Mechanisms of Vascular Remodeling in Hypertension

Abstract

Hypertension is both a cause and a consequence of central artery stiffening, which in turn is an initiator and indicator of myriad disease conditions and thus all-cause mortality. Such stiffening results from a remodeling of the arterial wall that is driven by mechanical stimuli and mediated by inflammatory signals, which together lead to differential gene expression and concomitant changes in extracellular matrix composition and organization. This review focuses on biomechanical mechanisms by which central arteries remodel in hypertension within the context of homeostasis—what promotes it, what prevents it. It is suggested that the vasoactive capacity of the wall and inflammatory burden strongly influence the ability of homeostatic mechanisms to adapt the arterial wall to high blood pressure or not. Maladaptation, often reflected by inflammation-driven adventitial fibrosis, not just excessive intimal–medial thickening, significantly diminishes central artery function and disturbs hemodynamics, ultimately compromising end organ perfusion and thus driving the associated morbidity and mortality. It is thus suggested that there is a need for increased attention to controlling both smooth muscle phenotype and inflammation in hypertensive remodeling of central arteries, with future studies of the often adaptive response of medium-sized muscular arteries promising to provide additional guidance.

Graphical Abstract

Graphical Abstract.

Hypertension is a significant risk factor for myriad cardiovascular, renovascular, and neurovascular diseases, and thus significant morbidity and mortality. Its incidence continues to rise worldwide. In the United States alone, the Centers for Disease Control (CDC) reports that nearly one-half of all adults are either hypertensive (systolic pressure ≥130 mm Hg or diastolic pressure ≥80 mm Hg) or being treated for such, with prevalence continuing to rise due to increasing obesity, diabetes, other autoimmune diseases, and the overall aging of our society.^1–4 Particularly alarming, however, is the increasing number of young people who are hypertensive, with early onset hypertension hastening the development of subsequent morbidities and increasing early mortality.^5–7 Independent of the segment of the population that is affected, hypertension typically associates with detrimental changes in central artery structure and function, including structural stiffening of the wall and endothelial dysfunction, both of which adversely affect the hemodynamics and thus perfusion of end organs,^8,9 ultimately driving the increased morbidity and mortality. Although there has been controversy as to whether stiffening of central arteries is a cause or consequence of hypertension, an insidious positive feedback loop appears to exist between stiffening and hypertension,¹⁰ perhaps rendering it irrelevant how the loop initiates. Regardless, early intervention appears critical given the difficulty of reversing many of the changes that characterize hypertension-induced arterial remodeling.

Sustained elevations of blood pressure elicit a host of mechanobiological responses by arteries that often result from phenotypic changes by the primary cells of the wall, reflected by a differential gene expression that drives remodeling. In addition, however, inflammation plays critical roles in mediating this remodeling, particularly in large arteries. Notwithstanding tremendous information gained from clinical studies, animal models continue to provide important insight into the mechanobiological and immunobiological mechanisms that dictate and/or are dictated by hypertension. This brief review focuses on lessons learned from mouse models of hypertension and, in particular, the associated remodeling of central arteries within the context of mechanical homeostasis—what promotes it and what prevents it. Differential remodeling across different sized vessels, from conductance to resistance vessels, emphasizes further the need to understand phenotypic differences among smooth muscle cells.

HEMODYNAMICS AND WALL MECHANICS

All primary cells of the arterial wall—endothelial, smooth muscle, fibroblasts, and resident macrophages—are exquisitely sensitive to changes in their mechanical environment, typically changing gene expression accordingly.¹¹ Simple examples include the up- or downregulation of nitric oxide synthase (eNOS) by endothelial cells in response to increases or decreases in flow, which affects vasoregulation of caliber, and an increased local production of angiotensin II (AngII) and transforming growth factor-beta by vascular smooth muscle cells in response to increases in cyclic stretch/stress, which increases rates of synthesis of extracellular matrix and increases wall thickness.^12,13 Indeed, such processes are often linked, with the vasoactive state of the arterial wall at the time of matrix turnover often dictating net biomechanical changes by defining the geometry in which the matrix turns over.^14,15 There is strong motivation, therefore, to quantify the biomechanical state of the arterial wall and how it changes with altered hemodynamic loading.

