This study describes several important mechanical design features of the aorta of a typical poikilothermic vertebrate. A strong functional similarity to the aorta of mammals is apparent, but some structural and mechanical differences are seen that reflect the lower pressure and simpler haemodynamics of the poikilothermic circulation.
The aorta is highly distensible, resilient and non-linearly elastic, giving it the requisite properties to act as an effective storage element in the arterial circulation.
An abrupt transition from high compliance (low elastic modulus) to relatively low compliance (high elastic modulus) takes place at pressures above the resting physiological range of 2-4 kPa. This behaviour reflects the composite nature of the artery wall in which rubbery elastin fibres and relatively rigid collagen fibres are the predominant elements.
The longitudinal tethering of the aorta when inflated is due primarily to anisotropy in elastic properties, rather than to links to the axial skeleton by branch vessels or connective tissue.
No significant changes in elastic properties or connective tissue content occur along the length of the toad arterial tree, in contrast to the situation in mammals.
The aorta is an important elastic element in the arterial circulation. The vessel wall expands during systole and recoils during diastole, thereby storing and releasing elastic strain energy and smoothing the pulsatile output of the heart. An important benefit is the reduction of the arterial pressure pulse and, consequently, protection of the small exchange vessels from the high shear forces associated with pulsatile flow, and a reduction of the total energy required to circulate the blood (Taylor, 1964). To perform this function effectively, the aorta must have non linear elastic properties in order to be a compliant reservoir at low pressures but be stiff enough to resist rupture at high pressures.
Most of our information on structure and mechanical properties of arteries comes from studies on mammalian tissues (Bergel, 1961; Milnor, 1982). The artery wall is a composite structure whose non-linear elastic properties result from the combination of rubber-like elastic and relatively inextensible collagen fibres. The transition from a highly compliant to a relatively stiff vessel takes place at about 10–12 kPa, the mean resting blood pressure. In mammals, the distensibility of the aorta decreases as the ratio of collagen to elastin increases along the arterial tree. The elastic properties of the artery wall are important in determining the haemodynamic behaviour of the arterial system.
Poikilothermic vertebrates generally have much lower blood pressures and often lower heart rates than do mammals (Shelton and Jones, 1968; Jones et al. 1974; Shelton and Burggren, 1976; Langille and Jones, 1977). Nevertheless, the aortas of reptiles, amphibians and fish appear to have non-linear elastic properties (Goto and Kimoto, 1966; Satchell, 1971; Gibbons and Shadwick, 1989) that are comparable to those observed in mammals, but presumably appropriate for function in a lower-pressure system. However, very little is known about the relationship between structure, connective tissue composition and the elastic properties of the aorta in any lower vertebrate species. The purpose of this investigation is to make a detailed study of the aortic mechanical properties and their structural basis in the toad Bufo marinus, and to compare the design features of this arterial system to that of mammals. In a subsequent study (Gibbons and Shadwick, 1991), the effects of arterial elasticity on haemodynamic properties in the toad will be examined.
MATERIALS AND METHODS
The experiments were performed on the toad, Bufo marinus L. Animals weighing 200-500 g were maintained in tanks at room temperature with access to water. They were fed mealworms weekly. The animals were killed by injection of MS-222 into the peritoneal cavity (Sandoz, 1:1000, 0.022 ml g−1 body mass). The heart, aortic arches and dorsal aorta were exposed by a ventral midline incision. Four regions were arbitrarily designated, as shown in Fig. 1, and their in situ lengths determined before excision of the whole vessel. Experiments were performed on these aortic segments at room temperature within a few hours of death.
Artery segments (about 5 cm long) were placed in a chamber containing amphibian saline, and cannulated at one end with a blunt 18-gauge syringe needle and connectors leading to a pressure reservoir and a variable-speed pump. Once the air bubbles had been cleared, the distal end and the branch segments were ligated, and the vessel was extended to its in vivo length. The preparations were generally leak-free. Slow inflation-deflation cycles (lasting 1–2 min) were performed using a variable-speed pump. Conditioning cycles (usually 2-3) were run until the pressure-diameter curves were stable. The pressure during one cycle was raised to approximately 10 kPa and lowered back to zero. Pressure was measured continuously using a P23Db Gould pressure transducer, while diameter was determined simultaneously with a video dimension analyzer system or VDA (Instrumentation for Physiology and Medicine, model 303), as described by Fung (1981). Some inflation tests were also performed on untethered vessel segments to measure the effect of increasing pressure on longitudinal extension. In this case, lengths were measured for step-wise pressure increments by using a microscope digital micrometer (Wild-Leitz MMS235).
