Secondary Logo

Journal Logo

Effect of aprotinin onin vitrocerebral endothelial ICAM-1 expression induced by astrocyte-conditioned medium

Harmon, D.*; Ghori, K.; Lan, W.; Shorten, G.

European Journal of Anaesthesiology: April 2005 - Volume 22 - Issue 4 - p 277–282
doi: 10.1017/S0265021505000463
Original Article

Background and objective: Aprotinin administration may decrease the incidence of stroke associated with coronary artery bypass surgery by an unknown mechanism. Astrocytes exposed to hypoxia produce proinflammatory cytokines and upregulate intercellular adhesion molecule (ICAM)-1 on cerebral endothelium. This study investigated the effects of aprotinin on cerebral endothelial activation by hypoxic astrocytes in vitro.

Methods: Mouse astrocytes were exposed to hypoxia in an anaerobic chamber for 4 h followed by reoxygenation for 24 h. Astrocyte-conditioned medium (ACM) collected from mouse astrocytes subjected to hypoxia/reoxygenation (HR) or normoxia were applied to mouse cerebral endothelial cell (MCEC) cultures for 4 and 24 h in normoxia. Endothelial cells were preincubated for 1 h with aprotinin (1600 KIU mL−1) prior to exposure to ACM. Flow cytometry was used to estimate endothelial ICAM-1 expression. Interleukin (IL)-1β space concentrations in ACM were estimated with enzyme-linked immunosorbent assay (ELISA). Repeated comparisons were made using analysis of variance (ANOVA) and post hoc Tukey test as appropriate. P < 0.05 was considered significant. Data is presented as mean (standard deviation, SD).

Results: MCEC ICAM-1 expression was greater after 24 h exposure to HR-ACM compared to normoxic-ACM (mean channel flouresence (MCF) 107.5 (4.5) vs. 74.3 (4.5), respectively, P < 0.001). ICAM-1 expression was decreased by aprotinin preincubation compared to control (MCF 91.0 (1.1) vs. 107.5 (2.1), P = 0.006). Supernatant IL-1β concentrations in astrocytes exposed to HR were greater than those exposed to normoxia (7.1 (0.2) vs. 4.1 (0.2), P = 0.01).

Conclusions: This may be a neuroprotective mechanism associated with aprotinin administration.

*Walton Centre for Neurology and Neurosurgery, Liverpool, UK

Cork University Hospital and University College Cork, Department of Anaesthesia and Intensive Care Medicine, Cork, Ireland

Cork University Hospital and University College Cork, Department of Surgery, Cork, Ireland

Correspondence to: Dominic Harmon, Walton Centre for Neurology and Neurosurgery, Lower Lane, Fazakerley, Liverpool L9 7LJ, UK. E-mail:; Tel: +353 21 4546400 ext 22566; Fax: +353 21 4546434

Accepted for publication January 2004

Aprotinin is a serine protease inhibitor derived from bovine lung, which can decrease blood loss during and after cardiac surgery [1]. In a post hoc analysis of 816 coronary artery bypass grafting (CABG) patients from a multicentre study, aprotinin administration was associated with a significantly (P = 0.04) decreased incidence of stroke (3.1% vs. 0.0%) [1]. A meta-analysis of placebo-controlled, randomized, double-blind studies of CABG patients receiving high-dose aprotinin or placebo has supported the hypothesis of a cerebroprotective effect of aprotinin administration, a reported stroke incidence of 4.2% vs. 0% [2]. Although the mechanism by which aprotinin confers neuroprotection is not known, an anti-inflammatory effect can be postulated.

An acute inflammatory response with influx of neutrophils exacerbates tissue injury associated with acute cerebral ischaemia [3,4]. This response is more intense following temporary (rather than permanent) middle cerebral artery occlusion [5,6]. After temporary occlusion of the middle cerebral artery in rats, inhibition of neutrophil or neutrophil-endothelial cell interactions decreased infarct size [7]. During ischaemia-reperfusion, inflammatory endothelial ICAM-1, E-selectin, and interleukin (IL)-8 are transcribed, resulting in an ‘activated’ endothelial cell [8]. The activated endothelial cell upregulates the surface expression of cell adhesion molecules and secretes cytokines, which act to sequester neutrophils in the ischaemic zone. Several therapeutic interventions have been explored to suppress post-ischaemic inflammatory response in attempts to decrease secondary injury following cerebral ischaemia-reperfusion [9].

