Role of the Renin Angiotensin System in Blood Pressure Allostasis-induced by Severe Food Restriction in Female Fischer rats

Severe food restriction (FR) is associated with blood pressure (BP) and cardiovascular dysfunction. The renin-angiotensin system (RAS) regulates BP and its dysregulation contributes to impaired cardiovascular function. Female Fischer rats were maintained on a control (CT) or severe FR (40% of CT) diet for 14 days. In response to severe FR, BP allostasis was achieved by up-regulating circulating Ang-[1–8] by 1.3-fold through increased angiotensin converting enzyme (ACE) activity and by increasing the expression of AT1Rs 1.7-fold in mesenteric vessels. Activation of the RAS countered the depressor effect of the severe plasma volume reduction (≥30%). The RAS, however, still underperformed as evidenced by reduced pressor responses to Ang-[1–8] even though AT1Rs were still responsive to the depressor effects of an AT1R antagonist. The aldosterone (ALDO) response was also inadequate as no changes in plasma ALDO were observed after the large fall in plasma volume. These findings have implications for individuals who have experienced a period(s) of severe FR (e.g., anorexia nervosa, dieters, natural disasters) and suggests increased activity of the RAS in order to achieve allostasis contributes to the cardiovascular dysfunction associated with inadequate food intake.

Severe food restriction (FR) causes abnormalities in the heart, vascular system and kidneys and includes hypotension, bradycardia, mitral valve prolapse and a prolonged QT interval 1,2 . Other medical complications include edema, electrolyte disturbances (e.g., hypokalemia and hypophosphatemia), nephrolithiasis and kidney failure 3 . Severe FR can be self-imposed as a result of eating disorders like anorexia nervosa (AN) 4 or in individuals who intentionally reduce their percent body fat for their occupation (e.g., models, athletes, dancers) or other reasons (e.g., dieters). Non-voluntary FR occurs in situations when people have inadequate access to food such as in war, natural disasters, periods of famine or low socio-economic conditions 5,6 . Severe FR can also occur in illnesses like cancer and AIDS due to inadequate food intake as a result of nausea or low palatability 7,8 . In these cases, the malnutrition as a result of the severe FR can independently add to the health complications of the disease.
We have shown that female Fischer rats maintained on a severe FR diet for 14 days (40% of a normal diet) developed malnutrition and shared numerous biochemical and physiological parameters found in humans after severe FR 9 . FR rats exhibited increased sensitivity to cardiopulmonary reflexes, the Bezold-Jarisch reflex, and α1-adrenoreceptor responses. We and others have also shown this model of FR leads to the development of cardiovascular abnormalities including ventricular hypertrophy and hypotension 9,10 .
The renin-angiotensin system (RAS) plays a critical role in blood pressure (BP) homeostasis and fluid and electrolyte balance and is a central regulator of cardiovascular and renal function. The activity of the octapeptide hormone angiotensin II (Ang- [1][2][3][4][5][6][7][8]) is widely targeted clinically in the treatment of hypertension and numerous cardiovascular and kidney diseases. In addition to its peripheral and central effects, Ang- [1][2][3][4][5][6][7][8] can act as a paracrine, autocrine, and intracrine hormone 11,12 . Ang- [1][2][3][4][5][6][7][8] exerts its vasoconstrictor effects by binding to Basal MAP and HR and plasma volume were lower in the FR group (Table 1) as we previously found 9,10 and as observed in self-imposed cases of FR in humans 1,2 . The hematocrit was higher in the FR compared to the CR group, which is also similar to what is found in individuals suffering from low food intake 18 . FR reduced urine volume by greater than 30% but had no effect on plasma ALDO or plasma potassium though plasma sodium showed a trend towards being elevated (Table 1). FR also caused reductions in sodium and potassium intake and excretion (U Na V, U K V) although net sodium (Na+ intake -UnaV) and potassium balance (K + intake -U K V) were maintained (Table 1). Under lengthier or more severe FR conditions, sodium and potassium balance can become impaired 3 , however, in our experimental conditions, we found FR for 14 days did not impair sodium or potassium balance. Thus, this protocol represents a moderate model of metabolic dysfunction.

α1-adrenergic receptor agonist responses.
To determine if FR selectively altered pressor responses mediated by the RAS, we studied the effect of FR on pressor responses to the α1-adrenergic receptor agonist L-phenylephrine. We found that FR had no effects on the pressor response to L-phenylephrine ( Fig. 9A,C); however, L-phenylephrine caused was a greater reduction in HR in the FR compared to CT group (P < 0.0001) (Fig. 9B,D).
