Citation
The relationship between human erythrocyte lipopolysaccharide receptor and an inhibitor of complement mediated lysis

Material Information

Title:
The relationship between human erythrocyte lipopolysaccharide receptor and an inhibitor of complement mediated lysis
Creator:
Jackson, Gloria Jean, 1939-
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Antibodies ( jstor )
Butanols ( jstor )
Complement system ( jstor )
Endotoxins ( jstor )
Erythrocyte membrane ( jstor )
Erythrocytes ( jstor )
Gels ( jstor )
Guinea pigs ( jstor )
Lipids ( jstor )
Sheep ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
22597758 ( ALEPH )
4748645 ( OCLC )

Downloads

This item has the following downloads:


Full Text












THE RELATIONSHIP BETWEEN
HUMAN ERYTHROCYTE LIPOPOLYSACCHARIDE RECEPTOR AND AN INHIBITOR OF COMPLEMENT MEDIATED LYSIS







By

GLORIA JEAN JACKSON












A DISSERTATION PRESENTED TO THE
GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY









UNIVERSITY OF FLORIDA 1978














ACKNOWLEDGMENTS


I wish to express sincere appreciation for the technical criticisms and patient guidance given me by Dr. Edward M. Hoffmann.

I also would like to extend thanks to Dr. L. 0. Ingram, Dr. A. S. Bleiweis, Dr. Paul W. Chun, Dr. Lester W. Clem, Dr. Paul Smith and Mr. Jim Milam for supplying constant help and encouragement.

To all of the members of my laboratory and departmental family over the last four and a half years, I offer my sincere gratitude for their constant encouragement, patience, caring and all of the happy times we've shared.

I wish also to add my thanks to Mrs. Sandy Cannella for her encouragement and her excellent and dedicated assistance in the typing of this manuscript.

Finally, but most important of all, for their continual prayers,

help, guidance, unfailing love and for always being there through the good and hard times, I thank my parents Mr. and Mrs. Charles and Sarah McKnight and my loving son Jacques. Without them this would never have been possible.













ii

















TABLE OF CONTENTS


PAGE

ACKNOWLEDGEMENTS.......................................... ii

LIST OF TABLES............................................ iv

LIST OF FIGURES........................................... v

GLOSSARY OF ABBREVIATIONS............................... vi

ABSTRACT.................................................. vii

INTRODUCTION .............................................. 1

MATERIALS AND METHODS..................................... 17

RESULTS................................................... 27

DISCUSSION................................................ 57

REFERENCES................................................ 62

BIOGRAPHY................................................. 68























iii















LIST OF TABLES


TABLE PAGE

1. Comparative Properties of the Classical
and Alternative Complement Pathways................ 8

2. Determination of Optimal LPS Concentration Used for Coating Human Erythrocytes................ 28

3. A comparison of LPS-receptor and IH Inhibitor
Activities of Several Crude Extracts of
Erythrocyte Stromata ............................... 30

4. Percent Recovery of LPS-receptor and IH Inhibitor Activity from Extracts of Human
Erythrocyte Stromata Following Sepharose 4B
and DEAE-Sephadex Purification..................... 34

5. Recovery of LPS-receptor and IH Inhibitor
Activity of Human Erythrocyte Stromata Extracted
at pH 5.3 Using the Procedure of Springer.......... 43

6. Recoveries of IH Inhibitor and LPS-receptor Activities From Erythrocyte Stromata Extracted
at pH 5.3 by the Procedure of Springer and
Subjected to DEAE-Sephadex Chromatography.......... 49

7. Consumption of Total Complement in Either
E Absorbed or E-LPS absorbed GP Serum by
Untreated and LPS Coated sheep Erythrocytes........ 51

















iv















LIST OF FIGURES


FIGURE PAGE

1. Probable structure of LPS of Escherichia coli...... 4

2. Schematic representation of purification
procedures of the LPS-receptor and IH inhibitor.... 19

3. IH inhibitor activities of crude butanol
extracted erythrocyte membranes .................... 32

4. A gel filtration profile of the butanol extracted human erythrocyte stromata .............. 36

5. DEAE-Sephadex chromatography of the
sepharose 4B active fractions from human stromata.. 38

6. Polyacrylamide disc gel electrophoresis
of human stromata extracts......................... 41

7. DEAE-Sephadex chromatography of
human erythrocyte stromata extracts
prepared at pH 5.3................................. 45

8. Treatment of partial purified extracts of human erythrocyte stromata with
sheep erythrocytes................................ 48

9. Complement consumption in LPS treated guinea pig serum........................... 53

10. Complement consumption of LPS
and IH inhibitor treated erythrocytes............. 56














v















GLOSSARY OF ABBREVIATIONS


C......................................... complement

Cl, C2, C3, C4, C5, C6, C7, C8, C91... complement components

E..................................... erythrocyte

A..................................... antibody

HA................................... hemagglutination

HAI................................... hemagglutination inhibition

EDTA.................................. ethylenediamine tetracetate



IThe nomenclature of complement used conforms to that proposed as a result of a series of World Health Organizations (Immunochemistry, 7:137, 1970).

























vi










Abstract of Dissertation Presented to the
Graduate Council of the University of Florida
in Partial Fulfillment of the Requirements for the
Degree of Doctor of Philosophy


THE RELATIONSHIP BETWEEN
HUMAN ERYTHROCYTE LIPOPOLYSACCHARIDE RECEPTOR AND AN INHIBITOR OF COMPLEMENT MEDIATED LYSIS

By

Gloria Jean Jackson

December, 1978


Chairman: Edward M. Hoffmann
Major Department: Microbiology and Cell Science


Human erythrocyte membrane extracts with receptor specificity for the lipopolysaccharides of gram negative bacteria were found to be rich in IH, a complement inhibitory substance originally isolated and described from human red cell membranes. The possibility that the LPS-receptor and IH inhibitor might be the same macromolecule was considered.

In this investigation it was demonstrated that although closely

associated, the two activities are biologically distinct and separable. This conclusion was supported by the results of five experimental approaches. Small, but distinct differences were observed in the elution profiles of the two activities when crude extracts of erythrocyte membranes were subjected to sepharose 4B, DEAE-Sephadex chromatography and polyacrylamide disc gel electrophoresis. Complete separation of the two activities was accomplished by a shift in the pH of the membrane extraction conditions. Further, differences between the two activities based on their ability to spontaneously bind to sheep erythrocytes were vii










demonstrated. Finally, evidence was presented indicating that the membranes of sheep erythrocytes consist of molecules with LPS-receptor specificity but were devoid of IH inhibitor activity.

Additional studies indicated that LPS bound to cell membranes can activate either the alternative or classical complement pathways and that IH inhibitor associated with cell membranes can block LPS-induced complement lysis of red cells despite the fact that complement activation has occurred.




































viii















INTRODUCTION


The endotoxins or lipopolysaccharides (LPS) of gram negative

bacteria are among numerous antigens known to be capable of fixing to the membranes of erythrocytes and other mammalian cells in vitro and under certain conditions, in vivo (1). Because of their unique ability to modulate the immune response in a wide variety of ways, they have emerged as a complex and fascinating class of macromolecules. Functionally, endotoxins have been shown to have many different properties. Due to their chemical makeup and localization in the outer membrane of the bacterial cell envelope they have been shown to play a major role in the establishment of a selective permeability barrier (2,3) and in serving as receptors for certain bacteriophages (4).

Of interest to the immunochemist, however, is the fact that interaction of LPS with components of the immune system may lead to a single or combination of physiological responses. These include toxicity, mitogenicity, immunogenicity, tolerance and activation of complement. Although much is known concerning the general nature of these responses, the mechanisms ofthe cell associated events responsible for their development in the presence of LPS are still not fully understood.

A great deal of information about the chemical structure of LPS, from a variety of organisms, has accumulated. Although it has been recently recognized that LPS isolated from a given organism is


1





2




heterogenous (5), LPS of most gram negative wildtype organisms appear to share the same basic molecular composition. As illustrated in Figure 1, all consist of three regions. The first region, the 0specific polysaccharide antigen which is made up of repeating units of five to eight monosaccharides, carries the main serologic specificity for a given organism. Numerous serological groups differing in 0antigen specificity are now recognized and the polysaccharides accordingly show wide inter- and intrageneric variations in composition (6). Of interest is the fact that natural antibodies to this region are found in most animal species but do not always appear to be protective, and in some cases a lethal gram negative bacteremia develops despite high titers of O-specific antibody (7). Region two consists of a short outer core which contains glucose (glc), galactose (gal), and N-acetylglucosamine (Glc NAC), and an inner core of L-glycero-Dmannoheptose, phosphate, ethanolamine and three molecules of 3-deoxyoctonate (KDO). Unique to the endotoxins of gram negatives, 3-deoxyoctonate provides a linkage site to the third region, the lipid A moiety. Lipid A is basically composed of a phosphorylated glucosamine backbone to which are attached fatty acids and ethanolamine residues. The nature and distribution of the lipid A fatty acids varies among bacterial groups with the inner core polysaccharide composition remaining constant (7). The complete LPS containing O-specific antigen is designated smooth (S), and all mutants lacking O-specific side chains are referred to as rough or R-forms.

Many early studies concerned with the interaction of endotoxins and biological systems were carried out using either whole bacteria














Col G1cNAc R R or R 1-4 1-2 1 3
[-Col1l-6G1cNAc131-2Gcl-2Gal]-G 1cc-2Gall1-3Gc-3Hepl-3al pa-5KDO(KDO,KDO)-GcN1-6GcN
n
ethanolamine 6 Fatty Acids I I~ P I L P antigenic side chain core lipid A


LPS1-n=11
LPS11-n=3.



*R1 =O;R2= Glcl-+4 GlcNh2;R3=Hep.





















Figure 1. Probable structure of LPS of Escherichia coli. There
are probably three separate core structures present
with R as shown for each. Pyrophosphate bridges probably cross link-adjacent lipid molecules yielding at least trimers. Abbreviations: Col, colitose; GlcNAc,
N-acetylglucosamine; Glc, glucose; Gal, galactose;
GlcN, glucosamine; Hep, heptose; and KDO, 2 keto-3deoxyoctanoci acid. A protein covalently linked to
LPS may also be present (6).






5



or partially purified LPS preparations in the fluid phase. This was in spite of the fact that it had been shown that heated LPS (1000C for 60 minutes) could be coated onto the surface of a number of cells including erythrocytes (8). The latter are often the model target cells for hemolytic assays. It has become increasingly clear that the biological consequence of either an in vivo or in vitro encounter with LPS, whether cell bound or partially purified, is dependent upon its fixation to various target cells (9). For example, it was reported that patients suffering from endotoxin shock and sepsis due to gram negative bacteria had greatly reduced levels of blood platelets and that the platelets contained LPS. It has further been established that human platelets possess an endotoxin binding receptor, which when triggered by interaction with LPS, results in the release of a vasoactive amine, 5-hydroxytryptamine (5-HT) and the unmasking of clotpromoting activity (10). In animal studies, it was observed that guinea pigs injected intravenously with LPS also show a 95% decrease in blood platelets with a concomitant shortening of the clotting time

(11). Additionally, it has been reported that there is a direct relationship between the susceptibility of different strains of mice to the lethal effects of Salmonella endotoxin and the affinity of their red cells for either heat-killed Salmonellae or free LPS (12).

Little was known about the nature of the attachment of LPS to

any cell until the late sixties, when G. Springer isolated an extract from the membranes of human erythrocytes having a high affinity and specificity for the lipopolysaccharides of a variety of gram negative






6



bacteria (13-15). This material, designated as an LPS-receptor, has now been purified to homogeneity and characterized. Springer has reported that the LPS-receptor is a lipoglycoprotein, rich in N-acetylneuraminic acid (NANA), galactose, hexosamine and contains about 61% protein (16). It appears to be a pentameric molecule with a molecular weight of about 228,000 daltons. The LPS-receptor functions by direct interaction with groups on the LPS molecule which provide an attachment site for tissue components (16). Strong evidence has accumulated suggesting that this attachment site is the lipid A moiety of LPS (17). This high affinity of the LPS-receptor for endotoxins is quite remarkable because both macromolecules are highly negatively charged: the receptor, because of its high NANA content and LPS because of its phosphoric acid radicals.

Because the immunological specificity of LPS bound to erythrocytes remains unchanged, Springer has suggested that the lipid A of LPS binds to the specific receptor via clusters of hydrophobic amino acids which makeup about 40% of the total peptide content of the receptor leaving the polysaccharide available for the reaction with antibodies (16). A complete understanding of the orientation of LPS on tissues, bound either by specific receptors or by non-specific mechanisms, maycome from studies involving the interaction of cell bound LPS with serum complement.

The anticomplementary effects of LPS have long been established. For some time, evidence seemed to suggest that the single most important factor in the development of a noxious response to endotoxins was the direct interaction of the lipid A region with biological






7



systems including the complement system (18). Whatever the mechanism, mounting evidence seems to implicate complement as a mediator of a number of the deleterious responses to endotoxins in experimental animals and man (8). More importantly, recent evidence indicates that the ability of LPS to initiate a complement response is not confined to the lipid A moiety but appears to involve the polysaccharide core as well (19).

It has become increasingly clear that a major role of the complement system during an immune response is the mediation of enhanced phagocytosis with a concomitant induction of an inflammatory response. This is accomplished by the sequential activation of the proteins which make up the complement system. The activation process can be divided into three major stages: a recognition stage, the generation of C3 cleavage enzymes and C3 activation stage, and a terminal or membrane attack stage (20-22). Under normal conditions, complement proteins exist in the serum as inactive precursors, and are activated by either of two pathways -- the classical or alternative. The component of these pathways and their reaction requirements are summarized in Table

1. The two share a number of similar characteristics but differ, basically, in the reactants and the sequence of reactions of the first two stages.

The capacity of endotoxins to activate the complement system by a mechanism which requires neither antibodies to LPS nor the participation of the early complement components was demonstrated more than twenty years ago (23). More recently, it was recognized that fluid phase activation of complement by LPS is not restricted to the















TABLE I

Comparative Properties of the Classical
and Alternative Complement Pathways



Classical Alternative


Activating agents

Immunoglobulins of human IgGl,IgG2, IgA and IgE IgG3;IgM
rabbit IgG IgM guinea pig IgG2 IgGl ruminant IgGl IgG2 Miscellaneous (Lipid A) Inulin, Zymosan,
Endotoxin or LPS, CVF Activating site Fc fragment F(ab) 2 (rabbit) or Fc fragment (human) Factors required to Cla Properdin, factor D generate C3 convertase C4 C3b C2 factor B


Total serum requirement Dilute Concentrated Divalent cation requirements Ca and Mg Mg aCl is a trimolecular complex of Clg, Clr and Cls. Classical pathway activation is initiated when Clq binds to immunoglobulin.






9



alternative pathway. Some preparations of LPS have been demonstrated to activate the classical pathway in either the presence or absence of specific antibodies. (8,24-26,27).

Normally, during classical pathway activation, the recognition stage is initiated by the fixation of the first component of the complement system, Cl, to IgG or IgM immunoglobulins of specific immune complexes. This fixation and activation of Cl generates a Cl esterase whose major substrates are the fourth and second components of the complement system.

Endotoxins in the fluid phase have been shown to initiate the activation of the complement system in vitro as described above in the presence of predominantly IgM antibodies directed against the 0-polysaccharide antigen (28). Unique to the endotoxins of certain gram negatives, however, is the capacity of the lipid A moiety of the molecule to fix Cl directly resulting in a antibody independent initiation of the classical complement cascade (8,29).

Alternative pathway activation, on the other hand, is triggered by direct activation of an initiation factor (IF) by endotoxins, or any one of a number of naturally occurring polysaccharides, and aggregates of IgA or IgE (30). It has been established that alternative pathway activation of the complement system involves the core polysaccharide region of the LPS molecule (31). Additionally,it has been shown that the length of the O-specific polysaccharide antigen and differences in the carbohydrate content may also play a role in this mechanism of activation (8).






10



In stage two of the classical pathway activation scheme, C3 convertase is generated when Cl esterase cleaves C4 and C2 in the presence of magnesium yielding the cell bound "C4b2a" enzyme. This convertase cleaves the third component of complement (C3) into two fragments, C3a and C3b. The latter fragment is unstable and has a highly reactive hydrophobic binding site. As a result of binding to target membranes at sites adjacent to the membrane bound C3 convertase, a new enzyme C5 convertase, whose major substrate becomes the fifth component of complement, is generated. In addition, C3b undergoes secondary changes giving rise to an immune-adherence site capable of binding to a variety of effector cells of the immune system which bear specific C3b receptors

(32). Additionally, C3b, either bound hydrophobically to a cell surface or free in solution, can produce further splitting of C3 via the C3 feedbackcycle of the alternative pathway described below. It is this amplification of the generation of C3b and its deposition via immune adherence onto the surface of specific target cells (and in some cases innocent bystanders), where the major function of the complement system is realized.

The mechanism of the alternative pathway generation of the C3

convertase is a bit more complicated in that there is a direct requirement for preformed C3b, the source of which is still not fully understood. This C3b in combination with factor B plus properdin and factor D forms the alternative pathway convertase "Factor BbC3b" capable of splitting C3 into C3a and C3b (33). This splits more native C3 giving rise to further C3b and the cycle is again repeated. This reaction






11



is magnesium dependent and unlike classical pathway activation it is inhibited by high concentrations of calcium (34). Under proper experimental conditions these cation requirements make serum chelated with either ethylendiamine tetraacetic acid (EDTA) or ethyleneglycol tetraacetic acid (EGTA) useful reagents for distinguishing between the two pathways (35). The former, being an effective chelator of both calcium and magnesium blocks the activation of the two pathways, while the latter, a less effective chelator of magnesium, preferentially blocks the classical pathway.

The last and final stage of complement activation is the same for both pathways and is initiated with the cleavage of C5 into two fragments, C5a and C5b by the C5 convertase. The larger C5b fragment then reacts sequentially with C6 and C7 to form either a cell bound or fluid phase trimolecular complex C5b67. The cell bound complex has the capacity to bind C8 and C9. If the cell to which this C5-C9 complex is associated is sensitive to complement mediated cytolysis, lysis ensues (36).

Fluid phase C5b67, C3a, and C5a are potent anaphylatoxins and chemotactins (37,38). Once leukocytes such as polymorphonuclear leukocytes (PMN) and macrophages have migrated to the site of complement activation, phagocytoses is initiated. As previously stated, the phagocytic process is enhanced by the fixation of complement components, especially C3b, onto the surface of particulate antigens or target tissues which promote adherence, thus facilitating ingestion (39). Release of liposomal hydrolases,either as a direct consequence of ingestion or expulsion of an indigestible target into the surrounding tissues, results in the generation of additional chemotactic factors (40,41). This





12



amplification represents an early stage in the development of an inflammatory response.

Severe tissue damage resulting from unchecked amplification of the activation of the complement system is probably minimized by the presence of two known serum inhibitors, Cl esterase inhibitor (42) and conglutinin activating factor (KAF) also known as C3b inactivator

(43). Perhaps there are other humoral or cellular inhibitory factors still to be defined.

Many of the studies cited here employed fluid phase LPS for in

vitro assays. Neter was the first to demonstrate that LPS coated onto the surface of sheep erythrocytes were sensitive to lysis by an antibody dependent classical complement pathway mechanism (44). Phillips and Mergenhagen employing sheep erythrocytes treated with LPS extracted by the Westphal-phenol procedure (45) confirmed Neter's observation showing the need for the presence of a naturally occurring y2 globulin for complement activation with a sparing consumption of the early components CI-C4 (46).

The in vivo consequences of the intravenous administration of

endotoxins to normal and complement deficient animals have been studied. Both complement pathways are activated with the classical pathway being required for the development of many of the pathophysiological symptoms such as the thrombocytopenia observed in many experimental endotoxin infections (11). Some investigators have suggested that classical pathway activation due to an antibody mediated fixation of LPS to various cells such as platelets and erythrocytes is more effective in






13





generating damage to normal cell membranes by inducing the selective release of a variety of vasoactive amines and cytolytic substances (47,48). Alternative pathway activation has not been shown to mediate the release of membrane damaging factors (49). In vivo systems are difficult to evaluate because of the large number of parameters which must be considered. Therefore, much study is needed before a complete understanding of the mechanism involved can be obtained.

Regardless of the nature of the system (in vitro or in vivo) or the state of the endotoxin (free or cell bound), the interaction of LPS with the complement system in the presence of specific antibody is extremely efficient, giving full response with sparing consumption of antibody and Cl-C4 (48). Virulence of gram negative bacteria would then appear to be related to the fate and site of free endotoxins released as a direct consequence of cell death due to phagocytosis and interaction of this free endotoxins with other cell systems in the presence of serum factors.

As previously stated, erythrocytes and now other cells have been shown to have membrane receptors which bind endotoxins. The biological role of these receptors is still not clear; however, fixation of LPS to the cell does appear to be a precondition for the triggering of many incitements of the immune system by LPS (49).

Hoffmann, in the late 1960s, isolated and described extracts from the membranes of human erythrocytes capable of inhibiting the hemolytic activity of complement when sensitized sheep erythrocytes were used as target cells (50). Extracts isolated at two different ionicities






14




yielded products having different binding affinities for the membranes of sheep erythrocytes. Material prepared at an ionic strength of 0.15 was shown to be capable of binding to sheep erythrocytes and protecting them from immune lysis, and has been designated IH inhibitor.

An extract prepared under the same conditions but at a lower

ionicity was incapable of binding to sheep erythrocytes but was capable of accelerating the decay of the complement component intermediate EAC142 to EAC14. This material was designated DAF for decay accelerating factor (51).

Preliminary studies indicate that the IH inhibitor is a large

molecule with a molecular weight greater than 250,000. It appears to contain at least two sugar moieties, glucose and galactose and it is about 4% protein by weight (50).

Existing data suggest that the IH inhibitor probably acts at the C3 convertase step. This was suggested by the finding that IH coated sheep erythrocytes in the intermediate state EAC142 consumed less C3 than untreated controls and that the inhibitory effects of IH ceased once activated C3 became fixed to the cell bound C3 convertase (51,52).

Attempts to more fully define the biochemical and biological properties of these macromolecules have been hampered by the inability to obtain them in a highly purified state.

Erythrocytes of different species differ in their susceptibilities as target cells in immune hemolysis, with sheep and chicken erythrocytes being far more sensitive than human and guinea pig erythrocytes. Differences within the same species have also been observed (53). For






15



example, it has been reported that patients suffering from paroxysmal nocturnal hemoglobinuria (PNH) have at least two populations of erythrocytes based on their susceptibility to immune hemolysis (54), with at least one subpopulation exhibiting extreme sensitivity to attack by an antibody-independent complement mediated mechanism. It has been demonstrated that extracts from the erythrocyte stromata of these patients have reduced levels of IH inhibitor activity.1 Additionally, evidence has been presented which suggests that there is a parallel between the presence of DAF on the membranes of the erythrocytes of certain species and resistance to the cytolytic effects of complement

(53).

A comparison of the isolation schemes for preparing the IH inhibitor and the LPS-receptor revealed a marked similarity between the two. Both activities are confined to that fraction of the membrane extractable by a butanol-water mixture at 40C and at an ionic strength of

0.15 or below. All of the activity of either material appeared to be localized only in the aqueous layer of the butanol extracted membranes. Ion exchange chromatography of extracts containing either the LPSreceptor or IH inhibitor activities indicate that both are eluted under identical conditions. These observations would suggest that the two activities may be similar or even identical.

The principle objective of this investigation was to determine if the IH inhibitor and LPS-receptor are either the same or closely



IData supported by personal experiments.






16



related. Although it has been observed that human erythrocytes are highly resistant to LPS mediated immune hemolysis, the reason for the refractoriness of these treated erythrocytes has not been defined. Sheep erythrocytes, as previously stated, are normally sensitive to immune lysis but may be rendered resistant by treatment with diluted extracts of the partially purified IH inhibitor. Therefore, the second objective of this investigation was to explore the biological consequences of the IH inhibitor in the interaction of LPS treated erythrocytes and serum complement, in an attempt to clearly establish a biological role for the IH inhibitor.















MATERIALS AND METHODS


Erythrocytes. Out-dated human blood (group 0, Rh positive) containing citrate-phosphate-dextrose as anticoagulant was obtained from the Civitan Regional Blood Center (Gainesville, FL). Whole sheep blood was taken by veniouncture from animals maintained at the Animal Research Laboratory of the J. Hillis Miller Health Center (Gainesville, FL). One volume of blood was mixed with an equal volume of sterile modified Alserver's solution (55) and the blood was stored at 40C for up to one month.

Preparation of erythrocyte stromata. Human and sheep erythrocyte stromata were prepared by the method of Springer et al. (56). Erythrocytes from whole blood were pelleted at 40C by centrifugation for ten minutes at 5COxg and the plasma and buffy coat were discarded. The packed cells were washed three times with phosphate buffered saline (0.13Msodium chloride plus 0.005 Mpotassium phosphate) at pH 7.4 and lysed in 10 volumes of distilled water at 40C. In the initial studies, the pH was adjusted to 5.3 with acetic acid and phenol was added to a final volume of 0.2%. The stromata were allowed to settle overnight at 40C and the supernatant fluid was removed. Ten volumes of cold distilled water were added, and the pH was readjusted to 5.3. The stromata were sedimented either by settling or centrifugation and the entire procedure was repeated six times with the addition of phenol

17




















Figure 2. A schematic representation of the extraction and
purification procedures that were employed in the
preparation of the LPS-receptor of Springer and
the IH inhibitor of Hoffmann.









WASHED PACKED RED BLOOD CELLS

(Springer) (Hoffma nn)

Water hemolysis, pH 5.3 Water hemolysis, pH 6.0
t + 0.2% phenol I

Er throcyte stroma Erythrocyte stroma

aqueous suspension (1:2), Phosphate buffer suspension
homogenization waring blender; (1:2), pH 7.5; butanol:water
butanol:water (1:2) extraction, (1:5) extraction, aqueous phase
16 hr, 40C, pH 8.0 (3X). adjusted to It of 0.15, repeat
Activity in aqueous phase extraction 6X.

Crude LPS-Receptor Crude IH

Centrifugation 33,000-151,000g DEAE column, elution 0.1 M phos>90% activity in aqueous top phate buffer, pH 7.0, with conlayer, sepharose 4B column; tinuous gradient, 0.1-1.0 M NaC1;
0.05 M Tris, pH 7.0. Sephadex G-200 column, PO4 buffer, pH 7.5, IH in void volume.
Purified LPS-Receptor Purified IH

DEAE-Sephadex A-25 column elution
0.05 M Tris, pH 7.0, with continuous gradient, 0.05-0.5 M
NaCl; Sucrose density gradient,
5%-25%; Sephadex G200 column, 0.05 M Tris, pH 7.0, receptor
in void volume.

Highly Purified LPS-Receptor





20




after every second water change. Following the final wash the stromata was collected by centrifugation, weighed and stored at -220C until use.

Isolation and purification of the LPS-receptor. The LPS-receptor was prepared as outlined in Figure 2 using a modification of the procedure of Springer et al. (15). A 50% aqueous stromata suspension was homogenized in a Waring blender and extracted overnight with two volumes of n-butyl alcohol at 40C for 16 hours at pH 8.2. Four phases, resolved after centrifugation at 2000xg for 30 minutes, organic, lipid, aqueous and solid. The aqueous phase was re-extracted twice with n-butanol, once for 30 minutes and again overnight and then thoroughly dialyzed against several changes of 0.05 M Tris-HC1 buffer (pH 7.0). An aqueous butanol extract low in LPS-receptor activity but high in IH inhibitor activity was obtained by shifting the pH of the butanol extraction from

8.2 to 5.3.

The dialyzed active crude butanol extract was centrifuged at 151,000 xg for 1-2 hours in a Spinco Model L2 preparative ultracentrifuge. Three phases resulted from the high speed centrifugation. Contrary to Springer's findings, the top aqueous layer possessed the highest LPSreceptor. After extensive dialysis, the aqueous top layer was applied to a 92 X 1.5 cm sepharose 4B column (Pharmacia Fine Chemicals, Piscataway, N.J.). The sepharose columns were washed with a 0.05 M Tris-HCI buffer at pH 7.0. Three milliliters fractions were collected and assayed for both LPS-receptor and IH inhibitor activities. The active sepharose fractions were pooled, concentrated tenfold by dialysis against 20% polyethyleneglycol in 0.05 M Tris-HCl buffer (pH 7.0) and applied to a 22.0 X 2.5 cm





21




DEAE-Sephadex A25 (Pharmacia Fine Chemicals, Piscataway, N.J.) column. After extensive washing of the column with the starting buffer, a linear sodium chloride gradient was initiated with 150 ml of 0.05 M pH

7.0 Tris-HCl buffer and 150 ml of 0.75 M, pH 7.5 Tris-HCl buffer. Three milliliter fractions were collected and assayed for the two activities.

Ether extraction. Sometimes LPS-receptor or binding activity could not be localized in the aqueous butanol phase after shift in the pH during extraction. The intermediate lipid phase was then further extracted with equal volumes of ether (Mallanchrodt, St. Louis, MO). Equal volumes of the lipid phase, suspended in 0.05 M Tris-Hcl buffer at a pH of 7.0 (1:2) and ether were vigorously mixed in a separatory funnel for 5-15 minutes. After phase separation the ether was removed from the aqueous and syrupy interphase layers by dialysis against phosphate buffered saline (pH 7.4) and from the organic phase by evaporation using a stream of nitrogen at 40C. Following evaporation, the residue remaining from the organic phase was reconstituted to its original volume with PBS and all phases were tested for LPS-receptor and IH inhibitor activities.

Polyacrylamide gel electrophoresis (PAGE). Disc polyacrylamide electrophoresis was performed using a modification of the method of Maurer (57). Extracts were applied to 7.5% acrylamide gels and were electrophoresed in a non-reducing Tris-glycine buffer system, pH 8.6, for 45 minutes at 40C. The gels were stained with 0.02% Coomassie blue containing 12.5% trichoracetic acid. For some studies, duplicate gels were run. One gel was stained as above-with the remaining gel being sliced for analysis for LPS-receptor and IH inhibitor activities.





22




Lipopolysaccharide (LPS). Lyophilized preparations of Salmonella typhimurium LPS extracted by the Boivin (58) and Westphal (45) procedures were purchased from Difco Laboratories (Detroit, MC). Heated (1000C for 3 hours) and unheated LPS stock solutions (1.0 mg/ml PBS at pH 7.4) were stored at -220C until used.

Antisera. Appropriate dilutions of Salmonella group B 0-antiserum obtained from Baltimore Biological Laboratories (Cockeysville, MD) were made in 0.01 M EDTA GVB=. Hemagglutination (HA) titers of the sera ranged from 64 to 256.

For hemolytic assays, rabit 19S antibodies to sheep erythrocyte stromata were obtained from Cordis Laboratories (Miami, FL). Stock solutions at a dilution of 1:100 in PBS (pH 7.4) were maintained until use at -220C.

Treatment of erythrocytes with LPS. Freshly acquired erythrocytes from a group 0, Rh positive adult were obtained and used in the coating and coating inhibition assays which were carried out as described by Springer et al. (14). For speed and economy, screening assays were assessed using a microtiter hemagglutination system. Briefly, the prodedure consisted of mixing equal volumes of either human or sheep erythrocytes at 2 X 108 cells/ml and dilutions of LPS for 45 minutes with shaking at 370C. After extensive washing, the smallest amount of LPS which afforded maximal hemagglutination by subsequently added antiserum was determined. This dilution, defined as the optimal coating dose, was used in all subsequent hemogglutination-inhibition assays.





23




LPS-receptor activity assay. LPS-receptor activity was determined by measuring the ability of a material to inhibit LPS fixation to erythrocytes. The procedure in the coating HA-inhibition assay differed from that in the coating test in that dilutions of LPS binding material were added to equal volumes of an optimal coating dose of LPS and incubated with shaking for 30 minutes at 370C. Erythrocytes were added, and HA titers determined as previously described. In each assay, a control consisting of LPS and erythrocytes but no LPS-receptor material followed by the subsequent addition of antiserum was included.

Isotonic buffer solutions employed in complement assays. The basic diluent for most hemolytic assays was the isotonic gelatin veronal buffer (GVB) described by Kabat and Mayer (55) which contained 0.00015 M CaC12, 0.0005 M MgCl2, and 0.1% gelatin at pH 7.5. In some cases, gelatin-veronal without CaC12 or MgC12, containing enough isotonic ethylenediaminetetra acetate (EDTA, pH 7.4) to bring the final concentration to either 0.01 M or 0.04 M, was employed. These buffers were designated as 0.01 M EDTA-GVB and 0.04 M EDTA-GVB respectively. In order to achieve maximum sensitivity, hemolytic assays involving individual complement components, and IH inhibitor assays were performed using a low ionic strength gelatin-veronal prepared by mixing equal volumes of 5.0% glucose with gelatin-veronal buffer containing twice the standard amount of CaCl2 and MgC12 (DGVB).

Sensitized sheep erythrocytes (EA). Sheep erythrocytes at a concentration of 109 per ml in 0.01 M EDTA-GVB were mixed with an equal volume of antibody to sheep stromata at-a final dilution of 1:500 in





24




the same buffer. The mixture was incubated with shaking for 30 minutes at 370C, and then at 00C for either 30 minutes or overnight. The cells were washed twice and standardized to the desired concentration before use.

