IFN-γ Contributes to the Immune Mechanisms of Hypertension

Abstract

Hypertension is the leading cause of cardiovascular disease and the primary risk factor for mortality worldwide. For more than half a century, researchers have demonstrated that immunity plays an important role in the development of hypertension; however, the precise mechanisms are still under investigation. The current body of knowledge indicates that proinflammatory cytokines may play an important role in contributing to immune-related pathogenesis of hypertension. Interferon gamma (IFN-γ), in particular, as an important cytokine that modulates immune responses, has been recently identified as a critical regulator of blood pressure by several groups, including us. In this review, we focus on exploring the role of IFN-γ in contributing to the pathogenesis of hypertension, outlining the various immune producers of this cytokine and described signaling mechanisms involved. We demonstrate a key role for IFN-γ in hypertension through global knockout studies and related downstream signaling pathways that IFN-γ production from CD8⁺ T cell (CD8T) in the kidney promoting CD8T-stimulated salt retention via renal tubule cells, thereby exacerbating hypertension. We discuss potential activators of these T cells described by the current literature and relay a novel hypothesis for activation.

Hypertension, a major cause of premature death, affects roughly half of the US population and 1.3 billion people worldwide (1,2). This condition has been subjected to intensive study for more than half a century leading to the development of a variety of classes of pharmaceutical options to lower BP; however, fewer than 50% of patients achieve good BP control (3). Further confounding clinical treatment of hypertension is the multifactorial dysfunctional aspect of this disease, wherein an estimated 90%–95% of treated patients present with unclear origin (essential hypertension) (4), and an estimated 20%–30% of cases are resistant to currently available treatment (resistant hypertension) (5). Thus, it is important to investigate unknown mechanisms causing essential hypertension and develop new therapeutic strategies against hypertension (6). Years of intense research have confirmed that the kidney plays a critical role in regulating BP (7), for example transplanting kidneys from hypertensive donors to normotensive recipients transfers hypertension (8). Guyton and others proposed that a physiologic defect in the kidney impairs its salt handling, which contributes to the development of hypertension (9,10). However, the exact identity of this kidney defect is still under investigation. One potential suspect being investigated recently is disorder of immunity, in particular adaptive immune cells, which contribute to hypertension through several possible mechanisms (11), including dysregulation of natriuresis and driving renal injury (12).

Published in the 1970s, Svendsen first demonstrated the role of the immune system in the classic deoxycorticosterone acetate (DOCA)+salt murine model of hypertension (13). In this experiment, nude mice (lacking thymus, immunodeficient) exhibited blunted BP increase compared with haired wild-type (WT) mice, along with fewer round cell infiltrations. When these nude mice received a thymus graft—thereby restoring mature immune cell production—the BP increase to DOCA+salt was restored to normal levels. This study identified a thymus-independent phase of BP elevation followed by a thymus-dependent phase (13). The particular immune cells driving the thymus-dependent phase of BP elevation had not been identified and became of interest to researchers. In 2007, Harrison and colleagues not only corroborated blunted BP elevation due to immunodeficiency through the use of RAG1^−/− mice (immunodeficient mice that produce no mature T or B cells) but also found that the adoptive transfer of T cells—but not B cells—would restore the hypertensive response to angiotensin II (AngII) (14). This landmark study was limited, wherein other labs found RAG1^−/− mice no longer showed blunted hypertensive response to AngII in later studies (15,16), and a discussion of this change and potential explanations were provided in an excellent editorial from Madhur et al. (17). Many other preclinical studies indicating the role of the immune system have been published in the last 20 years. A few highlights include the use of the immunosuppressant tacrolimus to ameliorate BP elevation in Dahl salt-sensitive rats fed a high salt diet (12) and the use of the immunosuppressants dexamethasone and etanercept to reduce renal fibrosis, albuminuria, T-cell infiltration, and NF-κB activation accompanying AngII-infused double-transgenic rats (18). As researchers began to characterize particular immune players involved, Kamat et al. demonstrated the critical roles of IFN-γ in AngII-mediated hypertension, wherein global knockout (KO) of the cytokine blunted BP elevation to this pressor (19). Known producers of IFN-γ include T cells, natural killer (NK) cells, monocytes/macrophages, neutrophils, and dendritic cells (DCs), and each of these immune cells have been implied as participating in the process of hypertension (11,20–25). Certain T cells (Th17) and macrophages have been specifically implicated to play a role through sodium-driven proinflammatory responses as described here (26); however, the purpose of this review is to highlight current research describing the proposed immune cells contributing to this “thymus-dependent” phase of BP elevation and outline the elucidated mechanisms, and provide insight into the particular role of IFN-γ in the immune cell-mediated stage of hypertensive development. For a comprehensive network of IFN-γ JAK/STAT signaling pathways, we refer the reader to this 2018 review by Bhat et al. (27), and these reviews for a discussion of IFN-γ transcription regulation (28,29).

