on the immune response promoting hypertension. Thus, we hypothesized that exogenous PGI₂ administration would blunt immune activation and vascular infiltration in systemic hypertension. We further hypothesized that this would be associated with sustained reductions in blood pressure in addition to any vasodilatory effects and demonstrate its potential use as a therapy for hypertension. Herein, we show that single nucleotide polymorphisms (SNP) in PTGIR (IP) in humans are associated with vascular and hypertensive pathologies, suggesting that a defect in IP signaling can contribute to disease pathogenesis. We also show that exogeneous administration of treprostinil in the setting of the angiotensin II (Ang II) model of hypertension, prevented the development of sustained elevated blood pressure and also reduced immune infiltration into the fat surrounding thoracic aorta. Further, we found that administration of treprostinil into mice with Ang II-induced established hypertension resulted in a blunting of blood pressure and a reduction in immune cells infiltrating the fat surrounding the thoracic aorta. Moreover, treprostinil reduced the percentage of Th1 cells while increasing the Treg:Th1 ratio, demonstrating direct immunomodulatory effects. Finally, treprostinil also prevented further development of aortic fibrosis when infused into mice with Ang II-induced established hypertension. We are the first to show that exogeneous PGI₂ signaling ameliorates sustained hypertension and associated aortic inflammation and fibrosis.

Methods:

Animals, Ang II infusion, and blood pressure measurement:

Animal experiments were conducted with IACUC approval at either Vanderbilt University Medical Center (Protocol number M2100025) or Indiana University School of Medicine (Protocol number 22067). Wild-type (C57Bl/6J) were either purchased from Jackson Laboratories, Bar Harbor, ME or were acquired from the In Vivo Therapeutics Core at the Indiana University School of Medicine. Male mice approximately 10 to 12 weeks of age were used for this study. Mice were anesthetized with isoflurane, and 4-week osmotic mini-pumps (ALZET) containing either Ang II (490 ng/kg/minute, Sigma Aldrich) or vehicle (0.08 mol/L sodium chloride/1% acetic acid solution) were implanted subcutaneously. At the time of Ang II pump insertion or two-weeks after Ang II pump insertion, a second 4-week osmotic mini-pump containing treprostinil (45ng/kg/min, commercial name – Remodulin, United Therapeutics) or its associated placebo (United Therapeutics) was also implanted. BP was measured non-invasively by tail-cuff plethysmography as previously described(24, 25). Prior to initiation of the study mice were trained for two days prior to baseline measurements of blood pressure that were recorded to enable acclimatization to the instrument. Blood pressure was then measured two days per week and averaged for the weekly measurement. At the end of the experiment, mice were euthanized with Fatal Plus (Patterson Veterinary Supply, Inc.). Following euthanasia, mice were perfused with cold saline until complete clearance of blood from circulation. Hearts, thoracic aortae, kidneys, and spleens were extracted and either fixed in 10% Formalin or placed into 5% FBS in 1640 RPMI (Corning) for further processing.

Flow cytometry of splenic and aortic leukocytes:

