Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Apolipoprotein E knock-out mice are highly susceptible to endotoxemia and Klebsiella pneumoniae infection

Apolipoprotein E knock-out mice are highly susceptible to endotoxemia and Klebsiella pneumoniae... Apolipoprotein E knock-out mice are highly susceptible to endotoxemia and Klebsiella pneumoniae infection Natasja de Bont,* Mihai G. Netea,* Pierre N. M. Demacker,* Ineke Verschueren,* † 1, Bart Jan Kullberg,* Ko Willems van Dijk, Jos W. M. van der Meer,* and Anton F. H. Stalenhoef Division of General Internal Medicine,* Department of Medicine, University Hospital Nijmegen, 6500 HB Nijmegen, The Netherlands, and Department of Human Genetics, Leiden University Medical Centre, Leiden, The Netherlands Abstract Lipoproteins are able to neutralize bacterial lipo- LPS forms complexes with all major lipoprotein classes polysaccharide (LPS) and thereby inhibit the proinflamma- in plasma, including chylomicrons, very low density lipo- tory cytokine response. In a previous study, we demonstrated proteins (VLDL), low density lipoproteins (LDL), high that hypercholesterolemic low density lipoprotein receptor density lipoproteins (HDL), and lipoprotein [a] (LP[a]) knock-out (LDLr2/2) mice are protected against lethal en- (2–5). Binding of endotoxin to any of these lipoproteins dotoxemia and gram-negative infection. In the present study results in neutralization of LPS and reduced release of we investigated the susceptibility of apolipoprotein E knock- TNFa, IL-1b, and IL-6 (4, 6, 7). Recent studies have shown out mice (apoE2/2) to LPS and to Klebsiella pneumoniae. that hyperlipoproteinemia in genetically modified mice These mice have increased plasma lipoprotein concentra- tions in the very low density lipoprotein (VLDL)-sized frac- (8, 9) as well as infusion of reconstituted HDL (rHDL) in tion. Despite 8-fold higher plasma cholesterol levels com- humans (10, 11) results in a diminished cytokine response, pared to controls, and in contrast to LDLr2/2 mice, leading to protection against the toxic effects of endotox- apoE2/2 mice were significantly more susceptible to endo- emia. Other studies have demonstrated that triglyceride- toxemia and to K. pneumoniae infection. Circulating TNFa rich lipoproteins, such as chylomicrons and VLDL, also concentrations after intravenously injected LPS were 4- to have the capacity to inactivate LPS and prevent endo- 5-fold higher in apoE2/2 mice, whereas IL-1a, IL-1b, and toxin-induced death in rodents (3, 12–14). IL-6 did not differ. This TNF response was not due to an in- It is not well understood which lipoprotein component creased cytokine production capacity of cells from apoE2/2 is most important for LPS neutralization. The ability of li- mice, as ex vivo cytokine production in response to LPS did not differ between apoE2/2 and control mice. The LPS- poproteins to bind and inactivate LPS is modulated by neutralizing capacity of apoE2/2 plasma was significantly apolipoproteins. Both apolipoprotein (apo) A-I, present less than that of controls. Most likely, the absence of apoE it- on HDL, and apoE, present on chylomicrons, VLDL, self in the knock-out mice explains the failure to neutralize HDL, and IDL, are capable of directly inactivating endo- LPS, despite the very high cholesterol concentrations.—de toxin and decreasing LPS-induced cytokine release (14 – Bont, N., M. G. Netea, P. N. M. Demacker, I. Verschueren, 17). An additional mechanism may be represented by the B. J. Kullberg, K. W. van Dijk, J. W. M. van der Meer, and lipid–lipid interactions between the lipid A component of A. F. H. Stalenhoef. Apolipoprotein E knockout-mice are LPS and cholesterol, triglycerides and/or phospholipids, highly susceptible to endotoxemia and Klebsiella pneumoniae infection. J. Lipid Res. 1999. 40: 680–685. leading to binding and neutralization of LPS (9, 18). In a previous study with LDL receptor-deficient (LDLr2/2) mice, we demonstrated that these mice with a Supplementary key words lipopolysaccharide • tumor necrosis factor • apolipoprotein E • hypercholesterolemia • neutralization hypercholesterolemia due to 7- to 9-fold increase of inter- mediate density lipoprotein (IDL) and LDL are protected against Gram-negative infections (8). The aim of the present study was to investigate the relative importance of Endotoxin, the lipopolysaccharide (LPS) component of the lipid component of VLDL and apoE for binding and the outer membrane of gram-negative bacteria, is the major neutralization of LPS. For this purpose, we investigated the pathogenic factor in gram-negative sepsis (1). When infused in vivo, LPS induces hypotension, disseminated intravascu- lar coagulation, and renal, hepatic and cerebral damage, Abbreviations: VLDL, very low density lipoproteins; LDL, low den- which may lead to shock and death. These actions are medi- sity lipoproteins; HDL, high density lipoproteins; apo E2/2, apolipo- ated by the production and release of proinflammatory cy- protein E deficient; LDLr2/2, LDL receptor deficient; CFU, colony tokines such as tumor necrosis factor a (TNFa) and inter- forming units. leukin-1b (IL-1b) by monocytes and macrophages. To whom correspondence should be addressed. 680 Journal of Lipid Research Volume 40, 1999 This is an Open Access article under the CC BY license. described elsewhere (8), upon stimulation with 100 ng/ml LPS susceptibility to endotoxin and K. pneumoniae of mice with a for 24 h. Samples were stored at 270 8C until assay. gene disruption for apo E. Due to the lack of apoE, these mice have hyperlipidemia as a result of decreased clearance Klebsiella pneumoniae infection K. pneumoniae and high levels of VLDL (19). As cholesterol-rich LDL and (ATCC 43816), a strain that produces a lethal infection in nor- HDL are able to neutralize LPS (12), one would expect 5 mal mice, was injected iv into the retro-orbital plexus (0.5 3 10 cholesterol-rich VLDL to neutralize LPS as well, rendering CFU/mouse) of apoE2/2 and control mice. After 90 min, sub- these mice resistant to LPS like the hypercholesterolemic groups of five mice were killed and blood was collected from the retro-orbital plexus for the measurement of cytokine concentra- LDLr2/2 mice (8). tions. In a separate group of mice, survival was assessed twice a day. Cytokine measurements MATERIALS AND METHODS TNFa, IL-1a, and IL-1b concentrations were measured by spe- cific radioimmunoassays (RIAs) described previously (21). IL-6 Animals concentrations were determined using a commercial ELISA kit Homozygous apolipoprotein E-deficient (apoE2/2) mice (6– (Biosource, Europe S.A.). Detection limits were 0.02 ng/ml for 8 weeks old) were used (Transgenic Facility of Leiden University IL-1a and IL-1b, and 0.04 ng/ml for TNFa. The accuracy of the Medical Center, Leiden, The Netherlands (19)). For the experi- cytokine assays was determined using reference preparations: ments, weight-matched C57BL/6J mice were selected. The ani- murine recombinant TNFa, IL-1a, and IL-1b obtained from the mals were kept under specific pathogen-free conditions. The ex- National Institute of Biological Standards and Control (NIBSC, periments were approved by the ethics committee on animal Hertfordshire, UK), for the IL-6 ELISA and, for the RPMI sam- experiments of the Catholic University Nijmegen. ples, control pools were made at our laboratory. All samples of each experiment were analyzed in the same run in duplicate. Us- Lipid measurements ing various batches of tracers the bias of all cytokine assays was ,20%; interassay variation ranged between 10 and 13% depend- For all mice, total and lipoprotein specific cholesterol and tri- ing on the concentration (n 5 22). glycerides in EDTA plasma were determined enzymatically with a Hitachi 747 analyzer 4 days preceeding the experiments men- LPS measurements tioned below. For determination of chemical composition and li- To determine whether the clearance of LPS in apoE2/2 mice poprotein cholesterol and triglyceride concentrations, samples is different from the clearance of LPS in C57BL/6 mice, six ani- from five mice were pooled. To isolate the VLDL 1 IDL fraction, mals of each group were bled 10 min after LPS injection and five plasma was brought to a density of 1.019 g/ml with d 5 1.10 g/ animals of each group after 90 min and 4 h into LPS-free EDTA- ml and centrifuged for 18 h at 36,000 g. After aspiration of the containing tubes. Tubes were centrifuged for 5 min at 13,000 g upper layer, the d . 