Categories
Uncategorized

First evidence of aryl hydrocarbon receptor as a druggable target in hypertension induced by chronic intermittent hypoxia

Abstract

Background and Purpose Obstructive sleep apnea (OSA) is associated to a high prevalence of resistant arterial hypertension (HTN) justifying the research on novel targets. Chronic
intermittent hypoxia (CIH) is a key feature in the development of OSA comorbidities, including HTN.

Experimental Approach We used arat model of CIH-induced HTN to disclose the hypothesis that the aryl hydrocarbon receptor (AHR) is activated by CIH once it shares the same binding
partner of HIF- 1α and promotes pro-oxidant, pro-inflammatory (NF-kB) and pro-fibrotic events in common with CIH.

Key Results Upon established hypertension (21 days exposure to CIH), we observed an increase in Cyp1a1 mRNA in kidney cortex (6-fold), kidney medulla (3-fold) and liver (3-fold), but not in other tissues. Increased renal expression of Ahr and markers of inflammation (Rela), epithelial to mesenchymal transition markers, the rate-controlling step of gluconeogenesis, Pepck1, and members of HIF-pathway, namely, Hif3a were also observed. Daily administration (14 days) of AHR antagonist, CH-223191 (5 mg.kg- 1.day- 1, gavage), simultaneously to CIH prevented the
increase in systolic blood pressure (SBP) by 53 ± 12% and in diastolic blood pressure (DBP) by 44 ± 16%. Moreover, its administration (14 days) upon already established HTN reversed the
increase in SBP by 52 ± 12%.

Conclusion and Implications CIH caused an activation of AHR signaling particularly in the kidney and its pharmacological blockade had a significant impact reverting already established HTN. This first evidence inspires innovative research opportunities for the understanding and treatment of this particular type of HTN.

Keywords: Secondary hypertension; CH-223191; AHR; kidney; CYP1A1; blood pressure;

1. Introduction

Obstructive sleep apnea (OSA) is a highly prevalent sleep-related breathing disorder [1]. It is characterized by recurrent episodes of airflow cessation or reduction, mainly due to mechanical obstruction in the upper airways. The main hallmark of OSA is chronic intermittent hypoxia (CIH) that is responsible for the vast majority of OSA-related comorbidities, namely cardio-metabolic diseases and particularly systemic hypertension (HTN) [2].Apart from being an independent risk factor for HTN, OSA is also a major secondary cause of resistant HTN to available antihypertensive drugs, including adrenergic blockers [3, 4] and justifying the identification of novel therapeutic drug classes. Typically, OSA patients lose the nocturnal physiological reduction in blood pressure (BP) and BP remains elevated in the absence of obstructive events during wakefulness [5]. There is substantial evidence that the mechanisms behind CIH-induced HTN include carotid body mediated long-term facilitation of sympathetic activity [6]. Also, emerging evidence supports additional origins for the autonomic imbalance behind CIH-induced HTN, such as aberrant reno-renal reflexes [7] and direct central activation of sympathetic tone to cardiovascular effectors. However, the molecular pathways that mediate diurnal chronic HTN after withdrawal of the nocturnal hypoxic stimuli remain uncertain.

Activation of transcription factors (triggered by sympathetic activation or not) might explain the gap between short-term hypoxic stimuli and long-term multisystem effects. Indeed, intermittent hypoxia increases hypoxia inducible factor 1α (HIF- 1α) stabilization [8- 11] and nuclear factor kappa B (NF-kB) activity [12, 13]. However, HTN is not a comorbidity particularly associated to other conditions characterized by systemic HIF- 1α stabilization (e.g. high altitude or chronic obstructive pulmonary disease) or NF-kB activations (e.g. rheumatoid arthritis, asthma). It is believed that at least some other tissue-specific mechanisms relevant to BP control maybe activated by CIH.

Herein, we have investigated the effects of CIH on the transcriptional pathways triggered by the aryl hydrocarbon receptor (AHR). AHR is a ligand-activated transcription factor that belongs to the basic helix-loop-helix/Per-Arnt-Sim (bHLH/PAS) family [14]. ARNT (also named HIF- 1ß) is a shared dimerization partner for AHR and other molecular pathways, such as HIF- 1 pathway [15]. The competition between AHR and HIF- 1α regarding dimerization to ARNT [16] might be relevant to consider in the context of CIH. Additionally, it is widely known that AHR and NF-kB subunits interact [17], namely the heterodimerization ofRelA-AHR, that is able to repress CYP1A1 activation [18]. Regulation of CYP1A1 expression is presently considered a sensitive marker of AHR activation [19].

