[A patient with inflammatory arthralgia, paraesthesia and atrial fibrillation]. / Ein Patient mit entzündlichen Arthralgien, Parästhesien und Vorhofflimmern.

A 61-year-old male patient had suffered from inflammatory arthralgia since 2014 and had been taking a proton pump inhibitor (PPI) for many years. Additionally, the patient was prescribed apixaban for atrial fibrillation. The blood tests on admission showed a marked hypomagnesemia with secondary hypocalcemia and hypokalemia. After discontinuation of the PPI and through substitution of magnesium, the blood results normalized completely and the patient was finally symptom-free. This case report underlines the important role of magnesium in rheumatology and presents the various mechanisms of action of magnesium.

Endothelial capillaries chemotactically attract tumour cells.

Directional migration of capillaries towards tumour implants is generally assumed to be regulated by chemotaxis. Preliminary evidence has also been presented for the existence of a reverse chemotactic signalling pathway, with capillaries attracting tumour cells via paracrine factors. By using a variety of endothelial cell types and tumour cell lines, this study has systematically investigated chemotaxis between endothelial cells and tumour cells in two- and three-dimensional systems. Checkerboard analysis revealed faint attraction of human umbilical vein endothelial cells (HUVECs), but not porcine aortic endothelial cells (PAECs), by tumour cells. In reverse, both PAECs and HUVECs potently induced chemotactic migration of tumour cells. Using a microcarrier-based fibrin gel assay, directional migration of endothelial cells towards tumour cells was not observed. In reverse, tumour cells were strongly attracted by endothelial cells. Identification of endothelium-derived chemotactic molecules may provide a valuable approach for the treatment of tumour metastasis.

Angiostatic effects of suramin analogs in vitro.

Suramin analogs are polyanionic naphthylureas structurally related to suramin, an antitumor agent with a narrow therapeutic window. The angiostatic activities of suramin and 16 suramin analogs were investigated using an easily quantifiable in vitro angiogenesis system. In addition, the antiproliferative activities of the analogs were studied in four different human tumor cell lines and in porcine aortic endothelial cells. The suramin analogs encompassed two main structural variations, i.e. their molecular size, and the number and substitution pattern of the sulfonate groups. Some suramin analogs with a reduced number of sulfonate groups (NF062, NF289 and NF326) showed significant dose-dependent angiostatic and also antiproliferative activities. The disulfonate NF062 was superior to suramin in inhibiting HT29 and T47D tumor cells while demonstrating a similar angiostatic potential as suramin. Therefore, the sulfonate groups in the para position of the amino groups of the naphthyl residues of suramin seem to be of special importance. The very small disulfonates (NF108, NF109, NF499, NF500 and NF241) and the asymmetric compound NF520, one half of the suramin molecule, are inactive. Therefore, a minimal molecule size seems to be essential for the biological activity. Suramin is a rather rigid molecule. The highly flexible analogs (NF527, NF528 and NF529) are inactive. This indicates that the molecular rigidity is important for the biological activity.

Are hyaluronan receptors involved in three-dimensional cell migration?

Hyaluronan (HA), an unbranched polysaccharide consisting of repeated glucuronic acid/N-acetylglucosamine disaccharide units, is ubiquitously present in the extracellular matrix of many tissues (for a more comprehensive review see: Fraser et al., 1997). Increased amounts of hyaluronan are produced by solid tumors and tumor-associated fibroblasts, and tumor-induced HA is correlated with poor prognosis. HA is well known to stimulate the migration of a large variety of cell types. Stimulation of cell migration by HA has been explained by different mechanisms. HA was shown to specifically bind to cell surface receptors, and inhibition of HA-receptor function was demonstrated to decrease cell migration and tumor growth. On the other hand, HA as a large hydrophilic molecule is also known to modulate the extracellular packing of collagen and fibrin, leading to increased fiber size and porosity of extracellular substrates. Hence a modified matrix architecture might similarly account for increased locomotion of cells. In this review, we attempted to summarize the available data on HA-induced cell migration, with particular emphasis on the role of HA receptors in three-dimensional cell migration. Although the HA receptor CD44 has been shown to mediate migration of cells over two-dimensional hyaluronan-coated surfaces in vitro, there is only little evidence that HA-binding to CD44 or other HA receptors has major impact on the locomotion of cells through three-dimensional matrices in vivo. We showed recently that the promigratory effect of HA in fibrin gels is largely due to HA-mediated modulation of fibrin polymerization. By increasing the porosity of fibrin gels, HA strongly accelerates cell migration. The porosity of matrices therefore appears as an important and probably underestimated determinant of cell migration and tumor spread.

