Spatial Distribution and Mechanical Function of Elastin in Resistance Arteries
A Role in Bearing Longitudinal Stress
Objective—Despite the role that extracellular matrix (ECM) plays in vascular signaling, little is known of the complex structural arrangement between specific ECM proteins and vascular smooth muscle cells. Our objective was to examine the hypothesis that adventitial elastin fibers are dominant in vessels subject to longitudinal stretch.
Methods and Results—Cremaster muscle arterioles were isolated, allowed to develop spontaneous tone, and compared with small cerebral arteries. 3D confocal microscopy was used to visualize ECM within the vessel wall. Pressurized arterioles were fixed and stained with Alexa 633 hydrazide (as a nonselective ECM marker), anti-elastin, or anti-type 1 collagen antibody and a fluorescent nuclear stain. Exposure of cremaster muscle arterioles to elastase for 5 minutes caused an irreversible lengthening of the vessel segment that was not observed in cerebral arteries. Longitudinal elastin fibers were demonstrated on cremaster muscle arterioles using 3D imaging but were confirmed to be absent in cerebral vessels. The fibers were also distinct from type I collagen fibers and were degraded by elastase treatment.
Conclusion—These results indicate the importance of elastin in bearing longitudinal stress in the arteriolar wall and that these fibers constrain vascular smooth muscle cells. Differences between skeletal muscle and cerebral small arteries may reflect differences in the local mechanical environment, such as exposure to longitudinal stretch.
The extracellular matrix (ECM) contains a number of proteins, including collagen, elastin, laminin, fibronectin, vitronectin, glycoproteins, and proteoglycans. In addition to providing a mechanically dynamic structural scaffold, the ECM is involved in physiological processes such as cell growth, differentiation, and migration. With respect to the vasculature, recent studies have demonstrated a number of these ECM proteins to actively signal through outside-in means into both vascular smooth muscle and endothelial cells, particularly conveying mechanical signals. For example, fibronectin binding through cell surface integrins modulates the activity of smooth muscle cell (SMC) ion channels (voltage-gated Ca2+ channels and large conductance Ca2+-activated K+ channels)1–3 and affects local cellular contractions.4 Similarly, ECM protein-integrin activation of various intracellular signaling mechanisms underlies endothelial cell mechanotransduction to stimuli such as shear stress.5,6 Despite these demonstrated roles of the ECM in vascular cell signaling, relatively little is known of the complexities of the in situ arrangement between specific ECM proteins and arteriolar SMCs. Given the above examples, it is likely that the structural arrangement of the vessel wall ECM proteins, particularly at the microvascular level, has an impact on how local mechanical forces are transmitted, sensed, and responded to and, ultimately, how effectively a vessel is able to alter its diameter.
Adding to difficulties in understanding the complexity of the extracellular components of the vessel wall is an apparent regional heterogeneity. Importantly, regional differences involve both variation in matrix composition and structural arrangement. Thus, in contrast to skeletal muscle arterioles, small cerebral arteries lack an external elastic lamina that is present in many peripheral vessels.7 Both vessel types exhibit an internal elastic lamina (IEL), although recent studies have suggested that heterogeneity at this level may underlie differences in myoendothelial gap junction–mediated communication and functional responses mediated through hyperpolarization.8,9
In regard to specific ECM proteins, Baumbach and colleagues have provided information on the composition of the arteriolar wall, particularly as it relates to the cerebral circulation under physiological conditions and in experimental models of hypertension.10–12 Those studies have emphasized the significant contribution that elastin makes to the overall composition of the vessel wall. From a functional perspective, elastin is an important determinant of arterial distensibility, and hence stiffness, in conduit vessels. Elastin's contribution is particularly relevant at low intraluminal pressures as compared with collagen, which exerts its effects on distensibility at higher pressures.13 Despite this, the functional role of elastin in small arteries and arterioles is less well defined.14
We hypothesized that an adventitial population of elastin-rich fibers exists in the vessel wall that are predominantly arranged in a longitudinal fashion and thereby bear and exert longitudinal mechanical stress in the arteriolar wall. To examine this hypothesis, we predicted that abluminal exposure of isolated arterioles to elastase would cause vessel lengthening. Furthermore, we predicted that this response to elastase would be absent in vessels lacking an overt external elastic lamina (such as small cerebral arteries).