T Cells

The specific role of the T cell, and not B cell, in contributing to hypertension was first outlined by Guzik et al. in 2007, wherein adoptive transfer of T cells but not B cells to RAG-1^−/− immunodeficient mice restored the blunted BP elevation of this model to AngII treatment. In that study, the authors also determined that NADPH oxidase contributed to complete development of hypertension via affecting T-cell infiltration (14). NAPDH oxidases (NOX)—particularly NOX2—mediate release of reactive oxygen species (ROS) alongside cellular mitochondria in a process termed “oxidative burst” to aid in elimination of invading micro-organisms (30,31). The role of NADPH oxidases and T cell–aggravated hypertension has been connected to the central nervous system in an AngII model of hypertension (32,33). Of interest, IFN-γ has been linked to upregulation of NAPDH oxidase in exposed epithelial cells (34,35) to concomitant upregulation of NADPH and IFN-γ in leukocytes in sickle cell disease (36), and to CD8⁺ enhanced ROS signaling through an NADPH oxidase-dependent CD39 expression mechanism (37,38). From these studies, and others, NAPDH oxidase activity within the immune cell has been tied to IFN-γ production and plays a role in cellular response to IFN-γ.

CD4⁺ Th1 T cells may play a role in the development of kidney damage in hypertension but not necessarily elevated BP itself as indicated by T-bet KO mice showing reduced renal damage but similarly elevated BP to WT mice in the AngII model of hypertension (39,40), Sun et al. further indicated the MR receptor on CD4⁺ T cells contributes to hypertension through regulation of IFN-γ (41). Our laboratory has found that CD8⁺ T cells (CD8Ts) play a direct role in the development of hypertension. After induction of hypertension in the DOCA+salt model (Cat. M-121, 50 mg/pellet in 21-day release formula and 1% NaCl drinking water) for 14–18 days, adoptive transfer of 1 × 10⁷ splenocyte-derived CD8Ts from these mice into uninephrectomized C57BL/6J male mice results in salt-sensitive hypertension in the recipient mice that can be alleviated through treatment with a thiazide diuretic (42). We found that these T cells remain within the kidney, directly interacting with the distal tubule and continuing to promote NCC expression and activity despite treatment with the diuretic (42). More recently, we found that this mechanism of hypertensive-derived CD8Ts interacting with the distal tubule thereby generating salt-sensitivity in the recipient mice is via the IFN-γ-programmed cell death ligand 1 (PDL1) pathway (43). IFN-γ KO mice demonstrate blunted BP in either the AngII model of hypertension (19) or DOCA+salt model (43). In performing adoptive transfer of DOCA+salt-induced hypertensive CD8Ts from WT or IFN-γ KO mice to the alternative strain (methods according to previously described laboratory protocols) (42,43), only the WT CD8Ts were able to generate salt-sensitive hypertension in the recipient mouse (Figure 1), indicating that IFN-γ from the CD8T itself is sufficient to drive salt-sensitive hypertension.

Figure 1.

Interferon -γ contributes to CD8T-mediated salt-sensitive hypertension. (A) Radiotelemetry recording of systolic BP in recipient wild-type (C57/B6) mice (blue) and IFN-γ knockout (B6 background) mice (red) after adoptive transfer of 1×10⁷ CD8Ts freshly isolated from deoxycorticosterone acetate+salt-treated IFN-γ knockout or wild-type mice. All mice were purchased from the Jackson Laboratory. Data were recorded every 15 minutes and averaged to six time points a day; n=5–6 mice per group. Statistical analysis was performed by two-way ANOVA. Significance for interaction effect: P<0.001; time effect: P<0.001; and strain recipient: P=0.02. (B) At the end point of BP recording in (A), immunohistochemistry staining (DAB) of kidney sections was performed using CD3 specific monoclonal antibody after the protocol we have published previously (42,43). In the quantification bar graph, each dot shown is the mean of four images taken per sectioned kidney. **P<0.05.

