IL-17A Contributes to the Angiotensin II-induced Neurovascular Coupling impairment through Oxidative Stress


 Hypertension, a multifactorial chronic inflammatory condition, is a risk factor for neurodegenerative diseases including stroke and Alzheimer’s disease. These diseases have been associated with higher concentration of blood interleukin (IL)-17A. However, the role that IL-17A plays in the relationship between hypertension and brain remains misunderstood. Cerebral blood flow regulation may be the crossroads of these conditions. Hypertension alters cerebral blood flow regulation including neurovascular coupling (NVC). In the present study, the effects of IL-17A on NVC in the context of hypertension induced by angiotensin (Ang) II will be examined. Our results show that the neutralization of IL-17A or the specific inhibition of its receptor prevent the Ang II- induced NVC impairment. These treatments reduce the Ang II-induced cerebral oxidative stress. Tempol and NOX-2 depletion prevent NVC impairment induced by IL-17A. These findings suggest that IL-17A, through superoxide anion production, is an important mediator of cerebrovascular dysregulation induced by Ang II.


INTRODUCTION
Hypertension, now considered a chronic inflammatory condition, is the most prevalent and modifiable risk factor for neurodegenerative diseases including stroke and Alzheimer's disease 1 .
However, although many studies have linked brain inflammation and cerebrovascular alterations with brain disorders [2][3][4][5] , the involvement of inflammation in the impact of hypertension on the brain remains largely unknown. Previous data from experimental model of hypertension have 2 shown important alteration of cerebral blood vessels and cerebral blood flow (CBF) regulation including neurovascular coupling (NVC) [6][7][8] . NVC is defined by an increase in CBF in response to neuronal activation. Since the brain has high energy needs, slight alterations of this mechanism can negatively impact cerebral protein synthesis and neuronal function 9 . Although the involvement of inflammation in hypertension and peripheral vascular injury is well documented, the impact on CBF regulation remains largely underinvestigated.
The impact of inflammation on NVC was previously revealed through anti-inflammatory treatments with Treg lymphocytes (CD4+/CD25+) or interleukin (IL)-10 in an experimental model of hypertension induced by mice chronic angiotensin (Ang) II perfusion. In this study, we have shown that these treatments prevent gliosis and cerebral oxidative stress 6 . Ang II plays a key role in inflammation and increases the activity of T effecter cells such as T helper 17 lymphocytes and gamma-delta T-cells, 10,11 enhancing the production of pro-inflammatory cytokines, such as IL-17A 10,12 . IL-17A can induce brain damage by acting directly on neurons 13 or indirectly through disruption of the BBB and neurovascular functions [14][15][16][17] . We hypothesized that these effects could be achieved through oxidative stress, since nicotinamide adenine dinucleotide phosphate (NADPH)-oxidase, specifically NOX-2, derived reactive oxygen species are well known to promote cerebrovascular dysfunctions [18][19][20] and to be involved in IL-17Ainduced vascular inflammation 21 .
IL-17A is involved in reduced resting CBF and NVC impairment induced by high salt diet in mice 16 . These findings raise the possibility that IL-17A may be involved in CBF impairment in other experimental models of hypertension. Therefore, we examined the role of IL-17A in the 3 neurovascular dysfunctions induced by Ang II. To explore this question, we have characterized the impact of IL-17A on CBF regulation in Ang II-induced hypertensive mice. Then, the potential role of NOX-2 and the subsequent superoxide anion production were evaluated to explain how IL-17A impairs NVC.

Neutralization of IL-17A or inhibition of its receptor prevents the Ang II-induced NVC impairment
To examine the involvement of IL-17A on neurovascular coupling (NVC) impairment induce by Ang II, IL-17A neutralizing antibody (Ab) was administered in mice in parallel with Ang II. As previously observed 8 , CBF increases in response to whiskers stimulations in Ang II hypertensive mice were significantly attenuated by 23.2% compared to their controls (Figure 1 A, C, p< 0.01).
Neutralizing IL-17A Ab prevents NVC impairment induced by Ang II (Figure 1 A, D, p< 0.01) without changes in resting CBF ( Figure 1B). IL-17A Ab does not elicit changes in cerebrovascular responses to neuronal stimulations in control mice.
We next examined whether the rescued cerebrovascular responses corresponded to a prevented increase in systolic blood pressure (SBP). In mice not receiving the IL-17A Ab, as expected 10,22,23 , Ang II significantly increased the SBP compared to its control at day 7 and 14.
This increase in SBP is of 37.4 and 35.9 mmHg respectively. IL-17A neutralization slightly attenuates the increase in SBP induced by Ang II at those same time point by 9.8 and 9.4 mmHg (Supplemental Figure 2).

