Complement-Dependent Shock and Tissue Damage Induced by Intravenous Injection of Cholesterol-Enriched Liposomes in Rats

 

Lajos Baranyi, PhD*

János Szebeni, MD, PhD*

Sándor Sávay, MD*

Michael Bodo, MD, PhD†

Milan Basta, MD, PhD‡

Timothy B. Bentley, PhD†

Rolf Bunger, MD, PhD§

Carl R. Alving, MD, PhD*

 

*Department of Membrane Biochemistry, Walter Reed Army Institute of Research, Silver Spring, Maryland

†Department of Resuscitative Medicine, Walter Reed Army Institute of Research, Silver Spring, Maryland

‡Epilepsy Research Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Maryland

§Department of Physiology, Anesthesiology, Molecular and Cell Biology, Uniformed Services University of Health Sciences, Bethesda, Maryland

 

Key words: C5a, cholesterol, complement activation, DIC, liposome, edema,
hemorrhage, shock, thromboxane A2, thrombus

ABSTRACT

A single intravenous injection of 4 to 8 mg/kg of large, multilamellar, cholesterol-enriched lipid vesicles (containing 71% cholesterol; HC–MLV) induced marked bradycardia, arrhythmia, and transient decrease in systemic blood pressure in rats. A single higher dose (15–20 mg/kg), or repeated injections of 4 to 8 mg/kg HC–MLV, resulted in severe pulmonary hemorrhage and edema (ARDS-like histologic changes), systemic microthrombus formation, hemorrhage in the kidneys and liver, as well as early signs of diffuse myocardial damage. Complement depletion with cobra venom factor, or the use of thromboxane A2 receptor inhibitor (SQ30741), prevented these adverse reactions, pointing to the involvement of C activation and TXA₂ release in the pathomechanism. The likely role of C activation was supported by the demonstration of strong C activation by HC–MLV in rat serum in vitro, along with the binding of natural IgG and IgM antibodies to these vesicles. The C activation by HC–MLV seems to proceed by way of the classic pathway mediated by natural antilipid antibodies. These studies present a novel, powerful method for inducing anaphylactoid shock and other C activation-related pathologic changes providing a model for multiple organ failure, semicrystalline cholesterol embolism, and C activation-related pseudoallergy.

INTRODUCTION

In the course of testing intravenously administered liposomes as potential drug carriers or as vehicles for hemoglobin in experimental blood substitutes, it has been discovered that liposomes could sometimes induce anaphylactoid reactions.1–3 The symptoms include hypo- and hypertension, bradycardia, and bradyarrhythmia. In some cases the reaction can be life-threatening. The reaction in pigs is characterized by a significant rise in pulmonary arterial pressure, followed by decrease in mean arterial blood pressure, bradyarrhythmia, transient thrombocytopenia, and thromboxane A2 release,3,4 a reaction that has been referred to as complement activation-related pseudoallergy (CARPA).4

Although IgE-mediated allergy has been excluded as a cause of this phenomenon, it has been found that the liposomal surface does have the ability to activate the classic complement pathway by binding naturally-occurring IgG and IgM antibodies,5 or by activating the classic pathway through C1q-binding,2 alternative pathway,6 or lectin pathway.7 If complement activation is solely responsible for signs and symptoms of CARPA, then by using liposomes that have enhanced complement-activating capability, it might be possible to increase the reaction proportionally to the extent of increased complement activation.

It has been reported earlier that liposomes containing 71% cholesterol have enhanced ability to activate the complement cascade in vitro when compared with liposomes having lesser amounts of cholesterol, such as 43% cholesterol.8 The data presented here indicate that intravenous injection of such cholesterol-enriched lipid vesicles triggers massive hemodynamic and other physiological changes that are also characteristic of a range of severe pathologic conditions, including hemorrhagic and septic shock, multiple organ failure, and semicrystalline cholesterol embolism. In addition to characterizing the HC–MLV-induced pathologic changes in rats, the goals of the present study were to explore the mechanism of C activation by HC–MLV and to test the efficacy of a TXA2 receptor inhibitor molecule in protecting against the described changes.

