1063
Monoclonal Antibody to Mouse Lipopolysaccharide Receptor Protects Mice against the Lethal Effects of Endotoxin David C. Morrison, Richard Silverstein, Stuart W. Bright, Tai-Ying Chen, Linda M. Flebbe, and Mei-Guey Lei
From the Departments of Microbiology, Molecular Genetics and Immunology and of Biochemistry and Molecular Biology, University of Kansas Medical Center, Kansas City
Endotoxic shock continues to pose a major clinical problem worldwide. Despite intensive efforts at therapeutic intervention, mortality resulting from profound shock remains unacceptably high at 25 %-40 % [1, 2]. Accumulated evidence from both in vitro and in vivo studies implicates bacterial endotoxic lipopolysaccharide (LPS) as a major contributing factor to the pathogenesis of septic shock (reviewed in [3]). The biologically active lipid A component of LPS, whose biochemical structure has been recently elucidated in detail [4], is well-recognized as a potent activator of peripheral blood mononuclear cells and tissue macrophages, and there is a growing consensus that products of LPS-stimulated monocytes, particularly interleukin-1 and tumor necrosis factor-a (TNFa), are key participants in the pathophysiologic responses to LPS [5]. In contrast to the considerable information available on the potential target cells responsive to LPS and the mediators produced by these cells in response to LPS, the mechanisms by which LPS initiates host-cell responses is not well understood. Recent studies in our laboratory have focused upon the identification and characterization of specific LPS receptors on murine lymphoreticular cells. Using photoactivatable radioiodinated disulfide-reducible LPS derivatives synthesized in our laboratory [6], we have identified and partially characterized a specific membrane-localized LPS binding protein
Received 7 February 1990; revised 5 June 1990. Financial support: National Institutes of Health (AI-22948-06, AI-2344705, RR-05373). T.Y.C. is a scholar of the Wesley Foundation Cancer Research and Training Grant to the Kansas universities. Reprints or correspondence: Dr. David C. Morrison, Department of Microbiology, Molecular Genetics and Immunology, University of Kansas Medical Center, Kansas City, KS 66103. The Journal of Infectious Diseases 1990;162:1063-1068 © 1990 by The University of Chicago. All rights reserved. 0022-1899/90/6205-0010$01.00
of 80-kDa molecular mass, which is expressed on mouse lymphocytes and macrophages [7, 8] . We have established methods for the partial purification of this protein and have generated several IgM hamster-mouse monoclonal antibodies (MAbs) with specificity for this protein [9]. In vitro studies have shown that one of these MAbs, MAb 5D3, functions as an agonist for LPS and activates macrophages to become cytotoxic for tumor cells [10]. Collectively, these results provide the first strong evidence that this 80-kDa membrane LPSbinding protein may function as a specific receptor for LPS. We investigated whether MAb 5D3 is protective when administered to mice before challenge with an otherwise lethal dose of endotoxin in the murine galactosamine-sensitization model [11]. We also assessed if that protection can be manifest in normal mice given an LDso dose of endotoxin.