Exposure of cerebral microvascular endothelial cells in vitro to hypoxia/reoxygenation (HR) ‘models’ ischaemia-reperfusion to the brain endothelium [10]. Cerebral endothelial cells (CECs) are similar to those of other organs in many respects [11]. In the brain, however, the endothelium is ensheathed by fine processes of astrocytes, which, in culture, produce marked changes in several endothelial parameters [11]. Astrocytes subjected in vitro to hypoxia and subsequent reoxygenation (HR) activate endothelium, including upregulation of intercellular adhesion molecule (ICAM)-1 and IL-1β [12].

Primary CECs from isolated microvessels in co-culture are treated with astrocyte-conditioned medium (ACM) offers an in vitro system for examination of the blood-brain barrier [11]. One limitation of this model is the contamination from non-endothelial cell types [11]. An alternative approach is to use immortalized cell lines [13]. Cloned, immortalized mouse brain endothelial cells (bEnd3) are phenotypically CECs and suitable for the study of astrocyte/endothelial cell interactions [14].

In this study, the effects of aprotinin on cerebral endothelial ICAM-1 expression induced by HR-ACM was investigated. This is particularly relevant in understanding potential neuroprotective effects associated with aprotinin administration during cardiac surgery.

Back to Top | Article Outline


Cell cultures

Mouse brain endothelial cell line bEnd3 (mouse cerebral endothelial cells, MCECs) (ATCC CRL-2299; Rockville, USA) were cultured in RPMI-1640 medium with 10% fetal calf serum, 100 U mL−1 penicillin G and 100 μg mL−1 streptomycin in an incubator with 37°C, 100% humidity, and 5% carbon dioxide (CO2). MCECs were seeded into 96-well culture plates at a cell concentration of 1 × 105 cells/well. Mouse type 1 astrocytes (ATCC CRL-2541; Rockville, USA) were cultured in Dulbecco's modified Eagle's medium with 4 mmol L-glutamine adjusted to contain 1.5 g L−1 sodium bicarbonate and 4.5 g L−1 glucose, 90%; fetal bovine serum, 10% at 37°C. All cell growth media were purchased from Sigma Aldrich, Poole, Dorset, UK.

Back to Top | Article Outline

Preparation of ACM

Mouse astrocytes were grown on 35 mm dishes to 90% confluence. Medium was removed and cells washed in Hanks' balanced salt solution (HBSS:Gibco BRL; Gaithersburg, Maryland, USA). A serum-free Dulbecco's modified Eagles's medium was then added to the cells and they were subjected to severe hypoxia (<2% oxygen) for 4 h in a hypoxic incubator (Billups-Rothenberg; Del Mar, California, USA) with a humidified, temperature-controlled incubator within the chamber. The entire system was purged with 95% N2/5% CO2 atmosphere. At the end of the hypoxic period, cells were reoxygenated in ambient air at 37°C for 24 h. The cell free medium (ACM) was then collected. MCECs were incubated in the absence (Groups A and C) or presence of aprotinin (Groups B and D) at 1600 KIU mL−1, a clinically relevant concentration [15] (see groups below) for 2 h prior to exposure to ACM. ACM collected from either normoxic (Con ACM) or hypoxic-reoxygenated mouse type 1 astrocyte (HR ACM) was applied to bEnd3 cells grown in 35 mm dishes (1 × 105 cells/well) to 90% confluence. MCECs were incubated with ACM for 4 and 24 h. These time points were chosen as ICAM-1 demonstrates a plateau of maximal expression at 8 h following stimulation with cytokines and remains elevated for at least 24 h [16]. The medium after centrifugation was then collected and MCECs were prepared for biochemical analysis. The supernatants of media were stored at −80°C until use. Three (n = 3) independent experiments were performed.