Another major finding was that FR caused a coordinate up-regulation in both ACE and ACE2 activity. Many studies have shown ACE and ACE2 are regulated in a ying-yang manner. For example, ACE mRNA & protein expression was increased while ACE2 mRNA & protein expression was reduced in the paraventricular nucleus of the hypothalamus of the spontaneously hypertensive rat when compared to Wistar Kyoto normotensive controls 28 . However, other studies have found ACE and ACE2 can act coordinately. The mRNA & protein expression of both ACE and ACE2 were reduced by diabetic nephropathy in kidney tubules 29 . Although FR caused a   Table 2. Effect of food restriction on plasma angiotensin peptides before and after Ang- [1][2][3][4][5][6][7][8][9][10] infusion. Blood angiotensinogen (n = 8/group) and angiotensin (Ang) peptide concentrations before (CT, n = 10; FR, n = 11/ group) and after Ang-[1-10] infusion (n = 8/group) in female rats subjected to control (CT) or food restricted (FR) diet for 14 days. ‡ P < 0.05 vs CT by two-way ANOVA and Bonferroni post-hoc test (drug; diet); # P < 0.05 vs vehicle, same group , by unpaired Student's t-test. (<) in front of a number indicates below threshold of signal-to-noise; nd, not detectable.
coordinate up-regulation of activity of both the enzyme that produces Ang- [1][2][3][4][5][6][7][8] (i.e., ACE) and the enzyme that degrades Ang- [1][2][3][4][5][6][7][8] (i.e., ACE2), ACE is the predominant RAS catabolic pathway activated by the FR diet since plasma Ang-[1-7] was below the assay detection limit. Few studies to date have investigated the effects of caloric or protein restriction on ACE and ACE2 activity; however, one study in ewes showed that restricting caloric intake by 30% in the periconceptional period increased the abundance of ACE protein and mRNA 30 . In contrast, individuals with AN were reported to have lower plasma ACE activity than age-matched controls 31 , however, the calorimetric method used to measure ACE in this human study has substrate specificity limitations that could impact the interpretation of the data. Therefore, it is not known whether the discrepancy between the human and animal data (rats and ewes) is due to species differences in the effects of FR on ACE or to methodological differences in assaying ACE activity.
Ang- [1][2][3][4][5][6][7][8] mediates its physiological actions by binding to its receptors in target tissues including the vasculature. In the present study, we found that the FR diet up-regulated mRNA expression of the AT 1 R. Our findings  are supported by a study showing that a protein restricted diet increased AT 1 R protein in rat aorta vessels 15 . This up-regulation of AT 1 R expression could be a compensatory response to the BP reduction induced by FR by serving to increase vasoconstriction in resistance vessels. This idea is supported by the AT 1 R antagonist infusion studies since we found losartan lowered BP to a greater extent in the FR animals compared to the CT group indicating that the activity of AT 1 Rs is greater in the FR rats. These results support a previous report showing that losartan caused a greater reduction in MAP in rats maintained on a low protein diet compared to controls 15 .
Given that the FR diet had no effect on AT 2 R or masR mRNA expression, it is unlikely that these receptors play a major role in the hypotension and bradycardia observed in this model. Only a few reports have investigated the effect of food or nutrient deficiency on AT 2 R expression and no studies to date have reported effects on the masR. One study in pregnant ewes fed a 50% nutrient-restricted diet during gestation found that the offspring were hypertensive and expressed high AT 2 R mRNA expression in the renal medulla 33 , however, the effect of nutrient restriction on AT 2 R expression in the mother was not investigated. Thus, it is not known whether the lack of effect of a FR diet on AT 2 R mRNA expression in mesenteric vessels is specific to rodents, target tissue differences (mesenteric vessels vs. renal medulla), the nature of the diet or other experimental differences.
Similar observations were seen in a study of patients with AN compared to age-matched controls; patients with AN had to be infused with a higher concentration of Ang- [1][2][3][4][5][6][7][8] to achieve a 20 mmHg increase in BP 17 . Furthermore, in a mouse model of hypotension induced by severe intravascular volume depletion and decreased renal Na + reabsorption, aortic vessels exhibited decreased contractile reactivity 34 . This study suggested the cause of the reduced vasoconstriction was the reduced vascular hypertrophy.