Lysophosphatidic acid activates nuclear factor kappa B and induces proinflammatory gene expression in endothelial cells.

The cellular phospholipid, lysophosphatidic acid (LPA), released by activated platelets and fibroblasts or, at high levels, from ovarian and cervical carcinomas is a powerful serum mitogen that may modulate several signaling pathways in endothelial cells (EC). Hence, LPA could function in a paracrine manner during EC-platelet interactions at sites of vascular injury. Here, we demonstrate activation of the transcription factor nuclear factor kappa B (NF-kappaB) in EC following exposure to LPA. EC activation was further characterized by increased levels of mRNA transcripts encoding E-selectin, Intercellular Adhesion Molecule-1, Interleukin-8 and Monocyte Chemoattractant Protein-1. These effects were inhibited by preincubating EC either in the presence of mepacrine (to block phospholipase A2) or of pertussis toxin (to increase ADP-ribosylation of Gi proteins). No inhibition was observed in the presence of putative LPA receptor antagonists suramin or thrombospondin. LPA induces a proinflammatory activation of endothelial cells that (i) involves Gi proteins; (ii) depends on phospholipase A2 activity; (iii) is associated with the activation of NF-kappaB and (iv) results in increased expression of proinflammatory genes. We propose that LPA release by activated platelets may directly modulate vascular inflammatory responses.

Hyaluronan stimulates tumor cell migration by modulating the fibrin fiber architecture.

The glycosaminoglycan hyaluronan, which supports tumor cell migration and metastasis, interferes with fibrin polymerization and leads to increased fiber size and porosity of fibrin clots. Here we have studied the proportionate effect of fibrin polymerization on hyaluronan-mediated migration of glioblastoma cells. The structural and physical properties of hyaluronan-containing fibrin gels were analyzed by turbidity measurement, laser scanning microscopy, compaction assay, and calculation of pore size by liquid permeation. When fibrin polymerized in the presence of hyaluronan or dextran, the resulting gels strongly stimulated cell migration, and migration significantly correlated with fiber mass-to-length ratios and pore diameters. In contrast, cell migration was not induced by addition of hyaluronan to supernatants of already polymerized gels. Hyaluronan-mediated migration was inhibited in fibrin gels by antibodies to alphav- and beta1integrins and the disintegrin echistatin, but not by antibodies to the hyaluronan receptor CD44 (up to 50 microg/ml). As a control, we show that anti-CD44 (10 microg/ml) inhibited cell migration on a pure hyaluronan matrix using a two-dimensional Boyden chamber system. In contrast to three-dimensional migration, the migration of cells on the surfaces of variably structured fibrin gels was not significantly different, indicating that increased gel permeability (porosity) may account for hyaluronan-mediated migration. We conclude that, in complex three-dimensional substrates, the predominant effect of hyaluronan on cell migration might be indirect and requires modulation of fibrin polymerization.

Contact-dependent inhibition of angiogenesis by cardiac fibroblasts in three-dimensional fibrin gels in vitro: implications for microvascular network remodeling and coronary collateral formation.

Angiogenesis and coronary artery collateral formation can improve blood flow and thereby prevent myocardial ischemia. The role of perivascular fibroblasts in neovascularization remains incompletely understood. Here we investigated the effects of epicardial and myocardial fibroblasts on angiogenesis in vitro by using a serum-free microcarrier-based fibrin gel angiogenesis system. To clearly distinguish between different cell types, we either stained endothelial cells or fibroblasts in the living with 1, 1'-dioctadecyl-3,3,3',3'-tetramethyl-indocarbocyanine-perchlorate (DiI). In cocultures, low numbers of heart fibroblasts stimulated endothelial sprouting, and capillary growth was also induced by fibroblast-conditioned media, indicating a paracrine mechanism. Capillary formation was decreased by increasing the density of fibroblasts in the cocultures, indicating contact-dependent inhibition. Using time-lapse studies, it turned out that close contacts between fibroblasts and endothelial cells resulted in rapid retraction of endothelial cells or, rarely, in cell death. Depending on the local ratio of fibroblasts to endothelial cell numbers, fibroblasts determined the location of capillary growth and the size of developing capillaries and thereby contributed to capillary network remodeling. In contrast to primary heart fibroblasts, NIH 3T3 fibroblasts did not display contact-dependent inhibition of endothelial sprouts. NIH fibroblasts were frequently seen in close association with endothelial capillaries, resembling pericytes. Contact-dependent inhibition of angiogenesis by epicardial fibroblasts could not be reversed by addition of neutralizing anti-TGF-beta1 antibodies, by addition of serum, of medium conditioned by hypoxic tumor cells or myocardium, by various cytokines or by growing cocultures under hypoxic conditions. Our results implicate a pivotal role of periendothelial mesenchymal cells for the regulation of microvascular network remodeling and collateral formation.

