The pancreatic islet endothelial cell: emerging roles in islet function and disease.

The pancreatic islets are one of the most vascularized organs of the body. This likely reflects the requirements of the organ for a rich supply of nutrients and oxygen to the tissue, as well as the need for rapid disposal of metabolites and secreted hormones. The islet endothelium is richly fenestrated to facilitate trans-endothelial transport of secreted hormones, has a unique expression of surface markers, and produces a number of vasoactive substances and growth factors. The islet endothelial cells play a critical role in the early phase of type 1 diabetes mellitus by increasing the expression of surface leucocyte-homing receptors, thereby enabling immune cells to enter the endocrine tissue and cause beta-cell destruction. Following transplantation, pancreatic islets lack a functional capillary system and need to be properly revascularized. Insufficient revascularization may severely affect the transport properties of the islet endothelial system, resulting in a dysfunctional islet graft.

Haemodialysis monocytopenia: differential sequestration kinetics of CD14+CD16+ and CD14++ blood monocyte subsets.

In peripheral blood the majority of circulating monocytes present a CD14highCD16- (CD14++) phenotype, while a subpopulation shows a CD14lowCD16+ (CD14+CD16+) surface expression. During haemodialysis (HD) using cellulosic membranes transient leukopenia occurs. In contrast, synthetic biocompatible membranes do not induce this effect. We compared the sequestration kinetics for the CD14+CD16+ and CD14++ monocyte subsets during haemodialysis using biocompatible dialysers. Significant monocytopenia, as measured by the leucocyte count, occurred only during the first 30 min. However, remarkable differences were observed between the different monocyte subsets. CD14++ monocyte numbers dropped to 77 +/- 13% of the predialysis level after 15 min, increasing to > or = 93% after 60 min. In contrast, the CD14+CD16+ subset decreased to 33 +/- 15% at 30 min and remained suppressed for the course of dialysis (67 +/- 11% at 240 min). Approximately 6 h after the end of HD the CD14+CD16+ cells returned to basal levels. Interestingly, the CD14+CD16+ monocytes did not show rebound monocytosis while a slight monocytosis of CD14++ monocytes was occasionally observed during HD. A decline in CD11c surface density paralleled the sequestration of CD14+CD16+ monocytes. Basal surface densities of important adhesion receptors differed significantly between the CD14+CD16+ and CD14++ subsets. In conclusion, during HD the CD14+CD16+ subset revealed different sequestration kinetics, with a more pronounced and longer disappearance from the blood circulation, compared with CD14++ monocytes. This sequestration kinetics may be due to a distinct surface expression of major adhesion receptors which facilitate leucocyte-leucocyte, as well as leucocyte-endothelial, interactions.

Immortalized human endothelial cells have a decreased response to IL-1 secondary to an IL-1 receptor defective expression restored by corticosteroids.

A human endothelial cell line is a convenient tool for exploring cell physiology and testing drugs and toxics. Several attempts have been made using SV40 to immortalize endothelial cells. We used human umbilical vein endothelial cells (HUVEC) transformed with a construct made of promoter of the vimentin gene and SV40 Tag. The proliferation of immortalized vascular endothelial cells (IVEC), as measured by [methyl-3H]thymidine incorporation, was compared to that of HUVEC in the presence of endothelial cell growth factor and cytokines: tumor necrosis factor-alpha (TNF-alpha), interleukin-1beta (IL-1beta) and interferon-gamma (IFN-gamma). Inhibition of [methyl-3H]thymidine incorporation by IL-1beta was lower than that observed with HUVEC, while TNF-alpha reduced the proliferation of IVEC and HUVEC to similar extents. Induction of intercellular adhesion molecule (ICAM-1), vascular cell adhesion molecule (VCAM-1) and E-selectin by TNF-alpha, measured by a radiometric technique, was similar in IVEC and HUVEC, while the induction of E-selectin by IL-1beta on IVEC was limited and significantly different from that observed on HUVEC (p<0.001). The number of 125I-IL-1beta binding sites on IVEC is 3-fold less than on HUVEC and the IL-1beta receptor number was reduced. Dexamethasone treatment of IVEC restored their reactivity to IL-1beta and corrected the IL-1beta binding and the receptor number. These results showed that the introduction of SV40 gene not only immortalized the cell but also altered IL-1 receptor expression. This alteration may be improved by addition of corticosteroids to the cell culture, which extends the possibility of using IVEC as a model of endothelial cells.