AHR has been shown to promote cardiovascular alterations duetoxenobiotic exposure in humans [20, 21] and in animals [22] and during development and/or homeostasis in animal models [23-26]. Polymorphisms of the AHR signaling pathway are reported to be closely associated with the pathogenesis of essential HTN [27]. Additionally, genome-wide association studies showed that human CYP1A1 polymorphisms are associated with kidney glomerular filtration rate [28] and increased risk of HTN [29]. This highlights the putative importance of the AHR-CYP1A1 axis in the homeostasis of BP. However, to the best of our knowledge, the role of the AHR pathway in OSA comorbidities has never been addressed.In this study, using arat model of CIH-HTN, we first investigated the effect of CIH on AHR pathway. Then, wepharmacologically targeted this pathway with an AHR antagonist CH-223191, in order to clarify how the manipulation of this molecular pathway interferes with BP and heart rate (HR) under CIH conditions, providing new opportunities for the understanding and the treatment of this particular type of HTN.

2. Methods
2.1 Study design

Five different sets of experiments with animals were performed. The animals were allocated randomly to each subset.A first set of animals was used to characterize the molecular pathway of AHR and other genes of interest after exposure to 21 days of CIH (n=8) after 2 days of chamber acclimatization under normoxic conditions (Nx, i.e. 21% O2 + 79% N2). Control animals (n=5) were exposed for 21 days to Nx in the same room as the CIH groups.The second and third sets of experiments were used to investigate the effect of the AHR antagonist CH-223191 [30] on BP and HR.Since the use of CH-223191 in rats in vivo is scarce, we decided to perform a first dose-exploratory experiment (n=3) to disclose the effects of 5 mg.kg- 1 once a day, by oralgavage in 1 mL of vegetable oil on BP and HR in Nx, followed by the its ability to prevent and revert HTN induced by CIH (second set, n=3, see study design scheme in Figure 5A). To the best of our knowledge there are only two studies that administered CH-223191 by gavage to rats: 8 mg.kg- 1 per day for 14 days in a model of pulmonary hypertension [31] and 5 mg.kg- 1 per day for 14 days in amodel of arthritis [32].

The third set of experiments (n=5) intended to validate thereversion in HTN with CH-223191 (see study design scheme in Figure 6A).A fourth set of experiments was used to assess the effect of the administration of CH-223191 (5 mg.kg- 1, by oralgavage, daily, for 14 days; n=3) in Nx (see study design scheme in supplementary figure 2).The last two groups of animals were used with the purpose of assessing the effects of CIH in the body weight growth: one group was exposed to Nx for 35 days (n=6) and another group with telemeters implanted was exposed to CIH for 35 days (n=4).

2.2 Animals

Thirty nine male Wistar Crl:WI (Han) (Rattus norvegicus L), aged 8- 12 weeks, with mean body weights of 249 ± 5 g, were obtained from the NOVA Medical School animal facility. Animals
were housed in polycarbonate cages with wire lids (Tecniplast, Buguggiate, Varese, Italy) and maintained under standard laboratory conditions: artificial 12 h light/dark cycles (lights on from 9 am – 9 pm), at room temperature (22 ± 2.0 °C) and a relative humidity of 60 ± 10%. Rats were given standard laboratory diet (SDS diets RM1) and reverse osmosis water ad libitum.
All applicable institutional and governmental regulations concerning the ethical use of animals were followed, according to the NIH Principles of Laboratory Animal Care (NIH Publication 85-23, revised 1985), the European guidelines for the protection of animals used for scientific purposes (European Union Directive 2010/63/EU) and the Portuguese regulation and laws on the
protection of animals used for scientific purposes (Lawnº 113/2013). All procedures were approved by the Ethical Committee of the NOVA Medical School (protocol nº15/2017/CEFCM).

