A first step towards recognizing the fundamental role of smooth muscle tone in large artery (dys)function?

Macro-/microstructural features and mechanical properties of blood vessels vary considerably at different locations in the systemic and pulmonary circulations, reflecting the specific function of these vessels in the cardiovascular system [1–3]. Historically, large arteries have been classified into two phenotypes: elastic and muscular. This differentiation is based on their composition, and consequently, their functional role in the circulation. Elastic arteries are close to the heart, present a high concentration of elastin, and have the function of smoothing the intermittent pulsation generated by the ventricles’ contraction into a more continuous pressure profile, thus maintaining flow during diastole (’Windkessel function’) [4–7]. On the other hand, muscular arteries are located more distally in the circulation, present a higher content of smooth muscle cells (SMCs), and thus, together with arterioles, play an important role in autoregulatory mechanisms [5,6]. Despite this classification, the boundary between these two arterial phenotypes is blurred as mechanical properties and wall composition change gradually while moving away from the heart [6–8]. Furthermore, SMCs constitute a considerable portion of the wall volume also in the aorta [6,8,9], so that, although often overlooked, their role in the mechanics of elastic arteries should not be neglected. Indeed, many animal studies have shown that aortic contraction can be elicited upon stimulation with vasoactive drugs/agents, both ex vivo and in vivo[5,9,10]. Furthermore, we have suggested that such contraction may play a pivotal role in preventing maladaptation and fibrotic remodelling of the aorta in angiotensin II-induced hypertension in mice [9]. Data on elastic artery contractility in humans is sparser than in animals. Stewart et al.[11] investigated the effect of nitroglycerin administration on the distensibility of the carotid artery. They found that, in hypertensive patients, a 22 mmHg decrease in brachial blood pressure (BP) was accompanied by a ∼0.65 mm increase in carotid diameter and no change in carotid distensibility, thus signifying a considerable alteration of carotid artery mechanics via relaxation of the SMCs. Similarly, Stefanadis et al.[12] studied the effect of Diltiazem administration on aortic function in normotensive and hypertensive patients. Invasive measurements of local diameter and pressure waveforms showed that, in both groups, SMC relaxation considerably alters the aortic pressure–diameter relationship and significantly increases aortic distensibility. In agreement with these findings, Faconti et al.[13] showed dissociation between acute changes in aortic and carotid–femoral pulse wave velocity (aPWV and cfPWV, respectively) and changes in BP following lower limb venous occlusion. Due to the nonlinear mechanical properties of the arterial wall, PWV is expected to acutely increase with increasing BP [14]. Inversion of this relationship, as observed by Faconti et al. following venous occlusion, might suggest that sympathetic activation in response to reduced venous return may trigger a mechanically significant SMC contraction in the human aorta. In summary, these findings warrant caution in dismissing the role of SMC contraction in the mechanics of elastic arteries. In biomechanical engineering settings, the complex nonlinear mechanical behaviour of the arterial wall is commonly modelled by mathematically describing the ability of the wall material to store elastic energy upon deformation [9,14,15]. Such modelling typically accounts for the summed contribution of the most mechanically relevant wall constituents: elastin, collagen, and if contractility data are available, SMCs. Ever since the pioneering work of Roach and Burton [16] on elastin/collagen-degraded arteries, elastin has been modelled as an elastomeric material, showing a concave pressure–diameter relationship (Fig. 1, grey area). Collagen fibres, on the other hand, because of their wavy arrangement in unloaded arteries, get recruited at higher pressures and their gradual uncrimping leads to a nearly exponential pressure–diameter relationship (Fig. 1, hatched area) [9,14,15,17]. Modelling SMC contraction is considerably more complex. Similar to skeletal muscle cells, because of the interaction between actin and myosin filaments, the ability of SMCs to generate force depends on the degree of stretch of the SMC [18]. Experimental work has shown that this SMC force–stretch relationship is nearly parabolic at most locations along the arterial tree, with the cells operating in the ascending arm of the parabola at physiological pressures [19,20]. Although models have been developed to effectively replicate this response [21,22], the limited amount of mechanical data that can be measured in vivo, together with the complexity of these formulations, strongly limits their use in clinical settings.FIGURE 1: Schematic representation of the pressure–diameter relationship of a highly compliant elastic artery with relaxed (solid lines) and contracted smooth muscle cells (SMCs, dashed lines). Fill patterns illustrate the contribution from elastin (grey solid), collagen (hatched), and SMCs (dotted). In the relaxed state, the systolic–diastolic pressure range (P s –P d) falls in the diameter range (Dsr−Ddr) where the gradual recruitment of collagen fibres takes place, thus conferring an exponential shape to the pressure–diameter relationship. When SMCs contract, the same pressure range is attained at considerably lower diameters (Dsc−Ddc). In this diameter range, the extracellular matrix (ECM) contribution is largely determined by elastin with minimal collagen recruitment. As a result, the exponential pressure–diameter relationship assumption may not hold for the contracted vessel.Single-exponential pressure–diameter relationships have been previously proposed as a practical and effective compromise between accurate description of the arterial wall behaviour and model simplicity to guarantee applicability in clinical settings [23,24]. Indeed, as shown in Fig. 1, the gradual recruitment of collagen fibres typically unfolds around