Materials and Methods
All studies were performed on arterial blood vessels obtained from male Sprague-Dawley rats (175–250 g, 6–9 weeks of age). Surgical and animal handling procedures were approved by the Animal Care and Use Committee of the University of Missouri.
Vessel Isolation, Cannulation, and Study
Cremaster first-order (1A) arterioles, small cerebral arteries, and mesenteric arteries were isolated from rats as previously described and cannulated for isobaric protocols15,16 (see supplemental material, available online at http://atvb.ahajournals.org). Arterioles were exposed for 5 minutes to porcine pancreatic elastase (0.05 U/mL) and then washed for 20 minutes. Vasomotor responses were then reassessed. In some experiments, vessels were superfused with a 0 mmol/L Ca2+ physiological salt solution containing 2 mmol/L EGTA to allow passive distensibility (pressure-diameter) relationships to be determined before and after elastase treatment. All elastase digestions were performed in the presence of a Ca2+-containing physiological salt solution.
Imaging and 3D Reconstruction/Data Analysis
Pressurized arterioles were fixed (2% paraformaldehyde) and stained with Alexa 633 hydrazide (0.2 μmol/L, excitation 633/emission 700 nm; nonselective matrix staining dye), anti-elastin antibody (primary antibody dilution 1:100; secondary antibody, 491/515 nm) and a nuclear stain Yo-Pro-1 iodide (491/515 nm) or 4′,6-diamidino-2-phenylindole (500 ng/mL, excitation: 355, emission: 400–450 nm).17 A subset of vessels were stained for type I collagen using an anti-rat polyclonal antibody (Millipore; 1:1000, 1 μg/mL) and a goat anti-rabbit Alexa Fluor 488 secondary antibody (20 μg/mL).
3D confocal microscopy was used to visualize ECM within the arteriole wall. Imaging was performed using a Leica TCS-SP5 microscope in conjunction with a Leica ×63 water immersion objective lens (numerical aperture 1.2) and Leica LAS AF image acquisition software. Images (512×512 pixels) were acquired at a resolution of 0.48 μm/pixel at 400 Hz. Line averaging (3 scans) was performed to reduce the low frequency noise. z-Dimension step size was 0.05 to 0.3 μm for all z-stacks. All acquisitions represent 8-bit TIFF gray scale images. Post-image-acquisition analyses were performed using ImageJ (National Institutes of Health, Bethesda, MD), Imaris (Bitplane Scientific Software, South Windsor, CT), and Image Pro (Media Cybernetics, Bethesda, MD).
See supplemental material for detailed Materials and Methods, including additional information relating to image acquisition and processing.
Chemicals and Reagents
Porcine elastase was obtained from Calbiochem (48 U/g, catalog #324682, lot #D00012477, EMD Chemicals, San Diego, CA) or from Sigma-Aldrich (5.9 U/mg, catalog #E1250, lot #041M7018V, St Louis, MO). Type II collagenase (553 U/mg) was purchased from Sigma-Aldrich. Anti-elastin antibody was obtained from Chemicon International (Millipore, Billerica, MA) and anti-collagen antibody directly from Millipore. Species appropriate secondary antibodies were obtained from Molecular Probes (Invitrogen, Carlsbad, CA). Alexa hydrazide, Yo-Pro-1 iodide, and 4′,6-diamidino-2-phenylindole were similarly obtained from Molecular Probes.
Results are presented as mean±SEM. Simple comparisons between 2 means were performed using a 2-tailed Student t test, and multiple group comparisons were performed using analysis of variance (ANOVA) with the Tukey post hoc test. Statistical significance was assumed at the P<0.05 level.
Treatment of Cremaster Arterioles With Elastase Causes Irreversible Lengthening
After developing myogenic tone, cannulated and pressurized (70 mm Hg) cremaster 1As were subjected to a 5-minute exposure to elastase (0.05 U/mL) delivered to the adventitial surface. Length of vessels was assessed before enzyme exposure and again following elastase treatment (and 20 minutes of washout) by adjusting the calibrated x translational direction of the micromanipulator holding the cannulation pipette. Lumenal diameter was monitored continually during elastase treatment and subsequent washout. The arterioles responded to the elastase by significantly lengthening by 35.6±2.3% (n=10; P<0.05) (Figure 1), whereas there was little apparent change in diameter (−5.2±0.8%). The lengthening, evident as a lateral bowing of the vessel between the cannulation pipettes, was not reversed by washout of the elastase, consistent with a direct effect on the vessel structure. An example time course for the elastase-induced lengthening of a cremaster muscle arteriole is shown in Figure 1A and 1B.