Rats were weighed at baseline and once a week throughout the entire study. Water and food intake were also measured once a week. At the end of the experiments, the rats were anaesthetized by intraperitoneal injection with a solution of medetomidine (0.5 mg.Kg- 1 body weight; Domitor®, Pfizer Animal Health, Auckland, New Zealand) and ketamine (75 mg.Kg- 1 body weight; Imalgene 1000®, Mérial, Lyon, France), and cardiac puncture was performed without thoracotomy. Death was confirmed by cervical dislocation before organs and tissue collection.

2.3 Cardiovascular parameters

Indwelling radiotelemeters (HD-S10, Data Sciences Corporation, USA) were inserted in the abdominal aorta, to allow the recording of BP and HR in freely moving rats, as previously
described in a recent paper by our group [33].SBP, DBP and HR values were obtained during 30-s sampling periods and were recorded for 20 min, everyday at 8am, during the light-off phase (rats’ active period), with exclusion of the first 5 minutes in each measurement for each rat. The light-off period was chosen to mimic more accurately the diurnal HTN in OSA patients. The development of stable HTN inthis model occurs in approximately 14-21 days of exposure to CIH [33].

2.4 Chronic intermittent hypoxia paradigm

Since the seminal work of Fletcher and colleagues [34], animal models of CIH have been extensively and successfully employed to assess changes in cardiovascular outcomes arisen from CIH [3]. The paradigm of CIH herein employed was already validated by our group [33] allows the lowering of O2 levels from 21% to ~5-6% in each cycle of CIH, and animals were exposed to 5.6 CIH cycles.h- 1 for 10.5 h.day- 1. This paradigm mimics the same pattern of CIH present in mild to moderate clinical OSA [33].Briefly, the animals exposed to CIH experiments were maintained in a eucapnic atmosphere inside medium A-chambers (Biospherix Ltd, NY, USA), three animals by chamber and oxygen concentrations were controlled through an OxyCycler A420C (Biospherix Ltd, NY, USA).Chambers were infused with 100% N2 for 3.5 minto quickly reduce the O2 concentration,following by an infusion with 100% O2 for 7 minto restore O2 to ambient levels until the start of the next CIH cycle. Rats were exposed to CIH in their sleep period (light phase), from 9.30 am to 8.00 pm, during 14, 21, 35 or 42 days, depending on the experiments. During the remaining hours of the day, the chambers were infused with 21% of O2.

2.5 Markers of AHR activation and quantitative real-time PCR

Liver and adipose tissue (used as comparator tissues not directly related to HTN) and kidney (medulla and cortex) were homogenized in Trizol® (Life Technologies) and stored at -80 ºC.
After total RNA extraction, cDNA was synthesized from 1 μg of RNA, using theNZY First-Strand cDNA synthesis kit (NZYTech, Lisbon, Portugal), according to the manufacturer’s directions. Quantitative realtime polymerase chain reaction (qPCR) was performed with 20 ng of cDNA, using SensiFAST™ SYBR® Hi-ROX Kit (Bioline, United Kingdom),with duplicates for each experiment. Supplementary Table 1 gives the gene-specific primers used for the genes analyzed. The relative amounts of mRNA were estimated using the ΔΔCT method with cyclophilin A (Cypa) as the reference gene [35]. We assessed mRNA levels of several genes associated with the AHR axis. The markers analyzed included those directly related with AHR pathway (Ahr, Arnt, Cyp1a1, Cyp1a2, Cyp1b1, and Pepck1); markers of fibrosis and epithelial-mesenchymal transition (EMT) (collagen 1A1 – Col1a1, α smooth muscle actin – aSma,fibronectin – Fn1, vimentin – Vim), epithelial markers (E-cadherin – Ecad) as well as markers of inflammation (interleukin 1b – Il1b and interleukin 6 – Il6) and endoplasmic reticulum stress (Grp78).

We have also quantified the transcription factors involved in CIH responses and related with AHR, namely NF-kB subunits, Rela and p49/p100, and members of HIF family, such asHif1a, Epas1 (HIF-2α), and Hif3a, and its associated target gene vascular endothelial growth factor A (Vegfa). Additionally, markers of the renin-angiotensin system, namely renin (Ren), angiotensin
II receptor type 1 (At1) and angiotensin-converting enzyme (Ace1) were also analyzed due to theirrelevance in BP control.