Activation of T cells—characterized by elevated expression of IFN-γ and TNF-α among other cytokines—has been linked to the development of salt-sensitive hypertension (44). In agreement with clinical data finding increased human hypertensive CD8T IFN-γ production compared with normotensive CD8Ts (45), we found that hypertensive murine CD8Ts demonstrated enhanced capacity of producing IFN-γ—but not TNF-α—compared with sham mouse–derived CD8Ts (43). This evidence, the ability of immunosuppressants to reduce BP in animal models of hypertension, and our co-culture model showing activated CD8Ts promoted greater NCC upregulation, sodium retention, and PDL1 expression through IFN-γ and the IFN-γ receptor in mouse distal convoluted tubular cells compared with naïve CD8Ts further supports the hypothesis that activated CD8Ts contribute to the development of hypertension. When we knocked down PDL1 within the renal tubule, BP elevation was blunted in both the DOCA+salt and CD8T adoptive transfer models of hypertension (43). Renal specific targeting of PDL1 may prove a promising clinical target in treating resistant hypertension.

During full activation of naïve CD8Ts, clonal expansion takes place after presentation of antigen, co-stimulatory factor, and a third signal (such as IL-12) wherein T-cell clones that are specific for individual antigens proliferate (46–48). Trott et al. performed a clonotype analysis of kidney CD8⁺ T-cell receptor sequences and found 3522±1049 unique T-cell receptor sequences in mice treated with AngII with three shared specific clonotypes in Vβ3, 8.1, and 17 families from four of the five mice that were not present in multiple sham mice; however, these clonal subtypes were not observed in other organs (48). The conclusion of this study suggested that rather than a single clonal population, an oligoclonal population of CD8Ts accumulate in the kidney and contribute to hypertension and sodium retention (48). The presence of more than one clonal population of CD8Ts presents at least two potential explanations: either (1) multiple unique antigens are being presented simultaneously leading to CD8T accumulation in the kidney through a IFN-γ-dependent mechanism or (2) activation of these CD8Ts is occurring via nonantigen-specific mechanisms, leading to IFN-γ production.

Some antigen-specific activation evidence includes this study by Rudemiller et al., wherein knocking down CD247 (responsible for coupling antigen recognition to intracellular transduction) (49,50) in Dahl salt-sensitive rats led to reduced T-cell infiltration into the kidneys and reduced BP (51). Blocking antigen presentation by DCs resulted in blunted BP elevation in two murine models of hypertension (52). The study by Hevia et al. found that ablation of CD11c⁺ antigen presenting cells prevented hypertension in an AngII+ high salt diet murine model (53). A few proposed antigens include isolevuglandin protein adducts (54,55), heat shock protein 70 (56,57), and Toll-like receptor (TLR) 4 or 2 activators such as C-reactive protein (58), uric acid (59), and others (60).

Experimental evidence suggests that such molecules may individually or collectively contribute to the role of T cells in hypertension; however, the presence of more than 3000 unique T-cell receptor sequences in AngII-induced hypertensive mice indicates that clonal expansion may not be occurring due to antigen presentation and recognition. Under certain conditions such as lymphopenia (61) or exposure to both IL-6 or IL-21 and IL-17 or IL-15 (62–65), CD8Ts can be activated without direct antigen presentation (66). IL-7 has also been implicated in the activation of autoimmune CD8Ts (62). In like manner, stimulation of T cells with phorbol myristate acetate (PMA) and the calcium ionophore ionomycin can bypass the T-cell membrane receptor complex and activate the T cell, leading to elevated expression of TNF-α and IFN-γ (67,68). Further research is needed to identify alternative pathways that may contribute to the activation of CD8Ts resulting in their infiltration within the kidney and stimulation of sodium retention.