4
In order to assess whether the IL-17A prevents NVC impairment through its receptor, mice received an IL-17A receptor antagonist (IL-17RA mAB) simultaneously with the chronic systemic administration of Ang II. Ang II attenuates the increases in CBF (13.6%) in response to whiskers stimulations compared to the control sham mice (18.3%) (Figure 2 A, C, p< 0.01).
Inhibiting the IL-17A receptor prevents the disruption of NVC induced by Ang II (18.6%) (Figure 2 A, D, p<0.01). The laser-Doppler probe was positioned to get similar resting CBF in between groups ( Figure 2B). The receptor antagonist does not elicit changes in cerebrovascular responses in control mice.
Then, the effect of IL-17A receptor inhibition on the SBP increase induced by Ang II was determined. The mean SBP in mice receiving Ang II was significantly higher compared with controls at day 7 and 14 with an increase of 51.7 and 53.7 mmHg, respectively. Inhibition of the IL-17A receptor partially attenuated the increase in SBP induced by Ang II at those same day point by 11.4 and 10.1 mmHg. (Supplemental Figure 3).

Neutralization of IL-17A or inhibition of its receptor prevents the superoxide anion production induced by Ang II
To determine whether the increase production of reactive oxygen species by Ang II is mediated by IL-17A, we investigated the effect of IL-17A Ab on superoxide anion production. As shown in Figure 3, the increased production of superoxide anion by Ang II seen in the somatosensory cortex (p = 0.12) and in the lacunosum moleculare (LMol), dentate gyrus (DG), cornu ammonis 1 (CA1) and cornu ammonis 3 (CA3) regions of the hippocampus (p< 0.01) was prevented 5 following IL-17A Ab administration (p< 0.05). Moreover, the production of superoxide anion in hypertensive mice receiving IL-17A Ab does not differ from that of control mice (Sham) with or without the IL-17A Ab administration.
In the same manner, somatosensory cortex and in all regions in hippocampus (LMol, DG, CA1, CA3,) showed a higher production of superoxide anion in mice with chronic administration of Ang II. IL-17RA mAB administration normalized these levels to the same as those of control mice (Sham) with or without the IL-17RA antagonist ( Figure 4).

Tempol treatment or NOX-2 depletion prevents NVC dysfunction induced by IL-17A
Since IL-17A neutralization and IL-17A receptor inhibition prevent NVC impairment and To better understand the mechanisms by which IL-17A induces cerebral oxidative stress, we investigated the implication of NOX-2, the isoform present in brain endothelial cells, microglia and astrocyte 7,19,24 , on NVC by using knock out mice (NOX-2 -/-). As obtain previously, IL-17A Rb administration in WT mice attenuated the response to whiskers stimulations by 24.9%

IL-17A
We then tested the efficiency of Tempol and NOX-2 to normalize the superoxide anion production induced by IL-17A.

DISCUSSION
Our major new findings are that IL-17A, through its receptor, IL-17RA, takes part in the NVC impairment induced by Ang II. This effect is obtained through a process mediated by the NOX-2-induced superoxide anion production in the hippocampus and the somatosensory cortex as demonstrated by a treatment with Tempol and deletion of the gene coding for NOX-2.
Ang II through the Ang II type 1 receptor (AT1R) signaling pathway is an important proinflammatory stimulus, triggering the production of cerebral and systemic proinflammatory cytokines 6,25-27 , chemokines 28 and reactive oxygen species 7,10,19,[29][30][31] . The presence of perivascular macrophages expressing AT1 receptors in the neurovascular unit led to the hypothesis that inflammatory factors may mediate the effect of Ang II. Perivascular macrophages depletion indeed rescued NVC in Ang II slow pressor hypertension 32 . The modulatory impact of cytokines on NVC was then demonstrated in the Ang model of hypertension 6 . In this study, the anti-inflammatory cytokine, IL-10, also prevented gliosis and cerebral oxidative stress 6 . Inversely, a systemic inflammatory state in mice characterized by higher circulating IL-17A levels and induced by a high salt diet, contributed to NVC impairment 16 . In the present study, blocking IL-17A or its receptor prevented the NVC impairment observed in Ang II slow pressor hypertension. Overall, these results suggest an important contribution of IL-17A in NVC impairment in models of hypertension. However, the mechanism by which IL-17 impairs NVC remains to be established.
Previous reports have shown an important link between the effect of Ang II on peripheral endothelial function and IL-17A 10,33 characterized by a lower increase in blood pressure in Ang II-induced hypertensive mice receiving IL-17A Ab or IL-17RA mAB administration 33 or in IL-17 -/mice 10 . Our results confirmed the reduction in blood pressure in mice receiving these treatments. However, these changes in blood pressure cannot explain the impact on NVC since it was previously demonstrated that the impact of Ang II on NVC is independent of its hypertensive effect 8,27 .
8 A more probable mechanism by which Ang II exerts its deleterious actions on NVC would be by activating its AT1R, at least on endothelial cells and perivascular macrophages 7,32 , which subsequently lead to the production of reactive oxygen species 7,19,24 . In the periphery, blood vessels from IL-17 −/− mice preserved vascular function, decreased superoxide production, and reduced T-cell infiltration in response to Ang II 10 . Therefore, we hypothesized that IL-17A could also modulate NVC in Ang II-induced hypertensive mice through a similar pathway. Our results confirm increased superoxide anion production in the hippocampus and somatosensory cortex in the slow pressor model induced by Ang II. Interestingly, neutralizing IL-17A or inhibiting its receptor normalizes the superoxide anion production. These results match those observed in the periphery by Madhur et al 10 . Overall, these findings suggest that IL-17A is involved in the superoxide anion production induced by Ang II, but that it is part of the mechanism by which Ang II impairs NVC.
To demonstrate that IL-17A can impair the cerebrovascular response, we tested the effect of the IL-17A Rb on NVC. We first showed that IL-17A Rb administration impaired NVC in a dose-