Studies of reactivity were performed to demonstrate that the vessels retained viability and that cellular contractile function was intact. Vessels held pressure and did not display leaks. Furthermore, vessels retained steady-state myogenic tone, although acute myogenic reactivity was depressed (Figure 2A). No difference in ability to dilate to acetylcholine (10−6 mol/L) or contract to phenylephrine (10−6 mol/L) was observed (Figure 2B).
To assess the effects of elastase treatment on the passive properties of the cannulated cremaster arterioles, pressure-diameter relationships were measured after superfusion with physiological salt solution lacking Ca2+ and containing 2 mmol/L EGTA. Elastase caused a significant leftward shift in the pressure-diameter relationship particularly at low pressures (Figure 2C) consistent with a major effect on the elastin components of the vessel wall as opposed to the collagen fibers.13,14
To quantify the degree of cremaster arteriole lengthening and to determine its effects on elements of the vessel wall, cell width and number of cells per unit length were determined pre- and postelastase treatment. To facilitate these measurements, a dye exclusion technique was used as previously described.18 Cell-impermeable carboxyfluorescein was placed in the vessel bath from which it subsequently diffused into the spaces between vascular smooth muscle cells (VSMCs). Acquired images are then digitally inverted to reveal the SMCs, which allowed cell shape to be readily discerned (Supplemental Figure I). Cell width increased significantly (P<0.05) following elastase treatment, whereas the number of cells per unit length decreased significantly (P<0.05) (Table). Lengthening was also associated with a change in shape of IEL holes, as indicated by a significant (P<0.05) increase in the long axis:short axis ratio (Table).
Because elastase is a serine protease, it was considered conceivable that the lengthening effect may reflect an action on multiple ECM targets. As a comparison, an additional set of cannulated vessels was similarly exposed to type II collagenase (30 U/mL, 5 minutes). In contrast to elastase, vessels treated with collagenase did not lengthen but showed significant vasodilation (34.1±4.0%; n=6; P<0.05; Figure 2D; see Supplemental Figure III for additional information). Also, in contrast to elastase treatment, the dilator effect of collagenase was reversed on washout (Figure 2D and supplemental material).
In Contrast, Small Cerebral Arteries Do Not Lengthen In Response to Elastase
In additional studies, the effects of elastase treatment were compared between cerebral arteries (n=8) and cremaster 1As (n=8). Vessels (taken from the same animals) were cannulated and treated in a similar fashion, and all developed spontaneous myogenic tone. Although cremaster vessels again showed a lengthening response, the cerebral vessels were largely unaffected by elastase treatment (Figure 3). Example images and the time course of elastin exposure are shown in Figure 1.
Despite the absence of lengthening of cerebral vessels following elastase treatment, their passive pressure-diameter relationships showed a leftward and upward shift (Figure 3), consistent with the enzyme treatment exerting distinct effects on the 2 vessel types.
In cerebral vessels, elastase treatment did not cause a measurable change in VSMC width or the number of cells per unit length (Table). Furthermore, cerebral VSMC width and numbers of cells per unit length were not significantly different from those of cremaster VSMCs under control conditions (Table). The IEL hole long:short axis ratio in cerebral arteries was unaffected by the elastase treatment (Table).
Cremaster Arterioles Show Longitudinal Adventitial Fibers That Contain Elastin and Are Degraded by Elastase
Cremaster 1A arterioles, stained with Alexa 633 hydrazide and counterstained for nuclei, showed longitudinally arranged adventitial (superficial) fibers that were branched in a dense and complex fashion forming a network traversing the circumference of the vessel (Figure 4A; Supplemental Movie Files I and II; Supplemental Figure IVa and IVb). From the reconstructed cross-sectional view, it is clear that the cremaster 1A has both an internal and external ECM layer (Figure 4A). The internal layer resembles previous descriptions of the IEL and exhibited holes that have previously been described to be sites where myoendothelial cell projections can be found9,17,19 (see also Supplemental Figure IV). The Alexa 633 hydrazide-stained adventitial ECM structure was absent in small cerebral arteries, consistent with the previous observation that these vessels lack a defined external elastic lamina (Figure 4B and Supplemental Movie Files III and IV).