Notwithstanding the complexity of the hemodynamics and wall mechanics, 2 simple relations provide considerable insight. Continuum interpretations of Newton’s second law of motion (i.e., linear momentum balance) provide relationships for 2 key components of the mechanical stress (a measure of the “force intensity” at a point): the primary flow-induced wall shear stress τ_w that stimulates endothelial cells and the circumferential (intramural) pressure-induced stress σ_θ that stimulates the smooth muscle cells of the media and fibroblasts of the adventitia (Figure 1a). Mean values of these 2 components of the stress can be written¹⁶

Figure 1.

(a) Schematic drawing showing 4 components of the mechanical stress and the areas on which they act: flow-induced wall shear stress τ_w ~ 5 Pa and pressure-induced radial σ_r ~ −6 kPa, circumferential σ_θ ~ 150 kPa, and axial σ_z ~ 150 kPa wall stress, with these values representative of the mouse descending thoracic aorta shown in a histological cross-section (note: 1 Pascal, Pa, is 1 Newton per meter squared). Although these values represent well the mean values, different values of the intramural stress exist in the media (higher) and adventitia (lower), suggesting different levels of mechano-sensitivity for all 3 primary cell types: endothelial, smooth muscle, and fibroblasts. (b) Representative biomechanical data emphasizing that luminal radius a (top row) and wall thickness h (bottom row) change acutely in response to both increased smooth muscle cell contractility (left) and distending pressure (right), both of which thereby govern mean circumferential wall stress, as, for example, via the Laplace equation, ${\sigma }_{\theta }=Pa(P,C)/h(P,C)$, where P is pressure and C is contractility. Data courtesy of Dr Sae-Il Murtada (author’s lab). Abbreviation: PE, phenylephrine.

$${\displaystyle \begin{array}{l}{\displaystyle {\tau }_w=\frac{4\mu Q}{\pi a^3},}\end{array}}$$ (1)

$${\displaystyle \begin{array}{l}{\displaystyle {\sigma }_{\theta }=\frac{Pa}{h},}\end{array}}$$ (2)

where Q is the volumetric flow rate (i.e., mean fluid velocity times the cross-sectional area through which the fluid flows), µ the viscosity of the blood (a measure of its resistance to flow), a the luminal radius (not an arbitrary radial location r in the wall, with $a\le r\le a+h$), P the distending pressure, and h the total wall thickness (not the intimal–medial thickness, which became a popular metric because imaging could not discern adventitial boundaries from perivascular tissue). It is helpful to appreciate that which equations (1) and (2) represent. It is intuitive, for example, that wall shear stress, which results from direct frictional interactions between the blood and endothelium, increases with increases in viscosity (e.g., hematocrit) and acute increases in flow, but decreases with luminal enlargement. Similarly, circumferential wall stress increases with acute increases in pressure and decreases with wall thickening, with luminal radius again playing an important role. Because of mechanisms of cell-mediated remodeling of the wall, however, the situation is very different for sustained than acute changes in flow or pressure, though often understandable in terms of the same simple relations for stress.

Let ε denote a sustained fold-change in flow (i.e., $\epsilon =Q/Q_o$ where subscript o denotes the original, or homeostatic, value) and γ a sustained fold-change in pressure ($\gamma =P/P_o$). It can be shown that both of the mean stresses in equations (1) and (2) can be maintained at or restored to original values if luminal radius a and wall thickness h adapt via

$${\displaystyle \begin{array}{l}{\displaystyle a\to {\epsilon }^{1/3}{a}_o,}\end{array}}$$ (3)

$${\displaystyle \begin{array}{l}{\displaystyle h\to \gamma {\epsilon }^{1/3}{h}_o.}\end{array}}$$ (4)

Note that the original values of the luminal radius a_o and wall thickness h_o are preserved for the special case of no change in flow or pressure from homeostatic (ε = 1, γ = 1), as they should. Many reports in the literature reveal that these 2 relations hold for modest alterations in hemodynamics.^14,17 In such cases, the vessels are said to have mechano-adapted; otherwise, they have maladapted mechanically. Whereas equations (3) and (4) define what should happen in a mechano-adaptation, they do not describe how this happens.

