Regional differences in statistical geometry of endothelial dense granules in human extremity veins

Introduction

In a previous study, our group quantitatively determined the amount of electron-dense vesicles using electron microscopy (EM) in rat limb vein endothelium subjected to different levels of orthostatic type gravitational stress. ¹ The appearance of these structures corresponded to secretory vesicles. They had round or oval shape with 40–250 nm size and an electron-dense content. Several of the vesicles exhibited the morphological characteristics of Weibel–Palade bodies (WPBs) with tubular structures and cigar-shaped appearance. ² These dense vesicles were localized mainly at the luminal side of the endothelial cells and could be distinguished from pinocytotic microvesicles which are much smaller and clearer structures. We tested and have proven the presence of endothelin (ET) and platelet-derived growth factor in them by immunohistochemistry. The vesicular density within the endothelial cells of saphenous veins decreased significantly after long-term (two weeks) orthostasis. ¹ These observations raised the possibility that production, development, and emptying of these dense vesicles are connected to gravitational stress.

Gravitational load applied in the form of keeping the animals chronically in tilted cages induces changes in venous network geometry, ³ increases sympathetic innervation density in the venous wall,^4,5 increases venous capacity,^6,7 reduces distensibility, augments myogenic tone,^8,9 and alters lumen volume regulation.^10,11 Human upper and lower body veins are working under much different gravitational and hemodynamic conditions. This is reflected in their different in vivo ¹² and in vitro biomechanics and histological structures. ¹³

Based on the above observations, the question arises whether chronic orthostatic stress results in similar effects on vesicle content of human extremity veins. In the present study, we made comparisons between different body regions exposed to different orthostatic loads to find out if the proportion of the vesicular area depends on the region. In addition to overall vesicle content, we also examined the shape of the vesicles using the techniques of statistical geometry.

Materials and methods

Results

Analysis of the electron micrographs of the specimens revealed that similar to the findings in the rat, human venous endothelium also contains dense vesicles. The vesicles identified in each vessel sample had the size of 30–600 nm (Figure 1(a) and (b)). Only a fraction of them showed a typical cigar-shaped Weibel–Palade body morphological appearance. OPEN IN VIEWER

The proportion of the total area of the dense vesicles to that of the endothelial cross section in all samples was 2.26 ± 0.34%. This value calculated for the upper extremity was 2.29 ± 0.49% and 2.23 ± 0.48% for lower limb without statistically significant difference. When lower limb samples were divided, vesicular densities were found to be 2.67 ± 0.94% in the groin and 1.85 ± 0.39% in the leg (n.s.). We calculated the vesicular densities of areas with different chronic orthostatic stress (upper extremity vs. leg) that also proved to be statistically insignificant (Figure 2). However, marked differences have been identified in granule geometry. Examining the geometry of the dense vesicles in leg and arm samples by measuring the ratio of long and short axes of 2540 granules, we found significant difference with the χ² test, p < 0.01 (Figure 3). Analyzing further that difference, we compared the relative frequencies of granules with different short and long axes from endothelial cells of human leg and arm veins. We found the vesicles from arm samples to be more circular and granules from the endothelium of leg veins more frequently elongated (Figure 4). The right block of Figure 4 demonstrates the difference between leg and arm plots. It gives the impression that what is missing in more circular granules in leg endothelial cells (blue spots) developed into more elongated cross sections (red spots). OPEN IN VIEWER

OPEN IN VIEWER

Figure 3.

Frequency of dense granules according to the ellipticity (the ratio of long/short axis) of their cross sections. Data based on measurement made on 2540 granules. Difference between leg and arm human small vein endothelial cells was significant with the χ² test, p < 0.01.

OPEN IN VIEWER

Figure 4.

Statistical analysis of dense granule geometry. Relative frequencies of granules with different short and long axes (normalized values in thousands). Left: in endothelial cells from human leg small veins; middle: in endothelial cells from small veins of human arm; right: their difference (leg minus arm values). Note higher frequency of more circular cross sections in arm vein endothelial cells and occurrence of more frequent elongated structures in the leg. Data based on measurement of 2540 granules.