Back to Top | Article Outline

Group assignment

A: No aprotinin, normoxic ACM (normal control);

B: Aprotinin, normoxic ACM (aprotinin control);

C: No aprotinin, HR ACM (experimental control);

D: Aprotinin, HR ACM (experimental group);

Back to Top | Article Outline

ICAM-1 expression

One hundred microlitres of stimulated MCEC suspension (1 × 106 cells mL−1) was stained with 10 μL of fluorescein-isothiocyanate conjugated (FITC) anti-CD54 (anti-ICAM-1) anti-mouse antibody (mAb) (Becton Dickinson, Cat. No. 553972, San Diego, CA, USA) or 10 μL of FITC-conjugated isotype IgG1 control mAb (Becton Dickinson, Cat No. 553253, San Diego, CA, USA) and incubated for 30 min at 4°C. ICAM-1 expression on MCEC was analysed with a Fluorescence-Activated Cell-sorter Scanner (FACScan cytofluorometer; Becton Dickinson, San Jose, CA, USA). The MCF intensity of stained cells was detected on the basis of a minimum number of 5000 cells collected, analysed using the software Lysis II.

IL-1β media supernatant concentrations were determined in ACM before incubation with MCECs using enzyme-linked immunosorbent assays (ELISAs) (Quantikine R&D systems, Europe Ltd, Abingdon, Oxfordshire, UK) according to manufacturer instructions. Concentrations were estimated at the end of HR. The sensitivity for IL-1β was 10 pg mL−1. The inter- and intra assay precisions for IL-1β for the range of values obtained in this study are 4.1-8.4% and 2.8-5.4%.

Back to Top | Article Outline


The Sigma Stat 2.0 for windows (SPSS, Inc., Chicago, IL, USA) software package was used for all statistical analyses. The data obtained were normally distributed and were analysed using repeated measures ANOVA and post hoc Tukey test or unpaired t-test as appropriate. P < 0.05 was considered significant. Data is presented as mean (SD).

Back to Top | Article Outline


MCEC ICAM-1 expression was similar in all 4 groups after 4-h exposure to HR ACM or normoxic ACM (Fig. 1). After 24 h exposure to HR ACM, MCEC ICAM-1 expression was greater in Group C (HR ACM) than in Group A (normoxic ACM) (MCF 107.5 (4.5) vs. 74.3 (4.5), P < 0.001) (Fig. 2).

Figure 1.

Figure 1.

Figure 2.

Figure 2.

MCEC ICAM-1 expression was similar in Group A (normoxic ACM) and Group B (aprotinin control) at 4 and 24 h (MCF, 73 (1) vs. 75 (0.7), P = 0.9) and (76 (5.7) vs. 79 (2), P = 0.9), respectively (Figs 1 and 2). This result rules out a direct drug effect of aprotinin.

After 24 h exposure to HR ACM, MCEC ICAM-1 expression was less in Group D (aprotinin) (MCF, 91.0 (1.1)) than Group C (no aprotinin) (MCF, 107.5 (2.1), (P = 0.006) (Fig. 2).

Supernatant astrocyte IL-1β concentrations (pg mL−1) were less in normoxic ACM than HR ACM (4 h hypoxia followed by 24 h reoxygenation) (7.1 (0.2) vs. 4.1 (0.2), P = 0.01).

Back to Top | Article Outline


This study demonstrates that hypoxic mouse astrocytes activate MCECs via mediator(s) secreted/released into cell media. Proinflammatory activation of MCECs by this paracrine pathway is manifest by increased expression of ICAM-1. This activation is prevented by aprotinin.

Close apposition of CEC and astrocytes, which invest 99% of the abluminal surface of the capillary membrane in the brain in vivo, provides for the proximity necessary to exact paracrine interactions. Evidence from grafting experiments, developmental studies, and culture models demonstrate inductive influences of perivascular astrocytes on brain endothelium and vice versa [17]. Therefore, cooperation between two cell types is likely to be involved in initiating microvascular and inflammatory responses to ischaemia. Peripheral neutrophil mobilization into the ischemic brain is critically dependent on the ‘inflammatory’ status of CEC, including CEC expression of adhesion molecules and chemokines. Stanimirovic and colleagues [18] have shown that CECs ‘activated’ by simulated in vitro ischaemia or exogenous IL-1β upregulate expression of ICAM-1, and avidly bind freshly isolated neutrophils in an ICAM-1/CD 18-dependent manner. Our results are consistent with those of Zhang and colleagues [12], who demonstrated increased CEC ICAM-1 expression and mouse type 1 astrocyte IL-1β concentrations in response to astrocytes exposed to HR. Zhang and colleagues [12] also demonstrated inhibition of these effects by the addition of IL receptor anatagonist (IL-Ra) and dexamethasone. Hypoxia increased IL-1β concentrations by upregulating IL-1β mRNA expression in astrocytes [12]. There was an increase in ICAM-1 expression in response to ACM (HR) at 24 h but not at 4 h in our study. ICAM-1 expression demonstrates a plateau of maximal expression at 8 h following stimulation with cytokines and remains elevated for at least 24 h [16].