Single-cell suspensions of the thoracic aorta with surrounding perivascular fat were prepared as previously described(24). Spleens were ground directly through a 70μM filter. Aortic homogenates were ground through a 70μM filter after digestion (RPMI 1640 media containing 5% FBS, 1mg/mL collagenase A, 1mg/mL collagenase B (Sigma)). Single-cell suspensions were stained for flow cytometry. Aortic and splenic suspensions were blocked with FC Shield (Cytek Biosciences). Aortic suspensions were stained using the following antibodies/stains: Aqua LIVE/DEAD Fixable Violet Dead Cell Stain (1:2000, Life Technologies), CD45 (1:200, PE-Cy5, 30-F11, BioLegend), CD3 (1: 200, PerCP-Cyanine 5.5, 145–2C11, Invitrogen), MHCII (1:1000, BV421, M5/114.15.2, BioLegend), CD11c (1:200, APC-eFluor 780, N418, Invitrogen) and F4/80 (1:200 BV711, T45–2342, BD Biosciences). Splenic Suspensions were stained using the following antibodies/stains: Aqua LIVE/DEAD Fixable Violet Dead Cell Stain (Life Technologies, CD45 (1:2000, PE-Cy5, 30-F11, BioLegend), CD3 (1:200, Pacific Blue, 145–2C11, BioLegend), CD4 (1:200, BV605, GK1.5, BioLegend), CD8 (1:1000, PE-Dazzle 594, 53–6.7, BioLegend), CD25 (1:200, Alexa Fluor 488, PC61.5, Invitrogen). Splenic samples were also stained using the following intracellular markers: Tbet (1:50, ebio4B10), Foxp3 (1:50, PE-Cy7, FJK-16s, Invitrogen), RORyt (1:50, BV786, Q31–378, BD Biosciences), IFNγ (1:50, PE, XMG1.2, BioLegend), IL-17A (1:50, Alexa Fluor 700, TC11–18H10.1, BioLegend). Splenic and Aortic suspensions were fixed with the eBioscience Foxp3/Transcription Factor Staining Buffer Set (Invitrogen). Splenic samples were counted prior to staining with a hemocytometer. A known quantity of counting beads (Count Bright Plus Absolute Counting Beads, Invitrogen) were added to each aortic sample prior to analysis. Samples were run on a Cytek Aurora and analyzed using FlowJo software (FlowJo, LLC). Aortic samples were normalized using the bead count and expressed as number of cells per thoracic aorta. Population percentages are displayed for splenic samples, ratios of percents were calculated as well and displayed.

Histology of hearts, thoracic aorta, and kidney:

Hearts, thoracic aorta, and kidneys were fixed in 10% formalin at 4°C for 5 days and then moved to 70% ethanol in cassettes. Cassettes were processed, embedded in paraffin blocks and sectioned at 5μm thickness at the histology core laboratory at Indiana University School of Medicine. Tissue sections were deparaffinized as follows: slides were washed 3 times in xylene for 4 minutes each; followed by two washes in 100% ETOH for 4 minutes each. This was followed by sequential washes in 95% ETOH; 70% ETOH, and 50% ETOH for 1 minute each. Slides were then placed in DIH₂O for 5 minutes. Slides were stained using the Masson Trichrome Kit (Epredia). Slides were first placed in Bouin’s solution and microwaved for 45 seconds and then left to stand for 6 minutes. Afterward, slides were rinsed under running DIH₂O for 5 minutes. Fresh Weigert’s iron hematoxylin solution was prepared by mixing part A and B in 1:1 ratio. Slides were incubated in the solution for 8 minutes then rinsed under running DIH₂O for 10 minutes. Nest, slides were then placed in Bieberich Scarlet Acid Fuchsin solution for 5 minutes and rinsed under running DIH₂O for 1 minute. For 5 minutes, slides were incubated in phosphotungstic-phosphomolybdic acid solution then moved to aniline blue stain solution for another 5 minutes. Finally, slides were placed in 1% acetic acid for one minute. After staining, slides were rinsed under running DIH₂O for 2 minutes, dehydrated with two washes in 100% ETOH for 2 minutes each; followed by three washes in xylene for 3 minutes each. Slides were then cover slipped using Cytoseal 60 mountant (Epredia). Slides were scanned and ImageJ software (NIH) was used to calculate percent fibrosis expressed as the fibrotic area over the total tissue area for hearts, thoracic aorta, and kidney. Assessment of renal injury was evaluated in the cortex on Masson’s Trichome Stained tissues. Histological changes were quantified by counting the following percentages of tubules and glomeruli that showed loss of brush border, tubule dilation, absence of Bowman’s space, thickening of the basement membrane, and increased mesangial matrix respectively. The quantitative percentage is as follows. 0 = none, 1 = <10%, 2 = 11–25%, 3 = 26–45%, 4 = 46–75%, and 5 = >76%. Histological analyses were performed in a blinded manner.

