Genetic Ablation of Prorenin Receptor in the Rostral Ventrolateral Medulla Influences Blood Pressure and Hydromineral Balance in Deoxycorticosterone-Salt Hypertension

Abstract Non-enzymatic activation of renin via its interaction with prorenin receptor (PRR) has been proposed as a key mechanism of local renin–angiotensin system (RAS) activation. The presence of renin and angiotensinogen has been reported in the rostral ventrolateral medulla (RVLM). Overactivation of bulbospinal neurons in the RVLM is linked to hypertension (HTN). Previous studies have shown that the brain RAS plays a role in the pathogenesis of the deoxycorticosterone (DOCA)-salt HTN model. Thus, we hypothesized that PRR in the RVLM is involved in the local activation of the RAS, facilitating the development of DOCA-salt HTN. Selective PRR ablation targeting the RVLM (PRRRVLM-Null mice) resulted in an unexpected sex-dependent and biphasic phenotype in DOCA-salt HTN. That is, PRRRVLM-Null females (but not males) exhibited a significant delay in achieving maximal pressor responses during the initial stage of DOCA-salt HTN. Female PRRRVLM-Null subsequently showed exacerbated DOCA-salt-induced pressor responses during the “maintenance” phase with a maximal peak at 13 d on DOCA-salt. This exacerbated response was associated with an increased sympathetic drive to the resistance arterioles and the kidney, exacerbated fluid and sodium intake and output in response to DOCA-salt, and induced mobilization of fluids from the intracellular to extracellular space concomitant with elevated vasopressin. Ablation of PRR suppressed genes involved in RAS activation and catecholamine synthesis in the RVLM but also induced expression of genes involved in inflammatory responses. This study illustrates complex and sex-dependent roles of PRR in the neural control of BP and hydromineral balance through autonomic and neuroendocrine systems. Graphical abstract


Introduction
Hypertension (HTN) is considered one of the most important health issues of the last few decades due to its high incidence and serious complications. According to the World Health Organization and the Centers for Disease Control and Pr ev ention, HTN affects more than 1.3 billion people w orldwide , including more than 100 million adults in the United States. 1 , 2 Importantly, HTN is responsible for 85 million deaths worldwide from heart disease, stroke, and kidney disease. Despite the wide variety of anti-hypertensive drugs available, there are still a large number of patients that are refractory to these conventional therapies. 3 , 4 The renin-angiotensin system (RAS) is one of the most studied mechanisms of b lood pr essur e (BP) regulation and an important pharmacological target widely used in clinical settings. Renin is the rate-limiting enzyme for angiotensin (ANG)-II production, the primary bioactive product of the RAS. However, only 15% of patients with essential HTN exhibit increased plasma renin activity. 5 Indeed, approximately 60% of patients with resistant HTN exhibit low plasma renin. 6 In addition to the classical circulating RAS, the tissue-specific RAS operates locally in the brain, heart, v asculatur e, adr enal gland, r e pr oducti v e tract, kidne ys, and man y other organs (re vie wed in 7 ). The existence of the brain RAS and its importance in BP regulation, hydromineral balance, autonomic function, and metabolism is well accepted. [8][9][10][11] Dysregulation of the brain RAS pla ys a ke y mechanistic role in the development and maintenance of high BP and energy balance in models exhibiting low plasma renin, such as DOC A-salt HTN . 12 This implies the existence of de novo generation of ANG-II within brain regions implicated in car dio vascular and metabolic control. Angiotensinogen (AGT), the only known renin substr ate , is constitutively expressed in astrocytes and some neurons. 13 Indeed, the brain's extracellular space is "bathed" in AGT. However, it remains contr ov ersial whether r enin, the key limiting enzyme of the ANG-II biosynthetic cascade, is necessary for the cleavage of AGT to generate ANG-I in the brain. 14 , 15 In the kidney, renin is translated as an inacti v e pr oenzyme (pr or enin) that r equir es hydr ol ytic cleav a ge of its prosegment to attain its full enzymatic activity . Intriguingly , evidence suggests that the enzymes responsible for the hydrolytic cleavage of prorenin may be absent in the brain. 16 , 17 In the last decade, several studies have demonstrated that pr or enin r ece ptor (PRR), a tr ansmembr ane protein encoded by Atp6ap2 , has the capacity to bind pr or enin and induce a nonenzymatic acti v ation of r enin without pr osegment cleav a ge. Studies hav e r e ported that expr ession of PRR in differ ent r egions of the brain participates in the regulation of BP and fluid balance through RAS-dependent mechanisms, and lack of PRR prevents the development of HTN in several preclinical models. [18][19][20] Howev er, r ecent studies hav e identified many other functions of PRR unrelated to the regulation of the RAS. For example, PRR is implicated in cellular processes, such as endosomal trafficking, autopha gy, neur oinflammation, and RAS-inde pendent signal transduction, that inv olv e differ ent mitogen-acti v ated pr otein kinases. [21][22][23][24] Thus, whether PRR mediates its effects through the acti v ation of the RAS r emains elusi v e.