2.6 Statistical analyses

GraphPad Prism version 5 (GraphPad Software Inc., San Diego, CA, USA) was used to perform Child immunisation all the statistical analysis. Data arepresented as the mean ± standard deviation or mean ±
standard error of the mean (SEM). Differences were considered significant atp<0.05 (*** p<0.001, ** p<0.01; * p<0.05) and obtained using one-way ANOVA or Mann–Whitney’s U-test whenever applicable. 3. Results
3.1 CIH activates AHR signaling particularly in kidney tissue

An overexpression of the Cyp1a1 gene (indicator of AHR activation) was observed primarily in the kidney cortex, but also in the kidney medulla and liver (Figure 1A). Among other target
genes of AHR, neither Cyp1b1 nor Cyp1a2 were modified by CIH in any of the four analyzed tissues. Ahr mRNA expression was also increased, except in visceral adipose tissue (VAT)
(Figure 1B). The binding partner of both AHR and HIF family pathways, Arnt, also known as HIF- 1β, was not upregulated by CIH in the kidney but exhibited increased expression in liver (2-
fold,). Overall, despite no exogenous exposure to any classical AHR ligands, we showed that CIH leads to activation of the AHR signaling pathway in several tissues (kidney, liver) important for xenobiotic metabolism and the development of HTN.

Figure 1. Chronic intermittent hypoxia (CIH) activates aryl hydrocarbon receptor (Ahr) and Cyp1a1 mRNA expression, particularly in renal cortex (RC), but also in renal medulla (RM) and
liver. Cyp1a1 mRNA was not expressed in visceral adipose tissue (VAT). Relative changes in transcriptional levels between samples from CIH-exposed (21 days, n=8) and normoxic animals
(n= 5) were obtained using comparative Ct method (2-∆∆Ct). Data arepresented as mean fold change ± SD; * p<0.05, ** p<0.01; Mann-Whitney test. Cyp1a1: cytochrome P450 family 1
subfamily A member 1; ND: not detectable.

3.2 CIH activates several members of HIF pathway, particularly in kidney

We assessed the effects of CIH on the expression of HIF family (1, 2 and 3 α) and vascular endothelial growth factor (Vegfa), a classical transcriptional target of HIF- 1α (Figure 2).
There were no changes in Hif1a mRNA expression in liver and visceral adipose tissue, but it was increased in both kidney cortex and medulla (Figure 2A). Accordingly, the same pattern was
observed for Vegfa (Figure 2B). CIH also altered the expression HIF-2α (Epas1), particularly in renal medulla (Figure 2C). The effect of CIH in Hif3a mRNA expression showed a similar
pattern to the one observed with HIF- 1α: no effect in liver and adipose tissue and increases in the kidney (Figure 2D), thus suggesting the activation of HIF pathways upon CIH exposure,
particularly in the kidney.

Figure 2. Chronic intermittent hypoxia (CIH) activates mRNA expression of hypoxia-inducible factors (HIF) and its target gene (Vegfa), in the kidney, but not in the liver and visceral adipose tissue (VAT). Relative changes in transcriptional levels between samples from CIH-exposed (21 days, n=8) and INCB024360 concentration normoxic animals (n=5) were obtained using comparative Ct method (2-ΔΔCt).Data are expressed as mean fold change ± SD; * p<0.05; Mann-Whitney test. Epas1: endothelial PAS domain protein 1; RC: renal cortex; RM: renal medulla; Vegfa: vascular endothelium growth factor A. 3.3 CIH impacts NF-kB inflammatory pathway and endoplasmic reticulum stress in kidney The mRNA levels of several inflammatory markers were evaluated, namely NF-kB subunits,Rela (also known asp65) and p49/p100 and the interleukinsIl1b and Il6 (Figure 3). Only Rela, which belongs to the classical pathway of NF-kB activation, was overexpressed in the kidney cortex (Figure 3A). In contrast, increased mRNA levels of p49/p100 were observed in VAT
(Figure 3B). Overall, our results suggest that pathways participating inflammation caused by CIH are tissue-specific. No changes for interleukins were observed in kidney and the Il6 was
only increased in VAT (Figure 3C and 3D). Endoplasmic reticulum (ER) stress can be a protective mechanism in early phases of CIH exposure, but overtime it can turn into a deleterious mechanism leading to cell death [36]. With this in mind, we have quantified the endoplasmic tissue stress-related GRP78 (Grp78 gene) mRNA, and found that its expression was upregulated in the kidney and liver tissues (Figure 3E).

