Jump to
Abstract
Abstract Cellular aspects of remodeling in intact arteries have not been fully investigated, mainly due to the lack of an appropriate methodology that allows for simple measurements. The aim of this study was to develop a method based on laser scanning confocal microscopy (LSCM), compare it with previous methodology, and apply it to the study of remodeling in hypertension. The morphology of mesenteric resistance arteries from stroke-prone spontaneously hypertensive rats (SHRSP) and Wistar-Kyoto rats (WKY) was determined with wire myography on one segment with a standardized diameter setting (0.9d100) and with perfusion myography on a second segment from the same artery at the calculated equivalent pressure. The second segments were stained with the nuclear dye Hoechst 33342 (live tissue) or propidium iodide (fixed tissue) and measured with LSCM and MetaMorph software. Compared with wire myography, perfusion myography showed similar differences from those previously reported. Compared with LSCM, perfusion myography showed a similar lumen but significantly smaller wall thickness in both live and fixed tissue, probably due to measurement underestimation. In the study with LSCM, arteries from SHRSP compared with those from WKY showed (1) reduced lumen, (2) altered cell density that was significantly increased in the adventitia, decreased in the media, and unchanged in the intima, (3) significantly increased medial volume, (4) significantly smaller endothelial cell nuclei, and (5) adventitial-like cells in the media. We conclude that (1) LSCM is a reliable and straightforward method to study morphology in intact vessels, (2) it provides new information on the cellular changes in remodeling, (3) adventitia might play an active role in the process of remodeling in hypertension, and (4) endothelium “remodels” in hypertension.
The vascular wall is an active organ composed of endothelial cells, SMCs, and adventitial cells, which are coupled to each other. The wall is able to change its structure through a process known as remodeling. Remodeling is usually an adaptive process in response to long-term changes in hemodynamic conditions, as occurs during development,1 but it can also contribute to the pathophysiology of vascular diseases and circulatory disorders.2 Over the past decade, there has been extensive research concerning the structure of vessels and the process of remodeling in both physiological and pathological conditions and with special interest at the level of the resistance vasculature.1–5
Two methods are currently used to study gross morphology and the function of live small precapillary arteries with diameters of <500 μm: wire and perfusion myography. In the wire myograph, the vessel is mounted on two wires passed through the lumen, and measurements of wall thickness and lumen are taken with an eyepiece attached to a light microscope.6,7 In the perfusion system,8,9 similar measurements of vessel wall and internal diameter are obtained with a VDA from a pressurized artery.
Despite this extensive research and the increasing interest in the remodeling process in pathological situations, little is known at the cellular level about the morphology of intact arteries and the structural changes associated with them. Cellular changes in remodeling have been mainly studied in SMCs in culture. Cell culture studies, although valuable, lack critical interactions with other vascular cell types and natural matrices found in the vessel wall, so the results cannot be extrapolated to the changes in intact vessels. On the other hand, histological information on vessel cellular structure also has several limitations: the tissue must be sectioned and then reconstructed, so it is very labor intensive and has the disadvantage of being unable to provide a study of live tissue.
Confocal microscopy produces “optical sections” through semitransparent tissue without the need for cutting thin slices. It eliminates the blur and flare of out-of-focus planes in an object, has improved axial resolution that enables accurate three-dimensional reconstruction, and has rapid speed of image acquisition.10–12 With LSCM, we recently developed methods that allow the study of the structure-function relationship in resistance arteries, as close as possible to the cellular level.13–15
The aim of the present work was to use LSCM to study in intact resistance arteries the cellular changes associated with remodeling in hypertension. We first assessed the ability of the method to obtain simple dimensional measurements compared with the two methods presently available: wire and perfusion myography. We then examined the cellular arrangement of the coats within the vascular wall by LSCM and found several features, not described before, in the three layers of mesenteric resistance arteries from SHRSP, including important changes in the adventitia and endothelium, layers usually neglected in structural studies in hypertension.
Methods
Animals
Rats were obtained from the Glasgow inbred colonies that have been maintained in the Department of Medicine and Therapeutics, University of Glasgow, since 1991.16 Thirteen WKY and eight SHRSP male or female 8- to 10-month-old rats were used. Systolic blood pressure was measured by the tail-cuff method (custom made; Department of Clinical Physics and Bio-Engineering, Western Infirmary, University of Glasgow). The rats were killed through inhalation of 3% to 4% halothane in oxygen for each experiment, according to institutional guidelines. The mesenteric bed was carefully removed and placed on a petri dish with PSS at 4°C for dissection of the third order branch of the superior mesenteric artery. Each artery was divided in two segments—one to be used on the wire and the other to be used on the perfusion myograph and subsequently on LSCM.