MATERIALS AND METHODS

Reagents

Dimyristoyl phosphatidylcholine (DMPC), dimyristoyl phosphatidylglycerol (DMPG), and cholesterol were purchased from Avanti Polar Lipids (Alabaster, AL). Thromboxane receptor A2 inhibitor SQ30741 (15-hydroxy-11 alpha, 9-alpha-[epoxymethano] prosta-5,13-dienoic acid) was provided by Dr. Gregory L. Stahl (Harvard Medical School, Boston, MA). Pooled, purified IgM (Venimmune) and pooled human IgG (Gamma Venin) were obtained from ZLB Bioplasma (Bern, Switzerland). Goat antihuman IgM–FITC, goat antihuman IgG–FITC, and cobra venom factor were purchased from Sigma (St. Louis, MO). Rabbit antirat IgG–FITC and rabbit antipig IgG–FITC were purchased from Pharmingen (San Diego, CA). The anticholesterol monoclonal mouse IgM antibody was described previously.9

Preparation and Size Distribution
of Liposomes

Large multilamellar vesicles containing 45:5:43 mole% of DMPC : DMPG : cholesterol (MLV) and 24:5:71 mole% of DMPC : DMPG : cholesterol (HC–MLV, or cholesterol-enriched liposomes) were prepared as described earlier.10 Measurement of liposome size by flow cytometry (vida infra) demonstrated that MLV and HC–MLV preparations had broad size distributions in the 0.8- to 5-µm range.

Experimental Protocols and Measurement of Hemodynamic Changes in Rats

Female Sprague–Dawley Rats (400–800 g) were anesthetized by injection of 50 mg/kg pentobarbital intraperitoneally and were breathing spontaneously. The core body temperature was monitored through a rectal thermal probe. The animals were placed on a homeothermic blanket (Harvard Apparatus, Holliston, MA) and the core temperature was kept at 36˚ to 37˚C. The internal jugular vein and carotid artery were cannulated. Liposomes were administered through the internal jugular vein in boluses of 0.05 to 1 mL using aliquots taken from 40 mg lipid/mL stock solutions. The injections took no more than 20 seconds, and liposomes were washed in the circulation with 0.2 mL phosphate-buffered saline (PBS). Repeated liposome injections were made at 15- to 30-minute intervals after all hemodynamic parameters normalized. The age and weight of rats did not influence the severity or reproducibility of liposome-induced shock. The arterial blood pressure was measured with a blood pressure analyzer (BPA-400A, Digi-Med, Louisville, KY) controlled by LabView-based software (LabView, National Instruments, Austin, TX; data acquisition program developed at the Walter Reed Army Institute of Research). The data were recorded at 200 Hz, and mean arterial blood pressure values were averaged for 5 seconds. Respiration was recorded with a type T miniature needle thermocouple connected to an electronic thermometer (BAT-12, Physitemp Instruments, NJ) and the measurements were digitized with a NI6023 DAQ card at 50-Hz sampling rate. A miniature thermocouple (needle probe 3721) was inserted into the tracheal tube that measured the temperature of the air in the tracheal tube. The maximum value represented the end of the expiration period; the minimum value the end of the inhalation period. The respiratory rate, inhalation and exhalation times were calculated from the thermocouple recordings.11 For statistical analysis, 10 inhalation–exhalation cycles were measured (inhalation time, exhalation time, amplitude, and peak-to-peak [P–P] distance) and compared with corresponding baseline recordings (Fig. 2D). Blood samples were obtained through the arterial catheter and were either allowed to clot for serum collection or treated with anticoagulant (8 U/mL heparin or 10 mM EDTA).

At the end of the observation period (1–3 injections of liposomes), selected animals (11 in total) were exsanguinated, and the liver, kidneys, and spleen samples were collected, cut into 4- to 5-mm thick slices, and preserved in 4% paraformaldehyde– PBS solution. The heart and the lung were perfused, filled with the same paraformaldehyde solution, and ligated in an inflated position. Samples were subsequently embedded in paraffin blocks, cut, and stained with hematoxylin–eosin.

Decomplementation

Decomplementation of rats was performed with an intraperitoneal injection of 20 U/kg cobra venom factor 24 hours and 1 hour before the injection of liposomes.