Complement (GPC). Fresh frozen guinea pig complement was obtained from Pel Freeze Laboratories (Rogers, AR). The serum was shipped in dry ice and was stored at -700C until use. In some studies, aliquots of GP serum were absorbed three times at O0C with either untreated or LPS treated sheep or human erythrocytes before use.

Complement Components. Guinea pig C1 and C2 were prepared by

methods described by Nelson et al. (59) and Ruddy and Austin (60,61). Individual lyophilized functionally pure guinea pig complement components C3, C5, C6, C7, C8 and C9 were purchased from Cordis Laboratories (Miami, FL).

Complement component intermediates. For IH inhibitor assays and complement consumption studies, sheep erythrocytes in the intermediate state EACT, EAC14, and EAC142 were prepared by the methods of Borsos and Rapp (62).

Determination of Tmax of EAC142. The kinetics of the generation of EAC142 was determined by the Tmax procedure described by Borsos et al. (35).

IH inhibitor preparation. Crude human erythrocyte stromata extracts, high in IH inhibitor activity, were isolated by a procedure described by Hoffmann (50). The method is outlined in Figure 2. Briefly, the essential differences in Hoffmann's preparation of butanol





25




erythrocyte stromata extracts high in IH inhibitor activity and extracts high in LPS-receptor activity as defined by Springer are: (1) stromata were prepared at pH of 6.0 7.0 without the addition of phenol; (2) crude washed stromata were suspended in equal volume of 0.005 M potassium phosphate buffer at a pH of 7.5 and extracted with n-butanol at a final concentration of 20% for 15 minutes; and, (3) after the first extraction the aqueous butanol phase was adjusted to an ionic strength of 0.15 by the addition of 3.0 M NaCl. Butanol extraction of the adjusted material was repeated until a lipid phase could no longer be separated. For further purification, concentrated material, active in IH activity, was subjected to gel filtration and DEAE chromatography.

Treatment of sheep erythrocytes with partially purified IH inhibitor material. Lipopolysaccharide coated and untreated sheep and human erythrocytes were treated with IH inhibitor material by the procedure described by Hoffmann (50). Equal volumes of erythrocytes at 108 cells/ ml in DGVB and extracts of IH inhibitor diluted 1:10 in DGVB were mixed at 00C. The mixture was transferred to a 300C water bath, incubated 30 minutes with shaking and was pelleted at 500xg for 10 minutes at 40C. The coated cells were washed two times, and standardized to the desired concentration in the appropriate buffer.

IH inhibitor activity. IH inhibitor activity was assessed using

the EAC142 inactivation assay described by Hoffmann (50). Sheep erythrocytes in the intermediate state EACI42 at 108 /ml were mixed with an equal volume of IH inhibitor material diluted in DGVB. The reaction mixtures were incubated at 300C for 15 minutes with constant shaking,





26




after which, three volumes of guinea pig complement diluted 1:25 in

0.04 M EDTA-GVB were added. The tubes were then incubated for 60 minutes at 370C with shaking. At the end of the incubation period, 10 volumes of ice-cold PBS was added to each reaction mixture. The cells were pelleted at 500xg for 10 minutes at 40C and the optical densities of the supernatant fluids were determined at a wave length of 414 nm.

Chelators. Stock solutions of disodium ethylenediametetra acetate (EDTA, Fisher Scientific Co., Fair Lawn, NJ) and ethyleneglycolbis (beta-amino-ethyl ether) N1,N tetraacetic acid (EGTA, Sigma Chemical Co., St. Louis, MO) were prepared as described by Fine et al. (35). The stock solutions were stored at 40C and diluted to a final concentration of 200 mM before use. Magnesium EGTA was prepared as described by Fine et al. (35).

Complement Consumption. The ability of erythrocytes treated

with LPS and/or IH inhibitor to consume complement was determined in reaction mixtures containing 0.1 ml of the treated cells (1 X 109 cells) or LPS and 0.9 ml of normal or absorbed guinea pig serum chelated with either EGTA or EDTA. The mixtures were incubated with shaking 60 minutes at 370C. Following the incubation period, the cells were pelleted at 500xg at 40C for 10 minutes. The supernatant fluids were reconstituted with magnesium and/or calcium and were analyzed for residual whole complement activity using a modification of the procedure as outlined by Kabat and Mayer (55).















RESULTS


Erythrocyte coating by LPS and its inhibition. Repeated titrations of Salmonella typhimurium LPS at concentrations ranging from 0.195 ug to 50.0 uIg/ml as determined with polyvalent and homologous Salmonella antiserum, employing erythrocytes at 2.0 X 108 cells/ml were carried out. The results of a representative experiment employing polyvalent antiserum are given in Table II. It can be seen that heating the LPS enhanced the erythrocyte coating capacity to a remarkable extent. Additionally, a maximum titer resulted when erythrocytes were exposed to at least 0.78 1g/ml of heated LPS. Therefore, an optimal coating unit (g) of heated LPS (defined as the reciprocal of the greatest dilution of LPS producing complete hemagglutination by either polyvalent or homologous antiserum) was taken as 0.78 ug/ml. Assays employing homologous antiserum to heated LPS yielded lower optimal coating doses of 0.78 ug/ml and 1.56 ug/ml, depending on the age of the antiserum. These results were identical for human and sheep erythrocytes. LPS extracted by the Westphal procedure resulted in an optimal coating dose of 0.39 ug/ml as also determined with homologous antiserum.

LPS-receptor activity, as evaluated in these studies, was based on the ability of a given erythrocyte preparation to inhibit the complete fixation of an optimal coating unit of LPS onto either sheep or human


27






28








TABLE II

Determination of Optimal LPS Concentration
Used for Coating Human Erythrocytes



Titersa
LPS
Boivin

(Pg/ml) unheated heatedb


50.0 80 ND 25.0 160 ND 12.5 80 ND

6.25 80 ND 3.125 80 160 1.563 80 160 0.781 40 160 0.390 20 80 0.195 10 40


aThe reciprocal of the dilution of anti LPS serum affording maximal
hemagglutination.
b
Stock solutions of LPS (1.0 mg/ml PBS) were heated 1000C for three
hours.





29




erythrocytes. Table III, Column A summarizes the results obtained with several crude aqueous butanol preparations of erythrocyte stromata. These data indicate that the range of LPS-receptor concentrations or dilutions needed to yield optimal inhibition of LPS erythrocyte coating varied with the source, concentration and condition (solvent) of the erythrocyte stromata extraction procedure.

IH inhibitor activity of crude butanol extracts of human and

sheep erythrocyte stromata. EAC142 inactivation by a crude erythrocyte stromal extract, assayed by the technique described in the section on materials and methods is shown in Figure 3. This procedure was used to determine the inhibitory potency of most extracts. Color controls for the presence of hemoglobin in the higher concentrations of crude preparations were necessary. The reciprocal of the dilution of a crude butanol preparation yielding greater than 50% inhibition of the lysis of EAC142 by C-EDTA are also shown in Table III. These results clearly indicate that LPS-receptor and IH inhibitor activities are contained in significant amounts in crude butanol stromal extracts obtained by either Springer's or Hoffmann's procedures. Of interest also, is that the potency of the two activities varied to the same extent.

A comparison of the chromatographic properties of the IH inhibitor and LPS-receptor from erythrocyte stromal extracts. The above data suggests that the LPS-receptor and IH inhibitor are either identical or closely related molecules, therefore, additional evidence to resolve this issue was sought. Human and sheep erythrocyte stromata were subjected to modifications of Springer's purification procedure as outlined






30









TABLE III

A Comparison of LPS-receptor and IH Inhibitor
Activities of Several Crude Extracts of Erythrocyte Stromata


Titers
Crude Stromata A B
Extracts LPS-receptora IH inhibitor


Human cells extracted
at pH 8.2 (Springer)
Crude aqueous butanol phase 40 40 Crude aqueous phase
high speedc top layer 320 320 high speed syrupy interphase 80 40 Human cells extracted
at pH 7.5 (Hoffmann)
Crude aqueous butanol phase 320 320 Sheep cells extracted
at pH 8.2 (Springer)
Crude aqueous butanol phase 0 0 Crude aqueous ether phase 0 0 Crude ether interphase 8 0 Crude ether organic phase 0 0

aThe smallest amount (dilution) giving complete inhibition of LPS coating. bThe reciprocal of the greatest dilution giving 50% inhibition of EAC142
lysis.
cHigh speed extracts were obtained by the centrifugation of the crude
aqueous butanol phase, 151,000xg for 2 hours.



















Figure 3. Inhibition of the hemolysis of EAC142 by C-EDTA in
the presence of various concentrations of three crude phases resulting from the butanol extraction of human
erythrocyte stromata prepared according to the procedure of Springer. The open circles show the activity of the crude untreated butanol extracted stromata;
the closed circles show the inhibition associated
with the top fraction obtained after the high speed'
centrifugation (40,000 rpm/2 hours) of the crude
extracted stromata; the closed squares show IH inhibitor activity of interphase obtained after the
high speed centrifugation of the extracted stromata.














100


60






100





LOG2 RECIPROCAL OF THE DILUTION
LOG2 RECIPROCAL OF THE DILUTION






33




in Figure 2. Figure 4 is a gel filtration elution profile on sepharose 4B, of the crude high speed top layer obtained from the crude butanol human erythrocyte stromal preparation. Fractions were monitored at 220 and 280 nm and were assayed for LPS-receptor and IH inhibitor activities as previously described. Two peaks were observed with both activities eluting in the peak following the void volume. Close examination of the sepharose 4B profile indicates that there is a slight displacement of the IH inhibitor activity to the left of the LPS-receptor activity. This would suggest that perhaps the two activities may be different.

Further attempts to separate and purify the two activities were accomplished using ion exchange chromatography. The sepharose 4B active peaks were pooled, dialyzed against the starting Tris-HCl buffer, and applied to a DEAE-Sephadex column. Fractionation was accomplished with a linear NaCl gradient. A typical chromatogram of the partially purified material(s) is shown in Figure 5. LPS-receptor and IH inhibitor activities eluted in a relatively narrow peak at about 0.3 MNaCl, again with the IH inhibitor slightly preceding the LPS-receptor activity.

The recovery of the LPS-receptor and IH inhibitor activities following sepharose 4B and DEAE-Sephadex chromatography is presented in Table IV. It should be noted that gel filtration on sepharose 48 yielded only about onefold increase in the purity of both activities with recovery of only 52% of the LPS-receptor activity and 65% of the IH inhibitor activity. DEAE-Sephadex was shown to result in as much as a 19-fold increase in activity, but resulted in a recovery of 50% of the specific LPS-receptor activity but only one third or 31% of the








TABLE IV

Recovery of LPS-receptor and IH Inhibitor Activity from Extracts
of Human Erythrocyte Stromata Following Sepharose 4B and DEAE-Sephadex Purification


Preparation volume Concentration Activity/ml Total Total Specificd Yielde Purificationf Conc.c Activity Activity %
(ml) (A220/ml) LPSRa IH50b (A220/ml) LPSR IH50 LPSR IH150 LPSR IH50 LPSR IH50 Sepharose 4B
input
(Crude high
speed top
layer conc) 4.5 479.2 320 320 2,156.4 1440 1440 0.67 0.67 Sepharose
4B recovery 47.0 18.5 16 20 871 752 940 0.862 1.0 52.2 65.3 1.3 1.5 DEAE-Sephadex
recovery 6.1 10 128 80 61.1 780 488 12.80 7.9 54.2 31.1 19.1 11.8 aThe smallest amount (dilution) giving maximum inhibition of LPS coating.

bThe reciprocal of the highest dilution giving 50% inhibition of EACT42 lysis. CConcentration.

dCalculated Dy dividing the value for total activity by the values for total concentration.

eCalculated by dividing the value for total activity after treatment, by the value tor the total activity before treatment, multiplied by 100.

fCalculated by dividing tne value for each of the specific activities by the value for the initial specific activity.


















Figure 4. A gel filtration profile of the butanol extracted human erythrocyte stromata prepared by the method of Springer. 4.5 ml of the
extracted material were applied to a sepharose 4B column (92 X 1.5 cm) and chromatographed at 4oC with 0.05 M Tris buffer, pH
7.0. The optical density at 220 nm of the input material was
479 and 4.0 ml fractions were collected. Optical densities of the fractions at 220 are shown by the closed circles, the closed
triangles indicate the.elution pattern of the IH inhibitor activity,and the shaded bars represent the fractions having LPSreceptor activity.









140 120 20 100 80



6010


40
A

0 20




10 vo 20 30 40 50

FRACTION NUMBER (4 Oml)




















Figure 5. A DEAE-Sephadex A-25 chromatography profile of the sepharose
4B active fractions from human stromata extracts prepared by
the method of Springer. 6.1 ml of input material, with an
optical density at 220 nm of 10.0 were applied to a 22 X 2.5 cm column. A linear sodium chloride gradient of 0.05-0.75 M Tris buffer, pH 7.0 was applied and 4 ml fractions were collected. Optical densities of the fractions at 220 nm are
shown by the closed circles. The closed triangles indicate the elution position of the fractions capable of inhibiting
lysis of sheep EACI42 and the shaded bars represent the
fractions having LPS-receptor activity.
















uo!i!q!qu I %






CC%
0 r



o / E .*


+ ..... .






................................



I I









39




original IH inhibitor activity. A comparison of the gel filtration and DEAE-Sephadex elution profiles (Figures 4 and 5) and the data in Table IV, revealing a 98% loss in the total mass but with a 19-fold increase in activity, indicate good purification of the two activities. These data do, however, suggest further differences in the two activities by the differences in their respective recoveries following sepharose 4B and DEAE-Sephadex chromatography.

The homogeneity of the active DEAE-Sephadex fractions was checked by electrophoresis on 7.5% polyacrylamide gels at pH 8.6. As shown in Figure 6, 6 12 Drotein bands were observed on the crude extracts with a sharp band and a diffused staining area localized at the top of the gel when partially purified DEAE-Sephadex fractions were electrophoresed. Analysis of duplicate gels indicated that most of the LPS-receptor activity was localized in an area about 6 mm into the gels with the IH inhibitor activity spread over a fairly large area at the top of the gel with a peak of activity at about 11.0 mm. (Figure 6).

IH inhibitor and LPS-receptor activities of sheep erythrocyte stromata. The aqueous phase of butanol extracted sheep erythrocyte stromata, prepared by Springer's extraction procedure at pH 8.2 and at pH 5.3, had neither detectable LPS-receptor nor IH inhibitor activities. LPS-receptor activity but no IH inhibitor was, however, observed in an ether soluble fraction of the n-butanol lipid phase as was shown in Table III.



















Figure 6. Correlation between the distribution of LPS-receptor
and IH inhibitor activities from extracts of human erythrocyte stromata when subjected to polyacrylamide disc gel electrophoresis. Extracts were applied to duplicate 7.5% gels and were electrophoresed in a non-reducing Tris-glycine buffer,
pH 8.6 for 45 minutes at 4C. After electrophoresis, one gel was stained for protein with coomassie blue and the other was cut into suitable
segments which were eluted and analyzed for IH inhibitor and LPS-receptor activities. The closed
triangles indicate the elution position of the IH
inhibitor and the shaded bars represent the elution
profile of the LPS-receptor activity.













4 40



3 30
320


20 .4a
a 2
I.


1 10




top bottom
5 10 15 20 25

DISTANCE (mm) FROM TOP OF GEL
+
tN .






42




A significant difference in the two activities was evident when it was serendipitously observed that a shift in the pH from 8.2 to 5.3 in the butanol extraction step of the crude erythrocyte stromata resulted in a preparation consisting of little or no detectable LPS-receptor activity, but high in IH inhibitor activity. The elution profile of the IH inhibitor activity following ion exchange chromatography on DEAE-Sephadex of the partially purified material, as shown in Figure 7, was similar to that observed for material extracted at pH 8.2. LPS-receptor activity assays were not carried out on the remaining butanol phases because these materials had been discarded before the impact of these observations were realized.

A shift in the pH to 5.3 of a crude butanol extract of erythrocyte stromata obtained at a pH of 8.2 affected neither the LPS-receptor nor IH inhibitor activities. This suggested the possibility that the LPSreceptor activity of the material extracted at a pH of 5.3 was not destroyed but was probably redistributed into another phase. Stromal extractions were carried out at pH 5.3 employing smaller volumes of erythrocytes stromata, using Springer's procedure, in an attempt to localize the LPS-receptor activity. As shown in Table V, some IH inhibitor activity was observed in the crude aqueous butanol phase when 10.0 ml of packed stromata were extracted. There was no detectable LPS-receptor in this butanol layer. The lipid phase mixed with 10 volumes of PBS was further subjected to an ether extraction resulting in four layers: a pellet, an aqueous layer, lipid interphase, and an organic layer. All of the LPS-receptor activity was recovered in the organic ether layer










TABLE V

Recovery of LPS-receptor and IH Inhibitor Activity of Human Erythrocyte Stromata Extracted at pH 5.3 Using the Procedure of Sprinqer


Preparation Quantity Total Concentration Total Activity Specific Activityd (A220/ml) LPSRa I50b LPSR IH50 Aqueous butanol 16.0 ml 15.0 0 32.0 0 2.1 Organic butanol 23.0 ml NDc 0 0 Lipid 1.2 g NDc Aqueous ether 5.0 6.32 0 160 0 25.3 Lipid interphase 4.5 ND 0 144 0 ND Organic ether 12.0 313.2 192 0 0.6 0 aThe smallest amount (dilution) giving maximum inhibition of LPS activity. bThe reciprocal of the highest dilution giving 50% inhibition of EACI24 lysis CNot determined, due to high levels of particulate matter. dCalculated by dividing the value for total activity by the value for total concentration.




















Figure 7. A DEAE-Sephadex A-25 chromatography profile of the
butanol extracted human erythrocyte stromata prepared at pH 5.3 by the method of Springer. 40.0 ml
of the input material with an optical density at
220 nm of 9.9 were applied to a 22 X 2.5 cm column.
4.0 ml fractions were eluted with a linear sodium
chloride gradient of 0.05-0.75'M,Tris buffer at
pH 7.0. Optical densities of'the fractions at 220
nm are shown by the closed circles. The closed triangles indicate the elution position of the fractions capable of inhibiting lysis of sheep
EAC142 and the shaded bars represent the fractions
having LPS-receptor activity.















A 220










z~r. ~ r

z




rm~




.. .... .





~14/h

C, Ia.C

rnIIhibition





46




with IH inhibitor activity distributed in both the aqueous and lipid interphases. A comparison of the purification tables (IV and VI) of crude extracts obtained at pH 8.2 and pH 5.3 indicates that, as was observed at pH 8.2, extracts obtained at pH 5.3 and subjected to DEAESephadex chromatography resulted in a substantially greater loss of total mass as estimated by the adsorbance at 220 nm and the yield of IH inhibitor activity.

Treatment of Springer's crude butanol extracts with sheep erythrocytes. To further establish that the two activities are distinctly different, equal volumes of sheep erythrocytes at either 108 or 109 cells/mi were mixed with equal volumes of a butanol high speed top layer extract prepared according to the procedure of Springer. This extract had an initial LPS-receptor titer of 128 and a IH50 inhibitor titer greater than 80. Control tubes consisting of equal volumes of buffer and the extracts were also prepared. All tubes were mixed at 300C for 30 minutes with shaking. The cells were pelleted by centrifugation and the supernatant fluids along with the buffer controls were diluted and assayed for LPS-receptor and IH inhibitor activities. As can be seen in Figure 8, the IH inhibitor activity was reduced substantially when extracts were treated with 109 cells/ml. In contrast, the LPSreceptor activity remained constant when the cell treated and buffer control supernatants fluids were compared.

The biological consequence of the IH inhibitor and LPS-receptor

on the erythrocyte membrane. The data in the previous section indicated two essential points. First, the membrane of the human erythrocyte,



























Figure 8. Treatment of partial purified extracts of human
erythrocyte stromata prepared by the method of
Springer. Either 109 sheep erythrocytes or buffer were mixed with an equal volumes of the extracted material. Following a 30 minute incubation period at 370C, the cell suspension was
pelleted and all supernatant fluids were assayed
for LPS-receptor and IH inhibitor activities.
The hatched bars represent the sheep erythrocyte treated extracts, and the opened bars the buffer
treated extracts.










100 --100
100









50 50= 50





0 -- I 0 -. R




10 20 40 80 160 320 10 20 40 80 160 320 RECIPROCAL OF THE DILUTION









TABLE VI

Recovery of IH Inhibitor and LPS-receptor Activities From Erythrocyte Stromata Extracted at pH 5.3 by the Procedure of Springer and Subjected to DEAE-Sephadex Chromatography


Preparation Volume Concentration Activity/mi Total Total Specificd Yied Purification Conc.c Activity Activity %
(ml) (A220/ml) LPSRa I1150b (A220/ml) LPSR I11150 LPSR IH50 LPSR 11150 LPSR H1150 DEAE-Sephadex
input 40 9.89 4 640 395.6 160 25600 0.4 64.7 -- -DEAE-Sephadex 5 13.74 4 160 68.7 20 800 0.29 11.6 12.5 3 1.4 5.5 aThe smallest amount (dilution) giving maximum inhibition of LPS coating. The reciprocal of the highest dilution giving 50% inhibition of EAC142 lysis.
Concentration.

dCalculated by dividing the value for total activity by the values for total concentration. eCalculated oy dividing the value for total activity after treatment, by the value for the total activity before treatment, multiplied by 100. fCalculated by dividing the value for each of the specific activities by the value for the initial specific activity.






50




which is fairly resistant to complement mediated lysis, possesses at least two distinctly different molecules with biologically apposing properties: lipoglycoproteins with a high affinity for lipopolysaccharides which are potent activators of the complement system, and a class of molecules shown to be potent inactivators of complement. Second, sheep cells which are normally extremely sensitive to immune lysis have been shown to be devoid of the IH inhibitor but possess molecules with an affinity for LPS which are confined to the lipid moiety of the membrane.

As shown in Figure 9, the interaction of free LPS (extracted by both the Boivin and Westphal procedures) with guinea pig serum which had been absorbed with sheep E coated with LPS (E-LPS), resulted in a substantial consumption of complement. Additionally, it can be seen that LPS (Boivin) appeared to be a much more efficient activator of the alternative complement pathway compared to LPS extracted by the Westphal procedure.

LPS (extracted by both procedures) coated onto the surfaces of

sheep erythrocytes showed a similar profile (Table VII), except erythrocytes coated with LPS extracted by the Westphal procedure were far more efficient activators of complement in the absence of natural antibodies to LPS. It was of interest, therefore, to determine if LPS on the surface of sheep erythrocytes, in the presence of the IH inhibitor and E-LPS absorbed guinea pig serum, would alter the complement consumption profile of E-LPS. To explore this possibility, erythrocytes were coated with LPS and IH inhibitor then reacted with guinea pig serum (absorbed






51









TABLE VII

Consumption of Total Complement in Either E Absorbed or E-LPS Absorbed GP Serum by Untreated and LPS Coated Sheep Erythrocytes


Cell Suspension Percent Consumed E absorbed E-LPS absorbed serum serum


E

ELPS Weszphal 21.3 16.89


LPS Boivin 47.9 6.38


















Figure 9. Consumption of complement in normal and E-LPS treated
guinea pig serum by fluid phase Salmonella typhimurium
LPS extracted by the Boivin and Westphal procedures.
EDTA, EGTA, and saline treated sera were treated with
the LPS preparations one hour at 370C, then sedimented, after which the CH50 titers were determined in the magnesium and calcium reconstituted sera. The top figure
represents the profile of the Boivin treated serum
and the lower figure, the Wesphal treated serum.





53


J UNTREATED SERUM
30
E30 L PS TREATED SERUM
COMPLEMENT 20 CONSUMED

10





BUFFER EGTA EDTA




30



COMPLEMENT 2.0 CONSUMED

10





BUFFER EGTA EDTA






54




with E-LPS Westphal). Five groups of cells were prepared; one group was coated with LPS only (E-LPS); a second group was coated with LPS first, then was treated with the IH inhibitor (E-LPS-IH); a third group was treated with IH inhibitor first, followed by the LPS-receptor (E-IHLPS); a fourth group was treated with IH inhibitor only (E-IH); and, a fifth group of cells (E) was treated with PBS under the same conditions and served as the control. The efficiency of the LPS coating procedure in the presence and absence of the IH inhibitor was evaluated by assaying the five groups of cells and their respective supernatant fluids for LPS activity, employing the hemagglutination assay (described in Materials and Methods) as determined with antiserum to Salmonella typhimurium group B. The results of these assays indicated that all of the LPS treated cells adsorbed equal quantities of LPS in the presence and absence of the IH inhibitor. The complement consumption profiles of the five groups of erythrocytes treateH with E-LPS Westphal absorbed guinea pig serum are presented in Figure 10. It can be seen that the presence of IH inhibitor on the cell caused about 50% reduction in the LPS mediated hemolytic action of complement. Of particular interest was the fate of the cells when complement was treated with IH LPS coated erythrocytes; significantly less cells were lysed. This was in contrast to the case where complement was mixed with E-LPS and the cells were completely lysed. These results suggested the possibility that complement activation had taken place, but that cells were protected from lysis by the presence of IH inhibitor on the membrane.




























Figure 10. Consumption of complement in E-LPS westphal
absorbed guinea pig serum by LPS and/or IH
inhibitor coated sheep erythrocytes. One tenth ml containing 109 sheep erythrocytes
(E) either untreated or treated with LPS
(E-LPS), IH inhibitor (E-IH) or LPS and IH
inhibitor (E-LPS-IH) were incubated with
0.9 ml of guinea pig serum at 370C for 1 hr.
Residual complement hemolytic activities
were assayed and the % of the available complement consumed was calculated.







40.



%o 30 COM PLEMENT CONSUMED 20 10





E E.LPS E.LPS.IH E.IH.LPS E.IH















DISCUSSION


The experiments reported here have demonstrated that extracts

from human erythrocyte membranes possessing LPS-receptor activity obtained by the method of Springer et al. (15) were also capable of inhibiting complement mediated lysis. The anticomplementary activity of these extracts was demonstrated to share many of the properties of the IH inhibitor previously described by Hoffmann (50).

Data are presented which strongly suggest that the two biologic activities are closely associated, but separable. Evidence for this was provided by the results of five different experimental approaches in the analysis of Springer human erythrocyte stromal extracts. The first was based on the chromatographic properties on sepharose 4B and DEAE-Sephadex where slight differences between the elution patterns and recoveries of the two activities were observed. The second piece of evidence came from the electrophoretic profile of the crude and partially purified extracts on 7.5% polyacrylamide gels under nonreducing conditions. The IH inhibitor activity was shown to cover a fairly large area at the top third of the gel with the LPS-receptor activity being localized in a narrow, single band with a peak of activity near the top of the gel. A third item of evidence was based on the redistribution and separation of the two activities into different phases when the pH during the crude stromal butanol extraction procedure




57






58




shifted from 8.2 to 5.3. The fourth approach, based on the high affinity of the IH inhibitor for the membranes of sheep erythrocytes, demonstrated that the IH inhibitor activity was partially removed leaving the LPSreceptor activity unchanged by the treatment of the stromal extracts with sheep erythrocytes. Finally, evidence was presented indicating that the binding specificity of sheep erythrocytes for LPS of gram negative bacteria is localized in a lipid moiety of the crude stromal extracts and is free of all detectable IH inhibitor activity.

It should be emphasized, however, that these experiments cannot exclude the possibility that both activities may be associated on the same macromolecule with the differences reported here being a consequence of experimental manipulation. That the two activities may be a function of a single macromolecule is certainly a major possibility. Springeret al.

(16), in assessing the chemical and physical properties of a homogenous preparation of the LPS-receptor, observed that both citraconylation and dissociating polyacryamide gel electrophoresis under standard conditions yielded two fragments, one of which absorbed significantly only at 230 nm. Decitraconylation of the citraconylated fragment restored high LPS-receptor activity to only one of the fragments. These studies are only suggestive and do not permit a decision on whether the activities are on the same molecule however.

In contrast, the data obtained from sheep erythrocytes which

completely lack IH inhibitor, but which possess LPS-receptor activity, would support the finding that the two macromolecules may be distinctly different. However, the evidence would suggest that the LPS-receptors





59





on sheep cells differ from those observed on human cells, since they are confined to the lipid moiety.

Good purification of the LPS-receptor and IH inhibitor activities following DEAE-Sephadex chromatography is indicated by the quantitative data tables. These results would suggest, however, that as a preparative purification step it should be modified to encourage higher yields of the two activities.

The observation that sheep erythrocyte membranes possess molecules with receptor specificity for LPS is not surprising for it has long been known that sheep cells could be modified by the presence of LPS of gram negative bacteria and that these modified cells are readily lysed in the presence of homologous antiserum to LPS and complement (44). In contrast, hemolysis was not observed when LPS treated human erythrocytes were treated under the same conditions. This raises the possibility that human erythrocytes, a natural source of the IH inhibitor even when treated with LPS, are extremely resistant to LPS mediated lysis because of the presence of inhibitor molecules.

The findings reported here generally agree with those of Phillips et al. (46) indicating that LPS treated sheep erythrocytes can activate the complement system in the presence of natural antibodies to the LPS. In contrast to their results, however, erythrocytes coated with a preparation of LPS extracted by the procedure of Westphal were shown to be capable of activating the complement system in the absence of natural antibodies to LPS. Additionally, fluid phase LPS extracted by the Boivin method (LPS-boivin) was shown to be a more effective activator of the






60




complement system than LPS extracted by the Westphal procedure (LPSwestphal) in the presence and absence of natural antibodies to LPS. However, once LPS extracted by the Boivin becomes cell associated, its capacity to activate the complement system in the absence of natural antibodies is greatly diminished. This is significant because LPS activation of the complement system by an antibody independent mechanism requires either an exposed lipid A moiety or polysaccharide core (8,19). This would suggest then that the orientation of LPS-boivin on the membrane may be different from that of LPS-westphal, resulting in the masking of active sites necessary for the activation of complement.

The fact that different preparationsof LPS from the same species, when coated onto the surface of sheep erythrocytes, activated the complement system :o different degrees and by different pathways, depending on the presence or absence of natural antibodies to LPS, introduces the possibility that the LPS activation of complement may require substances other than the LPS molecule alone. This especially appears to be true since LPS extracted by the Westphal procedure, which was shown to activate complement in the presence and absence of natural antibody to LPS, is known to contain less protein and lipoproteins than LPS extracted by the Boivin procedure.

Placing the IH inhibitor on LPS treated sheep erythrocytes

reduced the ability of erythrocytes to consume complement. The fact that the cell is protected even when complement is activated suggests that the IH inhibitor may occupy specific sites on the red cell membrane rendering it resistant to immune lysis, but also leaving it available






61




to partially inhibit immune lysis. This may be due either to the masking of LPS-receptors resulting in less LPS uptake (demonstrated not to be the case here), or the prevention of C3b fixation to the cell membrane thus blocking the initiation of the membrane attack system of serum complement. An analogous site on human erythrocytes is already occupied by the IH inhibitor rendering this cell naturally immune to LPS mediated lysis.

The findings presented here have led us to hypothesize that a necessary criterion for resistance to LPS induced complement mediated lysis would be the localization of LPS-receptor and IH inhibitor molecules on the same membrane. This would imply that any red cell devoid of IH inhibitor molecules would be far more susceptible to the cytolytic action of LPS activated complement.

LPS of gram negative bacteria are potent activators of the complement system. It is of great clinical interest,therefore, that substances on the surface of erythrocytes which bind LPS are found closely associated with materials capable of inhibiting complement mediated lysis. The symptoms of several infectious diseases, such as typhoid fever, have been observed to include very intensive erythrophagocytic activity by macrophages of lymph nodes. The consequence of this observation could have great clinical importance. Erythrocytes coated with LPS, in contact with serum complement, would naturally lead to activation of the complement system followed by increased phagocytosis with minimal cytolysis. This would lead naturally to an amplification of the activation of the complement cascade resulting in either clearance or a heightened inflammatory response.
















REFERENCES


1. Buston, A. 1959. The in-vivo sensitization of avian erythrocytes
with Salmonella gallinarum polysaccharide. Immunology. 2:203.

2. Leive, L., V. K. Shovlin and W. E. Mergenhagen. 1968. Physical,
chemical and immunological properties of LPS released from
Escherichia coli by ethylene diaminetetra acetate. J. Biol.
Chem. 243:6384.

3. Voll, M. J. and L. Leive. 1970. Release of LPS in Escherichia
coli resistant to be the permeability increase induced by
ethylenediamine tetraacetate. J. Biol. Chem. 245:1108.

4. Rapin, A. M. C. and H. M. Kalckar. 1971. Microbial Toxins.
Weinbaum, G., S. Kadis, and S. J. Ajl, eds. Academic Press,
New York. Vol 4, pp 267-303.

5. Morrison, D. C. and L. Leive. 1975. Fractions of lipopolysaccharide from Escherichia coli Qlll: prepared by two extraction procedures. J. of Biol. Chem. 250:2911.