Drugs and Solutions
All experiments were performed with PSS of the following composition (in mmol/L): NaCl 118.4, KCl 4.7, CaCl2 2.5, KH2PO4 1.2, MgSO4 1.2, NaHCO3 25, and glucose 11.1. The solution was bubbled with 95% O2/5% CO2 to pH 7.4 at 37°C. Propidium iodide and Hoechst 33342 trihydrochloride were purchased from Sigma Chemical Co, and halothane was purchased from ICI Pharmaceuticals.
Wire Myography
A 2-mm-long segment was mounted on a wire myograph (Myo-interface, model 410A; Aarhus). Briefly, two 40-μm tungsten wires were threaded through the lumen of the segment. One wire was attached to the stationary support driven by a micrometer, and the other was attached to an isometric force transducer and this in turn to a chart recorder (Linseis L6512B; Belmont Instruments) to measure force development. After a 30-minute equilibration period at 37°C, the morphological measurements were determined on the vessels held under just enough tension to hold the walls separated. In each segment, the following dimensions were measured with a light microscope (40× water-immersion objective, with a filar eyepiece, Zeiss; Welwyn): wall thickness (average of six measurements), distance between the inner edges of the wires (average of three measurements), and length (average of two measurements). The resting tension–internal circumference relation was determined for each segment according to Mulvany and Halpern.7 From this relation, we calculated the theoretical internal circumference of the artery if relaxed and under a transmural pressure of 100 mm Hg (L100). The corresponding internal diameter (d100) was calculated as L100/π. Each artery was then set to 0.9d100 calculated with a modified computer program. At this internal diameter wall thickness, wall-to-lumen ratio and equivalent pressure were calculated. CSA, equivalent to wall volume per unit length, was also calculated with the computer, assuming constant wall volume.17
Perfusion Myography
Mesenteric resistance arteries were prepared for morphological measurements with the use of pressure-perfusion myography (Living Systems).8 In brief, the arterial segment was secured between two glass cannulae using single strands of surgical nylon suture. One cannula was closed with a syringe and the other was connected to a system containing PSS, which in turn was linked to a pressure-servo unit. This system allowed the pressure (or flow) in the system to be precisely controlled and altered at will. The myograph was a 10-mL vessel chamber with an input-and-output channel to allow perfusion of PSS. One of the cannulae was attached to a micrometer, permitting adjustment to allow changes in longitudinal dimension of the mounted vessel without stretch or buckling. The artery was imaged using a video camera and analyzed with a VDA (Living Systems), which was linked to a chart recorder (Linseis L6512B; Belmont Instruments). Each segment was then pressurized at the equivalent pressure calculated from wire myograph experiments, as described before. After a 30-minute equilibration period at 37°C in oxygenated PSS, wall thickness and internal lumen diameter were measured at six equidistant points along the vessel segment with a VDA. At the end of the experiment, the arteries were fixed under pressure with 10% formal-saline solution for 30 minutes at room temperature and stored for further LSCM analysis. Fig 1A⇓ shows a photograph of the arterial wall from a perfusion myograph–mounted segment measured with VDA.
Schematic of a rat mesenteric artery showing the nuclei of the cells located in adventitia, media, and intima and the plane of focus for wall and lumen measurements in X-Y plane. A through C, Images of wall and lumen of a WKY rat mesenteric artery pressurized at 50 mm Hg on a perfusion myograph. Image dimensions are shown in each panel. A, Image taken from a perfusion myograph-mounted segment with a video camera attached to a light microscope (10× water-immersion objective). White line shows the measurement of wall taken with a VDA, and lines with arrow show the location of adventitia (Adv) and endothelium (End). B, Image taken with an LSCM on a slide (40× oil-immersion objective); inverted contrast. The vessel was stained with the nuclear dye propidium iodide. The location of the media (Med) is clearly defined by SMC nuclei in profile and delimited by the adventitia (Adv) and endothelium (End). C, Image of the lumen taken with an LSCM on slide (10× water-immersion objective). The vessel was stained with the nuclear dye propidium iodide.
Comparison of Live and Fixed Tissue
Five arteries from WKY were used to compare perfusion myography and LSCM measurements under live and fixed conditions. The arteries were first set at 50 mm Hg (mean equivalent pressure calculated from wire myograph previous experiments at 0.9d100). After a 30-minute equilibration period in PSS at 37°C, measurements were taken with the perfusion myograph VDA. The PSS was then substituted for one containing 0.01 mg/mL concentration of the vital dye Hoechst 3334218 for 30 minutes. To enhance the stain of the endothelium, the PSS with the dye was also perfused intraluminally for 2 minutes at a constant flow rate of 0.1 mL/min. The chamber was then placed on the stage of the LSCM, and images of the wall and lumen were acquired as described below. Finally, the vessel was fixed under pressure as previously described, and it was measured again with VDA and LSCM, after several washouts with PSS.
