Understanding sex differences in long-term blood pressure regulation: insights from experimental studies and computational modeling

Abstract

Sex differences in blood pressure and the prevalence of hypertension are found in humans and animal models. Moreover, there has been a recent explosion of data concerning sex differences in nitric oxide, the renin-angiotensin-aldosterone system, inflammation, and kidney function. These data have the potential to reveal the mechanisms underlying male-female differences in blood pressure control. To elucidate the interactions among the multitude of physiological processes involved, one may apply computational models. In this review, we describe published computational models that represent key players in blood pressure regulation, and highlight sex-specific models and their findings.

INTRODUCTION

Hypertension is a multifaceted disease affecting over 1.1 billion adults worldwide in 2015 and is one of the leading causes of deaths globally. In general, blood pressure in men exceeds that of age-matched women; therefore, men are at a greater risk for hypertension-related cardiovascular and renal diseases (58, 84). Despite these known sex differences, hypertensive men and women are typically given similar treatments. This one-size-fits-all approach yields a rather striking statistic: treated women achieve blood pressure control at a significantly lower rate (45%) compared with men (51%) (24).

Sex differences in blood pressure and the prevalence of hypertension are found in many mammalian and avian species (70). In humans (84) and in genetic models of hypertension such as spontaneously hypertensive rats (SHRs) and Dahl salt-sensitive rats (58), males develop earlier and more severe hypertension than females. To date, the mechanisms underlying male-female differences in blood pressure control remain incompletely understood.

In the past few years, an explosion of data has emerged concerning sex differences in nitric oxide (NO) bioavailability (11), the renin-angiotensin-aldosterone system (RAAS) (29, 38), inflammation (61, 78, 79), and kidney function (57, 68). To what extent do sex-based differences in renal transporter abundance, intrarenal RAAS components, and NO bioavailability underlie differences in blood pressure control between males and females? To elucidate the synergy among the multitude of physiological processes involved in blood pressure regulation, one may apply computational models. Below, we delineate published computational models that represent major players in blood pressure regulation, focusing on sex-specific models and simulation results.

SEX DIFFERENCES IN THE RAAS: DATA IN HUMANS AND RODENTS

The RAAS regulates blood pressure, in part, by adjusting glomerular filtration rate (GFR) and Na⁺ reabsorption in the nephrons of the kidney. ANG II, when bound to the ANG II type 1 receptor (AT₁R), stimulates Na⁺ reabsorption by increasing aldosterone secretion. AT₁R-bound ANG II also stimulates renal vasoconstriction, inflammation, and fibrosis (9). When bound to the angiotensin type 2 receptor (AT₂R), ANG II causes vasodilation and natriuresis (9). ANG (1–7), when bound to the MAS1 proto-oncogene (MAS) receptor, causes vasodilation and natriuresis and increases production of NO to reduce inflammation and fibrosis (9, 10). The renin-angiotensin reaction cascade is shown in Fig. 1.

Fig. 1.

Schematic diagram depicting the renin-angiotensin system reaction cascade. Blue and red arrows indicate higher abundance in males and females, respectively. Size of arrows roughly corresponds to abundance ratio between the two sexes, based on human and rat studies (14, 60, 67, 71, 73). Gray nodes indicate a lack of sex-specific data. AGT, angiotensinogen; ACE, angiotensin-converting enzyme; ANG I/II/IV/ (1–7), angiotensin I/II/IV/ (1–7); AT₁R/AT₂R, ANG II receptor type ½; NEP, neprilysin; PRA, plasma renin activity.