Our la borator y has pr eviousl y identified the existence of r enin-expr essing neur ons in close pr o ximity to AGT-e xpressing cells within the rostral ventrolateral medulla (RVLM) region. 25 This finding was not surprising as the RVLM is a region of the brainstem that plays a critical role in regulating car dio vascular function and BP via the sympathetic nervous system. The RVLM has been implicated in several car dio vascular disor ders, including HTN, heart failure, and arrhythmia. 26 Thus, we hypothesized that expression of PRR in the RVLM is r equir ed for the full development and maintenance of HTN through activation of the brain RAS. To test the physiological impact of PRR in the RVLM, compr ehensi v e car dio v ascular, r enal, and meta bolic studies wer e conducted in mice lacking PRR in the RVLM during DOCA-salt HTN . This stud y demonstrates a complex sex-specific role of PRR in the RVLM during the development and maintenance of DOCAsalt HTN.

Methods
The protocols for acute studies, neuronal cell culture experiments, telemetry recordings, power spectral analyses, biochemical assays, in situ hybridization, histological studies, acute saline c hallenge , tr anscutaneous glomerular filtr ation r ate (tGFR), non-inv asi v e pulse-w av e v elocity (PWV), n uclear ma gnetic resonance (NMR), bioimpedance spectroscopy (BIS), western blotting, and bulk RNA sequencing are described in the Supplemental Online Methods.

Animal Subjects
This study was conducted using C57BL/6J mice (Stock 000664, J ackson La boratories, ME) and mice carrying floxed Atp6ap2 encoding PRR, which was generously provided by Frederique Yiannikouris (Uni v ersity of K entuck y) with permission fr om Gene vie ve Nguyen, who developed the mouse model. 27 , 28 Both males and females were 12-14 wk of age at the onset of the experiments. Mice were provided standard la borator y chow (Envigo/Teklad #2920, IN), and chlorinated tap water was availa b le ad libitum during the experimental pr ocedur es. All animals were housed at standard room temperature ( ∼22 • C) under a 14:10 light-dark cycle (light onset at 5 am ). Surgical and experimental pr ocedur es adher ed to the National Institutes of Health's "Guide for the Care and Use of Laboratory Animals" and were approved by the Medical College of Wisconsin and the Uni v ersity of Iow a Animal Car e and Use Committees.

RVLM-Targeted Microinjection
Radiotelemetry-instrumented mice were anesthetized with isoflurane inhalation (1.5-2.0%) and secur el y placed in a stereotaxic fr ame . The RVLM coordinates were calculated using the Paxinos and F r anklin mouse br ain atlas: 1.25 mm mediolateral, 1.8 mm caudal to lambda, and 6.4 mm v entral fr om the dorsal surface of the skull. 29 Once the coordinate were localized, a 1.0 mm diameter hole was made using a hand drill to penetrate the skull and expose the brain. A 32-gauge Neuros syringe (Hamilton, NV) was used to deliver drugs or viruses. Before tr eatments, the corr ect placement of the injector was confirmed by observing a pr essor r esponse to L-glutamate (Sigma Aldrich, MO).

Conditional Ablation of PRR in the RVLM
Homozygous females and hemizygous males carrying exon 2floxed PRR allele in the X chromosome (PRR f/f and PRR f/y , respecti v el y), with or without a tdTomato Cr e r e porter allele (ROSA) were used. First, to validate the animal model, we bred PRR f/y with Ai14, tdTomato Cre reporter line (Stock #007914, Jackson Laboratories). These mice were subjected to stereotactic bilateral microinjections of either adenovirus (ad)-green fluorescent protein (GFP) or Ad-CRE recombinase fused with GFP using the coordinates described a bov e. After 2 wk, the targeted region was dissected under a fluorescent stereo microscope. Subsequentl y, DNA w as extracted fr om one-half of the collected tissues using the PicoPure DN A Extr action Kit (catalog # KIT0103, Applied Biosystems) and subjected to PCR validation. Excision of exon 2 was confirmed using PCR amplification with the following set of primers: 5 -A GCA CTCTCTTCCA GGT A TGTTGTG-3 and 5 -GCCCCTCTCTTA CA GTTCTATCA GT-3 . The presence of a 326 bp product indicated excision of PRR. Cycling conditions were the following: 94 • C for 3 minutes; 2 cycles of 94 • C for 1 minute, 62 • C for 1 minute, 72 • C for 1 minute; 30 cycles of 94 • C for 30 s, 62 • C for 30 s, 72 • C for 30 s, and 1 cycle of 72 • C for 1 minute. The other half of the tissues were homogenized with RIPA buffer containing phen ylmeth ylsulfon yl fluoride, sodium orthov anadate, and pr oteinase inhibitors (Catalog # sc-24948, Santa Cruz Biotechnology, TX), sonicated, and incubated at 4 • C for 30 minutes. The homogenate was centrifuged at 13 000 g for 10 minutes, and the supernatant was stored at -80 • C until protein analysis by western blotting, as described in the supplemental methods section.