1.019 g/ml fraction was brought to a density and plasma was transferred to clean LPS-free tubes. Subsequently, of 1.070 g/ml with d 5 1.225 g/ml to isolate the LDL fraction a chromogenic LAL assay (Endosafe Inc., Charleston, SC) was used during 24 h at 36,000 g. After aspirating LDL from the top layer, to measure endotoxin concentrations. Assays were performed HDL was separated from the serum proteins by centrifugation according to the manufacturer’s instructions. for 24 h at 39,000 g after raising the density to 1.18 g/ml. Subse- quently, cholesterol and triglyceride concentrations in all of the LPS neutralizing capacity lipoprotein fractions were determined with the Hitachi 747 ana- In order to assess the capacity of plasma of apoE2/2 and lyzer. Phospholipids and free cholesterol in all fractions were C57BL/6 mice to neutralize exogenous LPS in vitro, blood of measured using a kit (Boehringer Mannheim, Germany). Finally, apoE2/2 mice and their controls was drawn by cardiac punc- protein concentrations were ascertained by the method of Lowry ture and collected in LPS-free EDTA-containing tubes (Beckton et al. (20) and cholesteryl esters were calculated by subtracting Dickinson). Tubes were centrifuged for 10 min at 1800 g. Subse- free cholesterol from total cholesterol concentration. quently, plasma was centrifuged in an Eppendorf centrifuge at 13,000 g to acquire platelet-free plasma. LPS (100 ng/ml) was Lethality studies preincubated with the plasma at 37 8C. After 24 h, human periph- Lipopolysaccharide (LPS; Escherichia coli Serotype 055:B5) was eral blood mononuclear cells (obtained by Ficoll-separation) obtained from Sigma Chemical Co. (St. Louis, MO). Groups of at were added to the LPS and incubated at 378C once again. After least 10 mice were injected iv into the retro-orbital plexus with an additional 24 h incubation, cytokines in the supernatant were LPS (dose: 2 mg in 100 ml PBS/mouse). Survival in both groups measured by RIA and the production was expressed per ml of was assessed daily for at least 7 days. 6 medium containing 10 cells. Statistical analysis LPS-induced cytokine production in vivo Survival data were analyzed using the Kaplan-Meyer log rank In 3 separate experiments, apoE2/2 and control mice were test. Differences in concentrations of cytokines were analyzed us- injected iv into the retro-orbital plexus with LPS (50 mg in 100 ml ing the Mann-Whitney U test. Differences in lipid concentrations PBS/mouse). After 90 min (for TNFa production) and 4 h (for were analyzed with the Students’ t -test. Differences were consid- IL-1a and IL-1b production), groups of at least five animals were ered significant at P , 0.05. bled from the retro-orbital plexus into EDTA-containing tubes. Tubes were centrifuged for 5 min at 13,000 g and cytokines were measured in the plasma. RESULTS Ex vivo cytokine production Concentrations of lipids, lipoproteins, and relative In three separate experiments, resident peritoneal macro- chemical composition phages were isolated from groups of at least five apoE2/2 and Total plasma cholesterol concentrations were more than control mice. Subsequently, secreted and cell-associated produc- tion of TNFa, IL-1a, and IL-1b by these cells was determined as 8 times higher in apoE2/2 mice than in control C57BL/6 de Bont et al. Endotoxin susceptibility of apoE-deficient mice 681 TABLE 1. Plasma and lipoprotein cholesterol and triglyceride ity compared to control mice. All apoE2/2 mice died, concentrations (mmol/l) in apoE2/2 and control mice whereas only 40% of the controls were killed by the in- fection (Fig. 4). Ninety minutes after a K. pneumoniae Cholesterol Triglycerides inoculum of 5 3 10 CFU TNF concentration in the cir- ApoE2/2 Controls ApoE2/2 Controls culation of apoE2/2 mice was higher than those of mmol/l mmol/l control mice (12.3 6 3.4 ng/ml vs. 7.6 6 3.8 ng/ml), al- a b Plasma 16.14 6 3.73 1.87 6 0.15 1.13 6 0.40 0.85 6 0.30 though these differences did not reach statistical signifi- a c VLDL 10.2 6 1.14 0.12 6 0.03 0.81 6 0.14 0.48 6 0.16 cance (P 5 0.056). LDL 0.60 6 0.06 0.23 6 0.03 0.03 6 0.01 0.04 6 0.01 HDL 0.47 6 0.10 1.04 6 0.10 0.03 6 0.01 0.03 6 0.01 Ex vivo cytokine production Results are given as mean 6 SD for 30 animals per group. To investigate whether the increased in vivo production a b c P , 0.0001, P , 0.001, P , 0.01 (for comparison between of TNFa was due to an increased capacity of peritoneal C57BL/6 and apo E2/2). macrophages to produce cytokines, cells of both mouse strains were stimulated with LPS. No differences were found in either secreted or cell-associated TNFa, IL-1a, or mice, due to extremely high cholesterol concentrations of the VLDL + IDL fraction and, to a lesser extent, to LDL IL-1b between both strains (Table 2). cholesterol concentrations, whereas HDL concentration LPS clearance was lower (Table 1). Total triglyceride concentrations To test whether the LPS clearance differed in the two were slightly, though significantly, higher in the apoE2/2 mouse strains, we assessed LPS concentrations in the cir- mice (Table 1). The higher triglyceride concentrations culation. Ten min after an iv injection of 50 mg LPS, were reflected in the triglyceride concentrations of the apoE2/2 mice showed levels of endotoxin in plasma VLDL 1 IDL fraction (Table 1). similar to control mice in the two mouse strains (3.69 6 The chemical composition of all lipoprotein fractions 0.80 vs. 3.75 6 1.21 mg/ml). After 90 min, apoE2/2 showed differences (Fig. 1), especially in the cholesterol mice showed lower, albeit not significant, levels of endo- and triglyceride content of the VLDL 1 IDL and the LDL toxin in plasma compared with control animals (0.51 6 fraction. In the apoE2/2 mice, cholesterol represented the 0.03 vs. 0.66 6 0.21 mg/ml). Again no differences were most abundant component of lipoproteins, whereas control seen 4 h after LPS injection (0.34 6 0.03 vs. 0.37 6 0.09 mice have more triglycerides in these fractions (Fig. 1). mg/ml). Lethality and cytokines in vivo LPS neutralizing capacity Survival after an endotoxin injection is depicted in Fig. 2. To investigate whether the lipoproteins in the plasma of The mortality in the apoE2/2 mice (86%) was significantly apoE2/2 and control mice were equally potent in bind- higher than in controls (41%). Plasma concentrations of ing and neutralization of LPS, preincubation of LPS with TNFa 90 min after LPS administration were more than 4- plasma from the two mouse strains was performed. After fold higher in apoE2/2 mice than in control animals (Fig. 24 h, human PBMC were added and cytokine production 3). The circulating concentrations of IL-1a, IL-1b, and IL-6 measured after an additional 24 h incubation. IL-1b and were 10- to 100-fold lower than those of TNF, and no differ- TNFa production by PBMC were higher after preincuba- ences were detected 4 h after LPS injection (Fig. 3). tion of LPS with plasma obtained from apoE2/2 mice Klebsiella pneumoniae infection than with control plasma (Table 3), compatible with de- After an intravenous injection of 0.5 3 10 CFU of K. creased LPS-neutralizing capacity of lipoproteins from pneumoniae, apoE2/2 mice showed an increased mortal- apoE2/2 mice. Fig. 1. Chemical composition of lipoprotein fractions (VLDL 1 IDL, LDL and HDL) of apoE2/2 (solid bars) and control mice (open bars). The figure represents the percentage of free cholesterol (fc), cholesteryl esters (ce), triglycer- ides (tg), phospholipids (pl), and protein of each fraction. 