Sex differences in the circulating and intrarenal concentrations of the various forms of angiotensin and of the receptors are becoming increasingly well studied (18, 29, 85). Here, we focus on a few key differences. Sex hormones influence many parts of the RAAS. Generally, in males, the angiotensin-converting enzyme (ACE)/ANG II/AT₁R pathways are enhanced, whereas in females, the balance is shifted toward the ACE2/ANG (1–7)/Mas receptor and AT₂R pathways. In humans, estrogen causes females to have greater levels of circulating angiotensinogen (AGT) (18), while also decreasing circulating renin (71), ACE (20), AT₁R density, and aldosterone production (38). In rats, estrogen has been found to increase renal AT₂R expression and to decrease renal AT₁R synthesis (4, 55, 59, 71). On the other hand, in rats, testosterone increases renal AGT, AT₁R, and renin expression (13, 17, 33, 35). These effects of sex hormones likely lead to the greater AT₁R:AT₂R ratio found in male rats (8, 70, 85). Male rats also have greater ACE2 activity, while female rats have higher levels of ANG (1–7) (60).

The sex differences in the RAAS are not limited to peptide concentrations, as there are also differences in how the body reacts to changes within the RAAS. Changes in ANG II level produce significantly smaller changes in blood pressure in female rats than in male rats (18, 70), and ANG II infusion reduces renal perfusion significantly more in women than in men (21). The latter may be attributable to the fact that during ANG II infusion, GFR is maintained in men, while it drops in women (53).

These sex differences may lead to differences in drug treatment efficacy between the sexes. Angiotensin-converting enzyme inhibitors (ACEIs) reduce blood pressure more in male SHR than in female SHR (64). Women have been found to have better survival rates when taking angiotensin receptor blockers (ARBs) rather than ACEIs (31). That observation may be attributed to the possibility that the effectiveness of ACEIs in women decreases over time (74) or that ARBs reduce blood pressure more than ACEIs in women (65).

SEX DIFFERENCES IN THE RAAS: INSIGHTS FROM COMPUTATIONAL MODELS

A number of computational models have been developed to gain insights into the RAAS and to guide RAAS-targeted antihypertensive therapy. But despite the importance of sex differences throughout the RAAS, existing computational models of the RAAS are almost entirely sex neutral [with one exception (46)]. The “classical RAAS,” consisting of AGT, ANG I, ANG II, and AT₁R, is the most well-studied and modeled.

Takahashi et al. (76) modeled this “classical RAAS” along with bradykinin’s interactions with ACE and blood pressure. Their goal was to understand how ACE inhibitors can lower blood pressure, even when genetic variation in ACE and AGT expression does not substantially change blood pressure. They included blood pressure as an explicit variable, while other models below only used upstream ANG II levels as an indicator of drug treatment efficacy. With their model they were able to find that small changes in ACE and AGT, such as those caused by genetic differences, can be offset by bradykinin levels, so that ANG II levels stay the same. However, with ACE inhibition, the system deviates sufficiently far from baseline, such that bradykinin can no longer offset the drastic reduction in ANG I. As a result, ANG II levels, and, thus, blood pressure, drop as well.

Hong et al. (30) developed a pharmacodynamic model for renin, ANG I, and ANG II, together with a pharmacokinetic component describing the effects of aliskiren, a direct renin inhibitor (DRI). They found that aliskiren was best described with a two-compartment model with nonlinear capacity-limited distribution and elimination kinetics. Their model was able to show the greater increase in active renin from aliskiren treatment as compared with ACEI or ARB. They also discovered that there is a limit to the amount of RAAS blockade available from increasing doses of aliskiren. They presented a reduced model that does not rely on ANG II measurements, which can be difficult to attain.

Ramusovic and Laeer (63) also modeled this “classical RAAS” to compare and contrast the effects of a single dose of ACEI, ARB, and DRI. Even without the inclusion of circadian rhythm parameters as used in Hong et al. (30), Ramusovic and Laeer were able to achieve a sufficiently good fit for the plasma ANG I and ANG II levels following aliskiren treatment. Notably, model responses were matched to a range of dosages rather than just a single dosage.

