INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE SERUM PROTEIN OROSOMUCOID (₁-acid glycoprotein) is known to increase in concentration during, for instance, inflammatory conditions and cancer (25). It is also known to be essential for capillary permselectivity in several different capillary beds (6, 12, 14, 30). The data by Curry et al. (6) indicate that orosomucoid exerts its action by reinforcing the charge barrier, as Haraldsson and Rippe (12) originally predicted. The main source of orosomucoid is the liver, but synthesis of orosomucoid has been reported for leukocytes (2) and human breast epithelium (10). Our group has also recently shown that microvascular endothelial cells have a constituent synthesis of orosomucoid (28). During inflammation, synthesis of orosomucoid can be detected in the lung type II alveolar epithelial cells (5) and in the kidney (17). In addition, there are binding sites for orosomucoid at the surface of endothelial cells (26).

The Cytosensor microphysiometer technique (15, 19) can be used to study cells under well-controlled conditions. The method allows on-line recordings of proton excretion of cells (see Fig. 1). The proton excretion or extracellular acidification rate (ECAR) is a result of increased cellular metabolism activated by receptor-ligand interaction mediated by second messenger pathways. This increase in cellular metabolism gives rise to an increase in excretion of carbon dioxide and lactic acid that are transported out of the cells, resulting in an increased extracellular acidification. There are several important transporters of protons, one of which is the Na⁺/H⁺ exchanger, which is responsible for excreting protons when receptor activation takes place. Small changes (milli-pH unit/min) in proton concentration in the surrounding medium can be measured precisely by the microphysiometer assay system (23). Increased ECAR has been analyzed for a large number of different cell types and ligands. Among the many cytokines and inflammatory mediators that have been tested in the microphysiometer are interleukin-2, interleukin-4, interferon- (19), histamine (11), and bradykinin (8). Receptor activation of both native and transfected muscarinic, adrenergic, and dopaminergic receptors have also been extensively studied (19). The microphysiometer technique has been tested successfully on most types of classic second messenger pathways in cells, and data collected from such experiments compare excellently with data from more traditional assays (15). The cells used can be both transfected cell lines and primary cell cultures. The transfected cells have the advantage of responding with a stronger signal because the transfected receptors are so abundant. Primary cell cultures also perform well in the system, but a weaker signal is to be expected.

In the present study, we wanted to investigate whether orosomucoid had a direct effect on endothelial cells and, if so, through what mechanism. In addition, the modulatory effects of orosomucoid in inflammatory responses in endothelial cells were studied.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Preparation of cells and cell culture. Endothelial cells were obtained from human umbilical cord veins by a modification of the method of Jaffe et al. (16). The umbilical cord was recovered from the placenta, placed in a sterile 50-ml tube, and kept at 4°C until processing (kindly provided by Dr. T. Jansson and Dr. T. Powell). The umbilical vein was then cannulated with connecting tubes, and the tubes were secured by clamping the cord over the tubes with cable ties connected to a three-way stopcock. The vein was perfused free of blood using 0.01 M PBS. Prewarmed collagenase (C-0130; Sigma, St. Louis, MO) diluted to 0.1% in Hanks' balanced salt solution (Sigma) was infused into the umbilical vein, and the tubing was shut at the ends. The umbilical cord was placed in a bowl with PBS and incubated at 37°C for 10-15 min. After incubation, the collagenase solution containing the endothelial cells was flushed out by using 25 ml of medium M-131 (Cascade Biologics, Portland, OR) with 20% FCS (Harlan Sera-Lab, Loughborough, UK). The effluent was collected in a sterile 50-ml tube. The cells were sedimented at 170 g for 5 min, and the cell pellet was resuspended in 7 ml of fresh culture medium [M-131 containing 2% low-serum growth supplement and penicillin, streptomycin, and amphotericin B (Cascade Biologics)]. The cell suspension was transferred to a gelatin-coated 25-cm² flask (Costar, Cambridge, MA) and cultured in a humidified, 7% CO₂ atmosphere at 37°C. Passaging of cells was performed by washing them once in PBS, followed by 4-5 min of trypsination with 0.125% trypsin (art. 043-0509H, Life Technologies, Täby, Sweden). Two days before acidification rate measurements were taken, cells were plated onto the membrane of Transwell cell capsules (Molecular Devices, Sunnyvale, CA) with a density of 50,000-80,000 cells per capsule. Cells were starved 24 h before the experiment, using 10 times lower serum concentrations than normal (0.2%). Cells were used at passages 2-6.

Cell characterization. Immunohistochemistry was performed to determine the origin of the cells. We used a monoclonal antibody (MAb) against von Willebrand factor (factor VIII; Sigma) and also, in some experiments, an antibody against platelet endothelial cell adhesion molecule 1 (PECAM-1; DAKO, Glostrup, Denmark). The cells were tested for contamination of smooth muscle cells using an MAb against smooth muscle -actin (clone 1A4; Sigma). In addition, the cells were tested for mycoplasma infection.