In vitro expression of adhesion receptors and diapedesis by polymorphonuclear neutrophils during experimentally induced Streptococcus uberis mastitis.

The expression of adhesion receptors and diapedesis by polymorphonuclear neutrophils (PMN) were studied before and during experimentally induced Streptococcus uberis mastitis. Both quarters of the left half of the udders of five midlactation cows were inoculated with a suspension containing approximately 500 CFU of S. uberis 0140J. Clinical signs of an inflammatory reaction and leukocyte influx were observed 24 h after challenge. The expression of CD11b/CD18 adhesion receptors, determined by flow cytometry, was upregulated 24 h after challenge. A confluent monolayer of bovine secretory mammary epithelial cells on collagen-coated inserts was used to study PMN diapedesis. Bovine C5a was used as the chemoattractant. An 80% decrease in PMN diapedesis was observed 24 h after challenge. The decrease in diapedesis continued for 3 weeks after challenge.

Partial peripheral nerve injury leads to activation of astroglia and microglia which parallels the development of allodynic behavior.

Three different cell markers were studied in rats to note changes in the immunoreactivity (IR) in the lumbar spinal cord 1 to 84 days following partial sciatic nerve ligation (PSNL). Alteration in average IR were studied for complement receptor C3bi (OX42; microglia), major histocompatibility complex II (OX6; microglia), and glia fibrillary acidic protein (GFAP; astroglia). Thirty-four female rats underwent ligation of approximately 1/2 of the sciatic nerve (PSNL). This injury resulted in the development of mechanical allodynia, which was quantitated by the measurement of foot withdrawal threshold to the application of Von Frey filaments. Ipsilateral increase in IR of OX42 and GFAP was observed to occur within 2 days, maximized by day 14, and did not return to contralateral spinal gray matter IR levels by day 84 (time period of study). Increases in OX42 IR and GFAP IR were observed within the spinal segments innervated by the sciatic nerve. GFAP IR was not expressed in all lumbar segments. OX42 staining with the upper portion of the dorsal horn was found to localize within the areas of deafferentation demonstrated by loss of thiamine monophosphatase activity within the substantia gelatinosa. OX6 IR was only seen in a few cells within the ipsilateral gray matter, indicating that microglia did not become phagocytic. Both GFAP IR and OX42 IR were found to linearly correlate with allodynic behavior with OX42 IR being more statistically significant. Correlation of OX42 IR in only the upper portion of the dorsal horn (not including the neck) resulted in an even a greater level of significance. These findings demonstrate that microglia and astroglia are activated following PSNL and that their increase in IR correlates with the development of allodynic behavior.

Effect of linoleic acid, linoleic acid anilide, and arachidonic acid on the expression of adhesion molecules on human neutrophils.

The effects of linoleic acid, linoleic acid anilide, and arachidonic acid on the expression of CD11b/ CD18, CD11c/CD18 integrins and L-selectin on human neutrophils were studied by flow cytometry in a whole blood assay. None of these compounds had any effect on the basal expression of CD11b, CD11c, or L-selectin in the concentration range of 20-100 microM. However, linoleic acid at a concentration of 1000 microM slightly up-regulated CD11b and CD11c by a factor of 2.1 and 1.7, respectively. Linoleic acid, linoleic acid anilide, and arachidonic acid did not affect the formyl-methionyl-leucyl-phenylalanine induced up-regulation of CD11b or CD11c. However, linoleic acid and linoleic acid anilide slightly inhibited the phorbol myristate acetate (PMA)-induced expression of CD11b, which was decreased by 27 and 21% at concentrations of 100 and 1000 microM, respectively. Likewise, arachidonic acid at 40 microM inhibited the PMA-induced expression of CD11b by 19%. Our results suggest that linoleic acid, linoleic acid anilide, and arachidonic acid do not dramatically affect the expression of leukocyte adhesion molecules in a whole blood assay.