Lo et al. (48) developed a more comprehensive model of the RAAS containing AGT, ANG I, ANG II, ANG (1–7), ANG IV, AT₁R, and AT₂R. They validated the model using angiotensin infusion experiments and compared the effects of ACEI, ARB, and DRI. They extended their model to include not only the circulating plasma hormone concentrations, but intrarenal RAAS as well. Using this more complete and detailed model of the RAAS, they were able to measure the effects of drug treatments by monitoring AT₁R-bound ANG II levels, which is the actual effector of changes in the body, rather than plasma renin activity (PRA), which is the quantity commonly measured in clinical experiments. Although these models provide insight for understanding the RAAS and guiding RAAS-targeted antihypertensive therapy, none of them capture sex differences in the RAAS or in men and women’s differing response to RAAS inhibitors.

Recently, a sex-specific computational model of the RAAS was finally developed. Leete et al. (46) presented the first RAAS model for humans that takes into account known sex differences in the RAAS. In their model, the general effects of sex hormones were represented, but menopause and the estrous cycles were not variables. The model included the more extensive coverage of circulating angiotensin peptides as in Lo et al. (48) but was fitted separately for four groups: male normotensive rats, female normotensive rats, male hypertensive rats, and female hypertensive rats. They included sex differences in ANG I, ANG II, ANG (1–7), AT₁R-bound ANG II, and AT₂R-bound ANG II concentrations, and in ACE2 and neprilysin (NEP) activity. They identified that sex differences in peptide levels are attributable to differences in key reaction rate constants in the RAAS enzymatic cascade. By conducting model simulations, they proposed two explanations as to why ARBs are more effective than ACEI in women, while the opposite is true in men. First, AT₂R-bound ANG II may have a stronger beneficial effect in women; thus, ACEI-induced reduction in AT₂R-bound ANG II may be more detrimental in women. Second, renin production is regulated by AT₁R-bound ANG II via a negative feedback mechanism. Leete et al. (46) hypothesized that the strength of the AT₁R-bound ANG II-mediated feedback response of renin may be stronger in women. Thus, any ACEI-induced reduction in AT₁R-bound ANG II would be attenuated by the compensatory increase in renin production to a larger extent in women compared with men. Consequently, women’s AT₁R-bound ANG II would stay nearer to the hypertensive level, thereby reducing the benefits of ACEI. Through the same negative feedback mechanism, ARB would also elevate renin production more in women, resulting not only in higher ANG I and ANG II levels, but AT₂R-bound ANG II as well. Because AT₂R-bound ANG II may have more beneficial effects in women, its elevated level may explain the observed stronger blood pressure reduction in women. The sex-specific responses to RAAS inhibitors were further explored in a more comprehensive blood pressure model (45); see below.

SEX DIFFERENCES IN RENAL HEMODYNAMICS, EPITHELIAL TRANSPORT, AND KIDNEY FUNCTION: DATA IN HUMANS AND RODENTS

The kidney is a key (albeit not sole) determinant of blood pressure and of sex differences in hypertension (82). That role is evident in transplantation studies, where blood pressure “goes with the kidney”: transplanting a kidney from a hypertensive rat into a normotensive rat induces hypertension in the recipient (6). Essential for the kidney’s long-term control of blood pressure is the pressure-natriuresis mechanism, whereby increases in renal perfusion pressure lead to increases in Na⁺ excretion, which, in turn, lowers salt and water retention, reduces effective circulating volume, and lowers blood pressure. Females tend to exhibit a leftward shift in the pressure-natriuresis relation relative to males, i.e., females excrete the same amount of Na⁺ as males, but at lower arterial pressure under the same physiological conditions (28, 37). Pressure-natriuresis responses encompass multiple levels of Na⁺ transporter regulation (51, 52) and are substantially modulated by the RAAS (54).

In a post-mortem study involving 86 paired kidneys, Pourteyron (62) reported that men’s kidneys weigh 23% more than women’s. Compared with men, women, on average, have lower GFR, but that appears to be proportional to women’s smaller body size. Among age-matched, healthy adults, there were no significant differences in GFR between men and women after normalizing to 1.73 m² body surface area (50).