682 Journal of Lipid Research Volume 40, 1999 Fig. 2. Survival of apoE2/2 (r) and control mice (h) after LPS (2 mg/mouse iv). Sur vival was signifi- cantly impaired in the apoE2/2 mice. The figure shows pooled data of two experiments with at least 10 mice per group each; P , 0.01. tralization of endotoxin: the increased lipid concentra- DISCUSSION tion on the one hand, and the apolipoproteins on the In the present study we show that hyperlipidemic mice other. Both apoE2/2 and LDLr2/2 mice have increased deficient in apolipoprotein E are more susceptible to en- plasma lipid concentrations due to either increased cho- dotoxemia and to Klebsiella pneumoniae infection than con- lesterol-rich VLDL in apoE2/2 mice or high IDL and LDL trol mice. In the apoE2/2 mice, the severe cytokinemia, in LDLr2/2 mice (23). Because both strains have similar in particular TNFa, is most probably responsible for elevations in lipoprotein concentrations, and all lipopro- death. As macrophages of apoE2/2 and control mice tein subfractions have been shown to neutralize LPS (2–7), produced similar amounts of TNFa, the differences do the discrepancy in LPS neutralization cannot easily be ex- not seem to be due to differences in intrinsic cytokine plained by quantitative differences in hyperlipidemia. production capacity. The apoE2/2 plasma apparently exerts an impaired These results are in accordance with those of Roselaar LPS-neutralizing capacity despite high cholesterol. This is and Daugherty (22) who demonstrated that apolipoprotein in accordance with results of Harris et al. (12) showing E-deficient mice are more susceptible to Listeria monocyto- that cholesterol is not necessary for the interaction genes. However, their results are in marked contrast to those between lipoproteins and LPS, as a cholesterol-free lipid we obtained in hyperlipidemic LDL receptor knock-out emulsion, Soyacal also protected against endotoxin- (LDLr2/2) mice that had an increased survival towards a induced death. On the other hand, infusion of a com- Gram-negative challenge and a dampered proinflammatory mercially produced triglyceride-rich particle, Intralipid , endotoxin response by virtue of the endotoxin-neutralizing in humans did not result in a reduced cytokine release effects of lipoproteins (8). Moreover, the plasma of the (24). apoE2/2 mice appeared to have a low LPS-neutralizing ca- It should be taken into account that the lipoproteins in pacity, which was even less than that of normolipidemic con- the VLDL-sized fraction of apoE2/2 mice have a differ- trol mice. How should these differences be explained? ent chemical composition when compared to controls. Basically, two mechanisms may be responsible for neu- Conformational changes may have taken place, resulting Fig. 3. Plasma cytokine concentrations in apoE2/2 (black columns) and control mice (white columns) af- ter 50 mg of LPS iv. The plasma concentrations of TNFa (90 min after the LPS challenge) was significantly higher in apoE2/2 mice compared with control mice. No differences in IL 1a, IL-1b, or IL-6 plasma concen- trations (4 h after LPS challenge) were detected. The figure shows pooled data of three experiments with groups of at least 5 animals. Data are represented as proportion of the mean cytokine production of control animals; *P , 0.0001. de Bont et al. Endotoxin susceptibility of apoE-deficient mice 683 Fig. 4. Survival of apoE2/2 and control mice after iv infection with Klebsiella pneumoniae. An iv injection of 0.5 3 10 CFU K. pneumoniae was given to apoE2/2 (r) and control mice (h). Survival was significantly im- paired in apoE2/2 mice. The figure shows data of one experiment with 10 mice per group (P , 0.01). in a diminished availability of LPS-binding sites and a de- clearance from the plasma between the two mouse strains creased neutralizing capacity. indicates that the outcome after an LPS injection is deter- The second possible mechanism concerns the apolipo- mined in the liver and not in the blood. So, most proba- proteins. Both free apoA-I and apoE have strong LPS neu- bly, the LPS in apoE deficiency is only directed towards cy- tralizing effects (14–17). In apoE2/2 mice, the VLDL 1 tokine-producing Kupffer cells, whereas in the control IDL fraction is relatively poor in apolipoproteins. In addi- mice a part is directed towards parenchymal cells. How tion, the excess of lipids carried by the lipoprotein particles should we explain that we did not find differences in circu- in these mice may result in a change of the LPS-binding lating IL-1a, IL-1b, or IL-6 4 h after LPS injection? One ex- sites on their apolipoproteins. This, together with a 2-fold planation could be that TNF and IL-1 production are differ- decrease in HDL, may be an explanation why LPS in entially regulated (27). We have recently found that different apoE2/2 mice is neutralized to a limited extent. In addi- LPS receptors seem to be responsible for the production of tion to the LPS-binding capacity of apoproteins, apoE the various cytokines: CD14-dependent mechanisms medi- seems to be able to neutralize LPS differently. The LPS- ate TNFa production, while IL-1 synthesis is induced by detoxifying effects of chylomicrons are thought to be me- both CD14-dependent and CD14-independent mecha- diated by apoE present on this lipoprotein (14). Free en- nisms (28). Secondly, IL-1a and IL-1b are less prominent dotoxin, injected iv, is cleared from the plasma by the as circulating cytokines than TNFa. liver, where it is predominantly taken up by the Kupffer Let us now return to the question how the discrepancy cells resulting in production of cytokines (25, 26). How- in LPS neutralization between the two hyperlipidemic ever, when bound to apoE (free or associated with chylo- mouse strains should be explained. In comparison with microns), LPS can be shunted away from these Kupffer C57BL/6J mice, LDLr2/2 mice have extremely high IDL cells and directed to the parenchymal liver cells (14), re- and LDL concentrations and a 2-fold increased HDL- sulting in a reduction of peak cytokine serum levels (17). cholesterol as well (data not shown). ApoE is present on These observations are in accordance with ours, showing both IDL and HDL, hence increased in LDLr2/2 mice. that in apoE2/2 mice TNFa plasma concentrations were Therefore, absence of apoE in apoE2/2 but not in 4- to 5-fold higher than in controls after LPS challenge. LDLr2/2 mice may indeed explain the difference in out- Moreover, the fact that no difference was found in LPS come between the two mouse strains. TABLE 2. Secreted and cell-associated cytokine production after in vitro stimulation of peritoneal macrophages TABLE 3. Effect of apoE2/2 and control plasma on of apoE2/2 and C57BL/6 mice LPS-induced cytokine production Secreted Cell-associated TNFa IL-1b C57BL/6 ApoE2/2 C57BL/6 ApoE2/2 ApoE2/2 Control ApoE2/2 Control ng/ml ng/ml ng/ml ng/ml IL-1a 0.13 6 0.03 0.10 6 0.01 3.62 6 1.29 2.90 6 1.29 1 0.59 0.38 2.48 1.31 IL-1b 0.07 6 0.02 0.05 6 0.01 0.29 6 0.07 0.30 6 0.11 2 0.27 0.22 1.81 0.65 TNFa 0.79 6 0.19 0.73 6 0.20 0.19 6 0.04 0.24 6 0.08 3 0.44 0.36 1.88 0.84 Peritoneal macrophages (10 /well) were stimulated with LPS (100 LPS (100 ng/ml) preincubated with either plasma of apoE2/2 or ng/ml) for 24 h at 378C. The data represent mean 6 SD for six animals. of control mice was added to 5 3 10 PBMC/well and incubated for 24 The results do not differ significantly for the three cytokines measured. h at 37 8C (n 5 3). 