Microbial Pathogenesis 1990 ; 9 : 55-59
Short communications Dissociation between Limulus neutralisation and in vivo protection in monoclonal antibodies directed against endotoxin core structures J . S . McConnell,' B . J . Appelmelk2 and J . Cohen` 'Infectious Diseases Unit, Departments of Medicine and Bacteriology, Hammersmith Hospital, London W12 ONN, U .K. and 'Department of Medical Microbiology, Vrije Universiteit, Amsterdam
(Received June 14, 1989 ; accepted in revised form May 21, 1990)
McConnell, J . S . (Infectious Disease Unit, Departments of Medicine and Bacteriology, Hammersmith Hospital, London W1 2 ONN, U .K .), B . J . Appelmelk and J . Cohen . Dissociation between Limulus neutralisation and in vivo protection in monoclonal antibodies directed against endotoxin core structures . Microbial Pathogenesis 1990 ; 9 : 55-59 . Studies with rough mutants of certain Gram-negative bacteria have indicated that monoclonal antibodies (mAbs) to endotoxin core can protect animals and man from endotoxic shock . We assessed the ability of such antibodies to neutralise endotoxin in the Limulus amoebocyte lysate (LAL) assay, and compared this to their protective effect in a murine model of endotoxic shock. We evaluated 11 mAbs raised against Salmonella minnesota R595 . Endotoxin neutralisation in the LAL assay, expressed as 50% inhibition titres, ranged between 1/32 and 1/414 . However, there was no apparent relationship between the titre required to produce 50% inhibition of LAL and its ability to protect mice from endotoxic shock . We conclude that LAL neutralisation appears unrelated to biological activity ; in this system, LAL inhibition by mAb ascites cannot be used to predict protection in vivo . Key words : monoclonal antibodies ; lipopolysaccharide; Limulus assay .
Introduction In an effort to overcome the high mortality associated with Gram-negative sepsis and shock,' there has been extensive investigation into the protective effect of passive immunisation with antisera to lipopolysaccharide (LPS) in both animal models and in man .' With the development of monoclonal antibody (mAb) technology it was possible to overcome some of the drawbacks associated with polyclonal antisera to LPS .3 Several groups have prepared mAbs to the LPS core region in the hope that these would provide broad cross-protection from the effects of Gram-negative septicaemia 3-9 and some of these antibodies have proved to be effective in animal studies . 3-5,9,10 It would be useful to be able to predict by in vitro testing whether a particular anticore mAb is likely to be protective, as this would greatly aid the selection of antibodies for potential clinical use . There is evidence to suggest that the in vivo biological " Author to whom correspondence should be addressed at : Infectious Diseases Unit, Royal Postgraduate Medical School, Hammersmith Hospital, DuCane Road, London W12 ONN, U .K . 0882-4010/90/070055+05 $03 .00/0
© 1990 Academic Press Limited
56
J . S . McConnell eta/[
Table 1 Characteristics of 11 test and two control monoclonal antibodies Clone 14 17 19 20 21 22 25 26 27 28 29 Pen? 90 : A1
Subclass IgM IgG2b IgG3b IgM IgG3 IgG3 IgG3 IgG2a IgG3 IgM IgG2a IgG3 1gM
R595 binding
LAL titre
Protection factor
27 243 9 729 27 729 729 729 8 9 729
165 109 414 95 45 71 262 176 554 32 68
n .s . 91" 15' 46' 21 " n .s . n .s . 22" n .s . n .s, n .s .
0 0
43 23
R595 binding is the reciprocal (x 10 -3 ) of the titre obtained in a solid phase ELISA (see Methods) . The LAL titre is the reciprocal mean titre (n = 3) required to produce 50% LAL inhibition . The protection factor is the LD 50 obtained with the test ascites divided by the LD 50 obtained with Pen7 . P < 0 .05 compared to Pen 7 . n .s. = not significant .
response induced by a given LPS is related to its ability to stimulate the Limulus amoebocyte lysate (LAL) assay ." Thus it seemed possible that the LAL assay could be used to predict in vitro, the potential in vivo activity of a given mAb . Here we present the results of a study in which the LAL assay was used to assess the endotoxin neutralising ability of a range of mAbs .
