Densely arranged pericytes engird the endothelial tube of all coronary microvessels. Since the experimental access to these abundant cells in situ is difficult, a prerequisite for broader investigation is the availability of sufficient numbers of fully differentiated pericytes in homogenous culture. To reach this goal, we applied strictly standardized cell isolation techniques, optimized culture methods and specific histological staining. Approximately 1,000-fold enriched pericytes were proteolytically detached from highly purified coronary microvascular networks (density gradient centrifugation) of eight mammalian species including human. Addition of species-autologous fetal or neonatal serum (10-20% vol/vol) was a precondition for longer term survival of homogenous pericyte cultures. This ensured optimal growth (doubling time <14 h) and full expression of pericyte-specific markers. In 3-mo, 10(10) pericytes (15 g) could be cultivated from 1 bovine heart. Pericytes could be stored in liquid N(2), recultured, and passaged repeatedly without loss of typical features. In cocultures with EC or vascular smooth muscle cells, pericytes transferred fluorescent calcein to each other and to EC via their antler-like extensions, organized angiogenetic sprouting of vessels, and rapidly activated coagulation factors X and II via tissue factor and prothrombinase. The interconnected pericytes of the coronary system are functionally closely correlated with the vascular endothelium and may play key roles in the adjustment of local blood flow, the regulation of angiogenic processes, and the induction of procoagulatory processes. Their successful bulk cultivation enables direct experimental access under defined in vitro conditions and the isolation of pericyte specific antigens for the production of specific antibodies.

The barrier functions of myocardial precapillary arteriolar and postcapillary venular walls (PCA or PCV, respectively) are of considerable scientific and clinical interest (regulation of blood flow and recruitment of immune defense). Using enzyme histochemistry combined with confocal microscopy, we reexamined the cell architecture of human PCA and PVC and reconstructed appropriate in vitro models for studies of their barrier functions. Contrary to current opinion, the PCA endothelial tube is encompassed not by smooth muscle cells but rather by a concentric layer of pericytes cocooned in a thick, microparticlecontaining extracellular matrix (ECM) that contributes substantially to the tightness of the arteriolar wall. This core tube extends upstream into the larger arterioles, there additionally enwrapped by smooth muscle. PCV consist of an inner layer of large, contractile endothelial cells encompassed by a fragile, wide-meshed pericyte network with a weakly developed ECM. Pure pericyte and endothelial cell preparations were isolated from PCA and PCV and grown in sandwich cultures. These in vitro models of the PCA and PCV walls exhibited typical histological and functional features. In both plasma-like (PLM) and serum-containing (SCM) media, the PCA model (including ECM) maintained its low hydraulic conductivity (LP ϭ 3.24 Ϯ 0.52 · 10 Ϫ8 cm·s Ϫ1 · cmH2O Ϫ1 ) and a high selectivity index for transmural passage of albumin (SIAlb ϭ 0.95 Ϯ 0.02). In contrast, LP and SIAlb in the PCV model (almost no ECM) were 2.55 Ϯ 0.32 · 10 Ϫ7 cm·s Ϫ1 · cmH2O Ϫ1 and 0.88 Ϯ 0.03, respectively, in PLM, and 1.39 Ϯ 0.10 · 10 Ϫ6 cm·s Ϫ1 · cmH2O Ϫ1 and 0.49 Ϯ 0.04 in SCM. With the use of these models, systematic, detailed studies on the regulation of microvascular barrier properties now appear to be feasible.

We hypothesized that postcapillary venules play a central role in the control of the tightness of the coronary system as a whole, particularly under inflammatory conditions. Sandwich cultures of endothelial cells and pericytes of precapillary arteriolar or postcapillary venular origin from human myocardium as models of the respective vascular walls (sandwich cultures of precapillary arteriolar or postcapillary venular origin) were exposed to thrombin and components of the acutely activatable inflammatory system, and their hydraulic conductivity (L(P)) was registered. L(P) of SC-PAO remained low under all conditions (3.24 ± 0.52·10(-8)cm·s(-1)·cmH(2)O(-1)). In contrast, in the venular wall model, PGE(2), platelet-activating factor (PAF), leukotriene B(4) (LTB(4)), IL-6, and IL-8 induced a prompt, concentration-dependent, up to 10-fold increase in L(P) with synergistic support when combined. PAF and LTB(4) released by metabolically cooperating platelets, and polymorphonuclear leucocytes (PMNs) caused selectively venular endothelial cells to contract and to open their clefts widely. This breakdown of the barrier function was preventable and even reversible within 6-8 h by the presence of 50 μM quercetin glucuronide (QG). LTB(4) synthesis was facilitated by biochemical involvement of erythrocytes. Platelets segregated in the arterioles and PMNs in the venules of blood-perfused human myocardium (histological studies on donor hearts refused for heart transplantation). Extrapolating these findings to the coronary microcirculation in vivo would imply that the latter's complex functionality after accumulation of blood borne inflammatory mediators can change rapidly due to selective breakdown of the postcapillary venular barrier. The resulting inflammatory edema and venulo-thrombosis will severely impair myocardial performance. The protection afforded by QG could be of particular relevance in the context of cardiosurgical intervention.