684 Journal of Lipid Research Volume 40, 1999 Human very low density lipoproteins and chylomicrons can pro- In conclusion, in this study we demonstrated that hy- tect against endotoxin-induced death in mice. J. Clin. Invest. 86: percholesterolemia itself is not sufficient for protection 696–702. against endotoxemia. More likely, the presence of apoE is 13. Read, T. E., C. Grunfeld, Z. L. Kumwenda, M. C. Calhoun, J. P. essential in the process of LPS detoxification, either by cat- Kane, K. R. Feingold, and J. H. Rapp. 1995. Triglyceride-rich li- poproteins prevent septic death in rats. J. Exp. Med. 182: 267– alyzing the binding of LPS to the lipoprotein particle or by directing the LPS to the parenchymal cells away from cyto- 14. Rensen, P. C., M. Oosten, E. Bilt, M. Eck, J. Kuiper, and T. J. Ber- kine-producing Kupffer cells or by both mechanisms. kel. 1997. Human recombinant apolipoprotein E redirects lipo- polysaccharide from Kupffer cells to liver parenchymal cells in rats in vivo. J. Clin. Invest. 99: 2438–2445. 15. Flegel, W. A., M. W. Baumstark, C. Weinstock, A. Berg, and H. The authors thank Geert Poelen and Theo van der Ing for Northoff. 1993. Prevention of endotoxin-induced monokine re- their excellent assistance with the animal experiments. lease by human low and high density lipoproteins and by apolipo- Manuscript received 23 June 1998 and in revised form 26 October 1998. protein A-I. Infect. Immun. 61: 5140–5146. 16. Emancipator, K., G. Csako, and R. J. Elin. 1992. In vitro inactiva- tion of bacterial endotoxin by human lipoproteins and apolipo- proteins. Infect. Immun. 60: 596–601. 17. Rensen, P. C. N., M. van Oosten, E. van de Bilt, M. van Eck, J. REFERENCES Kuiper, and T. J. C. van Berkel. 1997. Human apolipoprotein E protects against the deleterious effects of bacterial lipopolysaccha- 1. Lynn, W. A., and J. Cohen. 1995. Adjunctive therapy for septic ride in rats in vivo. Circulation. 379 (Abstract). shock: a review of experimental approaches. Clin. Infect. Dis. 20: 18. Parker, T. S., D. M. Levine, J. C. Chang, J. Laxer, C. C. Coffin, and 143–158. A. L. Rubin. 1995. Reconstituted high-density lipoprotein neutral- 2. Van Lenten, B. J., A. M. Fogelman, M. E. Haberland, and P. A. Ed- izes gram-negative bacterial lipopolysaccharides in human whole wards. 1986. The role of lipoproteins and receptor-mediated en- blood. Infect. Immun. 63: 253–258. docytosis in the transport of bacterial lipopolysaccharide. Proc. 19. van Ree, J. H., W. J. van den Broek, V. E. Dahlmans, P. H. Groot, M. Natl. Acad. Sci. USA. 83: 2704–2708. Vidgeon-Hart, R. R. Frants, B. Wieringa, L. M. Havekes, and M. H. 3. Harris, H. W., C. Grunfeld, K. R. Feingold, T. E. Read, J. P. Kane, Hofker. 1994. Diet-induced hypercholesterolemia and atheroscle- A. L. Jones, E. B. Eichbaum, G. F. Bland, and J. H. Rapp. 1993. rosis in heterozygous apolipoprotein E-deficient mice. Atherosclero- Chylomicrons alter the fate of endotoxin, decreasing tumor necro- sis. 111: 25–37. sis factor release and preventing death. J. Clin. Invest. 91: 1028– 20. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. 1034. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 4. Flegel, W. A., A. Wolpl, D. N. Mannel, and H. Northoff. 1989. Inhi- 193: 265–275. bition of endotoxin-induced activation of human monocytes by 21. Vogels, M. T., E. J. Mensink, K. Ye, O. C. Boerman, C. M. Ver- human lipoproteins. Infect. Immun. 57: 2237–2245. schueren, C. A. Dinarello, and J. W. van der Meer. 1994. Differen- 5. Netea, M. G., N. de Bont, P. N. M. Demacker, B. J. Kullberg, L. E. H. tial gene expression for IL-1 receptor antagonist, IL-1, and TNF re- Jacobs, T. J. G. Verver-Jansen, A. F. H. Stalenhoef, and J. W. M. van ceptors and IL-1 and TNF synthesis may explain IL-1-induced der Meer. 1998. Lipoprotein[a] inhibits the lipopolysaccharide- resistance to infection. J. Immunol. 153: 5772–5780. induced tumor necrosis factor-alpha production by human mono- 22. Roselaar, S. E., and A. Daugherty. 1998. Apolipoprotein E-defi- nuclear cells. Infect. Immun. 66: 2365–2367. cient mice have impaired innate immune responses to Listeria 6. Cavaillon, J. M., C. Fitting, N. Haeffner-Cavaillon, S. J. Kirsch, and monocytogenes in vivo. J. Lipid. Res. 39: 1740–1743. H. S. Warren. 1990. Cytokine response by monocytes and mac- 23. Ishibashi, S., M. S. Brown, J. L. Goldstein, R. D. Gerard, R. E. rophages to free and lipoprotein-bound lipopolysaccharide. Infect. Hammer, and J. Herz. 1993. Hypercholesterolemia in low density Immun. 58: 2375–2382. lipoprotein receptor knockout mice and its reversal by adenovi- 7. Weinstock, C., H. Ullrich, R. Hohe, A. Berg, M. W. Baumstark, I. rus-mediated gene delivery [see comments]. J. Clin. Invest. 92: Frey, H. Northoff, and W. A. Flegel. 1992. Low density lipoproteins 883–893. inhibit endotoxin activation of monocytes. Arterioscler. Thromb. 12: 24. van der Poll, T., C. C. Braxton, S. M. Coyle, M. A. Boermeester, J. C. 341–347. Wang, P. M. Jansen, W. J. Montegut, S. E. Calvano, C. E. Hack, and 8. Netea, M. G., P. N. Demacker, B. J. Kullberg, O. C. Boerman, I. S. F. Lowry. 1995. Effect of hypertriglyceridemia on endotoxin re- Verschueren, A. F. Stalenhoef, and J. W. van der Meer. 1996. Low- sponsiveness in humans. Infect. Immun. 63: 3396–3400. density lipoprotein receptor-deficient mice are protected against 25. Praaning van Dalen, D. P., A. Brouwer, and D. L. Knook. 1981. lethal endotoxemia and severe gram-negative infections. J. Clin. Clearance capacity of rat liver Kupffer, endothelial, and parenchy- Invest. 97: 1366–1372. mal cells. Gastroenterology. 81: 1036–1044. 9. Levine, D. M., T. S. Parker, T. M. Donnelly, A. Walsh, and A. L. Ru- 26. Munford, R. S., J. M. Andersen, and J. M. Dietschy. 1981. Sites of bin. 1993. In vivo protection against endotoxin by plasma high tissue binding and uptake in vivo of bacterial lipopolysaccharide– density lipoprotein. Proc. Natl. Acad. Sci. USA. 90: 12040–12044. high density lipoprotein complexes: studies in the rat and squirrel 10. Hubsch, A. P., F. S. Powell, P. G. Lerch, and J. E. Doran. 1993. A re- monkey. J. Clin. Invest. 68: 1503–1513. constituted, apolipoprotein A-I containing lipoprotein reduces tu- 27. Zuckerman, S. H., and G. F. Evans. 1992. Endotoxin tolerance: in mor necrosis factor release and attenuates shock in endotoxemic vivo regulation of tumor necrosis factor and interleukin-1 synthe- rabbits. Circ. Shock. 40: 14–23. sis is at the transcriptional level. Cell. Immunol. 140: 513–519. 11. Pajkrt, D., J. E. Doran, F. Koster, P. G. Lerch, B. Arnet, T. van der 28. Netea, M. G., B. J. Kullberg, and J. W. M. van der Meer. 1998. Lipo- Poll, J. W. ten Cate, and S. J. H. van Deventer. 1996. Antiinflamma- polysaccharide-induced production of tumour necrosis factor and tory effects of reconstituted high density lipoprotein during hu- interleukin-1 is differtially regulated at the receptor level: the role man endotoxemia. J. Exp. Med. 184: 1601–1608. of CD14-dependent and CD14-independent pathways. Immunology. 12. Harris, H. W., C. Grunfeld, K. R. Feingold, and J. H. Rapp. 1990. 94: 340–344. de Bont et al. Endotoxin susceptibility of apoE-deficient mice 685 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Lipid Research Unpaywall

Apolipoprotein E knock-out mice are highly susceptible to endotoxemia and Klebsiella pneumoniae infection

Download PDF
6 pages

Loading next page...