6. Osborn, M. J. 1969. Structure and biosynthesis of the bacterial cell wall. Ann. Rev. Biochem. 38:501.

7. Luderitz, 0., V. L. Galanos, M. Nurminen, E. T. Rietschel, G.
Rosenfelder, M. Simon, and 0. Westphal. 1973. 'Lipid A.'
Chemical structure and biological activity. J. Infect. Dis.
128:817.

8. Morrison, D. C. and L. F. Kline. 1977. Activation of the
classical and properdin pathways of complement by bacterial
lipopolysaccharides (LPS). J. Immunol. 118:362.

9. Springer, George F. and J. C. Adye. 1975. Endotoxin-binding
substances from human leukocytes and platelets. Infect. and
Immun. 12:978.

10. Hawiger, J., A. Hawrger, and S. Timmons. 1975. Endotoxin
sensitive membrane components of human platelets. Nature.
256:125.

11. Kane, M. A., J. E. May, and M. M. Frank. 1973. Interaction


62











of the classical and alternate complement pathway with endotoxin lipopolysaccharide effect on platelets and blood coagulation. J. Clin, Invest. 52:370.

12. Hill, G. J. and D. W. Weiss. 1964. Relationships between
susceptibility of mice to heat-killed Salmonellae and
endotoxin and the the affinity of their red blood cells for killed organisms. In M. Landy and W. Braun, eds. Bacterial
Endotoxins. Institute of Microbiology, Rutgers, the State
University, New Brunswick, N.J. p422-427.

13. Springer, G. F., E. T. Wang, J. H. Nichols, J. M. Shear. 1966.
Relations between bacterial lipopolysaccharide structures and
those of human cells. An.. N.Y. Sci. 133:566.

14. Springer, G. F., V. Shankar, S. V. Huprikar, and E. Neter. 1970.
Specific inhibition of endotoxin coating of red cells by a human erythrocyte membrane component. Infec. Immun. 1:98.

15. Springer, G. F., J. C. Adye, A. Bezkorovainy, and J. R. Murthy.
1973. Functional aspects and nature of the lipopolysaccharidereceptor of human erythrocytes. J. Infect. Dis. S128:5202.

16. Springer, G F., J. C. Adye, A. Bezkorovainy, and B. Jirgensons.
1974. Properties and activity of the lipopolysaccharidereceptor from human erythrocytes. Biochem. 13:1379.

17. Loos, M., D. Bitter Suermann, and M. Dierich. 1974. Interaction
of the first (Cl) and second (C2) and fourth (C4) component
of complement with different preparations of bacterial lipopolysaccharide and with lipid A. J. Immunol. 112:935.

18. Herring, W. B., J. C. Herion, R. I. Walker, and J. G. Palmer.
1963. Distribution and clearance of circulating endotoxin.
J. Clin. Invest. 42:79.

19. Mergenhagen, S. E., R. Snyderman, and J. K. Phillips. 1973.
Activation of complement by endotoxin. J. Infect. Dis.
128:S86.

20. Mjeller-Eberhard, H. J. 1975. Complement. An.. Rev. Biochem.
44:697.

21. G6tze, 0. and H. J. MUeller-Eberhard. 1976. The alternative
pathway of complement activation. Advances in Immunol. 24:1.

22. McConnell, Ian and P. J. Lachmann. 1977. Complement receptors
and cell associated complement components. Immunol. Comm.
6:111.






64




23. Pillemer, L., M. D. Schoenberg, L. Blum, and L. Wurz. 1955.
Properdin system and immunity. II. Interaction of the
properdin system with polysaccharides. Science. 122:545.

24. Muschel, L. H., K. Schmoker, and P. M. Webb. 1964. Anticomplementary action of endotoxin. Proc. Soc. Exp. Biol.
Med. 117:639.

25. Dierich, M. P., D. Bitter Suermann, W. Konig, U. Hadding, C.
Galanos, and E. J. Rietscnel. 1973. Analysis of bypass
activation of C3 by endotoxin LPS and loss of this potency.
Immunology. 24:721.

26. Marcus, R. I., H. S. Shin, and M. M. Mayer. 1971. An alternative complement pathway. C3-cleaving activity not due to C4, 2a on endotoxic lipopolysaccharide after treatment with guinea pig serum; relation to properdin. Proc. Natl. Acad.
Sc. 68:1351.

27. Galanos, C. E., T. Rietschel, 0. LUdertiz, and 0. Westphal.
1971. Interaction of lipopolysaccharides and lipid A
with complement. Eur. J. Biochem. 19:143.

28. Mller, G. and G. Michael. 1971. Frequency of antigensensitive cells to thymus-independent antigens. Cell.
Immunol. 2:309.

29. Cooper, N. R. and D. C. Morrison. 1978. Binding and activation of the first component of human complement by the lipid
A region of lipopolysaccharides. J. of Immunol. 120:1862.

30. G6tze, 0. and H. J. MUller-Eberhard. 1971. The C3 activator
system: An alternate pathway of complement activation. J.
Exp. Med. 134:905.

31. Lay, W. H. and V. Nussenzwerg. 1968. Receptors for complement on leukocytes. J. Exp. Med. 128:991.

32. Snyderman, R. and M. C. Pike. 1975. Interaction of complex
polysaccharides with the complement system: Effect of calcium
depletion on terminal component consumption. Infect. Immun.
11:273.

33. Fearon, D. T., K. F. Austen, and S. Ruddy. 1973. Formation
of a hemolytically active cellular intermediate by the
interaction between properdin factors B and D and the activated third component of complement. J. Exp. Med.
138:1305.






65




34. Marcus, R. L., H. S. Shin, and M. M. Mayer. 1971. An alternate complement pathway: C3 cleaving activity not due to
C4, 2a, on endotoxic lipopolysaccharide after treatment
with guinea pig serum; relation to properdin (complement
components). Proc. Natl. Acad. Sci. USA. 68:1351.

35. Fine, D. P., S. R. Marney, D. G. Colley, J. S. Sergent, and
R. M. Des Prez. 1972. C3 shunt activation in human serum
chelated with EGTA. J. Immunol. 109:807.

36. Mayer, M. M. 1972. Mechanism of cytolysis by complement.
Proc. Nat. Acad. Sci. USA. 69:2954.

37. Jensen, J. 1967. Anaphylatoxin and its relation to the
complement system. Science. 155:1122.

38. Shin, H., R. Snyderman, E. Friedman, A. Mellors and M. Mayer.
1968. Chemotactic and anaphylatoxic fragment cleaved from
the fifth component of guinea pig complement. Science.
162:361.

39. May, J. E., M. A. Kane, and M. M. Frank. 1972. Immune adherence by the alternative complement pathway. Proc. Soc.
Exp. Biol. Med. 141:287.

40. Rother, K. 1972. Leukocyte mobilizing factor: A new biological activity derived from the third component of complement. Eur. J. Immunol. 2:550.

41. May, J. E. and M. M. Frank. 1972. Complement mediated tissue damage: Contribution of the classical and alternate
complement pathways in the Forssmann reaction. J. Immunol.
108:1517.

42. MUeller-Eberhard, H. J. and I. H. Lepow. 1965. C'l esterase
effect on activity and physicochemical properties of the
fourth component of complement. J. Exp. Med. 121:819.

43. Ruddy, S. and K. F. Austen. 1971. C3b inactivator of man.
II. Fragments of cell bound or fluid phase C3b. J.
Immunol. 107:742.

44. Neter, E. 1956. Bacterial hemagglutination and hemolysis.
Bacteriol. Rev. 20:166.

45. Westphal, 0. and J. Kann. 1965. Bacterial lipopolysaccharides. Extraction with phenol water and further applications of the procedure. In Methods in Carbohydrate Chemistry. R. L. Whistler, ed. Academic Press,
New York. Vol 5, p83.






66




46. Phillips, J. K., R. Snyderman, and S. E. Mergenhagen. 1972.
Activation of complement by endotoxin: A role for -2 globulin, Cl, C4 and C2 in the consumption of terminal complement components by endotoxin-coated erythrocytes. J. of
Immunol. 109:334.

47. May, J. E., M. A. Kane, and M. M. Frank. 1972. Host defense
against bacterial endotoxemia Contribution of the early
and late components of comolement to detoxification. J.
Immunol. 109:393.

48. Gewurz, H., H. S. Shin, and S. E. Mergenhagen. 1968. Interactions of the complement system with endotoxic lipopolysaccharide: Consumption of each of the six terminal complement components. J. Exp. Med. 128:1049.
49. May, J. E., M. A. Kane, and M. M. Frank. 1972. Immune
adherence by the alternate complement pathway. Proc. Soc.
Exp. Biol. Med. 141:287.

50. Hoffmann, Edward M. 1969. Inhibition of complement by a
substance isolated from human erythrocytes extraction
from human erythrocyte stromata. Immunochem. 6:391.

51. Hoffmann, E. M., W. C. Cheng, E. J. Tomeu, and C. M. Renk.
1974. Resistance of sheep erythrocytes to immune lysis by
treatment of the cells with a human erythrocyte extract:
Studies on the site of inhibition. J. of Immunol. 113:1501.

52. Hoffmann, E. M. 1969. Inhibition of complement by a substance isolated from human erythrocytes. II. Studies on
the site and mechanism of action. Immunochem. 6:405.

53. Hoffmann, E. M. and H. M. Etlinger. 1973. Extraction of
complement inhibitory factors from the erythrocytes of nonhuman species. J. of Immunol. 111:946.

54. Rosse, W. F. and J. V. Dacie. 1966. Immune lysis of normal
and PNH red cells. I. Sensitivity of PNH cells. J.
Clin. Invest. 45:736.

55. Kabat, E. A. and M. M. Mayer. 1961. Complement and complement fixation. In Experimental Immunochemistry.
Charles C. Thomas, Springfield, Ill. p149.

56. Springer, G. F., Y. Nagai, and H. Tegtmeyer. 1966.
Isolation and properties of human blood-group NN and
meconium-Vg antigens. Biochem. 5:3254.











57. Maurer, H. R. 1971. Disc Electrophoresis and Related Techniques of Polyacrylamide Gel-Electrophoresis. 2nd revised ed.
Walter de Gruyter, Berlin, N.Y.

58. Boivin, A. I. 1933. Extraction of bacterial O-antigens (endotoxins). Basic Exercises in Immunochemistry. A. Nowotny, ed.
Springer-Verlag, New York. 1969. p25.

59. Nelson, R. A., J. Jensen, I. Gigli, and N. Tamura. 1966.
Methods for the separation, purification, and measurement
of nine components of hemolytic complement in guinea pig
serum. Immunochem. 3:11.

60. Ruddy, S. and K. F. Austen. 1967. A stoichiometric assay for
the fourth component of complement in whole human serum using
EACgP and functionally pure human second component. J. Immunol.
99:1162.

61. Ruddy, S. and K. F. Austen. 1969. C3 inactivator of man.
I. Hemolytic measurement by the inactivation of cell bound
C3. j. Immunol. 102:533.

62. Boros, T. and H. J. Rapp. 1967. Immune hemolysis: A simplified method for preparation of EAC4 with guinea pig or with
human complement. J. Immunol. 99:263.















BIOGRAPHY


I, Gloria Jean Jackson, born on October 4, 1939 to Charles and Sarah McKnight, was the second of four children. The only girl, I attended and was graduated May 27, 1957, valedictorian of my class at Boylan Haven, a private school for girls located at that time in Jacksonville, Florida.

I later attended Bennett College, majoring in premedicine with a minor in psychology, graduating June of 1961. Lack of funds prevented entry into medical school, therefore, I was employed as a technician at University of Florida for one year. I later entered the University of Kansas at Lawrence, Kansas to pursue a master's degree in Microbiology specializing in microbial physiology. After a year and a half I completed the investigative requirements for the master's degree and was employed as a research technician in the Department of Microbiology, where I served as an assistant to a Microbial Geneticist. In October 1965 I was married to Virgil Lawrence Jackson, following which we relocated to Chicago, Illinois.

During my three years in Chicago, I worked as a research technician at the American Medical Association Biomedical Research Institute and later as assistant supervisor of the Microbiology Research Department at the Metropolitan Sanitary District. On May 16, 1969 I gave birth to a son, Jacques Duvall, following which we relocated to Parsippany, New Jersey.


68






69



During my stay in New Jersey I was employed at the Warner-Lambert Research Institute where I assisted in studies in oral microbiology and virology. Later relocating to Westen, Connecticut and finally Dover, Massachusetts where my husband was employed as vice president of a printing company.

Following a legal separation, I returned with my son, Jacques, to Gainesville to pursue a doctorate at the University of Florida, majoring in Microbiology and specializing in Immunology. I am presently employed at Abbott Labs of North Chicago as a product manager in research and development. In addition to immunology, I enjoy tennis, chess, bridge, skiing, and sailing.










I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.




Edward M. Hoffman, Chairman. Professor of Microbiology and Cell Science







I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.



Arnold S. Bleiweis Professor of
Microbiology and Cell Science







I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.




Lester W. Clem Professor of
Immunology and Medical Microbiology










I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.



Lonnie 0. Ingram
Associate Professor of Microbiology and Cell Science







I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.



Paul W. Chun Associate Professor of Biochemistry and Molecular Biology







This dissertation was submitted to the Graduate Faculty of the College of Liberal Arts and sciences and the Graduate Counciland was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy.

December 1978



Dean, College of Liberal Arts and Sciences



Dean, Graduate School




Full Text
50
which is fairly resistant to complement mediated lysis, possesses at
least two distinctly different molecules with biologically apposing
properties: 1 ipoglycoproteins with a high affinity for lipopoly-
saccharides which are potent activators of the complement system, and
a class of molecules shown to be potent inactivators of complement.
Second, sheep cells which are normally extremely sensitive to immune
lysis have been shown to be devoid of the IH inhibitor but possess
molecules with an affinity for LPS which are confined to the lipid
moiety of the membrane.
As shown in Figure 9, the interaction of free LPS (extracted
by both the Boivin and Westphal procedures) with guinea pig serum
which had been absorbed with sheep E coated with LPS (E-LPS), resulted
in a substantial consumption of complement. Additionally, it can be
seen that LPS (Boivin) appeared to be a much more efficient activator
of the alternative complement pathway compared to LPS extracted by
the Westphal procedure.
LPS (extracted by both procedures) coated onto the surfaces of
sheep erythrocytes showed a similar profile (Table VII), except erythro-
\ cytes coated with LPS extracted by the Westphal procedure were far more
efficient activators of complement in the absence of natural antibodies
to LPS. It was of interest, therefore, to determine if LPS on the sur
face of sheep erythrocytes, in the presence of the IH inhibitor and
E-LPS absorbed guinea pig serum, would alter the complement consumption
profile of E-LPS. To explore this possibility, erythrocytes were coated
with LPS and IH inhibitor then reacted with guinea pig serum (absorbed


2
heterogenous (5), LPS of most gram negative wildtype organisms appear
to share the same basic molecular composition. As illustrated in
Figure 1, all consist of three regions. The first region, the 0-
specific polysaccharide antigen which is made up of repeating units of
five to eight monosaccharides, carries the main serologic specificity
for a given organism. Numerous serological groups differing in 0-
antigen specificity are now recognized and the polysaccharides accord
ingly show wide inter- and intrageneric variations in composition (6).
Of interest is the fact that natural antibodies to this region are
found in most animal species but do not always appear to be protective,
and in some cases a lethal gram negative bacteremia develops despite
high titers of O-specific antibody (7). Region two consists of a
short outer core which contains glucose (glc), galactose (gal), and
N-acetylglucosamine (Glc NAC), and an inner core of L-glycero-D-
mannoheptose, phosphate, ethanolamine and three molecules of 3-deoxyoc-
tonate (KDO). Unique to the endotoxins of gram negatives, 3-deoxyoc-
tonate provides a linkage site to the third region, the lipid A moiety.
Lipid A is basically composed of a phosphorylated glucosamine backbone
to which are attached fatty acids and ethanolamine residues. The nature
and distribution of the lipid A fatty acids varies among bacterial
groups with the inner core polysaccharide composition remaining con
stant (7). The complete LPS containing O-specific antigen is designated
smooth (S), and all mutants lacking O-specific side chains are referred
to as rough or R-forms.
Many early studies concerned with the interaction of endotoxins
and biological systems were carried out using either whole bacteria


THE RELATIONSHIP BETWEEN
HUMAN ERYTHROCYTE LIPOPOLYSACCHARIDE RECEPTOR
AND AN INHIBITOR OF COMPLEMENT MEDIATED LYSIS
By
GLORIA JEAN JACKSON
A DISSERTATION PRESENTED TO THE
GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1978


39
original IH inhibitor activity. A comparison of the gel filtration
and DEAE-Sephadex elution profiles (Figures 4 and 5) and the data in
Table IV, revealing a 98% loss in the total mass but with a 19-fold
increase in activity, indicate good purification of the two activities.
These data do, however, suggest further differences in the two activi
ties by the differences in their respective recoveries following sepha-
rose 4B and DEAE-Sephadex chromatography.
The homogeneity of the active DEAE-Sephadex fractions was checked
by electrophoresis on 7.5% polyacrylamide gels at pH 8.6. As shown in
Figure 6, 6 12 protein bands were observed on the crude extracts with
a sharp band and a diffused staining area localized at the top of the
gel when partially purified DEAE-Sephadex fractions were electrophoresed.
Analysis of duplicate gels indicated that most of the LPS-receptor ac
tivity was localized in an area about 6 mm into the gels with the IH
inhibitor activity spread over a fairly large area at the top of the
gel with a peak of activity at about 11.0 mm. (Figure 6).
IH inhibitor and LPS-receptor activities of sheep erythrocyte
stromata. The aqueous phase of butanol extracted sheep erythrocyte
stromata, prepared by Springer's extraction procedure at pH 8.2 and at
pH 5.3, had neither detectable LPS-receptor nor IH inhibitor activities.
LPS-receptor activity but no IH inhibitor was, however, observed in an
ether soluble fraction of the n-butanol lipid phase as was shown in
Table III.


LIST OF TABLES
TABLE PAGE
1. Comparative Properties of the Classical
and Alternative Complement Pathways 8
2. Determination of Optimal LPS Concentration
Used for Coating Human Erythrocytes 28
3. A comparison of LPS-receptor and IH Inhibitor
Activities of Several Crude Extracts of
Erythrocyte Stromata 30
4. Percent Recovery of LPS-receptor and IH
Inhibitor Activity from Extracts of Human
Erythrocyte Stromata Following Sepharose 4B
and DEAE-Sephadex Purification 34
5. Recovery of LPS-receptor and IH Inhibitor
Activity of Human Erythrocyte Stromata Extracted
at pH 5.3 Using the Procedure of Springer 43
6. Recoveries of IH Inhibitor and LPS-receptor
Activities From Erythrocyte Stromata Extracted
at pH 5.3 by the Procedure of Springer and
Subjected to DEAE-Sephadex Chromatography 49
7. Consumption of Total Complement in Either
E Absorbed or E-LPS absorbed GP Serum by
Untreated and LPS Coated sheep Erythrocytes 51
iv


Col
1-4
GlcNAc
1-2
R R or R *
1 21 3
[-Cola! -6G1 cNAcgl -2(il cal-2Gal ]-61 cal -2Galal -3G1 el -3Hepal -3Hepal -5KD0(KD0,KD0)-GlcNl -6G1 cN
n
1 1 1
ethanol amine 6 Fatty Acids
p i i p i
1 . 1 1
antigenic side chain
core lipid A
LPS1-n=l 1
LPS11-n=3.
*R1 = 0;R2 = Glcl-^4 GlcNh2;R3=Hep.


INTRODUCTION
The endotoxins or 1ipopolysaccharides (LPS) of gram negative
bacteria are among numerous antigens known to be capable of fixing to
the membranes of erythrocytes and other mammalian cells in vitro and
under certain conditions, in vivo (1). Because of their unique ability
to modulate the immune response in a wide variety of ways, they have
emerged as a complex and fascinating class of macromolecules. Func
tionally, endotoxins have been shown to have many different properties.
Due to their chemical makeup and localization in the outer membrane of
the bacterial cell envelope they have been shown to play a major role
in the establishment of a selective permeability barrier (2,3) and in
serving as receptors for certain bacteriophages (4).
Of interest to the immunochemist, however, is the fact that inter
action of LPS with components of the immune system may lead to a single
or combination of physiological responses. These include toxicity,
mitogenicity, immunogenicity, tolerance and activation of complement.
Although much is known concerning the general nature of these responses
the mechanisms cfthe cell associated events responsible for their devel
opment in the presence of LPS are still not fully understood.
A great deal of information about the chemical structure of LPS,
from a variety of organisms, has accumulated. Although it has been
recently recognized that LPS isolated from a given organism is
1


63
of the classical and alternate complement pathway with endo
toxin 1ipopolysaccharide effect on platelets and blood co
agulation. J. Clin. Invest. 52:370.
12. Hill, G. J. and D. W. Weiss. 1964. Relationships between
susceptibility of mice to heat-killed Salmonellae and
endotoxin and the the affinity of their red blood cells for
killed organisms. In M. Landy and W. Braun, eds. Bacterial
Endotoxins. Institute of Microbiology, Rutgers, the State
University, New Brunswick, N.J. p422-427.
13. Springer, G. F., E. T. Wang, J. H. Nichols, J. M. Shear. 1966.
Relations between bacterial 1ipopolysaccharide structures and
those of human cells. An.. N.Y. Sci. 133:566.
14. Springer, G. F., V. Shankar, S. V. Huprikar, and E. Neter. 1970.
Specific inhibition of endotoxin coating of red cells by a
human erythrocyte membrane component. Infec. Immun. J_:98.
15. Springer, G. F., J. C. Adye, A. Bezkorovainy, and J. R. Murthy.
1973. Functional aspects and nature of the 1ipopolysaccharide-
receptor of human erythrocytes. J. Infect. Dis. SI28:5202.
16. Springer, G F., J. C. Adye, A. Bezkorovainy, and B. Jirgensons.
1974. Properties and activity of the 1ipopolysaccharide-
receptor from human erythrocytes. Biochem. 13:1379.
17. Loos, M., D. Bitter Suermann, and M. Dierich. 1974. Interaction
of the first (Cl) and second (C2) and fourth (C4) component
of complement with different preparations of bacterial 1ipopoly
saccharide and with lipid A. J. Immunol. 112:935.
18. Herring, W. B., J. C. Herion, R. I. Walker, and J. G. Palmer.
1963. Distribution and clearance of circulating endotoxin.
J. Clin. Invest. 42:79.
19. Mergenhagen, S. E., R. Snyderman, and J. K. Phillips. 1973.
Activation of complement by endotoxin. J. Infect. Dis.
128:S86.
20. Meller-Eberhard, H. J. 1975. Complement. An.. Rev. Biochem.
44:697.
21. Gotze, 0. and H. J. Mueller-Eberhard. 1976. The alternative
pathway of complement activation. Advances in Immunol. 24:1.
22. McConnell, Ian and P. J. Lachmann. 1977. Complement receptors
and cell associated complement components. Immunol. Comm.
6:111.


Figure 2. A schematic representation of the extraction and
purification procedures that were employed in the
preparation of the LPS-receptor of Springer and
the IH inhibitor of Hoffmann.


ACKNOWLEDGMENTS
I wish to express sincere appreciation for the technical criticisms
and patient guidance given me by Dr. Edward M. Hoffmann.
I also would like to extend thanks to Dr. L. 0. Ingram, Dr. A. S.
Bleiweis, Dr. Paul W. Chun, Dr. Lester W. Clem, Dr. Paul Smith and Mr.
Jim Milam for supplying constant help and encouragement.
To all of the members of my laboratory and departmental family over
the last four and a half years, I offer my sincere gratitude for their
constant encouragement, patience, caring and all of the happy times we've
shared.
I wish also to add my thanks to Mrs. Sandy Cannella for her encour
agement and her excellent and dedicated assistance in the typing of this
manuscript.
Finally, but most important of all, for their continual prayers,
help, guidance, unfailing love and for always being there through the good
and hard times, I thank my parents Mr. and Mrs. Charles and Sarah McKnight
and my loving son Jacques. Without them this would never have been
possible.


54
with E-LPS Westphal). Five groups of cells were prepared: one group
was coated with LPS only (E-LPS); a second group was coated with LPS
first, then was treated with the IH inhibitor (E-LPS-IH); a third group
was treated with IH inhibitor first, followed by the LPS-receptor (E-IH-
LPS); a fourth group was treated with IH inhibitor only (E-IH); and, a
fifth group of cells (E) was treated with PBS under the same conditions
and served as the control. The efficiency of the LPS coating procedure
in the presence and absence of the IH inhibitor was evaluated by assay
ing the five groups of cells and their respective supernatant fluids for
LPS activity, employing the hemagglutination assay (described in Mater
ials and Methods) as determined with antiserum to Salmonella typhimurium
group B. The results of these assays indicated that all of the LPS
treated cells adsorbed equal quantities of LPS in the presence and ab
sence of the IH inhibitor. The complement consumption profiles of the
five groups of erythrocytes treated with E-LPS Westphal absorbed guinea
pig serum are presented in Figure 10. It can be seen that the presence
of IH inhibitor on the cell caused about 50% reduction in the LPS med
iated hemolytic action of complement. Of particular interest was the
fate of the cells when complement was treated with IH LPS coated
erythrocytes; significantly less cells were lysed. This was in contrast
to the case where complement was mixed with E-LPS and the cells were
completely lysed. These results suggested the possibility that comple
ment activation had taken place, but that cells were protected from
lysis by the presence of IH inhibitor on the membrane.


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
r; : -t
Edward M. Hoffman, Chairman/
Professor of / y
Microbiology and Cell Science
UUri
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Arnold S. Bleiweis
Professor of
Microbiology and Cell Science
I certify that I
conforms to acceptable
adequate, in scope and
Doctor of Philosophy.
have read this study and that in my opinion it
standards of scholarly presentation and is fully
quality, as a dissertation for the degree of
Lester W. Clem
Professor of
Immunology and Medical Microbiology


15
example, it has been reported that patients suffering from paroxysmal
nocturnal hemoglobinuria (PNH) have at least two populations of erythro
cytes based on their susceptibility to immune hemolysis (54), with at
least one subpopulation exhibiting extreme sensitivity to attack by
an antibody-independent complement mediated mechanism. It has been
demonstrated that extracts from the erythrocyte stromata of these
patients have reduced levels of IH inhibitor activity.1 Additionally,
evidence has been presented which suggests that there is a parallel
between the presence of DAF on the membranes of the erythrocytes of
certain species and resistance to the cytolytic effects of complement
(53).
A comparison of the isolation schemes for preparing the IH inhibi
tor and the LPS-receptor revealed a marked similarity between the two.
Both activities are confined to that fraction of the membrane extract-
able by a butanol-water mixture at 4C and at an ionic strength of
0.15 or below. All of the activity of either material appeared to be
localized only in the aqueous layer of the butanol extracted membranes.
Ion exchange chromatography of extracts containing either the LPS-
receptor or IH inhibitor activities indicate that both are eluted under
identical conditions. These observations would suggest that the two
activities may be similar or even identical.
The principle objective of this investigation was to determine
if the IH inhibitor and LPS-receptor are either the same or closely
^Data supported by personal experiments.


TABLE VI
Recovery of IH Inhibitor and LPS-receptor Activities
From Erythrocyte Stromata Extracted at pH 5.3 by the Procedure
of Springer and Subjected to DEAE-Sephadex Ihromatography
Preparation Volume Concentration Activity/ml Total Total Specific^ Yield6"-
Conc.c Activity Activity t
(ml) (A22o/mD LPSRa IH50b (A22o/ml) LPSR IH50 LPSR IH50 LPSR IH50
Purification
LPSR
IH
50
DEAE-Sephadex
input
40
9.89
4
640
395.6
160
25600 0.4
64.7
DEAE-Sephadex
5
13.74
4
160
68.7
20
800 0.29
11.6
aThe smallest amount (dilution) giving maximum inhibition of LPS coating.
Che reciprocal of tne highest dilution giving 50% inhibition of EAC142 lysis.
Concentration.
Calculated by dividing the value for total activity by the values for total concentration.
Calculated by dividing the value for total activity after treatment, by the value for the total activity
before treatment, multiplied by 100.
Calculated by dividing the value for each of the specific activities by the value for the initial specific
activity.


28
TABLE II
Determination of Optimal LPS Concentration
Used for Coating Human Erythrocytes
Titers3
LPS
Boivin
(pg/ml)
unheated
heated13
50.0
80
ND
25.0
160
ND
12.5
80
ND
6.25
80
ND
3.125
80
160
1.563
80
160
0.781
40
160
0.390
20
80
0.195
10
40
aThe reciprocal of the dilution of anti LPS serum affording maximal
hemagglutination,
b
Stock solutions of LPS (1.0 mg/ml PBS) were heated 100C for three
hours.


25
erythrocyte stromata extracts high in IH inhibitor activity and extracts
high in LPS-receptor activity as defined by Springer are: (1) stromata
were prepared at pH of 6.0 7.0 without the addition of phenol; (2)
crude washed stromata were suspended in equal volume of 0.005 M potas
sium phosphate buffer at a pH of 7.5 and extracted with n-butanol at
a final concentration of 20% for 15 minutes; and, (3) after the first
extraction the aqueous butanol phase was adjusted to an ionic strength
of 0.15 by the addition of 3.0 M NaCl. Butanol extraction of the ad
justed material was repeated until a lipid phase could no longer be
separated. For further purification, concentrated material, active
in IH activity, was subjected to gel filtration and DEAE chromatography.
Treatment of sheep erythrocytes with partially purified IH inhibi
tor material. Lipopolysaccharide coated and untreated sheep and human
erythrocytes were treated with IH inhibitor material by the procedure
described by Hoffmann (50). Equal volumes of erythrocytes at 108 cells/
ml in DGVB and extracts of IH inhibitor diluted 1:10 in DGVB were mixed
at 0C. The mixture was transferred to a 30C water bath, incubated
30 minutes with shaking and was pelleted at 500xg for 10 minutes at
4C. The coated cells were washed two times, and standardized to the
desired concentration in the appropriate buffer.
IH inhibitor activity. IH inhibitor activity was assessed using
the EAC142 inactivation assay described by Hoffmann (50). Sheep erythro
cytes in the intermediate state EAC142 at 108 /ml were mixed with an
equal volume of IH inhibitor material diluted in DGVB. The reaction
mixtures were incubated at 30C for 15 minutes with constant shaking,


MATERIALS AND METHODS
Erythrocytes. Out-dated human blood (group 0, Rh positive) con
taining citrate-phosphate-dextrose as anticoagulant was obtained from
the Civitan Regional Blood Center (Gainesville, FL). Whole sheep
blood was taken by venipuncture from animals maintained at the Animal
Research Laboratory of the J. Hillis Miller Health Center (Gainesville,
FL). One volume of blood was mixed with an equal volume of sterile
modified Alserver's solution (55) and the blood was stored at 4C
for up to one month.
Preparation of erythrocyte stromata. Human and sheep erythrocyte
stromata were prepared by the method of Springer et al. (56). Erythro
cytes from whole blood were pelleted at 4C by centrifugation for ten
minutes at 500.xg and the plasma and buffy coat were discarded. The
packed cells were washed three times with phosphate buffered saline
(0.13Msodium chloride plus 0.005Mpotassium phosphate) at pH 7.4 and
lysed in 10 volumes of distilled water at 4C. In the initial studies,
the pH was adjusted to 5.3 with acetic acid and phenol was added to a
final volume of 0.2%. The stromata were allowed to settle overnight
at 4C and the supernatant fluid was removed. Ten volumes of cold
distilled water were added, and the pH was readjusted to 5.3. The
stromata were sedimented either by settling or centrifugation and the
entire procedure was repeated six times with the addition of phenol
17


36
FRACTION NUMBER (4 0ml)


Figure 1. Probable structure of LPS of Escherichia coli. There
are probably three separate core structures present
with R as shown for each. Pyrophosphate bridges pro
bably cross link-adjacent lipid molecules yielding at
least trimers. Abbreviations: Col, colitose; GlcNAc,
N-acetylglucosamine; Glc, glucose; Gal, galactose;
GlcN, glucosamine; Hep, heptose; and KDO, 2 keto-3-
deoxyoctanoci acid. A protein covalently linked to
LPS may also be present (6).


THE RELATIONSHIP BETWEEN
HUMAN ERYTHROCYTE LIPOPOLYSACCHARIDE RECEPTOR
AND AN INHIBITOR OF COMPLEMENT MEDIATED LYSIS
By
GLORIA JEAN JACKSON
A DISSERTATION PRESENTED TO THE
GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1978

ACKNOWLEDGMENTS
I wish to express sincere appreciation for the technical criticisms
and patient guidance given me by Dr. Edward M. Hoffmann.
I also would like to extend thanks to Dr. L. 0. Ingram, Dr. A. S.
Bleiweis, Dr. Paul W. Chun, Dr. Lester W. Clem, Dr. Paul Smith and Mr.
Jim Milam for supplying constant help and encouragement.
To all of the members of my laboratory and departmental family over
the last four and a half years, I offer my sincere gratitude for their
constant encouragement, patience, caring and all of the happy times we've
shared.
I wish also to add my thanks to Mrs. Sandy Cannella for her encour
agement and her excellent and dedicated assistance in the typing of this
manuscript.
Finally, but most important of all, for their continual prayers,
help, guidance, unfailing love and for always being there through the good
and hard times, I thank my parents Mr. and Mrs. Charles and Sarah McKnight
and my loving son Jacques. Without them this would never have been
possible.