Results All the test mAbs (but not the controls) bound to purified Salmonella minnesota R595 LPS in the solid phase ELISA . The reciprocal titres were between 8000 and 729000 (Table 1), but these titres did not correlate with the 50% LAL neutralisation titres or with the evidence of in vivo protection (see below) . The 50% inhibition titres of LAL by the mAbs ranged between 1/32 and 1/554 (Table 1) . The titres for the two control antibodies were : Pen 7 : 1/43, and 90 : A1 : 1/23 (mean of three experiments) . The protection factor (i .e . the increase in the LD 50 in animals receiving anti-endotoxin mAbs compared to those receiving the control mAb) varied between 1 and 91, but there was no correlation between the LAL inhibition titre and the protection factor (coefficient of linear regression r=-0 .23 P=0 .49) . Similarly, the immunoglobulin isotype appeared unrelated to activity as measured either by LAL neutralisation or in vivo protection . Of the six ascitic fluids characterised as IgG3, the neutralisation titres ranged between 1/71 and 1/554, and the protection factor between 1 and 21, but without any concordance . Clone 27 for instance had the lowest protection factor (1) and the highest inhibition titre (1 /554) .
Discussion Our aim in this study was to determine whether the ability of a range of anti-LPS core region mAbs to bind LPS and inhibit the LAL assay could be used to predict the
Monoclonal antibodies to endotoxin
57
protective effect of the antibody in vivo . We devised a measure of LAL neutralisation (the 50% inhibition titre) and compared this to the ability of the mAb to protect mice from endotoxic shock . There was an approximately 10-fold variation in the neutralising capacity of the ascitic fluids, and a range of approximately 100-fold in the protection factor (Table 1) . However, when the 50% inhibition titre was compared by linear regression with the protection factor for each mAb, no correlation was found . Two previous studies have considered the question of LAL neutralisation by antiLPS antibodies . Wells et a/. 12 found that equine polyclonal anti-LPS hyperimmune plasma reduced the apparent concentration of LPS in a 'spiked' plasma containing a fixed amount of E. coli 0111 : B4 LPS . The anti-LPS effect was directly related to the concentration of IgG present in the reaction mixture and a non-immune control plasma effected only a minimal reduction in the level of LPS . This anti-LPS preparation was reported to be effective in the therapy and prophylaxis of Gram-negative bacteraemia and endotoxic shock in several animal and human studies, and an ELISA showed it to be cross-reactive with LPS from several species of Gram-negative organism . Given the polyclonal nature of the antiserum tested, it is not clear whether the antibodies capable of LAL inhibition were the same as those which caused protection in vivo . In contrast, Appelmelk et a/. 13 found anti-lipid A mAbs to have no effect on the detection of lipid A by the LAL assay or on lipid-A-induced mortality . The lack of effect seen in both systems is surprising since the mAbs tested were directed against lipid A, that part of the LPS molecule which is responsible for both mortality and LAL activation ." The results reported here show that there is no predictable association between the ability of an antibody directed against LPS to neutralise LPS in the LAL assay and its protective efficacy in in vivo models of infection . This finding may be broadly interpreted in two ways . It may be that the clones demonstrating protection in vivo exert their effect by shielding and/or deforming that part of the lipid A which is responsible for mortality without influencing other parts of the molecule responsible for LAL activation . Another possible explanation for our findings is that the protective effects