Systolic blood pressure monitoring
Systolic blood pressure was monitored in awake mice using tail-cuff plethysmography (Kent Scientific Corp, Torrington, CT). Mice were warmed on a heating pad preheated at 37 °C for ten minutes before and during blood pressure recordings. Animals were habituated to the procedure three days before blood pressure assessment. Right before the implantation of osmotic minipump (day 0) and weekly until the NVC analysis, ten blood pressure assessment per mice were done and average. Blood pressure was monitored by the same person at the same time of the day.
(MilliporeSigma, USA). The depth of anesthesia was checked by testing corneal reflexes and motor responses to tail pinch. Mean blood pressure and blood sample collection for gases 13 assessment were monitored through the catheterization of the femoral artery. Mice were artificially ventilated with a nitrogen/oxygen/CO2 mixture through a tracheal intubation. Body temperature was maintained at 37 ºC throughout the experiment. CBF was monitored with a laser-Doppler probe (AD Instruments, USA) placed stereotaxically on the thinned skull above the whisker barrel area of the somatosensory cortex. The flowmeter and blood pressure transducer were connected to a computerized data acquisition system (MacLab; Colorado Springs, CO). Analysis of CBF responses began 30 minutes after the end of the surgery to allow blood gases to stabilize. Animals with a mean arterial blood pressure under 60 mmHg and/or blood gases outside normal range (pH: 7.35-7.40; pCO2: 33-45; pO2: 120-140) were eliminated from the study. CBF responses to neuronal activity were evaluated by whiskers stimulations.
Three whiskers stimulations sessions of one minute at 6 Hz were done on the contralateral side to CBF measurement. Three minutes resting periods were left between each stimulation. CBF values were acquired with the LabChart6 Pro software (v6.1.3, AD Instruments, USA). The percentage increase in CBF represents the peak CBF response relative to the resting CBF peak values during the 20 seconds before stimulations.

Superoxide anion production
Superoxide anion production was assessed by hydroethidine microfluorography as previously described 41 . Hydroethidine (dihydroethidium) is cell permeable and is oxidized to become the fluorophore ethidium bromide that intercalates in double-stranded DNA 42 . Mice were anesthetized with sodium pentobarbital (100 mg/kg body weight, CDMV) and perfused transcardially with PBS, pH 7.4. Brains were carefully isolated, frozen on dry ice and stored at -80°C until further analysis. Frozen brains were cut with a cryostat (20µm) and brain sections were mounted on slides and stored at -20°C overnight. The slides were air dried at room temperature for 15 min followed by 15 min on a slide warmer set at 45°C. The slides were then immersed in a dihydroethidium (DHE) solution (2µM, MilliporeSigma) dissolved in PBS at 37°C for 2 min. The slides were rinsed in PBS for 5 min and dried on a slide warmer for 20 minutes before they were coverslipped with Fluoromount-G mounting medium (Southern Biotech, USA). Images were acquired using an epifluorescence microscope Leica DM2000 with the same acquisition parameters for all groups. Analysis of relative fluorescence intensity was done using the Image J software (version 1.53; National Institutes of Health). DHE results are expressed relative to the control group.

Statistical analysis
Data analysis was performed with GraphPad Prism software (version 7.0, La Jolla, USA) and results are presented as mean ± SEM. CBF responses to whiskers stimulations, resting CBF, superoxide anion production and systolic blood pressure analysis were evaluated with an ANOVA for factorial design with repeated measures followed by a Bonferroni post-test for multiple group comparisons. The dose-response whiskers stimulations curve, resting CBF and mean arterial pressure from supplemental Figure 1 were analysed using a one-way-ANOVA followed by Dunnetʼs post-test. Significance was set at p<0.05. Sample size per group is presented in the results section as well as in the figure legends.