Measurement of Liposome-Induced Complement Consumption In Vitro

Liposomes (0.4 mM final phospholipid concentration) were incubated with undiluted rat serum at 37˚C while shaking at 80 rpm. After 30 minutes, vesicles were separated by centrifugation (at 18,000 rpm for 3 min) and the remaining hemolytic activity in the serum (CH₅₀ units) was determined.2

Flow-Cytometric Analysis of Immunoglobulin Binding to Liposomes

Liposomes (0.4 mM final phospholipid concentration) were incubated with serially diluted sera, purified pooled human IgG or IgM, or murine cholesterol-specific monoclonal IgM (clone, 2C8, developed at WRAIR, DC) for 30 minutes at 37˚ C with gentle agitation. After centrifugation, vesicles were washed, stained with FITC-conjugated rabbit antibodies directed against species-specific IgG and IgM (Sigma, St. Louis, MO), and analyzed with a FACSCalibur flow cytometer (Becton-Dickinson, Mountain View, CA). The FITC-labeled antibody without primary antibodies or sera served as negative control for nonspecific background staining.

Statistical Methods

Data are presented either as mean ± standard error or as individual observations. Differences between groups were examined using analysis of variance (ANOVA) and Student unpaired t tests, which was performed using GraphPad Prism for Windows V.3.00 (GraphPad Software, San Diego, CA).

RESULTS

Induction of Shock by
Intravenous Injection of
Cholesterol-Enriched Liposomes

Injection of low doses of HC–MLV resulted in anaphylactoid reactions characterized by severe hypotension, bradycardia, and tachypnea. As shown in Figure 1, after injection of 8 mg/kg HC–MLV the mean arterial blood pressure dropped by 55.6% (± 24.3% calculated from the analysis of data from 17 rats), and it reached a minimum within 2 to 4 minutes after the injection. Blood pressure remained low for an additional 3 to 5 minutes, and then gradually normalized (within 10–30 min). In surviving rats (27 of 31), bolus injections could be repeated 2 to 5 times with similar severe drops in blood pressure (Fig. 1), but no rat survived more than 6 injections. Usually the blood pressure became unstable after 3 to 5 injections and then gradually fell, leading to death of the animal. In contrast to HC–MLV, injections of comparable doses of MLV containing 43% cholesterol did not affect the arterial blood pressure (Fig. 1, inset). These blood pressure changes were associated with transient bradycardia with arrhythmia. A typical recording is shown on Figure 1, and inset D illustrates the parameters measured from the recordings.

Hypoventilation also occurred, as indicated by decreased frequency and irregularity of breathing rate (Fig. 2A–C). A statistically significant change in standard error values of peak-to-peak distance calculations clearly demonstrates the increased irregularity of breathing (Fig. 3, inset). Figure 3 also demonstrates that the amplitude of breathing slightly decreased (shallower breathing) by 15% to 20% within 2 minutes in rats injected with HC–MLV. The ratio of amplitudes measured before and after the HC–MLV injection also decreased slightly. Strong decrease in the ratio of pulse-to pulse distances is the result of decreased frequency of breathing in HC–MLV-treated rats, a change significant in each of the 8 cases analyzed (at P <0.001 level). Although the inhalation time did not change much, the time needed for exhalation also increased substantially, as the ratio of baseline and HC–MLV-treated values indicate (Fig. 3). Again, the changes are statistically significant in each of the 8 rats at the P <0.001 level. The amplitude subsequently normalized (Fig. 2C), but the inhalation/exhalation ratio remained depressed, indicating the onset of persistent bronchial constriction.

After the injection of HC–MLV, pulmonary fluid appeared in the intratracheal tube within 15 to 40 minutes, indicating severe permeability change and edema development in the lungs. In response to subsequent HC–MLV injection(s), the fluid became visibly pinkish in color, caused by extravasation of erythrocytes released during progressing pulmonary hemorrhage.