Uniaxial force-extension tests were made on a tensile testing machine (Monsanto Tensometer T10) in both the circumferential and longitudinal directions. For the circumferential testing, arterial rings were cut with 2 mm widths. They were placed around two metal L-shaped hooks. One hook was anchored at the base, while the other was connected to the force transducer attached to the moveable head. Longitudinal testing was done on vessel segments (2–3 cm long) held in vicetype clamps in the tensometer. Stress was then calculated as σ=F/A, where Fis the tensile force applied and A is the tissue cross-sectional area perpendicular to the force. Length changes (ΔL) were measured between surface markers by the video dimension analyzer. The markers were placed in the central region of the specimen to avoid any clamp effects. Strain was calculated as ϵ= ΔL/L0, where Lo was the initial length.
Connective tissue content
Histological sections were made from vessel segments fixed in Bouin’s solution (Humason, 1979) at constant pressures ranging from 0 to 4 kPa. Sections were stained with aldehyde-fuchsin or Verhoeff’s stain to identify elastin fibres. Picro-ponceau or Mallory’s Trichrome stain were used to identify collagen and muscle (Humason, 1979). The content and arrangement of the collagen and elastin were then examined in both the transverse and longitudinal directions. Sections were stained from arterial segments at the four arterial sites where inflation tests were made. For comparison, a fresh rat aorta was fixed at OkPa and stained using the same methods. To determine the smooth muscle disposition of the aortic wall, sections were stained with haematoxylin and counterstained with eosin.
The elastin content was determined by Lansing’s method. Freeze-dried vessel segments were rehydrated and then treated with 0.1 moll−1 sodium hydroxide for 45 min at 100 °C, followed by subsequent washings in distilled water. The residue was then dried and weighed (Lansing et al. 1952). The elastin content was this fraction’s dry mass expressed as a percentage of the total dry mass. Collagen content was estimated from the hydroxyproline content, which was determined with an automated amino acid analyser (LKB Alpha Plus, model 1215) or by a manual colorimetric assay (Berg, 1982). Values for collagen content were corrected to account for the hydroxyproline present in the elastin according to published information on the composition of elastin in Amphibia (Sage and Gray, 1979). The collagen content was calculated as a percentage of the total dry mass. A collagen to elastin ratio was also calculated.
Upon removal from the animal, the aorta shortened by approximately 30% in each of the designated regions, causing a corresponding increase in diameter. Interestingly, this is comparable to values reported by Bergel (1961) for the dog thoracic and abdominal aorta, which showed shortening of 32–34%. When left untethered during an inflation, the length increased with distending pressure until the upper physiological pressure was reached (Fig. 2). At this point, the aorta had regained its in vivo length. Increasing the pressure beyond the physiological range resulted in virtually no further lengthening. The vessels were stretched to the in vivo length for subsequent inflation tests.
A typical inflation cycle (Fig. 3) illustrates the non-linear and viscoelastic behaviour of the aorta. Compliance is relatively high up to the physiological pressure range (2–4 kPa), but thereafter the slope of the pressure-volume curve increases sharply as the artery wall becomes very stiff. The difference between the inflation and deflation portions of this curve indicates that some viscous energy losses occur during each cycle. The area under the ascending curve represents the total work done on inflation, while the area under the descending curve represents the strain energy recovered by elastic recoil. Hysteresis, the proportion of energy lost through viscous processes during each cycle, is the ratio of the area within the loop to the total area beneath the inflation limb. For the toad aorta, the hysteresis was low at each of the arterial sites tested (17% in this example). Resilience, the proportion of strain energy recovered, was correspondingly high, at 80-85%.
Fig. 4 shows the mean values of pressure versus circumferential strain from inflation tests of the four arterial sites, as shown in Fig. 1, in 10 animals. These curves are not statistically different, suggesting that the toad aorta has uniform elastic properties along its length. The circumferential strain of about 0.75 at the mean physiological pressure of 3 kPa is comparable to the level of strain in the rat thoracic aorta at its mean blood pressure of 11 kPa (Cox, 1978).
Stress-strain relationships, calculated from the data in Fig. 4, were all J-shaped and also showed no significant differences among the four arterial sites. The level of circumferential wall stress attained (approximately 1×105 Pa) at mean blood pressure was nearly equal to that in the aorta of the rat at its much higher mean blood pressure (Gibbons and Shadwick, 1989).