Expression of cGMP-dependent protein kinase I and its substrate VASP in neointimal cells of the injured rat carotid artery.

BACKGROUND: Neointimal fibroproliferative lesions after balloon angioplasty remain a major clinical problem, frequently leading to restenosis of initially successfully dilated coronary arteries. The cGMP-dependent protein kinase type I (cGMP-PK I) and its substrate vasodilator-stimulated phosphoprotein (VASP), molecular targets of the atrial natriuretic factor (ANF) and nitric oxide (NO) signalling pathways, are likely to be involved in various aspects of vascular wall regulation and restenosis formation. METHODS: To investigate the occurrence of cGMP-PK I and VASP in neointimal cells in situ, we performed immunohistochemistry and immunoblotting experiments on denuded rat carotid arteries. RESULTS: Although the soluble cGMP-PK I showed a homogeneous distribution throughout the neointima, VASP apparently was more concentrated in smooth muscle cells (SMCs) lining the artery lumen, possibly reflecting enhanced growth factor stimulation of luminal SMCs. The membrane-associated cGMP-PK type II could not be detected in both the non-injured vessel wall and the restenotic tissue. CONCLUSION: The presence of both cGMP-PK I and VASP, major regulators of the actin cytoskeleton and cell motility, in neointimal tissue suggest that this emerging signal transduction pathway could be a target for the regulation and control of restenosis.

Guided migration as a novel mechanism of capillary network remodeling is regulated by basic fibroblast growth factor.

To investigate mechanisms of capillary network remodeling, we developed a serum-free angiogenesis in vitro system in three-dimensional fibrin matrices which allows the study of directional growth of endothelial sprouts, anastomosis, and remodeling ('pruning') of the primitive plexus toward more elaborated capillary trees. To follow the movements of living endothelial cells by inverse-fluorescence microscopy, we cocultured unlabeled endothelial cells with endothelial cells labeled with the carbocyanine dye 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate (DiI). We show that elongation and retraction of neighboring capillary sprouts occurs simultaneously, resembling a tug-of-war by which endothelial cells are withdrawn from shortening sprouts to become incorporated in other sprouts nearby. For the first time, we directly demonstrate the long-suspected parallel sliding movement of endothelial cells. We show that cell migration persists within immature capillaries even after sprouts have merged to continuous capillary loops, leading to overlapping growth of opposing sprout tips. As a novel concept of capillary remodeling, we distinguish two types of endothelial cell migration: sprouting and guided migration. Sprouting is the de novo invasion of a matrix by endothelial cells, and guided migration is the locomotion of cells along preexistent capillary-like structures. We show that guided migration leads to remodeling of immature capillary networks and to the retraction of sprouts. We describe a method for quantification of sprouting versus guided migration in DiI-mosaic-labeled capillary networks, and we present evidence that endothelial cell-derived basic fibroblast growth factor serves as a chemotactic signal for other cells to migrate along a preestablished capillary-like structure.

The configuration of fibrin clots determines capillary morphogenesis and endothelial cell migration.

In the living organism, capillary growth frequently occurs in a fibrin-rich extracellular matrix. The structure and the mechanical properties of fibrin clots are influenced by various macromolecules (i.e., hyaluronic acid and thrombospondin) and also by pH, ionic strength, and thrombin concentrations of the milieu in which they polymerize. The configuration (three-dimensional architecture) and the rigidity of fibrin clots correlate with their opacity measured by spectrophotometric absorbance readings at 350 nm. By using bovine pulmonary artery endothelial cells and bovine fibrinogen, we show here that transparent fibrin clots (A(350) < 1.0), polymerized at > or = pH 7.5 or in the presence of increased thrombin or sodium chloride concentrations, strongly stimulated capillary morphogenesis in vitro. In contrast, opaque fibrin gels (A(350) > 1.5), polymerized at pH 7.2 or in the presence of dextran, stimulated only the migration of endothelial cells but not capillary morphogenesis. We demonstrate that the angiomorphogenic effects of basic fibroblast growth factor (bFGF) and vascular endothelial growth factor (VEGF) are strongly dependent on the structure of the fibrin clots. Our findings suggest that bFGF/VEGF primarily stimulate the proliferation of endothelial cells, whereas the three-dimensional architecture of the fibrin matrix is decisive for capillary morphogenesis.