LSCM
The fixed segments were incubated for 30 minutes in 10% formal-saline solution containing 10 μmol/L of the nuclear dye propidium iodide19 and washed overnight. The vessel was placed on a slide and covered with a coverslip attached to the slide by a thick layer of vacuum grease. This formed a small chamber for the artery, avoiding contact of the coverslip with the segment and therefore minimizing possible changes in wall shape. The vessel was then placed under an upright LSCM (Odyssey LSCM; Noran Instruments) with a Nikon microscope. The lumen was visualized with a 10× air objective (Nikon, NA 0.5), and the wall was visualized with a 40× oil immersion objective (Nikon, NA 1.3), using the argon ion 488-nm line with a 515-nm LP barrier filter of the LSCM (pinhole aperture, 15 μm).
Live tissue was stained with Hoechst 33342 as previously described, and the artery was visualized while mounted and pressurized on the myograph. The lumen was visualized with a 10× water objective (Nikon, NA 0.3), and the wall was visualized with a 40× water objective (Zeiss, Welwyn, NA 0.75), using the 364-nm line (no barrier filter) of the LSCM (pinhole aperture, 15 μm).
Lumen images were taken in the X-Y plane by placing the focus on the middle of the vessel (Fig 1⇑, schematic diagram). This allowed visualization of the lumen and the cell nuclei in profile (Fig 1C⇑). Six images of different regions were captured along the artery, and they were stored and coded for further analysis. This protocol was identical for fixed or live tissue.
Wall thickness measurements were taken in the X-Y plane and/or the Z axis. For X-Y measurements, the plane of focus was placed on the middle of the artery, and 12 images of the wall were captured along the segment. This protocol was identical for fixed or live tissue. The wall images clearly showed adventitial, medial, and intimal layers (Fig 1B⇑). However, the limits of the intima and adventitia were not always clear, and only the media was measured. A minimum of two measurements from each image of both lumen and media were taken with MetaMorph software (Universal Imaging Corporation). To determine the degree of consistency of X-Y measurements, they were obtained by three different observers and compared statistically with ANOVA test for interobserver reliability.
The different layers of small vessels stained with nuclear dyes can be clearly distinguished with confocal microscopy according to the shape and orientation of the cell nuclei.14,15 Nuclear dyes also provide the advantage of showing the cells as individual single objects, allowing for direct simple measurements with image analysis software, including the number of cells in each layer and the endothelial cell nuclei area and shape factor (a value from 0 to 1 that represents how closely the object resembles a circle).
Such a quantitative analysis can be done on the assumption that the dyes stain all nuclei of vascular cells. To prove this assumption, the pattern of nuclei stained with Hoechst 33342 was compared with that obtained with hematoxylin and eosin, a conventional histological protocol used for morphometric studies. For this purpose, an artery was stained first with Hoechst 33342 as described before, fixed under pressure, embedded in liver, and sectioned on a cryostat set at 10 μm. A section was then mounted on a slide and visualized with LSCM (40× water objective, 364 nm, no barrier filter). Ten optical slices (1 μm thick) were captured in Z axis from several regions of the section, and a three-dimensional image from each region was then reconstructed with MetaMorph software. The coverslip was subsequently removed from the preparation, and the section was then stained with hematoxylin and eosin and visualized with a Zeiss Axiophot microscope with a 20× air objective. Images from several regions of the vessel section were then taken with NIH Image 1.6 software. A comparison of Hoechst 33342– and hematoxylin and eosin–stained images at the same level of magnification showed a similar pattern of nuclei. In addition, LSCM images had a better resolution and allowed individual nuclei to be more clearly defined (Fig 2⇓).
Comparison of stain efficiency between Hoechst 33342 and hematoxylin and eosin (H-E) in a 10-μm-thick section of an SHRSP mesenteric resistance artery. The arteries were first stained with Hoechst 33342, fixed under pressure, embedded in liver tissue, and sectioned using a cryostat at 10 μm. A section was then visualized with an LSCM (364 nm, open filter), and 10 serial optical slices (1 μm thick) were acquired with MetaMorph software. The section was then subsequently stained with H-E and visualized with bright field microscopy. Left, Three-dimensional reconstruction of a region obtained from 10 serial optical slices with MetaMorph software (LSCM, Hoechst 33342). Right, An image from the same region taken with NIH Image 1.6 (bright field, H-E). Both dyes showed the same pattern of nuclei. LSCM Hoescht-stained images had a better resolution and individual nuclei could be more precisely identified (see arrowheads).