Test substances. Substances used for the experiments were as follows: highly purified orosomucoid (a generous gift from Immuno, Vienna, Austria) at a concentration of 0.01 g/l and histamine (Sigma) at a concentration of 1 µM. Membrane-soluble analogs to cAMP and cGMP (8-bromo-cAMP and 8-bromo-cGMP; Sigma) were used at a concentration of 0.5 mM each. 2,4-Dinitrophenol was used at a concentration of 1 mM. The test substances were suspended in the same medium as the one running in the system (Ham's F-12), and the pH was adjusted to 7.35.

Microphysiometry. Cells were placed into the sensor chambers of the microphysiometer (Cytosensor, Molecular Devices, Sunnyvale, CA). Culture medium was replaced by low-buffered, serum-free, bicarbonate-free Ham's F-12 medium, pH 7.35 (ICN, Costa Mesa, CA). The medium was flowing at 100 µl/min through the chamber with the cell capsules. The temperature of each sensor chamber was held at 37°C. The Cytosensor system uses a silicon-based, light-addressable potentiomeric sensor (LAPS) to continuously monitor minute changes in extracellular pH (23). A voltage signal proportional to pH was measured and recorded every second. To determine the acidification rate, the pump cycle was on for 80 s and off for 40 s, thus allowing accumulation of extracellular acidification metabolites produced by the cells as a result of cellular metabolism. Flow was then resumed and the acid flushed out of the chamber. The flow cycle was repeated continually, yielding one data point of the acidification rate (ECAR) at each flow-off period. Basal acidification rates were monitored for at least 50 min before the first application of test substance occurred. Data were recorded on-line on a Macintosh PowerPC computer using Cytosoft software, where 1 µV/s is equal to 1 × 10³ pH units/min. Acidification rates were expressed as percent change of the baseline activity before the administration of various stimuli (19) (see Fig. 1, A and B).

View larger version (1923K): [in this window] [in a new window]   Fig. 1.   Representative experiment. A: extracellular acidification rates (ECAR) measured as a percentage of baseline. Four chambers (represented by different symbols) of human umbilical vein endothelial cells (HUVEC) were exposed to orosomucoid (0.01 g/l) at time (t) = 11 min (arrow) for 50 s. Increase in acidification rate was +11 to +15% compared with baseline. B: raw data between 8 and 16 min (indicated by vertical lines in A) representing local pH measured in mV. To measure cellular acidification rate, we periodically halted fluid flow (pump off for 40 s and then on for 80 s), thus allowing a measurable buildup of acid metabolites produced by cells. It is important to notice that no pH shifts occurred when orosomucoid was introduced to the system.

Control experiments. As controls, we tested inflammatory agents such as bradykinin, substance P, thrombin and histamine on human endothelial cells of umbilical vein (HUVEC) or dermal origin (HMVEC). In addition, we tested the cell response with the mitochondrial uncoupler 2,4-dinitrophenol. Bradykinin did not produce any effect on the cells. The other agonists did have an effect; histamine gave the most reproducible effect and was therefore chosen for further studies.

Experimental protocol. After ~50-min measurements of basal acidification rate, the application of test substances commenced. At this point the cells exhibited a stable baseline. No pH shifts occurred when substances were applied. The cells were exposed to the test substances for 50 s, unless otherwise stated, and then to a 20-min wash period for washout of the active substance and to avoid desensitization. All experiments were ended with a control measurement of either histamine or 2,4-dinitrophenol to determine the cellular response. Each cell chamber had its own separate fluid path and was regarded as n = 1.

Statistics. Student's paired design t-test analysis was performed, and all measurements were conducted with the cells in each chamber serving as their own control.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Preparation of cells and cell culture. Most umbilical cord preparations yielded nice endothelial cell cultures; problems with contaminating smooth muscle cells were predominantly seen when the umbilical cord had been damaged in some way. Only clones of cells free from other contaminating cell types were used.

Characterization of endothelial cells. Staining with von Willebrand and PECAM-1 antibodies showed expression of these endothelial cell-specific markers. The cell cultures used in the experiments were negative for anti--smooth muscle actin and were free from mycoplasma.

Microphysiometer experiments. As an inclusion criterion, a positive response to histamine or 2, 4-dinitrophenol was required at the end of the experiment. Studies in which there was no significant increase in ECAR with these drugs were excluded.