Albumin plays a key role in controlling plasma osmolality besides plasma volume. Low plasma protein or sodium concentration will result in hypoosmolality causing water inside vessels to move into the intercellular space, resulting in edema and low blood volume 40 . Changes in blood osmolality and blood volume are well known to stimulate the vasoconstrictor arm of the RAS 41 . We previously showed that this FR model causes hypoalbuminemia 9,42 . Taken together, these studies suggest hypoalbuminemia contributes to both the hypotension and activation of the vasoconstrictor arm of the RAS in individuals experiencing inadequate levels of dietary protein as a result of AN, famine or other reasons 18 .
It is well known that the RAS can regulate the kidney function and BP. Another possible reason for the low BP in FR is due to the inability of the kidney to respond to the high circulating Ang- [1][2][3][4][5][6][7][8] levels. In animals with low plasma volume and intact RAS, Ang- [1][2][3][4][5][6][7][8] stimulates ALDO production and renal sodium and water reabsorption 43 . However, the FR rats in this study did not respond to the elevated Ang- [1][2][3][4][5][6][7][8] as evidenced by their net sodium balance ( Table 2) and ALDO levels ( Table 1). This indicates that kidneys of FR animals have become refractory to the actions of Ang- [1][2][3][4][5][6][7][8], the sum total of which results in reduced plasma volume and BP. One likely mechanism is desensitization of the AT 1 R, as it has been previously shown that mice with hypotension undergo renal AT 1 R desensitization 37 . Another result that can corroborate the possible AT 1 R desensitization is the normal basal plasma levels of ALDO in FR. It is well documented that Ang- [1][2][3][4][5][6][7][8] can act in the adrenal AT 1 R and release ALDO, which did not happened in FR rats. Another explanation for the lack of effect on plasma ALDO concentrations is a possible ceiling effect of calcium influx, which is necessary for the Ang-[1-8] ALDO release signaling 44 . Future studies will be needed to understand the mechanism on desensitization to Ang- [1][2][3][4][5][6][7][8] in FR.
As in patients with AN, HR was slower in the FR compared to CT animals. Bradycardia could be due to high parasympathetic output to the heart. Increased parasympathetic activation could explain the higher HR response after L-phenylephrine infusion in FR compared to CT animals due a change in the baroreflex. Furthermore, the HR response to Ang- [1][2][3][4][5][6][7][8][9][10] infusion was attenuated in the FR group. It is well known that in response to an acute increase in BP, HR is decreased by baroreflex stimulation in an attempt to bring BP back to normal 46 . Therefore, the Ang-[1-10]-induced increase in BP likely stimulates the baroreflex resulting in a compensatory decrease in HR. In conclusion, FR reduces plasma volume and BP. In order to maintain peripheral blood perfusion, the pressor arm of the RAS is activated increasing AGT and Ang- [1][2][3][4][5][6][7][8] levels, ACE activity and AT 1 R responsiveness. However, the RAS upregulation is inadequate to achieve full BP homeostasis. RAS activation is limited by attenuated Ang- [1][2][3][4][5][6][7][8] pressor responses, negative feedback to AGT and impaired ALDO responsiveness. Furthermore, once the period of severe FR ends, it is not clear if the RAS remains activated. If so, this up-regulation of the RAS may contribute to the long term adverse cardiovascular consequences observed in individuals who currently have or previously suffered from FR either voluntarily or not. Thus, it will be important to further investigate the mechanisms by which FR resets the RAS and the long term adverse consequences to the cardiovascular system.

Animals.
All experiments were conducted on female Fischer rats initially weighing 180-190 g at 3 months of age. Studies were conducted at both the Federal University of Ouro Preto (UFOP, Brazil) and Georgetown University (GU, USA). In Brazil, animals were bred in the University Center of Animal Sciences, UFOP. In the USA, animals were purchased from Envigo Corp. (Frederick, MD). All animals were housed in individual cages on a 12 hr light-dark cycle at room temperature (24 °C). The same person (AS) conducted the studies at UFOP and at GU. All of the experiments involving BP measurements were done at UFOP and all the samples processed for biochemical and molecular analysis were done at GU. For all studies conducted at GU, we confirmed that the FR diet had similar effects on BW, food and water intake, tissue weight, the estrous cycle, BP, and HR as we observed at UFOP. We chose to focus on a female model of severe food restriction because there are more women than men who experience periods of severe food restriction due to self-imposed diets and eating disorders 3 . Furthermore, while many studies have investigated the effects of malnutrition during pregnancy on the offspring 4,5 few have focused on the adverse consequences of severe FR to the mother prior to pregnancy.