The different wall layers (adventitia, media, and intima) were measured in Z axis. Stacks of 1-μm-thick serial optical slices were taken in Z axis from the first adventitial to the last endothelial nuclei. Fig 3A⇓ shows a schematic diagram that illustrates the process of stack capturing in Z axis. Three of these stacks were captured in different regions along the vessel length and stored for further analysis with MetaMorph software. This protocol was identical for fixed or live tissue. Under conditions used in the present study, Z-axis resolution, given by the LSCM microscope manufacturer, was 0.5 μm for 40× oil-immersion objective and 1.2 μm for 40× water-immersion objective.
A, Schematic of a rat mesenteric artery showing the nuclei of the cells located in adventitia, media, and intima and the principle of LSCM measurements in Z axis. B, Three-dimensional reconstructions of the adventitia, media, and intima layers of pressurized mesenteric resistance arteries from a WKY and an SHRSP. Reconstructions were obtained from 1-μm-thick optical sections taken in Z axis with an LSCM and analyzed with MetaMorph software. Top left, Image dimensions. Irregular adventitial-like cell nuclei can be observed within the media of the SHRSP rat mesenteric artery together with elongated SMC nuclei (middle right).
To have reliable and reproducible data, we defined the location of the first images from each layer based on computer measurement of fluorescence intensity. Thus, the first image of the adventitia in the stack was considered to be that which showed the first adventitial nucleus at its maximum intensity. In a similar way, the first images of the media and intima were defined. The last plane of the intima was considered to be that which showed the last endothelial nucleus at its maximum intensity. From these stacks of images, thickness and nuclei number were measured in each coat by image analysis.
In WKY arteries, there was an homogeneous distribution of endothelial cells. However, in SHRSP arteries, the number of endothelial nuclei was variable along the vessel length. Therefore, to have an accurate average number of the endothelial cells in the vessel, 10 images of the intima were also taken from each artery. This approximately covered the entire vessel length, excluding the extremes, which could have been damaged by the cannulae.
Wall CSA was calculated on the basis of the wall and lumen measurements obtained with VDA or LSCM, using the following formula: CSA=external CSA−internal CSA, where external CSA=π(lumen/2+wall)2 and internal CSA=π(lumen/2)2.
To allow comparison of control and remodeled vessels, the following calculations were performed on the basis of 1-mm-long segments: vessel volume (in mm3); volume=wall CSA (mm2)×1 mm; adventitial, medial, and intimal volumes (volume=respective layer CSA (mm2)×1 mm); total number of cells (adventitial, smooth muscle, or endothelial) (cell n=n of nuclei per stack×n of stacks per vessel volume); luminal surface calculated from vessel diameter (in mm2); luminal surface=2πdiameter/2; total number of endothelial cells was calculated per luminal surface of 1-mm-long vessel (endothelial cell n=endothelial cells/mm2×luminal surface).
Statistical and Data Analyses
Results are expressed as mean±SEM, and n denotes the number of animals used in each experiment. Statistical comparisons were made using Student’s t test for paired or unpaired experiments and one-way ANOVA for interobserver reliability experiments. A value of P<.05 was considered significant.
Results
Validation of LSCM for Morphological Analysis
Comparison Between Wire and Perfusion Myograph Measurements
In WKY, comparison between wire-mounted segments set at 0.9d100 and arteries pressurized at their equivalent pressure showed that lumen was significantly larger and wall thickness and wall-to-lumen ratio were significantly smaller in pressurized vessels (Fig 4⇓). CSAs were similar with both methods (wire myograph, 27 375±1160 μm2; perfusion myograph, 25 422±2490 μm2; n=7).
Comparison of WKY and SHRSP rat mesenteric resistance artery morphology measured with wire or perfusion myograph. Wire myograph-mounted segments were set at 0.9d100 and perfusion myograph-mounted segments were pressurized at a pressure equivalent to 0.9d100. Results are expressed as mean±SEM, with the number of animals given. ∗P<.05, compared with wire myograph; ++P<.01, compared with WKY.
In SHRSP arteries, there was no significant difference between wire or perfusion myography measurements of lumen, wall thickness, wall-to-lumen ratio (Fig 4⇑), or CSA (wire myograph, 42 525±5068 μm2; perfusion myograph, 40 277±4318 μm2).