TABLE OF CONTENTS
PAGE
ACKNOWLEDGEMENTS
LIST OF TABLES iv
LIST OF FIGURES v
GLOSSARY OF ABBREVIATIONS vi
ABSTRACT vi i
INTRODUCTION 1
MATERIALS AND METHODS 17
RESULTS 27
DISCUSSION 57
REFERENCES 62
BIOGRAPHY 68
i i i

LIST OF TABLES
TABLE PAGE
1. Comparative Properties of the Classical
and Alternative Complement Pathways 8
2. Determination of Optimal LPS Concentration
Used for Coating Human Erythrocytes 28
3. A comparison of LPS-receptor and IH Inhibitor
Activities of Several Crude Extracts of
Erythrocyte Stromata 30
4. Percent Recovery of LPS-receptor and IH
Inhibitor Activity from Extracts of Human
Erythrocyte Stromata Following Sepharose 4B
and DEAE-Sephadex Purification 34
5. Recovery of LPS-receptor and IH Inhibitor
Activity of Human Erythrocyte Stromata Extracted
at pH 5.3 Using the Procedure of Springer 43
6. Recoveries of IH Inhibitor and LPS-receptor
Activities From Erythrocyte Stromata Extracted
at pH 5.3 by the Procedure of Springer and
Subjected to DEAE-Sephadex Chromatography 49
7. Consumption of Total Complement in Either
E Absorbed or E-LPS absorbed GP Serum by
Untreated and LPS Coated sheep Erythrocytes 51
iv

LIST OF FIGURES
FIGURE PAGE
1.Probable structure of LPS of Escherichia coli 4
2. Schematic representation of purification
procedures of the LPS-receptor and IH inhibitor.... 19
3. IH inhibitor activities of crude butanol
extracted erythrocyte membranes 32
4. A gel filtration profile of the butanol
extracted human erythrocyte stromata 36
5. DEAE-Sephadex chromatography of the
sepharose 4B active fractions from human stromata.. 38
6. Polyacrylamide disc gel electrophoresis
of human stromata extracts 41
7. DEAE-Sephadex chromatography of
human erythrocyte stromata extracts
prepared at pH 5.3 45
8. Treatment of partial purified extracts
of human erythrocyte stromata with
sheep erythrocytes 48
9. Complement consumption in LPS
treated guinea pig serum 53
10. Complement consumption of LPS
and IH inhibitor treated erythrocytes 56
v

GLOSSARY OF ABBREVIATIONS
C complement
Cl, C2, C3, C4, C5, C6, C7, C8, C91... complement components
E erythrocyte
A antibody
HA hemagglutination
HAI hemagglutination inhibition
EDTA ethylenediamine tetracetate
^The nomenclature of complement used conforms to that proposed as a re
sult of a series of World Health Organizations (Immunochemistry, _7:137,
1970).
vi

Abstract of Dissertation Presented to the
Graduate Council of the University of Florida
in Partial Fulfillment of the Requirements for the
Degree of Doctor of Philosophy
THE RELATIONSHIP BETWEEN
HUMAN ERYTHROCYTE LIPOPOLYSACCHARIDE RECEPTOR
AND AN INHIBITOR OF COMPLEMENT MEDIATED LYSIS
By
Gloria Jean Jackson
December, 1978
Chairman: Edward M. Hoffmann
Major Department: Microbiology and Cell Science
Human erythrocyte membrane extracts with receptor specificity
for the 1ipopolysaccharides of gram negative bacteria were found to
be rich in IH, a complement inhibitory substance originally isolated
and described from human red cell membranes. The possibility that
the LPS-receptor and IH inhibitor might be the same macromolecule was
considered.
In this investigation it was demonstrated that although closely
associated, the two activities are biologically distinct and separable.
This conclusion was supported by the results of five experimental ap
proaches. Small, but distinct differences were observed in the elution
profiles of the two activities when crude extracts of erythrocyte mem
branes were subjected to sepharose 4B, DEAE-Sephadex chromatography
and polyacrylamide disc gel electrophoresis. Complete separation of
the two activities was accomplished by a shift in the pH of the membrane
extraction conditions. Further, differences between the two activities
based on their ability to spontaneously bind to sheep erythrocytes were
vii

demonstrated. Finally, evidence was presented indicating that the
membranes of sheep erythrocytes consist of molecules with LPS-receptor
specificity but were devoid of IH inhibitor activity.
Additional studies indicated that LPS bound to cell membranes can
activate either the alternative or classical complement pathways and
that IH inhibitor associated with cell membranes can block LPS-induced
complement lysis of red cells despite the fact that complement activa
tion has occurred.
vi 11

INTRODUCTION
The endotoxins or 1ipopolysaccharides (LPS) of gram negative
bacteria are among numerous antigens known to be capable of fixing to
the membranes of erythrocytes and other mammalian cells in vitro and
under certain conditions, in vivo (1). Because of their unique ability
to modulate the immune response in a wide variety of ways, they have
emerged as a complex and fascinating class of macromolecules. Func
tionally, endotoxins have been shown to have many different properties.
Due to their chemical makeup and localization in the outer membrane of
the bacterial cell envelope they have been shown to play a major role
in the establishment of a selective permeability barrier (2,3) and in
serving as receptors for certain bacteriophages (4).
Of interest to the immunochemist, however, is the fact that inter
action of LPS with components of the immune system may lead to a single
or combination of physiological responses. These include toxicity,
mitogenicity, immunogenicity, tolerance and activation of complement.
Although much is known concerning the general nature of these responses
the mechanisms cfthe cell associated events responsible for their devel
opment in the presence of LPS are still not fully understood.
A great deal of information about the chemical structure of LPS,
from a variety of organisms, has accumulated. Although it has been
recently recognized that LPS isolated from a given organism is
1

2
heterogenous (5), LPS of most gram negative wildtype organisms appear
to share the same basic molecular composition. As illustrated in
Figure 1, all consist of three regions. The first region, the 0-
specific polysaccharide antigen which is made up of repeating units of
five to eight monosaccharides, carries the main serologic specificity
for a given organism. Numerous serological groups differing in 0-
antigen specificity are now recognized and the polysaccharides accord
ingly show wide inter- and intrageneric variations in composition (6).
Of interest is the fact that natural antibodies to this region are
found in most animal species but do not always appear to be protective,
and in some cases a lethal gram negative bacteremia develops despite
high titers of O-specific antibody (7). Region two consists of a
short outer core which contains glucose (glc), galactose (gal), and
N-acetylglucosamine (Glc NAC), and an inner core of L-glycero-D-
mannoheptose, phosphate, ethanolamine and three molecules of 3-deoxyoc-
tonate (KDO). Unique to the endotoxins of gram negatives, 3-deoxyoc-
tonate provides a linkage site to the third region, the lipid A moiety.
Lipid A is basically composed of a phosphorylated glucosamine backbone
to which are attached fatty acids and ethanolamine residues. The nature
and distribution of the lipid A fatty acids varies among bacterial
groups with the inner core polysaccharide composition remaining con
stant (7). The complete LPS containing O-specific antigen is designated
smooth (S), and all mutants lacking O-specific side chains are referred
to as rough or R-forms.
Many early studies concerned with the interaction of endotoxins
and biological systems were carried out using either whole bacteria

Col
1-4
GlcNAc
1-2
R R or R *
1 21 3
[-Cola! -6G1 cNAcgl -2(il cal-2Gal ]-61 cal -2Galal -3G1 el -3Hepal -3Hepal -5KD0(KD0,KD0)-GlcNl -6G1 cN
n
1 1 1
ethanol amine 6 Fatty Acids
p i i p i
1 . 1 1
antigenic side chain
core lipid A
LPS1-n=l 1
LPS11-n=3.
*R1 = 0;R2 = Glcl-^4 GlcNh2;R3=Hep.

Figure 1. Probable structure of LPS of Escherichia coli. There
are probably three separate core structures present
with R as shown for each. Pyrophosphate bridges pro
bably cross link-adjacent lipid molecules yielding at
least trimers. Abbreviations: Col, colitose; GlcNAc,
N-acetylglucosamine; Glc, glucose; Gal, galactose;
GlcN, glucosamine; Hep, heptose; and KDO, 2 keto-3-
deoxyoctanoci acid. A protein covalently linked to
LPS may also be present (6).

5
or partially purified LPS preparations in the fluid phase. This was
in spite of the fact that it had been shown that heated LPS (100C
for 60 minutes) could be coated onto the surface of a number of cells
including erythrocytes (8). The latter are often the model target
cells for hemolytic assays. It has become increasingly clear that the
biological consequence of either an in vivo or in vitro encounter with
LPS, whether cell bound or partially purified, is dependent upon its
fixation to various target cells (9). For example, it was reported
that patients suffering from endotoxin shock and sepsis due to gram
negative bacteria had greatly reduced levels of blood platelets and
that the platelets contained LPS. It has further been established
that human platelets possess an endotoxin binding receptor, which when
triggered by interaction with LPS, results in the release of a vaso
active amine, 5-hydroxytryptamine (5-HT) and the unmasking of clot-
promoting activity (10). In animal studies, it was observed that
guinea pigs injected intravenously with LPS also show a 95% decrease
in blood platelets with a concomitant shortening of the clotting time
(11). Additionally, it has been reported that there is a direct
relationship between the susceptibility of different strains of mice
to the lethal effects of Salmonel1 a endotoxin and the affinity of their
red cells for either heat-killed Salmonellae or free LPS (12).
Little was known about the nature of the attachment of LPS to
any cell until the late sixties, when G. Springer isolated an extract
from the membranes of human erythrocytes having a high affinity and
specificity for the 1ipopolysaccharides of a variety of gram negative

6
bacteria (13-15). This material, designated as an LPS-receptor, has
now been purified to homogeneity and characterized. Springer has re
ported that the LPS-receptor is a lipoglycoprotein, rich in N-acetyl-
neuraminic acid (NANA), galactose, hexosamine and contains about 61%
protein (16). It appears to be a pentameric molecule with a molecular
weight of about 228,000 daltons. The LPS-receptor functions by direct
interaction with groups on the LPS molecule which provide an attachment
site for tissue components (16). Strong evidence has accumulated sug
gesting that this attachment site is the lipid A moiety of LPS (17).
This high affinity of the LPS-receptor for endotoxins is quite remark
able because both macromolecules are highly negatively charged: the
receptor, because of its high NANA content and LPS because of its
phosphoric acid radicals.
Because the immunological specificity of LPS bound to erythro
cytes remains unchanged, Springer has suggested that the lipid A of
LPS binds to the specific receptor via clusters of hydrophobic amino
acids which makeup about 40% of the total peptide content of the
receptor leaving the polysaccharide available for the reaction with
antibodies (16). A complete understanding of the orientation of LPS
on tissues, bound either by specific receptors or by non-specific
mechanisms, maycome from studies involving the interaction of cell
bound LPS with serum complement.
The anti complementary effects of LPS have long been established.
For some time, evidence seemed to suggest that the single most impor
tant factor in the development of a noxious response to endotoxins
was the direct interaction of the lipid A region with biological

7
systems including the complement system (18). Whatever the mechanism,
mounting evidence seems to implicate complement as a mediator of a
number of the deleterious responses to endotoxins in experimental
animals and man (8). More importantly, recent evidence indicates that
the ability of LPS to initiate a complement response is not confined
to the lipid A moiety but appears to involve the polysaccharide core
as well (19).
It has become increasingly clear that a major role of the com
plement system during an immune response is the mediation of enhanced
phagocytosis with a concomitant induction of an inflammatory response.
This is accomplished by the sequential activation of the proteins which
make up the complement system. The activation process can be divided
into three major stages: a recognition stage, the generation of C3
cleavage enzymes and C3 activation stage, and a terminal or membrane
attack stage (20-22). Under normal conditions, complement proteins
exist in the serum as inactive precursors, and are activated by either
of two pathways -- the classical or alternative. The component of
these pathways and their reaction requirements are summarized in Table
1. The two share a number of similar characteristics but differ, basic
ally, in the reactants and the sequence of reactions of the first two
stages.
The capacity of endotoxins to activate the complement system by
a mechanism which requires neither antibodies to LPS nor the partici
pation of the early complement components was demonstrated more than
twenty years ago (23). More recently, it was recognized that fluid
phase activation of complement by LPS is not restricted to the

8
TABLE I
Comparative Properties of the Classical
and Alternative Complement Pathways
Classical Alternative
Activating agents
Immunoglobulins of human
rabbit
guinea pig
ruminant
Miscellaneous
Activating site
Factors required to
generate C3 convertase
Total serum requirement
Divalent cation requirements
IgGl,IgG2,
IgA and IgE
IgG3; IgM
IgG
IgM
I gG2
IgGl
IgGl
IgG2
(Lipid A)
Inulin, Zymosan,
Endotoxin or LPS,
CVF
Fc fragment
F(ab)"2 (rabbit) i
or
Fc fragment (humai
n)
Cla
Properdin, factor
D
C4
C3b
C2
factor B
Dilute
Concentrated
Ca and Mg
Mg
aCl is a trimolecular complex of Clg, Clr and Cls. Classical pathway
activation is initiated when Clq binds to immunoglobulin.

9
alternative pathway. Some preparations of LPS have been demonstrated
to activate the classical pathway in either the presence or absence
of specific antibodies- (8,24-26,27).
Normally, during classical pathway activation, the recognition
stage is initiated by the fixation of the first component of the com
plement system, Cl, to IgG or IgM immunoglobulins of specific immune
complexes. This fixation and activation of Cl generates a Cl esterase
whose major substrates are the fourth and second components of the
complement system.
Endotoxins in the fluid phase have been shown to initiate the
activation of the complement system in vitro as described above in
the presence of predominantly IgM antibodies directed against the
O-polysaccharide antigen (28). Unique to the endotoxins of certain
gram negatives, however, is the capacity of the lipid A moiety of
the molecule to fix Cl directly resulting in a antibody independent
initiation of the classical complement cascade (8,29).
Alternative pathway activation, on the other hand, is triggered
by direct activation of an initiation factor (IF) by endotoxins, or
any one of a number of naturally occurring polysaccharides, and aggre
gates of IgA or IgE (30). It has been established that alternative
pathway activation of the complement system involves the core poly
saccharide region of the LPS molecule (31). Additionally,it has been
shown that the length of the O-specific polysaccharide antigen and
differences in the carbohydrate content may also play a role in this
mechanism of activation (8).

10
In stage two of the classical pathway activation scheme, C3 con-
vertase is generated when Cl esterase cleaves C4 and C2 in the presence
of magnesium yielding the cell bound "C4b2a" enzyme. This convertase
cleaves the third component of complement (C3) into two fragments,
C3a and C3b. The latter fragment is unstable and has a highly reactive
hydrophobic binding site. As a result of binding to target membranes
at sites adjacent to the membrane bound C3 convertase, a new enzyme C5
convertase, whose major substrate becomes the fifth component of com
plement, is generated. In addition, C3b undergoes secondary changes
giving rise to an immune-adherence site capable of binding to a variety
of effector cells of the immune system which bear specific C3b receptors
(32). Additionally, C3b, either bound hydrophobically to a cell surface
or free in solution, can produce further splitting of C3 via the C3
feedback cycle of the alternative pathway described below. It is this
amplification of the generation of C3b and its deposition via immune
adherence onto the surface of specific target cells (and in some cases
innocent bystanders), where the major function of the complement system
is realized.
The mechanism of the alternative pathway generation of the C3
convertase is a bit more complicated in that there is a direct require
ment for preformed C3b, the source of which is still not fully under
stood. This C3b in combination with factor B plus properdin and factor
D forms the alternative pathway convertase "Factor BbC3b" capable of
splitting C3 into C3a and C3b (33). This splits more native C3 giving
rise to further C3b and the cycle is again repeated. This reaction

n
is magnesium dependent and unlike classical pathway activation it is
inhibited by high concentrations of calcium (34). Under proper experi
mental conditions these cation requirements make serum chelated with
either ethylendiamine tetraacetic acid (EDTA) or ethyleneglycol tetra-
acetic acid (EGTA) useful reagents for distinguishing between the two
pathways (35). The former, being an effective chelator of both calcium
and magnesium blocks the activation of the two pathways, while the
latter, a less effective chelator of magnesium, preferentially blocks
the classical pathway.
The last and final stage of complement activation is the same for
both pathways and is initiated with the cleavage of C5 into two frag
ments, C5a and C5b by the C5 convertase. The larger C5b fragment then
reacts sequentially with C6 and C7 to form either a cell bound or fluid
phase trimolecular complex C5b67. The cell bound complex has the capa
city to bind C8 and C9. If the cell to which this C5-C9 complex is
associated is sensitive to complement mediated cytolysis, lysis ensues (36).
Fluid phase C5b67, C3a, and C5a are potent anaphylatoxins and chem-
otactins (37,38). Once leukocytes such as polymorphonuclear leukocytes
(PMN) and macrophages have migrated to the site of complement activa
tion, phagocytoses is initiated. As previously stated, the phagocytic
process is enhanced by the fixation of complement components, especially
C3b, onto the surface of particulate antigens or target tissues which
promote adherence, thus facilitating ingestion (39). Release of lipo
somal hydrolases, either as a direct consequence of ingestion or expul
sion of an indigestible target into the surrounding tissues, results
in the generation of additional chemotactic factors (40,41). This

12
amplification represents an early stage in the development of an inflam
matory response.
Severe tissue damage resulting from unchecked amplification of
the activation of the complement system is probably minimized by the
presence of two known serum inhibitors, Cl esterase inhibitor (42)
and conglutinin activating factor (KAF) also known as C3b inactivator
(43). Perhaps there are other humoral or cellular inhibitory factors
still to be defined.
Many of the studies cited here employed fluid phase LPS for in
vitro assays. Neter was the first to demonstrate that LPS coated onto
the surface of sheep erythrocytes were sensitive to lysis by an anti
body dependent classical complement pathway mechanism (44). Phillips
and Mergenhagen employing sheep erythrocytes treated with LPS extracted
by the Westphal-phenol procedure (45) confirmed Neter's observation
showing the need for the presence of a naturally occurring y2 globulin
for complement activation with a sparing consumption of the early com
ponents C1-C4 (46).
The in vivo consequences of the intravenous administration of
endotoxins to normal and complement deficient animals have been studied.
Both complement pathways are activated with the classical pathway being
required for the development of many of the pathophysiological symptoms
such as the thrombocytopenia observed in many experimental endotoxin
infections (11). Some investigators have suggested that classical
pathway activation due to an antibody mediated fixation of LPS to
various cells such as platelets and erythrocytes is more effective in

13
generating damage to normal cell membranes by inducing the selective
release of a variety of vasoactive amines and cytolytic substances
(47,48). Alternative pathway activation has not been shown to mediate
the release of membrane damaging factors (49). In vivo systems are
difficult to evaluate because of the large number of parameters which
must be considered. Therefore, much study is needed before a complete
understanding of the mechanism involved can be obtained.
Regardless of the nature of the system (in vitro or in vivo) or
the state of the endotoxin (free or cell bound), the interaction of
LPS with the complement system in the presence of specific antibody is
extremely efficient, giving full response with sparing consumption of
antibody and C1-C4 (48). Virulence of gram negative bacteria would
then appear to be related to the fate and site of free endotoxins re
leased as a direct consequence of cell death due to phagocytosis and
interaction of this free endotoxins with other cell systems in the
presence of serum factors.
As previously stated, erythrocytes and now other cells have been
shown to have membrane receptors which bind endotoxins. The biological
role of these receptors is still not clear; however, fixation of LPS
to the cell does appear to be a precondition for the triggering of
many incitements of the immune system by LPS (49).
Hoffmann, in the late 1960s, isolated and described extracts from
the membranes of human erythrocytes capable of inhibiting the hemolytic
activity of complement when sensitized sheep erythrocytes were used
as target cells (50). Extracts isolated at two different ionicities

14
yielded products having different binding affinities for the membranes
of sheep erythrocytes. Material prepared at an ionic strength of 0.15
was shown to be capable of binding to sheep erythrocytes and protecting
them from immune lysis, and has been designated IH inhibitor.
An extract prepared under the same conditions but at a lower
ionicity was incapable of binding to sheep erythrocytes but was capable
of accelerating the decay of the complement component intermediate
EAC142 to EAC14. This material was designated DAF for decay accelerating
factor (51).
Preliminary studies indicate that the IH inhibitor is a large
molecule with a molecular weight greater than 250,000. It appears to
contain at least two sugar moieties, glucose and galactose and it is
about 4% protein by weight (50).
Existing data suggest that the IH inhibitor probably acts at the
C3 convertase step. This was suggested by the finding that IH coated
sheep erythrocytes in the intermediate state EAC142 consumed less C3
than untreated controls and that the inhibitory effects of IH ceased
once activated C3 became fixed to the cell bound C3 convertase (51,52).
Attempts to more fully define the biochemical and biological pro
perties of these macromolecules have been hampered by the inability
to obtain them in a highly purified state.
Erythrocytes of different species differ in their susceptibilities
as target cells in immune hemolysis, with sheep and chicken erythro
cytes being far more sensitive than human and guinea pig erythrocytes.
Differences within the same species have also been observed (53). For

15
example, it has been reported that patients suffering from paroxysmal
nocturnal hemoglobinuria (PNH) have at least two populations of erythro
cytes based on their susceptibility to immune hemolysis (54), with at
least one subpopulation exhibiting extreme sensitivity to attack by
an antibody-independent complement mediated mechanism. It has been
demonstrated that extracts from the erythrocyte stromata of these
patients have reduced levels of IH inhibitor activity.1 Additionally,
evidence has been presented which suggests that there is a parallel
between the presence of DAF on the membranes of the erythrocytes of
certain species and resistance to the cytolytic effects of complement
(53).
A comparison of the isolation schemes for preparing the IH inhibi
tor and the LPS-receptor revealed a marked similarity between the two.
Both activities are confined to that fraction of the membrane extract-
able by a butanol-water mixture at 4C and at an ionic strength of
0.15 or below. All of the activity of either material appeared to be
localized only in the aqueous layer of the butanol extracted membranes.
Ion exchange chromatography of extracts containing either the LPS-
receptor or IH inhibitor activities indicate that both are eluted under
identical conditions. These observations would suggest that the two
activities may be similar or even identical.
The principle objective of this investigation was to determine
if the IH inhibitor and LPS-receptor are either the same or closely
^Data supported by personal experiments.

16
related. Although it has been observed that human erythrocytes are
highly resistant to LPS mediated immune hemolysis, the reason for the
refractoriness of these treated erythrocytes has not been defined.
Sheep erythrocytes, as previously stated, are normally sensitive to
immune lysis but may be rendered resistant by treatment with diluted
extracts of the partially purified IH inhibitor. Therefore, the second
objective of this investigation was to explore the biological conse
quences of the IH inhibitor in the interaction of LPS treated erythro
cytes and serum complement, in an attempt to clearly establish a bio
logical role for the IH inhibitor.

MATERIALS AND METHODS
Erythrocytes. Out-dated human blood (group 0, Rh positive) con
taining citrate-phosphate-dextrose as anticoagulant was obtained from
the Civitan Regional Blood Center (Gainesville, FL). Whole sheep
blood was taken by venipuncture from animals maintained at the Animal
Research Laboratory of the J. Hillis Miller Health Center (Gainesville,
FL). One volume of blood was mixed with an equal volume of sterile
modified Alserver's solution (55) and the blood was stored at 4C
for up to one month.
Preparation of erythrocyte stromata. Human and sheep erythrocyte
stromata were prepared by the method of Springer et al. (56). Erythro
cytes from whole blood were pelleted at 4C by centrifugation for ten
minutes at 500.xg and the plasma and buffy coat were discarded. The
packed cells were washed three times with phosphate buffered saline
(0.13Msodium chloride plus 0.005Mpotassium phosphate) at pH 7.4 and
lysed in 10 volumes of distilled water at 4C. In the initial studies,
the pH was adjusted to 5.3 with acetic acid and phenol was added to a
final volume of 0.2%. The stromata were allowed to settle overnight
at 4C and the supernatant fluid was removed. Ten volumes of cold
distilled water were added, and the pH was readjusted to 5.3. The
stromata were sedimented either by settling or centrifugation and the
entire procedure was repeated six times with the addition of phenol
17

Figure 2. A schematic representation of the extraction and
purification procedures that were employed in the
preparation of the LPS-receptor of Springer and
the IH inhibitor of Hoffmann.

WASHED PACKED RED BLOOD CELLS
(Springer)
(Hoffmann)
Water hemolysis, pH 5.3
+ 0.2% phenol
Water hemolysis, pH 6.0
Erythrocyte stroma
Erythrocyte stroma
aqueous suspension (1:2),
homogenization waring blender;
butanol:water (1:2) extraction,
16 hr, 4C, pH 8.0 (3X).
Activity in aqueous phase
Phosphate buffer suspension
(1:2), pH 7.5; butanol:water
(1:5) extraction, aqueous phase
adjusted to y of 0.15, repeat
extraction 6X.
Crude LPS-Receptor
Centrifugation 33,000-151,000g
>90% activity in aqueous top
layer, sepharose 4B column;
0.05 M Tris, pH 7.0.
Purified LPS-Receptor
DEAE-Sephadex A-25 column elution
0.05 M Tris, pH 7.0, with con
tinuous gradient, 0.05-0.5 M
NaCl; Sucrose density gradient,
5%-25%; Sephadex G200 column,
0.05 M Tris, pH 7.0, receptor
in void volume.
Crude IH
DEAE column, elution 0.1 M phos
phate buffer, pH 7.0, with con
tinuous gradient, 0.1-1.0 M NaCl;
Sephadex G-200 column, P0^ buffer,
pH 7.5, IH in void volume.
Purified IH
Highly Purified LPS-Receptor

20
after every second water change. Following the final wash the stromata
was collected by centrifugation, weighed and stored at -22C until use.
Isolation and purification of the LPS-receptor. The LPS-receptor
was prepared as outlined in Figure 2 using a modification of the pro
cedure of Springer et a 1. (15). A 50% aqueous stromata suspension was
homogenized in a Waring blender and extracted overnight with two volumes
of n-butyl alcohol at 4C for 16 hours at pH 8.2. Four phases, resolved
after centrifugation at 2000xg for 30 minutes, organic, lipid, aqueous
and solid. The aqueous phase was re-extracted twice with n-butanol,
once for 30 minutes and again overnight and then thoroughly dialyzed
against several changes of 0.05 M Tris-HCl buffer (pH 7.0). An aqueous
butanol extract low in LPS-receptor activity but high in IH inhibitor
activity was obtained by shifting the pH of the butanol extraction from
8.2 to 5.3.
The dialyzed active crude butanol extract was centrifuged at 151,000
xg for 1-2 hours in a Spinco Model L2 preparative ultracentrifuge.
Three phases resulted from the high speed centrifugation. Contrary
to Springer's findings, the top aqueous layer possessed the highest LPS-
receptor. After extensive dialysis, the aqueous top layer was applied
to a 92 X 1.5 cm sepharose 4B column (Pharmacia Fine Chemicals, Piscataway,
N.J.). The sepharose columns were washed with a 0.05 M Tris-HCl buffer at
pH 7.0. Three milliliters fractions were collected and assayed for both
LPS-receptor and IH inhibitor activities. The active sepharose fractions
were pooled, concentrated tenfold by dialysis against 20% polyethylene-
glycol in 0.05 M Tris-HCl buffer (pH 7.0) and applied to a 22.0 X 2.5 cm

21
DEAE-Sephadex A25 (Pharmacia Fine Chemicals, Piscataway, N.J.) column.
After extensive washing of the column with the starting buffer, a
linear sodium chloride gradient was initiated with 150 ml of 0.05 M pH
7.0 Tris-HCl buffer and 150 ml of 0.75 M, pH 7.5 Tris-HCl buffer. Three
milliliter fractions were collected and assayed for the two activities.
Ether extraction. Sometimes LPS-receptor or binding activity could
not be localized in the aqueous butanol phase after shift in the pH
during extraction. The intermediate lipid phase was then further ex
tracted with equal volumes of ether (Mallanchrodt, St. Louis, M0).
Equal volumes of the lipid phase, suspended in 0.05 M Tris-Hcl buffer
at a pH of 7.0 (1:2) and ether were vigorously mixed in a separatory
funnel for 5-15 minutes. After phase separation the ether was removed
from the aqueous and syrupy interphase layers by dialysis against phos
phate buffered saline (pH 7.4) and from the organic phase by evapora
tion using a stream of nitrogen at 4C. Following evaporation, the
residue remaining from the organic phase was reconstituted to its
original volume with PBS and all phases were tested for LPS-receptor
and IH inhibitor activities.
Polyacrylamide gel electrophoresis (PAGE). Disc polyacrylamide
electrophoresis was performed using a modification of the method of
Maurer (57). Extracts were applied to 7.5% acrylamide gels and were
electrophoresed in a non-reducing Tris-glycine buffer system, pH 8.6,
for 45 minutes at 4C. The gels were stained with 0.02% Coomassie blue
containing 12.5% trichoracetic acid. For some studies, duplicate gels
were run. One gel was stained as above'with the remaining gel being
sliced for analysis for LPS-receptor and IH inhibitor activities.

22
Lipopolysaccharide (LPS). Lyophilized preparations of Salmonella
typhimurium LPS extracted by the Boivin (58) and Westphal (45) proce
dures were purchased from Difco Laboratories (Detroit, MC). Heated
(100C for 3 hours) and unheated LPS stock solutions (1.0 mg/ml PBS
at pH 7.4) were stored at -22C until used.
Antisera. Appropriate dilutions of Salmonella group B 0-antiserum
obtained from Baltimore Biological Laboratories (Cockeysville, MD)
were made in 0.01 M EDTA GVB^. Hemagglutination (HA) titers of the sera
ranged from 64 to 256.
For hemolytic assays, rabit 19S antibodies to sheep erythrocyte
stromata were obtained from Cordis Laboratories (Miami, FL). Stock
solutions at a dilution of 1:100 in PBS (pH 7.4) were maintained until
use at -22C.
Treatment of erythrocytes with LPS. Freshly acquired erythrocytes
from a group 0, Rh positive adult were obtained and used in the coating
and coating inhibition assays which were carried out as described by
Springer et al. (14). For speed and economy, screening assays were as
sessed using a microtiter hemagglutination system. Briefly, the prodedure
consisted of mixing equal volumes of either human or sheep erythrocytes
at 2 X 10 cells/ml and dilutions of LPS for 45 minutes with shaking at
37C. After extensive washing, the smallest amount of LPS which afford
ed maximal hemagglutination by subsequently added antiserum was determined.
This dilution, defined as the optimal coating dose, was used in all sub
sequent hemogglutination-inhibition assays.