observed in vivo are not due to direct neutralisation of LPS (e .g . preventing induction of TNF or IL-1), but rather are the result of a different, less direct mechanism such as clearance of LPS-antibody complexes . Further study of these mechanisms may well shed light on the precise role of the LPS molecule in initiating lethal physiological changes ; meanwhile it is clear that LAL neutralisation cannot be used to screen mAbs for protective effect in vivo . Materials and methods Monoclonal antibodies . Mouse mAbs were produced in the laboratory of one of us (BJA) using methods described previously .' The mAbs used were identified as clones 14, 17, 19, 20, 21, 22, 25, 26, 27, 28, 29, and all were raised against LPS obtained from the Re mutant S. minnesota strain R595 . In addition, we studied two control mAbs, Pen 7 which binds to benzylpenicillin, 15 and 90 : Al which binds to Gram-positive cell wall structures . Immunoglobulin class and subclass were determined by ELISA using appropriate antisera (Nordic, Tilburg, The Netherlands) . The titre of each mAb was determined by an ELISA using a method described elsewhere 16 except that purified S . minnesota R595 LPS (List Biological Laboratories, Campbell, California) was used as the coating antigen . Mouse protection studies . For the mouse protection studies, groups of six random-bred Swiss-albino mice were given intravenous injections of S. minnesota R595 LPS in varying doses, mixed with 0 .2 ml ascites and 20 yg actinomycin D . Deaths were recorded for 1 week, and the LD 50 calculated . In this system, the LD 50 of endotoxin both for saline and Pen 7 (the
58
J . S . McConnell
et al.
control antibody) was 1 ng . A 'protection factor' was calculated for each mAb, defined as the LD 50 obtained with test mAb ascites divided by the LD 50 found for the Pen 7 control .
Endotoxin neutralisation . To assess the endotoxin neutralising ability of each antibody a fourfold dilution series of ascitic fluid was made in pyrogen free normal saline (PF-NS) containing R595 LPS, producing a final dilution range of antibody from neat to 1/16384, each dilution containing 500 pg/ml LPS . A parallel dilution series for each antibody was made which contained no added endotoxin . Both sets of ascites dilutions along with a positive control sample of 500 pg/ml LPS in PF-NS were assayed immediately for their endotoxin activity using LAL chromogenic substrate kits (Coatest, Kabi Vitrum, Uxbridge, U .K .) . The assay method was that described previously" with the following modifications : after reconstitution LAL was further diluted 1 in 3 in PFW and 50 pl added to 50 pl of sample in the wells of a flat bottom microtitre plate (Sterilin Ltd, Feltham, Middlesex, U .K .) ; following incubation at 37'C for 45 min 100 u1 of substrate S2423 at 0 .9 mm in the buffer supplied was added to each well ; the optical density (OD) was measured at 1 min and 2 min after addition of substrate and the change in OD (AOD) calculated for each sample . In order to quantify the neutralising ability of each antibody we calculated the titre of antibody required to produce a 50% inhibition of the maximum AOD produced by the positive control sample (AOD max ) . The neutralisation of 500 pg/ml LPS produced by each concentration of antibody (expressed as %AOD) was derived from the following equation : AOD A-AOD B % OD= x100 AOD max where AOD A = AOD of antibody dilutions containing 500 pg/ml LPS, AOD B = AOD of antibody alone, and AOD max = AOD of 500 pg/m1 LPS in PF-NS . Graphs were plotted of %AOD against log 10 percentage antibody concentration . From the graphs, the straight line part of the curve was used to calculate the 50% inhibition titres by regression analysis .
We- thank the Wellcome Trust for their support, and Dr H . deSoet (Amsterdam) for making available clone 90 : Al .