Under the imaging conditions used, autofluorescence of ECM proteins did not contribute greatly to the images of vessels stained with Alexa 633 hydrazide. Details of control experiments are shown in the supplemental material.
Staining with a specific elastin antibody showed overlap between its pattern and the structures stained by the Alexa hydrazide dye, suggesting that the adventitial fibers observed in cremaster vessels indeed contain elastin (Figure 5A–5C). Under the imaging conditions used, no fluorescence staining was detected when the elastin (primary) antibody was omitted (data not shown; also see supplemental material). Vessels treated with elastase showed only remnants of these fibers and areas of indentation where the fibers had likely been positioned before enzyme treatment (Figure 5D). It should be noted that, consistent with the observations of Briones et al,14 the conditions of the elastase treatment (addition to the adventitial surface, time of exposure, and concentration/activity of enzyme) were such that whereas the adventitial fibers were disrupted, the IEL remained largely intact.
Staining with a specific antibody for a Type 1 collagen showed an entirely different staining pattern from that of either Alexa 633 hydrazide or the elastin antibody (see Supplemental Figure V). In contrast to the distinct adventitial fibers shown by Alexa 633/elastin antibody staining, Type 1 collagen stained in a wavy, belt, or strap-like manner. A similar distribution of collagen was confirmed from autofluorescence images, as used by previous investigators20 (data not shown).
Small Mesenteric Arteries Contain Adventitial Fibers Resembling Those of Cremaster Arterioles
On the basis of the above, we hypothesized that elastin fibers support vessels prone to longitudinal stretch and as such predicted that small mesenteric arteries should contain an elastin fiber network more closely related to cremaster vessels than cerebral arteries. Similar to the cremaster vessels, small mesenteric arteries stained with Alexa 633 hydrazide and counterstained for nuclei showed longitudinally arranged adventitial (superficial) fibers that branched into a very dense and complex network traversing the adventitia (Figure 6 and Supplemental Movie Files V and VI).
Despite consisting of only a few cell layers in thickness, the walls of small arteries and arterioles present a complex biomechanical structure. Furthermore, as a contractile tissue, there is considerable interdependence of structure and cell function. Using confocal microscopy and 3D imaging approaches to reconstruct through-focus structural features of the vascular wall, the present study illustrates the complexity of the ECM while also demonstrating that particular matrix proteins may constrain or shield the environment within which SMCs reside and exert their contractile activity. In this regard, the data support a significant role for elastin-containing fibers in the mechanical properties of true resistance vessels14 and suggest that this property is not limited to larger conduit vessels. These conclusions are highlighted by brief elastase treatment causing degradation of adventitial matrix fibers and significant lengthening of cremaster muscle arterioles. Heterogeneity between vascular beds was, however, also observed. In particular, small cerebral arteries that are known to lack an external elastic lamina7 did not show lengthening to a similar elastase exposure.
In regard to the physiological significance of the apparent difference between adventitial structure of cremaster muscle and mesenteric arterioles versus small cerebral arteries, we speculate that this structure confers the ability to withstand longitudinal stretch. Thus, retraction and relaxation of the cremaster muscle would result in the centrally located arteriole being subjected to longitudinally directed strain, whereas cerebral vessels, located within a rigid skull, are largely protected from such elongation. To support this hypothesis, mesenteric arteries were examined as another vascular bed where vessels are commonly subjected to longitudinal changes in length. Similar to the cremaster 1A, small mesenteric arteries also showed a complex network of adventitial matrix protein fibers. Supporting structural heterogeneity within the vessel wall (in particular as it relates to the ECM) between vascular beds, elastin mRNA expression was significantly greater in cremaster and mesenteric vessels compared with small cerebral arteries (see Supplemental Figure VI). Further supporting heterogeneity, the relative amount of muscle cell protein (measured as α actin content) was greater in cerebral arteries as compared with both cremaster and mesenteric vessels (see Supplemental Figure VI). Conceivably, morphological differences could also exist at the level of the VSMCs; however, under control conditions, calculations of VSMC width and number of cells per unit length showed no differences between cremaster and cerebral small arteries.