More is known about the sex differences in kidney anatomy and function in rodents. The kidney mass of a female rat is approximately half of that of an age-matched male rat (49, 68). Glomerulus population is similar in both sexes (57). Single-nephron GFR (SNGFR) in female rat kidneys is lower than that in the male (57, 66), whereas urinary outputs are not substantially different between the sexes after adjusting for body weight (68). Furthermore, NO bioavailability is known to be higher in female than male SHRs (75), possibly due to the significantly higher abundance of renal NOS1 and NOS2 in females compared with males (11). NO contributes to blood pressure control by regulating vascular tone and hemodynamics. Additionally, NO also inhibits tubular Na⁺ reabsorption, an action that promotes natriuresis and diuresis.

In a recent study, Veiras et al. (81) reported sexually dimorphic patterns in transporter abundance in rodents. Specifically, they compared the expression, covalent modifications, and regulators of electrolyte transporters, channels, and claudins in male and female Sprague-Dawley rats (81). Their findings demonstrated that female (in comparison to male) rat nephrons exhibit: 1) lower abundance of proximal tubule Na⁺-phosphate cotransporter 2 (NaPi2), aquaporin (AQP)-1, and claudin-2; 2) greater Na⁺-H⁺ exchanger 3 (NHE3) phosphorylation, which is a marker of distribution at the base of the microvilli in an inactive domain (7), and 3) lower Na⁺ and $\text{H}\text{C}{\text{O}}_3^{-}$ reabsorption and increased volume flow from the proximal tubule. Additionally, Sabolic et al. (69) reported higher mSGLT2 (SGLT2, sodium-glucose cotransporter 2) protein abundance level in female Wistar rats compared with males. No correlation was found between transporter abundance and the estrous cycle stage (81). A caveat is that the above findings indicate sex differences in transporter expression levels, which do not directly translate into transporter activities. As discussed below, computational models can be used to assess the extent to which individual sex differences, in morphology or transporter activities, contribute to the observed difference in solute and water transport along the nephron in male versus female rats.

SEX DIFFERENCES IN RENAL HEMODYNAMICS, EPITHELIAL TRANSPORT, AND KIDNEY FUNCTION: INSIGHTS FROM COMPUTATIONAL MODELS

Over the past couple of decades, classes of computational models have been developed to investigate different aspects of the kidney: some models focus on whole animal or whole kidney function and metabolism (e.g., Refs. 15a, 19, 27, 39, and 56), some examine renal autoregulation and hemodynamics (e.g., Refs. 42 and 72), whereas others consider nephron-level function and simulate membrane transport kinetics (e.g., Refs. 16, 40, and 83). Almost all models are based on the rat [one exception being a human nephron model (41)], or more specifically, the male rat. Very recently, computational models were finally developed for the kidney of the female rat; see below.

What are the functional implications of the sexually dimorphic renal transporter abundance revealed in rodent kidneys, as reported in Refs. 69 and 81? Epithelial transport models of the nephron can be used to provide insights. Given a set of model parameters, renal epithelial transport models can predict tubular fluid and solute flow, water, and solute fluxes through individual transporters or channels, as well as urine flow and solute excretion rates.

As previously noted, existing epithelial transport models have, until recently, been built exclusively for males (e.g., Refs. 43, 44, and 83). To analyze epithelial transport in both sexes of the rat, Li et al. (47) developed the first sex-specific computational epithelial transport models for the proximal convoluted tubule of the rat. A schematic diagram of the proximal tubule cell is shown in Fig. 2. The models accounted for the sex differences in expression levels of the apical and basolateral transporters (81), in SNGFR, and in tubular dimensions. Model simulations predicted that the substantially lower fractional volume reabsorption in female (about half of male) can be attributed to their smaller transport area and lower aquaporin-1 expression level. The latter also results in a larger contribution of the paracellular pathway to water transport. Model simulations also predicted that the similarly lower fractional Na⁺ reabsorption in female is due primarily to their smaller transport area and lower Na⁺/H⁺-exchanger (NHE3) and claudin-2 expression levels (Fig. 3). Notably, unlike most Na⁺ transporters, whose expression levels are lower in female, SGLT2 expression levels are 2.5-fold higher in female (69). Model simulations suggested that the higher SGLT2 expression in female may compensate for its lower tubular transport area to achieve a similar hyperglycemic tolerance as male (47).