Variant Analysis:

BioVu at Vanderbilt University Medical Center was utilized for this study. The data in BioVu are all deidentified. We conducted a PheWAS using the rare coding variants in the PTGIR. Our preliminary cohort consisted of 35,480 adult subjects from BioVU who were genotyped on the Exome BeadChip by Illumina, created to capture functional exonic variants. This cohort predominately included individuals with electronic health record reported non-Hispanic White race, 29,795 (84%) and female gender 19,604 (55.3%). From this population we restricted analysis to the 29,712 subjects identified as having genetically determined European ancestry based on principal component analysis. The Exome BeadChip directly genotyped 11 PTGIR (IP) coding variants, nine of which had at least one heterozygote. These nine variants were either missense or nonsense variants with an allele frequency less than 0.5%. Owing to the infrequent nature of rare coding variants in this population we were underpowered for analyses stratified by biological sex, however, all regression analyses were adjusted for age, sex, and a study specific identifier. The study specific identifier was used because the exome chip population included samples from various studies, therefore it was vital to adjust for any potential biases that may have existed due to the initial sample collection. We conducted a phenome-wide association study (PheWAS) using standard methodology adjusting for the prespecified covariates and required at least two unique entry dates for cases. Phecodes version 1.2 was used for this analysis and each phecode represents an aggregate grouping of International classifiers of Disease billing codes version 9 and 10 designed to represent a clinical disease. We used a gene-based burden test to collectively analyze these rare variants. In total, we tested 1362 phenotypes, of which 52 associations achieved a nominally significant p-value of < 0.05. The SNPs evaluated in the PheWAS are available in Supplementary Table 1 and the full results report is available in Supplementary Table 2.

Statistics:

2–4 independent experiments were conducted per figure. Data are represented as mean ± SEM. Data are analyzed by 2-way ANOVA followed by a Tukey post hoc test to correct for multiple comparisons in GraphPad Prism 10. Blood pressure data are analyzed by 2-way ANOVA with repeated measures followed by a Holm-Sidak post hoc test to correct for multiple comparisons in GraphPad Prism 10. Fisher’s exact test followed by a per-SNP Bonferroni’s correction was used to evaluate variant association. Data was considered significant at P < 0.05.

Results:

PTGIR missense or nonsense variants are associated with vascular and hypertension associated pathologies in humans.

PGI₂ is a known vasodilator and has vascular protective capabilities. Thus, given our interest in PGI₂ signaling and its immunomodulatory role in hypertension, we sought to determine if polymorphisms in the receptor for PGI₂, IP, in humans (gene name PTGIR) were associated with hypertension, hypertension-associated pathologies, or vascular pathologies. Utilizing Vanderbilt University’s biobank, BioVu (26), we examined associations between rare genetic variants that were in the coding region of PTGIR with different disease phenotypes, effectively conducting a gene Phenome Wide Association Study (PheWAS). 9 variants in PTGIR had at least one heterozygote in our ancestry dataset, the variants included are listed in Supplementary Table 1. The variant with the most heterozygotes is PTGIR p.P226T, which is a missense mutation located in the third intracellular loop of IP, a region known to interact with the heterotrimeric G proteins (27). We found that these rare PTGIR variants were significantly associated with several vascular and hypertensive related pathologies including hypertensive chronic kidney disease, ventricular fibrillation and flutter, unstable angina, heart valve disorders, and embolism and thrombosis of the abdominal aorta (Table 1). These results demonstrate that mutations in the PGI₂ receptor may be a risk factor for vascular pathologies and hypertension.

Table 1:

Gene based burden test for association. Table of phecodes related to vascular function and hypertension associated with PTGIR variants.

gene	phecode	phecode_string	cases	controls	Heterozygote cases	OR	P
PTGIR	401.22	Hypertensive Chronic Kidney Disease	1737	14362	23	1.86	0.0297
	427.41	Ventricular Fibrillation and Flutter	110	18026	5	5.45	0.000474
	411.1	Unstable Angina	987	21791	15	2	0.0204
	395	Heart Valve Disorders	2509	24494	32	1.49	0.043
	444.2	Embolism and thrombosis of abdominal aorta	54	23810	2	1.23	0.04797

Treprostinil is protective in an Ang II-induced mouse model of hypertension.