Figure 3. Chronic intermittent hypoxia (CIH) activates mRNA expression of NF-kB related genes (Rela, p49/p100, IL1b, IL6) and endoplasmic reticular stress marker (Grp78), in a tissue-
dependent manner. Relative changes in transcriptional levels between samples from CIH-exposed (21 days, n=8) and normoxic animals (n= 5) were obtained using comparative Ct method (2-ΔΔCt). Data arepresented as mean fold change ± SD; * p<0.05; Mann-Whitney test.Il1b: interleukin 1 beta; Il6: interleukin 6; Grp78: glucose regulated protein 78; p49/p100:nuclear factor kappa B subunit 2; Rela: v-rel avian reticuloendotheliosis viral oncogene homolog A; RC: renal cortex; RM: renal medulla; VAT: visceral adipose tissue. 3.4 CIH impacts renin-angiotensin system in kidney Higher expressions of renin (Ren) in the renal medulla (3.2-fold change) and of AT1 receptor (At1) in the renal cortex (2-fold change) were observed upon CIH exposure (supplementary figure 1). In contrast, no changes of Ace1 (angiotensin I-converting enzyme type I) in kidney tissue were found. These results are in line with the experiments described by others [37, 38],when analyzing kidney tissue as a whole, validating our CIH paradigm. 3.5 CIH impacts epithelial to mesenchymal transition process in kidney Increased expression of vimentin and fibronectin (mesenchymal markers) is traditionally observed during EMT while E-cadherin (epithelial marker) expression is decreased. While Vimentin (Figure 4A) and Fn1 (Figure 4B) were upregulated in kidney tissue, the mRNA expression of Ecad (Figure 4C) and of the fibrotic markers Col1a1 (Figure 4D) and aSma (Figure 4E) showed no difference in kidney tissue upon CIH exposure. Overall, this suggests the acquisition of mesenchymal characteristics by kidney tissue in response to 21 days of CIH. Figure 4. Chronic intermittent hypoxia (CIH) oncology medicines activates amesenchymal phenotype in the kidney,but not in the liver and visceral adipose tissue (VAT). (A-B) Mesenchymal, (C) epithelial and
(D-E) fibrosis markers were analyzed. Relative changes in transcriptional levels between samples from CIH-exposed (21 days, n=8) and normoxic animals (n= 5) were obtained using comparative Ct method (2-ΔΔCt). Data arepresented as mean fold change ± SD; * p<0.05; Mann-Whitney test. aSma: alpha smooth muscle actin; Col1a1: collagen type I alpha 1; Fn1:fibronectin 1; Ecad: E-cadherin; RC: Renal cortex; RM: Renal medulla; Vim: vimentin. 3.6 AHR antagonist CH-223191 prevented CIH-induced increase in blood pressure The molecular results concerning the likely activation of the AHR pathway after 21 days of CIH exposure (first set of experiments) directed the work towards the in vivo use of an AHR
antagonist to counteract the effects of CIH (second set of experiments). Since the use of CH-223191 in rats in vivo is scarce, we decided to perform a first dose-exploratory experiment to
clarify the effect of this compound on BP and HR and its potential to impact the HTN induced by CIH. Then=3 is justified considering that this was a first preliminary dose-exploratory experiment and that the administration of CH-22319, particularly in rats is limited by its prohibitive costs as previously reported [31].

Figure 5A displays the timeline of the experiment, that started by measuring the baseline values for BP and HR in normoxia (Nx; chamber acclimatization). Next,the animals were treated with the CH-223191 for 7 days (5 mg.Kg- 1, once a day), to assess if the compound could alter these cardiovascular parameters in Nx conditions (Nx+CH) and no changes were observed: 124 ± 5
mmHg vs. 126 ± 4 mmHg for SBP (Figure 5B) and 85 ± 7 mmHg vs. 88 ± 6 mmHg for DBP (Figure 5C). In a separate group of experiments, we confirmed that a longer exposure to CH-223191 (14 days) did not change basal BP in Nx conditions (supplementary figure 2). Then, CH-223191 was administered concomitantly with CIH (14 days), and amoderate non-significant increase in BP (CIH+CH, 131 ± 5mmHg) was observed (Figure 5B and 5C). In order to investigate if the AHR antagonist was preventing a larger increase in BP induced by CIH, we withdrew the compound, while maintaining the animals exposed to CIH for more 14 days (CIH only). Over this period, a bigger and significant increase in the BP was observed (142 ± 1 mmHg), suggesting that the AHR antagonist was able to prevent (CIH+CH vs. CIH) HTN caused by CIH in 56 ± 15% and 44 ± 16% of SBP and DBP, respectively (Figure 5B and C).More importantly, after the rise of the BP that followed the withdrawal of the antagonist (CIH+CH vs. CIH), re-administration of CH-223191 reverted the increase in the BP caused by CIH in 61 ± 18% and 45 ± 13% in SBP (142 ± 1 mmHg vs. 131± 4 mmHg) and DBP (99 ± 4 mmHg vs. 93 ± 4 mmHg), respectively (CIH vs. CIH+CH) (Figure 5B and C). No changes were observed through this longitudinal study in heart rate (Figure 5 D).