MECHANICAL HOMEOSTASIS

Structural and functional responses of arteries to sustained alterations in hemodynamic loading often represent particular manifestations of a mechanical homeostasis that exists at subcellular, cellular, and tissue levels in the vasculature.¹¹ For example, in vitro studies of smooth muscle cells and fibroblasts reveal that mechanical stresses at focal adhesions (i.e., clusters of integrins) tend to be maintained at or restored to original values when perturbed¹⁸ and similarly overall smooth muscle cell stiffness tends to increase with increased in-plane loading but then return toward original values.¹⁹ Note that material stiffness relates linearly with stress for the highly nonlinear (exponential) stress–strain behaviors exhibited by most vascular cells and tissues, as observed by Y.C. Fung in the 1960s and widely accepted since.¹⁶ Hence, an experimentally observed restoration of either wall stress or material stiffness likely reflects a common underlying mechano-regulatory process by the cells. Indeed, vessel-level studies suggest that the circumferential material stiffness tends to be highly mechano-regulated in arteries, both across species in health²⁰ and within species under different conditions.²¹ Together material stiffness and geometry (especially wall thickness) determine the structural stiffness of the wall, which dictates the hemodynamics. It is typically the structural stiffness that is observed clinically (via measures of distensibility or pulse wave velocity) to increase in central arteries in hypertension, often due to wall thickening. Importantly, these measures of structural stiffness are pressure dependent, hence their interpretation in hypertension must account for the blood pressure at which the metric was measured.²²

The general concept of homeostasis—that key physiological variables tend to be regulated within predefined ranges via negative feedback—was introduced by Cannon in the 1920s and augmented by introduction of the concept of set-points by Hardy in the 1950s. Common examples include regulation of core body temperature (macroscale) and interstitial fluid pH (microscale). Because most constituents of the extracellular matrix turnover slowly but continuously (e.g., the normal half-life of vascular collagen is on the order of 70 days), homeostasis plays important roles even in the long-term maintenance of arterial structure and function under normal conditions,²³ not just in response to sustained alterations in hemodynamic loads or during disease progression or wound healing. Figure 2 emphasizes that mechanical homeostasis requires mechanical stimuli, states, sensors, set-points, and a system that is responsive to the perceived inputs. Ubiquitous mechanical stimuli are, of course, blood flow and pressure.

Figure 2.

Schema of a negative feedback system characteristic of mechanical homeostasis in arteries, defined by the mechanical stimuli that give rise to the mechanical state (e.g., values of mechanical stress) that are sensed by the cells (e.g., via integrins), with the perceived state compared with homeostatic set-points (target values) to drive possible cell and tissue turnover at stress-modulated rates, determined in part by system gains (sensitivities). Inflammation can promote or prevent homeostasis, depending on the inflammatory burden, the former usually via acute and the latter via chronic inflammation. Note: min{deviations} is the homeostatic process by which the deviations from set points are resolved.

One way to quantify the mechanical state is to calculate mechanical stresses, which ultimately derive from 3 quantities: the mechanical loads, the geometry (e.g., a cylindrical tube defined by luminal radius a and wall thickness h), and the material properties. Whereas viscosity (a property of blood) appears explicitly in the equation for flow-induced shear stress at high shear rates (equation (1)), the effect of material properties appears implicitly in the equation for pressure-induced circumferential wall stress (equation (2)). That is, luminal radius a depends on both the acute pressure-dependent passive stiffening of the wall and the active contractile state, C, which can be conceptualized via the relations $a=a(P,C)$ and similarly for wall thickness $h=h(P,C)$, as illustrated in Figure 1b though usually not shown explicitly in the stress equation. Recall from above that smooth muscle contractility, and its partial paracrine control by mechano-sensitive endothelial cells, is critical in dictating the mechanical state in which cell and matrix turnover occurs. A simple case is flow-induced remodeling wherein inward vs. outward flow depends directly on vasoregulation of the lumen about which the matrix is remodeled.²⁴ It is emphasized further that smooth muscle cell actomyosin activity is not only important for vasoregulation at the vessel level, it also enables the cellular mechano-sensing and mechano-regulation of matrix that is critical in remodeling.²³

Sensors are needed to interpret the often changing mechanical state (Figure 2). Although we do not yet know precisely how vascular cells sense changes in their mechanical environment, integrins play central roles in the sensing of intramural stresses by smooth muscle cells and fibroblasts^25,26 while complexes such as VE-cadherin—PECAM-1 play central roles in the sensing of wall shear stress by endothelial cells.²⁷ These sensors can be compromised in some cases, thus cells may perceive states of stress that differ from the actual states. As an example, dysfunctional mechano-sensing appears to be fundamental in the development and progression of some thoracic aortic aneurysm,^28,29 noting that uncontrolled hypertension is a key risk factor for these aortopathies. Whether well sensed or not, negative feedback characteristic of mechanical homeostasis requires that the sensed states be compared with set-points to determine possible deviations (Figure 2). Consider, deviations $\mathrm{\Delta }(\dots )$ in wall shear and intramural stress, namely,