Brain endothelial cells in vivo are exposed to various paracrine sources of proinflammatory mediators including those released by smooth muscle cells, pericytes, and perivascular macrophages. In vivo, small amounts of mediator released can reach high concentrations. Mediators released into ACM, however, are substantially diluted when applied to CECs. This may account for a decreased HR ACM effect in this study.

Cerebral endothelium is an important target for IL-1β produced by hypoxic perivascular astrocytes. The ability of IL-Ra to decrease ischaemic cerebral injury in vivo [19] is probably in part the result of both attenuated CEC activation via IL-1β produced by perivascular astrocytes and decreased neutrophil infiltration into the brain. Previously, therapies that targeted endothelial IL-1R1 or production/release of IL-1β by the glial cell compartment have been suggested as strategies to target post-ischaemic brain inflammation and secondary brain injury [12]. In this study, supernatant astrocyte IL-1β concentrations were less in normoxic ACM than HR ACM. Our study does not investigate a link between the increased IL-1β in astrocyte culture supernatant and the endothelial upregulation of ICAM-1, or demonstrate a relationship between IL-1β and the effect of aprotinin in downregulating ICAM-1 expression. The effect of HR on astrocyte IL-1β supernatant concentrations is reported, as it is presumably the source of IL-1β which has been shown to increase cerebral endothelial ICAM-1 expression [18]. We cannot exclude the possibility that proinflammatory activation of MCEC is aided by other factors secreted in hypoxic ACM. Zhang and colleagues [20] have demonstrated that both IL-8 and tumour necrosis factor alpha (TNFα) mRNA are increased in human astrocytes exposed to in vitro ischaemic conditions. TNFα has also been shown to increase human CEC ICAM-I expression associated with ischaemic conditions [21].

Aprotinin has both soluble and cell-associated targets within the inflammatory system [22]. It has been shown to significantly decrease neutrophil activation at 15-60 min following cardiopulmonary bypass (CPB), as assessed by diminished expression of Mac-1 (CD11b/CD18) [23]. It decreases neutrophil azurophilic granule release and blocks secretion of myeloperoxidase and neutrophil elastase induced by neutrophil chemoattractants [15]. Aprotinin therefore exerts a potent inhibitory effect on neutrophils. Aprotinin exerts no effect on either rolling or adhesion of neutrophils, but it significantly inhibits the passage of neutrophils through the endothelial wall [15]. In vitro, aprotinin acts to inhibit ICAM-1 expression and neutrophil transmigration in endothelial monolayers stimulated with TNFα [24]. A protease-sensitive pathway has been postulated to play a necessary role in optimal endothelial ICAM-1 expression [24]. Endothelial cells express protease-activated receptors (1 and 2) and as aprotinin is known to block these in vitro [25], aprotinin may thus prevent autocrine upregulation of CEC ICAM-1 by this mechanism.

In animals, systemic inflammatory mediators trigger extensive changes in (central nervous system) CNS inflammatory gene expression, neurochemistry, neuroendocrine status, thermoregulation, behaviour, and cognition [26]. For example, systemic inflammatory mediators trigger expression of IL-1 [27], IL-6 [30], TNFα [27], complement components [31], inducible cyclooxygenase [28], and inducible nitric oxide synthase [31] in the brain parenchyma, cerebral vasculature, and/or perivascular microglia. CNS responses to systemic inflammatory mediators rapidly alter CNS gene expression and functional status, and, by their nature, are likely to augment CNS injury from any coexisting perioperative neurologic insults [32].