Fig. 2.

Schematic diagram of the proximal convoluted cell model (47) of a normotensive rat, showing major Na⁺, K⁺, and Cl⁻ transport pathways. Flux values that were computed at the midpoint of the proximal convoluted tubule are shown in blue and red for male and female rats, respectively. Because of the lower activities of the Na⁺-H⁺ exchanger 3 (NHE3) and Na⁺-phosphate cotransporter 2 (NaPi2) in female, the corresponding predicted fluxes are substantially lower in female. [Reprinted from (47).]

Fig. 3.

Simulation results obtained by the sex-specific rat proximal convoluted tubule computational model in Ref (47). A–G: predicted solute flow profiles along the proximal tubule of the male (solid lines) and female (dashed lines) rat kidney. H: tubular fluid pH. I: water flow profile. Substantially less Na⁺, K⁺, Cl⁻, and water are reabsorbed along the female rat proximal tubule compared with the male, in large part, due to the smaller transport area and lower transporter activity in female. TA, titratable acid. [Reprinted from (47).]

At the organ level, how do the sex differences in renal anatomy and hemodynamics affect kidney function? Chen et al. (11) developed the first sex-specific model for the whole medulla. Specifically, they applied computational models of oxygenation and transport of solutes in the kidneys of male and female SHRs to investigate the sex differences in NO levels. The models represent renal medullary structures, including the loops of Henle, collecting ducts, vasa recta, and capillaries (Fig. 4). To simulate renal oxygenation, the model vasa recta and capillaries are divided into plasma and red blood cell compartments. The models represent the transport of water and major solutes NaCl and urea (the active transport of NaCl consumes O₂), as well as O₂, NO, and superoxide. Solute and water transport is represented using the single-barrier approach (i.e., detailed epithelial transport is not represented). The model predicts tubular and vascular flows, luminal and interstitial solute concentrations, transmural water and solute fluxes, and basal and transport-driven metabolism at all levels of the renal medulla. The male and female SHR models differ in blood flow, SNGFR, and size, while tubular transport and metabolism were assumed the same [this study was published before the sexual dimorphism in renal transporter pattern was clearly characterized (81)].

Fig. 4.

Schematic diagram of sex-specific computational model of the rat renal medulla in Ref (11). The diagram shows a short loop of Henle, a representative long loop, turning at x₂ within the inner medulla, and two representative descending vasa recta (DVR), terminating at x₁ and x₂. Although only one long loop and two DVR are shown, the model represents one long loop and one DVR at every spatial point in the inner medulla and renal medulla, respectively. A representative collecting duct is shown, with “branches” representing the coalescence of the inner-medullary collecting ducts. Collecting duct outflow becomes urine. [Reprinted from (11).]

Model simulations indicated that medullary O₂ and NO bioavailability would be similar in male and female rats if they had the same transport and metabolic parameters (Fig. 5). That result suggested that the greater NO levels observed in female SHR compared with males (75) could not be attributed entirely to the sex differences in kidney size and blood flow. Instead, the greater NO bioavailability in female SHR might be the consequence of their greater NO production or lower levels of NO scavenging by superoxide. Simulation results also suggested that medullary NO levels and tissue O₂ tension (Po₂) were highly sensitive to endothelial NO generation rates (Fig. 5). Significantly higher medullary NO and Po₂ levels could be obtained even when the higher NO generation rates are limited to specific vascular segments. In contrast, NO and Po₂ levels were not sensitive to the potentially lower superoxide production rate in female (and, thus, lower medullary superoxide levels and lower levels of NO scavenging by superoxide).