References 1 . Hamill RJ, Maki DG . Endotoxin shock in man caused by Gram-negative bacilli-etiology, clinical features, diagnosis, natural history and prevention . In : Proctor RA, ed . Handbook of endotoxin . Vol . 4 . Clinical aspects of endotoxin shock . Amsterdam : Elsevier, 1986; 55-126 . 2 . Proctor RA. Role of antibody in the prevention and pathogenesis of endotoxin and Gram-negative septic shock . In : Proctor RA, ed . Handbook of endotoxin, Vol . 4 . Clinical aspects of endotoxin shock . Amsterdam : Elsevier, 1986; 161-84 . 3 . Appelmelk BJ . Antibodies to the LPS core region and their protective role in Gram-negative sepsis . Free University Press, Amsterdam, 1987 . 4 . Dunn DL, Bogard WC Jr, Cerra FB . Efficacy of type-specific and cross-reactive murine monoclonal antibodies directed against endotoxin during experimental sepsis . Surgery 1985 ; 98 : 283-90 . 5 . Dunn DL, Ewald DC, Chandan N, Cerra FB . Immunotherapy of Gram-negative bacterial sepsis . Arch Surg 1986 ; 121 : 58-62 . 6 . Mutharia LM, Crockford G, Bogard WC Jr, Hancock REW . Monoclonal antibodies specific for Escherichia co/i J5 lipopolysaccharide : cross reaction with other Gram-negative bacterial species . Infect Immun 1984; 45 : 631-6 . 7 . Nelles MJ, Niswander CA . Mouse monoclonal antibodies reactive with J5 lipopolysaccharide exhibit extensive serological cross-reactivity with a variety of Gram-negative bacteria . Infect Immun 1984 ; 46 : 677-81 . 8 . Miner KM, Manyak CL, Williams E et al Characterization of murine monoclonal antibodies to Escherichia coli J5 . Infect Immun 1986 ; 52 : 56-62 . 9 . Teng NNH, Kaplan HS, Hebert JM et al Protection against Gram-negative bacteremia and endotoxemia with human monoclonal IgM antibodies . Proc Natl Acad Sci USA 1985 ; 82 :1790-4. 10 . Silva AT, Appelmelk BJ, Buurman WA, Bayston KA, Cohen J . Monoclonal antibody to endotoxin core protects mice from Escherichia coli sepsis by a mechanism independent of tumor necrosis factor and interleukin 6 . J Infect Dis (in press) . 11 . Weary, ME, Donohue G, Pearson FC, Storey K. Relative potencies of four reference endotoxin standards as measured by the Limulus amoebocyte lysate and USP rabbit pyrogen tests . Appi Environ Microbiol 1980; 40 : 1148-51 . 12 . Wells MT, Gaffin SL, Gregory M, Coovadia Y. Properties of equine anti-lipopolysaccharide hyperimmune
Monoclonal antibodies to endotoxin
13 .
14 .
15 . 16 .
17 .
59
plasma : binding to lipopolysaccharide and activity against Gram-negative bacteria . J Med Microbiol 1987 ;24 :187-96 . Appelmelk BJ, Verweij-van Vaught AMJJ, Maaskant JJ et al . Production and characterization of mouse monoclonal antibodies reacting with the lipopolysaccharide core region of Gram-negative bacilli . J Med Microbiol 1988 ; 26 : 107-14 . Galanos C, Luderitz 0, Rietschel ET, Westphal 0 . Newer aspects of the chemistry and biology of bacterial lipopolysaccharides, with special reference to their lipid A component . In : Goodwin TW, ed . International review of biochemistry . Vol . 14 . Biochemistry of lipids II . Baltimore : University Park Press, 1977 ;239-335 . de Haan P, de Jonge AJR, Verbrugge T, Boorsma DM . Three epitope-specific monoclonal antibodies against the hapten penicillin . Int Arch Allergy Appi Immunol 1985 ; 76 : 42-6 . Moore RH, Lampert IA, Chia Y, Aber V, Cohen J . Influence of endotoxin on graft-versus host disease after bone marrow transplantation across major histocompatibility barriers in mice . Transplantation 1987 ; 43 : 731-6 . Cohen J, McConnell JS . Observations on the measurement and evaluation of endotoxemia by a quantitative Limulus lysate microassay . J Infect Dis 1984 ; 150 : 916-24 .