Figure 5. A first dose-exploratory experiment (n=3) to clarify the effect of CH-223191 on blood pressure and heart rate. The AHR antagonist, CH-223191 (CH, 5 mg.Kg- 1.day- 1, gavage)
prevents and reverts the increase in blood pressure caused by chronic intermittent hypoxia (CIH) exposure. (A) Longitudinal study design. The effect of CH-223191 was assessed in (B) systolic blood pressure, (C) diastolic blood pressure and (D) heart rate. Data arepresented as mean ± SEM (n=3); * p<0.05, paired one-way ANOVA with Bonferroni post-test. Only the last 8 days of each condition were considered for statistical analysis. Bpm: beats per minute; mmHg: millimeter of mercury; Nx: normoxia. 3.7 AHR antagonist CH-223191 reverted CIH-induced hypertension Next, we aimed to confirm that AHR blocking reverses CIH-HTN. For that,a third set of experiments was performed (see also material and methods to compare each set),where the administration of this AHR antagonist was initiated when HTN had been already established (i.e.after 21 days of CIH) (Figure 6A). CH-223191 (5 mg.Kg- 1 orally, during 14 days) reduces the increase in SBP caused by CIH (21 days) in 52 ± 10% (141 ± 4 mmHg vs. 134 ± 4 mmHg,respectively), while the reduction in DBP (41 ± 13%, 104 ± 4 vs. 99 ± 5 mmHg, respectively) did not achieve statistical significance (Figures 6B and 6C).This effect of CH-223191 in reverting high BP cannot be attributed to its vehicle (vegetable oil) because in control experiments, where the antagonist was replaced by its vehicle, no changes in the effect of CIH in BP and HR were observed (supplementary figure 3). Figure 6. The AHR antagonist, CH-223191 (CH, 5 mg.Kg- 1.day- 1, gavage) reverts the blood pressure increase caused by chronic intermittent hypoxia (CIH). (A) Longitudinal study design
(n=5). The effect of CH-223191 was assessed in (B) systolic blood pressure, (C) diastolic blood pressure and in (D) heart rate. Data are expressed as mean ± SEM (n=5); * p<0.05; **p<0.01;
*** p<0.001; paired one-way ANOVA with Bonferroni post-test. Only the data of the last 8 days of each sub-group (Nx, CIH or CIH+CH), were used for statistical analysis. Bpm: beats per
minute; mmHg: millimeter of mercury; Nx: normoxia.

3.8 AHR antagonist CH-223191 interference in other CIH-induced cardiometabolic changes

Although no impact of CIH or CH-223191 on HR was observed in second set of experiments (figure 5 D), in thereversion protocol (third set) (figure 6), the exposure of CIH together with
CH-223191 decreased the HR in 9±2% (from 407 ± 21 beats per minute (bpm) to 370 ± 24 bpm),when compared to Nx (figure 6D). An effect confirmed also in the vehicle-treated animals, with
HR decreasing with CIH (415 ± 12 bpm vs 375 ± 11 bpm, Nx and CIH, respectively) and maintained with the vehicle administration (364 ± 15 mmHg) (supplementary figure 3C).