The objective of this study was to examine if hypoxic mouse astrocytes activate MCECs via mediator(s) secreted/released in to cell media and to assess the effect of aprotinin on cerebral endothelial ICAM-1 expression. Further studies should investigate whether IL-1β added to endothelial cultures in the concentrations measured in stressed astrocyte media result in endothelial ICAM-1 upregulation, and whether this is suppressed by aprotinin. It would be a novel finding if aprotinin actually inhibits IL-1β release by stimulated astrocytes. The implication would be that if ICAM-1 expression and the resulting endothelial injury, results from astrocyte IL-1β release, aprotinin would thus have a protective effect at two points in the system. Also the protection produced by aprotinin should be investigated to determine if it was additive to that produced by optimal doses of IL1β-Ra, which would suggest that additional pathways are involved.

Patients undergoing cardiac surgery demonstrate a marked generalized inflammatory response [33]. Based on animal investigations [34], it is likely that non-specific inflammation exacerbates the injury associated with focal cerebral ischaemia following microgaseous or macroatheromatous cerebral embolization, as occur during CABG and CPB. Decreased cerebral endothelial ICAM-1 expression in response to activated glial cell compartment by aprotinin, as demonstrated in this study, may decrease post-ischaemic brain inflammation and secondary brain injury associated with cardiac surgery and CPB.