Histopathologic Changes in Rat
Organs After Cholesterol-Enriched Liposome Injection

The major pathologic change observed in all organs was a marked congestion of capillary blood and interstitial microhemorrhage typically after 3 injections of 8 mg HC–MLV in approximately 20-minute intervals, although some rats developed the symptoms after a single injection. Typical changes are illustrated for kidney, lung, liver, and heart (Fig. 4). In the kidney, congested tortuous glomeruli were seen, and a large number of erythrocytes were found in the interstitium surrounding the tubuli (Fig. 4A). In addition to diffuse hemorrhage in the pulmonary parenchyma, there were areas of major hemorrhage, and granulocytes could be observed within the alveoli (Fig. 4B). In the liver, endothelial injury, necrosis, and parenchymal degeneration with hemorrhage, as well as the presence of erythrocytes between the sinusoids, were found (Fig. 4C). Figure 4D shows myocardial morphological alterations resulting from liposome-induced damage. The apical and septal regions of ventricles exhibited areas with darker, seemingly degenerated muscle. These areas were confirmed by histopathologic examination to be necrotic, containing a large number of myocytes with disrupted muscle fibers. These regions also showed congested blood vessels with severe endothelial injury and leakage. In addition, erythrocytes were found between muscle fibers in a diffuse fashion (Fig. 4D). No histologic changes were seen in rats treated with comparable doses of MLV.

These observations suggest the possibility that thrombogenesis and blockade of microvessels might have occurred in the animals that were injected with HC–MLV. Microaggregates and microthrombi consisting of platelets, liposomes, and erythrocytes were regularly found in all organs, particularly in the lung (Fig. 4E). However, considering the remarkable reversibility of the changes in blood pressure and respiration rate in most of the animals during the first injections, it appears that initially the congestion might have been transient and only one of the possible effects that contributed to development of the symptoms. Additional factors had to be considered, and the question remained open as to why cholesterol-enriched liposomes induced far more severe reactions and caused permanent damage in rats than did liposomes with lower cholesterol content.

In Vitro Complement Activation
by Liposomes

Increasing the cholesterol content of the liposomes to 71% (HC–MLV) rendered the particles highly effective complement activators when compared with low-cholesterol (MLV) liposomes (Fig. 5). Under the conditions used in this study, cholesterol-enriched liposomes consumed 18% of total hemolytic complement activity in rat sera within 15 minutes, whereas less than 2% consumption was measured in samples incubated with control liposomes (MLV). Differences were even more striking in samples in which human and pig sera were used. Cholesterol-enriched liposomes consumed all of the available hemolytic activity in pig sera and more than 70% of the activity in human sera, whereas the cholesterol-poor version of the liposomes (MLV) consumed approximately 20% of pig or human hemolytic activity. The ANOVA analysis indicated that the difference in complement source (rat, pig, or human) is responsible for approximately 70% of variance in the experiment, whereas the difference in liposomes contributes approximately 25%. The difference between MLV and HC–MLV is also statistically significant by unpaired t test (P <0.05) in each case.

One possible reason for enhanced complement activation by HC–MLV might be that HC–MLV could bind more of the naturally occurring anticholesterol antibodies that are known to be present in blood.1,12 Liposomes with either 71% or 43% cholesterol were treated with murine monoclonal anticholesterol antibodies, and the binding of the antibodies to the liposomes was determined by flow cytometry using an indirect immunofluorescence staining method (Fig. 6). The HC–MLV vesicles were stained with monoclonal anticholesterol antibodies more strongly than MLV of the same size. The mean fluorescence channel value for anticholesterol antibodies was 373 in the case of HC–MLV and 41 in the case of MLV after subtraction of nonspecific binding by FITC-labeled secondary antibody. Similar results were observed when the HC–MLV and MLV vesicles were exposed to normal human sera instead of monoclonal anticholesterol antibodies. The experiment was repeated with purified polyclonal human IgG and IgM, and the anti-IgM antibodies stained stronger than the anti-IgG (data not shown), indicating a potential predominance of IgM class natural anticholesterol antibodies in human sera, as previously described.12

Inhibition of the Complement System

Injection of 20 units of cobra venom factor 24 hours and 1 hour before the experiments was used as a method to block the complement cascade in 3 rats. Decomplementation proved to be highly protective and prevented the hypotensive reaction normally seen in HC–MLV-treated rats. As shown by a representative experiment (Fig. 7), rats (n = 3) survived the injection of an otherwise lethal dose of liposomes (30 mg/kg), thereby demonstrating that complement played a critical role in the cardiovascular distress induced by HC–MLV.