4. Discussion

Experiments to address how CIH modifies AHR-CYP1A1 pathway were performed using arat model exposed to a mild paradigm of CIH for 21 days. This represents the time needed for HTN development in this model. Animals kept 21 days in normoxia (Nx) were used as control group. Cyp1a1 mRNA expression was used as a sensitive indicator of AHR activation and assessed in
several tissues (kidney cortex, kidney medulla, liver and visceral adipose tissue). Then we validated the protocol investigating the mRNA expression of classical biomarkers of CIH
including HIF, NF-kB and the renin-angiotensin-aldosterone (RAA) pathway that have also been shown to be related to AHR. As AHR not only regulates xenobiotic metabolism, we have also
analyzed several panels of markers of AHR-related processes such as fibrosis [39], EMT [35],inflammation and gluconeogenesis [40].

The identification of new pathways related to OSA-associated HTN is needed to deal with the high prevalence of resistant hypertension in this population. Our hypothesis was based on the
assumption that the AHR pathway is activated by CIH and that this can contribute to HTN associated to this condition. Overall, the results confirmed our initial hypothesis and provided
new insights about AHR signaling linking different co-morbidities associated to OSA. In fact we demonstrated that CIH overexpresses the classical AHR target gene, the xenobiotic metabolizing enzymes (Cyp1a1), particularly in renal tissues. Moreover, blockade of the AHR with CH-223191 prevented and reverted the HTN induced by CIH. The overexpression of EMT genes
found in the kidney of animals exposed to CIH is also in line with an endogenous activation of the AHR pathway by CIH. The increase in Pepck1 mRNA-expression (the rate-controlling step
of gluconeogenesis) also suggests that AHR pathway might also be linked with other comorbidities associated to CIH such as glucose intolerance. In contrast, the inflammation of the
visceral adipose tissue induced by CIH might represent an AHR-independent event, because it showed a completely different pattern of genes induced by CIH: no change in the typical AHR
pathway induced genes; a selective overexpression of Il6 without overexpression of the canonical RelA/NF-kB activation and an overexpression of the p49/p100 NF-kB. We also found a remarkable increase in the mRNA expression of Hif3a in the kidney. This finding suggests that CIH might stimulate the inhibitory control of HIF3α over HIF- 1/2α or that HIF3α has
transactivation activity partially overlapping that of HIF- 1α, as already proposed by others [41-43].

Inference of AHR activation was mainly based on the assumption that CYP1A1 is a hallmark of AHR activation [19] and that CYP1A2 and CYP1B1 are also regulated in an AHR- independent
manner [44]. To the best of our knowledge, the effects of CIH on CYP1A1 have never been addressed. A decrease (0.5-fold) in hepatic CYP1A2 mRNA and protein levels caused by CIH was described in mice by Zhang and co-authors [45], while in rats CYP1A2 mRNA did not change. AHR mRNA expression was slightly increased in kidney and liver. Its binding partner, shared with HIF family pathways, ARNT, was not upregulated by CIH in the kidney, but exhibit increased expression in liver (2.1-fold) and visceral adipose tissue (2.4-fold). The meaning of these results is unknown once the regulation of ARNT expression is far from being elucidated.

As CIH seems to activate the AHR pathway, we logically performed in vivo experiments with an antagonist of the receptor to assess the effects of the AHR on the development of HTN. CH-
223191 has been developed in 2006 and shown to prevent activation of the AHR by 2,3,7,8-(HAHs) [30, 46]. More recently the first in vivo study in rats showed that an oral administration
of CH-223191 (8 mg.Kg- 1) reversed the development of pulmonary HTN caused by the AHR agonist Sugen, followed by exposure to sustained hypoxia, normalizing as well pulmonary CYP1A1 expression [31]. The phenotype of sustained hypoxia differs from CIH because it has been associated to pulmonary HTN mediated by HIF- 1α stabilization, not causing systemic HTN
[47]. In contrast, the high prevalence of cardiovascular disease in OSA patients has been attributed to a predominant activation of inflammatory (e.g. NF-kB) over HIF- 1 dependent-
transcription in experimental models of CIH andin OSA patients [12]. In fact, the increase of However, the finding that the AHR antagonist reverses both pulmonary [31] and systemic HTN
(present work) point to the need of an in-depth understanding of HIF-AHR interactions in conditions associated to hypoxia.