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{\Delta }{\tau }_w=\frac{{\tau }_w-{\tau }_w^o}{{\tau }_w^o},}\end{array}}$$ (5)

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{\Delta }{\sigma }_{\theta }=\frac{{\sigma }_{\theta }-{\sigma }_{\theta }^o}{{\sigma }_{\theta }^o},}\end{array}}$$ (6)

where ${\sigma }_{\theta }^o$ and ${\tau }_w^o$ are set-points, with superscript o denoting an original (homeostatic) value. See, for example, Baeyens et al.³⁰ for a demonstration of the shear stress set-point and its dependence on a vascular endothelial growth factor receptor (VEGFR3), which interacts with VE-cadherin to enable mechano-sensing. Division of the stress differences in equations (5) and (6) by the set-points is simply for convenience, thus normalizing the deviations which proves convenient because circumferential stress is typically on the order of 150 kPa whereas wall shear stress is only on the order of 5 Pa in the murine aorta (1.5 Pa in the human aorta).³¹ These very different ranges of stress suggest that different cell types developed independent methods for sensing the mechanical states to which they are exposed; there is much more to learn in this regard, noting that endothelial cells are subjected directly to the flow-induced shear stress whereas the intramural smooth muscle cells and fibroblasts are partly stress shielded from the full extent of the wall stress by the extracellular matrix in which they are embedded. Nevertheless, to account for cells that may not accurately sense the actual mechanical state, the wall shear stress in equation (5) can be premultiplied by a factor $(1-\xi )$, yielding $(1-\xi ){\tau }_w$ in the first term, and similarly circumferential stress in equation (6) can be premultiplied by $(1-\delta )$, yielding $(1-\delta ){\sigma }_{\theta }$. Values of $\xi =0,\delta =0$ capture the case of perfect mechano-sensing; values of $\delta <1,\xi <1$ capture cases of compromised mechano-sensing. Hence, sensors are critical to mechanical homeostasis; cells respond to what they perceive the stress to be, not to what the stress is actually. There is thus a pressing need for an increased understanding of the precise mechanisms of cell sensing and what can compromise such sensing.

Note further from equations (5) and (6) that no action is needed by a cell (system) if both deviations are zero, that is if $\mathrm{\Delta }{\sigma }_{\theta }=0$ and $\mathrm{\Delta }{\tau }_w=0$. In contrast, if either or both of the deviations are nonzero, the appropriate cell type will typically alter gene expression in an attempt to reduce the deviations back toward zero via negative feedback, that is, to restore the mechanical state toward homeostatic (Figure 2). For example, if the flow-induced wall shear stress is greater than its set-point, endothelial cells will typically increase their production of nitric oxide to dilate the wall and reduce the shear (cf. equation (1)); if the pressure-induced intramural stress is greater than its set-point, intramural cells will typically increase in mass (hypertrophy or hyperplasia) and/or produce more extracellular matrix, both to increase wall thickness and reduce the circumferential stress (cf. equation (2)). Recall that equations (3) and (4) reveal the associated geometric outcomes if such responses are fully mechano-adaptive. The rates at which the cells respond to deviations of perceived states from set-points can be thought of as controlled by system gains, that is, mechano-sensitivities (Figure 2).