Short communications Dissociation between Limulus neutralisation and in vivo protection in monoclonal antibodies directed against endotoxin core structures J . S . McConnell,' B . J . Appelmelk2 and J . Cohen` 'Infectious Diseases Unit, Departments of Medicine and Bacteriology, Hammersmith Hospital, London W12 ONN, U .K. and 'Department of Medical Microbiology, Vrije Universiteit, Amsterdam
(Received June 14, 1989 ; accepted in revised form May 21, 1990)
McConnell, J . S . (Infectious Disease Unit, Departments of Medicine and Bacteriology, Hammersmith Hospital, London W1 2 ONN, U .K .), B . J . Appelmelk and J . Cohen . Dissociation between Limulus neutralisation and in vivo protection in monoclonal antibodies directed against endotoxin core structures . Microbial Pathogenesis 1990 ; 9 : 55-59 . Studies with rough mutants of certain Gram-negative bacteria have indicated that monoclonal antibodies (mAbs) to endotoxin core can protect animals and man from endotoxic shock . We assessed the ability of such antibodies to neutralise endotoxin in the Limulus amoebocyte lysate (LAL) assay, and compared this to their protective effect in a murine model of endotoxic shock. We evaluated 11 mAbs raised against Salmonella minnesota R595 . Endotoxin neutralisation in the LAL assay, expressed as 50% inhibition titres, ranged between 1/32 and 1/414 . However, there was no apparent relationship between the titre required to produce 50% inhibition of LAL and its ability to protect mice from endotoxic shock . We conclude that LAL neutralisation appears unrelated to biological activity ; in this system, LAL inhibition by mAb ascites cannot be used to predict protection in vivo . Key words : monoclonal antibodies ; lipopolysaccharide; Limulus assay .
Introduction In an effort to overcome the high mortality associated with Gram-negative sepsis and shock,' there has been extensive investigation into the protective effect of passive immunisation with antisera to lipopolysaccharide (LPS) in both animal models and in man .' With the development of monoclonal antibody (mAb) technology it was possible to overcome some of the drawbacks associated with polyclonal antisera to LPS .3 Several groups have prepared mAbs to the LPS core region in the hope that these would provide broad cross-protection from the effects of Gram-negative septicaemia 3-9 and some of these antibodies have proved to be effective in animal studies . 3-5,9,10 It would be useful to be able to predict by in vitro testing whether a particular anticore mAb is likely to be protective, as this would greatly aid the selection of antibodies for potential clinical use . There is evidence to suggest that the in vivo biological " Author to whom correspondence should be addressed at : Infectious Diseases Unit, Royal Postgraduate Medical School, Hammersmith Hospital, DuCane Road, London W12 ONN, U .K . 0882-4010/90/070055+05 $03 .00/0
© 1990 Academic Press Limited
56
J . S . McConnell eta/[
Table 1 Characteristics of 11 test and two control monoclonal antibodies Clone 14 17 19 20 21 22 25 26 27 28 29 Pen? 90 : A1
Subclass IgM IgG2b IgG3b IgM IgG3 IgG3 IgG3 IgG2a IgG3 IgM IgG2a IgG3 1gM
R595 binding
LAL titre
Protection factor
27 243 9 729 27 729 729 729 8 9 729
165 109 414 95 45 71 262 176 554 32 68
n .s . 91" 15' 46' 21 " n .s . n .s . 22" n .s . n .s, n .s .
0 0
43 23
R595 binding is the reciprocal (x 10 -3 ) of the titre obtained in a solid phase ELISA (see Methods) . The LAL titre is the reciprocal mean titre (n = 3) required to produce 50% LAL inhibition . The protection factor is the LD 50 obtained with the test ascites divided by the LD 50 obtained with Pen7 . P < 0 .05 compared to Pen 7 . n .s. = not significant .
response induced by a given LPS is related to its ability to stimulate the Limulus amoebocyte lysate (LAL) assay ." Thus it seemed possible that the LAL assay could be used to predict in vitro, the potential in vivo activity of a given mAb . Here we present the results of a study in which the LAL assay was used to assess the endotoxin neutralising ability of a range of mAbs .