Comparison Between VDA and LSCM Measurements in Fixed Arteries
Lumen dimensions were similar when taken with VDA or LSCM in both WKY (Fig 5A⇓) and SHRSP (data not shown).
Effect of fixation on WKY rat mesenteric resistance artery morphology. All segments were pressurized at 50 mm Hg and measured with a VDA and an LSCM before and after fixation. For LSCM measurements, the segments were stained with the vital nuclear dye Hoechst 33342. A, Comparison of lumen measurements with both methods before and after fixation. B, Comparison of wall measurements with both methods before and after fixation. C, Comparison of adventitia, media, intima, and wall thickness measurements taken with LSCM before and after fixation. Stacks of images of 1-μm-thick optical slices were acquired from the adventitia to the lumen with a 40× oil-immersion objective, and images of the lumen were taken with a 10× air objective. Thickness was measured in Z-axis and lumen dimensions in X-Y plane with MetaMorph software. Results are expressed as mean±SEM ++P<.01, +P<.05, compared with live tissue (paired Student’s t test); ∗P<.01, compared with VDA (unpaired Student’s t test), with the number of animals given.
In WKY, wall thickness was significantly greater when measured with LSCM compared with VDA (Fig 5B⇑). CSA was also significantly larger when measured with LSCM compared with VDA (CSA LSCM=38 104±2577 μm2, CSA VDA=25 422±2490 μm2; n=5, P<.05, paired experiments). LSCM measurements of the medial layer were similar whether measured in the X-Y (19.8±1.1 μm) or Z (21.3±1.6 μm, n=5, paired experiments) direction. In addition, these LSCM measurements of media were similar to the “wall” measurements estimated with perfusion myography VDA (23.0±0.25 μm, n=5, paired experiments). Similar results were found in SHRSP rats (data not shown).
Effect of Fixation on WKY Artery Structure
The effect of fixation was tested only in arteries from WKY. VDA and LSCM measurements were compared.
Fixation produced a small but significant reduction of lumen, which was similar for the two methods (VDA, 12%; LSCM, 13%) (Fig 5A⇑).
Fixation had no significant effect on wall thickness or CSA when measured by either method (Fig 5B⇑). With LSCM, it was possible to distinguish between cell types within the vessel wall by the shape of the nuclei and therefore to measure the thickness of the different layers (adventitia, media, and intima) in the Z axis. Fixation produced no significant difference in the thickness of any of the layers (Fig 5C⇑).
Accuracy of LSCM Measurements
Measurements in Z axis were calculated with the computer program on the basis of fluorescence intensities of stained objects in the image. To determine the reliability of measurements in X-Y plane (more susceptible to subjectivity), the samples were blindly taken and the images were measured by three different observers. Statistical analysis with one-way ANOVA showed no significant difference between observers in measurements of lumen (observer A, 369.46 pixels; observer B, 360.8 pixels; and observer C, 354.9 pixels; P=.93) or media (observer A, 166.7 pixels; observer B, 173.7 pixels; and observer C, 198.2 pixels; P=.33).
Comparison of WKY and SHRSP Mesenteric Artery Morphology
Systolic blood pressure of SHRSP was significantly higher compared with that of WKY (SHRSP, 164±9.8 mm Hg; WKY, 122±4.3 mm Hg; P<.001).
All three methods showed a significantly smaller lumen and larger wall thickness and wall-to-lumen ratio in SHRSP compared with WKY (Fig 4⇑ and Table 1⇓). With LSCM, it was also possible to distinguish between cell types within the vessel wall on the basis of the shape of the nuclei and therefore to determine the structure of the different coats (adventitia, media, and intima) in Z axis. The changes observed in each layer are described below.
LSCM Measurements of Mesenteric Resistance Artery Morphology in WKY and SHRSP
Changes in the Morphology of the Adventitia
SHRSP mesenteric arteries showed similar adventitial thickness (Fig 6A⇓, Table 1⇑) and volume (Fig 6B⇓) compared with WKY arteries. The total number of cells (Figs 3B⇑ and 6C⇓) and density of cells in the adventitia (Fig 6D⇓) were significantly higher in SHRSP arteries.
Comparison of morphology in intact mesenteric resistance arteries from WKY and SHRSP rats using LSCM. The arteries were pressurized, fixed under pressure, and stained with the nuclear dye propidium iodide. Stacks of images of 1-μm-thick optical slices were taken from the adventitia to the lumen with a 40× oil-immersion objective. MetaMorph software was used to measure, in Z axis, thickness of each layer and for cell counting. A, Layer thickness. B, Layer volume. C, Total number of cells. D, Density of cells per layer. The parameters were calculated on the basis of a 1-mm-long artery. Results are expressed as mean±SEM, with the number of animals given. ∗∗P<.01, ∗P<.05, compared with WKY.