Figure 3 shows illustrative data from a recent report from the author’s laboratory for the time-course of near mechano-adaptation of the infrarenal abdominal aorta in a mouse model of AngII-induced hypertension.³² As it can be seen, the early increase in systolic blood pressure (ultimately to γ = 186/112 = 1.66-fold at 4 weeks, noting that mean and diastolic pressures increased as well) initially increased the passive circumferential stress, but the intramural cells subsequently restored this stress toward its original value (to within 14% at 4 weeks) due in large part to mechano-adaptive thickening of the wall (h/h_o = 1.66 at 4 weeks, recall equation (4)) mainly via an increased deposition of fibrillar collagens. This return of wall stress toward its normal value, that is the set-point, reflects both the importance of cell mechano-sensing and homeostatic control (noting Cannon’s appropriate choice of the prefix “homeo,” meaning similar, in contrast to “homo,” meaning same). Indeed, a prior study showed that elimination of the α₁ subunit of the α₁β₁ integrin attenuated adventitial thickening of the common carotid artery in AngII-induced hypertension in the mouse.³³ As seen in Figure 3, luminal radius remained increased (a/a_o= 1.11 at 4 weeks, recall equation (3)) despite presumed maintenance of cardiac output ($\epsilon =1=Q/Q_o$); axial stretch decreased (down by 9% at 4 weeks), which also helped to reduce the multiaxial wall stress (not shown). A net effect of this multidimensional remodeling response was a reduction in the elastic energy storage capability of the infrarenal aorta (down by 18% at 4 weeks), which is to say that even in cases of near mechano-adaptive remodeling (equations (3) and (4)), hypertension can compromise the mechanical functionality of central arteries—one of the primary functions of an elastic artery is to store elastic energy when deforming during systole and to use this energy to work on the blood during diastole to augment flow.³⁴ Interestingly, the thoracic aorta in these same mice maladapted, with wall stresses reduced to well below normal values by 4 weeks due to an over-thickening of the wall that was ascribed mainly to adventitial fibrosis, namely exuberant collagen deposition. The primary difference between the 2 aortic regions appeared to be infiltration of abundant inflammatory cells into the wall of the thoracic aorta after the first 2 weeks of AngII infusion but much less so in the infrarenal abdominal aorta.³²

Figure 3.

Time-course of restorative (homeostatic) responses of the mouse infrarenal abdominal aorta to 28 days of angiotensin II-induced hypertension (top left, systolic pressure changes over 4 weeks, noting that mean and diastolic pressures increased as well). Shown too are normalized (to values at day 0, prior to inducing hypertension) values of luminal radius (top, middle) and wall thickness (top, right); in contrast to the acute contraction- and pressure-dependent changes geometry in Figure 1b, these changes are entrenched due to remodeling. Shown, too, are evolving values of circumferential wall stress (bottom, left) and material stiffness (bottom, middle) as well as elastic energy storage (bottom, right). Note that wall stress and stiffness first increase with increasing pressure, then tend back toward (but not to) the homeostatic values (horizontal dotted line). This partial restoration was due in large part to the increase in wall thickness, but also a decrease in axial stretch (not shown). Note that basal tone (not included) would bring both stress and stiffness even closer to original values. Data replotted from Bersi et al.³²

INFLAMMATION

Up to this point we have focused on obvious roles of biomechanical factors in hypertensive remodeling. Inflammation also plays critical roles, however. Much has been written on this subject, including excellent reviews published over the past 2 decades,^1,4,35–37 hence, consider here a different aspect of inflammation. Inflammatory processes evolved to protect against life-threatening insults, including bacterial and viral, and thus are prioritized.³⁸ It is becoming increasingly clear that prioritized processes can override homeostatic processes, including mechanical, by changing the extent of the biological response especially by resetting homeostatic set-points, gains, and rates (cf. Figure 2). We observed such resetting in the aforementioned maladaptive hypertensive remodeling of the thoracic aorta in AngII infusion—actually, we were initially uncomfortable with computational model predictions that homeostatic set-points and gains had to change for our model to describe these data,³⁹ but we then found independent support that inflammation can reset homeostatic control parameters.⁴⁰ Cells involved in this maladaptive aortic remodeling included T cells⁴¹ and macrophages,⁴² both of which emerged after, not before, the initial mechano-responses to the elevated blood pressure.³² These findings, and many others like them, are consistent with the concept that local inflammatory processes can engage if normal homeostatic processes are unable to respond completely or quickly enough to the perturbation.⁴³

Like other primary arterial cells, macrophages are highly sensitive to changes in their mechanical environment.^44,45 Recalling that material stiffness increases acutely and linearly with increases in mechanical stress, many studies have shown the importance of material stiffness in controlling macrophage phenotype, with increases in stiffness/stress tending to drive a proinflammatory phenotype.^46–48 Again, therefore, a possible insidious positive feedback loop can emerge: acute increases in stiffness/stress can drive an increased turnover of matrix that is mediated in part by macrophages, with this remodeling possibly entrenching a stiffer matrix, which in turn can drive further macrophage activity, and so forth. Yet, if early increases in stiffness/stress are appropriately resolved or at least controlled (e.g., in the infrarenal abdominal aorta in AngII infusion; Figure 3), then inflammation can be transient and promote mechanical homeostasis. Such inflammation has been termed “para-inflammation” to emphasize its role in promoting, not preventing, other homeostatic processes (Figure 4).