Results All the test mAbs (but not the controls) bound to purified Salmonella minnesota R595 LPS in the solid phase ELISA . The reciprocal titres were between 8000 and 729000 (Table 1), but these titres did not correlate with the 50% LAL neutralisation titres or with the evidence of in vivo protection (see below) . The 50% inhibition titres of LAL by the mAbs ranged between 1/32 and 1/554 (Table 1) . The titres for the two control antibodies were : Pen 7 : 1/43, and 90 : A1 : 1/23 (mean of three experiments) . The protection factor (i .e . the increase in the LD 50 in animals receiving anti-endotoxin mAbs compared to those receiving the control mAb) varied between 1 and 91, but there was no correlation between the LAL inhibition titre and the protection factor (coefficient of linear regression r=-0 .23 P=0 .49) . Similarly, the immunoglobulin isotype appeared unrelated to activity as measured either by LAL neutralisation or in vivo protection . Of the six ascitic fluids characterised as IgG3, the neutralisation titres ranged between 1/71 and 1/554, and the protection factor between 1 and 21, but without any concordance . Clone 27 for instance had the lowest protection factor (1) and the highest inhibition titre (1 /554) .
Discussion Our aim in this study was to determine whether the ability of a range of anti-LPS core region mAbs to bind LPS and inhibit the LAL assay could be used to predict the
Monoclonal antibodies to endotoxin
57
protective effect of the antibody in vivo . We devised a measure of LAL neutralisation (the 50% inhibition titre) and compared this to the ability of the mAb to protect mice from endotoxic shock . There was an approximately 10-fold variation in the neutralising capacity of the ascitic fluids, and a range of approximately 100-fold in the protection factor (Table 1) . However, when the 50% inhibition titre was compared by linear regression with the protection factor for each mAb, no correlation was found . Two previous studies have considered the question of LAL neutralisation by antiLPS antibodies . Wells et a/. 12 found that equine polyclonal anti-LPS hyperimmune plasma reduced the apparent concentration of LPS in a 'spiked' plasma containing a fixed amount of E. coli 0111 : B4 LPS . The anti-LPS effect was directly related to the concentration of IgG present in the reaction mixture and a non-immune control plasma effected only a minimal reduction in the level of LPS . This anti-LPS preparation was reported to be effective in the therapy and prophylaxis of Gram-negative bacteraemia and endotoxic shock in several animal and human studies, and an ELISA showed it to be cross-reactive with LPS from several species of Gram-negative organism . Given the polyclonal nature of the antiserum tested, it is not clear whether the antibodies capable of LAL inhibition were the same as those which caused protection in vivo . In contrast, Appelmelk et a/. 13 found anti-lipid A mAbs to have no effect on the detection of lipid A by the LAL assay or on lipid-A-induced mortality . The lack of effect seen in both systems is surprising since the mAbs tested were directed against lipid A, that part of the LPS molecule which is responsible for both mortality and LAL activation ." The results reported here show that there is no predictable association between the ability of an antibody directed against LPS to neutralise LPS in the LAL assay and its protective efficacy in in vivo models of infection . This finding may be broadly interpreted in two ways . It may be that the clones demonstrating protection in vivo exert their effect by shielding and/or deforming that part of the lipid A which is responsible for mortality without influencing other parts of the molecule responsible for LAL activation . Another possible explanation for our findings is that the protective effects