Fig. 5.

Simulation results obtained for varying NO generation rate (denoted G_NO) (11). Predicted effects on interstitial [NO] (A), [O₂⁻] (B), and oxygen tension Po₂ (C). Results were obtained for male spontaneously hypertensive rats (SHR) using baseline G_NO, and for female SHR using G_NO at baseline and at twice and three times its value. These results suggest that medullary NO levels and tissue Po₂ are highly sensitive to endothelial NO generation rates. [Reprinted from Ref (11).]

SEX DIFFERENCES IN BLOOD PRESSURE REGULATION IN HUMANS AND ANIMALS

Sex differences in blood pressure are similarly well documented in various animal models. Male SHRs have been reported to exhibit higher blood pressure than age-matched females (12, 22), as do Dahl salt-sensitive rats (15) and New Zealand genetically hypertensive rats (2). Compared with age-matched premenopausal women, men have higher blood pressure, are at higher risk for cardiovascular and renal diseases (36, 84), and are associated with higher incidence of controlled hypertension (1). Following menopause, however, blood pressure rises in women. Sex differences in blood pressure control and hypertension have been discussed in a number of review articles (e.g., Refs. 23, 64, and 85).

SEX DIFFERENCES IN LONG-TERM BLOOD PRESSURE REGULATION: INSIGHTS FROM COMPUTATIONAL MODELING

Given the multiple systems involved, studying whole body blood pressure regulation requires a systems biology approach. Thus, computational modeling is a useful tool. Computational modeling of the circulatory system for blood pressure regulation was pioneered by Guyton et al. (26) in their seminal work in 1972, hereafter referred to as the “Guyton model.” Each model component had previously been the subject of research investigation, and this work provided a mathematical framework to show how the different regulators operate together in the overall circulatory system. Model components include cardiovascular function, circulatory dynamics, renal hemodynamics, kidney function, respiratory function, neurohormonal feedback, autonomic nervous system activity, and electrolyte balance. The key regulator of blood pressure is the kidney, whereby the kidney increases Na⁺ and water excretion (decreases reabsorption) in response to elevated blood pressure. As a result, extracellular fluid volume decreases, which, in turn, lowers blood pressure (26).

Since its publication in 1972, the Guyton model has undergone a series of extensions and revisions, and several alternatives have been published. We will summarize a few recent ones. Karaaslan et al. (34) integrated the Guyton model (26) with Uttamsingh et al.’s (80) model for detailed renal function; they also added an explicit representation of the renal sympathetic nerve activity (RSNA). The goal of this expansion was to account for more detailed mechanisms of the kidney as it is the primary regulator of blood pressure, and to account for RSNA since RSNA-related mechanisms are involved in hypertension, congestive heart failure, and cirrhosis. Thomas et al. (77) expanded the Guyton model in a different direction. They incorporated into the model Ikeda et al.’s detailed representation of body fluid regulation (32) and also added a detailed model of the kidney. That expansion allowed for more detailed mechanisms of body fluid homeostasis, as this is intertwined with blood pressure regulation. None of the models discussed so far considered the RAAS in detail. That changed when Guillaud and Hannaert (25) augmented the model of Thomas et al. (77) by adding a detailed model of the RAAS. Similarly, Hallow et al. (27) added a detailed model of the RAAS to the model by Karaaslan et al. (34). Here, the representation of the RAAS is even more detailed than that of Guillaud and Hannaert (25), and it was included to simulate antihypertensive drugs that target specific pathways in the RAAS.