Interestingly, although elastase did not cause lengthening of the small cerebral arteries both cremaster and cerebral vessels showed shifts in their passive pressure-diameter relationships on enzyme treatment. We speculate that this may be explained by differences in the longitudinal and circumferential properties of the vessel wall. Thus, the length changes are a function of the longitudinally arranged adventitial elastin fibers, whereas the shifts in the pressure-diameter relationships may relate to effects of elastase on other elements, such as the IEL.
IEL holes have recently received considerable attention as sites of endothelial-SMC connectivity via myoendothelial junctions.8,19 Additional analysis was undertaken to determine whether the elastase treatment affected the shape of IEL holes. To quantify changes, the long axis:short axis ratios of the holes were determined. For measurement of the IEL hole size, the vessel image stacks were processed to extract the image data associated with IEL layer alone. Elastase treatment of cremaster vessels led to a significant (P<0.05) increase in the long:short axis ratio of the IEL holes as the vessel lengthened. This could reflect either stretch of the IEL or a direct effect of the elastase on the IEL. In contrast to the cremaster arterioles, cerebral vessels neither lengthened nor changed the measured characteristics of the IEL holes. Indirectly, this supports the idea that cremaster arterioles are normally constrained in the axial direction by adventitially located elastin fibers. Cleavage of these fibers leads an overt increase in vessel length, an increase in cell width, and a stretching of wall structures such as the IEL.
The present studies did not consider any structural role played by other elements that contribute to the adventitia. Images stained with either 4′,6-diamidino-2-phenylindole or propidium iodide showed nuclei in the region of the adventita. In addition, the orientation of these cells indicated that they were unlikely to be either smooth muscle or endothelial cells. As these cells were apparent in cremaster vessels before and after elastase treatment (data not shown) it is unlikely that these elements contributed to the lengthening process. Nevertheless, future studies should be aimed at identifying the nature of such cells.
Consistent with elastase causing lengthening of the cremaster arteriole, arterial tortuosity has been reported in studies administering elastase in vivo to induce aneurysm formation21–23 and in transgenic animal models where elastin levels are genetically manipulated.24 Such studies have largely focused on conduit vessels because of the role of elastin in elastic recoil of arteries and the clinical significance of large vessel aneurysms. In addition to common findings with respect to vessel length, disruption of elastin appears to decrease the distensibility of both large arteries and, at least, first-order arterioles from skeletal muscle. Also consistent with a role for elastin in regulating small vessel dimensions, Briones et al14 showed elastase treatment to decrease the ex vivo distensibility of small mesenteric arteries. Interestingly, in the present imaging studies, mesenteric vessels exhibited a complex adventitial fibrillar network similar to that observed in the walls of the cremaster muscle arterioles.
Additional studies will be required to delineate the exact involvement of specific matrix proteins. We initially used Alexa 633 hydrazide staining as a nonselective marker for ECM proteins. However, it is evident from the specific elastin staining that this dye preferentially binds elastin-containing structures. Further supporting its restricted staining was the observation that a different staining pattern is observed with an antibody known to be specific for Type 1 collagen. Similarly, cleavage of the fibers with elastase led to a leftward shift in the passive pressure-diameter relationship (for cremaster arterioles) at low pressures, consistent with the contribution of elastin to vascular mechanics as compared with that of collagen.13,14 This, however, does not indicate that the stained adventitial fibers are composed of elastin alone and they presumably contain other proteins, such as fibrillin.24–26
A question that arose from the present studies is whether or not the staining pattern of elastin-containing fibers found in the cremaster arterioles is representative of small arteries and arterioles in other skeletal muscles. Relevant to this, we have observed a complex fiber network in arterioles isolated from hamster gracilis muscle, suggesting that a similar structure exists in differing skeletal muscles and across species (data not shown).
Important questions arising from our studies include how VSMCs interact with the matrix scaffold and whether the differences in regional matrix structure imply differences in how local mechanical forces are transmitted, sensed, and responded to. Clearly, the lengthening of cremaster muscle 1As and increasing SMC width following elastase exposure suggest that under resting conditions, SMCs are constrained along the long axis of the vessel or across the short axis of individual SMCs. Importantly, the enzyme treatment did not markedly alter vascular reactivity to vasoactive agents (acetylcholine and phenylephrine), and the vessels remained free of leaks to the intraluminal pressure. Furthermore, although the lengthening response was seen with elastase treatment, this was not seen following collagenase exposure. As such, we do not believe that our observations can be simply explained by nonselective destruction of vessel integrity per se.