Figure 4.

Schema representing degrees of deviation in the perceived state from the homeostatic set-point (cf. Figure 2). When the deviation is modest, a mechanically homeostatic (stress) response can be sufficient; when the deviation is slightly greater, a homeostasis-promoting para-inflammatory response can yield a stable state, often adaptive. If the degree of deviation in the perceived state is far from the homeostatic set-point, homeostasis-preventing inflammation may result in maladaptation. Motivated by concepts presented by Chovatiya and Medzhitov.⁴³

Tissue resident macrophages, in particular, play critical homeostatic roles via the removal of apoptotic cells, cellular debris, and degraded matrix,^49,50 noting that matrix turnover always involves deposition and degradation, the latter primarily via matrix metalloproteinases (MMPs) but other proteases as well.^51,52 Importantly, increased mechanical stretch/stress can increase cellular synthesis of extracellular matrix¹³ and production/activation of the MMPs that degrade it,⁵³ again connecting hypertensive remodeling of extracellular matrix directly to changes in the mechanical environment experienced by the cells. The fundamental role of macrophages in vascular homeostasis was first demonstrated in central arteries by 2 groups in 2008, one for flow-induced inward remodeling and one for outward remodeling^54,55; recall again the importance of smooth muscle contractility in dictating the state in which matrix remodeling is entrenched.²⁴ A follow-up study by one of these groups showed further the importance of resident, not just recruited, macrophages.⁵⁶ Briefly, both of the initial studies used carotid artery ligation to decrease or increase (contralateral) blood flow to drive remodeling.^54,55 Importantly, the homeostasis-promoting inflammation (negative feedback) was transient, first increasing then quickly decreasing back to baseline with parallel increases and then decreases in heightened MMP activity. Such transient responses are to be contrasted with the chronic inflammation that appears to drive maladaptative remodeling (positive feedback), which in the aforementioned AngII-induced remodeling of the thoracic aorta did not resolve even after 7 months following stoppage of the AngII infusion.³²

In summary, inflammation can arise from many sources, including invasion of pathogens or autoimmune diseases that manifest first in tissues and organs other than arteries, and adversely affect central artery function.⁴ In contrast, para-inflammation can arise locally simply to promote homeostasis when perturbations in mechanical loading cannot be addressed quickly or completely via mechanical homeostasis (Figure 4). A mouse study of the natural history of the in vivo development of neovessels in tissue engineering revealed clearly the potentially delicate balance between favorable and unfavorable effects of inflammation on the vasculature.⁵⁷ There is, therefore, a pressing need to understand better the diverse roles of inflammation and the cellular phenotypes that define it in the vasculature.

MURINE MODELS—FOCUS ON TAC

Early rodent models of hypertension included Goldblatt surgical models to affect renal function and, of course, the spontaneously hypertensive rat. Notwithstanding all that has been learned, and continues to be learned, from rat models,⁵⁸ mice have emerged as the model of choice in much of vascular research, including hypertension. Many different methods have been used to induce hypertension in mice,^59,60 including use of DOCA salt or mini-osmotic pumps that provide continuous infusion of norepinephrine or AngII. High salt diets combined with chronic l-NAME to block endothelial derived nitric oxide synthase are also common. Each model has advantages and disadvantages. Given the focus herein on biomechanical factors and mechanical homeostasis, however, consider findings from the so-called TAC (transverse aortic constriction) model of proximal central arterial hypertension.

Briefly, in the TAC model one places a constricting ligature around the vessel of choice, often the aortic arch between the brachiocephalic and left common carotid artery, which narrows the aorta locally and mechanically increases resistance to blood flow; blood pressure thus increases proximal to the ligature to preserve flow ($\mathrm{\Delta }P=QR$where $\mathrm{\Delta }P$ is the pressure drop that drives flow Q against resistance R). More specifically, despite modest changes in mean pressure proximal and distal to the ligature, this technique primarily increases pulse pressure proximally and decreases it distally. Indeed, subsequent remodeling of the hypertensive right relative to the left common carotid artery correlates better with changes in pulse than mean arterial pressure.⁶¹ This remodeling is characterized, similar to that in AngII-induced hypertension, by increases in luminal radius and wall thickness, the latter due in large part to an increased deposition of fibrillar collagens and accumulation of glycosaminoglycans. Monocyte chemoattractant protein-1 (MCP-1) increases as well, consistent with monocyte/macrophage involvement in the remodeling.