Averina et al. (3) challenged the role of the kidney in blood pressure regulation. While the Guyton model and its descendants assume that sodium excretion is determined by arterial pressure, Averina et al. assumed that sodium excretion does not depend on arterial pressure. They stated that both the hypertensive hemodynamic profile and the pressure-natriuresis relationship can be reproduced with natriuresis driven by sodium intake and sympathetic nerve activity being the primary regulators of blood pressure. They developed a very simple model of the cardiovascular system and kidney function to demonstrate this. This model is for qualitative and theoretical purposes to test an alternative hypothesis to the classic Guyton model.

Beard et al. (5) were concerned with developing a tractable model, where all variables, parameters, equations, and experimental data used to identify the parameters are explicitly stated. With this goal in mind, they developed a simplified model of blood pressure regulation that included mechanics of the heart, kinetics of baroreflex firing, circulatory dynamics, and neurohormonal feedback. Although they did not deny the kidney’s role in blood pressure regulation as Averina et al. (3), they did diminish its role, with greater significance of blood pressure regulation given to the baroreflex arc and the RAAS (5). This simple model is useful for gaining a bare bones understanding of mechanisms involved in blood pressure regulation, but it lacks sufficient details to simulate certain clinically interesting variables or scenarios.

The Guyton lineage models for blood pressure regulation are all sex neutral. To investigate the sexual dimorphism in blood pressure regulation and the implications in antihypertensive therapy, Leete and Layton (45) developed the first sex-specific computational models of blood pressure regulation. Their models account for sexual dimorphism based upon published observations, which included 1) the vasodilation of the renal afferent arterial by AT₂R-bound angiotensin II in the RAS (which is assumed to occur only in females), 2) sex differences in the RAAS, and 3) the less excitable and more easily repressed female RSNA. A schematic diagram is shown in Fig. 6, where components outlined in red are sex-specific.

Fig. 6.

Schematic model of blood pressure regulation. Pink nodes denote variables that describe cardiovascular function; green nodes denote renal hemodynamics; orange nodes denote renal Na⁺ handling and urine production; blue nodes denote the renin-angiotensin system (RAS). Red outlines denote sex-specific model components. ADH, anti-diuretic hormone; AGT, angiotensinogen; ALD, aldosterone; ANG I, angiotensin I; ANG II, angiotensin II; AT₁R-bound ANG II, angiotensin II type 1 receptor bound angiotensin II; AT₂R-bound ANG II, angiotensin II type 2 receptor bound; MAP, mean arterial pressure; PRA, plasma renin activity; PRC, plasma renin concentration; RSNA, renal sympathetic nerve activity. [Modified from Ref (45) with permission from Elsevier: Comp Biol Med 104: 139–148, 2019.]

Leete and Layton applied the models to assess men and women’s responses to various hypertensive stimuli. To that end, they induced hypertension, in both the male and female models, by individually increasing the following model parameters: systemic vascular resistance, afferent arteriole resistance, proximal tubule Na⁺ reabsorption, distal tubule Na⁺ reabsorption, collecting duct Na⁺ reabsorption, renin secretion, aldosterone secretion, and RSNA. With each parameter change, the predicted blood pressures were recorded for the male and female models. Key results are summarized in Fig. 7. These results suggested that, for a given pathophysiological perturbation, the severity of hypertension, thus, induced may vary significantly between the sexes. Moreover, both the male and female models predict that stiffening of the afferent arterioles induces the largest increase in extracellular fluid volume (results not shown) and, thus, blood pressure. Model simulations further indicate that the stronger RSNA-mediated regulation of afferent arteriole tone in women is primarily responsible for their resistance to developing hypertension (in response to afferent arteriole stiffening).

Fig. 7.

Simulation results using the sex-specific blood pressure regulation model in Ref (45). Results show predicted changes in mean arterial pressure (MAP) in response to various hypertensive stimuli, obtained separately for men and women. The chosen increase in afferent arteriole resistance yields the largest rise in MAP. [Reprinted from Ref (45) with permission from Elsevier: Comp Biol Med 104: 139–148, 2019.]