Fig. 5 shows the incremental elastic modulus as a function of distending pressure. These plots demonstrate that the arterial wall stiffness is highly dependent on pressure and increases considerably over the physiological range, from about 0.2 to 0.7 MPa. At the mean resting blood pressure of 3 kPa, the elastic modulus is approximately 0.4 MPa. This is similar to values reported for the aorta of other poikilotherms (Gibbons and Shadwick, 1989) as well as for the aorta and upper thoracic aorta of mammals (See Fig. 10) at their respective blood pressures. The similarity of the four plots in Fig. 5 indicates that the toad aorta, unlike that of mammals, has uniform elastic properties along its entire length.
Potential vascular muscle activators, such as acetylcholine and epinephrine (West and Burrgren, 1984), were applied to the arterial wall and perfused through the vessel in an attempt to stimulate muscle activity. These tests were made within a few minutes of death. Although many different concentrations of these drugs were tested, the dorsal aorta showed no measurable mechanical response to these chemicals.
The stress-strain curves resulting from uniaxial tests (Fig. 6) show that the aorta is much stiffer longitudinally than circumferentially. This is in agreement with biaxial tests in which changes in pressure produced larger strain increments circumferentially than longitudinally (Figs 2, 4).
Structure and histology
Table 1 shows average dimensions of the outside radius (R) and wall thickness (h) of the toad aorta at a mean pressure of 3 kPa. Both R and h are important in determining the mechanical properties of the arterial wall. This aorta is surprisingly uniform in these dimensions and in the R/h ratio along its length, i.e. there is virtually no ‘geometric taper’ as there is in the mammalian aorta (see Fig. 9). The average radius of 1.27 mm and wall thickness of 0.055 mm give an R/h ratio of 23 at mean blood pressure. In contrast, the R/h ratio for a mammalian thoracic aorta at mean blood pressure is only about 10 (McDonald, 1974). The relative thinness of the toad aorta is illustrated in Fig. 7; while the diameter of the aorta from the toad and a similar-sized rat are about the same, the wall of the former is only half as thick as that of the latter.
In both the transverse and longitudinal sections of the toad aorta, the elastin (stained black) appears as 4–5 layers within the wall (Fig. 7). These sections together show that the elastin must be arranged as concentric cylindrical sheets. In uninflated vessels, the luminal elastica is thicker and wavier than the other elastin layers found throughout the wall. Smooth muscle cells, indicated by the stained nuclei in Fig. 7D, are not very abundant. In both transverse and longitudinal planes, the elastin layers were relatively thin and sometimes appeared incomplete or fused to adjacent layers. In a rat aorta of similar diameter, the elastin lamellae were more numerous (9–11), more uniform and well-defined, and thicker than in the toad aorta (Fig. 7). In the latter there is no distinct outer layer of collagen; rather, it occurs in large amounts among the elastin lamellae. There is relatively less elastin and more collagen in the arterial wall of the toad than in that of the rat. Fig. 8 shows sections of toad aorta fixed at pressures of 1.0, 1.7, 2.5 and 4.0 kPa. With increasing pressure, the elastin fibres become straighter and more taut as they are loaded within the physiological pressure range. Concomitantly, the interlamellar space and lamellar thickness decrease.
The elastin content of the toad aorta was approximately 15% of the total dry mass, while the collagen content was about 35% of the dry mass. These values were consistent at all four sites (Table 1), indicating that the aorta has a uniform composition along its length. The collagen to elastin ratio of 2.3 in the toad aorta is more than five times that found at the level of the aortic arch in mammals (see Fig. 10).
This study describes several important mechanical design features of the aorta of a typical poikilothermic vertebrate. The high distensibility, resilience and nonlinear elasticity are the requisite properties for the vessel to act as an effective storage element in the arterial circulation. In this respect, a strong functional similarity to the aorta of mammals is apparent, although specific structural and mechanical differences are seen that reflect the much lower level of pressure and simpler haemodynamics for which the toad vessel is suited.