In another set of animals, the RVLM of PRR f/y was targeted with either adeno-associated virus (AA V) 2/2-CMV -GFP or AA V2/2-CMV -CRE-GFP, and br ain punc hes from the RVLM were collected. These tissues were subjected to total RN A extr action using TRIzol r ea gent (Thermo Fisher Scientific, NA), followed b y re verse transcription using the Superscript III r ev erse tran-scriptase kit (Invitrogen, CA). To detect PRR ( Atp6ap2 ) and 18s (housekeeping gene), a real-time PCR was performed using the Fast SYBR Green Master Mix (Invitrogen, CA) and the following primer set: Atp6ap2 : 5 -TCTCTCCGAA CTGCAA GTGCAA CA-3 and 5 -CC AAACCTGCC AGCTCC AA TGAA T-3 ; 18s : 5 -CGCTT CCTT ACCTGGTTGA T-3 and 5 -GAGCGACC AAAGGAACC A T A-3 , r especti v el y.

Chronic Studies in DOCA-Salt Model
Male and female 12-14-wk-old PRR flox mice or wildtype (WT) littermates wer e di vided into 4 gr oups: (a) male PRR f/y + AA V -GFP (male WT), (b) male PRR f/y + AA V -CRE-GFP (male PRR RVLM-Null ), (c) female PRR f/f + AA V -GFP (female WT), and (d) female PRR f/f + AA V -CRE-GFP (female PRR RVLM-Null ). The first cohort was instrumented with radiotelemetric devices for BP recordings, as described in the supplemental section. After 48 hours, viruses wer e deli v er ed bilaterall y into the RVLM as afor ementioned. Pr essor r esponse to a small dose of L-glutamate at the injection site was used to confirm the correct placement of the injector in the RVLM. At least 3 wk were allowed to induce expression of Cre. After a 3-d baseline recording, DOCA-salt HTN was induced by a subcutaneous implant of a 50 mg DOCA pellet (Deoxycorticosterone acetate, Cat #D7000, Millipore-Sigma, MO), after which 0.15 M NaCl saline (dissolved in deionized water) was offered as the only source of drinking water. BP, heart rate (HR), and activity were recorded for 21 d. Power spectral analysis of HR variability and diastolic BP v aria bility w as performed at baseline, day 3 (earl y sta ge), and day 13 (late sta ge) using HemoLa b softw ar e as pr eviousl y described. 30 , 31 A second set of animals was subjected to comprehensi v e meta bolic and hydr omineral balance phenotyping. Mice were single-housed in metabolic cages (item #52-6756, Hatteras Instruments model MMC100, Harvard Apparatus, MA) to assess food and fluid intake and for urine and feces collection at baseline and days 3, 6, and 13 on the DOCA-salt regimen. Twenty-four-hour urine samples were processed to evaluate urine electr ol ytes, catecholamines, aldoster one, and cope ptin. The urine sodium and potassium concentr ations w ere measured using a flame photometer (J enw ay PFP7 or a BWB XP). For nor e pine phrine/e pine phrine, aldoster one, and cope ptin measur ements, commerciall y av aila b le ELISA kits fr om Abnov a (catalog # KA1877, Taiwan), Cayman Chemicals (catalog # 501090, MI), and Cloud-Clone Corp (catalog # CEA365Mu, PRC) were used, r especti v el y.