Inhibition of Thromboxane A2 Receptor

One of the known consequences of complement activation is the release of anaphylatoxin (C5a), a mediator that has been shown to be involved in liposome-induced cardiovascular reactions in pigs.10 C5a also contributes to the release of thromboxane A2 by platelets or macrophages, a known accompaniment of C5a-mediated shock.11 Because thromboxane A2 is one of the most potent smooth muscle contractors, a thromboxane A2 receptor inhibitor might have the beneficial effect of ameliorating HC–MLV-induced cardiopulmonary distress. To test this, rats were pretreated with 1 mg/kg SQ30741, a thromboxane A2 receptor inhibitor. This pretreatment prevented the HC–MLV-induced reactions, resulting in no significant change in blood pressure (Fig. 8, representative of 3 independent experiments). Although tissue damage and pulmonary edema were not observed at these liposome doses, the protective effects of the thromboxane receptor inhibitor SQ30741 were gradually overcome when the dose of liposomes was increased to 40 mg/kg or higher (data not shown).

DISCUSSION

The data presented in this study demonstrate that intravenous injection of cholesterol-enriched liposomes cause severe cardiovascular reactions, characterized by severe bradycardia, arrhythmia, hypotension, vascular leakage, hemorrhage, and rapid death of the rats. The reaction was the result of an increased ability of cholesterol-enriched liposomes, as compared with non-enriched liposomes, to activate the complement system.

The MLV and HC–MLV used in our experiments were relatively large (0.8–5 µm) and had similar size distributions. This is relevant because of our previous observation that liposomes having small sizes activate the complement system to a lesser extent than large liposomes.10 Size differences are therefore excluded as the source of the differences between liposomes in the present study. Although liposomal-negative charge (resulting from DMPG) is also known to correlate with the intensity of complement activation and liposome-induced cardiovascular reaction (CARPA) in pigs,10 the DMPG/DMPC ratio was identical for MLV and HC–MLV used in the present study.

The cause of the increased cardiovascular reactivity by HC–MLV apparently lies primarily in the increased cholesterol content of the vesicles. Earlier studies have indicated that separation of lipid mixtures into cholesterol-rich and cholesterol-poor domains rapidly occurs at room temperature.14 The existence of an altered membrane structure is also suggested by the observation that cholesterol-enriched liposomes can be used to induce, and can efficiently bind, cholesterol-specific monoclonal antibodies9,15 (Fig. 6). The enhanced ability of cholesterol-enriched liposomes to bind anticholesterol antibodies suggests that the 3-OH group, the epitope, to which the antibody is known to bind,16 is less cryptic in the cholesterol-enriched liposomes.

The presence of natural antiphospholipid and anticholesterol antibodies in human blood and in blood from other species is well-documented,17,18 and induction of similar antibodies by immunization with liposomes containing lipid A has also been demonstrated.15,19,20 Increased charge density outside the cholesterol islands on the membrane surface might also play a role, because phospholipids are excluded from these areas, thus possibly enhancing the binding the naturally occurring antiphospholipid antibodies. We have confirmed the enhanced binding of antibodies, primarily IgM, to HC–MLV by flow cytometry using either murine monoclonal antibodies, or naturally occurring antibodies in human, pig, or rat sera.

In addition to complement activation by antiphospholipid and anticholesterol antibodies, complement activation might occur in the absence of antigen-specific immunoglobulins.21 Complement activation results in opsonization and massive release of anaphylatoxins C5a and C3a, which are split products that have wide ranges of physiological effects.22–24 Our observation that decomplementation of rats with cobra venom factor abrogates the cardiopulmonary symptoms elicited by injection of the HC–MLV strongly suggests that complement activation is the primary source of the symptoms.

An additional consequence of the lack of membrane-associated complement regulatory molecules on the surface of liposomes (and other complement activating particles, such as the semicrystalline cholesterol released from ruptured atherosclerotic plaques) is a fast and extensive opsonization by C3b (and C4b). These opsonins are high-affinity ligands for CR1, which normally facilitates the clearance of opsonized particles from the circulation.25,26 CR1 is abundant in the membrane of erythrocytes, granulocytes as well as platelets, and binding of highly opsonized HC–MLV to C5a-activated, CR1 receptor-bearing cells might result in formation of metastable rosettes of liposomes, erythrocytes, and C5a-activated granulocytes and platelets,25–29 which could serve as a core for formation of microthrombi found in rats treated with HC–MLV.