Development of pericyte-like cells during angiogenesis in quail chick chimeras as detected by combined Feulgen reaction and immunohistochemistry.

Perivascular fibroblasts have been proposed as possible precursor cells for microvascular pericytes. To investigate the development of pericytes during angiogenesis we examined interspecific grafts between chick and quail embryos. Limb buds of three-day old quail embryos were transferred to the chorioallantoic membrane (CAM) of ten to fourteen day-old chick embryos. Six days after grafting, the limb buds were explanted and histologically examined by combined Feulgen reaction and immunohistochemistry using an antibody to quail endothelial and hemopoietic cells (QH-1). Limb buds were found to be vascularized by a network of capillaries which were partially derived from sprouts of the chick CAM microvasculature. Numerous hybrid capillaries were detected, consisting of host endothelial cells (chick) and graft pericytes (quail). These results provide further support for the idea that microvascular pericytes can evolve from perivascular fibroblasts.

A microcarrier-based cocultivation system for the investigation of factors and cells involved in angiogenesis in three-dimensional fibrin matrices in vitro.

Angiogenesis in situ includes coordinated interactions of various microvascular cell types, i.e., endothelial cells, pericytes and perivascular fibroblasts. To study the cellular interactions of microvascular cells in vitro, we have developed a microcarrier-based cocultivation system. The technical details of this method include seeding of endothelial cells on unstained cytodex-3 microcarriers and seeding of pericytes, fibroblasts or vascular smooth muscle cells on microcarriers which have been labeled by trypan blue staining. A mixture of both unstained and trypan blue-stained microcarriers was subsequently embedded in a three-dimensional fibrin clot. The growth characteristics of each cell type could be conveniently observed since the majority of cells left their supporting microcarriers in a horizontal direction to migrate into the transparent fibrin matrix. As differently stained microcarriers were randomly arranged in the fibrin matrix, the characteristic patterns of the microcarriers allowed location of particular points of interest at different developmental stages, facilitating the observation of cellular growth over the course of time. One further advantage of this microcarrier-based system is the possibility of reliably quantifying capillary growth by determination of average numbers of capillary-like formations per microcarrier. Thus, this model allows convenient evaluation of the effects of non-endothelial cells on angiogenesis in vitro. By using this coculture system, we demonstrate that endothelial capillary-like structures in vitro do not become stabilized by contacting vascular smooth muscle cells or pericytes during the initial stages of capillary formation.

A novel, microcarrier-based in vitro assay for rapid and reliable quantification of three-dimensional cell migration and angiogenesis.

Angiogenesis in situ occurs within the interstitial extracellular matrix. The complexity of currently used three-dimensional in vitro angiogenesis systems makes it difficult to quantify cellular growth and neovessel formation. To overcome this problem we were interested to develop an angiogenesis system which allows rapid and reliable quantification of three-dimensional neovessel formation in vitro. Endothelial cells were seeded on gelatine-coated microcarriers (MCs). Cell-coated MCs were suspended in a solution of fibrinogen which was then induced to polymerize by addition of thrombin. By this way, MCs were entrapped in a three-dimensional fibrin matrix. Within a few hours, endothelial cells began to leave their supporting microcarriers and to migrate into the fibrin gel. Without addition of stimulators of angiogenesis, endothelial cells showed incoherent migration into the matrix. In contrast, in response to fibronectin, basic fibroblast growth factor (bFGF), or vascular endothelial growth factor (VEGF), respectively, endothelial cells assembled to form multicellular capillary-like structures occasionally exceeding 1000 microns in length. Each MC gave rise to a limited number of capillaries. A single culture dish contained hundreds of MCs, ensuring that a sufficient number of random samples was present for a reliable statistical evaluation. The angiogenic response could be easily quantified by determination of the average number of capillary-like formations per MC (cap/MC). The capillary count for macrovascular endothelial cells from the bovine pulmonary artery was 0.14 cap/MC when no angiogenic stimulators were contained within the fibrin gel. Addition of 200 micrograms/ml fibronectin increased capillary formation to 0.63 cap/MC (P < 0.0001) at Day 6. Already after 3 days, addition of bFGF (30 ng/ml) yielded a capillary count of 1.05 and addition of VEGF (100 ng/ml) resulted in 0.91 cap/MC. In contrast, addition of hyaluronic acid stimulated migration of dispersed endothelial cells into the fibrin matrix without leading to significant capillary formation (0.09 cap/MC). Hydrocortisone alone or in combination with heparin led to a significant inhibition of bFGF-stimulated angiogenesis. We thus have developed a convenient angiogenesis in vitro system which allows reliable quantification of capillary formation in a three-dimensional environment. Based on this assay we conclude that apart from proliferation and migration of endothelial cells, angiogenesis additionally requires the assembly of cells to form multicellular capillaries. This process is strongly induced by the extracellular matrix protein fibronectin. Hyaluronic acid, on the other hand, promotes migration but not capillary formation (assembly).