NK Cells

Group 1 innate lymphocytes, NK cells arise from the same family as T and B cells (69) and have been implicated in both AngII-mediated hypertension and pulmonary hypertension. Depleting NK1.1 cells in WT C57BL/6 mice before AngII treatment resulted in blunted vascular dysfunction. Through a mechanism involving IFN-γ and T-bet, Kossmann et al. described a mutual activation pathway involving IL-12-secreting monocytes stimulating NK cells, leading to increased IFN-γ production and AngII-mediated hypertension (70). This relationship between monocyte and NK cell differentiation has been elucidated to involve TXb21 and IL-15R in a tumor environment (71). IL-12 driving NK production of IFN-γ has been described for some time (72). In a rat model of pulmonary hypertension, NK cells have been proposed to play a protective role rather than pathologic (73); as such, the role of NK cells likely involves environment-specific driven regulation. Such tissue specific regulation has not yet been fully elucidated.

Monocytes and Macrophages

In addition to their role in promoting NK differentiation, monocytes (in particular macrophages) have been described as contributing to the pathogenesis of hypertension. An excellent review was provided by Rucker and Crowley; as such, we refer the reader to this publication for a thorough description of the contribution of macrophages to hypertension (74). Macrophages have been discovered to infiltrate and remain within the renal interstitium in the AngII model of hypertension, correlating with elevated TGF-β and monocyte chemotactic protein (MCP), even after cessation of AngII (75). LysM⁺ monocytes have been described to contribute to—and be increased by—AngII-driven arterial hypertension through an experiment wherein depletion of LysM⁺ myelomonocytic cells reduced AngII-induced BP elevation and blunted vascular dysfunction, nitric oxide bioactivity, and vascular oxidative stress. Reconstitution with WT monocytes but not neutrophils restored the AngII-mediated disease phenotype (76).

Due to the ability of immunosuppressants to, at minimum, reduce renal damage elicited in several animal models of hypertension and the correlation between macrophage infiltration and inflammation in models of hypertension (77,78), it would follow that at least a connection can be established between proinflammatory macrophages and tissue inflammation and damage in hypertension. Such a relationship was discussed in the excellent review by Rodriguez-Iturbe et al. (79) in reference to the studies by Bravo et al. (80) and Quiroz et al. (81), wherein macrophage numbers and oxidative stress were correlated in two different animal models of hypertension. The excellent review provided by Rucker and Crowley described the elucidated role of TNF-α from macrophages driving renal inflammation (74); as such, we will focus on the relationship between IFN-γ and macrophages in hypertension.

In addition to aiding in the transition from innate immunity to adaptive immunity, IFN-γ can influence its surroundings directly through actions such as local dilation of blood vessels allowing for immune cells to localize at sites of inflammation (21). However, of particular note is the effect of IFN-γ on macrophages. After release by other immune cells such as NK cells or T cells, IFN-γ stimulates macrophages and primes them for response by inducing a proinflammatory activation and stimulating the release of other cytokines causing the M1 phenotype seen in macrophages (82,83). The M1 phenotype name was derived to match Th1 release of IFN-γ; however, various cytokines and stimuli can affect the activation state of macrophages. Because stimulation with a combination of factors or high salt can shift the activation spectrum of macrophages, the classification of M1 and M2 phenotypes may become insufficient to distinguish the activation status (82,83). Generally, this classic activation by T cells through IFN-γ leads to macrophage production of inflammatory response genes and cytokines such as IL-12, IL-23, and NO generating the proinflammatory macrophage type (82,84). Macrophages also have the ability to secrete IFN-γ to the same extent as T cells via stimulation with certain cytokines such as through simultaneous activation by IL-12 and IL-18 (24), leading to the potential of generating IFN-γ in an autocrine manner (24,85). Accordingly, macrophage activation by IFN-γ is multifaceted in the source and the activation state on the basis of other factors and cytokines present.