23
LPS-receptor activity assay. LPS-receptor activity was determined
by measuring the ability of a material to inhibit LPS fixation to
erythrocytes. The procedure in the coating HA-inhibition assay differed
from that in the coating test in that dilutions of LPS binding material
were added to equal volumes of an optimal coating dose of LPS and in
cubated with shaking for 30 minutes at 37C. Erythrocytes were added,
and HA titers determined as previously described. In each assay, a
control consisting of LPS and erythrocytes but no LPS-receptor material
followed by the subsequent addition of antiserum was included.
Isotonic buffer solutions employed in complement assays. The basic
diluent for most hemolytic assays was the isotonic gelatin veronal
buffer (6VB) described by Kabat and Mayer (55) which contained 0.00015
M CaCl^* 0.0005 M MgCl2> and 0.1% gelatin at pH 7.5. In some cases,
gelatin-veronal without CaCl2 or MgCl2, containing enough isotonic
ethylenediaminetetra acetate (EDTA, pH 7.4) to bring the final concen
tration to either 0.01 M or 0.04 M, was employed. These buffers were
designated as 0.01 M EDTA-GVB and 0.04 M EDTA-GVB respectively. In order
to achieve maximum sensitivity, hemolytic assays involving individual
complement components, and IH inhibitor assays were performed using a
low ionic strength gelatin-veronal prepared by mixing equal volumes of
5.0% glucose with gelatin-veronal buffer containing twice the standard
amount of CaCl2 and MgCl2 (DGVB).
Sensitized sheep erythrocytes (EA). Sheep erythrocytes at a con
centration of 109 per ml in 0.01 M EDTA-GVB were mixed with an equal
volume of antibody to sheep stromata at-a final dilution of 1 :500 in

24
the same buffer. The mixture was incubated with shaking for 30 minutes
at 37C, and then at 0C for either 30 minutes or overnight. The cells
were washed twice and standardized to the desired concentration before
use.
Complement (GPC). Fresh frozen guinea pig complement was obtained
from Pel Freeze Laboratories (Rogers, AR). The serum was shipped in
dry ice and was stored at -70C until use. In some studies, aliquots
of GP serum were absorbed three times at 0C with either untreated or
LPS treated sheep or human erythrocytes before use.
Complement Components. Guinea pig Cl and C2 were prepared by
methods described by Nelson et al. (59) and Ruddy and Austin (60,61).
Individual lyophilized functionally pure guinea pig complement compo
nents C3, C5, C6, C7, C8 and C9 were purchased from Cordis Laboratories
(Miami, FL).
Complement component intermediates. For IH inhibitor assays and
complement consumption studies, sheep erythrocytes in the intermediate
state EAC1, EAC14, and EAC142 were prepared by the methods of Borsos
and Rapp (62).
Determination of Tmax of EAC142. The kinetics of the generation
of EAC142 was determined by the Tmax procedure described by Borsos
et al. (35).
IH inhibitor preparation. Crude human erythrocyte stromata ex
tracts, high in IH inhibitor activity, were isolated by a procedure
described by Hoffmann (50). The method is outlined in Figure 2.
Briefly, the essential differences in Hoffmann's preparation of butanol

25
erythrocyte stromata extracts high in IH inhibitor activity and extracts
high in LPS-receptor activity as defined by Springer are: (1) stromata
were prepared at pH of 6.0 7.0 without the addition of phenol; (2)
crude washed stromata were suspended in equal volume of 0.005 M potas
sium phosphate buffer at a pH of 7.5 and extracted with n-butanol at
a final concentration of 20% for 15 minutes; and, (3) after the first
extraction the aqueous butanol phase was adjusted to an ionic strength
of 0.15 by the addition of 3.0 M NaCl. Butanol extraction of the ad
justed material was repeated until a lipid phase could no longer be
separated. For further purification, concentrated material, active
in IH activity, was subjected to gel filtration and DEAE chromatography.
Treatment of sheep erythrocytes with partially purified IH inhibi
tor material. Lipopolysaccharide coated and untreated sheep and human
erythrocytes were treated with IH inhibitor material by the procedure
described by Hoffmann (50). Equal volumes of erythrocytes at 108 cells/
ml in DGVB and extracts of IH inhibitor diluted 1:10 in DGVB were mixed
at 0C. The mixture was transferred to a 30C water bath, incubated
30 minutes with shaking and was pelleted at 500xg for 10 minutes at
4C. The coated cells were washed two times, and standardized to the
desired concentration in the appropriate buffer.
IH inhibitor activity. IH inhibitor activity was assessed using
the EAC142 inactivation assay described by Hoffmann (50). Sheep erythro
cytes in the intermediate state EAC142 at 108 /ml were mixed with an
equal volume of IH inhibitor material diluted in DGVB. The reaction
mixtures were incubated at 30C for 15 minutes with constant shaking,

26
after which, three volumes of guinea pig complement diluted 1:25 in
0.04 M EDTA-GVB were added. The tubes were then incubated for 60 min
utes at 37C with shaking. At the end of the incubation period, 10
volumes of ice-cold PBS was added to each reaction mixture. The cells
were pelleted at 500xg for 10 minutes at 4C and the optical densities
of the supernatant fluids were determined at a wave length of 414 nm.
Chelators. Stock solutions of disodium ethylenediametetra acetate
(EDTA, Fisher Scientific Co., Fair Lawn, NJ) and ethyleneglycolbis
(beta-amino-ethyl ether) N^,N tetraacetic acid (EGTA, Sigma Chemical
Co., St. Louis, M0) were prepared as described by Fine et al. (35).
The stock solutions were stored at 4C and diluted to a final concen
tration of 200 mM before use. Magnesium EGTA was prepared as described
by Fine et al. (35).
Complement Consumption. The ability of erythrocytes treated
with LPS and/or IH inhibitor to consume complement was determined in
reaction mixtures containing 0.1 ml of the treated cells (1 X 10^ cells)
or LPS and 0.9 ml of normal or absorbed guinea pig serum chelated with
either EGTA or EDTA. The mixtures were incubated with shaking 60
minutes at 37C. Following the incubation period, the cells were pel
leted at 500xg at 4C for 10 minutes. The supernatant fluids were
reconstituted with magnesium and/or calcium and were analyzed for
residual whole complement activity using a modification of the pro
cedure as outlined by Kabat and Mayer (55).

RESULTS
Erythrocyte coating by LPS and its inhibition. Repeated titra
tions of Salmonella typhimurium LPS at concentrations ranging from
0.195 ug to 50.0 ug/ml as determined with polyvalent and homologous
Salmonella antiserum, employing erythrocytes at 2.0 X 10 cells/ml
were carried out. The results of a representative experiment em
ploying polyvalent antiserum are given in Table II. It can be seen
that heating the LPS enhanced the erythrocyte coating capacity to a
remarkable extent. Additionally, a maximum titer resulted when erythro
cytes were exposed to at least 0.78 ug/ml of heated LPS. Therefore,
an optimal coating unit (u) of heated LPS (defined as the reciprocal
of the greatest dilution of LPS producing complete hemagglutination by
either polyvalent or homologous antiserum) was taken as 0.78 ug/ml.
Assays employing homologous antiserum to heated LPS yielded lower
optimal coating doses of 0.78 ug/ml and 1.56 ug/ml, depending on the
age of the antiserum. These results were identical for human and sheep
erythrocytes. LPS extracted by the Westphal procedure resulted in an
optimal coating dose of 0.39 ug/ml as also determined with homologous
antiserum.
LPS-receptor activity, as evaluated in these studies, was based on
the ability of a given erythrocyte preparation to inhibit the complete
fixation of an optimal coating unit of LPS onto either sheep or human
27

28
TABLE II
Determination of Optimal LPS Concentration
Used for Coating Human Erythrocytes
Titers3
LPS
Boivin
(pg/ml)
unheated
heated13
50.0
80
ND
25.0
160
ND
12.5
80
ND
6.25
80
ND
3.125
80
160
1.563
80
160
0.781
40
160
0.390
20
80
0.195
10
40
aThe reciprocal of the dilution of anti LPS serum affording maximal
hemagglutination,
b
Stock solutions of LPS (1.0 mg/ml PBS) were heated 100C for three
hours.

29
erythrocytes. Table III, Column A summarizes the results obtained with
several crude aqueous butanol preparations of erythrocyte stromata.
These data indicate that the range of LPS-receptor concentrations or
dilutions needed to yield optimal inhibition of LPS erythrocyte coating
varied with the source, concentration and condition (solvent) of the
erythrocyte stromata extraction procedure.
IH inhibitor activity of crude butanol extracts of human and
sheep erythrocyte stromata. EAC142 inactivation by a crude erythrocyte
stromal extract, assayed by the technique described in the section on
materials and methods is shown in Figure 3. This procedure was used to
determine the inhibitory potency of most extracts. Color controls for
the presence of hemoglobin in the higher concentrations of crude prepara
tions were necessary. The reciprocal of the dilution of a crude butanol
preparation yielding greater than 50% inhibition of the lysis of EAC142
by C-EDTA are also shown in Table III. These results clearly indicate
that LPS-receptor and IH inhibitor activities are contained in signifi
cant amounts in crude butanol stromal extracts obtained by either Spring
er's or Hoffmann's procedures. Of interest also, is that the potency
of the two activities varied to the same extent.
A comparison of the chromatographic properties of the IH inhibitor
and LPS-receptor from erythrocyte stromal extracts. The above data sug
gests that the LPS-receptor and IH inhibitor are either identical or
closely related molecules, therefore, additional evidence to resolve
this issue was sought. Human and sheep erythrocyte stromata were sub
jected to modifications of Springer's purification procedure as outlined

30
TABLE III
A Comparison of LPS-receptor and IH Inhibitor
Activities of Several Crude Extracts of Erythrocyte Stromata
Titers
Crude Stromata
Extracts
A
LPS-receptora
IH
B .
inhibitorD
Human cells extracted
at pH 8.2 (Springer)
Crude aqueous butanol phase
40
40
Crude aqueous phase
high speedc top layer
320
320
high speed syrupy interphase
80
40
Human cells extracted
at pH 7.5 (Hoffmann)
Crude aqueous butanol phase
320
320
Sheep cells extracted
at pH 8.2 (Springer)
Crude aqueous butanol phase
0
0
Crude aqueous ether phase
0
0
Crude ether interphase
8
0
Crude ether organic phase
0
0
aThe smallest amount (dilution) giving complete inhibition of LPS coating.
^The reciprocal of the greatest dilution giving 50% inhibition of EAC142
lysis.
c
High speed extracts were obtained by the centrifugation of the crude
aqueous butanol phase, 151,OOOxy for 2 hours.

Figure 3. Inhibition of the hemolysis of EAC142 by C-EDTA in
the presence of various concentrations of three crude
phases resulting from the butanol extraction of human
erythrocyte stromata prepared according to the pro
cedure of Springer. The open circles show the activ
ity of the crude untreated butanol extracted stromata;
the closed circles show the inhibition associated
with the top fraction obtained after the high speed
centrifugation (40,000 rpm/2 hours) of the crude
extracted stromata; the closed squares show IH in
hibitor activity of interphase obtained after the
high speed centrifugation of the extracted stromata.

100
60
30
10
log2 reciprocal of the dilution
CO
PO

33
in Figure 2. Figure 4 is a gel filtration elution profile on sepharose
4B, of the crude high speed top layer obtained from the crude butanol
human erythrocyte stromal preparation. Fractions were monitored at 220
and 280 nm and were assayed for LPS-receptor and IH inhibitor activities
as previously described. Two peaks were observed with both activities
eluting in the peak following the void volume. Close examination of
the sepharose 4B profile indicates that there is a slight displacement
of the IH inhibitor activity to the left of the LPS-receptor activity.
This would suggest that perhaps the two activities may be different.
Further attempts to separate and purify the two activities were
accomplished using ion exchange chromatography. The sepharose 4B active
peaks were pooled, dialyzed against the starting Tris-HCl buffer, and
applied to a DEAE-Sephadex column. Fractionation was accomplished with
a linear NaCl gradient. A typical chromatogram of the partially puri
fied material(s) is shown in Figure 5. LPS-receptor and IH inhibitor
activities eluted in a relatively narrow peak at about 0.3MNaCl, again
with the IH inhibitor slightly preceding the LPS-receptor activity.
The recovery of the LPS-receptor and IH inhibitor activities
following sepharose 4B and DEAE-Sephadex chromatography is presented
in Table IV. It should be noted that gel filtration on sepharose 4B
yielded only about onefold increase in the purity of both activities
with recovery of only 52% of the LPS-receptor activity and 65% of the
IH inhibitor activity. DEAE-Sephadex was shown to result in as much as
a 19-fold increase inactivity, but resulted in a recovery of 50%
of the specific LPS-receptor activity but only one third or 31% of the

TABLE IV
Recovery of LPS-receptor and IH Inhibitor Activity from Extracts
of Human Erythrocyte Stromata Following Sepharose 4B
and DEAE-Sephadex Purification
Preparation Volume
Concentration Activity/ml Total
Total
Specific0* Yielde
Purification*"
Conc.c
Activity
Activity
%
(ml)
(A220/ml) LPSRa IH5Qb (A220/ml)
LPSR IH50
LPSR IH50
LPSR IH go
LPSR IH50
Sepharose 4B
input
(Crude high
speed top
layer cone) 4.5
479.2
320
320
2,156.4
1440
1440
0.67
0.67


Sepharose
4B recovery 47.0
18.5
16
20
871
752
940
0.862
1.0
52.2
65.3
DEAE-Sepha
dex
recovery 6.1
10
128
80
61.1
780
488
12.80
7.9
54.2
31.1
aThe smallest amount (dilution) giving maximum inhibition of LPS coating.
bThe reciprocal of the highest dilution giving 50% inhibition of EAC142 lysis.
c
Concentration.
Calculated oy dividing the value for total activity by the values for total concentration.
0
Calculated oy dividing the value for total activity after treatment, by the value tor the total activity before
treatment, multiplied by luO.
^Calculated by dividing the value for each of the specific activities by the value for the initial specific activity.

Figure 4. A gel filtration profile of the butanol extracted human erythro
cyte stromata prepared by the method of Springer. 4.5 ml of the
extracted material were applied to a sepharose 4B column (92 X
1.5 cm) and chromatographed at 4C with 0.05 M Tris buffer, pH
7.0. The optical density at 220 nm of the input material was
479 and 4.0 ml fractions were collected. Optical densities of
the fractions at 220 are shown by the closed circles, the closed
triangles indicate the elution pattern of the IH inhibitor ac
tivity, and the shaded bars represent the fractions having LPS-
receptor activity.

36
FRACTION NUMBER (4 0ml)

Figure 5. A DEAE-Sephadex A-25 chromatography profile of the sepharose
4B active fractions from human stromata extracts prepared by
the method of Springer. 6.1 ml of input material, with an
optical density at 220 nm of 10.0 were applied to a 22 X 2.5
cm column. A linear sodium chloride gradient of 0.05-0.75 M
Tris buffer, pH 7.0 was applied and 4 ml fractions were col
lected. Optical densities of the fractions at 220 nm are
shown by the closed circles. The closed triangles indicate
the elution position of the fractions capable of inhibiting
lysis of sheep EAC42 and the shaded bars represent the
fractions having LPS-receptor activity.

u
20 170 190 210 230
FRACTION NUMBER (4.0ml)
CO
CO
% Inhibition

39
original IH inhibitor activity. A comparison of the gel filtration
and DEAE-Sephadex elution profiles (Figures 4 and 5) and the data in
Table IV, revealing a 98% loss in the total mass but with a 19-fold
increase in activity, indicate good purification of the two activities.
These data do, however, suggest further differences in the two activi
ties by the differences in their respective recoveries following sepha-
rose 4B and DEAE-Sephadex chromatography.
The homogeneity of the active DEAE-Sephadex fractions was checked
by electrophoresis on 7.5% polyacrylamide gels at pH 8.6. As shown in
Figure 6, 6 12 protein bands were observed on the crude extracts with
a sharp band and a diffused staining area localized at the top of the
gel when partially purified DEAE-Sephadex fractions were electrophoresed.
Analysis of duplicate gels indicated that most of the LPS-receptor ac
tivity was localized in an area about 6 mm into the gels with the IH
inhibitor activity spread over a fairly large area at the top of the
gel with a peak of activity at about 11.0 mm. (Figure 6).
IH inhibitor and LPS-receptor activities of sheep erythrocyte
stromata. The aqueous phase of butanol extracted sheep erythrocyte
stromata, prepared by Springer's extraction procedure at pH 8.2 and at
pH 5.3, had neither detectable LPS-receptor nor IH inhibitor activities.
LPS-receptor activity but no IH inhibitor was, however, observed in an
ether soluble fraction of the n-butanol lipid phase as was shown in
Table III.

Figure 6. Correlation between the distribution of LPS-receptor
and IH inhibitor activities from extracts of human
erythrocyte stromata when subjected to polyacryla
mide disc gel electrophoresis. Extracts were ap
plied to duplicate 7.5% gels and were electro-
phoresed in a non-reducing Tris-glycine buffer,
pH 8.6 for 45 minutes at 4C. After electro
phoresis, one gel was stained for protein with co-
omassie blue and the other was cut into suitable
segments which were eluted and analyzed for IH in
hibitor and LPS-receptor activities. The closed
triangles indicate the elution position of the IH
inhibitor and the shaded bars represent the elution
profile of the LPS-receptor activity.

40
30
20
10
bottom
DISTANCE (mm) FROM TOP OF GEL
% Inhibition

42
A significant difference in the two activities was evident when
it was serendipitously observed that a shift in the pH from 8.2 to 5.3
in the butanol extraction step of the crude erythrocyte stromata resulted
in a preparation consisting of little or no detectable LPS-receptor activ
ity, but high in IH inhibitor activity. The elution profile of the IH
inhibitor activity following ion exchange chromatography on DEAE-Sephadex
of the partially purified material, as shown in Figure 7, was similar
to that observed for material extracted at pH 8.2. LPS-receptor activity
assays were not carried out on the remaining butanol phases because these
materials had been discarded before the impact of these observations were
realized.
A shift in the pH to 5.3 of a crude butanol extract of erythrocyte
stromata obtained at a pH of 8.2 affected neither the LPS-receptor nor
IH inhibitor activities. This suggested the possibility that the LPS-
receptor activity of the material extracted at a pH of 5.3 was not de
stroyed but was probably redistributed into another phase. Stromal ex
tractions were carried out at pH 5.3 employing smaller volumes of erythro
cytes stromata, using Springer's procedure, in an attempt to localize
the LPS-receptor activity. As shown in Table V, some IH inhibitor activ
ity was observed in the crude aqueous butanol phase when 10.0 ml of
packed stromata were extracted. There was no detectable LPS-receptor
in this butanol layer. The lipid phase mixed with 10 volumes of PBS
was further subjected to an ether extraction resulting in four layers:
a pellet, an aqueous layer, lipid interphase, and an organic layer.
All of the LPS-receptor activity was recovered in the organic ether layer

TABLE V
Recovery of l.PS-receptor and IH Inhibitor Activity of Human
Erythrocyte Stromata Extracted at pH 5.3 Using the Procedure of Springer
Preparation
Quantity
Total Concentration
(A220/ml)
Total Activity
LPSRa IH5Qb
Specific
LPSR
Activityd
ih50
Aqueous butanol
16.0 ml
15.0
0
32.0
0
2.1
Organic butanol
23.0 ml
NDC
0
0
Lipid
1.2 g
NDC
Aqueous ether
5.0
6.32
0
160
0
25.3
Lipid interphase
4.5
ND
0
144
0
ND
Organic ether
12.0
313.2
192
0
0.6
0
aThe smallest amount (dilution) giving maximum inhibition of LPS activity.
bThe reciprocal of the highest dilution giving 50% inhibition of EAC124 lysis
cNot determined, due to high levels of particulate matter.
Calculated by dividing the value for total activity by the value for total concentration.

Figure 7. A DEAE-Sephadex A-25 chromatography profile of the
butanol extracted human erythrocyte stromata pre
pared at pH 5.3 by the method of Springer. 40.0 ml
of the input material with an optical density at
220 nm of 9.9 were applied to a 22 X 2.5 cm column.
4.0 ml fractions were eluted with a linear sodium
chloride gradient of 0.05-0.75!M, Tris buffer at
pH 7.0. Optical densities of the fractions at 220
nm are shown by the closed circles. The closed
triangles indicate the elution position of the
fractions capable of inhibiting lysis of sheep
EAC142 and the shaded bars represent the fractions
having LPS-receptor activity.

FRACTION NUMBER (4.0ml)
A 220
co 4* en
o
% Inhibition

46
with IH inhibitor activity distributed in both the aqueous and
lipid interphases. A comparison of the purification tables (IV and VI)
of crude extracts obtained at pH 8.2 and pH 5.3 indicates that, as was
observed at pH 8.2, extracts obtained at pH 5.3 and subjected to DEAE-
Sephadex chromatography resulted in a substantially greater loss of
total mass as estimated by the adsorbance at 220 nm and the yield of IH
inhibitor activity.
Treatment of Springer's crude butanol extracts with sheep erythro
cytes To further establish that the two activities are distinctly dif
ferent, equal volumes of sheep erythrocytes at either 10^ or 109 cell s/ml
were mixed with equal volumes of a butanol high speed top layer ex
tract prepared according to the procedure of Springer. This extract
had an initial LPS-receptor titer of 128 and a IH50 inhibitor titer
greater than 80. Control tubes consisting of equal volumes of buffer
and the extracts were also prepared. All tubes were mixed at 30C for
30 minutes with shaking. The cells were pelleted by centrifugation
and the supernatant fluids along with the buffer controls were diluted
and assayed for LPS-receptor and IH inhibitor activities. As can be
seen in Figure 8, the IH inhibitor activity was reduced substantially
when extracts were treated with 109 cells/ml. In contrast, the LPS-
receptor activity remained constant when the cell treated and buffer
control supernatants fluids were compared.
The biological consequence of the IH inhibitor and LPS-receptor
on the erythrocyte membrane. The data in the previous section indicated
two essential points. First, the membrane of the human erythrocyte,

Figure 8. Treatment of partial purified extracts of human
erythrocyte stromata prepared by the method of
Springer. Either lo" sheep erythrocytes or buf
fer were mixed with an equal volumes of the ex
tracted material. Following a 30 minute incu
bation period at 37C, the cell suspension was
pelleted and all supernatant fluids were assayed
for LPS-receptor and IH inhibitor activities.
The hatched bars represent the sheep erythrocyte
treated extracts, and the opened bars the buffer
treated extracts.

10 20 40 80 160 320
RECIPROCAL
100 -
5 50
O
JB
MB
JE
C
I
i
I
¡
I
ii
1
10 20 40 80 160 320
THE DILUTION

TABLE VI
Recovery of IH Inhibitor and LPS-receptor Activities
From Erythrocyte Stromata Extracted at pH 5.3 by the Procedure
of Springer and Subjected to DEAE-Sephadex Ihromatography
Preparation Volume Concentration Activity/ml Total Total Specific^ Yield6"-
Conc.c Activity Activity t
(ml) (A22o/mD LPSRa IH50b (A22o/ml) LPSR IH50 LPSR IH50 LPSR IH50
Purification
LPSR
IH
50
DEAE-Sephadex
input
40
9.89
4
640
395.6
160
25600 0.4
64.7
DEAE-Sephadex
5
13.74
4
160
68.7
20
800 0.29
11.6
aThe smallest amount (dilution) giving maximum inhibition of LPS coating.
Che reciprocal of tne highest dilution giving 50% inhibition of EAC142 lysis.
Concentration.
Calculated by dividing the value for total activity by the values for total concentration.
Calculated by dividing the value for total activity after treatment, by the value for the total activity
before treatment, multiplied by 100.
Calculated by dividing the value for each of the specific activities by the value for the initial specific
activity.

50
which is fairly resistant to complement mediated lysis, possesses at
least two distinctly different molecules with biologically apposing
properties: 1 ipoglycoproteins with a high affinity for lipopoly-
saccharides which are potent activators of the complement system, and
a class of molecules shown to be potent inactivators of complement.
Second, sheep cells which are normally extremely sensitive to immune
lysis have been shown to be devoid of the IH inhibitor but possess
molecules with an affinity for LPS which are confined to the lipid
moiety of the membrane.
As shown in Figure 9, the interaction of free LPS (extracted
by both the Boivin and Westphal procedures) with guinea pig serum
which had been absorbed with sheep E coated with LPS (E-LPS), resulted
in a substantial consumption of complement. Additionally, it can be
seen that LPS (Boivin) appeared to be a much more efficient activator
of the alternative complement pathway compared to LPS extracted by
the Westphal procedure.
LPS (extracted by both procedures) coated onto the surfaces of
sheep erythrocytes showed a similar profile (Table VII), except erythro-
\ cytes coated with LPS extracted by the Westphal procedure were far more
efficient activators of complement in the absence of natural antibodies
to LPS. It was of interest, therefore, to determine if LPS on the sur
face of sheep erythrocytes, in the presence of the IH inhibitor and
E-LPS absorbed guinea pig serum, would alter the complement consumption
profile of E-LPS. To explore this possibility, erythrocytes were coated
with LPS and IH inhibitor then reacted with guinea pig serum (absorbed

51
TABLE VII
Consumption of Total Complement
or E-LPS Absorbed
by Untreated and LPS Coated
in Either E Absorbed
GP Serum
Sheep Erythrocytes
Cell Suspension
Percent
Consumed
E absorbed
E-LPS absorbed
serum
serum
E
^LPS Westphal
21.3
16.89
ULPS Boivin
47.9
6.38

Figure 9. Consumption of complement in normal and E-LPS treated
guinea pig serum by fluid phase Salmonella typhimurium
LPS extracted by the Boivin and Westphal procedures,
EDTA, EGTA, and saline treated sera were treated with
the LPS preparations one hour at 37C, then sedimented,
after which the CH50 titers were determined in the mag
nesium and calcium reconstituted sera. The top figure
represents the profile of the Boivin treated serum
and the lower figure, the Wesphal treated serum.

53
%
COMPLEMENT

CONSUMED
UNTREATED
SERUM
%
COMPLEMENT
CONSUMED
BUFFER EGTA EDTA

54
with E-LPS Westphal). Five groups of cells were prepared: one group
was coated with LPS only (E-LPS); a second group was coated with LPS
first, then was treated with the IH inhibitor (E-LPS-IH); a third group
was treated with IH inhibitor first, followed by the LPS-receptor (E-IH-
LPS); a fourth group was treated with IH inhibitor only (E-IH); and, a
fifth group of cells (E) was treated with PBS under the same conditions
and served as the control. The efficiency of the LPS coating procedure
in the presence and absence of the IH inhibitor was evaluated by assay
ing the five groups of cells and their respective supernatant fluids for
LPS activity, employing the hemagglutination assay (described in Mater
ials and Methods) as determined with antiserum to Salmonella typhimurium
group B. The results of these assays indicated that all of the LPS
treated cells adsorbed equal quantities of LPS in the presence and ab
sence of the IH inhibitor. The complement consumption profiles of the
five groups of erythrocytes treated with E-LPS Westphal absorbed guinea
pig serum are presented in Figure 10. It can be seen that the presence
of IH inhibitor on the cell caused about 50% reduction in the LPS med
iated hemolytic action of complement. Of particular interest was the
fate of the cells when complement was treated with IH LPS coated
erythrocytes; significantly less cells were lysed. This was in contrast
to the case where complement was mixed with E-LPS and the cells were
completely lysed. These results suggested the possibility that comple
ment activation had taken place, but that cells were protected from
lysis by the presence of IH inhibitor on the membrane.

Figure 10. Consumption of complement in E-LPS westphal
absorbed guinea pig serum by LPS and/or IH
inhibitor coated sheep erythrocytes. One
tenth ml containing 10^ sheep erythrocytes
(E) either untreated or treated with LPS
(E-LPS), IH inhibitor (E-IH) or LPS and IH
inhibitor (E-LPS-IH) were incubated with
0.9 ml of guinea pig serum at 37C for 1 hr.
Residual complement hemolytic activities
were assayed and the % of the available com
plement consumed was calculated.

40
0/
/o
COM PLEMENT
CONSUMED
E.LPS.IH E.IH.LPS E.IH
en
en

DISCUSSION
The experiments reported here have demonstrated that extracts
from human erythrocyte membranes possessing LPS-receptor activity ob
tained by the method of Springer et al. (15) were also capable of inhi
biting complement mediated lysis. The anticomplementary activity of
these extracts was demonstrated to share many of the properties of the
IH inhibitor previously described by Hoffmann (50).
Data are presented which strongly suggest that the two biologic
activities are closely associated, but separable. Evidence for this
was provided by the results of five different experimental approaches
in the analysis of Springer human erythrocyte stromal extracts. The
first was based on the chromatographic properties on sepharose 4B and
DEAE-Sephadex where slight differences between the elution patterns
and recoveries of the two activities were observed. The second piece
of evidence came from the electrophoretic profile of the crude and
partially purified extracts on 7.5% polyacrylamide gels under non
reducing conditions. The IH inhibitor activity was shown to cover
a fairly large area at the top third of the gel with the LPS-receptor
activity being localized in a narrow, single band with a peak of
activity near the top of the gel. A third item of evidence was based on
the redistribution and separation of the two activities into different
phases when the pH during the crude stromal butanol extraction procedure
57

58
shifted from 8.2 to 5.3. The fourth approach, based on the high affinity
of the IH inhibitor for the membranes of sheep erythrocytes, demonstrated
that the IH inhibitor activity was partially removed leaving the LPS-
receptcr activity unchanged by the treatment of the stromal extracts
with sheep erythrocytes. Finally, evidence was presented indicating
that the binding specificity of sheep erythrocytes for LPS of gram nega
tive bacteria is localized in a lipid moiety of the crude stromal ex
tracts and is free of all detectable IH inhibitor activity.
It should be emphasized, however, that these experiments cannot
exclude the possibility that both activities may be associated on the
same macromolecule with the differences reported here being a consequence
of experimental manipulation. That the two activities may be a function
of a single macromolecule is certainly a major possibility. Springeret al.
(16), in assessing the chemical and physical properties of a homogenous
preparation of the LPS-receptor, observed that both citraconylation and
dissociating polyacryamide gel electrophoresis under standard conditions
yielded two fragments, one of which absorbed significantly only at 230
nm. Decitraconylation of the citraconylated fragment restored high
LPS-receptor activity to only one of the fragments. These studies are
only suggestive and do not permit a decision on whether the activities
are on the same molecule however.
In contrast, the data obtained from sheep erythrocytes which
completely lack IH inhibitor, but which possess LPS-receptor activity,
would support the finding that the two macromolecules may be distinctly
different. However, the evidence would suggest that the LPS-receptors

59
on sheep cells differ from those observed on human cells, since they
are confined to the lipid moiety.
Good purification of the LPS-receptor and IH inhibitor activities
following DEAE-Sephadex chromatography is indicated by the quantitative
data tables. These results would suggest, however, that as a pre
parative purification step it should be modified to encourage higher
yields of the two activities.
The observation that sheep erythrocyte membranes possess mole
cules with receptor specificity for LPS is not surprising for it has
long been known that sheep cells could be modified by the presence
of LPS of gram negative bacteria and that these modified cells are
readily lysed in the presence of homologous antiserum to LPS and com
plement (44). In contrast, hemolysis was not observed when LPS treated
human erythrocytes were treated under the same conditions. This raises
the possibility that human erythrocytes, a natural source of the IH
inhibitor even when treated with LPS, are extremely resistant to LPS
mediated lysis because of the presence of inhibitor molecules.
The findings reported here generally agree with those of Phillips
et al. (46) indicating that LPS treated sheep erythrocytes can activate
the complement system in the presence of natural antibodies to the LPS.
In contrast to their results, however, erythrocytes coated with a pre
paration of LPS extracted by the procedure of Westphal were shown to be
capable of activating the complement system in the absence of natural
antibodies to LPS. Additionally, fluid phase LPS extracted by the Boivin
method (LPS-boivin) was shown to be a more effective activator of the

60
complement system than LPS extracted by the Westphal procedure (LPS-
westphal) in the presence and absence of natural antibodies to LPS.
However, once LPS extracted by the Boivin becomes cell associated, its
capacity to activate the complement system in the absence of natural
antibodies is greatly diminished. This is significant because LPS
activation of the complement system by an antibody independent mech
anism requires either an exposed lipid A moiety or polysaccharide core
(8,19). This would suggest then that the orientation of LPS-boivin on
the membrane may be different from that of LPS-westphal, resulting in
the masking of active sites necessary for the activation of complement.
The fact that different preparations of LPS from the same species,
when coated onto the surface of sheep erythrocytes, activated the com
plement system to different degrees and by different pathways, depending
on the presence or absence of natural antibodies to LPS, introduces
the possibility that the LPS activation of complement may require sub
stances other than the LPS molecule alone. This especially appears to
be true since LPS extracted by the Westphal procedure, which was shown
to activate complement in the presence and absence of natural antibody
to LPS, is known to contain less protein and lipoproteins than LPS
extracted by the Boivin procedure.
Placing the IH inhibitor on LPS treated sheep erythrocytes
reduced the ability of erythrocytes to consume complement. The fact
that the cell is protected even when complement is activated suggests
that the IH inhibitor may occupy specific sites on the red cell membrane
rendering it resistant to immune lysis, but also leaving it available

61
to partially inhibit immune lysis. This may be due either to the masking
of LPS-receptors resulting in less LPS uptake (demonstrated not to be
the case here), or the prevention of C3b fixation to the cell membrane
thus blocking the initiation of the membrane attack system of serum
complement. An analogous site on human erythrocytes is already occupied
by the IH inhibitor rendering this cell naturally immune to LPS mediated
lysis.
The findings presented here have led us to hypothesize that a
necessary criterion for resistance to LPS induced complement mediated
lysis would be the localization of LPS-receptor and IH inhibitor mole
cules on the same membrane. This would imply that any red cell devoid
of IH inhibitor molecules would be far more susceptible to the cytolytic
action of LPS activated complement.
LPS of gram negative bacteria are potent activators of the com
plement system. It is of great clinical interest, therefore, that sub
stances on the surface of erythrocytes which bind LPS are found closely
associated with materials capable of inhibiting complement mediated lysis.
The symptoms of several infectious diseases, such as typhoid fever, have
been observed to include very intensive erythrophagocytic activity by
macrophages of lymph nodes. The consequence of this observation could
have great clinical importance. Erythrocytes coated with LPS, in contact
with serum complement, would naturally lead to activation of the com
plement system followed by increased phagocytosis with minimal cytolysis.
This would lead naturally to an amplification of the activation of the
complement cascade resulting in either clearance or a heightened inflam
matory response.