The effect of fibroblasts, vascular smooth muscle cells, and pericytes on sprout formation of endothelial cells in a fibrin gel angiogenesis system.

We have recently shown that during angiogenesis in situ, sprouting and newly formed capillaries appear to be composed of two cell types, endothelial cells and nonendothelial, pericyte-like cells. The effect of pericytes on the process of neovessel formation is largely unknown. To study the influence of nonendothelial cell types on endothelial tubule formation, we have performed coculture experiments in a fibrin-clot angiogenesis system. When seeded below a critical density on the surface of fibrin gels, endothelial cells (from macro- or microvascular origin) did not show spontaneous formation of sprouts. However, in superconfluent cell cultures or after stimulation of endothelial cells with basic fibroblast growth factor (bFGF), endothelial cells frequently acquired an elongated shape. By stimulation of endothelial cells with both bFGF and vascular endothelial growth factor (VEGF), development of short capillary-like structures was induced. When endothelial cells were cocultivated with a cell type of high fibrinolytic potential, i.e., fibroblasts, development of capillary-like formations could not be detected. Cocultivation of endothelial cells with vascular smooth muscle cells or with retinal pericytes also did not increase the number of capillary-like formations in fibrin gels. In contrast, vascular smooth muscle cells on their own could be demonstrated to give rise to branched capillary-like networks in fibrin, which easily could be mistaken for true capillaries. Our results indicate that periendothelial cells contribute to angiogenesis not only by fibrinolysis and proteolytic permeation of the extracellular matrix. Rather, the interactions of endothelial cells and pericyte-like cells, as frequently observed during neovessel formation in situ, appear to be more specific and may require factors hitherto unknown.

Different sequences of expression of band 3, spectrin, and ankyrin during normal erythropoiesis and erythroleukemia.

Expression of the erythrocyte anion exchanger band 3, and ankyrin and spectrin, two cytoskeletal proteins of the red blood cell membrane, was studied by immunofluorescence using: 1) smears of human bone marrow from healthy donors and from a patient with erythroleukemia, 2) human red blood cell precursors grown in cell culture, and 3) murine erythroleukemia cells grown in cell culture. Double immunostaining with antibodies to band 3 in combination with spectrin or ankyrin revealed that these proteins become expressed synchronously during normal human erythropoiesis. In contrast, both murine erythroleukemia cells (induced by fibronectin and dimethyl sulfoxide to differentiate in vitro) and erythroblasts from a patient suffering from erythroleukemia displayed distinct asynchronicity in expression of these proteins, ie, ankyrin and spectrin were synthesized first, followed by band 3 at a later stage of erythroid development. After the onset of band 3 expression in human erythroleukemia cells, an increase of membrane-associated fluorescence was detectable for both ankyrin and spectrin, supporting the general view that band 3 promotes assembly of the membrane cytoskeleton. These findings indicate that the current concept of a sequential expression of spectrin/ankyrin and band 3 is valid only for erythroleukemia cells or transformed erythropoietic cell lines but does not occur in normal erythropoiesis, during which these proteins become expressed simultaneously.

Pericyte involvement in capillary sprouting during angiogenesis in situ.