Our PheWAS data demonstrate that SNPs in IP (PTGIR), the receptor for PGI₂, are risk factors for vascular and hypertension-associated pathologies in humans. Based on this data in combination with our previous work demonstrating that PGI₂ is anti-inflammatory and promotes the formation of Treg in pulmonary arterial hypertension, we hypothesized that IP signaling protects against the development of hypertension and associated aortic inflammation. To test this hypothesis, we implanted osmotic mini pumps containing Ang II or vehicle into mice at the same time as a second pump was implanted containing the PGI₂ analog treprostnil (TPL) or its associated placebo. This experimental design is shown in Figure 1A. Prior to implanting pumps, mice were trained, and baseline blood pressure was measured via tail-cuff plethysmography. Blood pressure was monitored weekly until euthanasia on day 28. We found that mice infused with both Ang II and treprostinil were protected from the development of hypertension compared to mice challenged with Ang II and treated with treprostinil vehicle. At 4 weeks, blood pressure in mice infused with Ang II and treprostinil were significantly reduced compared to mice infused with Ang II and treprostinil vehicle (Figure 1B). We next examined the immune cells that infiltrated the perivascular fat surrounding the thoracic aorta. We found that mice infused with treprostinil and Ang II had significantly reduced numbers of CD3+ T cells, MHC II+ CD11c+ dendritic cells, and F4/80+ monocytes and macrophages surrounding their aortae than mice infused with Ang II and treprostinil vehicle (Figure 1C–F). Upstream gates are shown in Supplementary Figure 1. Together, these data demonstrate that administration of treprostinil at the onset of infusion of a hypertensive stimulus protects the mice against the development of sustained elevated blood pressure and associated aortic inflammation.

Figure 1:

Infusion of treprostinil (TPL) alongside angiotensin II blocks the development of hypertension and associated aortic inflammation. A) Experimental model B) Tail cuff blood pressure measurement (n=3–5) C) Representative flow cytometric analysis of CD3+ T cells, MHC II+CD11c+ DCs, and F4/80+ monocytes/macrophages. Total number of D) CD3+ T cells E) MHC II+ CD11c+ DCs and F) F4/80+ monocytes/macrophages (n=3–4).. Data are represented as mean ± SEM. *P<0.05, ****P<0.0001

Treprostinil blunts blood pressure and reduces aortic inflammation in established hypertension.

To determine if IP (PTGIR), the receptor for PGI₂, signaling reduces blood pressure and associated inflammation in established hypertension and thus may serve as a viable therapeutic for uncontrolled systemic hypertension, we infused mice with Ang II or its vehicle for two weeks via osmotic mini-pump and then inserted a second osmotic mini-pump at day 14 containing treprostinil (PGI₂ analog) or its vehicle. Mice were euthanized on day 28. The experimental design is shown in Figure 2A. For this experiment, we also trained mice and measured baseline blood pressure and then blood pressure weekly using tail-cuff plethysmography. We found that administration of treprostinil exogenously in mice that were hypertensive blunted their blood pressure. Specifically, mice infused with treprostinil secondary to Ang II infusion had significantly lower blood pressure at 4 weeks compared to mice infused with treprostinil vehicle secondary to Ang II (Figure 2B). We next looked at infiltration of immune cells around the aorta to assess vascular inflammation. We found that mice infused with treprostinil two weeks after initiation of Ang II infusion had significantly reduced CD3+ T cells, MHCII+CD11c+ dendritic cells, and F4/80+ monocytes and macrophages compared to mice that received treprostinil vehicle two weeks after initiation of Ang II infusion (Figure 2C–F). Further, we examined both CD3+CD4+ and CD3+CD8+ T cell subpopulations within the aorta. We found that both CD3+CD4+ and CD3+CD8+ T cell subpopulations were significantly reduced in mice infused with treprostinil two weeks after initiation of Ang II infusion (Figure 2G–I). Together, these data demonstrate that exogenous PGI₂ acts to reduce blood pressure and attenuate associated aortic inflammation in established hypertension.

Figure 2:

Infusion of treprostinil after establishment of hypertension blunts blood pressure and reduces associated aortic inflammation. A) Experimental model B) Tail cuff blood pressure measurement (n=12) C) Representative flow cytometric analysis of CD3+ T cells, MHC ll+CD11c+ DCs, and F4/80+ monocytes/macrophages. Total number of D) CD3+ T cells E) MHC II+ CD11 c+ DCs and F) F4/80+ monocytes/macrophages (n=12). G) Representative flow cytometric analysis of CD4+ T cells and CD8+ T cells. Total number of H) CD3+CD4+ T cells and I) CD3+CD8+ T cells (n=7–9). Data are represented as mean ± SEM. *P<0.05, **P<0.01

Treprostinil reduces Th1 cells in established hypertension.