REFERENCES
1. Buston, A. 1959. The in-vivo sensitization of avian erythrocytes
with Salmonella gall inarum polysaccharide. Immunology. 2;203.
2. Leive, L., V. K. Shovlin and W. E. Mergenhagen. 1968. Physical,
chemical and immunological properties of LPS released from
Escherichia coli by ethylene diaminetetra acetate. J. Biol.
Chem. 243:6384.
3. Voll, M. J. and L. Leive. 1970. Release of LPS in Escherichia
coli resistant to be the permeability increase induced by
ethylenediamine tetraacetate. J. Biol. Chem. 245:1108.
4. Rapin, A. M. C. and H. M. Kalckar. 1971. Microbial Toxins.
Weinbaum, G., S. Kadis, and S. J. Ajl, eds. Academic Press,
New York. Vol 4, pp 267-303.
5. Morrison, D. C. and L. Leive. 1975. Fractions of lipopoly-
saccharide from Escherichia col i Qlll: prepared by two ex
traction procedures. J. of Biol. Chem. 250:2911.
6. Osborn, M. J. 1969. Structure and biosynthesis of the bac
terial cell wall. Ann. Rev. Biochem. 38:501.
7. Luderitz, 0., V. L. Galanos, M. Nurminen, E. T. Rietschel, G.
Rosenfelder, M. Simon, and 0. Westphal. 1973. 'Lipid A.1
Chemical structure and biological activity. J. Infect. Dis.
128:817.
8. Morrison, D. C. and L. F. Kline. 1977. Activation of the
classical and properdin pathways of complement by bacterial
1ipopolysaccharides (LPS). J. Immunol. 118:362.
9. Springer, George F. and J. C. Adye. 1975. Endotoxin-binding
substances from human leukocytes and platelets. Infect, and
Immun. 12:978.
10. Hawiger, J., A. Hawrger, and S. Timmons. 1975. Endotoxin
sensitive membrane components of human platelets. Nature.
256:125.
11. Kane, M. A., J. E. May, and M. M. Frank. 1973. Interaction
62

63
of the classical and alternate complement pathway with endo
toxin 1ipopolysaccharide effect on platelets and blood co
agulation. J. Clin. Invest. 52:370.
12. Hill, G. J. and D. W. Weiss. 1964. Relationships between
susceptibility of mice to heat-killed Salmonellae and
endotoxin and the the affinity of their red blood cells for
killed organisms. In M. Landy and W. Braun, eds. Bacterial
Endotoxins. Institute of Microbiology, Rutgers, the State
University, New Brunswick, N.J. p422-427.
13. Springer, G. F., E. T. Wang, J. H. Nichols, J. M. Shear. 1966.
Relations between bacterial 1ipopolysaccharide structures and
those of human cells. An.. N.Y. Sci. 133:566.
14. Springer, G. F., V. Shankar, S. V. Huprikar, and E. Neter. 1970.
Specific inhibition of endotoxin coating of red cells by a
human erythrocyte membrane component. Infec. Immun. J_:98.
15. Springer, G. F., J. C. Adye, A. Bezkorovainy, and J. R. Murthy.
1973. Functional aspects and nature of the 1ipopolysaccharide-
receptor of human erythrocytes. J. Infect. Dis. SI28:5202.
16. Springer, G F., J. C. Adye, A. Bezkorovainy, and B. Jirgensons.
1974. Properties and activity of the 1ipopolysaccharide-
receptor from human erythrocytes. Biochem. 13:1379.
17. Loos, M., D. Bitter Suermann, and M. Dierich. 1974. Interaction
of the first (Cl) and second (C2) and fourth (C4) component
of complement with different preparations of bacterial 1ipopoly
saccharide and with lipid A. J. Immunol. 112:935.
18. Herring, W. B., J. C. Herion, R. I. Walker, and J. G. Palmer.
1963. Distribution and clearance of circulating endotoxin.
J. Clin. Invest. 42:79.
19. Mergenhagen, S. E., R. Snyderman, and J. K. Phillips. 1973.
Activation of complement by endotoxin. J. Infect. Dis.
128:S86.
20. Meller-Eberhard, H. J. 1975. Complement. An.. Rev. Biochem.
44:697.
21. Gotze, 0. and H. J. Mueller-Eberhard. 1976. The alternative
pathway of complement activation. Advances in Immunol. 24:1.
22. McConnell, Ian and P. J. Lachmann. 1977. Complement receptors
and cell associated complement components. Immunol. Comm.
6:111.

64
23. Pillemer, L., M. D. Schoenberg, L. Blum, and L. Wurz. 1955.
Properdin system and immunity. II. Interaction of the
properdin system with polysaccharides. Science. 122:545.
24. Muschel, L. H., K. Schmoker, and P. M. Webb. 1964. Anti
complementary action of endotoxin. Proc. Soc. Exp. Biol.
Med. 117:639.
25. Dierich, M. P., D. Bitter Suermann, W. Konig, U. Hadding, C.
Galanos, and E. J. Rietschel. 1973. Analysis of bypass
activation of C3 by endotoxin LPS and loss of this potency.
Immunology. 24:721.
26. Marcus, R. I., H. S. Shin, and M. M. Mayer. 1971. An alter
native complement pathway. C3-c1eaving activity not due to
C4, 2a on endotoxic 1ipopolysaccharide after treatment with
guinea pig serum; relation to properdin. Proc. Natl. Acad.
Sc. 68:1351 .
27. Galanos, C. E., T. Rietschel, 0. Liidertiz, and 0. Westphal.
1971. Interaction of 1ipopolysaccharides and lipid A
with complement. Eur. J. Biochem. 19:143.
28. Moller, G. and G. Michael. 1971. Frequency of antigen-
sensitive cells to thymus-independent antigens. Cell.
Immunol. j?:309.
29. Cooper, N. R. and D. C. Morrison. 1978. Binding and activa
tion of the first component of human complement by the lipid
A region of 1ipopolysaccharides. J. of Immunol. 120:1862.
30. Gotze, 0. and H. J. M1ler-Eberhard. 1971. The C3 activator
system: An alternate pathway of complement activation. J.
Exp. Med. 134:905.
31. Lay, W. H. and V. Nussenzwerg. 1968. Receptors for com
plement on leukocytes. J. Exp. Med. 128:991.
32. Snyderman, R. and M. C. Pike. 1975. Interaction of complex
polysaccharides with the complement system: Effect of calcium
depletion on terminal component consumption. Infect. Immun.
U:273.
33. Fearon, D. T., K. F. Austen, and S. Ruddy. 1973. Formation
of a hemolytically active cellular intermediate by the
interaction between properdin factors B and D and the
activated third component of complement. J. Exp. Med.
138:1305.

65
34. Marcus, R. L., H. S. Shin, and M. M. Mayer. 1971. An alter
nate complement pathway: C3 cleaving activity not due to
C4, 2a, on endotoxic 1ipopolysaccharide after treatment
with guinea pig serum; relation to properdin (complement
components). Proc. Natl. Acad. Sci. USA. 68:1351.
35. Fine, D. P., S. R. Mamey, D. G. Colley, J. S. Sergent, and
R. M. Des Prez. 1972. C3 shunt activation in human serum
chelated with EGTA. J. Immunol. 109:807.
36. Mayer, M. M. 1972. Mechanism of cytolysis by complement.
Proc. Nat. Acad. Sci. USA. 69:2954.
37. Jensen, J. 1967. Anaphylatoxin and its relation to the
complement system. Science. 155:1122.
38. Shin, H., R. Snyderman, E. Friedman, A. Mellors and M. Mayer.
1968. Chemotactic and anaphylatoxic fragment cleaved from
the fifth component of guinea pig complement. Science.
162:361.
39. May, J. E., M. A. Kane, and M. M. Frank. 1972. Immune ad
herence by the alternative complement pathway. Proc. Soc.
Exp. Biol. Med. 141:287.
40. Rother, K. 1972. Leukocyte mobilizing factor: A new bio
logical activity derived from the third component of com
plement. Eur. J. Immunol. 2^:550.
41. May, J. E. and M. M. Frank. 1972. Complement mediated tis
sue damage: Contribution of the classical and alternate
complement pathways in the Forssmann reaction. J. Immunol.
108:1517.
42. Meller-Eberhard, H. J. and I. H. Lepow. 1965. C"1 esterase
effect on activity and physicochemical properties of the
fourth component of complement. J. Exp. Med. 121:819.
43. Ruddy, S. and K. F. Austen. 1971. C3b inactivator of man.
II. Fragments of cell bound or fluid phase C3b. J.
Immunol. 107:742.
44. Neter, E. 1956. Bacterial hemagglutination and hemolysis.
Bacterio!. Rev. 20:166.
45. Westphal, 0. and J. Kann. 1965. Bacterial lipopoly-
saccharides. Extraction with phenol water and further
applications of the procedure. In Methods in Carbohy
drate Chemistry. R. L. Whistler, ed. Academic Press,
New York. Vol 5, p83.

66
46.Phillips, J. K., R. Snyderman, and S. E. Mergenhagen. 1972.
Activation of complement by endotoxin: A role for yZ glo
bulin, Cl, C4 and C2 in the consumption of terminal com
plement components by endotoxin-coated erythrocytes. J. of
Immunol. 109:334.
47.May, J. E., M. A. Kane, and M. M. Frank. 1972. Host defense
against bacterial endotoxemia Contribution of the early
and late components of complement to detoxification. J.
Immunol. 109:393.
48. Gewurz, H., H. S. Shin, and S. E. Mergenhagen. 1968. Inter
actions of the complement system with endotoxic lipopoly-
saccharide: Consumption of each of the six terminal com
plement components. J. Exp. Med. 128:1049.
49. May, J. E., M. A. Kane, and M. M. Frank. 1972. Immune
adherence by the alternate complement pathway. Proc. Soc.
Exp. Biol. Med. 141:287.
50. Hoffmann, Edward M. 1969. Inhibition of complement by a
substance isolated from human erythrocytes extraction
from human erythrocyte stromata. Immunochem. 6^:391.
51. Hoffmann, E. M., W. C. Cheng, E. J. Tomeu, and C. M. Renk.
1974. Resistance of sheep erythrocytes to immune lysis by
treatment of the cells with a human erythrocyte extract:
Studies on the site of inhibition. J. of Immunol. 113:1501.
52. Hoffmann, E. M. 1969. Inhibition of complement by a sub
stance isolated from human erythrocytes. II. Studies on
the site and mechanism of action. Immunochem. 6;405.
53. Hoffmann, E. M. and H. M. Etlinger. 1973. Extraction of
complement inhibitory factors from the erythrocytes of non
human species. J. of Irrmunol. Ill:946.
54. Rosse, W. F. and J. V. Dacie. 1966. Immune lysis of normal
and PNH red cells. I. Sensitivity of PNH cells. J.
Clin. Invest. 45:736.
55. Kabat, E. A. and M. M. Mayer. 1961. Complement and com
plement fixation. In Experimental Immunochemistry.
Charles C. Thomas, Springfield, Ill. pi49.
56. Springer, G. F., Y. Nagai, and H. Tegtmeyer. 1966.
Isolation and properties of human blood-group NN and
meconium-Vg antigens. Biochem. 5^:3254.

67
57. Maurer, H. R. 1971. Disc Electrophoresis and Related Techni
ques of Polyacrylamide Gel-Electrophoresis. 2nd revised ed.
Walter de Gruyter, Berlin, N.Y.
58. Boivin, A. I. 1933. Extraction of bacterial O-antigens (endo
toxins). Basic Exercises in Immunochemistry. A. Nowotny, ed.
Springer-Verlag, New York. 1969. p25.
59. Nelson, R. A., J. Jensen, I. Gigli, and N. Tamura. 1966.
Methods for the separation, purification, and measurement
of nine components of hemolytic complement in guinea pig
serum. Immunochem. _3:^-
60. Ruddy, S. and K. F. Austen. 1967. A stoichiometric assay for
the fourth component of complement in whole human serum using
EACSP and functionally pure human second component. J. Immunol.
99:1162.
61. Ruddy, S. and K. F. Austen. 1969. C3 inactivator of man.
I. Hemolytic measurement by the inactivation of cell bound
C3. J. Immunol. 102:533.
62.Boros, T. and H. J. Rapp. 1967. Immune hemolysis: A simpli
fied method for preparation of EAC4 with guinea pig or with
human complement. J. Immunol. 99:263.

BIOGRAPHY
I, Gloria Jean Jackson, born on October 4, 1939 to Charles and
Sarah Mcknight, was the second of four children. The only girl, I
attended and was graduated May 27, 1957, valedictorian of my class
at Boylan Haven, a private school for girls located at that time in
Jacksonville, Florida.
I later attended Bennett College, majoring in premedicine with a
minor in psychology, graduating June of 1961. Lack of funds prevented
entry into medical school, therefore, I was employed as a technician at
University of Florida for one year. I later entered the University of
Kansas at Lawrence, Kansas to pursue a master's degree in Microbiology
specializing in microbial physiology. After a year and a half I com
pleted the investigative requirements for the master's degree and was
employed as a research technician in the Department of Microbiology,
where I served as an assistant to a Microbial Geneticist. In October
1965 I was married to Virgil Lawrence Jackson, following which we re
located to Chicago, Illinois.
During my three years in Chicago, I worked as a research technician
at the American Medical Association Biomedical Research Institute and
later as assistant supervisor of the Microbiology Research Department
at the Metropolitan Sanitary District. On May 16, 1969 I gave birth
to a son, Jacques Duvall, following which we relocated to Parsippany,
New Jersey.
68

69
During my stay in New Jersey I was employed at the Warner-Lambert
Research Institute where I assisted in studies in oral microbiology
and virology. Later relocating to Westen, Connecticut and finally
Dover, Massachusetts where my husband was employed as vice president
of a printing company.
Following a legal separation, I returned with my son, Jacques, to
Gainesville to pursue a doctorate at the University of Florida, major
ing in Microbiology and specializing in Immunology. I am presently
employed at Abbott Labs of North Chicago as a product manager in re
search and development. In addition to immunology, I enjoy tennis,
chess, bridge, skiing, and sailing.

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
r; : -t
Edward M. Hoffman, Chairman/
Professor of / y
Microbiology and Cell Science
UUri
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Arnold S. Bleiweis
Professor of
Microbiology and Cell Science
I certify that I
conforms to acceptable
adequate, in scope and
Doctor of Philosophy.
have read this study and that in my opinion it
standards of scholarly presentation and is fully
quality, as a dissertation for the degree of
Lester W. Clem
Professor of
Immunology and Medical Microbiology

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
u/ X;
y //
Lonnie 0. Ingram
Associate Professor of
Microbiology and Cell Science
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Associate Professor of
Biochemistry and Molecular Biology
This dissertation was submitted to the Graduate Faculty of the College of Lib
eral Arts and Sciences and the Graduate Council and was accepted as partial
fulfillment of the requirements for the degree of Doctor of Philosophy.
December 1978
Dean, College of Liberal Arts and Sciences
Dean, Graduate School



Figure 5. A DEAE-Sephadex A-25 chromatography profile of the sepharose
4B active fractions from human stromata extracts prepared by
the method of Springer. 6.1 ml of input material, with an
optical density at 220 nm of 10.0 were applied to a 22 X 2.5
cm column. A linear sodium chloride gradient of 0.05-0.75 M
Tris buffer, pH 7.0 was applied and 4 ml fractions were col
lected. Optical densities of the fractions at 220 nm are
shown by the closed circles. The closed triangles indicate
the elution position of the fractions capable of inhibiting
lysis of sheep EAC42 and the shaded bars represent the
fractions having LPS-receptor activity.


100
60
30
10
log2 reciprocal of the dilution
CO
PO


7
systems including the complement system (18). Whatever the mechanism,
mounting evidence seems to implicate complement as a mediator of a
number of the deleterious responses to endotoxins in experimental
animals and man (8). More importantly, recent evidence indicates that
the ability of LPS to initiate a complement response is not confined
to the lipid A moiety but appears to involve the polysaccharide core
as well (19).
It has become increasingly clear that a major role of the com
plement system during an immune response is the mediation of enhanced
phagocytosis with a concomitant induction of an inflammatory response.
This is accomplished by the sequential activation of the proteins which
make up the complement system. The activation process can be divided
into three major stages: a recognition stage, the generation of C3
cleavage enzymes and C3 activation stage, and a terminal or membrane
attack stage (20-22). Under normal conditions, complement proteins
exist in the serum as inactive precursors, and are activated by either
of two pathways -- the classical or alternative. The component of
these pathways and their reaction requirements are summarized in Table
1. The two share a number of similar characteristics but differ, basic
ally, in the reactants and the sequence of reactions of the first two
stages.
The capacity of endotoxins to activate the complement system by
a mechanism which requires neither antibodies to LPS nor the partici
pation of the early complement components was demonstrated more than
twenty years ago (23). More recently, it was recognized that fluid
phase activation of complement by LPS is not restricted to the


5
or partially purified LPS preparations in the fluid phase. This was
in spite of the fact that it had been shown that heated LPS (100C
for 60 minutes) could be coated onto the surface of a number of cells
including erythrocytes (8). The latter are often the model target
cells for hemolytic assays. It has become increasingly clear that the
biological consequence of either an in vivo or in vitro encounter with
LPS, whether cell bound or partially purified, is dependent upon its
fixation to various target cells (9). For example, it was reported
that patients suffering from endotoxin shock and sepsis due to gram
negative bacteria had greatly reduced levels of blood platelets and
that the platelets contained LPS. It has further been established
that human platelets possess an endotoxin binding receptor, which when
triggered by interaction with LPS, results in the release of a vaso
active amine, 5-hydroxytryptamine (5-HT) and the unmasking of clot-
promoting activity (10). In animal studies, it was observed that
guinea pigs injected intravenously with LPS also show a 95% decrease
in blood platelets with a concomitant shortening of the clotting time
(11). Additionally, it has been reported that there is a direct
relationship between the susceptibility of different strains of mice
to the lethal effects of Salmonel1 a endotoxin and the affinity of their
red cells for either heat-killed Salmonellae or free LPS (12).
Little was known about the nature of the attachment of LPS to
any cell until the late sixties, when G. Springer isolated an extract
from the membranes of human erythrocytes having a high affinity and
specificity for the 1ipopolysaccharides of a variety of gram negative


53
%
COMPLEMENT

CONSUMED
UNTREATED
SERUM
%
COMPLEMENT
CONSUMED
BUFFER EGTA EDTA


13
generating damage to normal cell membranes by inducing the selective
release of a variety of vasoactive amines and cytolytic substances
(47,48). Alternative pathway activation has not been shown to mediate
the release of membrane damaging factors (49). In vivo systems are
difficult to evaluate because of the large number of parameters which
must be considered. Therefore, much study is needed before a complete
understanding of the mechanism involved can be obtained.
Regardless of the nature of the system (in vitro or in vivo) or
the state of the endotoxin (free or cell bound), the interaction of
LPS with the complement system in the presence of specific antibody is
extremely efficient, giving full response with sparing consumption of
antibody and C1-C4 (48). Virulence of gram negative bacteria would
then appear to be related to the fate and site of free endotoxins re
leased as a direct consequence of cell death due to phagocytosis and
interaction of this free endotoxins with other cell systems in the
presence of serum factors.
As previously stated, erythrocytes and now other cells have been
shown to have membrane receptors which bind endotoxins. The biological
role of these receptors is still not clear; however, fixation of LPS
to the cell does appear to be a precondition for the triggering of
many incitements of the immune system by LPS (49).
Hoffmann, in the late 1960s, isolated and described extracts from
the membranes of human erythrocytes capable of inhibiting the hemolytic
activity of complement when sensitized sheep erythrocytes were used
as target cells (50). Extracts isolated at two different ionicities


12
amplification represents an early stage in the development of an inflam
matory response.
Severe tissue damage resulting from unchecked amplification of
the activation of the complement system is probably minimized by the
presence of two known serum inhibitors, Cl esterase inhibitor (42)
and conglutinin activating factor (KAF) also known as C3b inactivator
(43). Perhaps there are other humoral or cellular inhibitory factors
still to be defined.
Many of the studies cited here employed fluid phase LPS for in
vitro assays. Neter was the first to demonstrate that LPS coated onto
the surface of sheep erythrocytes were sensitive to lysis by an anti
body dependent classical complement pathway mechanism (44). Phillips
and Mergenhagen employing sheep erythrocytes treated with LPS extracted
by the Westphal-phenol procedure (45) confirmed Neter's observation
showing the need for the presence of a naturally occurring y2 globulin
for complement activation with a sparing consumption of the early com
ponents C1-C4 (46).
The in vivo consequences of the intravenous administration of
endotoxins to normal and complement deficient animals have been studied.
Both complement pathways are activated with the classical pathway being
required for the development of many of the pathophysiological symptoms
such as the thrombocytopenia observed in many experimental endotoxin
infections (11). Some investigators have suggested that classical
pathway activation due to an antibody mediated fixation of LPS to
various cells such as platelets and erythrocytes is more effective in


67
57. Maurer, H. R. 1971. Disc Electrophoresis and Related Techni
ques of Polyacrylamide Gel-Electrophoresis. 2nd revised ed.
Walter de Gruyter, Berlin, N.Y.
58. Boivin, A. I. 1933. Extraction of bacterial O-antigens (endo
toxins). Basic Exercises in Immunochemistry. A. Nowotny, ed.
Springer-Verlag, New York. 1969. p25.
59. Nelson, R. A., J. Jensen, I. Gigli, and N. Tamura. 1966.
Methods for the separation, purification, and measurement
of nine components of hemolytic complement in guinea pig
serum. Immunochem. _3:^-
60. Ruddy, S. and K. F. Austen. 1967. A stoichiometric assay for
the fourth component of complement in whole human serum using
EACSP and functionally pure human second component. J. Immunol.
99:1162.
61. Ruddy, S. and K. F. Austen. 1969. C3 inactivator of man.
I. Hemolytic measurement by the inactivation of cell bound
C3. J. Immunol. 102:533.
62.Boros, T. and H. J. Rapp. 1967. Immune hemolysis: A simpli
fied method for preparation of EAC4 with guinea pig or with
human complement. J. Immunol. 99:263.


RESULTS
Erythrocyte coating by LPS and its inhibition. Repeated titra
tions of Salmonella typhimurium LPS at concentrations ranging from
0.195 ug to 50.0 ug/ml as determined with polyvalent and homologous
Salmonella antiserum, employing erythrocytes at 2.0 X 10 cells/ml
were carried out. The results of a representative experiment em
ploying polyvalent antiserum are given in Table II. It can be seen
that heating the LPS enhanced the erythrocyte coating capacity to a
remarkable extent. Additionally, a maximum titer resulted when erythro
cytes were exposed to at least 0.78 ug/ml of heated LPS. Therefore,
an optimal coating unit (u) of heated LPS (defined as the reciprocal
of the greatest dilution of LPS producing complete hemagglutination by
either polyvalent or homologous antiserum) was taken as 0.78 ug/ml.
Assays employing homologous antiserum to heated LPS yielded lower
optimal coating doses of 0.78 ug/ml and 1.56 ug/ml, depending on the
age of the antiserum. These results were identical for human and sheep
erythrocytes. LPS extracted by the Westphal procedure resulted in an
optimal coating dose of 0.39 ug/ml as also determined with homologous
antiserum.
LPS-receptor activity, as evaluated in these studies, was based on
the ability of a given erythrocyte preparation to inhibit the complete
fixation of an optimal coating unit of LPS onto either sheep or human
27


WASHED PACKED RED BLOOD CELLS
(Springer)
(Hoffmann)
Water hemolysis, pH 5.3
+ 0.2% phenol
Water hemolysis, pH 6.0
Erythrocyte stroma
Erythrocyte stroma
aqueous suspension (1:2),
homogenization waring blender;
butanol:water (1:2) extraction,
16 hr, 4C, pH 8.0 (3X).
Activity in aqueous phase
Phosphate buffer suspension
(1:2), pH 7.5; butanol:water
(1:5) extraction, aqueous phase
adjusted to y of 0.15, repeat
extraction 6X.
Crude LPS-Receptor
Centrifugation 33,000-151,000g
>90% activity in aqueous top
layer, sepharose 4B column;
0.05 M Tris, pH 7.0.
Purified LPS-Receptor
DEAE-Sephadex A-25 column elution
0.05 M Tris, pH 7.0, with con
tinuous gradient, 0.05-0.5 M
NaCl; Sucrose density gradient,
5%-25%; Sephadex G200 column,
0.05 M Tris, pH 7.0, receptor
in void volume.
Crude IH
DEAE column, elution 0.1 M phos
phate buffer, pH 7.0, with con
tinuous gradient, 0.1-1.0 M NaCl;
Sephadex G-200 column, P0^ buffer,
pH 7.5, IH in void volume.
Purified IH
Highly Purified LPS-Receptor


Figure 3. Inhibition of the hemolysis of EAC142 by C-EDTA in
the presence of various concentrations of three crude
phases resulting from the butanol extraction of human
erythrocyte stromata prepared according to the pro
cedure of Springer. The open circles show the activ
ity of the crude untreated butanol extracted stromata;
the closed circles show the inhibition associated
with the top fraction obtained after the high speed
centrifugation (40,000 rpm/2 hours) of the crude
extracted stromata; the closed squares show IH in
hibitor activity of interphase obtained after the
high speed centrifugation of the extracted stromata.


REFERENCES
1. Buston, A. 1959. The in-vivo sensitization of avian erythrocytes
with Salmonella gall inarum polysaccharide. Immunology. 2;203.
2. Leive, L., V. K. Shovlin and W. E. Mergenhagen. 1968. Physical,
chemical and immunological properties of LPS released from
Escherichia coli by ethylene diaminetetra acetate. J. Biol.
Chem. 243:6384.
3. Voll, M. J. and L. Leive. 1970. Release of LPS in Escherichia
coli resistant to be the permeability increase induced by
ethylenediamine tetraacetate. J. Biol. Chem. 245:1108.
4. Rapin, A. M. C. and H. M. Kalckar. 1971. Microbial Toxins.
Weinbaum, G., S. Kadis, and S. J. Ajl, eds. Academic Press,
New York. Vol 4, pp 267-303.
5. Morrison, D. C. and L. Leive. 1975. Fractions of lipopoly-
saccharide from Escherichia col i Qlll: prepared by two ex
traction procedures. J. of Biol. Chem. 250:2911.
6. Osborn, M. J. 1969. Structure and biosynthesis of the bac
terial cell wall. Ann. Rev. Biochem. 38:501.
7. Luderitz, 0., V. L. Galanos, M. Nurminen, E. T. Rietschel, G.
Rosenfelder, M. Simon, and 0. Westphal. 1973. 'Lipid A.1
Chemical structure and biological activity. J. Infect. Dis.
128:817.
8. Morrison, D. C. and L. F. Kline. 1977. Activation of the
classical and properdin pathways of complement by bacterial
1ipopolysaccharides (LPS). J. Immunol. 118:362.
9. Springer, George F. and J. C. Adye. 1975. Endotoxin-binding
substances from human leukocytes and platelets. Infect, and
Immun. 12:978.
10. Hawiger, J., A. Hawrger, and S. Timmons. 1975. Endotoxin
sensitive membrane components of human platelets. Nature.
256:125.
11. Kane, M. A., J. E. May, and M. M. Frank. 1973. Interaction
62


16
related. Although it has been observed that human erythrocytes are
highly resistant to LPS mediated immune hemolysis, the reason for the
refractoriness of these treated erythrocytes has not been defined.
Sheep erythrocytes, as previously stated, are normally sensitive to
immune lysis but may be rendered resistant by treatment with diluted
extracts of the partially purified IH inhibitor. Therefore, the second
objective of this investigation was to explore the biological conse
quences of the IH inhibitor in the interaction of LPS treated erythro
cytes and serum complement, in an attempt to clearly establish a bio
logical role for the IH inhibitor.


10 20 40 80 160 320
RECIPROCAL
100 -
5 50
O
JB
MB
JE
C
I
i
I
¡
I
ii
1
10 20 40 80 160 320
THE DILUTION


demonstrated. Finally, evidence was presented indicating that the
membranes of sheep erythrocytes consist of molecules with LPS-receptor
specificity but were devoid of IH inhibitor activity.
Additional studies indicated that LPS bound to cell membranes can
activate either the alternative or classical complement pathways and
that IH inhibitor associated with cell membranes can block LPS-induced
complement lysis of red cells despite the fact that complement activa
tion has occurred.
vi 11


TABLE IV
Recovery of LPS-receptor and IH Inhibitor Activity from Extracts
of Human Erythrocyte Stromata Following Sepharose 4B
and DEAE-Sephadex Purification
Preparation Volume
Concentration Activity/ml Total
Total
Specific0* Yielde
Purification*"
Conc.c
Activity
Activity
%
(ml)
(A220/ml) LPSRa IH5Qb (A220/ml)
LPSR IH50
LPSR IH50
LPSR IH go
LPSR IH50
Sepharose 4B
input
(Crude high
speed top
layer cone) 4.5
479.2
320
320
2,156.4
1440
1440
0.67
0.67


Sepharose
4B recovery 47.0
18.5
16
20
871
752
940
0.862
1.0
52.2
65.3
DEAE-Sepha
dex
recovery 6.1
10
128
80
61.1
780
488
12.80
7.9
54.2
31.1
aThe smallest amount (dilution) giving maximum inhibition of LPS coating.
bThe reciprocal of the highest dilution giving 50% inhibition of EAC142 lysis.
c
Concentration.
Calculated oy dividing the value for total activity by the values for total concentration.
0
Calculated oy dividing the value for total activity after treatment, by the value tor the total activity before
treatment, multiplied by luO.
^Calculated by dividing the value for each of the specific activities by the value for the initial specific activity.


Figure 4. A gel filtration profile of the butanol extracted human erythro
cyte stromata prepared by the method of Springer. 4.5 ml of the
extracted material were applied to a sepharose 4B column (92 X
1.5 cm) and chromatographed at 4C with 0.05 M Tris buffer, pH
7.0. The optical density at 220 nm of the input material was
479 and 4.0 ml fractions were collected. Optical densities of
the fractions at 220 are shown by the closed circles, the closed
triangles indicate the elution pattern of the IH inhibitor ac
tivity, and the shaded bars represent the fractions having LPS-
receptor activity.


8
TABLE I
Comparative Properties of the Classical
and Alternative Complement Pathways
Classical Alternative
Activating agents
Immunoglobulins of human
rabbit
guinea pig
ruminant
Miscellaneous
Activating site
Factors required to
generate C3 convertase
Total serum requirement
Divalent cation requirements
IgGl,IgG2,
IgA and IgE
IgG3; IgM
IgG
IgM
I gG2
IgGl
IgGl
IgG2
(Lipid A)
Inulin, Zymosan,
Endotoxin or LPS,
CVF
Fc fragment
F(ab)"2 (rabbit) i
or
Fc fragment (humai
n)
Cla
Properdin, factor
D
C4
C3b
C2
factor B
Dilute
Concentrated
Ca and Mg
Mg
aCl is a trimolecular complex of Clg, Clr and Cls. Classical pathway
activation is initiated when Clq binds to immunoglobulin.


10
In stage two of the classical pathway activation scheme, C3 con-
vertase is generated when Cl esterase cleaves C4 and C2 in the presence
of magnesium yielding the cell bound "C4b2a" enzyme. This convertase
cleaves the third component of complement (C3) into two fragments,
C3a and C3b. The latter fragment is unstable and has a highly reactive
hydrophobic binding site. As a result of binding to target membranes
at sites adjacent to the membrane bound C3 convertase, a new enzyme C5
convertase, whose major substrate becomes the fifth component of com
plement, is generated. In addition, C3b undergoes secondary changes
giving rise to an immune-adherence site capable of binding to a variety
of effector cells of the immune system which bear specific C3b receptors
(32). Additionally, C3b, either bound hydrophobically to a cell surface
or free in solution, can produce further splitting of C3 via the C3
feedback cycle of the alternative pathway described below. It is this
amplification of the generation of C3b and its deposition via immune
adherence onto the surface of specific target cells (and in some cases
innocent bystanders), where the major function of the complement system
is realized.
The mechanism of the alternative pathway generation of the C3
convertase is a bit more complicated in that there is a direct require
ment for preformed C3b, the source of which is still not fully under
stood. This C3b in combination with factor B plus properdin and factor
D forms the alternative pathway convertase "Factor BbC3b" capable of
splitting C3 into C3a and C3b (33). This splits more native C3 giving
rise to further C3b and the cycle is again repeated. This reaction


61
to partially inhibit immune lysis. This may be due either to the masking
of LPS-receptors resulting in less LPS uptake (demonstrated not to be
the case here), or the prevention of C3b fixation to the cell membrane
thus blocking the initiation of the membrane attack system of serum
complement. An analogous site on human erythrocytes is already occupied
by the IH inhibitor rendering this cell naturally immune to LPS mediated
lysis.
The findings presented here have led us to hypothesize that a
necessary criterion for resistance to LPS induced complement mediated
lysis would be the localization of LPS-receptor and IH inhibitor mole
cules on the same membrane. This would imply that any red cell devoid
of IH inhibitor molecules would be far more susceptible to the cytolytic
action of LPS activated complement.
LPS of gram negative bacteria are potent activators of the com
plement system. It is of great clinical interest, therefore, that sub
stances on the surface of erythrocytes which bind LPS are found closely
associated with materials capable of inhibiting complement mediated lysis.
The symptoms of several infectious diseases, such as typhoid fever, have
been observed to include very intensive erythrophagocytic activity by
macrophages of lymph nodes. The consequence of this observation could
have great clinical importance. Erythrocytes coated with LPS, in contact
with serum complement, would naturally lead to activation of the com
plement system followed by increased phagocytosis with minimal cytolysis.
This would lead naturally to an amplification of the activation of the
complement cascade resulting in either clearance or a heightened inflam
matory response.


40
30
20
10
bottom
DISTANCE (mm) FROM TOP OF GEL
% Inhibition


Figure 9. Consumption of complement in normal and E-LPS treated
guinea pig serum by fluid phase Salmonella typhimurium
LPS extracted by the Boivin and Westphal procedures,
EDTA, EGTA, and saline treated sera were treated with
the LPS preparations one hour at 37C, then sedimented,
after which the CH50 titers were determined in the mag
nesium and calcium reconstituted sera. The top figure
represents the profile of the Boivin treated serum
and the lower figure, the Wesphal treated serum.