Although the release of the anaphylatoxins results in smooth muscle contraction, and would be expected to reduce the lumen of arterioles, resulting in increased blood pressure, we actually measured a rapid fall in mean blood pressure. This anomaly might have been caused by the introduction of HC–MLV into circulation through the jugular vein, with the lung therefore being the first organ affected by rapid complement activation. Vasoconstriction in the lungs was detected in pigs on injection of liposomes. A rapid 30% to 200% increase in pulmonary blood pressure was seen within 3 to 5 minutes of the injections. This was quickly followed by a severe drop in mean arterial blood pressure and a significant decrease in cardiac output, because the lungs failed to provide sufficient blood volume for maintaining the systemic blood pressure, a status rapidly exacerbated by cardiac ischemia that is known to follow this sequence of events.4

Anaphylatoxins and other factors accumulating during local hypoxia created by impaired microcirculation might also contribute to accumulation and activation of circulating granulocytes22–24,27 or mast cells.23,28 Granulocyte activation might be accompanied by release of granzymes, a set of proteolytic enzymes, all of which could be responsible for the observed tissue damage, vascular leakage, and hemorrhage,28–30 which might play a role in the physiological reaction induced by intravenous injection of HC–MLV.

We conclude that intravenous injection of cholesterol-enriched lipid vesicles having an enhanced ability to activate the complement systems results in escalation of complement activation-related pseudoallergic reaction (CARPA)4 to a fatal form of an anaphylactoid reaction, which is characterized not only by changes in blood pressure and bradyarrhythmia, but also thromboxane A2 release. This phenomenon shows prominent ARDS-like features30–32 and symptoms seen in systemic crystalline cholesterol syndrome,33,34 both of which represent severe and clinically important conditions.

REFERENCES

1. Wassef NM, Johnson SH, Graeber GM, et al: Anaphylactoid reactions mediated by autoantibodies to cholesterol in miniature pigs. J Immunol 143:2990‑2995, 1989.

2. Szebeni J, Wassef NM, Spielberg H, Rudolph AS, Alving CR: Complement activation in rats by liposomes and liposome-encapsulated hemoglobin: evidence for anti-lipid antibodies and alternative pathway activation. Biochem Biophys Res Commun 205:255‑263, 1994.

3. Szebeni J: The interaction of liposomes with the complement system. Crit Rev Ther Drug Carrier Syst 15:57‑88, 1998.

4. Szebeni J, Fontana JL, Wassef NM, et al: Hemodynamic changes induced by liposomes and liposome-encapsulated hemoglobin in pigs: a model for pseudoallergic cardiopulmonary reactions to liposomes. Role of complement and inhibition by soluble Cr1 and anti-C5a antibody. Circulation 99:2302‑2309, 1999.

5. Szebeni J, Alving CR: Complement-mediated acute effects of liposome-encapsulated hemoglobin. Artif Cells Blood Substit Immobil Biotechnol 27:23‑41, 1999.

6. Funato K, Yamashita C, Kamada J, Tominaga S, Kiwada T: Plasma factor triggering alternative complement pathway activation by liposomes. Pharm Res 11:372–376, 1993.

7. Chonn A, Semple SC, Cullis PR: Separation of large unilamellar liposomes from blood components by spin column procedure: towards identifying plasma proteins which mediate liposome clearance in vivo. Biochim Biophys Acta 1070:215–222, 1991.

8. Alving CR, Richards RL, Guirguis AA: Cholesterol-dependent human complement activation resulting in damage to liposomal model membranes. J Immunol 118:342‑347, 1977.

9. Swartz GM, Gentry MK, Amende LM, Blanchette-Mackie EJ, Alving CR: Antibodies to cholesterol. Proc Natl Acad Sci U S A 85:1902‑1906, 1988.

10. Szebeni J, Baranyi L, Savay S, et al: Liposome-induced pulmonary hypertension: properties and mechanism of a complement-mediated pseudoallergic reaction. Am J Physiol Heart Circ Physiol 279:H1319‑H1328, 2000.

11. Gaddes LA: Respiration. In: Branzino JD, Ed: The biomedical engineering handbook. Boca Raton, FL: CRC Press; 1995:1231–1239.