To investigate the participation of microvascular pericytes in the process of capillary sprouting, we examined whole-mount preparations of the rat mesentery by use of a double immunofluorescence approach. Angiogenesis was induced by intraperitoneal injections of either the mast cell-degranulating substance compound 48/80 or tumor cell-conditioned medium. Capillary sprouts were visualized by staining with rhodamine-conjugated phalloidin and pericytes were simultaneously stained by an antibody to the intermediate filament protein desmin. Developing pericytes were negative for the smooth-muscle isoform of alpha-actin, but were clearly reactive for desmin. Pericytes appear to be involved in the earliest stages of capillary sprouting. Pericytes were regularly found lying at and in front of the advancing tips of endothelial sprouts. At many sites pericytes were seen to bridge the gap between the leading edges of opposing endothelial sprouts, which were apparently preparing to merge, suggesting that pericytic processes may serve as guiding structures aiding outgrowth of endothelial cells.

Adducin in erythrocyte precursor cells of rats and humans: expression and compartmentalization.

Adducin is a calmodulin-binding protein involved in the assembly of the erythrocyte membrane skeleton. To investigate the expression of adducin during human erythropoiesis, we performed immunofluorescence studies on smears of cultured human erythroblasts. Adducin immunoreactivity was found in the early stages of erythropoiesis. Proerythroblasts were the first erythroid precursor cells positive for adducin. The adducin signal was very similar to the signal of erythroid beta-spectrin in that both proteins lined the membrane of erythroid precursor cells. Cell fractionation experiments were performed to further analyze the intracellular distribution of adducin in erythroid cells. In erythroblasts, about 60% of total cellular adducin appeared in the Triton-soluble fraction. In reticulocytes, the Triton-soluble fraction decreased to 30% of total reticulocyte adducin. Erythrocytes had no detectable amount of adducin in the Triton-soluble pool. Instead, adducin was quantitatively bound to the Triton-insoluble erythrocyte cytoskeleton. Our results suggest that adducin is expressed early in the development of the erythrocyte membrane skeleton, whereas stable assembly onto the membrane skeleton does not occur before the final stages of mammalian erythropoiesis.

Demonstration of actin filament stress fibers in microvascular endothelial cells in situ.

We have developed a method for immunostaining the microvascular tree of rat mesenteric windows in situ. The procedure consists of three steps, i.e., mild fixation with formaldehyde, controlled proteolytic digestion of the mesothelial layer, and permeabilization with acetone. Discrimination between different microvascular segments was possible by double-fluorescent staining with antibodies to the smooth muscle isoform of alpha-actin and to nonmuscle myosin from platelets. Antibodies to nonmuscle myosin labeled numerous longitudinally oriented cables in endothelial cells of all microvascular segments (arterioles, metarterioles, pre-, mid-, and postcapillaries, small venules). Occasionally, the myosin-containing cables displayed the interrupted sarcomere-like staining pattern that is diagnostic for stress fibers. In contrast, staining of actin filaments with phalloidin-rhodamin resulted in a noninterrupted, continuous fluorescence of the stress fibers. A possible functional role of microvascular endothelial stress fibers is to serve as a tensile cytoskeletal scaffold that stabilizes the tubular, three-dimensional geometry of microvessels and, in addition, to help the endothelium resist the shear forces created by blood flow and by collision with red and white blood cells.

Heterogeneity of microvascular pericytes for smooth muscle type alpha-actin.

Microvascular pericytes are believed to be involved in various functions such as regulation of capillary blood flow and endothelial proliferation. Since pericytes represent a morphologically heterogeneous cell population ranging from circular smooth musclelike to elongated fibroblast-like morphology it is possible that regulation of blood flow (via contractility) and control of endothelial proliferation (as well as other metabolic functions) may be accomplished by different subsets of pericytes. In the present study we provide evidence for heterogeneity of pericytes at the molecular level by using two novel technical approaches. These are (a) immunostaining of whole mounts of the microvascular beds of the rat mesentery and bovine retina and (b) immunoblotting studies of microdissected retinal microvessels. We show that pericytes of true capillaries (midcapillaries) apparently lack the smooth muscle isoform of alpha-actin whereas transitional pericytes of pre- and postcapillary microvascular segments do express this isoform. Thus, regulation of capillary blood flow may be accomplished by the smooth muscle-related pre- and postcapillary pericytes whereas the nonmuscle pericytes of true capillaries may play a role in other functions.