9
alternative pathway. Some preparations of LPS have been demonstrated
to activate the classical pathway in either the presence or absence
of specific antibodies- (8,24-26,27).
Normally, during classical pathway activation, the recognition
stage is initiated by the fixation of the first component of the com
plement system, Cl, to IgG or IgM immunoglobulins of specific immune
complexes. This fixation and activation of Cl generates a Cl esterase
whose major substrates are the fourth and second components of the
complement system.
Endotoxins in the fluid phase have been shown to initiate the
activation of the complement system in vitro as described above in
the presence of predominantly IgM antibodies directed against the
O-polysaccharide antigen (28). Unique to the endotoxins of certain
gram negatives, however, is the capacity of the lipid A moiety of
the molecule to fix Cl directly resulting in a antibody independent
initiation of the classical complement cascade (8,29).
Alternative pathway activation, on the other hand, is triggered
by direct activation of an initiation factor (IF) by endotoxins, or
any one of a number of naturally occurring polysaccharides, and aggre
gates of IgA or IgE (30). It has been established that alternative
pathway activation of the complement system involves the core poly
saccharide region of the LPS molecule (31). Additionally,it has been
shown that the length of the O-specific polysaccharide antigen and
differences in the carbohydrate content may also play a role in this
mechanism of activation (8).


Abstract of Dissertation Presented to the
Graduate Council of the University of Florida
in Partial Fulfillment of the Requirements for the
Degree of Doctor of Philosophy
THE RELATIONSHIP BETWEEN
HUMAN ERYTHROCYTE LIPOPOLYSACCHARIDE RECEPTOR
AND AN INHIBITOR OF COMPLEMENT MEDIATED LYSIS
By
Gloria Jean Jackson
December, 1978
Chairman: Edward M. Hoffmann
Major Department: Microbiology and Cell Science
Human erythrocyte membrane extracts with receptor specificity
for the 1ipopolysaccharides of gram negative bacteria were found to
be rich in IH, a complement inhibitory substance originally isolated
and described from human red cell membranes. The possibility that
the LPS-receptor and IH inhibitor might be the same macromolecule was
considered.
In this investigation it was demonstrated that although closely
associated, the two activities are biologically distinct and separable.
This conclusion was supported by the results of five experimental ap
proaches. Small, but distinct differences were observed in the elution
profiles of the two activities when crude extracts of erythrocyte mem
branes were subjected to sepharose 4B, DEAE-Sephadex chromatography
and polyacrylamide disc gel electrophoresis. Complete separation of
the two activities was accomplished by a shift in the pH of the membrane
extraction conditions. Further, differences between the two activities
based on their ability to spontaneously bind to sheep erythrocytes were
vii


46
with IH inhibitor activity distributed in both the aqueous and
lipid interphases. A comparison of the purification tables (IV and VI)
of crude extracts obtained at pH 8.2 and pH 5.3 indicates that, as was
observed at pH 8.2, extracts obtained at pH 5.3 and subjected to DEAE-
Sephadex chromatography resulted in a substantially greater loss of
total mass as estimated by the adsorbance at 220 nm and the yield of IH
inhibitor activity.
Treatment of Springer's crude butanol extracts with sheep erythro
cytes To further establish that the two activities are distinctly dif
ferent, equal volumes of sheep erythrocytes at either 10^ or 109 cell s/ml
were mixed with equal volumes of a butanol high speed top layer ex
tract prepared according to the procedure of Springer. This extract
had an initial LPS-receptor titer of 128 and a IH50 inhibitor titer
greater than 80. Control tubes consisting of equal volumes of buffer
and the extracts were also prepared. All tubes were mixed at 30C for
30 minutes with shaking. The cells were pelleted by centrifugation
and the supernatant fluids along with the buffer controls were diluted
and assayed for LPS-receptor and IH inhibitor activities. As can be
seen in Figure 8, the IH inhibitor activity was reduced substantially
when extracts were treated with 109 cells/ml. In contrast, the LPS-
receptor activity remained constant when the cell treated and buffer
control supernatants fluids were compared.
The biological consequence of the IH inhibitor and LPS-receptor
on the erythrocyte membrane. The data in the previous section indicated
two essential points. First, the membrane of the human erythrocyte,


20
after every second water change. Following the final wash the stromata
was collected by centrifugation, weighed and stored at -22C until use.
Isolation and purification of the LPS-receptor. The LPS-receptor
was prepared as outlined in Figure 2 using a modification of the pro
cedure of Springer et a 1. (15). A 50% aqueous stromata suspension was
homogenized in a Waring blender and extracted overnight with two volumes
of n-butyl alcohol at 4C for 16 hours at pH 8.2. Four phases, resolved
after centrifugation at 2000xg for 30 minutes, organic, lipid, aqueous
and solid. The aqueous phase was re-extracted twice with n-butanol,
once for 30 minutes and again overnight and then thoroughly dialyzed
against several changes of 0.05 M Tris-HCl buffer (pH 7.0). An aqueous
butanol extract low in LPS-receptor activity but high in IH inhibitor
activity was obtained by shifting the pH of the butanol extraction from
8.2 to 5.3.
The dialyzed active crude butanol extract was centrifuged at 151,000
xg for 1-2 hours in a Spinco Model L2 preparative ultracentrifuge.
Three phases resulted from the high speed centrifugation. Contrary
to Springer's findings, the top aqueous layer possessed the highest LPS-
receptor. After extensive dialysis, the aqueous top layer was applied
to a 92 X 1.5 cm sepharose 4B column (Pharmacia Fine Chemicals, Piscataway,
N.J.). The sepharose columns were washed with a 0.05 M Tris-HCl buffer at
pH 7.0. Three milliliters fractions were collected and assayed for both
LPS-receptor and IH inhibitor activities. The active sepharose fractions
were pooled, concentrated tenfold by dialysis against 20% polyethylene-
glycol in 0.05 M Tris-HCl buffer (pH 7.0) and applied to a 22.0 X 2.5 cm


65
34. Marcus, R. L., H. S. Shin, and M. M. Mayer. 1971. An alter
nate complement pathway: C3 cleaving activity not due to
C4, 2a, on endotoxic 1ipopolysaccharide after treatment
with guinea pig serum; relation to properdin (complement
components). Proc. Natl. Acad. Sci. USA. 68:1351.
35. Fine, D. P., S. R. Mamey, D. G. Colley, J. S. Sergent, and
R. M. Des Prez. 1972. C3 shunt activation in human serum
chelated with EGTA. J. Immunol. 109:807.
36. Mayer, M. M. 1972. Mechanism of cytolysis by complement.
Proc. Nat. Acad. Sci. USA. 69:2954.
37. Jensen, J. 1967. Anaphylatoxin and its relation to the
complement system. Science. 155:1122.
38. Shin, H., R. Snyderman, E. Friedman, A. Mellors and M. Mayer.
1968. Chemotactic and anaphylatoxic fragment cleaved from
the fifth component of guinea pig complement. Science.
162:361.
39. May, J. E., M. A. Kane, and M. M. Frank. 1972. Immune ad
herence by the alternative complement pathway. Proc. Soc.
Exp. Biol. Med. 141:287.
40. Rother, K. 1972. Leukocyte mobilizing factor: A new bio
logical activity derived from the third component of com
plement. Eur. J. Immunol. 2^:550.
41. May, J. E. and M. M. Frank. 1972. Complement mediated tis
sue damage: Contribution of the classical and alternate
complement pathways in the Forssmann reaction. J. Immunol.
108:1517.
42. Meller-Eberhard, H. J. and I. H. Lepow. 1965. C"1 esterase
effect on activity and physicochemical properties of the
fourth component of complement. J. Exp. Med. 121:819.
43. Ruddy, S. and K. F. Austen. 1971. C3b inactivator of man.
II. Fragments of cell bound or fluid phase C3b. J.
Immunol. 107:742.
44. Neter, E. 1956. Bacterial hemagglutination and hemolysis.
Bacterio!. Rev. 20:166.
45. Westphal, 0. and J. Kann. 1965. Bacterial lipopoly-
saccharides. Extraction with phenol water and further
applications of the procedure. In Methods in Carbohy
drate Chemistry. R. L. Whistler, ed. Academic Press,
New York. Vol 5, p83.


51
TABLE VII
Consumption of Total Complement
or E-LPS Absorbed
by Untreated and LPS Coated
in Either E Absorbed
GP Serum
Sheep Erythrocytes
Cell Suspension
Percent
Consumed
E absorbed
E-LPS absorbed
serum
serum
E
^LPS Westphal
21.3
16.89
ULPS Boivin
47.9
6.38


59
on sheep cells differ from those observed on human cells, since they
are confined to the lipid moiety.
Good purification of the LPS-receptor and IH inhibitor activities
following DEAE-Sephadex chromatography is indicated by the quantitative
data tables. These results would suggest, however, that as a pre
parative purification step it should be modified to encourage higher
yields of the two activities.
The observation that sheep erythrocyte membranes possess mole
cules with receptor specificity for LPS is not surprising for it has
long been known that sheep cells could be modified by the presence
of LPS of gram negative bacteria and that these modified cells are
readily lysed in the presence of homologous antiserum to LPS and com
plement (44). In contrast, hemolysis was not observed when LPS treated
human erythrocytes were treated under the same conditions. This raises
the possibility that human erythrocytes, a natural source of the IH
inhibitor even when treated with LPS, are extremely resistant to LPS
mediated lysis because of the presence of inhibitor molecules.
The findings reported here generally agree with those of Phillips
et al. (46) indicating that LPS treated sheep erythrocytes can activate
the complement system in the presence of natural antibodies to the LPS.
In contrast to their results, however, erythrocytes coated with a pre
paration of LPS extracted by the procedure of Westphal were shown to be
capable of activating the complement system in the absence of natural
antibodies to LPS. Additionally, fluid phase LPS extracted by the Boivin
method (LPS-boivin) was shown to be a more effective activator of the



PAGE 1

THE RELATIONSHIP BETWEEN HUMAN ERYTHROCYTE LIPOPOLYSACCHARIDE RECEPTOR AND AN INHIBITOR OF COMPLEMENT MEDIATED LYSIS By GLORIA JEAN JACKSON A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1978

PAGE 2

ACKNOWLEDGMENTS I wish to express sincere appreciation for the technical criticisms and patient guidance given me by Dr. Edward M. Hoffmann. I also would like to extend thanks to Dr. L. 0. Ingram, Dr. A. S. Bleiweis, Dr. Paul W. Chun, Dr. Lester W. Clem, Dr. Paul Smith and Mr. Jim Milam for supplying constant help and encouragement. To all of the members of my laboratory and departmental family over the last four and a half years, I offer my sincere gratitude for their constant encouragement, patience, caring and all of the happy times we've shared. I wish also to add my thanks to Mrs. Sandy Cannella for her encouragement and her excellent and dedicated assistance in the typing of this manuscript. Finally, but most important of all, for their continual prayers, help, guidance, unfailing love and for always being there through the good and hard times, I thank my parents Mr. and Mrs. Charles and Sarah McKnight and my loving son Jacques. Without them this would never have been possible. 1i

PAGE 3

TABLE OF CONTENTS PAGE ACKNOWLEDGEMENTS LIST OF TABLES LIST OF FIGURES v GLOSSARY OF ABBREVIATIONS vi ABSTRACT vii INTRODUCTION 1 MATERIALS AND METHODS 17 RESULTS 27 DISCUSSION 57 REFERENCES 62 BIOGRAPHY 68 1 i i

PAGE 4

LIST OF TABLES TABLE p^e^ 1. Comparative Properties of the Classical and Alternative Complement Pathways 8 2. Determination of Optimal LPS Concentration Used for Coating Human Erythrocytes 28 3. A comparison of LPS-receptor and IH Inhibitor Activities of Several Crude Extracts of Erythrocyte Stromata 3q 4. Percent Recovery of LPS-receptor and IH Inhibitor Activity from Extracts of Human Erythrocyte Stromata Following Sepharose 4B and DEAE-Sephadex Purification 34 5. Recovery of LPS-receptor and IH Inhibitor Activity of Human Erythrocyte Stromata Extracted at pH 5.3 Using the Procedure of Springer 43 6. Recoveries of IH Inhibitor and LPS-receptor Activities From Erythrocyte Stromata Extracted at pH 5.3 by the Procedure of Springer and Subjected to DEAE-Sephadex Chromatography 49 7. Consumption of Total Complement in Either E Absorbed or E-LPS absorbed GP Serum by Untreated and LPS Coated sheep Erythrocytes 51 iv

PAGE 5

LIST OF FIGURES FIGURE PAGE 1. Probable structure of LPS of Escherichia coli 4 2. Schematic representation of purification procedures of the LPS-receptor and IH inhibitor 19 3. IH inhibitor activities of crude butanol extracted erythrocyte membranes 32 4. A gel filtration profile of the butanol extracted human erythrocyte stromata 36 5. DEAE-Sephadex chromatography of the sepharose 4B active fractions from human stromata.. 38 6. Pol yacryl amide disc gel electrophoresis of human stromata extracts 41 7. DEAE-Sephadex chromatography of human erythrocyte stromata extracts prepared at pH 5.3 45 8. Treatment of partial purified extracts of human erythrocyte stromata with sheep erythrocytes 48 9. Complement consumption in LPS treated guinea pig serum 53 10. Complement consumption of LPS and IH inhibitor treated erythrocytes 56 V

PAGE 6

GLOSSARY OF ABBREVIATIONS C complement CI, C2, C3, C4, C5, C6, C7, C8, C9^ complement components E erythrocyte A antibody HA hemagglutination HAI hemagglutination inhibition EDTA ethyl enediamine tetracetate The nomenclature of complement used conforms to that proposed as a result of a series of World Health Organizations ( Immunochemistry, 7^:137, 1970). .< — vi

PAGE 7

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE RELATIONSHIP BETWEEN HUMAN ERYTHROCYTE LIPOPOLYSACCHARIDE RECEPTOR AND AN INHIBITOR OF COMPLEMENT MEDIATED LYSIS By Gloria Jean Jackson December, 1978 Chairman: Edward M. Hoffmann Major Department: Microbiology and Cell Science Human erythrocyte membrane extracts with receptor specificity for the 1 ipopolysaccharides of gram negative bacteria were found to be rich in IH, a complement inhibitory substance originally isolated and described from human red cell membranes. The possibility that the LPS-receptor and IH inhibitor might be the same macromolecule was considered. In this investigation it was demonstrated that although closely associated, the two activities are biologically distinct and separable. This conclusion was supported by the results of five experimental approaches. Small, but distinct differences were observed in the elution profiles of the two activities when crude extracts of erythrocyte membranes were subjected to sepharose 4B, DEAE-Sephadex chromatography and polyacryl amide disc gel electrophoresis. Complete separation of the two activities was accomplished by a shift in the pH of the membrane extraction conditions. Further, differences between the two activities based on their ability to spontaneously bind to sheep erythrocytes were vi i

PAGE 8

demonstrated. Finally, evidence was presented indicating that the membranes of sheep erythrocytes consist of molecules with LPS-receptor specificity but were devoid of IH inhibitor activity. Additional studies indicated that LPS bound to cell membranes can activate either the alternative or classical complement pathways and that IH inhibitor associated with cell membranes can block LPS-induced complement lysis of red cells despite the fact that complement activation has occurred. vi i i

PAGE 9

INTRODUCTION The endotoxins or 1 ipopolysaccharides (LPS) of gram negative bacteria are among numerous antigens known to be capable of fixing to the membranes of erythrocytes and other mammalian cells in vitro and under certain conditions, in vivo (1). Because of their unique ability to modulate the immune response in a wide variety of ways, they have emerged as a complex and fascinating class of macromolecules. Functionally, endotoxins have been shown to have many different properties. Due to their chemical makeup and localization in the outer membrane of the bacterial cell envelope they have been shown to play a major role in the establishment of a selective permeability barrier (2,3) and in serving as receptors for certain bacteriophages (4). Of interest to the immunochemist however, is the fact that interaction of LPS with components of the immune system may lead to a single or combination of physiological responses. These include toxicity, mitogenicity, immunogenicity tolerance and activation of complement. Although much is known concerning the general nature of these responses the mechanisms cf the cell associated events responsible for their devel opment in the presence of LPS are still not fully understood. A great deal of information about the chemical structure of LPS, from a variety of organisms, has accumulated. Although it has been recently recognized that LPS isolated from a given organism is .1

PAGE 10

2 heterogenous (5), LPS of most gram negative wildtype organisms appear to share the same basic molecular composition. As illustrated in Figure 1, all consist of three regions. The first region, the 0specific polysaccharide antigen which is made up of repeating units of five to eight monosaccharides, carries the main serologic specificity for a given organism. Numerous serological groups differing in 0antigen specificity are now recognized and the polysaccharides accordingly show wide interand intrageneric variations in composition (6). Of interest is the fact that natural antibodies to this region are found in most animal species but do not always appear to be protective, and in some cases a lethal gram negative bacteremia develops despite high titers of 0-specific antibody (7). Region two consists of a short outer core which contains glucose (glc), galactose (gal), and N-acetyl glucosamine (Glc NAC), and an inner core of L-glycero-Dmannoheptose, phosphate, ethanolamine and three molecules of 3-deoxyoctonate (KDO). Unique to the endotoxins of gram negatives, 3-deoxyoctonate provides a linkage site to the third region, the lipid A moiety. Lipid A is basically composed of a phosphorylated glucosamine backbone to which are attached fatty acids and ethanolamine residues. The nature and distribution of the lipid A fatty acids varies among bacterial groups with the inner core polysaccharide composition remaining constant (7). The complete LPS containing 0-specific antigen is designated smooth (S), and all mutants lacking 0-specific side chains are referred to as rough or R-forms. Many early studies concerned with the interaction of endotoxins and biological systems were carried out using either whole bacteria

PAGE 11

4 1 r— U CD 1 O o o o o in 1 O Od) CO -K 1 '*"" CO o CY Odl %n: — 00 O ai 1 — u en CO 'a ta cs CM 1 U CM 'a < 1 O z I— 1 — u —CD CD <=1I O I CJ (/I a •r— O >> +-> c: re -t-> QJ rO CD CM I r — o CM u
PAGE 12

I -f-> o O (O cC sz: Q. CT) O C 1— to -ICD OJ T3 CTli— D-T3 Ol O LU 3 CO >> O M l/l •!Ol I— ^ O 3 o o I o CO +-> r 0 -a +J Ol 01 ^ to 1 — 1 — o o o x: E o CL O -D T•• >) Q. CO C o QJ CO O +-I (J (T3 •!I— C\J I— fC CD >, O I— • Q r— i.^: E to Ol CD -o — c to fO > o +-> c Ol o So CJ Q. (U -(-) ro iJT to O Qto (D QJ CO QJ QJ O to X2 sQJ IfI ea; 3 S^ •)-> ^ c: E O +J S -I• a o -t-l 1 — CO CO J3 to CO QJ X5 to r— C -Q tCH to Q. J3 QJ (/) O u 3 O o QJ CO O -r1 — -)-> QJ O) CL.4-J QJ O I— CO SCO QJ CO E O -1S_ S(J +-> o CD QJ QJ -1>^4-> in O QJ 4-> XI to SS•!— lO OJ a. to 3 XI o o cn 3 to 1— I— +-> >-> CD O +-) Q) CJ to I -M C QJ (/> 0) sa. QJ X o CO to >^ (O o X o 00 QJ Cl_ 1— z: CD -o _j QJ S3

PAGE 13

5 or partially purified LPS preparations in the fluid phase. This was in spite of the fact that it had been shown that heated LPS (lOCC for 60 minutes) could be coated onto the surface of a number of cells including erythrocytes (8). The latter are often the model target cells for hemolytic assays. It has become increasingly clear that the biological consequence of either an in vivo or in vitro encounter with LPS, whether cell bound or partially purified, is dependent upon its fixation to various target cells (9). For example, it was reported that patients suffering from endotoxin shock and sepsis due to gram negative bacteria had greatly reduced levels of blood platelets and that the platelets contained LPS. It has further been established that human platelets possess an endotoxin binding receptor, which when triggered by interaction with LPS, results in the release of a vasoactive amine, 5hydroxy tryptamine (5-HT) and the unmasking of clotpromoting activity (10). In animal studies, it was observed that guinea pigs injected intravenously with LPS also show a 95% decrease in blood platelets with a concomitant shortening of the clotting time (11). Additionally, it has been reported that there is a direct relationship between the susceptibility of different strains of mice to the lethal effects of Salmonell a endotoxin and the affinity of their red cells for either heat-killed Salmonel lae or free LPS (12). Little was known about the nature of the attachment of LPS to any cell until the late sixties, when G. Springer isolated an extract from the membranes of human erythrocytes having a high affinity and specificity for the 1 ipopolysaccharides of a variety of gram negative

PAGE 14

6 bacteria (13-15). This material, designated as an LPS-receptor, has now been purified to homogeneity and characterized. Springer has reported that the LPS-receptor is a 1 ipoglycoprotein rich in N-acetylneuraminic acid (NANA), galactose, hexosamine and contains about 61% protein (16). It appears to be a pentameric molecule with a molecular weight of about 228,000 daltons. The LPS-receptor functions by direct interaction with groups on the LPS molecule which provide an attachment site for tissue components (16). Strong evidence has accumulated suggesting that this attachment site is the lipid A moiety of LPS (17). This high affinity of the LPS-receptor for endotoxins is quite remarkable because both macromolecules are highly negatively charged: the receptor, because of its high NANA content and LPS because of its phosphoric acid radicals. Because the immunological specificity of LPS bound to erythrocytes remains unchanged. Springer has suggested that the lipid A of LPS binds to the specific receptor via clusters of hydrophobic amino acids which makeup about 40% of the total peptide content of the receptor leaving the polysaccharide available for the reaction with antibodies (16). A complete understanding of the orientation of LPS on tissues, bound either by specific receptors or by non-specific mechanisms, may come from studies involving the interaction of cell bound LPS with serum complement. The anticomplementary effects of LPS have long been established. For some time, evidence seemed to suggest that the single most important factor in the development of a noxious response to endotoxins was the direct interaction of the lipid A region with biological

PAGE 15

7 systems including the complement system (18). Whatever the mechanism, mounting evidence seems to implicate complement as a mediator of a number of the deleterious responses to endotoxins in experimental animals and man (8). More importantly, recent evidence indicates that the ability of LPS to initiate a complement response is not confined to the lipid A moiety but appears to involve the polysaccharide core as well (19). It has become increasingly clear that a major role of the complement system during an immune response is the mediation of enhanced phagocytosis with a concomitant induction of an inflanmatory response. This is accomplished by the sequential activation of the proteins which make up the complement system. The activation process can be divided into three major stages: a recognition stage, the generation of C3 cleavage enzymes and C3 activation stage, and a terminal or membrane attack stage (20-22). Under normal conditions, complement proteins exist in the serum as inactive precursors, and are activated by either of two pathways -the classical or alternative. The component of these pathways and their reaction requirements are summarized in Table 1. The two share a number of similar characteristics but differ, basically, in the reactants and the sequence of reactions of the first two stages The capacity of endotoxins to activate the complement system by a mechanism which requires neither antibodies to LPS nor the participation of the early complement components was demonstrated more than twenty years ago (23). More recently, it was recognized that fluid phase activation of complement by LPS is not restricted to the

PAGE 16

8 TABLE I Comparative Properties of the Classical and Alternative Complement Pathways Classical Alternative Activating agents Immunoglobulins of human rabbit guinea pig IgG2 ruminant Miscellaneous Activating site Factors required to generate C3 convertase IgGl ,IgG2, IgG3;IgM IgG IgGl (Lipid A) Fc fragment Cl^ C4 C2 IgA and IgE IgM IgGl IgG2 Inulin, Zymosan, Endotoxin or LPS, CVF F(ab)'2 (rabbit) or Fc fragment (human) Properdin, factor D C3b factor B Total serum requirement Dilute Divalent cation requirements Ca and Mg Concentrated Mg Cl is a trimolecular complex of Clg, Clr and Cls. Classical pathway activation is initiated when Clq binds to immunoglobulin.

PAGE 17

9 alternative pathway. Some preparations of LPS have been demonstrated to activate the classical pathway in either the presence or absence of specific antibodies(8,24-26,27). Normally, during classical pathway activation, the recognition stage is initiated by the fixation of the first component of the complement system, CI, to IgG or IgM immunoglobulins of specific immune complexes. This fixation and activation of CI generates a CI esterase whose major substrates are the fourth and second components of the complement system. Endotoxins in the fluid phase have been shown to initiate the activation of the complement system in vitro as described above in the presence of predominantly IgM antibodies directed against the 0-polysaccharide antigen (28). Unique to the endotoxins of certain gram negatives, however, is the capacity of the lipid A moiety of the molecule to fix CI directly resulting in a antibody independent initiation of the classical complement cascade (8,29). Alternative pathway activation, on the other hand, is triggered by direct activation of an initiation factor (IF) by endotoxins, or any one of a number of naturally occurring polysaccharides, and aggregates of IgA or IgE (30). It has been established that alternative pathway activation of the complement system involves the core polysaccharide region of the LPS molecule (31). Additionally, it has been shown that the length of the 0-specific polysaccharide antigen and differences in the carbohydrate content may also play a role in this mechanism of activation (8).

PAGE 18

10 In stage two of the classical pathway activation scheme, C3 convertase is generated when CI esterase cleaves C4 and C2 in the presence of magnesium yielding the cell bound "C4b2a" enzyme. This convertase cleaves the third component of complement (C3) into two fragments, C3a and C3b. The latter fragment is unstable and has a highly reactive hydrophobic binding site. As a result of binding to target membranes at sites adjacent to the membrane bound C3 convertase, a new enzyme C5 convertase, whose major substrate becomes the fifth component of complement, is generated. In addition, C3b undergoes secondary changes giving rise to an immune-adherence site capable of binding to a variety of effector cells of the immune system which bear specific C3b receptors (32). Additionally, C3b, either bound hydrophobically to a cell surface or free in solution, can produce further splitting of C3 via the C3 feedback cycle of the alternative pathway described below. It is this amplification of the generation of C3b and its deposition via immune adherence onto the surface of specific target cells (and in some cases innocent bystanders), where the major function of the complement system is realized. The mechanism of the alternative pathway generation of the C3 convertase is a bit more complicated in that there is a direct requirement for preformed C3b, the source of which is still not fully understood. This C3b in combination with factor B plus properdin and factor D forms the alternative pathway convertase "Factor BbC3b" capable of splitting C3 into C3a and C3b (33). This splits more native C3 giving rise to further C3b and the cycle is again repeated. This reaction

PAGE 19

n is magnesium dependent and unlike classical pathway activation it is inhibited by high concentrations of calcium (34). Under proper experimental conditions these cation requirements make serum chelated with either ethylendiamine tetraacetic acid (EDTA) or ethyl eneglycol tetraacetic acid (EGTA) useful reagents for distinguishing between the two pathv/ays (35). The former, being an effective chelator of both calcium and magnesium blocks the activation of the two pathways, while the latter, a less effective chelator of magnesium, preferentially blocks the classical pathway. The last and final stage of complement activation is the same for both pathways and is initiated with the cleavage of C5 into tv/o fragments, C5a and C5b by the C5 convertase. The larger C5b fragment then reacts sequentially with C6 and C7 to form either a cell bound or fluid phase trimolecular complex C5b67. The cell bound complex has the capacity to bind C8 and C9. If the cell to which this C5-C9 complex is associated is sensitive to complement mediated cytolysis, lysis ensues (36) Fluid phase C5b67, C3a, and C5a are potent anaphylatoxins and chemotactins (37,38). Once leukocytes such as polymorphonuclear leukocytes (PMN) and macrophages have migrated to the site of complement activation, phagocytoses is initiated. As previously stated, the phagocytic process is enhanced by the fixation of complement components, especially C3b, onto the surface of particulate antigens or target tissues which promote adherence, thus facilitating ingestion (39). Release of liposomal hydrolases, either as a direct consequence of ingestion or expulsion of an indigestible target into the surrounding tissues, results in the generation of additional chemotactic factors (40,41). This

PAGE 20

12 amplification represents an early stage in the development of an inflammatory response. • Severe tissue damage resulting from unchecked amplification of the activation of the complement system is probably minimized by the presence of two known serum inhibitors, CI esterase inhibitor (42) and conglutinin activating factor (KAF) also known as C3b inactivator (43). Perhaps there are other humoral or cellular inhibitory factors still to be defined. Many of the studies cited here employed fluid phase LPS for in vitro assays. Neter was the first to demonstrate that LPS coated onto the surface of sheep erythrocytes were sensitive to lysis by an antibody dependent classical complement pathway mechanism (44). Phillips and Mergenhagen employing sheep erythrocytes treated with LPS extracted by the Westphal-phenol procedure (45) confirmed Neter's observation showing the need for the presence of a naturally occurring yl globulin for complement activation with a sparing consumption of the early components C1-C4 (46). The in vivo consequences of the intravenous administration of endotoxins to normal and complement deficient animals have been studied. Both complement pathways are activated with the classical pathway being required for the development of many of the pathophysiological symptoms such as the thrombocytopenia observed in many experimental endotoxin infections (11). Some investigators have suggested that classical pathway activation due to an antibody mediated fixation of LPS to various cells such as platelets and erythrocytes is more effective in

PAGE 21

13 generating damage to normal cell membranes by inducing the selective release of a variety of vasoactive amines and cytolytic substances (47,48). Alternative pathway activation has not been shown to mediate the release of membrane damaging factors (49). In vivo systems are difficult to evaluate because of the large number of parameters which must be considered. Therefore, much study is needed before a complete understanding of the mechanism involved can be obtained. Regardless of the nature of the system ( in vitro or in vivo ) or the state of the endotoxin (free or cell bound), the interaction of LPS with the comolement system in the presence of specific antibody is extremely efficient, giving full response with sparing consumption of antibody and C1-C4 (48). Virulence of gram negative bacteria would then appear to be related to the fate and site of free endotoxins released as a direct consequence of cell death due to phagocytosis and interaction of this free endotoxins with other cell systems in the presence of serum factors. As previously stated, erythrocytes and now other cells have been shown to have membrane receptors which bind endotoxins. The biological role of these receptors is still not clear; however, fixation of LPS to the cell does appear to be a precondition for the triggering of many incitements of the immune system by LPS (49). Hoffmann, in the late 1960s, isolated and described extracts from the membranes of human erythrocytes capable of inhibiting the hemolytic activity of complement when sensitized sheep erythrocytes were used as target cells (50). Extracts isolated at two different ioni cities

PAGE 22

14 yielded products having different binding affinities for the membranes of sheep erythrocytes. Material prepared at an ionic strength of 0.15 was shown to be capable of binding to sheep erythrocytes and protecting them from immune lysis, and has been designated IH inhibitor. An extract prepared under the same conditions but at a lower Tonicity was incapable of binding to sheep erythrocytes but was capable of accelerating the decay of the complement component intermediate EAC142 to EACH. This material was designated DAF for decay accelerating factor (51 ). Preliminary studies indicate that the IH inhibitor is a large molecule with a molecular weight greater than 250,000. It appears to contain at least two sugar moieties, glucose and galactose and it is about 4% protein by weight (50). Existing data suggest that the IH inhibitor probably acts at the C3 convertase step. This was suggested by the finding that IH coated sheep erythrocytes in the intermediate state EAC142 consumed less C3 than untreated controls and that the inhibitory effects of IH ceased once activated C3 became fixed to the cell bound C3 convertase (51,52). Attempts to more fully define the biochemical and biological properties of these macromolecules have been hampered by the inability to obtain them in a highly purified state. Erythrocytes of different species differ in their susceptibilities as target cells in immune hemolysis, with sheep and chicken erythrocytes being far more sensitive than human and guinea pig erythrocytes. Differences within the same species have also been observed (53). For

PAGE 23

15 example, it has been reported that patients suffering from paroxysmal nocturnal hemoglobinuria (PNH) have at least two populations of erythrocytes based on their susceptibility to immune hemolysis (54), with at least one subpopulation exhibiting extreme sensitivity to attack by an antibody-independent complement mediated mechanism. It has been demonstrated that extracts from the erythrocyte stromata of these patients have reduced levels of IH inhibitor activityJ Additionally, evidence has been presented which suggests that there is a parallel between the presence of DAF on the membranes of the erythrocytes of certain species and resistance to the cytolytic effects of complement (53). A comparison of the isolation schemes for preparing the IH inhibitor and the LPS-receptor revealed a marked similarity between the two. Both activities are confined to that fraction of the membrane extractable by a butanol -water mixture at 4C and at an ionic strength of 0.15 or below. All of the activity of either material appeared to be localized only in the aqueous layer of the butanol extracted membranes. Ion exchange chromatography of extracts containing either the LPS: receptor or IH inhibitor activities indicate that both are eluted under identical conditions. These observations would suggest that the two activities may be similar or even identical. The principle objective of this investigation was to determine if the IH inhibitor and LPS-receptor are either the same or closely Data supported by personal experiments.