To determine whether treprostinil (PGI₂ analog) reduced inflammation by altering CD3+CD4+ T cell subtypes when administered secondary to Ang II, we again employed the experimental model shown in Figure 2A. We then isolated spleens at day 28 and examined the presence of Tbet+ Th1 cells, RORyt+ Th17 cells, and Foxp3+CD25+ Treg cells. We found a significantly reduced percentage of Tbet+ Th1 cells of total CD4+ T cells in the spleens of mice infused with treprostinil two weeks after initiation of Ang II infusion compared to mice infused with treprostinil vehicle two weeks after initiation of Ang II infusion (Figure 3A,B). Th17 and Treg cells remained unchanged (Figure 3A, C–D). Upstream gates are shown in Supplementary Figure 2. Further, we evaluated IFNγ and IL-17A expression in splenic CD4+ T cells to determine if the cytokine production of the Th1 or Th17 T cell subsets was reduced. We found that the percentage of IFNγ+CD4+ T cells, but not IL-17A+CD4+ T cells, was reduced in mice infused with treprostinil two weeks after initiation of Ang II infusion compared to mice infused with treprostinil vehicle two weeks after Ang II infusion (Supplementary Figure 3). Upstream gates are also shown in Supplementary Figure 2. The observed reduction in IFNγ+CD4+ T cells in mice infused with treprostinil two weeks after initiation of Ang II corroborates the inferred effect of treprostinil on Th1 cells identified through Tbet staining. Further, the lack of a significant reduction in IL-17A+CD4+ T cells in mice infused with treprostinil two weeks after initiation of Ang II is consistent with the absence of effect of treprostinil on Th17 cells identified through RORyt staining. Although we did not see differences in Treg percentage, we evaluated both Treg:Th17 (Figure 3E) and Treg:Th1 (Figure 3F) ratios which can be indicative of changes in Treg suppressive functionality. Fewer Tregs and an imbalance of Tregs and proinflammatory T cells subsets is associated with the pathogenesis of hypertension(12). While the Treg:Th17 was not different (Figure 3E), we found a greater Treg:Th1 ratio in mice that received treprostinil two weeks after initiation of Ang II infusion compared to mice that received treprostinil vehicle two weeks after initiation of Ang II (Figure 3F). Together, these data demonstrate that IP (PTGIR), the receptor for PGI₂, signaling is promoting an anti-inflammatory environment in the setting of Ang II through reduction of Th1 cells and is associated with an increase in Treg:Th1 ratios.

Figure 3:

Infusion of treprostinil after establishment of hypertension reduces Th1 cells. A) Representative flow cytometric analysis of Th subsets B) Percentage of CD4+Tbet+ Th1 C) Percentage of CD4+RORyt+ Th17 D) Percentage of CD4+Foxp3+CD25+ Treg E) Treg:Th17 ratio F) Treg:Th1 ratio (n=7–17). Data are represented as mean ± SEM. *P<0.05

Treprostinil (or IP signaling) blunts the development of aortic fibrosis in established hypertension.

To determine if IP signaling (PTGIR), the receptor for PGI₂ was also protective against Ang II-induced end-organ damage, we evaluated thoracic aortae and hearts for fibrosis via Masson’s trichrome staining after mice underwent the protocol in Figure 2A. We found that mice infused with treprostinil two weeks after initiation of Ang II infusion had significantly reduced aortic fibrosis compared to mice infused with treprostinil vehicle (Figure 4A–B). Interestingly, we found no significant difference in cardiac fibrosis between treprostinil treated mice and those receiving vehicle (Figure 4C–D). Further, we found a trend toward a decrease in kidney fibrosis but an increase in loss of tubular architectural integrity in mice that received treprostinil following Ang II infusion compared to mice that received Ang II and treprostinil vehicle (Supplementary Figure 4A–C). Together, these data suggest that exogeneous IP signaling is protective against the development of vascular fibrosis in established hypertension.