Regulation of expression of the LFA-1 and p150,95 leukocyte integrins: involvement of the CD11a and CD11c gene promoters.

Human Lymphocyte Associated Antigen-1 (LFA-1, CD11a/CD18, alpha L/beta 2) and p150,95 (CD11c/CD18, alpha X/beta 2) are cell surface alpha/beta heterodimers that, together with Mac-1 (CD11b/CD18, alpha M/beta 2) comprise the leukocyte-restricted beta 2 subfamily of integrins. LFA-1 is the only integrin expressed on all leukocyte lineages while p150,95 is exclusively expressed on cells of the myeloid lineage and on activated B lymphocytes and natural killer cells. The expression of the leukocyte integrins is regulated during cell activation and differentiation by transcriptional mechanisms. To dissect the molecular basis for the tissue-restricted and developmentally regulated expression of LFA-1 and p150,95, the promoter regions of their corresponding alpha subunits (CD11a and CD11c) were isolated and functionally characterized. Both promoters lack TATA and CAAT boxes, but exhibit initiator-like sequences at their major transcriptional start sites. Transient expression of CD11a- and CD11c-based reporter gene constructs have demonstrated the involvement of both promoters in the tissue-specific expression of LFA-1 and p150,95. Furthermore, a combination of DNAse I protection experiments and mobility band shift assays have revealed the existence of numerous DNA-protein interactions at the proximal region of both promoters, some of which overlap with consensus binding sequences for known transcription factors and correlate with the pattern of expression of both integrins.

Radiodetoxified endotoxin-induced tolerance alters monocyte but not neutrophil CD11b and CD18 expression in response to lipopolysaccharide.

OBJECTIVES: To test the hypothesis that pretreatment with radiodetoxified endotoxin (RDE) may mitigate the deleterious effects of subsequent infection, in part by modifying leukocyte adhesion receptor expression, and to investigate the cellular mechanisms of endotoxin tolerance induced by RDE. DESIGN: To assess the effect of RDE pretreatment on mortality from bacterial peritonitis, rats were implanted with an intraperitoneal, barium-fecal inoculum at intervals of 0, 1, 3, and 5 days after RDE injection. Experiments were then conducted to test the effect on leukocyte adhesion receptor expression. Two groups of mice received saline solution, and one group, RDE. After 72 hours, one group received saline solution (saline/saline group), the others, lipopolysaccharide (LPS) (saline/LPS and RDE/LPS groups). Peripheral leukocytes were obtained 1 hour after injection and were analyzed for CD11b and CD18 expression by flow cytometry. SETTING: Laboratory animal study. RESULTS: Survival rates were not improved in rats that were pretreated with RDE 0 and 24 hours before inoculum (0% and 7%, respectively). In rats that were pretreated 72 hours and 120 hours before inoculum, 47% (P < .01) and 60% (P < .01) survived, respectively. CD18 expression on polymorphonuclear leukocytes increased twofold in the RDE/LPS (mean +/- SEM, 300.3 +/- 32.9) and the saline/LPS (mean +/- SEM, 360.4 +/- 59.9) groups compared with controls (mean +/- SEM, 176.4 +/- 18.9) (P < .05). CD11b expression on polymorphonuclear leukocytes increased threefold in the RDE/LPS (mean +/- SEM, 91.3 +/- 8.1) and the saline/LPS (mean +/- SEM, 89.8 +/- 11.4) groups compared with controls (mean +/- SEM, 32.1 +/- 1.8) (P < .05). CD18 expression on monocytes decreased in the saline/LPS group (mean +/- SEM, 134.2 +/- 14.2) and was unchanged in the RDE/LPS group (mean +/- SEM, 200.2 +/- 17.2) compared with controls (mean +/- SEM, 217.6 +/- 16.5) (P < .05). CD11b expression on monocytes decreased in the saline/LPS group (mean +/- SEM, 25.8 +/- 2.2) and was unchanged in the RDE/LPS group (mean +/- SEM, 36.4 +/- 0.9) compared with controls (mean +/- SEM, 39.7 +/- 3.9) (P < .05). CONCLUSIONS: Radiodetoxified endotoxin reduces mortality rates from bacterial peritonitis when given at least 72 hours prior to a bacterial inoculum. Tolerance to subsequent LPS challenge is associated with an abrogation of the reduced peripheral monocyte CD11b and CD18 expression observed in native LPS-stimulated mice but is not associated with changes in polymorphonuclear leukocyte CD11b and CD18 expression. The mechanism of the observed RDE-induced monocyte hyporesponsiveness to LPS and its possible protective effect is uncertain and requires further investigation.