Materials and Methods Animals. For most experiments 3- to 6-month-old outbred CF1 mice (Charles River Laboratories, Wilmington, MA) were used. In some studies, the mice were 3- to 9-month-old female C3Heb/FeJ (Jackson Laboratories, Bar Harbor, ME) and endotoxin unresponsive C3H/He (Bantin and Kingman, Freemont, CA). The C3H/He mice were essentially equivalent to the C3Heb/HeJ strain from Jackson Laboratories; the latter were unavailable because of a fire at the laboratories and thus mice from an alternative vendor were used. LPS and other reagents. LPS from Escherichia coli 0111:B4 was extracted and purified by the phenol-water procedure of Westphal et al. [12] as modified by Morrison and Leive [13]. LPS from Salmonella enteritidis (Difco, Detroit) was also extracted by aqueous phenol. Purified recombinant TNFa was provided by Chiron (Emeryville, CA). D-galactosamine was purchased from Sigma Chemical (St. Louis). Monoclonal and polyclonal antibodies. Hamster-mouse IgM MAb 5D3 was prepared as described [9]. The MAb was purified from hybridoma culture supernatants by ion-exchange chromatog-
Downloaded from http://jid.oxfordjournals.org/ at The University of British Colombia Library on July 13, 2015
Specific endotoxic lipopolysaccharide (LPS) binding sites on the cell membranes of murine lymphocytes and macrophages that may serve as functional receptors for LPS have recently been identified using photoactivatable cross-linking LPS derivatives. A monoclonal antibody (MAb 5D3) with specificity for this 80-kDa protein has also been generated and characterized. The capacity of MAb 5D3 to protect mice against the lethal effects of endotoxin was investigated. Pretreatment ofCA mice with as little as 15 /Lg ofMAb 5D3 provided virtually complete protection against a dose ofendotoxin 10-fold greater than that required to kill all mice in an untreated control group using the galactosamine sensitization model. Significant protection was also afforded normal mice given MAb 5D3 relative to saline. Several lines of evidence suggest that MAb 5D3-mediated protection is due to the agonist properties of this antibody rather than a receptor r blockade mechanism.
1064
Morrison et al.
Results
Protective efficacy ofMAb 5D3 against endotoxin lethality in D-galactosamine-treated mice. We first assessed the capacity of the MAb 5D3 anti-LPS receptor antibody to provide protection against an otherwise lethal dose of LPS. For these experiments mice were sensitized to LPS by simultaneous treatment with o-galactosamine. Groups of mice were given various graded doses of MAb 5D3 or saline ip, and /'\J80 min later challenged with 12 ng of LPS and 18 mg of o-galactosamine, also given ip. Survival was monitored over the next 2 days. The cumulative results of several experiments (table 1) show that as little as 7.5 p.g of MAb 5D3 provided complete protection against endotoxin lethality in the o-galactosamine model. In contrast, no statistically significant pro-
Table 1. Protective efficacy of monoclonal antibody (MAb) 5D3 in D-galactosamine-treated mice. Survivors/total by experiment Pretreatment
Challenge*
Saline MAb 503 (Jlg) 75 7.5 0.75 None
12 ng LPS 12 ng LPS 12 ng LPS 12 ng LPS PBS
2
Total (%)
2/15
8/16
10/31 (34)
10/10
10/10 10/10
20/20 (100) 10/10 (100) 3/5 (60) 10/10 (100)
5/5
3/5 5/5
* All mice were simultaneously given 18 mg ofo-galactosamine. LPS saccharide.
= lipopoly-
tection was provided to mice pretreated with the pyrogen-free PBS (34% ~urvival; P > .50). Freudenberg and Galanos [15] showed that LPS pretreatment also protects against subsequent LPS lethality in the n-galactosamine model, reportedly by a mechanism involving macrophage activation. Thus, it was important to establish that our observed protection was not the result of endotoxin contamination of MAb 5D3. Heat treatment at 100°C was used to distinguish MAb 5D3 activity from potential contaminating endotoxin activity. Further, to exclude nonspecific effects of the antibody, normal hamster IgM was included in one experiment as an antibody isotype control. Table 2 summarizes results of four experiments. Both LPS and MAb 5D3 