TAC elevates blood pressure mechanically, but increased wall stresses in the hypertensive segments drive local AngII production and transforming growth factor-beta expression⁶² consistent with earlier studies on mechanically stressed smooth muscle cells.¹³ Local production of AngII may also be augmented by renal sensing of a decreased pulse/perfusion pressure due to the proximal aortic constriction. Regardless, the media thickens without a change in medial cell density while the adventitia thickens with a marked increase in cell density, with possible differentiation of fibroblasts to the myofibroblast phenotype and infiltration of inflammatory cells. There are also significant increases in fibrillar collagens as well as increases in MCP-1 that parallel increases in interleukin-6 (which also plays key roles in AngII-induced hypertension⁴²) and MMPs. The selective AngII type I receptor (AT1R) antagonist losartan attenuates, but does not eliminate, the hypertension-induced remodeling. In a follow-up study, these investigators showed further the angiotensin converting enzyme inhibitor captopril failed to reduce adverse aortic remodeling in this TAC model. Rather, a combined treatment, captopril plus C21, an AT2R agonist, reduced MMP9 expression more than captopril alone, resulting in slowed remodeling and less aortic dilatation. It is the central role of AngII in hypertensive remodeling of central arteries, even when induced mechanically, that has driven the widespread use of chronic AngII infusion in mice, which provides important insight though complicating interpretation because exogenous AngII is proinflammatory, not just a consistent elevator of blood pressure.

Another common aortic banding model, actually full ligation, places an occluding ligature around the abdominal aorta between the renal arteries, directly engaging the renin–angiotensin system. In a nice longitudinal study, it was found that early remodeling (<8 days) of the common carotid artery was driven by changes in smooth muscle cell contractility whereas later remodeling (~56 days) was driven more by matrix remodeling with vascular tone restored toward normal.⁶³ This study supports further the aforementioned observation that vasoregulatory changes often define the new mechanical state in which matrix is subsequently entrenched via heightened turnover.^14,15 Hence, the phenotype of the smooth muscle cell appears to be modulated temporally in hypertensive remodeling in central arteries, with augmented contractile then synthetic capability returning to the more quiescent matrix phenotype at long times. Longitudinal studies are thus critical for full understanding, though often not performed in otherwise revealing studies.

SMOOTH MUSCLE PHENOTYPE

Notwithstanding critical roles of endothelial cells, adventitial fibroblasts, and possibly inflammatory cells, both resident and recruited, it is the smooth muscle cells that ultimately serve as the central node that dictates arterial function. It should thus not be surprising that vascular remodeling in hypertension differs dramatically across large (central) arteries, medium-sized (muscular) arteries, and small (arterioles) arteries (Figure 5). Notably, in each case there is a tendency to restore the mean circumferential stress toward normal via appropriate thickening of the wall (equation (4)), but large arteries tend to outward remodel while arterioles tend to inward remodel, encroaching on the lumen.^59,64 That is, only the medium-sized muscular arteries tend to mechano-adapt fully, both in hypertension and aging.⁹ One cannot fail to note that these differential responses reflect the basal smooth muscle phenotypes across these types of vessels: a matrix phenotype in central arteries that is primarily responsible for mechano-sensing and mechano-regulating the matrix of the wall to ensure proper resilience and yet appropriate compliance and strength,⁶⁵ a contractile phenotype in muscle arteries that enables coarse mechano-regulation of flow, and an enhanced contractile phenotype in arterioles that enables fine mechano-regulation of flow that includes an integrin-dependent myogenic (pressure-dependent) contractility.⁶⁶ As in flow-induced remodeling,²⁴ it is the contractile state in which matrix turns over that appears to dictate the differential outcomes across the 3 types of arteries. Modest contractility in central arteries does not prevent the pressure-induced elastic distension that allows matrix turnover at a dilated state whereas exuberant myogenic responses in arterioles allow matrix turnover in a narrowed state. Interesting, it was recently found that differential contractile capacity of the thoracic aorta from mouse-to-mouse associates with adaptive vs. maladaptive remodeling in AngII-induced hypertension,⁶⁷ suggesting that increased contractility reduces wall stress (via the aforementioned ${\sigma }_{\theta }=Pa(P,C)/h(P,C)$) and thus the mechano-stimulus for resident or recruited cell-mediated matrix turnover. Studies using single-cell RNA sequencing have great promise to detail phenotypic differences across vessel types in health as well as hypertension,⁵⁸ but should ultimately be combined with detailed studies of the layer-specific mechanics to delineate mechanobiological and immunobiological effects, the latter of which appear to be less in small than large arteries in the spontaneously hypertensive rat⁵⁸ and similarly appear to be less in medium-sized than large arteries in patients having autoimmune diseases.⁴