Our initial observation that a longitudinal recoil of about 30% occurred when the aorta was excised is surprisingly similar to the behaviour of the aorta of mammals. With inflation to the normal physiological pressure of about 3 kPa, an unrestrained toad aorta regains its in vivo length (Fig. 2), i.e. the longitudinal stress due to pressure alone causes the lengthening, and tethering to the axial skeleton is unimportant. In fact, there are no branch vessels between the subclavian and coeliacomesenteric arteries, and very few along the rest of the aorta. During a normal pressure pulse, the aorta will lengthen only slightly, owing to the relatively limited longitudinal extensibility at pressures above 2 kPa. Similarly, in the mammalian thoracic aorta, longitudinal strains of only 1% or less occur with each heart beat (Patel and Fry, 1964), although this behaviour has generally been attributed to a high degree of tethering by the intercostal arteries and connective tissues (Milnor, 1982). However, Van Loon et al. (1977) demonstrated that there is a maximal length to which mammalian arteries elongate when pressurized without tethering, and that this is very close to the in vivo length. Thus, to a large extent, the relatively constant aortic length in vivo results from the mechanical anisotropy of the wall material, i.e. the stiffness is greater longitudinally than it is circumferentially (Fenn, 1957; Patel et al. 1969; Vaishnav et al. 1972; Dobrin, 1983). This type of anisotropy in the toad aorta is evident from the uniaxial tests shown in Fig. 6. Interestingly, similar mechanical behaviour has been observed for the aorta of octopods and squid (Gosline and Shadwick, 1982; Shadwick and Gosline, 1985; Shadwick and Nilsson, 1990), soft-bodied animals in which no skeletal tethering of the aorta is possible.
The non-linear and viscoelastic properties of the toad aorta reflect its composite structure; it consists mainly of highly extensible elastin and relatively inextensible collagen. With inflation, the elastic lamellae are loaded and become straightened as the pressure approaches the physiological range (Fig. 8). Above 3 kPa there is an abrupt decrease in the vessel distensibility and an increase in the circumferential elastic modulus (Figs 3, 5), above what could be attributed to elastin alone (Aaron and Gosline, 1981). In the mammalian aorta, an abrupt transition from high compliance to high stiffness also occurs, but at much higher mean blood pressures of 11–12 kPa (Bergel, 1961; Gibbons and Shadwick, 1989). In both toad and mammal, inflation of the thoracic aortas to their respective physiological mean pressures results in circumferential strains of 0.6–0.8, an elastic modulus of about 0.4 MPa and mechanical hysteresis of 15–20% (Bergel, 1961; Cox, 1978; Gibbons and Shadwick, 1989). Thus, the aorta of the toad appears to be designed to have essentially the same functional properties as that of mammals, but at much lower pressures.
A significant finding in this study is that, in contrast to mammals, there is no geometric or elastic ‘tapering’ in the arterial tree of the toad. The dimensions and elastic properties are relatively uniform along the length of the aorta, whereas the mammalian aorta exhibits a progressive decrease in radius and increase in elastic modulus peripherally (Figs 9, 10). The distribution of mechanical properties in these aortas is correlated with their connective tissue composition, i.e. the toad aorta has a relatively constant quantity of elastin and collagen along its length, while in the mammalian aorta an increase in the ratio of collagen to elastin occurs distal to the heart (McDonald, 1974; Fig. 10). In the mammalian system, this elastic and geometric tapering together cause a continuous increase in pressurewave velocity and aortic impedance distal to the heart, giving the aorta the properties of a non-uniform transmission line (McDonald, 1974; Milnor, 1982). The lack of any significant elastic or geometric taper in the toad aorta, by contrast, suggests that this vessel is more suited to function as a simple Windkessel than as a complex transmission line. These haemodynamic features are considered in more detail in the following paper.
The question of how the non-linear elastic properties occur at much lower pressures in the toad aorta than in that of the mammal can be addressed by considering structural and connective tissue compositional differences. The aortic R/h ratio is nearly 2.5 times higher in the toad, resulting in higher wall stress (equation 1) and, consequently, higher modulus for any given pressure than in the mammalian vessel. The higher fraction of collagen in the former may also contribute to the transition from high to low compliance at the relatively low physiological pressure in the toad compared to that in the mammalian system. In a comparative study, Wolinsky and Glagov (1967) established that the number of elastin layers, N, in the mammalian aorta was a function of body size, or specifically of aortic radius at mean blood pressure. Consequently, the circumferential wall tension per layer (=Pr/N) remains nearly constant at about 2 Pa m in all species. In contrast, the quantity and mechanical behaviour of the elastin layers in the toad aorta do not fit this general pattern, instead becoming straightened at a relatively low pressure (Fig. 8) that generates a tension per layer of only about 0.5Pam.
In conclusion, specific differences in structure, connective tissue architecture and mechanical properties are demonstrated in the toad aorta compared to that of mammals. Overall, however, the mechanical behaviour of the aorta is functionally comparable in the two taxa, although their range of physiological pressures is very different. Thus, the aorta has the same capability to act as an effective elastic energy storage element in the circulation of this poikilothermic vertebrate as in that of mammals. The contribution of these elastic properties to arterial haemodynamics is explored in the following study.
This work was supported by a Graduate Research Grant from the University of Calgary to C.A.G. and by an operating grant and Research Fellowship from the Natural Science and Engineering Research Council of Canada to R.E.S.