Figure 4:

Infusion of treprostinil after establishment of hypertension blunts aortic fibrosis. A) Representative Masson’s Trichrome stained aortic slices for each group. B) Quantification of aortic fibrosis (n=10–20) C) Representative Masson’s Trichrome stained hearts for each group D) Quantification of cardiac fibrosis (n=9–15). Data are represented as mean ± SEM. *P<0.05, ***P<0.001

Discussion:

In this study, we evaluated the effects of exogenous IP (PTGIR), the receptor for PGI₂, signaling on blood pressure, end-organ fibrosis, and inflammation using Ang II-induced experimental hypertension models in mice. We demonstrate, for the first time, that exogenous IP signaling restrains the cellular inflammatory response in an experimental model of mice with established hypertension, identifying this pathway as a target for development of potential therapeutics to mitigate the inflammatory component of hypertension. Further, we connected SNPs in IP in humans with a greater risk of development of vascular or hypertension-associated pathologies. This finding may have important precision medicine implications. Those individuals that harbor one of these SNPs may not respond as well, or at all, to PGI₂ analog therapy due to the SNPs resulting in missense or nonsense mutations in the PTGIR receptor that may disrupt its ability to adequately signal. The list of diseases associated with the PTGIR polymorphisms (Table 1) is interesting. In particular, hypertensive chronic kidney disease, a condition due to damage to the kidneys over time from elevated blood pressure, and not hypertension itself, is identified. Hypertension affects many patients with chronic kidney disease but the relationship between the two is quite complex, as hypertension can cause kidney injury leading to chronic kidney disease with the reverse also being true, development of chronic kidney disease can also lead to hypertension(28). Of note, hypertension itself was not associated with PTGIR variants in this study, this is likely due to the high throughput phenotyping method that is used by PheWAS. In cases where common phenotypes, like hypertension, are examined for association with rare variants, noise from the assay can lead to false negatives. This could be occurring in our study due to how rare PTGIR variants are in the population. It is further possible that hypertensive chronic kidney disease, being a more severe hypertensive phenotype, is what drove that association in our study. Future studies would also involve leveraging additional datasets to derive a larger cohort of individuals with PTGIR variants to test association with hypertension and to also evaluate sex-specific, and age-specific differences, all current limitations of this study.

The ability of treprostinil (PGI₂ analog) to completely prevent the maintenance of hypertension in the Ang II-induced model, when administered alongside Ang II at its initiation was interesting. However, this finding is not necessarily surprising. For one, as mentioned in the introduction, a previous study has demonstrated that loss of IP (PTGIR), the PGI₂ receptor, in mice leaves them susceptible to salt-sensitive hypertension and associated cardiac fibrosis (22). Importantly, Ang II is a potent vasoconstrictor while PGI₂ is a potent vasodilator. These effects likely counteract one another. Specifically, PGI₂ counteracts the effects of thromboxane A2, a vasoconstrictor and prothrombotic mediator that contributes to the development of Ang II-induced hypertension (29). Further, the interactions between PGI₂ and the renin angiotensin system (RAS) pathway have been extensively studied. In the kidney, prostaglandin generation is increased by Ang II specifically due to increases in cyclooxygenase II (COX-2) in the kidney as a counterregulatory mechanism (30). Increases in COX-2, an enzyme responsible for processing arachidonic acid, lead to increases in PGI₂ in specific tissues that express PGI₂ synthase (PTGIS). As a part of this pathway, PGI₂ stimulates renin release which leads to aldosterone release and sodium reabsorption. This is counter to the sodium release that would be needed to lower blood pressure via pressure natriuresis. However, PGI₂ acts as a potent vasodilator, important for maintaining glomerular filtration rate (GFR), increasing renal blood flow, and promoting release of sodium. These other effects of PGI₂ could be negating the PGI₂-induced increases in renin release in our model to exert a renoprotective effect and help to lower blood pressure (30, 31). Another important vasodilator, nitric oxide (NO), often acts in parallel with PGI₂ to exert its protective roles. However, NO can stimulate both COX-1 and COX-2 to produce prostaglandins including PGI₂ (32). Nonetheless, bioavailability of NO is reduced in Ang II-induced models of hypertension due to associated elevated levels of oxidative stress (33). Regardless, the complex interactions between PGI₂ and both the RAS and NO pathways are important to consider when attempting to understand how physiologically treprostinil is exerting its effects in our model of Ang II-induced hypertension. It is also worth considering that PGI₂ is preventing immune cell activation in this model in favor of a sustained state of tolerance to Ang II. Further, it is worth noting that what also may be occurring is a lack of an immune response in this model due to the offset of Ang II by PGI₂.