Sta tistical Anal ysis
All data are presented as mean ± SEM. The data were analyzed using GraphPad Prism software (Version 9.5.0, Boston, MA). Two-way ANOVA or mixed-effects model with r e peated measures as appropriate, followed by Sidak's multiple comparison pr ocedur es or a 2-tailed unpaired t -test was performed. Survi v al rate curv es wer e anal yzed using the Logrank comparison test and the pairwise comparison of individual survival curves using the Kaplan-Meier method. Outliers were identified by either the interquartile range method and the ROUT method with Q = 5%. A P-value of less than 0.05 was considered significant. Individual data points were plotted in dot/whisker plots if appropriate. In all experiments, animals were randomly assigned into two experimental groups, either WT or PRR RVLM-Null . The systolic BP change following acute intracer ebr ov entricular infusion of PRR antagonist (PRO20), losartan, and control scramble peptide in C57BL/6J mice under deoxycorticosterone-acetate (DOCA)-salt-induced hypertension. Systolic BP was measured at baseline and during DOCA-salt administration. The change in systolic BP (from baseline, Systolic BP) to ICV drug infusion w as anal yzed by r e peated measur es two-w ay ANOVA with Dunnett m ultiple comparison post-hoc test. Time P < .01, Drug P = .12, and TxD P = .16. The area under the curve (AUC) of Systolic BP was calculated and analyzed b y one-wa y ANOVA followed b y the Tukey post-hoc test. One-way ANOVA P = 0.02, * P < .01 losartan versus scr amble , * * P < .01 losartan versus scr amble , and # P < .01 PRO20 versus scr amble . (B) Acute unilateral injection of artificial cer ebr ospinal fluid (aCSF), recombinant prorenin, or angiotensin II (ANG II) targeting the RVLM. The Systolic BP w as anal yzed by r e peated measur es two-w ay ANOVA with Dunnett m ultiple comparison post-hoc test. Time P = .01, Drug P = .15, TxD P < .01. The AUC w as anal yzed by one-w ay ANOVA followed b y Tuke y post-hoc test. One-wa y ANOVA P < .01, * P < .01 ANG II versus aCSF, and # P < .01 recombinant prorenin versus aCSF. After the treatments, fluor escent micr ospher es wer e deli v er ed, and the br ains w ere sectioned to confirm the correct placement of the injector in the RVLM region with e pi-fluor escent microscopy.

PPR is Expressed in the Brainstem and is Implicated in BP Control
To confirm the importance of PRR in the neurogenic control of BP, we evaluated whether pharmacological blockade of PRR in the br ain amelior ates esta b lished HTN in C57BL/6J mice subjected to DOCA-salt HTN. Similar to what was reported by Li et al. 32 , intracer ebr ov entricular (ICV) infusion of PRR antagonist (PRO20) elicited an acute reduction in systolic BP compara b le to ICV-losartan ( Figure 1 A). Given the premises that: (1) administration of ANG-II in the RVLM elicits pressor and autonomic responses 33 and (2) neurons with renin promoter acti vity ar e localized in close pr o ximity to AGT-e xpressing cells in the RVLM 25 , we hypothesized that PRR contributes to local RAS acti v ation in this brain r e gion. F irst, w e performed acute microinjections of recombinant prorenin targeting the RVLM in anesthetized mice. Vehicle and ANG-II were used as negati v e and positi v e contr ols, r especti v el y. RVLM-targeted administration of recombinant prorenin elicited a transient 7.5 ± 3.1 mmHg increase in systolic BP, which peaked at 80 s post-injection. In contrast, ANG-II elicited a 16.8 ± 3.4 mmHg increase , whic h peaked at 50 s post-injection. Vehicle did not elicit significant changes in BP ( Figure 1 B). These data confirm that local acti v ation of the RAS in the RVLM is sufficient to increase BP and also suggest that the RVLM likely contains the machinery to activate prorenin. Furthermore, administration of r ecombinant pr or enin in cultur ed brainstem neur ons elicited intracellular signaling acti v ation e videnced b y ele vated phosphor ylated extracellular signal-r egulated kinase 1/2 (ERK) normalized by total ERK ( Figure S1A) and incr eased super oxide pr oduction (Figur e S1B). Pr otein anal ysis of brain homogenates confirmed that PRR is expressed in the hypothalamus, brainstem, and cortex ( Figure 2 A). Furthermore, the multicolor in situ mRN A hybridization tec hnique r ev ealed that PRR is expr essed in neuronal and microglial cell populations within the RVLM ( Figure 2 B). We also observed that PRR is expressed in both inhibitory GABAergic-and excitatory glutaminergic neurons (   PRR RVLM-Null exhibited a further increase in systolic BP compared to female WT. There was no difference in HR or activity between these groups. Since only females lacking PRR exhibited a prominent phenotype, all subsequent studies only employed females. PRR RVLM-Null exhibited a significant increase in pulse pr essur e upon DOCA-salt ( Figure S3A). Therefore, we sought to measure PWV as an indication of arterial stiffness. As expected, 13 d of DOCA-salt treatment induced a significant increase in pulse w av e v elocity in both gr oups, but no significant difference w as observ ed betw een PRR RVLM-Null and WT (F igure S3B).
To evaluate autonomic balance to the heart, power spectral anal yses of HR v aria bility wer e performed at baseline (day 0), the early stage (day 3), and the late stage (day 13) on DOCAsalt in female WT or PRR RVLM-Null . Both groups exhibited an elevated low frequenc y/high frequenc y ratio (which estimates a r atio betw een the sympathetic to parasympathetic activity) in the early stage and a low ratio in the late stage. Howev er, ther e wer e no significant differ ences among the gr oups ( Figure 5 A). Supporting these observations, PRR RVLM-Null did not exhibit any significant difference in response to either βadr energic or m uscarinic b lockade with pr opranolol or atr opine, r especti v el y ( Figur e 5 B). PRR RVLM-Null exhibited a diastolic BP v aria bility (a surr ogate of sympathetic outflow to resistance arterioles) of 3.67 ± 1.06 compared to 1.99 ± 0.46 in the WT.