Salivary up-regulation of human polymorphonuclear leucocyte chemotaxis and adhesion-molecule expression.

Chemotaxis and adhesion molecules were investigated in peripheral blood neutrophils incubated with saliva collected from healthy donors. A salivary concentration of 20% increased chemotactic responses and CD11a/CD18 and CD11b/CD18 expression, but had no effect on formyl-methionyl-leucyl-phenylalanine receptors.

Leukocyte CD11b and CD18 expression are increased in blood salvaged for autotransfusion.

To investigate whether activated leukocytes are present in salvaged blood, we measured complete blood counts and quantified the surface expression of the leukocyte adhesion receptors CD11b and CD18 in salvaged blood and arterial blood from six male patients undergoing elective abdominal aortic aneurysm repair. Salvaged blood contained 5,450 +/- 1010 leukocytes/microL and 7600 +/- 6200 platelets/microL and had a hematocrit of 50.6 +/- 3.7%. CD 11b expression was 3.3 +/- 0.5 fold higher on neutrophils and 3.8 +/- 1.0 fold higher on monocytes from salvaged blood compared with arterial blood (P < 0.05 for both). CD18 expression was increased 3.2 +/- 0.2 fold on neutrophils and 2.5 +/- 0.4 fold on lymphocytes (P < 0.05) in salvaged compared to arterial blood (P < .05). Monocyte expression of CD18 was increased 4.50 +/- 1.1 fold in salvaged blood, but this difference was not statistically significant. We conclude that a substantial number of activated leukocytes are present in salvaged blood. Because activated leukocytes could potentially be detrimental to the recipient, our findings raise theoretical concerns about the use of salvaged blood and emphasize the need for further refinement of the procedure.

Induction of intercellular adhesion molecule-1 by lipopolysaccharide in canine alveolar macrophages.