Serum IFN-γ levels have been clinically shown to be a predictor of high systolic BP (86). With diastolic BP, there was not only significance with IFN-γ but also MCP-1 levels—after corrections for variables such as age and sex (86). Within models of hypertension with IFN-γ knocked out, such as AngII mini osmotic pump implantation, some labs have demonstrated a reduction in monocyte infiltration within the aorta and a reduction in MCP-1 and other cytokines such as macrophage inflammatory protein 1α, and P-selectin ligand (75,87). Due to MCP-1’s role in recruiting macrophages to tissues, this suggests a potential interaction between hypertension, the renin-angiotensin system, and macrophage activation and recruitment to tissue.

Within the kidney, macrophage infiltration with hypertension has been noted by several research groups (75,88,89). Infiltration not only occurs during high levels of AngII but can also persist after AngII and systolic BP return to the normal range with lingering MCP-1 and TGF-β (75). Macrophages within the kidneys have been suggested to be involved in kidney injury seen with hypertension. For instance, depletion of macrophages with liposome-encapsulated clodronate has been demonstrated to reduce the kidney injury seen in Ang II-induced hypertension (88), and a similar observation was also found in Dahl salt-sensitive rats (88). Continuing to elucidate the ties between hypertension and macrophages remains, including the forms of activation macrophages may experience from IFN-γ release whether by other immune cells or other cytokines. Additionally, further research to expand on the role of infiltrating macrophages in the kidney and the damage induced through inflammation during hypertension is also needed.

Neutrophils

In the innate response to bacterial infection (such as Salmonella-induced colitis), neutrophils—phagocytic cells from the innate immune system (90)—function as a critical source of IFN-γ production, and depletion of these cells results in relief of many IFN-γ-induced disease symptoms (91,92). In contrast, depletion of neutrophils by administration of RB6–8C5 resulted in hypotension in WT C57BL/6 mice but not in IFN-γ or iNOS deficient mice, indicating neutrophils may also maintain physiologic BP via suppression of IFN-γ-dependent iNOS expression (93). Neutrophils play a role in the maintenance of homeostasis (93), and disruption could lead to elevated inflammation and BP dysregulation as evidenced by neutrophil/lymphocyte ratios functioning as predictors of hypertension (94–96); however, the specific relationship between neutrophil-derived IFN-γ and hypertension in an inflammatory situation is still unclear. Additionally, WT monocytes, not neutrophils, restored the AngII-mediated disease phenotype in LysM+-depleted mice (76). This excellent review by Araos et al. further describes the role of neutrophils in hypertension, and we refer the reader here for further information (97).

Myeloid-Derived Suppressor Cells

Immature and heterogeneous myeloid cells have been recently demonstrated to regulate the immune system (77,98). Myeloid-derived suppressor cells (MDSCs) can be identified by expression of CD11B and Gr-1 surface antigens (99). In addition to playing a role in amino acid metabolism and ROS production, it is known that MDSCs are capable of downregulating immune-system facilitated T-cell response in vivo and in vitro (99). This alludes to a possible role of MSDCs in regulating BP. It had been previously observed in tumor models that the T-cell suppression acts by a mechanism that is independent of traditional MHC complex function—meaning there is no classic antigen presentation by these cells (100). In a recent study utilizing AngII-induced hypertension, it was discovered that MSDCs undergo a phenotypic change after the onset of hypertension; MDSCs harvested from hypertensive mice lost the expression of their surface CD80 and MHC-II (77). Concomitantly, these same cells exhibited increased expression of the IFN-γ receptor, IFN-γR1 (77). Hypertension results in an increase of CD8⁺T-cells that express IFN-γ (43,45). This increase of T cells was attenuated by the presence of MDSCs (77). Further, antibody-mediated depletion of MDSCs resulted in a great increase of IFN-γ-expressing CD4T and CD8T cells (77). Although the complete mechanism of action of MDSCs remains unknown, there is evidence to support that MDSCs might induce development of regulatory T cells. These cells suppress immune response by inhibiting T-cell proliferation and cytokine expression (101). Although MDSCs may have other activities that have still not been discovered, these promising results point to the possibility that MDSCs are capable of inducing regulatory T-cell differentiation to suppress inappropriate T-cell proliferation and attenuate its effects.