Unexpectedl y, this differ ence did not r each statistical significance ( Figure 5 C). Given the variability in these analyses, we sought to measure norepinephrine and epinephrine in urine samples collected at day 13-14. Consistent with the previous findings, female PRR RVLM-Null exhibited significantly elevated urinar y nor e pine phrine ( Figur e 5 D). At day 21, mice wer e sacrificed, and tissues were weighed and collected for further analysis. No differences in the relative kidney, heart, and spleen weight were found among the groups ( Figure S4). We next evaluated whether PRR RVLM-Null exhibited an impaired capacity to excrete sodium. We gave an intraperitoneal injection of isotonic saline (0.9% NaCl) equi v alent to 10% of body mass in a separate cohort of mice that were not instrumented with radiotelemeters. Inter estingl y, PRR RVLM-Null showed an impaired capacity to excrete the excess sodium at baseline and in the late stage of DOCA-salt ( Figure 6 A). Thus, we hypothesized that PRR RVLM-Null might exhibit either impaired renal function or a sodium-depleted state. To evaluate renal function, we measured GFR by transcutaneous fluorescein isothiocyanate-sinistrin elimination method. No difference in tGFR w as observ ed between groups either at baseline or after DOCA-salt ( Figure 6 B). Then, we evaluated whether expression of transporters affecting renal sodium r ea bsorption w as alter ed. Total and phosphorylated expression of various tubular sodium tr ansporters w er e measur ed by western b lotting. Contrar y to our hypothesis, we observed that total and phosphorylated NCC w ere par adoxically downre gulated in PRR RVLM-Null . No difference in NKCC2, total and phosphorylated NHE3, and αENaC were observ ed ( Figur e 6 C and D, Figur e S4).

Genetic Ablation of PRR Targeting the RVLM Augments the Effect of DOCA-Salt on Hydromineral Intake and Output and Mobilizes Fluid Between Intracellular and Extracellular Compartments
Based on the observations that female PRR RVLM-Null exhibited normal tGFR and downregulated phosphorylated and total NCC, we concluded that lower sodium excretion in response to acute saline challenge might not be due to impaired renal function. We hypothesized that lower sodium excretion after acute load occurs as a response to a sodium and water deficient state, where physiological systems attempt to retain sodium and water to reac h hydrominer al balance . Therefore , a new cohort was employed to assess fluid and sodium balance at differ ent sta ges of DOCA-salt. Mice wer e placed in meta bolic cages for a comprehensive 24-h evaluation of sodium input versus output. At baseline, fluid and sodium intake, urinary volume, and sodium excretion between WT and PRR RVLM-Null did not differ. Administration of DOCA-salt induced a prominent increase in sodium and fluid intake as well as urine volume and sodium excretion in both groups. However, female PRR RVLM-Null exhibited an exaggerated increase in fluid and sodium intake and urinary volume and sodium excretion to DOCA-salt ( Figure 7 A and B). No major differences in potassium excretion, sodium balance, plasma sodium, and potassium concentr ation w er e observ ed ( Figur e 7 C, D, and E). Then, mice were subjected to combined time-domain NMR and BIS to assess body mass and composition, total body water, and fluid distribution betw een extr acellular and intr acellular spaces. 34 No differences between groups in body mass, fat mass, fat-free mass, food and caloric intake, and feces w eight w er e observ ed at baseline. DOCA-salt administration in both groups induced a transient increase in fat mass concomitant with a decrease in fat-free mass and total body water, which normalized after day 13 ( Figure 8 A). No difference in food and caloric intake and feces weight was observed ( Figure S6). Although no difference in body water was observed among groups, the NMR-BIS anal ysis r ev ealed that a b lation of PRR in the RVLM results in an imbalance between extracellular and intracellular fluids. That is, PRR RVLM-Null exhibited a lower extracellular and higher intracellular fluid ( Figure 8 B) when compared to WT. We hypothesized that an imbalance betw een extr acellular and intracellular fluid might be sufficient to trigger v asopr essin r elease. Thus, urinar y cope ptin, a meta bolicall y sta b le surr ogate marker of v asopr essin r elease w as quantified. Supporting our hypothesis, 24 h-copeptin lev els wer e significantl y elev ated in PRR RVLM-Null compared to WT suggesting an elevated vasopressin r elease ( Figur e 8 C). Ther e w as no differ ence in urinar y aldoster one ( Figur e 8 D).