observed in vivo are not due to direct neutralisation of LPS (e .g . preventing induction of TNF or IL-1), but rather are the result of a different, less direct mechanism such as clearance of LPS-antibody complexes . Further study of these mechanisms may well shed light on the precise role of the LPS molecule in initiating lethal physiological changes ; meanwhile it is clear that LAL neutralisation cannot be used to screen mAbs for protective effect in vivo . Materials and methods Monoclonal antibodies . Mouse mAbs were produced in the laboratory of one of us (BJA) using methods described previously .' The mAbs used were identified as clones 14, 17, 19, 20, 21, 22, 25, 26, 27, 28, 29, and all were raised against LPS obtained from the Re mutant S. minnesota strain R595 . In addition, we studied two control mAbs, Pen 7 which binds to benzylpenicillin, 15 and 90 : Al which binds to Gram-positive cell wall structures . Immunoglobulin class and subclass were determined by ELISA using appropriate antisera (Nordic, Tilburg, The Netherlands) . The titre of each mAb was determined by an ELISA using a method described elsewhere 16 except that purified S . minnesota R595 LPS (List Biological Laboratories, Campbell, California) was used as the coating antigen . Mouse protection studies . For the mouse protection studies, groups of six random-bred Swiss-albino mice were given intravenous injections of S. minnesota R595 LPS in varying doses, mixed with 0 .2 ml ascites and 20 yg actinomycin D . Deaths were recorded for 1 week, and the LD 50 calculated . In this system, the LD 50 of endotoxin both for saline and Pen 7 (the
58
J . S . McConnell
et al.
control antibody) was 1 ng . A 'protection factor' was calculated for each mAb, defined as the LD 50 obtained with test mAb ascites divided by the LD 50 found for the Pen 7 control .
Endotoxin neutralisation . To assess the endotoxin neutralising ability of each antibody a fourfold dilution series of ascitic fluid was made in pyrogen free normal saline (PF-NS) containing R595 LPS, producing a final dilution range of antibody from neat to 1/16384, each dilution containing 500 pg/ml LPS . A parallel dilution series for each antibody was made which contained no added endotoxin . Both sets of ascites dilutions along with a positive control sample of 500 pg/ml LPS in PF-NS were assayed immediately for their endotoxin activity using LAL chromogenic substrate kits (Coatest, Kabi Vitrum, Uxbridge, U .K .) . The assay method was that described previously" with the following modifications : after reconstitution LAL was further diluted 1 in 3 in PFW and 50 pl added to 50 pl of sample in the wells of a flat bottom microtitre plate (Sterilin Ltd, Feltham, Middlesex, U .K .) ; following incubation at 37'C for 45 min 100 u1 of substrate S2423 at 0 .9 mm in the buffer supplied was added to each well ; the optical density (OD) was measured at 1 min and 2 min after addition of substrate and the change in OD (AOD) calculated for each sample . In order to quantify the neutralising ability of each antibody we calculated the titre of antibody required to produce a 50% inhibition of the maximum AOD produced by the positive control sample (AOD max ) . The neutralisation of 500 pg/ml LPS produced by each concentration of antibody (expressed as %AOD) was derived from the following equation : AOD A-AOD B % OD= x100 AOD max where AOD A = AOD of antibody dilutions containing 500 pg/ml LPS, AOD B = AOD of antibody alone, and AOD max = AOD of 500 pg/m1 LPS in PF-NS . Graphs were plotted of %AOD against log 10 percentage antibody concentration . From the graphs, the straight line part of the curve was used to calculate the 50% inhibition titres by regression analysis .
We- thank the Wellcome Trust for their support, and Dr H . deSoet (Amsterdam) for making available clone 90 : Al .