 
/lp/unpaywall/apolipoprotein-e-knock-out-mice-are-highly-susceptible-to-endotoxemia-0KUsPm85Gl

References (28)

Publisher
Unpaywall
ISSN
0022-2275
DOI
10.1016/s0022-2275(20)32147-7
Publisher site
See Article on Publisher Site

Abstract

Apolipoprotein E knock-out mice are highly susceptible to endotoxemia and Klebsiella pneumoniae infection Natasja de Bont,* Mihai G. Netea,* Pierre N. M. Demacker,* Ineke Verschueren,* † 1, Bart Jan Kullberg,* Ko Willems van Dijk, Jos W. M. van der Meer,* and Anton F. H. Stalenhoef Division of General Internal Medicine,* Department of Medicine, University Hospital Nijmegen, 6500 HB Nijmegen, The Netherlands, and Department of Human Genetics, Leiden University Medical Centre, Leiden, The Netherlands Abstract Lipoproteins are able to neutralize bacterial lipo- LPS forms complexes with all major lipoprotein classes polysaccharide (LPS) and thereby inhibit the proinflamma- in plasma, including chylomicrons, very low density lipo- tory cytokine response. In a previous study, we demonstrated proteins (VLDL), low density lipoproteins (LDL), high that hypercholesterolemic low density lipoprotein receptor density lipoproteins (HDL), and lipoprotein [a] (LP[a]) knock-out (LDLr2/2) mice are protected against lethal en- (2–5). Binding of endotoxin to any of these lipoproteins dotoxemia and gram-negative infection. In the present study results in neutralization of LPS and reduced release of we investigated the susceptibility of apolipoprotein E knock- TNFa, IL-1b, and IL-6 (4, 6, 7). Recent studies have shown out mice (apoE2/2) to LPS and to Klebsiella pneumoniae. that hyperlipoproteinemia in genetically modified mice These mice have increased plasma lipoprotein concentra- tions in the very low density lipoprotein (VLDL)-sized frac- (8, 9) as well as infusion of reconstituted HDL (rHDL) in tion. Despite 8-fold higher plasma cholesterol levels com- humans (10, 11) results in a diminished cytokine response, pared to controls, and in contrast to LDLr2/2 mice, leading to protection against the toxic effects of endotox- apoE2/2 mice were significantly more susceptible to endo- emia. Other studies have demonstrated that triglyceride- toxemia and to K. pneumoniae infection. Circulating TNFa rich lipoproteins, such as chylomicrons and VLDL, also concentrations after intravenously injected LPS were 4- to have the capacity to inactivate LPS and prevent endo- 5-fold higher in apoE2/2 mice, whereas IL-1a, IL-1b, and toxin-induced death in rodents (3, 12–14). IL-6 did not differ. This TNF response was not due to an in- It is not well understood which lipoprotein component creased cytokine production capacity of cells from apoE2/2 is most important for LPS neutralization. The ability of li- mice, as ex vivo cytokine production in response to LPS did not differ between apoE2/2 and control mice. The LPS- poproteins to bind and inactivate LPS is modulated by neutralizing capacity of apoE2/2 plasma was significantly apolipoproteins. Both apolipoprotein (apo) A-I, present less than that of controls. Most likely, the absence of apoE it- on HDL, and apoE, present on chylomicrons, VLDL, self in the knock-out mice explains the failure to neutralize HDL, and IDL, are capable of directly inactivating endo- LPS, despite the very high cholesterol concentrations.—de toxin and decreasing LPS-induced cytokine release (14 – Bont, N., M. G. Netea, P. N. M. Demacker, I. Verschueren, 17). An additional mechanism may be represented by the B. J. Kullberg, K. W. van Dijk, J. W. M. van der Meer, and lipid–lipid interactions between the lipid A component of A. F. H. Stalenhoef. Apolipoprotein E knockout-mice are LPS and cholesterol, triglycerides and/or phospholipids, highly susceptible to endotoxemia and Klebsiella pneumoniae infection. J. Lipid Res. 1999. 40: 680–685. leading to binding and neutralization of LPS (9, 18). In a previous study with LDL receptor-deficient (LDLr2/2) mice, we demonstrated that these mice with a Supplementary key words lipopolysaccharide • tumor necrosis factor • apolipoprotein E • hypercholesterolemia • neutralization hypercholesterolemia due to 7- to 9-fold increase of inter- mediate density lipoprotein (IDL) and LDL are protected against Gram-negative infections (8). The aim of the present study was to investigate the relative importance of Endotoxin, the lipopolysaccharide (LPS) component of the lipid component of VLDL and apoE for binding and the outer membrane of gram-negative bacteria, is the major neutralization of LPS. For this purpose, we investigated the pathogenic factor in gram-negative sepsis (1). When infused in vivo, LPS induces hypotension, disseminated intravascu- lar coagulation, and renal, hepatic and cerebral damage, Abbreviations: VLDL, very low density lipoproteins; LDL, low den- which may lead to shock and death. These actions are medi- sity lipoproteins; HDL, high density lipoproteins; apo E2/2, apolipo- ated by the production and release of proinflammatory cy- protein E deficient; LDLr2/2, LDL receptor deficient; CFU, colony tokines such as tumor necrosis factor a (TNFa) and inter- forming units. leukin-1b (IL-1b) by monocytes and macrophages. To whom correspondence should be addressed. 680 Journal of Lipid Research Volume 40, 1999 This is an Open Access article under the CC BY license. described elsewhere (8), upon stimulation with 100 ng/ml LPS susceptibility to endotoxin and K. pneumoniae of mice with a for 24 h. Samples were stored at 270 8C until assay. gene disruption for apo E. Due to the lack of apoE, these mice have hyperlipidemia as a result of decreased clearance Klebsiella pneumoniae infection K. pneumoniae and high levels of VLDL (19). As cholesterol-rich LDL and (ATCC 43816), a strain that produces a lethal infection in nor- HDL are able to neutralize LPS (12), one would expect 5 mal mice, was injected iv into the retro-orbital plexus (0.5 3 10 cholesterol-rich VLDL to neutralize LPS as well, rendering CFU/mouse) of apoE2/2 and control mice. After 90 min, sub- these mice resistant to LPS like the hypercholesterolemic groups of five mice were killed and blood was collected from the retro-orbital plexus for the measurement of cytokine concentra- LDLr2/2 mice (8). tions. In a separate group of mice, survival was assessed twice a day. Cytokine measurements MATERIALS AND METHODS TNFa, IL-1a, and IL-1b concentrations were measured by spe- cific radioimmunoassays (RIAs) described previously (21). IL-6 Animals concentrations were determined using a commercial ELISA kit Homozygous apolipoprotein E-deficient (apoE2/2) mice (6– (Biosource, Europe S.A.). Detection limits were 0.02 ng/ml for 8 weeks old) were used (Transgenic Facility of Leiden University IL-1a and IL-1b, and 0.04 ng/ml for TNFa. The accuracy of the Medical Center, Leiden, The Netherlands (19)). For the experi- cytokine assays was determined using reference preparations: ments, weight-matched C57BL/6J mice were selected. The ani- murine recombinant TNFa, IL-1a, and IL-1b obtained from the mals were kept under specific pathogen-free conditions. The ex- National Institute of Biological Standards and Control (NIBSC, periments were approved by the ethics committee on animal Hertfordshire, UK), for the IL-6 ELISA and, for the RPMI sam- experiments of the Catholic University Nijmegen. ples, control pools were made at our laboratory. All samples of each experiment were analyzed in the same run in duplicate. Us- Lipid measurements ing various batches of tracers the bias of all cytokine assays was ,20%; interassay variation ranged between 10 and 13% depend- For all mice, total and lipoprotein specific cholesterol and tri- ing on the concentration (n 5 22). glycerides in EDTA plasma were determined enzymatically with a Hitachi 747 analyzer 4 days preceeding the experiments men- LPS measurements tioned below. For determination of chemical composition and li- To determine whether the clearance of LPS in apoE2/2 mice poprotein cholesterol and triglyceride concentrations, samples is different from the clearance of LPS in C57BL/6 mice, six ani- from five mice were pooled. To isolate the VLDL 1 IDL fraction, mals of each group were bled 10 min after LPS injection and five plasma was brought to a density of 1.019 g/ml with d 5 1.10 g/ animals of each group after 90 min and 4 h into LPS-free EDTA- ml and centrifuged for 18 h at 36,000 g. After aspiration of the containing tubes. Tubes were centrifuged for 5 min at 13,000 g upper layer, the d . 