Tr anscriptomic Anal yses Re veal Canonical Pa thways and Putati v e Gene Candida tes Leading to Exagger a ted Responses to DOCA-Salt in PRR RVLM-Null
Using the P alko vits br ain punc h tec hnique , w e obtained total RNA from female RVLM at baseline and after 13 d of DOCAsalt. Samples were subjected to bulk RNA sequencing to identify putati v e candidates that could e xplain physiolo gical outcomes. Differ ential expr ession anal yses using the generalized-linear r egr ession appr oach r ev ealed an upr egulation of 434 genes and downregulation of 128 genes in DOCA-salt HTN compared to the baseline control. Ablation of PRR in DOCA-salt induced upregulation of 258 genes and downregulation of 147 genes compared to mice targeted with a control virus ( Figure S7A-C). RNAseq gene expression data were then analyzed using Ingenuity Pathw ay Anal ysis to infer the influence of PRR deletion on biological networks and signaling pathways. The top 5 canonical pathways and upstream regulators are listed in Figure S7D and E. Then, we sought to selecti v el y ev aluate key candidate genes indi viduall y. First, as expected, ther e w as a tr end tow ard a decrease ( P = .08) in PRR expression in DOCA-salt-treated PRR RVLM-Null compar ed to WT ( Figur e 9 A). Inter estingl y, ther e was no difference at baseline, presumably due to insufficient time to induce a complete extinction of mRNA from every cell in the sample that express PRR. In WT mice, DOCA-salt incr eased expr ession of sev eral RAS genes, including Agtr1a , catec holamine synthesis ( T h , Dbh , Pnmt ) and serotonin biosynthesis ( Tph2 , Slc6a4 , Aaad ) ( Figure 9 A and B). PRR a b lation pr ev ented the effects of DOCA-salt on several of these genes. Inter estingl y, PRR RVLM-Null exhibited upregulation of Ace2 , considered a protecti v e arm of the RAS . Moreo ver, the lack of PRR in the RVLM r esulted in elev ated genes inv olv ed in inflammator y r esponses, including Nfkb1 , Nlrp3 , P22phox , and Il18r1 ( Figure 9 C). These data might indicate that a b lation of PRR in the RVLM is sufficient to suppress local RAS activation and induction of catecholamine and serotonin biosynthesis in the RVLM by DOCA-salt. Howev er, the acti v ation of neur oinflammator y signals might trigger the exacerbated fluid and pressor responses to DOCA-salt in PRR RVLM-Null . Kidney l ysates fr om DOCA-tr eated mice wer e subjected to western blot analysis of p-NCC, total NCC, NKCC2, p-NHE3, total NHE3, and αENAC. A r e pr esentati v e b lot is shown. Complete blots and loading controls are provided in supplemental Figure S5. (D) Quantification of bands was performed using Image J. Data are expressed as mean ± SEM. Data were analyzed using an unpaired t -test. * P < .05 and * * * P < 0.0005 versus WT + DOCA.

Discussion
The main finding of this study is that the localized a b lation of PRR in the RVLM region results in a sex-dependent phenotype c har acterized by exacerbated dipsogenic, natriuretic, and pressor responses to DOCA-salt during the late "maintenance" phase of the model. In addition, this phenotype was also associated with e xacerbated dipso genic r esponse, decr eased r elati v e distribution betw een extr acellular and intr acellular fluid compartmentalization, elev ated urinar y cope ptin (an index of arginine v asopr essin), and incr eased sympathetic tone to the kidney and resistance arterioles.