the physiological pressure range, conferring an exponential shape to pressure–diameter curves in this range [16]. Although the model accuracy may be less than optimal in elastin-dominated portions of the pressure–diameter relationships (Fig. 1), these portions concern subphysiological pressures (<50 mmHg), and are, hence, of limited interest in most clinical settings. A possibly more clinically relevant limitation consists in the fact that these models assume that arteries exhibit a constant mechanical behaviour (i.e. pressure relates directly and uniquely to diameter), while, as argued above, this can potentially be dramatically altered by SMC contraction. The current issue of the Journal of Hypertension features an elegant study that expands the single-exponential modelling approach to capture the arterial wall response at different levels of SMCs contraction. Not unlike the classical biomechanical modelling described above, Pewowaruk and Gepner [25] theorize that the arterial wall pressure–diameter relationship can be modelled as the sum of two exponential terms: the first describes the response of the extracellular matrix (ECM) (i.e. mainly collagen and elastin) and preserves the form of the exponential function proposed by Hayashi et al.[23], whereas the second represents the contribution of the SMC contraction. The complete model requires estimation of only four parameters – two distinct stiffness indices for the ECM and SMC, a SMC tone parameter quantifying the degree of contraction, and a reference diameter – and can be easily applied to any clinical dataset in which haemodynamic data (e.g. pressure–diameter relationship, pulse wave velocity, or arterial distensibility) have been acquired at at least two different states of SMC contraction. In addition to the formulation and applicability of the model itself, the main physiologically relevant finding of Pewowaruk and Gepner's study is the potentially different functional role played by SMC contraction at different locations in the arterial tree. In elastic arteries, such as the aorta and carotid artery, the SMC stiffness parameter is lower than that of the ECM and contraction leads to an overall softening of the arterial wall despite considerable reduction of the luminal diameter. The opposite holds for more muscular arteries (e.g. brachial, femoral, and tibial arteries), where vasoconstriction is accompanied by considerable wall stiffening. The pragmatic single-exponential nature of the proposed model also represents its major limitation. As argued above, exponential models most effectively describe the pressure–diameter relationship of arteries in the pressure range characterized by the gradual collagen recruitment. It is relatively easy to realize, however, that this pressure range is not independent from the SMCs contraction state but shifts towards higher pressures with increasing vascular tone [26]. As illustrated in Fig. 1, SMC contraction entails achieving a lower diameter for any given pressure level (from Dsr−Ddr to Dsc−Ddc). Therefore, at any pressure, collagen fibres will be less stretched when the vessel is contracted compared with the fully relaxed state. In highly compliant, elastin-rich arteries, such as the aorta of young individuals, the onset of collagen recruitment happens in the physiological pressure range (e.g. 120/80 mmHg systolic/diastolic) in the relaxed state (Fig. 1, hatched area) [27]. SMC contraction, however, may reduce the operating diameter range to below the point of collagen recruitment (i.e. to the range where elastin is the mechanically dominant passive constituent). As elastin behaves elastomerically, this potentially weakens the validity of the exponential pressure–diameter relationship assumption (Fig. 1, dashed line). The impact of this limitation is expected to be much less significant in stiffer arteries (e.g. more peripheral arteries and arteries of older individuals), where collagen fibres play a major role throughout most of the low-to-high pressure range. A second potential limitation of the approach proposed by Pewowaruk and Gepner consists in the use of an exponential function to model the SMC contribution to the pressure–diameter relationship. As discussed above, the contractile force generated by SMCs is dependent on the stretch each SMC is subjected to, following a parabola [19,20]. Although SMCs typically operate in the ascending arm of this parabolic relationship, their ability to generate force is expected to reach a maximum at a critical diameter distension and then steadily decline. Clearly, this type of stretch-dependent response cannot be captured with an exponential relationship, which is monotonically increasing in nature. Gade et al.[28] recently proposed an alternative approach to model SMC contraction in vivo based on Rachev's biomechanical model [21], which accounts for the aforementioned parabolic force–stretch relationship. A careful analysis, however, suggests that this critical SMC stretch is reached at pressures well above the physiological range and that the formulation proposed by Pewowaruk and Gepner is conceptually equivalent to that used by Gade et al.[28] at physiological pressures. Furthermore, some of the parameters of Rachev's model, such as the SMCs stretch corresponding to their peak force, cannot be estimated in vivo, thus requiring the postulation of assumptions based on literature data [28], which renders the model less patient-specific. The modelling approach proposed by Pewowaruk and Gepner constitutes an important step towards unravelling the important role played by SMCs in arterial (dys)function in humans, providing a simple and effective tool that can be easily applied to clinical data. Fitting models to sparse in-vivo data mandates that the fitted models are as simple as possible; the aforementioned limitations are, hence, a logical consequence of this. Pewowaruk and Gepner were not put off by these limitations but rather worked with them, resulting in this innovative work. After all, all models are wrong, but some are useful[29]. ACKNOWLEDGEMENTS This work was supported by the European Union's Horizon 2020 Research and Innovation program (grant 793805 to B.S.). Conflicts of interest There are no conflicts of interest.