1.019 g/ml fraction was brought to a density and plasma was transferred to clean LPS-free tubes. Subsequently, of 1.070 g/ml with d 5 1.225 g/ml to isolate the LDL fraction a chromogenic LAL assay (Endosafe Inc., Charleston, SC) was used during 24 h at 36,000 g. After aspirating LDL from the top layer, to measure endotoxin concentrations. Assays were performed HDL was separated from the serum proteins by centrifugation according to the manufacturer’s instructions. for 24 h at 39,000 g after raising the density to 1.18 g/ml. Subse- quently, cholesterol and triglyceride concentrations in all of the LPS neutralizing capacity lipoprotein fractions were determined with the Hitachi 747 ana- In order to assess the capacity of plasma of apoE2/2 and lyzer. Phospholipids and free cholesterol in all fractions were C57BL/6 mice to neutralize exogenous LPS in vitro, blood of measured using a kit (Boehringer Mannheim, Germany). Finally, apoE2/2 mice and their controls was drawn by cardiac punc- protein concentrations were ascertained by the method of Lowry ture and collected in LPS-free EDTA-containing tubes (Beckton et al. (20) and cholesteryl esters were calculated by subtracting Dickinson). Tubes were centrifuged for 10 min at 1800 g. Subse- free cholesterol from total cholesterol concentration. quently, plasma was centrifuged in an Eppendorf centrifuge at 13,000 g to acquire platelet-free plasma. LPS (100 ng/ml) was Lethality studies preincubated with the plasma at 37 8C. After 24 h, human periph- Lipopolysaccharide (LPS; Escherichia coli Serotype 055:B5) was eral blood mononuclear cells (obtained by Ficoll-separation) obtained from Sigma Chemical Co. (St. Louis, MO). Groups of at were added to the LPS and incubated at 378C once again. After least 10 mice were injected iv into the retro-orbital plexus with an additional 24 h incubation, cytokines in the supernatant were LPS (dose: 2 mg in 100 ml PBS/mouse). Survival in both groups measured by RIA and the production was expressed per ml of was assessed daily for at least 7 days. 6 medium containing 10 cells. Statistical analysis LPS-induced cytokine production in vivo Survival data were analyzed using the Kaplan-Meyer log rank In 3 separate experiments, apoE2/2 and control mice were test. Differences in concentrations of cytokines were analyzed us- injected iv into the retro-orbital plexus with LPS (50 mg in 100 ml ing the Mann-Whitney U test. Differences in lipid concentrations PBS/mouse). After 90 min (for TNFa production) and 4 h (for were analyzed with the Students’ t -test. Differences were consid- IL-1a and IL-1b production), groups of at least five animals were ered significant at P , 0.05. bled from the retro-orbital plexus into EDTA-containing tubes. Tubes were centrifuged for 5 min at 13,000 g and cytokines were measured in the plasma. RESULTS Ex vivo cytokine production Concentrations of lipids, lipoproteins, and relative In three separate experiments, resident peritoneal macro- chemical composition phages were isolated from groups of at least five apoE2/2 and Total plasma cholesterol concentrations were more than control mice. Subsequently, secreted and cell-associated produc- tion of TNFa, IL-1a, and IL-1b by these cells was determined as 8 times higher in apoE2/2 mice than in control C57BL/6 de Bont et al. Endotoxin susceptibility of apoE-deficient mice 681 TABLE 1. Plasma and lipoprotein cholesterol and triglyceride ity compared to control mice. All apoE2/2 mice died, concentrations (mmol/l) in apoE2/2 and control mice whereas only 40% of the controls were killed by the in- fection (Fig. 4). Ninety minutes after a K. pneumoniae Cholesterol Triglycerides inoculum of 5 3 10 CFU TNF concentration in the cir- ApoE2/2 Controls ApoE2/2 Controls culation of apoE2/2 mice was higher than those of mmol/l mmol/l control mice (12.3 6 3.4 ng/ml vs. 7.6 6 3.8 ng/ml), al- a b Plasma 16.14 6 3.73 1.87 6 0.15 1.13 6 0.40 0.85 6 0.30 though these differences did not reach statistical signifi- a c VLDL 10.2 6 1.14 0.12 6 0.03 0.81 6 0.14 0.48 6 0.16 cance (P 5 0.056). LDL 0.60 6 0.06 0.23 6 0.03 0.03 6 0.01 0.04 6 0.01 HDL 0.47 6 0.10 1.04 6 0.10 0.03 6 0.01 0.03 6 0.01 Ex vivo cytokine production Results are given as mean 6 SD for 30 animals per group. To investigate whether the increased in vivo production a b c P , 0.0001, P , 0.001, P , 0.01 (for comparison between of TNFa was due to an increased capacity of peritoneal C57BL/6 and apo E2/2). macrophages to produce cytokines, cells of both mouse strains were stimulated with LPS. No differences were found in either secreted or cell-associated TNFa, IL-1a, or mice, due to extremely high cholesterol concentrations of the VLDL + IDL fraction and, to a lesser extent, to LDL IL-1b between both strains (Table 2). cholesterol concentrations, whereas HDL concentration LPS clearance was lower (Table 1). Total triglyceride concentrations To test whether the LPS clearance differed in the two were slightly, though significantly, higher in the apoE2/2 mouse strains, we assessed LPS concentrations in the cir- mice (Table 1). The higher triglyceride concentrations culation. Ten min after an iv injection of 50 mg LPS, were reflected in the triglyceride concentrations of the apoE2/2 mice showed levels of endotoxin in plasma VLDL 1 IDL fraction (Table 1). similar to control mice in the two mouse strains (3.69 6 The chemical composition of all lipoprotein fractions 0.80 vs. 3.75 6 1.21 mg/ml). After 90 min, apoE2/2 showed differences (Fig. 1), especially in the cholesterol mice showed lower, albeit not significant, levels of endo- and triglyceride content of the VLDL 1 IDL and the LDL toxin in plasma compared with control animals (0.51 6 fraction. In the apoE2/2 mice, cholesterol represented the 0.03 vs. 0.66 6 0.21 mg/ml). Again no differences were most abundant component of lipoproteins, whereas control seen 4 h after LPS injection (0.34 6 0.03 vs. 0.37 6 0.09 mice have more triglycerides in these fractions (Fig. 1). mg/ml). Lethality and cytokines in vivo LPS neutralizing capacity Survival after an endotoxin injection is depicted in Fig. 2. To investigate whether the lipoproteins in the plasma of The mortality in the apoE2/2 mice (86%) was significantly apoE2/2 and control mice were equally potent in bind- higher than in controls (41%). Plasma concentrations of ing and neutralization of LPS, preincubation of LPS with TNFa 90 min after LPS administration were more than 4- plasma from the two mouse strains was performed. After fold higher in apoE2/2 mice than in control animals (Fig. 24 h, human PBMC were added and cytokine production 3). The circulating concentrations of IL-1a, IL-1b, and IL-6 measured after an additional 24 h incubation. IL-1b and were 10- to 100-fold lower than those of TNF, and no differ- TNFa production by PBMC were higher after preincuba- ences were detected 4 h after LPS injection (Fig. 3). tion of LPS with plasma obtained from apoE2/2 mice Klebsiella pneumoniae infection than with control plasma (Table 3), compatible with de- After an intravenous injection of 0.5 3 10 CFU of K. creased LPS-neutralizing capacity of lipoproteins from pneumoniae, apoE2/2 mice showed an increased mortal- apoE2/2 mice. Fig. 1. Chemical composition of lipoprotein fractions (VLDL 1 IDL, LDL and HDL) of apoE2/2 (solid bars) and control mice (open bars). The figure represents the percentage of free cholesterol (fc), cholesteryl esters (ce), triglycer- ides (tg), phospholipids (pl), and protein of each fraction. 