The physiological outcomes of brain RAS acti v ation ar e highl y de pendent on the neur al cir cuits inv olv ed. This study was focused on the RVLM, a key region of car dio vascular regulation and sympathetic contr ol, wher e we pr eviousl y identified r enin pr omoter acti vity. 25 The RVLM r egion is situated within the blood-brain barrier; although it is unlikely that circulating ANG-II can reach the RVLM neurons directly-especially in the context of DOCA-salt HTN where circulating levels of ANG-II are low-the RVLM remains subject to trans-endothelial signaling and thus potential ANG-II transcytosis. 7 , 35 , 36 The PRR has been proposed as a key player in the generation of ANG-II in several tissues by facilitating the non-pr oteol ytic activ ation of (pr o)r enin. 37 Studies designed to inv estigate the specific role of PRR in specific brain regions such as the subfornical organ or paraventricular nucleus emerged recently. 18 , 38 , 39 Using in situ hybridization tec hnique , w e observed that PRR is widel y expr essed in RVLM and is present in a wide variety of cell types, including glutaminergic and GABAergic neurons as well as in microglia. Stereotactic injections of recombinant mouse pr or enin targeting the RVLM r esulted in acute pr essor r esponses, indicating that the RVLM region contains the machinery to activate and respond to e xo genous prorenin. Mor eov er, r ecombinant pr or enin elicited intracellular signaling and generation of r eacti v e oxygen species in brainstem neurons in vitro. Whether these effects of pr or enin ar e RAS-de pendent still needs to be evaluated. Then, we ster eotacticall y deli v er ed AA V -CRE into the RVLM of mice carrying floxed PRR alleles to perform region-specific gene deletions. Since previous studies r e ported that the functions of PRR ar e sex-de pendent, we included both sexes in this study. [40][41][42] Likewise, sex differences in the development of DOCA-salt HTN include dimorphism in BP, v asopr essin r elease, v ascular r eacti vity, inflammator y pr ofile, bar or eflex function, and importantly the brain RAS activity. [43][44][45][46][47][48] Similarly, evidence supports that development of DOCAsalt HTN is influenced by the estrous cycle. 49 We acknowledge this r e pr esents a potential confounding factor that was not considered in the study. Contrary to our expectations, males lacking PRR did not exhibit any difference in BP compared to the WT control upon DOCA-salt c hallenge . In contrast, upon DOCAsalt stimulation, female PRR RVLM-Null mice displayed significantly lower diastolic BP during the early phase (first week). Subsequently, female PRR RVLM-Null mice on DOCA-salt displayed an exacerbated increase in systolic BP during the late "maintenance" phase (after the second week). These experiments led us to two main conclusions. First, the unexpected exacerbation of BP increases after the dela yed de velopment of DOCA-salt HTN in female PRR RVLM-Null mice suggests that differ ent r oles of PRR might exist depending on the cell type and brain region inv olv ed. Second, it is well-accepted that the development of HTN to DOCA-salt inv olv es differ ent physiological mechanisms that change throughout time. During the initial days of DOCAsalt treatment, fluid retention due to an imbalance in renal sodium handling concomitant with excessi v e salt intake is critical for the full esta b lishment of HTN. Then, the slow increase of BP and sustained HTN, which occur during the next few weeks, is thought to be mediated by neural and humoral factors, including elevated sympathetic tone to resistance arterioles and v asopr essin r elease, r especti v el y. 50 We consider ed that in the model DOCA-salt HTN, the role of PRR in the RVLM is bi-phasic gi v en that RVLM-specific a b lation of PRR in the earl y dev elopmental sta ge of DOCA-salt HTN is pr otecti v e, while during the late maintenance stage, seems to be detrimental. Importantly, the increase in BP in the female PRR RVLM-Null mice might be linked to sympathoexcitation, as the power spectral analysis of diastolic BP variability was elevated in these animals. We acknowledge that pow er spectr al analysis is an indir ect measur ement of sympathetic dri v e. Thus, alternati v el y, we also measured urine norepinephrine, which was elevated in the female PRR RVLM-Null mice. Ideally, renal nerve recording might confer additional direct measurement of sympathetic dri v e . How ever, this tec hnique becomes a limitation gi v en the need for anesthetics and the size restraint of mice. An alternati v e mechanism causing elevated BP can be explained by the observation that female PRR RVLM-Null mice exhibited exacerbated salt and water intake in response to DOCA-salt. Ther e w as no significant change in the daily sodium balance, pr esuma b l y because ther e w as a compensator y r eduction in the expression and activity of sodium transporters, leading to elev ated urinar y excr etion. It has been r e ported that conditions such as DOCA-salt HTN induces a shift of fluid from the extracellular to intracellular spaces. 51 Similarly, our study r ev ealed that females lacking PRR showed a reduction of fluid in the extracellular and increase in the intracellular compartment. Cianci et al. have reported that the increase in total body water content in hypertensi v e patients is attributa b le to the expansion of the intracellular fluid. Mor eov er, intracellular w ater corr elates with BP. 52 It has been also indicated that hypertensi v e patients also exhibit a reduction in the extracellular fluid volume. 