12. Alving CR, Swartz GM, Wassef NM: Naturally-occurring autoantibodies to cholesterol in humans. Biochem Soc Trans 17:637‑639, 1989.

13. Marceau F, Lundberg C, Hugli TE: Effects of the anaphylatoxins on circulation. Immunopharmacology 14:67‑84, 1987.

14. Worthman LA, Nag K., Davis PJ, Keough KM: Cholesterol in condensed and fluid phosphatidylcholine monolayers studied by epifluorescence microscopy. Biophys J 72:2569–2580, 1997.

15. Alving CR, Swartz GM: Antibodies to cholesterol, cholesterol conjugates and liposomes: implications for atherosclerosis and autoimmunity. Crit Rev Immunol 10:441‑453, 1991.

16. Dijkstra J, Swartz GM, Raney JJ, et al: Interaction of anti-cholesterol antibodies with human lipoproteins. J Immunol 157:2006‑2013, 1996.

17. Alving CR: Antibodies to lipids and lipid membranes: reactions with phosphatidylcholine, cholesterol, liposomes, and bromelin-treated erythrocytes. In: Harris N, Exner T, Hughes GRV, Asherson R, Eds: Phospholipid binding antibodies. Boca Raton, FL: CRC Press; 1991:73–95.

18. Alving CR, Wassef NM: Naturally occurring antibodies to cholesterol: a new theory of LDL cholesterol metabolism. Immunol Today 20:362‑366, 1999.

19. Alving CR: Natural antibodies against phospholipids and liposomes in humans. Biochem Soc Trans 12:342‑344, 1984.

20. Alving CR, Swartz GM, Wassef NM, et al: Immunization with cholesterol-rich liposomes induces anti-cholesterol antibodies and reduces diet-induced hypercholesterolemia and plaque formation. J Lab Clin Med 127:40‑49, 1996.

21. Marjan J, Xie Z, Devine DV: Liposome-induced activation of the classical complement pathway does not require immunoglobulin. Biochim Biophys Acta 1192:35‑44, 1994.

22. Ember JA, Hugli TE: Complement factors and their receptors. Immunopharmacology 38:3‑15, 1997.

23. Erdei A, Kerekes K, Pecht I: Role of C3a and C5a in the activation of mast cells. Exp Clin Immunogenetics 14:16‑18, 1997.

24. Gerard C, Gerard NP: C5a anaphylatoxin and its seven transmembrane-segment receptor. Annu Rev Immunol 12:775‑808, 1994.

25. Fearon DT: Identification of the membrane glycoprotein that is the C3b receptor of the human erythrocyte, polymorphonuclear leukocyte, B lymphocyte, and monocyte. J Exp Med 152:20‑30, 1980.

26. Loughrey HC, Bally MB, Reinish LW, Cullis PR: The binding of phosphatidylglycerol liposomes to rat platelets is mediated by complement. Thromb Haemostasis 64:172–176, 1990.

27. Birmingham DJ: Erythrocyte complement receptors. Crit Rev Immunol 15:133‑154, 1995.

28. Michiels C, Arnouldt T, Remacle J: Endothelial cell response to hypoxia: initiation of cascade of cellular interactions. Biochim Biophys Acta 1497:1–10, 2000.

29. Hammerschmidt DE, Harris PD, Wayland JH, Craddock PR, Jacob HS: Complement-induced granulocyte aggregation in vivo. Am J Pathol 102:146‑150, 1981.

30. Cochrane CG: Role of granulocytes in immune complex-induced tissue injuries. Inflammation 2:319‑333, 1977.

31. Parsley EL: Acute respiratory distress syndrome. Cellular biology and pathology.            Respir Care Clin North Am 4:583‑609, 1998.

32. Ketai LH, Godwin JD: A new view of pulmonary edema and acute respiratory distress syndrome.     J Thorac Imaging 13:147‑171, 1998.

33. Hammerschmidt DE, Greenberg CS, Yamada O, Craddock PR, Jacob HS: Cholesterol and atheroma lipids activate complement and stimulate granulocytes: a possible mechanism for amplification of ischemic injury in atherosclerotic states. J Lab Clin Med 98:68‑77, 1981.