The authors further applied the models to investigate the higher effectiveness of ARB compared with ACEI in blood pressure reduction in hypertensive women (a difference not found in men). Model results attributed that difference to a potential vasodilatory effect of AT₂R-bound ANG II, which was assumed to be present only in females. More specifically, this finding stemmed from the blood pressure-lowering effect of AT₂R-bound ANG II: it decreases afferent and efferent arteriole resistance (opposite effect of AT₁R-bound ANG II) (85). The AT₁R:AT₂R ratio has generally been reported to be higher in males than females (73, 85); in fact, AT₂R have been reported to be absent, or nearly so, in the adult male rat kidney (74). These observations have led to the hypothesis that AT₂R-bound ANG II may play a significant role in blood pressure regulation and in hypertensive patients’ response to ACEI and ARB—a hypothesis that is supported by model results (45). In summary, by accounting for the sexual dimorphism in blood pressure regulation, sex-specific blood pressure regulation models [such as (45)], can be used to gain insights into the sex-specific pathophysiology of hypertension. Nonetheless, limitations of the model (45) should be noted. A number of players in blood pressure are not represented, including immune cell infiltration and inflammation, endothelin, epoxyeicosatrienoic acid, 20-hydroxyeicosatetraenoic acid, obesity-mediated sympathetic activation, and age. In the female case, menopause and the estrous cycle were not considered variables. With the incorporation of these mechanisms, a powerful tool can be developed to tailor antihypertensive therapy to individuals.

CONCLUSIONS

Recent mandates from the National Institutes of Health (NIH) requiring the consideration of sex as a biological variable in NIH-funded research has resulted in a significant increase of females in experimental studies. Indeed, the topic of sex and gender is of growing interest to many in the scientific community, including heart and circulatory physiologists. However, while there have been key advances in our understanding of the mechanisms underlying sex differences in blood pressure since they were first reported decades ago, our understanding remains incomplete. Because of the central role of the kidney in the development of hypertension, gaining a better understanding of fundamental differences between the sexes in renal function in health and disease is important for gaining insight into the basis of blood pressure control in both males and females. Our inadequate knowledge regarding the molecular mechanisms responsible for observed sex differences in blood pressure control and renal function is likely due, in large part, to the scarcity of available data and analysis in females.

It is no longer sufficient for investigators to simply assess the existence of “sex differences”. Although it is important for a phenotypic sex difference to be reported, to advance toward improved treatment strategies for disease, studies, whether experimental or computational, should also be designed to understand the mechanistic basis underlying the observations. More modeling studies that include in-depth analysis of sex-specific cardiovascular function are needed. Indeed, a recent search for “sex-specific computational model blood pressure” on PubMed yields only two publications (11, 45), both of which were published by the authors’ research group.

As a result of the current lack of sex-specific knowledge, the standard of care for most cardiovascular and renal diseases, including hypertension, has been designed on the basis of clinical studies conducted primarily in men, with the assumption that the treatment effects in women will be similar to those observed in men. This assumption, however, is becoming increasingly challenged as the scientific community learns more about how the sex of the individual impacts basic physiology, the pathophysiology of numerous diseases, and pharmacology.

Studies on sex differences generally fall into two categories: they either describe variations in cardiovascular or renal function (e.g., in urinary excretion or blood pressure following a given treatment), or report differences in the expression/activity of certain markers. Computational models have the potential to bridge the existing gap by establishing the link between mechanisms at the molecular/cellular level and observations at the organ level.

GRANTS

This research was supported by the Canada 150 Research Chair program and National Institute of Diabetes and Digestive and Kidney Diseases Grant R01-DK-106102.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

A.T.L. conceived and designed research; S.A., R.H., J.L., and A.T.L. analyzed data; S.A. and J.L. prepared figures; S.A., R.H., J.L., and A.T.L. drafted manuscript; S.A., R.H., J.L., and A.T.L. edited and revised manuscript; S.A., R.H., J.L., and A.T.L. approved final version of manuscript.