Previous studies have demonstrated that alveolar macrophages (AMphis) adherent to plastic release interleukin-8 in response to lipopolysaccharide (LPS). We sought to determine whether LPS could also alter surface adhesion molecule expression and thus modulate additional AMphi adhesive interactions. Canine AMphis obtained by bronchoalveolar lavage of excised lung were adhered onto tissue culture plastic and exposed to LPS (0.01 to 100 ng/ml). Expression of beta 2 integrins and intercellular adhesion molecule-1 (ICAM-1) was subsequently determined by flow cytometry, a cDNA probe to canine ICAM-1 was used to quantify ICAM-1 mRNA, and changes in adhesion molecule function were assessed by evaluating the extent of homotypic aggregation. ICAM-1 and CD11a/CD18 were present on freshly isolated AMphis. CD11b/CD18 and CD11c/CD18 were expressed at lower levels. Nonadherent AMphis expressed a pattern of beta 2 integrin and ICAM-1 comparable to adherent cells. During short-term LPS stimulation (3 h), adherent AMphis increased both the synthesis and expression of ICAM-1. CD18 expression was either decreased or remained unchanged with LPS stimulation. LPS stimulation in vitro (> 0.01 ng/ml) enhanced the homotypic aggregation of adherent AMphis. Aggregation was blocked by monoclonal antibodies to ICAM-1 (CL18/6) and CD11a (R7.1) and CD18 (R15.7). Similar kinetics were found for expression of ICAM-1 and homotypic aggregation, suggesting that up-regulation of ICAM-1 is a major determinant of the LPS-stimulated aggregation of AMphis.

Bovine leukocyte adhesion deficiency: in vitro assessment of neutrophil function and leukocyte integrin expression.

Bovine leukocyte adhesion deficiency (BLAD) was identified in a two-month-old Holstein heifer calf using DNA-polymerase chain reaction analysis of the affected calf and other clinical parameters. Neutrophil integrin expression (CD18, CD11a, CD11c), aggregation, and transendothelial migration were studied in vitro. Neutrophils were isolated from the affected calf and from normal, healthy, age-matched control Holstein calves. Neutrophils isolated from the affected BLAD calf had decreased expression of leukocyte integrins on their cell surface, decreased ability to aggregate in response to chemotactic stimuli, and decreased ability to migrate across bovine endothelial cell monolayers in vitro. Transendothelial migration of neutrophils from normal calves was reduced to levels comparable to the BLAD neutrophils by treatment with an anti-CD18 monoclonal antibody (MAb 60.3). Peripheral-blood lymphocytes from the BLAD calf also expressed negligible levels of leukocyte integrins, similar to their neutrophil counterparts. Our experimental findings in vitro correlate well with the clinical observations of decreased leukocyte trafficking and diminished host defense in leukocyte adhesion-deficient animals. The syndrome of BLAD may be a suitable model for one of the human leukocyte adhesion deficiency disorders.

IL-4 inhibits binding and cytotoxicity of NK cells to vascular endothelium.

We investigated the influence of IL-4 on the interaction between Natural Killer (NK) cells and vascular endothelial cells (EC). Pretreatment of NK cells with IL-4 inhibited the adhesion of NK cells on resting or IL-1-activated EC. The inhibitory action of IL-4 was observed on both unstimulated NK cells as well as on cells concomitantly activated with IL-2 or with phorbol ester. IL-4 also inhibited the cytotoxicity of IL-2 activated NK cells on EC. Binding of NK cells to vascular EC involves the LFA1/ICAM-1 and 2, and VLA-4/VCAM-1 adhesion pathway. Anti-CD18 mAb had a lower inhibitory effect in IL-4-treated NK cells compared to control. The levels of inhibition with resting vs IL-1-activated EC, as well as antibody blocking experiments, suggested that IL-4 exerts an inhibitory influence predominantly on the beta 2-integrin (CD18)-dependent adhesion pathway; nevertheless IL-4 did not affect the expression of CD18/CD11a and VLA-4 on the NK cell surface, as assessed by flow cytometry. NK cells are potent producers of IFN-gamma and there is evidence that they play a role in orienting immunity to the TH1-type responses. The inhibition by IL-4 of NK cell binding to EC and subsequent recruitment and activation may thus contribute to development of TH2 responses.

Remodeling of neutrophil phospholipids with 15(S)-hydroxyeicosatetraenoic acid inhibits leukotriene B4-induced neutrophil migration across endothelium.