682 Journal of Lipid Research Volume 40, 1999 Fig. 2. Survival of apoE2/2 (r) and control mice (h) after LPS (2 mg/mouse iv). Sur vival was signifi- cantly impaired in the apoE2/2 mice. The figure shows pooled data of two experiments with at least 10 mice per group each; P , 0.01. tralization of endotoxin: the increased lipid concentra- DISCUSSION tion on the one hand, and the apolipoproteins on the In the present study we show that hyperlipidemic mice other. Both apoE2/2 and LDLr2/2 mice have increased deficient in apolipoprotein E are more susceptible to en- plasma lipid concentrations due to either increased cho- dotoxemia and to Klebsiella pneumoniae infection than con- lesterol-rich VLDL in apoE2/2 mice or high IDL and LDL trol mice. In the apoE2/2 mice, the severe cytokinemia, in LDLr2/2 mice (23). Because both strains have similar in particular TNFa, is most probably responsible for elevations in lipoprotein concentrations, and all lipopro- death. As macrophages of apoE2/2 and control mice tein subfractions have been shown to neutralize LPS (2–7), produced similar amounts of TNFa, the differences do the discrepancy in LPS neutralization cannot easily be ex- not seem to be due to differences in intrinsic cytokine plained by quantitative differences in hyperlipidemia. production capacity. The apoE2/2 plasma apparently exerts an impaired These results are in accordance with those of Roselaar LPS-neutralizing capacity despite high cholesterol. This is and Daugherty (22) who demonstrated that apolipoprotein in accordance with results of Harris et al. (12) showing E-deficient mice are more susceptible to Listeria monocyto- that cholesterol is not necessary for the interaction genes. However, their results are in marked contrast to those between lipoproteins and LPS, as a cholesterol-free lipid we obtained in hyperlipidemic LDL receptor knock-out emulsion, Soyacal also protected against endotoxin- (LDLr2/2) mice that had an increased survival towards a induced death. On the other hand, infusion of a com- Gram-negative challenge and a dampered proinflammatory mercially produced triglyceride-rich particle, Intralipid , endotoxin response by virtue of the endotoxin-neutralizing in humans did not result in a reduced cytokine release effects of lipoproteins (8). Moreover, the plasma of the (24). apoE2/2 mice appeared to have a low LPS-neutralizing ca- It should be taken into account that the lipoproteins in pacity, which was even less than that of normolipidemic con- the VLDL-sized fraction of apoE2/2 mice have a differ- trol mice. How should these differences be explained? ent chemical composition when compared to controls. Basically, two mechanisms may be responsible for neu- Conformational changes may have taken place, resulting Fig. 3. Plasma cytokine concentrations in apoE2/2 (black columns) and control mice (white columns) af- ter 50 mg of LPS iv. The plasma concentrations of TNFa (90 min after the LPS challenge) was significantly higher in apoE2/2 mice compared with control mice. No differences in IL 1a, IL-1b, or IL-6 plasma concen- trations (4 h after LPS challenge) were detected. The figure shows pooled data of three experiments with groups of at least 5 animals. Data are represented as proportion of the mean cytokine production of control animals; *P , 0.0001. de Bont et al. Endotoxin susceptibility of apoE-deficient mice 683 Fig. 4. Survival of apoE2/2 and control mice after iv infection with Klebsiella pneumoniae. An iv injection of 0.5 3 10 CFU K. pneumoniae was given to apoE2/2 (r) and control mice (h). Survival was significantly im- paired in apoE2/2 mice. The figure shows data of one experiment with 10 mice per group (P , 0.01). in a diminished availability of LPS-binding sites and a de- clearance from the plasma between the two mouse strains creased neutralizing capacity. indicates that the outcome after an LPS injection is deter- The second possible mechanism concerns the apolipo- mined in the liver and not in the blood. So, most proba- proteins. Both free apoA-I and apoE have strong LPS neu- bly, the LPS in apoE deficiency is only directed towards cy- tralizing effects (14–17). In apoE2/2 mice, the VLDL 1 tokine-producing Kupffer cells, whereas in the control IDL fraction is relatively poor in apolipoproteins. In addi- mice a part is directed towards parenchymal cells. How tion, the excess of lipids carried by the lipoprotein particles should we explain that we did not find differences in circu- in these mice may result in a change of the LPS-binding lating IL-1a, IL-1b, or IL-6 4 h after LPS injection? One ex- sites on their apolipoproteins. This, together with a 2-fold planation could be that TNF and IL-1 production are differ- decrease in HDL, may be an explanation why LPS in entially regulated (27). We have recently found that different apoE2/2 mice is neutralized to a limited extent. In addi- LPS receptors seem to be responsible for the production of tion to the LPS-binding capacity of apoproteins, apoE the various cytokines: CD14-dependent mechanisms medi- seems to be able to neutralize LPS differently. The LPS- ate TNFa production, while IL-1 synthesis is induced by detoxifying effects of chylomicrons are thought to be me- both CD14-dependent and CD14-independent mecha- diated by apoE present on this lipoprotein (14). Free en- nisms (28). Secondly, IL-1a and IL-1b are less prominent dotoxin, injected iv, is cleared from the plasma by the as circulating cytokines than TNFa. liver, where it is predominantly taken up by the Kupffer Let us now return to the question how the discrepancy cells resulting in production of cytokines (25, 26). How- in LPS neutralization between the two hyperlipidemic ever, when bound to apoE (free or associated with chylo- mouse strains should be explained. In comparison with microns), LPS can be shunted away from these Kupffer C57BL/6J mice, LDLr2/2 mice have extremely high IDL cells and directed to the parenchymal liver cells (14), re- and LDL concentrations and a 2-fold increased HDL- sulting in a reduction of peak cytokine serum levels (17). cholesterol as well (data not shown). ApoE is present on These observations are in accordance with ours, showing both IDL and HDL, hence increased in LDLr2/2 mice. that in apoE2/2 mice TNFa plasma concentrations were Therefore, absence of apoE in apoE2/2 but not in 4- to 5-fold higher than in controls after LPS challenge. LDLr2/2 mice may indeed explain the difference in out- Moreover, the fact that no difference was found in LPS come between the two mouse strains. TABLE 2. Secreted and cell-associated cytokine production after in vitro stimulation of peritoneal macrophages TABLE 3. Effect of apoE2/2 and control plasma on of apoE2/2 and C57BL/6 mice LPS-induced cytokine production Secreted Cell-associated TNFa IL-1b C57BL/6 ApoE2/2 C57BL/6 ApoE2/2 ApoE2/2 Control ApoE2/2 Control ng/ml ng/ml ng/ml ng/ml IL-1a 0.13 6 0.03 0.10 6 0.01 3.62 6 1.29 2.90 6 1.29 1 0.59 0.38 2.48 1.31 IL-1b 0.07 6 0.02 0.05 6 0.01 0.29 6 0.07 0.30 6 0.11 2 0.27 0.22 1.81 0.65 TNFa 0.79 6 0.19 0.73 6 0.20 0.19 6 0.04 0.24 6 0.08 3 0.44 0.36 1.88 0.84 Peritoneal macrophages (10 /well) were stimulated with LPS (100 LPS (100 ng/ml) preincubated with either plasma of apoE2/2 or ng/ml) for 24 h at 378C. The data represent mean 6 SD for six animals. of control mice was added to 5 3 10 PBMC/well and incubated for 24 The results do not differ significantly for the three cytokines measured. h at 37 8C (n 5 3). 