In WKY arteries, there was a clear separation between adventitial and medial coats; adventitial cells were confined to its layer, and the maximum depth within the media at which adventitial cells could be found was 1.45±0.6 μm. However, in SHRSP, cells with an irregular nuclei shape similar to those of adventitial cells were found in deep layers of the medial coat; the maximum penetration within the media of these adventitia-like cells was 14.4±3.9 μm (P<.01 compared with WKY). This abnormality can be observed in the three-dimensional reconstruction of the media in Fig 3B⇑ (middle); in SHRSP, elongated SMC nuclei coexist with irregular adventitia-like nuclei (top), whereas in WKY arteries, only SMC nuclei were found.
Changes in Morphology of the Media
In SHRSP arteries, medial thickness (Fig 6A⇑, Table 1⇑) and volume (Fig 6B⇑) were significantly greater compared with WKY vessels. There was no difference in the total SMC number between WKY and SHRSP (Figs 3B⇑ and 6C⇑). The density of SMC in the media was significantly reduced in SHRSP (Fig 6D⇑).
Changes in Morphology of the Intima
Intima thickness and volume were similar in both rat strains (Fig 6A⇑ and 6B⇑), but the luminal surface was reduced in SHRSP (Table 2⇓). In SHRSP mesenteric arteries, there was a significant reduction in the total number of endothelial cells (Fig 6C⇑); however, the density of endothelial cells per mm2 (Table 2⇓) remained unchanged (ie, the cell number was reduced in proportion of the decrease in luminal surface). In SHRSP segments, the cells were not homogeneously distributed along the artery (data not shown). In SHRSP, there was a significant reduction in nuclei area, and the shape of the endothelial nuclei was variable; however, the mean of the shape factor was not statistically different between strains (Table 2⇓, Fig 3B⇑).
Endothelial Layer Observed With LSCM in WKY and SHRSP Mesenteric Resistance Arteries
Discussion
The aim of this study was to make a direct quantitative analysis of the cellular basis of vascular remodeling in hypertension using a method based on LSCM in intact vessels and compare it with previous methodology.
LSCM provides a rapid and accurate new method for determining the morphology of whole resistance arteries in live and fixed conditions. It correlates well with the more limited data available from wire and perfusion myography.
LSCM has several advantages over previous methodology for morphological measurements. First, it provides a higher resolution of image. This allows for straightforward measurements at the cellular level with image analysis software, including automated counting of different types of cells as well as determination of several cell parameters such as nuclear shape, orientation, and area. Second, it can be performed in intact arteries without the need for histological section. Third, it gives objective and bias-free data on basic dimensions. Therefore, LSCM represents an important advantage over labor-intensive histological procedures; even in improved methods such as the three-dimensional dissector, the tissue has to be sectioned and photographed, and pictures must be assessed by eye.20
In addition, LSCM showed differences that had not been found previously in the cellular characteristics of adventitial, medial, and intimal coats of mesenteric resistance arteries from SHRSP compared with WKY. This finding provides new and potentially important information on the remodeling process in hypertension and other vascular diseases that is relevant to understanding the pathogenesis and developing new therapeutic approaches.
Remodeling of SHRSP Mesenteric Resistance Arteries
It is well established that hypertension is associated with structural changes in the resistance vasculature, such as reduced lumen and increased wall-to-lumen ratio. These changes can be associated with net growth of the vessel or rearrangement of wall material (remodeling).3 This previous concept of remodeling has been redefined, and remodeling is now considered to be a more complex process that may include growth as well as rearrangement and either reduction or increase in vessel lumen.21 Our results suggest that mesenteric resistance arteries from SHRSP present hypertrophic inward remodeling, resulting from a reduced vessel lumen and an increased wall volume. We use “hypertrophic” here to refer to tissue volume, not to cell size.
There is controversy about the nature of structural vascular changes associated with genetic and experimental hypertension in rats. Some reports show that the media of small mesenteric arteries from spontaneously hypertensive rats is thicker due to an increased number of SMCs.22 On the contrary, in New Zealand genetically hypertensive23 and renal hypertensive24 rats, the changes in wall growth are not due to SMC hyperplasia. Our results with LSCM, in which the integrity of wall structure is preserved, also point in this latter direction. In SHRSP mesenteric arteries, the volume of the media was greater by 32% compared with WKY, without changes in the number of SMCs. Within a given length of artery, there is an increase in the average volume occupied by each SMC and its associated extracellular space/matrix. Because in the present study we measured the number of nuclei (hence number of cells) per unit volume, we cannot distinguish whether there is an increase in cell volume or in extracellular space/matrix, both of which have been proposed before.23 Rearrangement of SMCs was found in the cerebral vasculature of SHRSP,25 so rearrangement of a constant number and size of SMCs within a larger coat volume remains an attractive possibility.