Diet.
After receiving the animals, they stayed in the animal facility for 2 weeks to recover from potential travel stress and to enable acclimatization to the environment. After 2 weeks, the animals were single housed and food intake and BW were determined daily at 5:00 pm for each animal for two weeks. After this period, the animals were randomly divided into two groups: control (CT) or food-restricted (FR). The average food consumption determined for individual animals during the previous 2 weeks was used to calculate the amount of food given to the FR group, which was 40% of the average of their normal consumption (determined two weeks prior to the beginning of the study). BW was measured daily before replenishing the food, as we previously described 9 . The CT group had free access to food for the duration of the study period. All animals received a standard rat diet (Rodent diet 20, #5053, LabDiet, Marlborough, MA) and had ad libitum access to water.
Catheter Implantation. After two weeks on CT or FR diets, the animals were subjected to catheter implantation, as described previously 47 . In brief, rats were anaesthetized (2.5% isoflurane at 3 L/min O 2 ) and polyethylene catheters were inserted into the femoral artery for cardiovascular measurements and into the femoral vein for drug infusions. The catheters were tunneled subcutaneously and exteriorized at the back of the neck. After surgery, analgesics (4 mg/kg Ketoflex, 0.1 mL/300 g, s.c) and antibiotics (0.2 mL/100 g, penicillin, streptomycin, dihydrostreptomycin, s.c.) were administered post-surgery and 24 hrs after surgery if any sign of pain was observed. While the catheter inserted into the femoral artery could reduce blood flow to the lower leg, the rats quickly shift flow to collateral vessels to compensate for the loss of blood flow from the femoral artery and they recover well with minimum distress using this widely used method 9,10 . Experimental procedures began 48 h after recovery from the anesthesia. Separate animals were used for drug infusion and molecular studies, except the group that received Ang- [1][2][3][4][5][6][7][8][9][10] infusions since blood was collected from these animals to measure Ang peptides.

MAP & HR.
All experiments recording MAP and HR were made in conscious freely moving rats. The rat arterial catheter was connected to a pressure transducer (MLT0699; ADI Instruments, Bella Vista, Australia) and a signal amplifier (ETH-400; CB Sciences Inc., Milford, MA, USA). The analog signal from the amplifier was digitized by a 12-bit analog-to-digital converter (PowerLab/400; ADI Instruments), and the pulsatile arterial pressure was recorded at 1000 Hz using Chart 7.0 for Windows software (ADI Instruments). MAP and HR were derived on-line from the pulsatile arterial pressure measurements using pulse-to-pulse analysis 15 . Data for HR and MAP values were recorded continuously. Baseline MAP and HR were determined during the two min-period that preceded drug injections and are expressed as the mean ± SEM. Responses to treatments are calculated from continual 2 min averages after drug infusions and are expressed as the change in MAP or HR from baseline.
Plasma Sodium and Potassium Concentration and Balance. This experimental group was acclimated to the metabolic cage for 48 hours beginning on day 11. On day 13, 24 h collections began. Rats were given ad libitum access to water and the amount of powdered rat chow (Rodent diet 20, #5053, LabDiet, Marlborough, MA -sodium 0.3% and potassium 1.10%) was given according to the diet protocol (see above), either ad libitum for CT rats (n = 7) or 40% of the CT diet for FR rats (n = 7). Food intake and urine volume were measured gravimetrically. At the end of the experiment, the animals were anaesthetized with isoflurane (2.5% isoflurane at 3 L/min O 2 ) and the femoral artery and vein were catheterized for plasma volume measurements (see below). Blood was collected by abdominal aorta puncture and samples were centrifuged at 13,000 × g for 10 min to separate the plasma, which was stored in sterilized tubes at −80 °C until analysis. Sodium and potassium concentrations were determined by flame photometry (Model #2655-10; Cole-Parmer, Vernon Hills, IL). Plasma Volume. Plasma volume was measured by the Evan's blue dye technique as previously described 48 .