References 1 . Hamill RJ, Maki DG . Endotoxin shock in man caused by Gram-negative bacilli-etiology, clinical features, diagnosis, natural history and prevention . In : Proctor RA, ed . Handbook of endotoxin . Vol . 4 . Clinical aspects of endotoxin shock . Amsterdam : Elsevier, 1986; 55-126 . 2 . Proctor RA. Role of antibody in the prevention and pathogenesis of endotoxin and Gram-negative septic shock . In : Proctor RA, ed . Handbook of endotoxin, Vol . 4 . Clinical aspects of endotoxin shock . Amsterdam : Elsevier, 1986; 161-84 . 3 . Appelmelk BJ . Antibodies to the LPS core region and their protective role in Gram-negative sepsis . Free University Press, Amsterdam, 1987 . 4 . Dunn DL, Bogard WC Jr, Cerra FB . Efficacy of type-specific and cross-reactive murine monoclonal antibodies directed against endotoxin during experimental sepsis . Surgery 1985 ; 98 : 283-90 . 5 . Dunn DL, Ewald DC, Chandan N, Cerra FB . Immunotherapy of Gram-negative bacterial sepsis . Arch Surg 1986 ; 121 : 58-62 . 6 . Mutharia LM, Crockford G, Bogard WC Jr, Hancock REW . Monoclonal antibodies specific for Escherichia co/i J5 lipopolysaccharide : cross reaction with other Gram-negative bacterial species . Infect Immun 1984; 45 : 631-6 . 7 . Nelles MJ, Niswander CA . Mouse monoclonal antibodies reactive with J5 lipopolysaccharide exhibit extensive serological cross-reactivity with a variety of Gram-negative bacteria . Infect Immun 1984 ; 46 : 677-81 . 8 . Miner KM, Manyak CL, Williams E et al Characterization of murine monoclonal antibodies to Escherichia coli J5 . Infect Immun 1986 ; 52 : 56-62 . 9 . Teng NNH, Kaplan HS, Hebert JM et al Protection against Gram-negative bacteremia and endotoxemia with human monoclonal IgM antibodies . Proc Natl Acad Sci USA 1985 ; 82 :1790-4. 10 . Silva AT, Appelmelk BJ, Buurman WA, Bayston KA, Cohen J . Monoclonal antibody to endotoxin core protects mice from Escherichia coli sepsis by a mechanism independent of tumor necrosis factor and interleukin 6 . J Infect Dis (in press) . 11 . Weary, ME, Donohue G, Pearson FC, Storey K. Relative potencies of four reference endotoxin standards as measured by the Limulus amoebocyte lysate and USP rabbit pyrogen tests . Appi Environ Microbiol 1980; 40 : 1148-51 . 12 . Wells MT, Gaffin SL, Gregory M, Coovadia Y. Properties of equine anti-lipopolysaccharide hyperimmune
Monoclonal antibodies to endotoxin
13 .
14 .
15 . 16 .
17 .
59
plasma : binding to lipopolysaccharide and activity against Gram-negative bacteria . J Med Microbiol 1987 ;24 :187-96 . Appelmelk BJ, Verweij-van Vaught AMJJ, Maaskant JJ et al . Production and characterization of mouse monoclonal antibodies reacting with the lipopolysaccharide core region of Gram-negative bacilli . J Med Microbiol 1988 ; 26 : 107-14 . Galanos C, Luderitz 0, Rietschel ET, Westphal 0 . Newer aspects of the chemistry and biology of bacterial lipopolysaccharides, with special reference to their lipid A component . In : Goodwin TW, ed . International review of biochemistry . Vol . 14 . Biochemistry of lipids II . Baltimore : University Park Press, 1977 ;239-335 . de Haan P, de Jonge AJR, Verbrugge T, Boorsma DM . Three epitope-specific monoclonal antibodies against the hapten penicillin . Int Arch Allergy Appi Immunol 1985 ; 76 : 42-6 . Moore RH, Lampert IA, Chia Y, Aber V, Cohen J . Influence of endotoxin on graft-versus host disease after bone marrow transplantation across major histocompatibility barriers in mice . Transplantation 1987 ; 43 : 731-6 . Cohen J, McConnell JS . Observations on the measurement and evaluation of endotoxemia by a quantitative Limulus lysate microassay . J Infect Dis 1984 ; 150 : 916-24 .