pretreatment provided significant protection against a lethal challenge dose of LPS; however, unlike LPS, protection afforded by MAb 5D3 treatment was heat labile. That heated MAb 5D3 did not, in some way, nonsp.ecifically trap contaminating endotoxin was shown by an additional experiment in which LPS was added to MAb 5D3 before heating. Further, no protection was observed in mice given an equivalent dose of normal hamster IgM or IgG (data not shown). These studies establish that MAb 5D3 pretreatment protected mice against subsequent lethal endotoxin challenge and that the observed protection was most likely not the result of endotoxin contamination. We also directly assessed by Limulus amoebocyte lysate assays the amount ofendotoxin contamination in both the MAb 5D3 preparation and normal hamster IgM. Compared with an E. coli 0111:B4 LPS standard, 15 p.g of MAb 5D3 contains
Monoclonal Antibody to Mouse Lipopolysaccharide Receptor Protects Mice against the Lethal Effects of Endotoxin David C. Morrison, Richard Silverstein, Stuart W. Bright, Tai-Ying Chen, Linda M. Flebbe, and Mei-Guey Lei
From the Departments of Microbiology, Molecular Genetics and Immunology and of Biochemistry and Molecular Biology, University of Kansas Medical Center, Kansas City
Endotoxic shock continues to pose a major clinical problem worldwide. Despite intensive efforts at therapeutic intervention, mortality resulting from profound shock remains unacceptably high at 25 %-40 % [1, 2]. Accumulated evidence from both in vitro and in vivo studies implicates bacterial endotoxic lipopolysaccharide (LPS) as a major contributing factor to the pathogenesis of septic shock (reviewed in [3]). The biologically active lipid A component of LPS, whose biochemical structure has been recently elucidated in detail [4], is well-recognized as a potent activator of peripheral blood mononuclear cells and tissue macrophages, and there is a growing consensus that products of LPS-stimulated monocytes, particularly interleukin-1 and tumor necrosis factor-a (TNFa), are key participants in the pathophysiologic responses to LPS [5]. In contrast to the considerable information available on the potential target cells responsive to LPS and the mediators produced by these cells in response to LPS, the mechanisms by which LPS initiates host-cell responses is not well understood. Recent studies in our laboratory have focused upon the identification and characterization of specific LPS receptors on murine lymphoreticular cells. Using photoactivatable radioiodinated disulfide-reducible LPS derivatives synthesized in our laboratory [6], we have identified and partially characterized a specific membrane-localized LPS binding protein
Received 7 February 1990; revised 5 June 1990. Financial support: National Institutes of Health (AI-22948-06, AI-2344705, RR-05373). T.Y.C. is a scholar of the Wesley Foundation Cancer Research and Training Grant to the Kansas universities. Reprints or correspondence: Dr. David C. Morrison, Department of Microbiology, Molecular Genetics and Immunology, University of Kansas Medical Center, Kansas City, KS 66103. The Journal of Infectious Diseases 1990;162:1063-1068 © 1990 by The University of Chicago. All rights reserved. 0022-1899/90/6205-0010$01.00
of 80-kDa molecular mass, which is expressed on mouse lymphocytes and macrophages [7, 8] . We have established methods for the partial purification of this protein and have generated several IgM hamster-mouse monoclonal antibodies (MAbs) with specificity for this protein [9]. In vitro studies have shown that one of these MAbs, MAb 5D3, functions as an agonist for LPS and activates macrophages to become cytotoxic for tumor cells [10]. Collectively, these results provide the first strong evidence that this 80-kDa membrane LPSbinding protein may function as a specific receptor for LPS. We investigated whether MAb 5D3 is protective when administered to mice before challenge with an otherwise lethal dose of endotoxin in the murine galactosamine-sensitization model [11]. We also assessed if that protection can be manifest in normal mice given an LDso dose of endotoxin.