Back to Top | Article Outline


1. Alderman EL, Levy JH, Rich JB, et al. Analyses of coronary graft patency after aprotinin use: results from the International Multicenter Aprotinin Graft Patency Experience (IMAGE) trial. J Thorac Cardiovasc Surg 1998; 116: 716-730.
2. Murkin JM. Attenuation of neurologic injury during cardiac surgery. Ann Thorac Surg 2001; 72: S1838-S1844.
3. Barone FC, Feuerstein GZ. Inflammatory mediators and stroke: new opportunities for novel therapeutics. J Cereb Blood Flow Metab 1999; 19: 819-834.
4. Dirnagl U, Iadecola C, Moskowitz MA. Pathobiology of ischaemic stroke: an integrated view. Trends Neurosci 1999; 22: 391-397.
5. Barone FC, Schmidt DB, Hillegass LM, et al. Reperfusion increases neutrophils and leukotriene B4 receptor binding in rat focal ischemia. Stroke 1992; 23: 1337-1347.
6. Clark RK, Lee EV, White RF, Jonak ZL, Feuerstein GZ, Barone FC. Reperfusion following focal stroke hastens inflammation and resolution of the ischemic injured tissue. Brain Res Bull 1994; 35: 387-392.
7. Chopp M, Zhang RL, Chen H, Li Y, Jiang N, Rusche JR. Postischemic administration of an anti-Mac-1 antibody reduces ischemic cell damage after transient middle cerebral artery occlusion in rats. Stroke 1994; 25: 869-876.
8. Wang X, Feuerstein GZ. Induced expression of adhesion molecules following focal brain ischemia. J Neurotrauma 1995; 12: 825-832.
9. Stanimirovic D, Satoh K. Inflammatory mediators of cerebral endothelium: a role in ischemic brain inflammation. Brain Pathol 2000; 10: 113-126.
10. Nagy Z, Kolev K, Csonka E, Pek M, Machovich R. Contraction of human brain endothelial cells induced by thrombogenic and fibrinolytic factors. An in vitro cell culture model. Stroke 1995; 26: 265-270.
11. Joo J. Endothelial cells of the brain and other organ systems: some similarities and differences. Prog Neurobiol 1996; 48: 255-273.
12. Zhang W, Smith C, Howlett C, Stanimirovic D. Inflammatory activation of human brain endothelial cells by hypoxic astrocytes in vitro is mediated by IL-1 beta. J Cereb Blood Flow Metab 2000; 20: 967-978.
13. Tan KH, Dobbie MS, Felix RA, Barrand MA, Hurst RD. A comparison of the induction of immortalized endothelial cell impermeability by astrocytes. NeuroReport 2001; 12: 1329-1334.
14. Yoder EJ. Modifications in astrocyte morphology and calcium signalling induced by a brain capillary endothelial cell line. Glia 2002; 38: 137-145.
15. Asimakopoulos G, Thompson R, Nourshargh S, et al. An anti-inflammatory property of aprotinin detected at the level of leukocyte extravasation. J Thorac Cardiovasc Surg 2000; 120: 361-369.
16. McEver RP. Selectins: novel receptors that mediate leukocyte adhesion during inflammation. Thromb Haemostasis 1991; 65: 223-228.
17. Abbott NJ, Revest PA, Romero IA. Astrocyte-endothelial interaction: physiology and pathology. Neuropathol Appl Neurobiol 1992; 18: 424-433.
18. Stanimirovic DB, Ball R, Durkin JP. Stimulation of glutamine uptake and Na, K-ATPase activity in rat astrocytes exposed to ischemia-like insults. Glia 1997; 19: 123-134.
19. Garcia JH, Liu KF, Relton JK. Interleukin-1 receptor antagonist decreases the number of necrotic neurons in rats with middle cerebral artery occlusion. Am J Pathol 1995; 147: 1477-1486.
20. Zhang W, Smith C, Monette R, Shapiro A, Hutchinson J, Stanimirovic D. Increased expression of bioactive chemokines in human cerebromicrovascular endothelial cells and astrocytes subjected to simulated ischemia in vitro. J Neuro-immunol 1999; 101: 148-160.
21. Stanimirovic DB, Shapiro A, Wong J, Hutchison J, Durkin JP. The induction of ICAM-I in human cerebromicrovascular endothelial cells (HCEC) by ischemic-like conditions promotes enhanced neutrophil/HCEC adhesion. J Neuroimmunol 1997; 76: 193-205.
22. Mojcik CF, Levy JH. Aprotinin and the systemic inflammatory response after cardiopulmonary bypass. Ann Thorac Surg 2001; 71: 745-754.
23. Hill GE, Alonso A, Spurzem JR, Stammers AH, Robbins RA. Aprotinin and methylprednisolone equally blunt neutrophil cardiopulmonary bypass-induced inflammation in humans. J Thorac Cardiovasc Surg 1995; 110: 1658-1662.
24. Asimakopoulos G, Lidington EA, Mason J, Haskard DO, Taylor KM, Landis RC. Effect of aprotinin on endothelial cell activation. J Thorac Cardiovasc Surg 2001; 122: 123-128.
25. Pintigny D, Prigent DJ. Aprotinin can inhibit the proteolytic activity of thrombin: a fluorescent and enzymatic study. Eur J Biochem 1992; 207: 89-95.
26. Dantzer R, Bluthe RM, Gheusi G, et al. Molecular basis of sickness behavior. Ann NY Acad Sci 1998; 856: 132-138.
27. Turrin NP, Gayle D, Ilyin SE, et al. Pro-inflammatory and anti-inflammatory cytokine mRNA induction in the periphery and brain following intraperitoneal administration of bacterial lipopolysaccharide. Brain Res Bull 2001; 54: 443-453.
28. Vallieres L, Rivest S. Interleukin-6 is a needed proinflammatory cytokine in the prolonged neural activity and transcriptional activation of corticotrophin-releasing factor during endotoxemia. Endocrinology 1999; 140: 3890-3903.
29. Nadeau S, Rivest S. The complement system is an integrated part of the natural innate immune response in the brain. FASEB J 2001; 15: 1410-1412.
30. Lacroix S, Rivest S. Effect of acute systemic inflammatory response and cytokines on the transcription of the genes encoding cyclooxygenase enzymes (COX-1 and COX-2) in the rat brain. J Neurochem 1998; 70: 452-466.
31. Wong ML, Bongiorno PB, al-Shekhlee A, Esposito A, Khatri P. IL-1 beta, IL-1 receptor type I and iNOS gene expression in rat brain vasculature and perivascular areas. NeuroReport 1996; 7: 2445-2448.
32. Allan SM, Rothwell NJ. Cytokines and acute neuro degeneration. Nat Revs Neurosci 2001; 2: 734-744.
33. Paparella D, Yau TM, Young E. Cardiopulmonary bypass induced inflammation: pathophysiology and treatment. An update. Eur J Cardiothorac Surg 2002; 21: 232-244.
34. del Zoppo GJ, Schmid-Schonbein GW, Mori E, Copeland BR, Chang CM. Polymorphonuclear leukocytes occlude capillaries following middle cerebral artery occlusion and reperfusion in baboons. Stroke 1991; 22: 1276-1283.

INFLAMMATION; CELL ADHESION MOLECULES; intercellular adhesion molecule-1; PROTEINS; aprotinin; CENTRAL NERVOUS SYSTEM; neuroprotective agents; hypoxia-ischaemia; brain

© 2005 European Society of Anaesthesiology