Effects of histamine and orosomucoid. Histamine induced an increase in metabolic rate in a dose-dependent manner in HUVEC (Fig. 2). Thus stimulating the HUVEC for 50 s with histamine (1 µM solution) gave rise to an increased metabolic response (+19 ± 2%, n = 22, P < 0.001). The response rapidly diminished on prolonged stimulation and vanished after 4 min (see Fig. 3). The metabolic activity of HUVEC increased significantly in the presence of 0.01 g/l orosomucoid (+14 ± 1%, n = 25, P < 0.001) (Fig. 3). Tachyphylaxia was also observed for this protein, and exposing the cells to 0.01 g/l orosomucoid resulted in an acidification response that peaked rapidly and then declined back to baseline within 4 min (Fig. 3). Pretreatment with orosomucoid for 6 min, followed by stimulation with histamine in the presence of orosomucoid, resulted in a significant decrease of the response to histamine (+6 ± 2%, n = 25, P < 0.001) (Fig. 4). The anti-inflammatory effect of orosomucoid was reversible because stimulation with histamine after 20 min of orosomucoid-free perfusion gave a response of +14 ± 1%.

View larger version (10K): [in this window] [in a new window]   Fig. 2.   Dose-response curve for histamine. HUVEC were stimulated with histamine at concentrations of 1 (n = 8), 0.1 (n = 8), or 0.01 µM (n = 7) for 50 s. Stimulation was followed by a 20-min perfusion with agonist-free media, which is the time required to fully restore histamine response after a period of desensitization. Another concentration of histamine could then be introduced.

View larger version (10K): [in this window] [in a new window]   Fig. 3.   Effects of histamine and orosomucoid, followed by desensitization. HUVEC were continuously exposed to either orosomucoid (0.01 g/l, n = 25) or histamine (1 µM, n = 22). Both substances gave rise to potent activation of endothelial cells. A pronounced desensitization was seen in a similar manner for both substances, with negligible responses after 4 min of stimulation. ***P < 0.001.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Attenuating effect of orosomucoid on histamine response. ECAR increased significantly when cells were exposed to 1 µM histamine. Pretreatment with orosomucoid (0.01 g/l) for 6 min resulted in a significantly lower ECAR in response to 1 µM histamine in presence of orosomucoid (n = 25). Exposure time was 50 s, and washout time was 20 min between each exposure. ***P < 0.001.

Effects of cAMP and cGMP. Another set of experiments was performed to find the possible second messenger pathway mediating the effect of orosomucoid. Cells were first shown to give rise to an increase in ECAR when orosomucoid was applied (+20 ± 2%, n = 18) (Fig. 5). The cells were then pretreated with cAMP (0.5 mM 8-bromo-cAMP) for 20 min, giving an ECAR of +19 ± 2%. The effect of cAMP was transient, and the cell signal returned to baseline after 3 min and thereafter remained stable. Orosomucoid was then applied in the presence of cAMP, and the ECAR was +5 ± 1%, which was significantly less than the ECAR without 8-bromo-cAMP (n = 16, P < 0.001). The cells were then perfused with medium free from orosomucoid and cAMP for 20 min. A new stimulation with orosomucoid gave rise to a response almost as high as before cAMP (+17 ± 2%) (Fig. 5). A similar protocol was used for cGMP (0.5 mM 8-bromo-cGMP), which did not seem to affect the signal transduction of the orosomucoid response in HUVEC (Fig. 6). Pretreatment with cAMP or cGMP using similar protocols abolished the effect of histamine on the endothelial cells. Thus, after pretreatment with cAMP, the cellular response to histamine was 10 ± 0% (n = 2), whereas cGMP gave a histamine response of 8 ± 4% (n = 6).

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Involvement of cAMP in orosomucoid response. HUVEC responded with a significant rise in ECAR to an initial exposure of 0.01 g/l orosomucoid. Application of 0.5 mM 8-bromo-cAMP induced a similar increase in ECAR. ECAR returned to baseline and remained stable within 20 min of prolonged exposure to 8-bromo-cAMP. A second application of orosomucoid in the presence of 8-bromo-cAMP induced a significantly lower increase in ECAR (n = 16). After 20 min of agonist-free perfusion, a final stimulation with orosomucoid resulted in ECAR almost as high as that for the first orosomucoid exposure. ***P < 0.001.

View larger version (22K): [in this window] [in a new window]   Fig. 6.   Response to cGMP. HUVEC were stimulated with orosomucoid (0.01 g/l), 8-bromo-cGMP (0.5 mM), and finally with orosomucoid plus 8-bromo-cGMP using a protocol similar to that described in Fig. 5. No significant alteration of ECAR was seen (n = 6).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This is the first report demonstrating effects of the plasma protein orosomucoid on endothelial cells. We have studied HUVEC. The cellular response to orosomucoid was rapid and could be reduced by pretreatment with 8-bromo-cAMP, indicating receptor activation. The technique used, microphysiometry, has been well established for studies of other cells (15, 19, 24). It is based on the principle that receptor activation will induce an increase in metabolic activity and, hence, an increase in ECAR. Moreover, our results strongly suggest that cAMP is involved in the cellular signaling pathway for orosomucoid because the cellular response to orosomucoid was diminished by pretreatment with the soluble cAMP analog 8-bromo-cAMP for 20 min. In contrast, cGMP does not seem to be involved in the anti-inflammatory effect of orosomucoid.