684 Journal of Lipid Research Volume 40, 1999 Human very low density lipoproteins and chylomicrons can pro- In conclusion, in this study we demonstrated that hy- tect against endotoxin-induced death in mice. J. Clin. Invest. 86: percholesterolemia itself is not sufficient for protection 696–702. against endotoxemia. More likely, the presence of apoE is 13. Read, T. E., C. Grunfeld, Z. L. Kumwenda, M. C. Calhoun, J. P. essential in the process of LPS detoxification, either by cat- Kane, K. R. Feingold, and J. H. Rapp. 1995. Triglyceride-rich li- poproteins prevent septic death in rats. J. Exp. Med. 182: 267– alyzing the binding of LPS to the lipoprotein particle or by directing the LPS to the parenchymal cells away from cyto- 14. Rensen, P. C., M. Oosten, E. Bilt, M. Eck, J. Kuiper, and T. J. Ber- kine-producing Kupffer cells or by both mechanisms. kel. 1997. Human recombinant apolipoprotein E redirects lipo- polysaccharide from Kupffer cells to liver parenchymal cells in rats in vivo. J. Clin. Invest. 99: 2438–2445. 15. Flegel, W. A., M. W. Baumstark, C. Weinstock, A. Berg, and H. The authors thank Geert Poelen and Theo van der Ing for Northoff. 1993. Prevention of endotoxin-induced monokine re- their excellent assistance with the animal experiments. lease by human low and high density lipoproteins and by apolipo- Manuscript received 23 June 1998 and in revised form 26 October 1998. protein A-I. Infect. Immun. 61: 5140–5146. 16. Emancipator, K., G. Csako, and R. J. Elin. 1992. In vitro inactiva- tion of bacterial endotoxin by human lipoproteins and apolipo- proteins. Infect. Immun. 60: 596–601. 17. Rensen, P. C. N., M. van Oosten, E. van de Bilt, M. van Eck, J. REFERENCES Kuiper, and T. J. C. van Berkel. 1997. Human apolipoprotein E protects against the deleterious effects of bacterial lipopolysaccha- 1. Lynn, W. A., and J. Cohen. 1995. Adjunctive therapy for septic ride in rats in vivo. Circulation. 379 (Abstract). shock: a review of experimental approaches. Clin. Infect. Dis. 20: 18. Parker, T. S., D. M. Levine, J. C. Chang, J. Laxer, C. C. Coffin, and 143–158. A. L. Rubin. 1995. Reconstituted high-density lipoprotein neutral- 2. Van Lenten, B. J., A. M. Fogelman, M. E. Haberland, and P. A. Ed- izes gram-negative bacterial lipopolysaccharides in human whole wards. 1986. The role of lipoproteins and receptor-mediated en- blood. Infect. Immun. 63: 253–258. docytosis in the transport of bacterial lipopolysaccharide. Proc. 19. van Ree, J. H., W. J. van den Broek, V. E. Dahlmans, P. H. Groot, M. Natl. Acad. Sci. USA. 83: 2704–2708. Vidgeon-Hart, R. R. Frants, B. Wieringa, L. M. Havekes, and M. H. 3. Harris, H. W., C. Grunfeld, K. R. Feingold, T. E. Read, J. P. Kane, Hofker. 1994. Diet-induced hypercholesterolemia and atheroscle- A. L. Jones, E. B. Eichbaum, G. F. Bland, and J. H. Rapp. 1993. rosis in heterozygous apolipoprotein E-deficient mice. Atherosclero- Chylomicrons alter the fate of endotoxin, decreasing tumor necro- sis. 111: 25–37. sis factor release and preventing death. J. Clin. Invest. 91: 1028– 20. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. 1034. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 4. Flegel, W. A., A. Wolpl, D. N. Mannel, and H. Northoff. 1989. Inhi- 193: 265–275. bition of endotoxin-induced activation of human monocytes by 21. Vogels, M. T., E. J. Mensink, K. Ye, O. C. Boerman, C. M. Ver- human lipoproteins. Infect. Immun. 57: 2237–2245. schueren, C. A. Dinarello, and J. W. van der Meer. 1994. Differen- 5. Netea, M. G., N. de Bont, P. N. M. Demacker, B. J. Kullberg, L. E. H. tial gene expression for IL-1 receptor antagonist, IL-1, and TNF re- Jacobs, T. J. G. Verver-Jansen, A. F. H. Stalenhoef, and J. W. M. van ceptors and IL-1 and TNF synthesis may explain IL-1-induced der Meer. 1998. Lipoprotein[a] inhibits the lipopolysaccharide- resistance to infection. J. Immunol. 153: 5772–5780. induced tumor necrosis factor-alpha production by human mono- 22. Roselaar, S. E., and A. Daugherty. 1998. Apolipoprotein E-defi- nuclear cells. Infect. Immun. 66: 2365–2367. cient mice have impaired innate immune responses to Listeria 6. Cavaillon, J. M., C. Fitting, N. Haeffner-Cavaillon, S. J. Kirsch, and monocytogenes in vivo. J. Lipid. Res. 39: 1740–1743. H. S. Warren. 1990. Cytokine response by monocytes and mac- 23. Ishibashi, S., M. S. Brown, J. L. Goldstein, R. D. Gerard, R. E. rophages to free and lipoprotein-bound lipopolysaccharide. Infect. Hammer, and J. Herz. 1993. Hypercholesterolemia in low density Immun. 58: 2375–2382. lipoprotein receptor knockout mice and its reversal by adenovi- 7. Weinstock, C., H. Ullrich, R. Hohe, A. Berg, M. W. Baumstark, I. rus-mediated gene delivery [see comments]. J. Clin. Invest. 92: Frey, H. Northoff, and W. A. Flegel. 1992. Low density lipoproteins 883–893. inhibit endotoxin activation of monocytes. Arterioscler. Thromb. 12: 24. van der Poll, T., C. C. Braxton, S. M. Coyle, M. A. Boermeester, J. C. 341–347. Wang, P. M. Jansen, W. J. Montegut, S. E. Calvano, C. E. Hack, and 8. Netea, M. G., P. N. Demacker, B. J. Kullberg, O. C. Boerman, I. S. F. Lowry. 1995. Effect of hypertriglyceridemia on endotoxin re- Verschueren, A. F. Stalenhoef, and J. W. van der Meer. 1996. Low- sponsiveness in humans. Infect. Immun. 63: 3396–3400. density lipoprotein receptor-deficient mice are protected against 25. Praaning van Dalen, D. P., A. Brouwer, and D. L. Knook. 1981. lethal endotoxemia and severe gram-negative infections. J. Clin. Clearance capacity of rat liver Kupffer, endothelial, and parenchy- Invest. 97: 1366–1372. mal cells. Gastroenterology. 81: 1036–1044. 9. Levine, D. M., T. S. Parker, T. M. Donnelly, A. Walsh, and A. L. Ru- 26. Munford, R. S., J. M. Andersen, and J. M. Dietschy. 1981. Sites of bin. 1993. In vivo protection against endotoxin by plasma high tissue binding and uptake in vivo of bacterial lipopolysaccharide– density lipoprotein. Proc. Natl. Acad. Sci. USA. 90: 12040–12044. high density lipoprotein complexes: studies in the rat and squirrel 10. Hubsch, A. P., F. S. Powell, P. G. Lerch, and J. E. Doran. 1993. A re- monkey. J. Clin. Invest. 68: 1503–1513. constituted, apolipoprotein A-I containing lipoprotein reduces tu- 27. Zuckerman, S. H., and G. F. Evans. 1992. Endotoxin tolerance: in mor necrosis factor release and attenuates shock in endotoxemic vivo regulation of tumor necrosis factor and interleukin-1 synthe- rabbits. Circ. Shock. 40: 14–23. sis is at the transcriptional level. Cell. Immunol. 140: 513–519. 11. Pajkrt, D., J. E. Doran, F. Koster, P. G. Lerch, B. Arnet, T. van der 28. Netea, M. G., B. J. Kullberg, and J. W. M. van der Meer. 1998. Lipo- Poll, J. W. ten Cate, and S. J. H. van Deventer. 1996. Antiinflamma- polysaccharide-induced production of tumour necrosis factor and tory effects of reconstituted high density lipoprotein during hu- interleukin-1 is differtially regulated at the receptor level: the role man endotoxemia. J. Exp. Med. 184: 1601–1608. of CD14-dependent and CD14-independent pathways. Immunology. 12. Harris, H. W., C. Grunfeld, K. R. Feingold, and J. H. Rapp. 1990. 94: 340–344. de Bont et al. Endotoxin susceptibility of apoE-deficient mice 685

Journal

Journal of Lipid Research – Unpaywall

Published: Apr 1, 1999

There are no references for this article.