Discussion

In a previous work from our laboratory, ¹ we found diminished amount of dense secretory granules of the saphenous vein endothelium in rats subjected to gravitational stress, while a similar alteration in the brachial venous endothelium, not affected by the gravitation was not seen. We have attributed it to a higher intensity rate of secretion in the leg induced by gravitation. As a logical next step of investigations, in the present work, we compared the amount of secretory granules present in human venous endothelial cells of upper and lower limbs. In addition to the volume occupied by secretory granules, the shape of their cross sections was also examined by statistical image analysis techniques.

No difference in the secretory granule volume could be identified; in fact, the similarity of these values is suggesting that (1) human veins are adapted to their common gravitational load and (2) there should be existing some control mechanism ensuring a typical 2.2–2.3% of cell volume in venous endothelial cells. However, the shape of the granules differed between the upper and lower body groups; in the leg venous endothelium, they were significantly more elongated. This might reflect differences in secretory content or different rates or phases of the maturing process in granule development. Based on our statistical analysis, we suppose that the originally spherical granules of the leg vein endothelium become more elongated.

There are several products known to be synthetized in the endothelium and stored in granules. The von Willebrand factor (VWF) is one of the most examined substances stored in organelles called WPBs. These bodies are described as rod-shaped, 0.1 µm wide, and up to 3 µm long tubular organelles that are dense on EM images. ² According to the latest results, VWF serves as major force behind the biogenesis of WPBs. During maturation of these special vesicles, VWF becomes highly multimerized and compacted causing tubular structure, elliptic shape, and remarkable increase of electron density of WPBs.^14–16

Besides VWF, WPBs contain agents that regulate hemostasis, inflammation, and hemodynamics. VWF primarily promotes platelet adhesion and aggregation. Coagulation factor VIII and XIII are substances with crucial roles in hemostasis. Other proven components, P-selectin and interleukin-8 have significant role in the inflammation process. The major effect of ET-1 and ET-converting enzyme is vasoconstriction, while calcitonin gene-related peptide induces vasodilation.^14,17,18 The content of the WPBs, the distribution of the different agents is not stable, depends on the localization of the endothelium and the physiological circumstances. ¹⁵ Exocytosis of WPBs is induced by hypoxia, ischemia, or radiation injury. Several proteins such as fibrin, thrombin, vascular endothelial growth factor or serotonin, histamine, and epinephrine activate WPB release.^14,17,19,20 Some other products, tissue plasminogen activator (tPA), protein S, and tissue factor pathway inhibitor have been detected to be stored in dense granules different from WPBs. These organelles are smaller than WPBs with the maximum diameter of 0.25 µm.^18,21

In a recent study from our laboratory, we compared the biomechanical properties of human vein samples also from gravitationally divergent body areas. In human superficial small leg veins, we found lower wall stress, higher distensibility at higher pressure range, lower elastic modulus, higher spontaneous or maximum contraction, and lower endothelial dilation compared to upper body vein samples. ¹³ In the current study, we could demonstrate significant difference in ellipticity between the two regions. One can speculate that the granules of the two regions undergo a different maturation process. This might contribute to the differences of thrombotic and hemodynamic properties of leg and upper body veins. We cannot exclude the possibility that pathological differences contributed to the differences observed in the two groups.

In summary, we demonstrated the presence of endothelial dense vesicles in human extremity superficial veins. Most probably, they are secretory vesicles, some of them having the morphological characteristics of WPBs.

To investigate the contribution of endothelial vesicles to adaptation to orthostatic stress in human extremity veins, we quantified these organelles. The overall mean ratio of the vesicular area to that of the endothelium proved to be 2.26%. Comparing the ratios of different regions with similar gravitational load (upper extremity and groin), we found no significant difference between the results. When we calculated the vesicular densities of areas with different orthostatic stress (upper limb vs. leg), the difference was also statistically insignificant. According to the outcome of these results, vesicular density within endothelial cells does not depend on the vascular region or orthostatic load. The lack of difference between areas of different gravitational stress might be attributed to the lifestyle and/or exposure of dynamic changes of hydrostatic pressure during daily activity in human. However, remarkable differences have been identified in granule ellipticity between areas exposed to different gravitational stresses. Measuring the ratio of long and short axes of granules from leg and arm samples, we found significant difference (p < 0.01). Comparing the relative frequencies of granules with different short and long axes, we found the vesicles from leg samples to be more elongated. It is probable that leg granules develop into more elongated ones, contributing to differences in thrombotic and hemodynamic behavior.