53 Gi v en that extr acellular dehydr ation stim ulates drinking and v asopr essin release, it might be possible that the altered fluid distribution explains in part the elev ated urinar y cope ptin lev els and exacerbated drinking response to DOCA-salt in PRR RVLM-Null mice. 54 To identify key genes affected by the r emov al of PRR in the RVLM, we employed transcriptomic approaches in brain punches collected from the RVLM of female WT or PRR RVLM-Null mice at baseline and during the maintenance late phase of DOCA-salt. Using Ingenuity pathway analysis, which allows a compr ehensi v e gene expr ession data anal ysis to infer the underlying causes of the observed changes, we identified three important canonical pathways that could be implicated in this phenomenon: (1) G-protein coupled r ece ptor signaling, (2) DNA methylation and transcriptional r e pr ession, and (3) phagosome formation. PRR appears to be essential for the maintenance of critical cellular functions as it is r equir ed for the assemb l y of the lysosomal proton-transporting V-type ATPase, protein de gr adation, and pha gol ysosome formation. 55 Inter estingl y, we identified POU4F1 as the main upstr eam r egulator. It has been shown that POU4F1 plays a key role in the development of the nucleus ambiguus, and mice lacking POU4F1 die prematur el y due to defects in this nucleus. 56 Thus, the absence of PRR might be lethal for certain cell types within the RVLM. Future studies that use advanced techniques to ablate specific cell types and neuronal populations within the RVLM region, such as the nucleus ambiguus, would be r equir ed. Differ ential analysis of genes that are associated with the regulation of the RAS and sympathetic activity in the RVLM as well as inflammator y cascades wer e performed. Although female PRR RVLM-Null mice exhibited higher BP when the RVLM punches were collected, we observed a paradoxical decrease in the expression of "detrimental" RAS genes and elevated expression of "protective" RAS genes. This implies that despite the a b lation of PRR resulting in the local inhibition of the RAS in the RVLM, this protecti v e effect was insufficient to counterbalance the hypertensive responses elicited by alternative mechanisms. Moreover, the a b lation of PRR upr egulated genes associated with inflammation. Ther efor e , w e hypothesize that a b lation of PPR might target POU4F1 and impair mechanisms inv olving G-pr otein signaling, DNA methylation, and phagosome formation leading to a localized inflammatory response in the brainstem. We acknowledge that ther e ar e certain limitations. First, the RVLM is not a defined anatomical structure with clear boundaries and is defined as Figure 9. Bulk RNA sequencing data analysis of RVLM brain punches collected from WT and PRR RVLM-Null females at baseline or day 13 with DOCA-salt hypertension. Expr ession pr ofiles of key genes inv olv ed in (A) RAS acti v ation, (B) catecholamine and ser otonin biosynthesis, and (C) neur oinflammation ar e shown. Data ar e expr essed as mean ± SEM and were analyzed using 2-way ANOVA followed by Sidak's multiple comparison post-hoc test. * P < .05, * * P < .01 and * * * P < .001. a region within the medullary reticular formation. This region comprises sev eral subr egions and heterogeneous cell types, including the C1 adrenergic cell group, cholinergic neurons, and many others. 26 Importantly, PRR is expressed in almost every cell type, including the microglia within the RVLM. 24 Studies have shown the presence of PRR in other cell types in the brain. For instance, PRR is expressed in astrocytes and plays a role in oligodendrogenesis and endothelial cell polarity. Thus, we consider a possible role for PRR to not only be expressed in neurons and microglia but also in astrocytes, oligodendrocytes, and endothelial cells. Although this was not directly measured, analysis of genes related to the functions of these cells by RNA sequencing did not suggest significant alterations. 21 , 57 Gi v en the lack of clear anatomical boundaries, the histological confirmation of the corr ect deli v er y of the viruses has been challenging.
Ther efor e , tr ansient pr essor r esponses to small amounts of glutamate, an excitatory neurotransmitter, were used as criteria to confirm the correct injection site. In addition to the AA V -GFP contr ol gr oup, we consider ed other contr ols such as off-target AA V -cre injections into areas in the brain presumed to play no role in BP or AA V -cre injections in wildtype mice. However, we felt that these additional experiments would not add significant impact to the study as the results would largely be negati v e. We consider it likely that complete a b lation of PRR was not achiev ed. Our pr evious experience in other br ain re gions indicates that the efficiency of gene recombination never reaches 100%. In this study, the reduction in PRR expression in the RVLM punches was ∼30%, but it did not reach statistical significance ( P = .08). Of note, however, the area of the brain punch is relati v el y bigger than the RVLM; thus, some surrounding tissue that did not undergo viral transduction and thus cannot undergo PRR a b lation w as sampled. Thus, the decrease in PRR expression is likely an underestimate, perhaps lar gely so . Additionally, we recognize that the serotype of the AAV used might not possess the most efficient tropism to infect the variety of cells comprised in the RVLM. Finally, we could not conduct cell-specific a b lation of PRR in the RVLM region. Given the ubiquitous expression of PRR and the heterogeneity of the RVLM region, it is likely that cell-specific a b lation of PRR could lead to different physiological outcomes. Future studies will address whether conditional ablation of PRR in cholinergic neurons, catecholaminergic C1 neur ons, non-C1 neur ons, or glial cells recapitulates the phenotype of PRR RVLM-Null mice.
In conclusion, we demonstrated that PRR in the RVLM region plays an important role in BP and hydromineral balance in a model of low plasma renin HTN. Importantly, the role of PRR is sex-dependent and cell-type-specific. Given the extensive and ubiquitous expression of PRR in brain regions of car dio vascular regulation, it is of utmost importance to consider that there might be off-target effects in ther apeutic str ate g ies targ eting PRR.