DCs

CD11c⁺ DCs have been demonstrated to play an important role in hypertension, wherein blocking antigen presentation by DCs via antagonizing CD80 and CD86 with CTLA4-Ig or B7 deficient mice blunted BP elevation in both the AngII and DOCA+salt models of hypertension (52). Depleting DCs in the mice prevented the development of hypertension in the AngII model, which could be restored by adoptive transfer of WT CD11c⁺ antigen presenting cells (53). DCs are found activated in hypertensive animals to stimulate T-cell production of IFN-γ and IL-17A (102–104). However, the mechanisms regulating of DCs in hypertension is still under investigation. For example, sodium has been suggested to play a role in DC activation through amiloride-sensitive channels and the serum/glucocorticoid kinase 1 (105). Other evidence suggest that isoketal-modified proteins (proteins oxidatively modified by highly reactive γ-ketoaldehydes that accumulate in DCs during hypertension) are responsible for DC activation and subsequent T-cell activation in hypertension (102), and more recently, using DC-specific KO of AT1R, Lu et al. demonstrated a protective role of AT1 receptor on DCs against hypertension and T-cell activation (106). Nevertheless, how DCs are regulated in hypertension and what are the antigen(s) they present to activate T cells to exacerbate hypertension are critical questions that need to be further studied.

Limitations

The relationship between IFN-γ and tissue damage and fibrosis appears to be somewhat contextual, wherein blocking IFN-γ signaling in AngII-infused mice can reduce cardiac remodeling and immune cell infiltration (107,108), but alternative models propose a protective role for IFN-γ (109,110). Further exploration is needed for delineation of these effects. In addition, environmental exposure and subsequent gut microbiota alterations play a role directing the development of murine immune systems and host response to various challenges; as such, laboratory mice do not necessarily model clinical situations (111,112). The microbiota, in particular, has become increasingly implicated in regulating the murine response to hypertension in many experimental models (113,114) and may contribute to cardiovascular phenotypes in some immune-deficient mice. To this end, the preclinical established link between the immune cell activation and hypertension may involve gut microbiome alterations that regulate immune responses in such a way that are not inherently translatable to the clinic; however, such gut microbiome dysregulations have been described in clinical studies (115). We refer the reader to this excellent review by Avery et al. for further discussion of the gut microbiome and hypertension (116).

Clinical Translation

Clinical immunosuppression and hypertension have a rather complicated relationship due to adverse drug effects. In renal transplants, immunosuppressant calcineurin inhibitors have been linked to exacerbated BP (117), but BP post renal transplant has been ameliorated through non-nephrotoxic immunosuppression with mycophenolate mofetil with or without rapamycin (118). Immunosuppression (excluding calcineurin inhibitors) has been shown to be favorable in a clinical trial for patients with CKD (119). In agreement, several proinflammatory factors have been found to be upregulated in patients with treatment-resistant hypertension cases of CKD (120). Further clinical studies are needed to verify the efficacy of targeting key immune players in hypertension (121).

In Summary

IFN-γ plays a key role in the development and maintenance of hypertension. Global KO of IFN-γ results in blunted BP elevation in both the AngII and DOCA+salt models of hypertension. Potential mechanistic relationships between IFN-γ production and hypertension have been noted with NK cells, neutrophils, macrophages, DCs, and T cells (Figure 2). The source(s) of activation—particularly for CD8Ts—driving elevated IFN-γ production are still being researched; nonetheless, a potential novel mechanism involving activators appears promising in light of the identification of an oligoclonal population of CD8Ts in the kidneys of hypertensive mice.

Figure 2.

Graphical representation of proposed mechanisms by which immune cells contribute to the development of hypertension through enhanced immune cell infiltration, inflammation, and sodium retention.

Disclosures

All authors have nothing to disclose.

Funding

This study was supported by NIH/NHLBI (Grant R01-HL146713) to S. Mu, AHA (Grant 15BGIA25730047) to S. Mu, UAMS Medical Research Endowment Awards to S. Mu, and the Systems Pharmacology and Toxicology Training Program (Grant T32-GM106999) to L. Benson.

Author Contributions

L.N. Benson and Y. Liu were responsible for the investigation; L.N. Benson, K.S. Deck, and C. Mora wrote the original draft of the manuscript; S. Mu, L.N. Benson, and Y. Liu were responsible for the methodology; and S. Mu was responsible for the conceptualization, funding acquisition, supervision, and validation, and reviewed and edited the manuscript.