The effects of CIH on HR demonstrate some inconsistencies. In fact, both the seminal works of Fletcher’s group [34, 52] and a more recent one [33] found no changes in HR in animals
submitted to 35 days of CIH, but other works found elevated values of HR in both male [53] and female [54] rats exposed to CIH during 35 days. In all the studies, HR recordings were obtained in conscious rats, however, while the two former studies used implanted telemeters/catheters and daily monitoring, the later analyzed HR through intra-femoral catheters introduced at the end of the CIH exposure, in anaesthetized animals.

We have paid particular attention to the kidney due to its pivotal importance in BP regulation,because renal sympathetic nerve ablation has been a common strategy to control resistant HTN
[55, 56] and because the effects of CIH have been poorly investigated in the kidney. A few studies have addressed the renal effects of CIH in animal models [10, 36, 49, 57]. Long-term IH
exposure (2-4 months) induces kidney damage, with the presence of renal inflammation,oxidative damage, endoplasmic reticulum (ER) stress, glomerular hypertrophy,mesangial matrix expansion, increased expression of glomerular growth factors (TGF-β1, CTGF and VEGF) cell death and fibrosis [57]. These observations were also reported by others [48, 58].

The added value of the present work is to note that the link between CIH and ER stress is already apparent in the kidney and liver after 21 days of CIH exposure and to highlight the putative association of EMT/fibrosis caused by CIH with the endogenous activation of AHR and Cyp1a1 overexpression.The effect of CIH on body weightherein observed was expected and consistent with CIH causing a reduction in rats’ growth curve, an effect well established in this animal model [33,59]. However, our results suggest that the decrease in body weight can be attributed to a decrease in appetite/food intake and not only to a higher metabolism and energy expenditure as proposed before by others [60]. Apparently, the growth delay and reduction in food consumption caused by CIH are not mediated by activation of AHR, since CH-223191 (5 mg.Kg- 1.day- 1, 2 weeks) did not modify it. In contrast, CH-223191 (10 mg.Kg- 1.day- 1, 5 weeks) suppressed severe wasting with significant weight loss characteristic of AHR agonists, such as TCDD [30]. This inconsistency might reflect the need of higher doses of CH-223191. As a consequence, different doses and intensities of CIH exposure should be further investigated before discarding the participation of AHR activation on the metabolic effects triggered by CIH. In the present work we found an increase in liver and kidney mRNA expression of Pepck1, a major regulator of hepatic gluconeogenesis controlled by AHR [61, 62] among others. Although we ignore at which extent this increase in PEPCK involves changes in AHR activation, it seems that CIH shares with AHR agonists a decrease in appetite but the opposite effects in PEPCK activity [55, 57]. The increase in Pepck1 expression caused by CIH (21 days) might explain why animals submitted to CIH show increased levels of insulin [9, 63] and is consistent with the pattern of intolerance to glucose observed in sleep apnea patients [64]. These results support not only the activation of AHR pathway by CIH but also its participation in CIH-cardio-metabolic comorbidities other than HTN. However, since renal PEPCK is also regulated by other transcriptional factors such as NF-kB [65], we cannot exclude the involvement of other pathways.

This study has several limitations related to the experimental model used (whole animal in vivo) that is more oriented to functional drug effects than to characterization of molecular mechanisms.Although we have provided evidence that CIH stimulates AHR signaling, particularly in the kidney, we cannot state that the anti-hypertensive effect of the AHR antagonist, CH-223191 was mediated by a direct effect on the kidney. CH-223191 is a highly lipophilic drug, and we did not investigate the effect of CIH on AHR pathway in the central nervous system and carotid body,key players in BP control. Other limitations of the study include the use of only one antagonist in a single dose and a relative short time exposure to CH-223191 and the lack of data on protein quantification of AHR hallmarks. Anyway, the functional effect of CH-223191 in HTN is relevant enough to support further exploitation of the mechanisms behind.
In conclusion, the pharmacological blockade of AHR had a significant impact reverting already established HTN caused by CIH without hypotensive effects in normoxic conditions. We also
noted that AHR pathway might also be a common link with other comorbidities associated to CIH such as glucose intolerance.

Systemic HTN is a major cause of resistant HTN in OSA patients. Knowing the increasing prevalence associated with this condition, the research on novel therapeutic approaches to treat
this condition is of utmost importance. Thus, our work provides an innovative view over the molecular mechanism of HTN-CIH, presenting AHR signaling as a new player in the pathophysiology of HTN-CIH, calling upon more research efforts to unveil the exact mechanisms linking AHR with HTN-CIH.