Figure 5.

Schema of the differential responses of large (elastic), medium-sized (muscular), and small (arterioles) arteries to a sustained increase in blood pressure (hypertension, HTN). All 3 types of vessels tend to thicken in an attempt to restore wall stress back toward normal (cf. Figure 3), yet the luminal response is very different, likely reflecting differential smooth muscle phenotypes. Interestingly, dilatation of the elastic arteries is not ideal, but it can help reduce the increased pulse wave velocity that results from structural stiffening; similarly, luminal encroachment in the arterioles is not ideal, but it may help attenuate the penetration of pulse pressure waves into the microcirculation.^9,10

DISCUSSION

The term hypertension tends first to elicit thoughts of high blood pressure, but it was originally coined to emphasize the “excessive tension” in the wall of an artery that results from elevated pressure. This early choice of words was remarkably insightful for it was not until the mid-1970s that cell culture studies revealed that all cell types of the arterial wall are exquisitely sensitive to their local mechanical environment, including increased tension or stress. Hence, mechanobiological mechanisms often drive initial remodeling in response to early increases in wall stress due to high blood pressure, with the degree of smooth muscle contractility and stiffness of the extracellular matrix playing equally important (though perhaps different temporal) roles in determining the degree to which wall stress increases. When these initial mechanobiological responses are either not rapid enough or not sufficient, para-inflammation appears to play a complementary role in seeking to promote homeostasis. For reasons not yet clear, however, inflammation can quickly transition in central arteries from promoting to preventing homeostasis, the latter of which is often characterized by adventitial fibrosis and an associated marked structural stiffening that adversely affects central hemodynamics and ultimately end organ perfusion via an insidious positive feedback loop, thus increasing morbidity and mortality. There is, therefore, continued need to understand arterial remodeling in hypertension in terms of mechanical homeostasis and its loss, particularly how smooth muscle cell phenotype (contractile, synthetic, and degradative⁶⁸) and inflammatory burden play central roles.

Although we focused on 2 primary components of stress (equations (1) and (2)), all arteries exist in vivo under an axial prestretch (or prestress), which contributes significantly to the overall mechanical state sensed by the intramural cells.⁶⁹ Direct manipulation increasing this axial stretch reveals that intramural cells again respond quickly and dramatically to remodel the wall homeostatically.⁷⁰ Although difficult to assess in vivo, hypertensive remodeling results in marked reductions in axial stretch,³² which are evidenced when transecting the vessel and which help reduce the increased multiaxial stress in the wall. Extreme reductions in this axial prestress can promote arterial tortuosity, however, which can adversely affect the hemodynamics.⁷¹ Again, therefore, that which separates beneficial vs. adverse remodeling is often a matter of the degree, not type, of response. Computational models allow differential effects of multiaxial changes in wall stress on arterial remodeling to be predicted, pointing to experimental manipulations that can be used to test resulting hypotheses.³⁹

In summary, mechanical homeostasis preserves central artery structure, properties, and function over long periods in the face of modest perturbations. In hypertension, the vasoactive capacity of the wall and inflammatory burden strongly influence the ability of homeostatic mechanisms to adapt the arterial wall or not, with different contractile capacity in large, medium-sized, and resistance vessels influencing their differential responses. Aortic maladaptation, primarily via adventitial fibrosis, significantly diminishes biomechanical function and impairs hemodynamics, ultimately compromising end organ perfusion and thus driving the associated morbidity and mortality. There is, therefore, a need for increased attention to controlling smooth muscle phenotype and inflammation in hypertensive remodeling of central arteries.

FUNDING

This work was supported, in part, by grants from the US National Institutes of Health (R01 HL105297, P01 HL134605).

DISCLOSURE

The author declared no conflict of interest.