In an earlier study, Spofford and Chilian27 reported differing responses of cerebral and mesenteric small arteries with respect to pressure-induced changes in c-fos expression. Increasing intraluminal pressure led to a decrease in cerebral vessel c-fos expression while causing an increase in mesenteric vessels. Interestingly the pressure-induced effect on c-fos expression was mediated through the elastin-laminin receptor and could be blocked by decoy peptides that are presumed to inhibit normal matrix protein binding to elastin. These authors further found differences in the matrix protein composition between the cerebral and mesenteric vessels suggesting that the differences in pressure responsiveness may reflect regional variation in the matrix environment. Specifically, consistent with the present study, these authors reported an absence of elastin in the adventitia of the cerebral vessels and a presence of elastin in the adventitia of mesenteric vessels. Cerebral vessels, however, were reported to show comparatively greater levels of elastin within the medial layer.
In addition to understanding the physiological structure of the vessel wall, the relevance to pathophysiology should also be noted. Although currently largely limited to studies of conduit vessels and subcutaneous resistance arteries, age-related changes in the ratio of elastin to collagen affect vascular stiffness.28–31 Furthermore, in aging and metabolic disturbances (for example, in diabetes mellitus), long-lived proteins such as elastin and collagen are subject to modification by glycation and oxidation.32,33 Presumably, such changes may alter not only the mechanical properties of the ECM proteins but also the relationship between the ECM and cellular elements of the vessel wall, thus potentially affecting matrix protein-mediated cellular signaling.
Collectively, the data support our hypothesis that elastin is an important ECM protein for bearing longitudinal stress in the arteriolar wall. Furthermore, when present, these fibers act to constrain or restrict the longitudinal dimensions of the vessel. Differences between arterioles from skeletal muscle and mesentery versus small arteries from the cerebral circulation may be related to differences in the local mechanical environments in which these vessels reside. Skeletal muscle and mesenteric tissues are continuously exposed to longitudinally oriented stretching forces that are transferred to the vasculature. By comparison, cerebral tissue is encapsulated and the brain vasculature is not normally exposed to these types of stretching forces. Knowledge of the 3D architecture of the ECM components of the vessel wall is also of importance to our understanding of how these elements are affected in aging and disease states, particularly where matrix proteins are damaged, degraded,34 or posttranslationally modified.32,33
Sources of Funding
This study was supported by National Institutes of Health Grants HL92241-02 (to M.A.H.), P01-HL-095486 (to G.A.M. and M.J.D.), and HL088105-02 (to L.M.L.). Dr Clifford was supported by a Research Career Enhancement Award from the American Physiological Society.
- Received April 19, 2011.
- Accepted September 20, 2011.
- © 2011 American Heart Association, Inc.
- Yang Y,
- Wu X,
- Gui P,
- Wu J,
- Sheng JZ,
- Ling S,
- Braun AP,
- Davis GE,
- Davis MJ
- Sun Z,
- Martinez-Lemus LA,
- Hill MA,
- Meininger GA
- Shyy JY,
- Chien S
- Hajdu MA,
- Heistad DD,
- Siems JE,
- Baumbach GL
- Dobrin PB
- Hill MA,
- Zou H,
- Davis MJ,
- Potocnik SJ,
- Price S
- Martinez-Lemus LA,
- Hill MA,
- Bolz SS,
- Pohl U,
- Meininger GA
- Carta L,
- Wagenseil JE,
- Knutsen RH,
- Mariko B,
- Faury G,
- Davis EC,
- Starcher B,
- Mecham RP,
- Ramirez F
- Gaballa MA,
- Jacob CT,
- Raya TE,
- Liu J,
- Simon B,
- Goldman S
- Intengan HD,
- Deng LY,
- Li JS,
- Schiffrin EL
- Intengan HD,
- Thibault G,
- Li JS,
- Schiffrin EL
- Akhtar K,
- Broekelmann TJ,
- Miao M,
- Keeley FW,
- Starcher BC,
- Pierce RA,
- Mecham RP,
- Adair-Kirk TL