5-Lipoxygenase products, such as leukotrienes, are important stimuli for leukocyte-mediated tissue injury in acute inflammation. 15-Hydroxyeicosatetraenoic acid (15-HETE) is an eicosanoid generated by a variety of cell types via the actions of 15-lipoxygenases and, in addition, cyclooxygenases and epoxygenases. 15-HETE levels are frequently elevated at sites of inflammation, and extracellular 15(S)-HETE is esterified rapidly into neutrophil (PMN) phospholipids in vitro to levels that are comparable with arachidonic acid. We present evidence that remodeling of PMN phospholipids with 15(S)-HETE stereoselectively inhibits PMN migration across endothelium in response to leukotriene B4 (LTB4) and other chemoattractants. Esterified 15(S)-HETE causes a striking reduction in the affinity of LTB4 cell-surface receptors for their ligand and inhibition of LTB4-triggered stimulus-response coupling. As a result of these actions, esterified 15(S)-HETE attenuates the cytoskeletal rearrangements and CD11/CD18-mediated adhesive events that subserve directed locomotion of PMN across endothelium. These observations indicate that products of the 5-lipoxygenase and 15-lipoxygenase pathways can exert counterbalancing influences on PMN trafficking across endothelium. They suggest that 15(S)-HETE may be a potent endogenous inhibitor of PMN-endothelial interactions in vivo and serve to limit or reverse acute inflammation.

Bovine leukocyte adhesion deficiency in Holstein cattle.

Two Holstein heifers with persistent and recurrent infections including ulcerative gingivitis, periodontitis, pneumonia, loss of teeth and stunted growth associated with marked neutrophilia were evaluated clinically and for neutrophil function, CD18 expression on neutrophils and CD18 genotype analysis by DNA-polymerase chain reaction (PCR) test. Adherence to nylon fibers and phagocytic activity of neutrophils from affected animals were significantly (p < 0.05) impaired as compared with those of controls. Neutrophils from affected heifers had decreased chemiluminescent (CL) responses when stimulated with opsonized zymosan, compared with those of controls. In contrast, neutrophils from affected heifers produced increased CL responses when stimulated with latex beads and phorbol myristate acetate compared with those of controls. The clinical findings, functional leukocyte abnormalities, deficiency in expression of CD18 on neutrophils, and the D128G mutation detected by DNA-PCR testing of affected heifers demonstrated that these heifers have bovine leukocyte adhesion deficiency (BLAD). Although both animals were confirmed to be homozygotes for BLAD by DNA-PCR test, they had differences in clinical, hematological and neutrophil functional characteristics.

The expression of CD11/CD18 molecules on rabbit leucocytes: identification of monoclonal antibodies to CD18 and their effect on cellular adhesion processes.

The monoclonal antibodies, L13/64 and RCN1/21, raised against rabbit leucocytes, have been shown, by sequential immunoprecipitation and immunoblotting, to react with the rabbit CD18 molecule. They recognise not only surface-expressed CD18 but also an intracellular form which appears to be partially glycosylated. The expression of the CD11 and CD18 glycoproteins on a wide variety of rabbit leucocyte populations has been investigated by flow cytometry, using these two monoclonal antibodies (Mabs), together with others which recognise human CD11 and CD18 proteins but cross-react with rabbit tissues. The distribution of these leucocyte integrin molecules has been shown to be similar to that observed in humans and determination of the N-terminal sequence of rabbit CD11b shows strong homology with human and mouse sequences. Of four anti-rabbit CD18 Mabs tested, only one, L13/64, has been shown to be capable of inhibiting the adhesion of fMLP-stimulated neutrophils to gelatin coated plastic and the homotypic aggregation of PMA-stimulated T cells, both of which assays have been shown to be CD18-dependent. RCN1/21 causes aggregation of unstimulated neutrophils, but it is not known whether this is due to cellular activation or agglutination.

Expression of human CD18 in murine granulocytes and improved efficiency for infection of deficient human lymphoblasts.