The present observation that orosomucoid induces a cAMP-dependent signal in the endothelial cells supports and extends the findings made by Schnitzer and Pinney (26) of receptor-like binding sites in bovine lung microvascular endothelial cells. These authors demonstrated that the binding rapidly reached equilibrium, was independent of calcium, and was rapidly reversible. Moreover, binding sites for orosomucoid were not found in other vascular wall associated cells such as fibroblasts or smooth muscle cells. Orosomucoid binding has also been reported in the rat prostate gland (22) and macrophages (4). Unfortunately, an orosomucoid receptor has not yet been characterized or cloned.

The second important and new finding in this paper is that orosomucoid has an anti-inflammatory action directly on the endothelial cells. Thus the response of the cells to histamine was markedly attenuated after a 6-min pretreatment with orosomucoid (see Fig. 4). Indeed, such an effect is to be expected, because we found that cAMP is involved in the action of orosomucoid. Several previous studies have shown that increased intracellular cAMP can block inflammatory response both in vivo (27) and in cell monolayers (1). It also has been reported that cAMP has effects on the basal permeability of certain capillaries. Thus increased concentrations of cAMP reduce the microvascular permeability in chorioallantoic membranes of chick embryos (7), in cultured endothelial cell populations (29), and in intact microvessels of frog mesentery (1). There is one previous report of microphysiometry data on a primary culture of endothelial cells. In that paper the effect of 1 µM histamine was studied on the same cell type as in the present study, HUVEC, giving an ECAR of +30% (11).

Histamine, thrombin, bradykinin, and other proinflammatory substances act on endothelial cells in a complex manner. After activation of specific receptors on the endothelial plasmalemma, a flux of calcium seems to be induced by receptor-operated channels (29). Indeed, in a previous study on isolated perfused rat hindquarters, we demonstrated that the inflammatory reaction is dependent on the extracellular concentration of calcium (13). Thus reducing the extracellular concentration of calcium to 0.1 mM completely prevents the effect of histamine. Inhibition of voltage-operated channels (1 × 10⁶ M felodipin) did not affect the histamine response, and neither did extracellular depletion of magnesium (13). The rise of intracellular calcium induces activation of myosin light-chain kinase and actin-myosin interaction, giving rise to endothelial contraction. Hereby, interendothelial "leaks" or "large pores" are created, allowing plasma proteins to be filtered from blood to interstitium. The elevated intracellular calcium also seems to induce a rise in nitric oxide (NO) concentration, which in turn will elevate cGMP concentrations acting through a negative feedback loop. Thus cGMP reduces the calcium influx and increases cAMP, thereby inhibiting the histamine response. Indeed, NO has been suggested to interfere with the inflammatory reaction in a variety of conditions (3, 9).

Muchitsch et al. (21) have recently shown that orosomucoid decreases capillary leakage in inflammatory conditions evoked by several inflammatory agents such as histamine, thrombin, and platelet-activating factor in guinea pigs in vivo. Suggestions have been made that the elevated concentration of orosomucoid during inflammation could act as a negative feedback, limiting the extent of the damage caused by the inflammatory process. A local expression of orosomucoid could thus act to protect inflamed tissue locally (5). Thus Libert et al. (18) have shown that mice are significantly protected from lethal shock induced by tumor necrosis factor- or bacterial endotoxin (lipopolysaccharide) when orosomucoid is given as a bolus injection (resulting in a serum concentration of 1.5 mg/ml). In a recent study, orosomucoid increased the survival rate after septic peritonitis in rats. However, the protection did not affect the outcome of several other shock models in mice and rats (20). Interestingly, orosomucoid seems to be beneficial for rats with puromycin aminonucleoside (PAN)-induced nephrotic syndrome. In that study (20), treatment with 600 mg · kg¹ · day¹ orosomucoid at days 6-9 after injection of PAN resulted in reduced proteinuria and lower plasma creatinine values. These observations of anti-inflammatory actions of orosomucoid fit well with the present finding that the protein affects the endothelial cells per se and diminishes the effect of histamine.

We conclude that orosomucoid affects endothelial cells through a cAMP-dependent mechanism, indicating activation of a receptor. This could be, in part, an internal local control mechanism, because the microvascular endothelial cells have been shown to produce orosomucoid themselves (28). However, orosomucoid elicits an anti-inflammatory effect directly on the endothelial cells, reducing the response to histamine. Thus orosomucoid is required for maintenance of a high capillary permselectivity, and the protein also has an anti-inflammatory effect directly on the endothelial cells.