All previous studies on resistance artery vascular morphology have been focused on the changes in the media, probably due to the difficulty in quantifying adventitia and intima. With LSCM, which allows the study of these layers, we have been able to demonstrate almost doubling of adventitial cell number in the mesenteric vasculature of SHRSP rats compared with WKY. Similar alteration has been previously reported in coronary arteries after balloon injury, suggesting that adventitia can play an important role in the process of repair.5 Adventitial thickness remained unchanged, and therefore this layer did not contribute to the overall increase in wall volume; however, the increase in adventitial cell number suggests that this layer might contribute to the changes in wall architecture and thus play an active role in the remodeling process. The finding of irregularly shaped nuclei in the medial layer is also interesting. One possible explanation is that these nuclei correspond to adventitial cells; if that were the case, this would imply a migration of these cells from the outermost part of the vessel into the media and further supports the hypotheses of an active role of the adventitia in the remodeling process. An second explanation is that these nuclei correspond to disoriented SMCs, similar to the alteration found in the SHRSP cerebral vasculature, in which there was a rearrangement of SMCs in X-Y plane.25 However, the spherical shape of the nuclei found in the medial coat in the present study supports the hypotheses of a change in orientation in Z axis rather than in X-Y plane. A third alternative would be that those irregular nuclei correspond to a different phenotype of SMCs.
Endothelium is potentially an important element in the genesis and development of remodeling.2 It is an important determinant in vivo of the lumen dimensions and overall vessel structure; structural changes in vessel diameter due to changes in blood flow are dependent on an intact endothelium.26 In addition, the endothelium is in contact with circulating substances, including mitogens, and therefore may be the first to undergo structural changes in abnormal situations. In SHRSP, the total number of endothelial cells was reduced by half compared with WKY arteries. However, when the number of endothelial cells was expressed per mm2, there was no difference between strains due to the reduction in luminal surface in SHRSP arteries. Thus, this change in the endothelium could reflect a primary change as part of the remodeling process associated with hypertension, but it could also indicate a secondary adaptation of the endothelium to the narrower lumen, such as to maintain the density of endothelial cells.
Another notable feature was the heterogeneous distribution of endothelial cells. Some regions of the luminal surface from SHRSP arteries were almost devoid of endothelial nuclei, whereas others had normal numbers of cells. In SHRSP arteries, the surface areas of the individual nuclei were also significantly smaller, and the nuclear shape was variable. It is possible that these abnormalities could be associated with endothelial dysfunction such as the impairment of endothelium-dependent relaxations, which has been described in hypertension.27,28
Differences Between Wire and Perfusion Myography
In a few studies, wire and perfusion myography have compared.29–31 Like these previous reports, our study shows a larger lumen, smaller wall thickness, and smaller wall-to-lumen ratio for perfusion compared with wire myography. The gold standard usually used in wire myography, defined by Mulvany and Halpern,7 and on which we have based the conditions for our study, is the theoretical internal circumference that the artery should have if relaxed and under a transmural pressure of 100 mm Hg: the actual measurements on which this calculation is based are taken with the wires just under tension. In contrast, in the pressure myograph, the vessel morphology is estimated directly with a VDA. It has been argued that part of the difference may thus lie in the theoretical assumptions necessary for the wire myography, exacerbated perhaps by other physical differences such as the axial distension with increasing pressure that occurs in the perfusion myograph, but not experienced on a wire myograph.30
Our objective was to validate our basic myography technique against the known standards. This was accomplished in relation to both the differences between the two established methods and the gross “remodeling” in the hypertensive model shown by changes in lumen, wall thickness, and wall-to-lumen ratio, which was demonstrated by either method. We then chose the perfusion myograph, a more physiological method, to apply LSCM because the measurements are directly comparable with the VDA system: the comparison is based on real measurements of the same physical situation using different measuring instruments and allows the maintenance of the vessel shape at different distending pressures.31
Differences Between LSCM and Perfusion Myography