34. Cate H, Schoenmakers SH, Franco R, Timmermann JJ, Spek CA, Reitsman PH: Microvascular coagulopathy and disseminated intravascular coagulation. Crit Care Med 29:95–97, 2001.

 

 

Figure 1. Repeated injection of rats with cholesterol-enriched liposomes. The mean blood pressure (MBP) in the rats injected with boluses of 8 mg/kg liposomes containing 43% cholesterol did not change significantly (inset). However, the same dose, and even lower doses, of HC–MLV provoked a strong, but transient decrease in blood pressure. Parallel to blood pressure change, in rats treated with HC–MLV severe bradycardia was measured (HR, thin line, right Y axis). Arrows indicate the intravenous injection of liposome boluses. The rat injected with cholesterol-enriched liposomes survived 5 injections and then died (representative data from 6 experiments).

 

 

A

 

B

 

C

 

Figure 2. Changes in respiration rate. The figure shows representative recording of 8 animals from different phases of the reaction induced by HC–MLV injection. Respiratory changes can be seen within 2 to 3 minutes after injecting rats with cholesterol-enriched liposomes. The breathing becomes arrhythmic, the frequency decreases, and the normalization took more than 15 minutes, when it occurred at all. Table D shows the calculated parameters amplitude is in mV, proportional with respiration intensity. IN, inhalation time; EX, exhalation time; and P–P, distance is the peak-to-peak distance, all in seconds.

 

D

 

 

Figure 3. Changes in respiration. Ratios of inhalation and exhalation time, peak-to-peak distances, and amplitudes obtained from recording thermocouple inserted into the tracheal cannula before and after injection of HC–MLV. Each of the box-and-whiskers plot represents the average of 10 successive inhalation/exhalation episodes from 8 individual rats. The P–P distance ratios and exhalation time ratios refer to changes that are statistically significant at P <0.001 level in paired Student t test. The inset shows that the breathing become irregular a sensitive measure of which is the standard deviation of the P–P distance values, that increased significantly (P <0.0015, n = 8).

 

Figure 4. Tissue sections from rats treated with cholesterol-enriched liposome. The major pathologic change observed in all organs was marked congestion of blood vessels and interstitial microhemorrhage with extended injury of endothelium. (A) Kidney, (B) lung, (C) liver, (D) heart. Hematoxylin–eosin staining, magnification ¥ 90. The animal of which the tissue sections are presented here was injected with 8 mg HC–MLV 3 times and was moribund at the time of euthanasia. No histologic changes were seen in rats treated with similar doses of MLV. Inset E shows an electromicrograph of a microthrombus, an aggregate of erythrocytes, thrombocytes, and granulocytes in the lung. Liposome ghosts can also be noticed as transparent holes scattered in the bulk of thrombus (¥ 6000).

 

Figure 5. Complement activation by HC–MLV. Individual sera (3-3) were incubated with 0.4 mM liposomes composed of either MLV or with same amount of cholesterol-enriched MLV (HC–MLV) at 37˚C. After centrifugal removal of liposomes, CH50 values were determined and expressed as percentage of untreated control. Although the presence of MLV resulted in low-level complement consumption, the cholesterol-enriched liposomes consumed almost all of the hemolytic activity in pig and human sera and substantial reduction has been found in rat sera as well (P <0.05 in each instance).

 

 

Figure 6. Antibody binding by HC–MLV and MLV. Monoclonal anticholesterol antibody stained the HC–MLV stronger than the MLV particles of corresponding size distribution after subtraction of nonspecific binding by FITC control. Similarly, enhanced staining was found with purified pooled normal human IgM (2.5 µg/mL Venimmune), indicating an enhanced antibody binding to HC–MLV.

 

Figure 7. Effect of decomplementation. Decomplementation using cobra venom factor protected rats from even the lethal doses of HC–MLV (representative experiment from 3 rats, all similarly surviving the injection of lethal doses of HC–MLV).

 

 

Figure 8. Effect of thromboxane A2 receptor inhibitor SQ70431 on HC–MLV-induced cardiopulmonary distress. Blocking the thromboxane A2 receptors reduced the duration and decreased the severity of the HC–MLV-induced decrease in arterial blood pressure, and the rats survived the injection of otherwise lethal doses of liposome (representative of 3 experiments).