In brief, after baseline blood collection, Evan's blue dye (0.3 mg/ml) was infused into the venous catheter. Blood collections (0.2 mL), were taken from the arterial catheter at 5 and 10 min post-infusion. Plasma was collected following centrifugation of whole blood. Standards were made in 1% plasma at 0, 1, 5, 10, and 20 g/ml of Evan's blue dye. The concentration of Evan's blue dye was measured in a 100 μL plasma sample using a plate reader (FLUOstar Omega, BMG Labtech Inc., Cary, NC) at 610 nm at baseline before dye administration and at 5-and 10-min following dye administration. Background corrections were made by subtracting the observed concentration at baseline from the observed concentrations of the 5-and 10-min time points. Average plasma volumes were calculated by dividing the amount of administered Evan's blue dye (75 g) by the background corrected concentrations.
Plasma AGT. Parallel animal groups were used for all plasma and urine measurements except plasma volume and angiotensin peptide concentrations. Plasma concentrations of AGT were measured using an ELISA kit (Immuno-Biological Laboratories, Minneapolis, MS) 49 . In brief, 100 μl/well of standard rodent AGT (0.08-5.0 ng/mL) or plasma samples (1: 2,500 diluted in ELISA buffer) were added to the ELISA plate and incubated at 37 °C for 1 h. After the incubation, the plates were rinsed seven times with Washing Buffer provided by the kit. The plates were then incubated with 100 μl/well of horseradish peroxidase-labeled C-terminal antibody (1:30 diluted in antibody solution) at 37 °C for 30 min. After the incubation, the plates were rinsed nine times with Washing Buffer. Next, the plates were incubated with 100 μl/well of 3,3′,5,5′-tetramethylbenzidine solution under light-protected conditions at room temperature for 30 min. Finally, 100 μl of sulfuric acid (0.5 mol/l) was added to each well to stop the reaction. The absorbance values were measured at 450 nm. Plasma ALDO. Plasma ALDO was measured by ELISA (Enzo Life Sciences, NY). 100 μL of plasma diluted 1:5 was added to the plate immediately after the addition of antibody (50 μL) and the samples, the plate was incubated overnight at 4 °C. The next day, the plate was washed 3 times and 200 μL of substrate was added followed by a 1 h incubation at room temperature. The reaction was stopped by the addition of trisodium phosphate solution.
The absorbance values were measured at 405 nm.
PRA. Plasma renin activity was measured using the 5-FAM/QXL ® 520 fluorescence resonance energy transfer (FRET) peptide 52 . Reactions were conducted in 96 well microtiter plates containing 40 μL of plasma and the product formation was determined at 37 °C by following the fluorescence as a function of time using a fluorescence plate reader (FLUOstar Omega) at an excitation wavelength of 490 nm and an emission wavelength of 520 nm. Data were recorded at 2 min intervals and initial velocities were determined from the rate of fluorescence increased over the 10-100 min time course, which corresponded to the linear range of the assay. The data were expressed as the rate of change in fluorescence, using the product of 5-carboxyfluorescein (5-FAM) as a marker for PRA (n = 10-8/group).
qPCR. The mesenteric cascade, including the superior mesenteric artery was dissected from the intestinal wall using ice cold phosphate buffered saline (PBS) and this tissue was removed from the same rat used for plasma measurements. After the fat and veins were removed from the vessels using an Olympus dissecting microscope, the tissue was snap frozen in a dry ice-methanol bath and stored at −80 °C until use. The mesenteric vessels were homogenized using ceramic microbeads and total RNA was extracted using RNeasy according to the manufacturer's instructions (Bio-Rad). Purified RNA (1 μg) was reverse transcribed with a high-capacity cDNA reverse transcription kit. Quantitation of specific rRNA was performed by real-time PCR using the ABI Prism 7700 Sequence Detection System (Applied Biosystems, Foster City, CA). The PCR reaction mixture consisted of RNase-free water, SYBR green supermix and 300 nmol/L specific primers as previously described 53  Statistical Analysis. Prism 7.0 (GraphPad Software, La Jolla, CA, USA) was used to analyze all data and to construct the graphs. The data are expressed as mean ± standard error of the mean (SEM). The data for basal parameters characterization, enzyme activity, mRNA expression and peptide concentrations were analyzed first for normality using the Shapiro-Wilk normality test and when following the normality, were analyzed using the Student's unpaired t-test to assess differences between groups. All the results expressed as a curve after drug stimulation over time were compared by two-way (time and diet as factors) analysis of variance (ANOVA) followed by Bonferroni post-test using all the time-points showed in the graph. To test treatment effects within the same group, the Student's t-test was used when the data followed a normal distribution. The significance threshold level was set at 0.05.