Materials and Methods Animals. For most experiments 3- to 6-month-old outbred CF1 mice (Charles River Laboratories, Wilmington, MA) were used. In some studies, the mice were 3- to 9-month-old female C3Heb/FeJ (Jackson Laboratories, Bar Harbor, ME) and endotoxin unresponsive C3H/He (Bantin and Kingman, Freemont, CA). The C3H/He mice were essentially equivalent to the C3Heb/HeJ strain from Jackson Laboratories; the latter were unavailable because of a fire at the laboratories and thus mice from an alternative vendor were used. LPS and other reagents. LPS from Escherichia coli 0111:B4 was extracted and purified by the phenol-water procedure of Westphal et al. [12] as modified by Morrison and Leive [13]. LPS from Salmonella enteritidis (Difco, Detroit) was also extracted by aqueous phenol. Purified recombinant TNFa was provided by Chiron (Emeryville, CA). D-galactosamine was purchased from Sigma Chemical (St. Louis). Monoclonal and polyclonal antibodies. Hamster-mouse IgM MAb 5D3 was prepared as described [9]. The MAb was purified from hybridoma culture supernatants by ion-exchange chromatog-
Downloaded from http://jid.oxfordjournals.org/ at The University of British Colombia Library on July 13, 2015
Specific endotoxic lipopolysaccharide (LPS) binding sites on the cell membranes of murine lymphocytes and macrophages that may serve as functional receptors for LPS have recently been identified using photoactivatable cross-linking LPS derivatives. A monoclonal antibody (MAb 5D3) with specificity for this 80-kDa protein has also been generated and characterized. The capacity of MAb 5D3 to protect mice against the lethal effects of endotoxin was investigated. Pretreatment ofCA mice with as little as 15 /Lg ofMAb 5D3 provided virtually complete protection against a dose ofendotoxin 10-fold greater than that required to kill all mice in an untreated control group using the galactosamine sensitization model. Significant protection was also afforded normal mice given MAb 5D3 relative to saline. Several lines of evidence suggest that MAb 5D3-mediated protection is due to the agonist properties of this antibody rather than a receptor r blockade mechanism.
1064
Morrison et al.
Results
Protective efficacy ofMAb 5D3 against endotoxin lethality in D-galactosamine-treated mice. We first assessed the capacity of the MAb 5D3 anti-LPS receptor antibody to provide protection against an otherwise lethal dose of LPS. For these experiments mice were sensitized to LPS by simultaneous treatment with o-galactosamine. Groups of mice were given various graded doses of MAb 5D3 or saline ip, and /'\J80 min later challenged with 12 ng of LPS and 18 mg of o-galactosamine, also given ip. Survival was monitored over the next 2 days. The cumulative results of several experiments (table 1) show that as little as 7.5 p.g of MAb 5D3 provided complete protection against endotoxin lethality in the o-galactosamine model. In contrast, no statistically significant pro-
Table 1. Protective efficacy of monoclonal antibody (MAb) 5D3 in D-galactosamine-treated mice. Survivors/total by experiment Pretreatment
Challenge*
Saline MAb 503 (Jlg) 75 7.5 0.75 None
12 ng LPS 12 ng LPS 12 ng LPS 12 ng LPS PBS
2
Total (%)
2/15
8/16
10/31 (34)
10/10
10/10 10/10
20/20 (100) 10/10 (100) 3/5 (60) 10/10 (100)
5/5
3/5 5/5
* All mice were simultaneously given 18 mg ofo-galactosamine. LPS saccharide.
= lipopoly-
tection was provided to mice pretreated with the pyrogen-free PBS (34% ~urvival; P > .50). Freudenberg and Galanos [15] showed that LPS pretreatment also protects against subsequent LPS lethality in the n-galactosamine model, reportedly by a mechanism involving macrophage activation. Thus, it was important to establish that our observed protection was not the result of endotoxin contamination of MAb 5D3. Heat treatment at 100°C was used to distinguish MAb 5D3 activity from potential contaminating endotoxin activity. Further, to exclude nonspecific effects of the antibody, normal hamster IgM was included in one experiment as an antibody isotype control. Table 2 summarizes results of four experiments. Both LPS and MAb 5D3 pretreatment provided significant protection against a lethal challenge dose of LPS; however, unlike LPS, protection afforded by MAb 5D3 treatment was heat labile. That heated MAb 5D3 did not, in some way, nonsp.ecifically trap contaminating endotoxin was shown by an additional experiment in which LPS was added to MAb 5D3 before heating. Further, no protection was observed in mice given an equivalent dose of normal hamster IgM or IgG (data not shown). These studies establish that MAb 5D3 pretreatment protected mice against subsequent lethal endotoxin challenge and that the observed protection was most likely not the result of endotoxin contamination. We also directly assessed by Limulus amoebocyte lysate assays the amount ofendotoxin contamination in both the MAb 5D3 preparation and normal hamster IgM. Compared with an E. coli 0111:B4 LPS standard, 15 p.g of MAb 5D3 contains