Using tissues from WKY, we found significantly larger wall thickness and CSA measurements with LSCM, under both live and fixed conditions. This was not due to a general calibration problem of the LSCM because lumen measurements were similar when taken with both methods, provided the conditions were the same (live or fixed). This is also supported by the similarity of LSCM measurements of the media whether taken on the X-Y plane or the Z axis. We suggest that the difference observed in wall measurements between methods is due to an underestimation of wall thickness when measured with the perfusion myograph VDA. Adventitia is more translucent than the media, and it is difficult to detect when the vessel is brightly illuminated for optimal visualization of the lumen. As seen in Fig 1A⇑, it is likely that much of the adventitia and intima are not included in the “wall” measurement taken with the VDA. Therefore, what has been defined as “wall” in previous studies probably was mainly media, resulting in underestimation of adventitia and endothelium. In addition, the terminology used in the literature to define “wall” measurements is confusing, reflecting the difficulty of clearly measuring different layers. Some studies use the term “wall thickness,”7,29,31,32 others use “media thickness,”24,33,34 and one established that “media” was clearly discernible but considered it “wall.”30
In the images captured with LSCM in the X-Y plane, the adventitia and endothelium were more clearly discernible than with VDA. Nevertheless, we considered the images to be inadequate for providing an exact measurement of these layers because the outer limits of adventitia and intima were not always clearly shown. Conversely, the stacks of sharp nuclei, obtained at sequential Z-axis depths, allow the construction of accurate three-dimensional images. The ability to distinguish between cell types with the use of nuclear dyes14,15 allows visualization and separate measurement within the different layers in the wall. The limiting factor for LSCM measurements is the quality of the images, which is in turn dependent on the thickness of the specimen. That means that the thickness of the wall represents a limitation of the technique, and the present version of this method for intact pressurized vessels is restricted to resistance arteries or thinner vessels. Another important factor is the capacity of the different cell nuclei to capture the dye; this is always less for the endothelial nuclei, even when the dye is applied intraluminally.
Effect of Fixation
We considered it important to determine the effect of fixation on lumen and wall dimensions for three reasons. First, the ability to use fixed tissue greatly facilitates the experimental process because it can be stored for later analysis, particularly when combined with histochemical methods, which are not practicable in live tissue. Second, the use of water-immersion lenses, which are necessary for live tissue measurements, limits the quality of image because they generally are not available with Plan Apochromatic corrections and cannot be made with a high NA. This is an essential feature, especially for Z-axis measurements, because the axial resolution rises with the square of the NA.10 Third, fixation allows the use of a wider range of dyes, including membrane-impermeable dyes such as propidium iodide, that can be detected with a red-green filter with a very good signal.19
Fixatives are known to shrink tissues. Formaldehyde, the fixative used in our study, is superior to others, minimizing this problem, especially when prepared in saline solution.35,36 In our experimental conditions, the adventitia, media, and intima thickness were not affected by this fixative. However, lumen demonstrated a small reduction, an effect that was equally observed with VDA or LSCM measurements. This effect could be partially due to a slight contraction of the vessel produced by the change in temperature (from the PSS at 37°C to the formal-saline solution at room temperature). This temperature-induced contraction could be further minimized by heating the fixative to 37°C before its addition to the perfusion bath.37
In conclusion, LSCM uncovered previously unknown aspects of the remodeling of the vascular wall in SHRSP rats, a genetic model of hypertension. This included changes in adventitia, a layer that has not been previously assessed in studies of small vessel physiology. LSCM is a reliable method with which to measure morphology of intact resistance arteries in live and fixed conditions, and it has several advantages over previous methodology. Limitations of the technique are the development of fluorescent dyes and the improvement of objective lenses and of software for image analysis. We are confident that the further development of LSCM for measurements of vascular morphology will help us to gain knowledge about the structure of resistance arteries and its changes during remodeling in different pathological conditions.
Selected Abbreviations and Acronyms
| CSA | = | cross-sectional area |
| LSCM | = | laser scanning confocal microscopy |
| NA | = | numerical aperture |
| SHRSP | = | stroke-prone spontaneously hypertensive rat(s) |
| SMC | = | smooth muscle cells |
| VDA | = | video dimension analyzer |
| WKY | = | Wistar-Kyoto rat(s) |
Acknowledgments
This work was supported by grants from the British Heart Foundation (PG 95123 and FS93025), Medical Research Council Clinical Research Initiative in Heart Failure (PG 9307850), and Wellcome Trust (045924/Z/95). Dr González was a recipient of a short-term fellowship PR95–392 from DGICYES, Spain. We are grateful to Craig Daly for his advice with LSCM and Dr Ian Montgomery for his help with histology.
- Received June 9, 1997.
- Revision received June 27, 1997.
- Accepted June 27, 1997.
References
This Issue
Jump to
Article Tools
Share this Article
- Cellular Aspects of Vascular Remodeling in Hypertension Revealed by Confocal Microscopy