The CD18 gene encodes the beta 2-subunit of leukocyte integrins, and mutations in this gene cause extreme host susceptibility to bacterial and fungal infection. Because expression of CD18 is restricted to bone marrow-derived cells, this disorder is considered an excellent candidate for somatic gene therapy utilizing ex vivo infection of bone marrow stem cells. We have constructed a retroviral vector expressing CD18 with the Moloney murine leukemia virus (Mo-MLV) long terminal repeat (LTR) as the promoter, and high-titer ecotropic and amphotropic producer cell lines were isolated using the GP+E-86 and GP+envAM12 safe packaging cell lines. Infection of CD18-deficient lymphoblasts resulted both in expression of immunodetectable CD18 at 35-40% of normal levels on 55-60% of cells and in functional restoration of CD18-dependent aggregation. All of 16 mice transplanted with syngeneic bone marrow infected with the CD18 retrovirus expressed human CD18 on 17-36% of granulocytes at 2 weeks after transplantation, and expression was appropriately up-regulated in response to stimulation with zymosan-activated serum. This recombinant retrovirus should prove useful for further studies of somatic gene therapy for CD18 deficiency.

Modulation, in vivo and in vitro, of surface expression of CD18 by bovine neutrophils.

A series of experiments was designed to elucidate some of the factors that may influence surface expression of CD18 by bovine neutrophils. Expression of CD18 was determined by immunofluorescence flow cytometry. Neutrophils recovered from the uterus of cows (n = 9) after intrauterine administration of sterile oyster glycogen solution expressed (mean +/- SD) 123 +/- 21% more CD18 than did circulating neutrophils recovered simultaneously from the same cows (P = 0.003). In 8 cows given 20 mg of dexamethasone IM daily for 3 days, expression of CD18 on blood neutrophils was 29.6 +/- 8% less after treatment than before treatment (P = 0.0078). Neutrophils from 12 cows or bulls exposed to phorbol myristate acetate in vitro increased expression of CD18 by 137 +/- 37% (P = 0.0035). Likewise, exposure of neutrophils from 8 cattle to zymosan-activated bovine plasma increased CD18 exposure by 10.6 +/- 3.8% (P = 0.029). These findings indicate that expression of CD18 by bovine neutrophils is a dynamic system, capable of responding to inflammatory stimuli. Inadvertent activation of neutrophils may be responsible for some of the variance in expression observed when examining large groups of cattle for CD18 expression by neutrophils. The ability of bovine neutrophils to respond rapidly to various stimuli by increasing surface expression of CD18 indicates that a pool of intracellular CD18 may be available for inclusion in the plasmalemma, as has been reported for human neutrophils.

Expression of leucocyte adhesion molecules at the human blood-brain barrier (BBB).

The expression of leucocyte adhesion molecules was studied on cerebral endothelia by immunocytochemistry. In peritumoral "normal" brain tissue we found low endothelial expression of ICAM1, LFA3, CD44, and CD9, whereas VLA1 was present on vessels in high incidence and density. LFA1, CD2, and CR3 were found on intraluminal and parenchymal leucocytes, but were absent on brain vessels. In brain tumors and inflammatory brain lesions, we observed an up-regulation of endothelial ICAM1 and LFA3 expression, whereas other adhesion molecules on endothelial cells remained unchanged. Within the brain parenchyma, ICAM1 and LFA3 were found on astrocytes and tumor cells; on the contrary, LFA1 was expressed on microglial cells similar to CR3. CD44 and CD9 showed a diffuse neuropil expression in normal and tumoral tissue, whereas VLA1 was not expressed on any parenchymal cells. Our data show that multiple different adhesion molecules are present on blood-brain barrier endothelium (BBB) under normal conditions and some adhesion molecules are up-regulated in brain tumors and under inflammatory conditions. The presence of adhesion molecules in the vessel walls as well as on parenchymal cells like astrocytes and microglia may guide inflammatory cells into and through the brain in the course of immune surveillance and inflammation.