While the kidney was not the immediate focus of our study, we found a paradoxical trend toward a decrease in renal fibrosis and an increase in loss of architectural integrity in the cortex in mice that received treprostinil (PGI₂ analog) following induction of hypertension through Ang II infusion. The trend toward a decrease in fibrosis is in line with previously published data demonstrating PGI₂ to be renoprotective and protect the kidneys from fibrosis in models of renal fibrosis (34). The increase in loss of architectural integrity and definition of tubular borders was initially surprising; however, we believe we may be detecting the early stages of kidney repair in our scoring method, as renal epithelial cells can dedifferentiate and lose some architectural integrity during the initial stages of repair following injury (35). Thus, we hypothesize that administration of treprostinil following Ang II may prevent further renal damage and increase renal blood flow allowing for initiation of repair processes. Further studies are needed to better understand the renal effects of treprostinil in our model. In line with PGI₂ having a renoprotective effect, a previous kidney-focused study evaluated the ability of beraprost sodium (an orally active PGI₂ analog) to mitigate renal injury in Dahl salt-sensitive rats, a model of salt-induced renal injury. In this study, the Dahl salt-sensitive rats were given beraprost sodium in the drinking water upon initiation of high salt diet to trigger the model. Ultimately beraprost sodium administration was sufficient to prevent proteinuria and renal injury in the Dahl salt-sensitive rats. Further, beraprost sodium was also able to prevent the upregulation in expression of several pro-inflammatory cytokines in the kidney (36). This study did not evaluate the ability of beraprost sodium to reduce established effects of high salt on the Dahl-salt sensitive rat.

The model in which treprostinil was effectively administered as a therapeutic for established hypertension is important and holds the promise as an avenue of exploration for treatment targets. Hypertension is termed the “silent killer”. Up to 40% of US adults are unaware they have hypertension and of those who are aware they have the disease, greater than 50% do not have their blood pressure completely under control (37). There are many potential reasons for this, one being lack of treatment of the underlying inflammatory component that’s driving the disease(8). Potential therapeutic targets in the PGI₂ signaling cascade are interesting and worth evaluating as they could exhibit a dual effect, as shown by our data and published reports, they could modulate the immune and inflammatory response in individuals with hypertension and push the milieu toward a more anti-inflammatory state and also reduce blood pressure additively through direct vasodilation.

A previous study directly examined PGI₂ in the setting of hypertension. However, the study design and goals differed substantially from ours. It is important to note that Scholkens et al. used rats that developed spontaneous and chronic renal hypertension and their experiments were acute in nature(23). Specifically, PGI₂ was injected into anesthetized rats and outcomes were assessed at the latest 45 minutes after injection(23). In contrast, our study aimed to directly evaluate the effect of exogenous PGI₂ on cellular inflammation associated with hypertension as well as evaluate it as a potential therapeutic in mice over an extended period of time, mimicking a clinically relevant treatment scenario in humans.

Perspectives:

Our results demonstrate an important role for exogenous PGI₂ in mitigating cellular inflammation associated with hypertension and lowering blood pressure, specifically through the modulation of Th1 cell populations. This highlights the potential ability to repurpose FDA approved PGI₂ analogs or identify mediators in the PGI₂ signaling pathway as therapeutic targets to treat systemic hypertension through reduction in inflammation.

Clinical Perspectives:

1. This study directly investigates a potential therapeutic for the reduction of inflammation associated with hypertension, an area not treated by current therapeutics.

2. This study finds that exogeneous PGI₂ mitigates inflammation and reduces blood pressure in established hypertension and that SNPs in PTGIR in humans are associated with increased risk for the development of certain vascular and hypertensive-related pathologies.

3. Clinically, PGI₂ analogs are currently FDA approved for the treatment of pulmonary arterial hypertension; these would quickly be available for essential hypertension. Of note, systemic use of PGI₂ analogs for the treatment of essential hypertension will require monitoring of blood pressure and bleeding-risk in patients.

Data Availability Statement:

Summary statistics are provided as a supplemental table in this manuscript. Additional data requests can be made to the authors and will be shared pending institutional approval from Vanderbilt University Medical Center. Scripts for the PheWAS analysis and phecodes are available for download from GitHub (https://github.com/PheWAS/PheWAS/tree/master/R).