PAGE 24

16 related. Although it has been observed that human erythrocytes are highly resistant to LPS mediated immune hemolysis, the reason for the refractoriness of these treated erythrocytes has not been defined. Sheep erythrocytes, as previously stated, are normally sensitive to immune lysis but may be rendered resistant by treatment with diluted extracts of the partially purified IH inhibitor. Therefore, the second objective of this investigation was to explore the biological consequences of the IH inhibitor in the interaction of LPS treated erythrocytes and serum complement, in an attempt to clearly establish a biological role for the IH inhibitor. .1

PAGE 25

MATERIALS AND METHODS Erythrocytes Out-dated human blood (group 0, Rh positive) containing citrate-phosphate-dextrose as anticoagulant v^as obtained from the Civitan Regional Blood Center (Gainesville, FL). Whole sheep blood was taken by venipuncture from animals maintained at the Animal Research Laboratory of the J. Hillis Miller Health Center (Gainesville, FL). One volume of blood was mixed with an equal volume of sterile modified Alserver's solution (55) and the blood was stored at 4C for up to one month. Preparation of erythrocyte stromata Human and sheep erythrocyte stromata were prepared by the method of Springer et al. (56). Erythrocytes from whole blood were pelleted at 4C by centrifugation for ten minutes at SOO.xg and the plasma and buffy coat were discarded. The packed cells were washed three times with phosphate buffered saline (0.13 M sodium chloride plus 0.005 M potassium phosphate) at pH 7.4 and lysed in 10 volumes of distilled water at 4C. In the initial studies, the pH was adjusted to 5.3 with acetic acid and phenol was added to a final volume of 0.2%. The stromata were allowed to settle overnight at 4^0 and the supernatant fluid was removed. Ten volumes of cold distilled water were added, and the pH was readjusted to 5.3. The stromata were sedimented either by settling or centrifugation and the entire procedure was repeated six times with the addition of phenol 17

PAGE 26

OJ c: +-> T3 (0 an c o s'r— T3 Q. Q. X E (/) OJ O) ^(U o ^ s+-> 5 or MO +-> Q. fO q; E j=: u to O -M Q) E •f— M to 1 ro q; 00 O 4-> Q. :ji C _J O) o •4 i. c o o o c •ro +-> •rC fO ro -M E u (13 (D •rHto t— t U a. CO S01 0) Sx: < QQ.-P s-

PAGE 27

19 '4) 00 _i _i LU o Q O o _1 CO Q o X a. (/) •r— M >V O E c O) c: x: (O E S<*M-(-> O ZT. t-> ro 0) QJ L./ +-) Q. fa t/l OI VI 3 f~ O QJ O) Q. o 3 (/) c CT 3 rO (/I Ut n o S0) o o t+E +-> X o o sjQ (O o -M S+J JZ 4-> o -l-> a. X X! cu fO -M -C +j o >1 CL m o t/1 CM un 3 s_ o c M So X x: Q. n3 I 1/1 o I— O) I O <4E (T3 Mo 3 O -C s: j:: ^ 4-J o O O) •r. a. E S r3 UJ o 1 1— C o E O > O • O 3 •r-M •> O 3 or o r— Q. cz > 0) QJ O •.O c ST3 C\J re 1 Et SC3 a: 1 — t I— 3 X O -Q c/1 O) O 3 T3 LO O) O 13 LU +-> 3 x: a: IT3 C CL UJ ^ •rOJ :r Q Q.-l-> (/I CL o Q. tn CO • Lf) n: Q. O # E CO ^ O E SGJ o (U CD + c i•r— Ol S4-> Q. re 3 c 4-> 1 +J so o 3 E E •> i. 0) o Q. r— O S O; E o -a +J o O (U (J •!E +-> E U *-> LO "CJ 3 a. re OJ c ex: • re 1 — cu Sto LT) to E E +-> o so tj CO S +-> re 3 3 3 -r1 CD U 0) SX X x: 1 O .2 ID so 1— CD O) CO n. o OJ O o o >,o ~ — E o O "•-(-> o a. to o 5o O O -1CM o cu X QC 10 3 •f— OJ zn o <: 31 c T3 CL cu T3 a. cu oo re > O CL X •(cu re E Q. E CL'ra; o +-> i. OJ O) •> "o 1/1 x: •* 3 _l o m -(-> (-> E 4-> 4-> •rre to u XJ to re o Q. to 1 — i. 3 re re O a. re > x: •!(U re Tsscu o TD 4-> 1/5 rsi 3 o (U to >l o 3 +-> OJ t— CLI— 3 1— rt/5 E +-> 4O to lyi cu I/) CO •• MOJ 3 O) o q; re o. CO s: 3 s: •r— 4-> O cn i> S_ _j 1 o Ln O S>1 0) o re CO +-> a> Ln LU tn 3 1 — CM Ln > 3 (J 3 E -M Q. E o >>o o cC O E O 1 o CL. o cr o 3 O _l :z Ln CD cu +-> T3 •r— >1 -y 3 ^ CT 3 •rLlJ o a. 2:

PAGE 28

20 after every second water change. Following the final wash the stromata was collected by centrifugation, weighed and stored at -22C until use. Isolation and purification of the LPS-receptor The LPS-receptor was prepared as outlined in Figure 2 using a modification of the procedure of Springer et al. (15). A 50% aqueous stromata suspension was homogenized in a Waring blender and extracted overnight with two volumes of n-butyl alcohol at 4C for 16 hours at pH 8.2. Four phases, resolved after centrifugation at 2000xg for 30 minutes, organic, lipid, aqueous and solid. The aqueous phase was re-extracted twice with n-butanol once for 30 minutes and again overnight and then thoroughly dialyzed against several changes of 0.05 M Tris-HCl buffer (pH 7.0). An aqueous butanol extract low in LPS-receptor activity but high in IH inhibitor activity was obtained by shifting the pH of the butanol extraction from 8.2 to 5.3. The dialyzed active crude butanol extract was centrifuged at 151,000 xg for 1-2 hours in a Spinco Model L2 preparative ul tracentrifuge. Three phases resulted from the high speed centrifugation. Contrary to Springer's findings, the top aqueous layer possessed the highest LPSreceptor. After extensive dialysis, the aqueous top layer was applied to a 92 X 1.5 cm sepharose 4B column (Pharmacia Fine Chemicals, Piscataway, N.J.). The sepharose columns were washed with a 0.05 M Tris-HCl buffer at pH 7.0. Three milliliters fractions were collected and assayed for both LPS-receptor and IH inhibitor activities. The active sepharose fractions were pooled, concentrated tenfold by dialysis against 20% polyethyl eneglycol in 0.05 M Tris-HCl buffer (pH 7.0) and applied to a 22.0 X 2.5 cm

PAGE 29

21 DEAE-Sephadex A25 (Pharmacia Fine Chemicals, Piscataway, N.J.) column. After extensive washing of the column with the starting buffer, a linear sodium chloride gradient was initiated with 150 ml of 0.05 M pH 7.0 Tris-HCl buffer and 150 ml of 0.75 M, pH 7.5 Tris-HCl buffer. Three milliliter fractions were collected and assayed for the two activities. Ether extraction Sometimes LPS-receptor or binding activity could not be localized in the aqueous butanol phase after shift in the pH during extraction. The intermediate lipid phase was then further extracted with equal volumes of ether (Mallanchrodt, St. Louis, MO). Equal volumes of the lipid phase, suspended in 0.05 M Tris-Hcl buffer at a pH of 7.0 (1:2) and ether were vigorously mixed in a separatory funnel for 5-15 minutes. After phase separation the ether was removed from the aqueous and syrupy interphase layers by dialysis against phosphate buffered saline (pH 7.4) and from the organic phase by evaporation using a stream of nitrogen at 4C. Following evaporation, the residue remaining from the organic phase was reconstituted to its original volume with PBS and all phases were tested for LPS-receptor and IH inhibitor activities. Polyacrylamide gel electrophoresis (PAGE) Disc polyacryl amide electrophoresis was performed using a modification of the method of Maurer (57). Extracts were applied to 7.5% acrylamide gels and were el ectrophoresed in a non-reducing Tris-glycine buffer system, pH 8.6, for 45 minutes at 4C. The gels were stained with 0.02% Coomassie blue containing 12.5% trichoracetic acid. For some studies, duplicate gels were run. One gel was stained as above' with the remaining gel being sliced for analysis for LPS-receptor and IH inhibitor activities.

PAGE 30

22 Lipopolysaccharide (LPS) Lyophilized preparations of Salmonella t.yphimurium LPS extracted by the Boivin (58) and Westphal (45) procedures were purchased from Difco Laboratories (Detroit, MC). Heated (100C for 3 hours) and unheated LPS stock solutions (1.0 mg/ml PBS at pH 7.4) were stored at -22C until used. Antisera Appropriate dilutions of Salmonel la group B 0-antiserum obtained from Baltimore Biological Laboratories (Cockeysville, MD) were made in 0.01 M EDTA GVB". Hemagglutination (HA) titers of the sera ranged from 64 to 256. For hemolytic assays, rabit 193 antibodies to sheep erythrocyte stromata were obtained from Cordis Laboratories (Miami, FL). Stock solutions at a dilution of 1:100 in PBS (pH 7.4) were maintained until use at -22C. Treatment of erythrocytes with LPS Freshly acquired erythrocytes from a group 0, Rh positive adult were obtained and used in the coating and coating inhibition assays which were carried out as described by Springer et al. (14). For speed and economy, screening assays were assessed using a microtiter hemagglutination system. Briefly, the prodedure consisted of mixing equal volumes of either human or sheep erythrocytes at 2 X 10^ cells/ml and dilutions of LPS for 45 minutes with shaking at 37C. After extensive washing, the smallest amount of LPS which afforded maximal hemagglutination by subsequently added antiserum was determined. This dilution, defined as the optimal coating dose, was used in all subsequent hemogglutination-inhibition assays.

PAGE 31

23 LPS-receptor activity assay LPS-receptor activity was determined by measuring the ability of a material to inhibit LPS fixation to erythrocytes. The procedure in the coating HA-inhibition assay differed from that in the coating test in that dilutions of LPS binding material were added to equal volumes of an optimal coating dose of LPS and incubated with shaking for 30 minutes at 37C. Erythrocytes were added, and HA titers determined as previously described. In each assay, a control consisting of LPS and erythrocytes but no LPS-receptor material followed by the subsequent addition of antiserum was included. Isotonic buffer solutions employed in complement assays The basic diluent for most hemolytic assays was the isotonic gelatin veronal buffer (GVB) described by Kabat and Mayer (55) which contained 0.00015 M CaCl^* 0.0005 M MgCl2. and 0.1% gelatin at pH 7.5. In some cases, gelatin-veronal without CaCl2 or MgCl2> containing enough isotonic ethyl enediaminetetra acetate (EDTA, pH 7.4) to bring the final concentration to either 0.01 M or 0.04 M, was employed. These buffers were designated as 0.01 M EDTA-GVB and 0.04 M EDTA-GVB respectively. In order to achieve maximum sensitivity, hemolytic assays involving individual complement components, and IH inhibitor assays were performed using a low ionic strength gelatin-veronal prepared by mixing equal volumes of 5.0% glucose with gelatin-veronal buffer containing twice the standard amount of CaCl2 and MgCl2 (DGVB). Sensitized sheep erythrocytes (EA) Sheep erythrocytes at a concentration of 10^ per ml in 0.01 M EDTA-GVB were mixed with an equal volume of antibody to sheep stromata ata final dilution of 1:500 in

PAGE 32

24 the same buffer. The mixture was incubated with shaking for 30 minutes at 37C, and then at 0C for either 30 minutes or overnight. The cells were washed twice and standardized to the desired concentration before use. Complement (GPC) Fresh frozen guinea pig complement was obtained from Pel Freeze Laboratories (Rogers, AR). The serum was shipped in dry ice and was stored at -70C until use. In some studies, aliquots of GP serum were absorbed three times at 0C with either untreated or LPS treated sheep or human erythrocytes before use. Complement Components Guinea pig CI and C2 were prepared by methods described by Nelson et al (59) and Ruddy and Austin (60,51). Individual lyophilized functionally pure guinea pig complement components C3, C5, C6, C7, C8 and C9 were purchased from Cordis Laboratories (Miami, FL). ^ Complement component intermediates For IH inhibitor assays and complement consumption studies, sheep erythrocytes in the intermediate state EACT, EACU, and EACT42 were prepared by the methods of Borsos and Rapp (62). Determination of Tmax of EACT42 The kinetics of the generation of EACT42 was determined by the Tmax procedure described by Borsos et al. (35). ^ IH inhibitor preparation Crude human erythrocyte stromata ex. tracts, high in IH inhibitor activity, were isolated by a procedure described by Hoffmann (50). The method is outlined in Figure 2. Briefly, the essential differences in Hoffmann's preparation of butanol

PAGE 33

25 erythrocyte stromata extracts high in IH inhibitor activity and extracts high in LPS-receptor activity as defined by Springer are: (1) stromata were prepared at pH of 6.0 7.0 without the addition of phenol; (2) crude washed stromata were suspended in equal volume of 0.005 M potassium phosphate buffer at a pH of 7.5 and extracted with n-butanol at a final concentration of 20% for 15 minutes; and, (3) after the first extraction the aqueous butanol phase was adjusted to an ionic strength of 0.15 by the addition of 3.0 M NaCl. Butanol extraction of the adjusted material was repeated until a lipid phase could no longer be separated. For further purification, concentrated material, active in IH activity, was subjected to gel filtration and DEAE chromatography. Treatment of sheep erythrocytes with partially purified IH inhibitor material Lipopolysaccharide coated and untreated sheep and human erythrocytes were treated with IH inhibitor material by the procedure described by Hoffmann (50). Equal volumes of erythrocytes at 10^ cells/ ml in DGVB and extracts of IH inhibitor diluted 1:10 in DGVB were mixed at 0C. The mixture was transferred to a 30C water bath, incubated 30 minutes with shaking and was pelleted at 500xg for 10 minutes at 4C. The coated cells were washed two times, and standardized to the desired concentration in the appropriate buffer. IH inhibitor activity IH inhibitor activity was assessed using the EAC142 inactivation assay described by Hoffmann (50). Sheep erythrocytes in the intermediate state EAC142 at 10^ /ml were mixed with an equal volume of IH inhibitor material diluted in DGVB. The reaction mixtures were incubated at 30C for 15 minutes with constant shaking.

PAGE 34

26 after which, three volumes of guinea pig complement diluted 1:25 in 0.04 M EDTA-GVB were added. The tubes were then incubated for 60 minutes at 37C with shaking. At the end of the incubation period, 10 volumes of ice-cold PBS was added to each reaction mixture. The cells were pelleted at 500xg for 10 minutes at 4C and the optical densities of the supernatant fluids were determined at a wave length of 414 nm. Chelators Stock solutions of disodium ethyl enediametetra acetate (EDTA, Fisher Scientific Co., Fair Lawn, NJ) and ethyl eneglycol bis (beta-amino-ethyl ether) N^,N tetraacetic acid (EGTA, Sigma Chemical Co., St. Louis, MO) were prepared as described by Fine et al (35). The stock solutions were stored at 4C and diluted to a final concentration of 200 mM before use. Magnesium EGTA was prepared as described by Fine et al (35). Complement Consumption The ability of erythrocytes treated with LPS and/or IH inhibitor to consume complement was determined in reaction mixtures containing 0.1 ml of the treated cells (1 X 10^ cells) or LPS and 0.9 ml of normal or absorbed guinea pig serum chelated with either EGTA or EDTA. The mixtures were incubated with shaking 60 minutes at 37C. Following the incubation period, the cells were pelleted at 500xg at 4C for 10 minutes. The supernatant fluids were reconstituted with magnesium and/or calcium and were analyzed for residual whole complement activity using a modification of the procedure as outlined by Kabat and Mayer (55).

PAGE 35

RESULTS Erythrocyte coating by LPS and its inhibition Repeated titrations of Salmonella typhimurium LPS at concentrations ranging from 0.195 ug to 50.0 ug/ml as determined with polyvalent and homologous Salmonella antiserum, employing erythrocytes at 2.0 X 10^ cells/ml were carried out. The results of a representative experiment employing polyvalent antiserum are given in Table II. It can be seen that heating the LPS enhanced the erythrocyte coating capacity to a remarkable extent. Additionally, a maximum titer resulted when erythrocytes were exposed to at least 0.78 ug/ml of heated LPS. Therefore, an optimal coating unit in) of heated LPS (defined as the reciprocal of the greatest dilution of LPS producing complete hemagglutination by either polyvalent or homologous antiserum) was taken as 0.78 yg/ml Assays employing homologous antiserum to heated LPS yielded lower optimal coating doses of 0.78 yg/ml and 1.56 ug/ml, depending on the age of the antiserum. These results were identical for human and sheep erythrocytes. LPS extracted by the Westphal procedure resulted in an optimal coating dose of 0.39 ug/ml as also determined with homologous antiserum. LPS-receptor activity, as evaluated in these studies, was based on the ability of a given erythrocyte preparation to inhibit the complete fixation of an optimal coating unit of LPS onto either sheep or human 27

PAGE 36

28 TABLE II Determination of Optimal LPS Concentration Used for Coating Human Erythrocytes Titers^ LPS Boivin (yg/mi ) unheated heatedb 50.0 80 m 25.0 • 160 • m 12.5 6.25 80 NO 3.125 80 160 1.563 80 160 0.781 40 160 0.390 20 80 0.195 10 40 ^The reciprocal of the dilution of anti LPS serum affording maximal hemagglutination. b Stock solutions of LPS (1.0 mg/ml PBS) were heated 100C for three hours.

PAGE 37

29 erythrocytes. Table III, Column A summarizes the results obtained with several crude aqueous butanol preparations of erythrocyte stromata. These data indicate that the range of LPS-receptor concentrations or dilutions needed to yield optimal inhibition of LPS erythrocyte coating varied with the source, concentration and condition (solvent) of the erythrocyte stromata extraction procedure. IH inhibitor activity of crude butanol extracts of human and sheep erythrocyte stromata EAC142 inactivation by a crude erythrocyte stromal extract, assayed by the technique described in the section on materials and methods is shown in Figure 3. This procedure was used to determine the inhibitory potency of most extracts. Color controls for the presence of hemoglobin in the higher concentrations of crude preparations were necessary. The reciprocal of the dilution of a crude butanol preparation yielding greater than 50% inhibition of the lysis of EAC142 by C-EDTA are also shown in Table III. These results clearly indicate that LPS-receptor and IH inhibitor activities are contained in significant amounts in crude butanol stromal extracts obtained by either Springer's or Hoffmann's procedures. Of interest also, is that the potency of the two activities varied to the same extent. A comparison of the chromatographic properties of the IH inhibitor and LPS-receptor from erythrocyte stromal extracts The above data suggests that the LPS-receptor and IH inhibitor are either identical or closely related molecules, therefore, additional evidence to resolve this issue was sought. Human and sheep erythrocyte stromata were subjected to modifications of Springer's purification procedure as outlined

PAGE 38

30 TABLE III A Comparison of LPS-receptor and IH Inhibitor Activities of Several Crude Extracts of Erythrocyte Stromata Crude Stromata Extracts Titers A B LPS-receptor IH inhibitor' Human cells extracted at pH 8,2 (Springer) Crude aqueous butanol phase Crude aqueous phase high speed^ top layer high speed syrupy interphase Human cells extracted at pH 7.5 (Hoffmann) Crude aqueous butanol phase Sheep cells extracted at pH 8.2 (Springer) Crude aqueous butanol phase Crude aqueous ether phase Crude ether interphase Crude ether organic phase 40 320 80 320 0 0 8 0 40 320 40 320 0 0 0 0 *The smallest amount (dilution) giving complete inhibition of LPS coating. The reciprocal of the greatest dilution giving 50% inhibition of EAC142 lysis 'High speed extracts were obtained by the centrifugation of the crude aqueous butanol phase, 151,000xg for 2 hours.

PAGE 39

0) £Z re D re > 1 • E 1 •rre re iO -t-> E CD •rtJ J= ^U O T3 CD CD re Cl re sCD Q. CD 31 E CU 4— -t-> +-> CO T3 +J o ho 0) to re 3 O sSSUJ jc c +-> -t-> -D (J cn (J o CD to 1 -4-) o (U o +-> o •1— o 3: +-> CO x: CD to HTD +-> +-> o o to ^ re 4-' to Sin o (J c 4Q) CD re +J to X o C Sc ^ o X -o OJ a) Sre •r+-> r— •r— 0) s0) 3 re X +-> o o +J f" O" -M CD o CD 4-> 3 •r— 0) (D to (J cu CU J2 £ t/1 re 4to c: SQ. +-> OO %. >, cu cu +J SCD o r — lO 4-> I. re S c CD +-) c Q. 1— CD o o x: 4-> E o E x: c o +-> re (U o re to o o cn s_ i+-> c: •r3 (0 re s3 00 +-> o re o 4a; > E GJ CD CJ +-> •rcn c CD re re >) s+-> 14s. a (_> SE +-> +-> o -)-> 3 i. 4c: o c f4J to il •r— o s> O 4-" to 4-> c to Ol o +-> re U o o 01 QJ CD re CD (/) u O 4-J to CD 3 CD CD +J ai o C x: 4+-> SQ. •rsto S0) 44J •r— O o to X3 Q. s_ O o jre 4-> •r— to +-> 3 x: +j SJZ 0) re >.-a >^ XI cn x: sCU JZ •r— CD X •r•r•4-> Q. o 4-> 3 (J CD x: CO 3

PAGE 40

32

PAGE 41

33 in Figure 2. Figure 4 is a gel filtration elution profile on sepharose 4B, of the crude high speed top layer obtained from the crude butanol human erythrocyte stromal preparation. Fractions were monitored at 220 and 280 nm and were assayed for LPS-receptor and IH inhibitor activities as previously described. Two peaks were observed with both activities eluting in the peak following the void volume. Close examination of the sepharose 4B profile indicates that there is a slight displacement of the IH inhibitor activity to the left of the LPS-receptor activity. This would suggest that perhaps the two activities may be different. Further attempts to separate and purify the two activities were accomplished using ion exchange chromatography. The sepharose 4B active peaks were pooled, dialyzed against the starting Tris-HCl buffer, and applied to a DEAE-Sephadex column. Fractionation was accomplished with a linear NaCl gradient. A typical chromatogram of the partially purified material (s) is shov;n in Figure 5. LPS-receptor and IH inhibitor activities eluted in a relatively narrow peak at about 0.3MNaCl, again with the IH inhibitor slightly preceding the LPS-receptor activity. The recovery of the LPS-receptor and IH inhibitor activities following sepharose 48 and DEAE-Sephadex chromatography is presented in Table IV. It should be noted that gel filtration on sepharose 4B yielded only about onefold increase in the purity of both activities with recovery of only 52% of the LPS-receptor activity and 65% of the IH inhibitor activity. DEAE-Sephadex was shown to result in as much as a 19-fold increase inactivity, but resulted in a recovery of 50% of the specific LPS-receptor activity but only one third or 31% of the

PAGE 42

3^ M +* U t +J CQ LU O) E , c +-> QJ •IOO C > O •rO)-!O s_ I— O r— 4J O •rLL. o C <: Q. (T3 (T3 Q. T3 00 C O) I rO +-> UJ >i< SO LU O O O -M S_ Q.^ -a (U +-> c O >, n: QJ SSLU I c _l B o u 8 ^ — J lX3 CO O Lf) •>3: r~ > -M 1— 1 lli J. fO O o o •r41a: CO • • •01 o 1 — 13 Q. f — >> 4a. _j 1 — >" M— (J QJ Q. V) u o cn 'e' ^ r— +J O) >CM CM (0 &. o •-> •1" C •1^ 00 1 ~ e Ol OJ a. 1 CM 0) j:= £: _j LT) u +J +j s. o u • w o "4— T3 O •rOJ 0) o CO •r+J :r > • >) — 4.1l-H O O •r> +J > > U •!Qi OJ QJ +j cc: CM o •r— i. a. (J 00 r-. in 00 +-> >O 00 et Q. CO (0 o 4_l c o CM o o UJ to >, OJ (/) OJ o o o 00 oo 4>, un dCO OO 1— !-> n: fO "1— 1— 1 CTi ^ _J c >
  • +- > o •>'+o o •r— t— 4-> OC o CM o -M r+J O fO u O 00 ID 00 c r** o j:3 _l 1 — •r— +-> 5 J3 ^— ^ OJ •r" r— • • •r— -t-> o E 4u 03 O O r— CO o > OJ -l-> C CVJ r> LD Q. O O CM C\J 10 1— O r— XI +J o O o o X ta 4>, IX) CM CM 00 -tJ O 4-> n: CO LI o ^— r* — t +j JZ > CTi o ^-> CJ •rta •r— s_ o fO -M a: M .> o +j QJ u OO o CO > 3 c S<: DCM 1 — CM r— >o —1 CO 1 — cn •r0) o 4-o 3 QJ O ^ cu • •r-M 13 O 3 4-> --^ o to > — ^ <0 r'1— QJ 03 1— > SE •4-" QJ > M ^ C O CM in CD CU ^ QJ jC Q) CM cn 00 o •r— •M U OJ ( — f— T3 D1 O) s= < O — o C +J o •r— r— o .— QJ 4> •tCL ?= o E O r— > 1> 3 LT) o 1 — • ID 1+-> r— ^— • • • (0 f— • T3 r— TD o = rO >5 >, > — CQ ^ — +J O O o f— a O 1 OJ s_ -l-> o o a o S"3 1 — Q. fD QJ Q) 4-> QJ 1— QJ o o +-> C 4-> +J to 4-> O 1/1 > Q. QJ (O CJ +-> (O (O QJ (O to O M O O 0) > E 0) r— E sS:3 QJ o sSo OO O CO sOJ 3 •<-> 3 (O QJ Q) QJ 1 X CJ u o O fO O J= OJ >, x: 1LU QJ QJ QJ QJ i— QJ QJ CL S_ Q. eC -O So fO S_ rO S. OJ CO 1 — QJ ca LU 1— o C_> -t-> :_5 oo Q o "O QJ 4-

    PAGE 43

    0 ^ s+-> X a: 4> t|CM ro >> O CTl 5 5— S01 rO) 1— MCU C (.1 o to 1 — cu o ro Ol -l-l u ro s4J X o E ro +-> C s. Q. o -i-> ro 4-> 4QJ CD • r— J3 O O t/) • CO si O) CT) CD •ro Ln sso Q. ro oo ^ o Q4QJ ^ O on +-> T3 ro S o -COO +J +-> o CD E -D OJ +J CD -rro CD E -t-J CD ro X! i>i E CD 4-> ro u u sO-T•!n +-> o •ro. o -o O) CD 4-> • O o I— CD O O 4-> O CD E CD -C c: I— -t-> I I ro a. _j tO CT/ JO > •Iro c •r(/) ;c o I — I 'r— +J CD <~) ^ ro +-) s_ o CM OJ Q. CD >) ro jQ Q. CLi — ro a' ro E -ro Q. SSCD JZ ro +-> o -(-> ro ro E "O E c o "o ro iCD 4-> -l-J'-^ to o E ro u O) i+->+-> IT) >, X • O O) 1 — -E >5 O U CL' CD s m c o -!-> U O ro CM SCM •1-1 — ro E 1,-5 CL E o o o CD ^ +-> I— ro E S_ ro <4O CD 4-) O CD E -t-J i. OJ +-> -(-^ c 4-3 CD ro t/5 Cl CD SE Q. O CD •r54-1 zs m — sCD ro ja O) -E CD ro • +J -E >, ro to +j o -r•rCD > T3 -E -rE +-> 4-> •Io "D ro to E (D ro S1— o cn ^4-> E +-> Q. ro -rCD •I> <_) i-ra; •4-J +J S-

    PAGE 45

    0) >. in 1 Ol cn o c CM Ln c 4-> E s_ o o (T3 •r(D O) X sO JSZ So (/) re o c nh +-> Q. n o to o (/) E o c CM 0) +J E SO (0 QJ u (O fC cu Q. s+J c fO O) CD •4-J fO 0) s(/) s_ sX E a. E +J ra Ol Q.-D O to -(-> ro (D •r— -D (0 sO ra 3 s"e +-> 0) o (0 s+J Q. cu U O o Sre GJ CD -(-> ro reto -o xz^u^u S1 — Ol — "u+^i — ratofo cn E E I— -E Ol Oro Ot-MM-OjS(->E'— ^ 1 — OfCS "OEQE-EID 30.1/10)^ Q. o E-'-rooj— +->T3a) S_E ET3 -r-O EO • Oto-US-M-fOO) o i. O in ro o, S_ 4CD CM 2 to (J CM 1 Ln CD CM SE E Na00 CM to E ro o QJ T3 oil— Q1 E +-> QJ • -a Ol •rO _l t o Sro E to 4-> < CL 1— o •1LU Dl X +-> to >> r— :e ro r— to E Ol u +-> o u O Q. "D re 4•r•a: •r— Q. QJ > re So to •*-> QJ QJ ro 4E s. E SZ Q. -o OJ • QJ O -M o to OJ o -D E 4'r— m U~l > JE E "+>,-(-> 4c 1 -l-> • J3 3 O o LU +j OJ ro I— J2 X3 •r•a; o E o O QJ E QJ t/) -(-> LU ro •r— CJ to •4-> 2 U Q QJ +J o o QJ to ro Q. E SCLl ^ JZ >5 SCC +-> O O 1— 1 — to 4-> I— 4in 0)

    PAGE 46

    38

    PAGE 47

    39 original IH inhibitor activity. A comparison of the gel filtration and DEAE-Sephadex elution profiles (Figures 4 and 5) and the data in Table IV, revealing a 98% loss in the total mass but with a 19-fold increase in activity, indicate good purification of the two activities. These data do, however, suggest further differences in the two activities by the differences in their respective recoveries following sepharose 4B and DEAE-Sephadex chromatography. The homogeneity of the active DEAE-Sephadex fractions was checked by electrophoresis on 7.5?'. polyacryl amide gels at pH 8.6. As shown in Figure 6, 5 12 protein bands were observed on the crude extracts with a sharp band and a diffused staining area localized at the top of the gel when partially purified DEAE-Sephadex fractions were el ectrophoresed. Analysis of duplicate gels indicated that most of the LPS-receptor activity was localized in an area about 6 mm into the gels with the IH inhibitor activity spread over a fairly large area at the top of the gel with a peak of activity at about 11.0 mm. (Figure 6). IH inhibitor and LPS-receptor activities of sheep erythrocyte stromata The aqueous phase of butanol extracted sheep erythrocyte stromata, prepared by Springer's extraction procedure at pH 8.2 and at pH 5.3, had neither detectable LPS-receptor nor IH inhibitor activities. LPS-receptor activity but no IH inhibitor was, however, observed in an ether soluble fraction of the n-butanol lipid phase as was shown in Table III.

    PAGE 48

    So D. ro O) E 0 3 cu ^ 1 4CO O Ci. I in 4J 4u O ro S_ C -(-> o X •I +-> T4-) C (J O) ro O) 5 i*-> o c +J -rro O) >— s_ s-a o c o ro I ro I 1— Q. >) ro sU O) ro s>i cu CL LTl — o u +-> ro o) i. s"O 4-> CD QJ X 5 •M LlJ 0 "D 01 C •>-> • ro J3 to 3 •.tn to to 1 — C SCD q; o -C ^ 5^ S O-LD O • ro sr-. ro <_> oj E O) +J o r— ro s> o U 00 o O -r+J s-o x: XI QJ E QSOl I to M13 4-) ^ <_) O) O) r— C (U U i>, O) 1— 4-> Cn4I cr (/i •r— hej o •r4-) (_> ro 3 X! to QJ Ol S+J I 3 C C 0 •!C E ro Ln o -O <4OJ to KD 01 • SCO o :r: CL n. I o (J OP +-> ro •14_> 2 -r3 C to QJ O +J 4-1 O C SQ4J S3 0 o <+1/1 X! ro QJ 3 £ •Ii_ ro QJ 4-> x: t>0 4-> o to ro QJ 2 -C 4-> r— QJ XJ 01 C ro QJ C QJ O 3 to •rQJ to -rQJ to o ro J=: E CL o c I o •r-O •— 1 4-> Q) 3 n: to QJ 1— t— 1 C ^ OJ I— 4-> SU QJ o x: <+O) o +-> X I— QJ N • 1 — to ro QJ C -rro +-> X > C -rro 4-> o XJ ro QJ 4-> S3 O r— +-) Q) Q. QJ QJ (_> +-> C QJ 1/1 >, QJ 4-> S•!Q. > QJ TO to ro •r— x: XJ S C ro to 4-) s. c o QJ 4-> E Ol QJ Tto x: so +-> Q. QJ O QJ SI LO 4-J QJ Q, ro x: _j U 4-> •r0) X X x: C E 4-> •Iro QJ X QJ X ro x: to to QJ so o r— 4-> a; cn-ir— C X2 •!ro -r4•r^ O SE i. 4-> •!Q. (U

    PAGE 49

    41

    PAGE 50

    42 A significant difference in the two activities was evident when it was serendipitously observed that a shift in the pH from 8.2 to 5.3 in the butanol extraction step of the crude erythrocyte stromata resulted in a preparation consisting of little or no detectable LPS-receptor activity, but high in IH inhibitor activity. The elution profile of the IH inhibitor activity following ion exchange chromatography on DEAE-Sephadex of the partially purified material, as shown in Figure 7, was similar to that observed for material extracted at pH 8.2. LPS-receptor activity assays were not carried out on the remaining butanol phases because these materials had been discarded before the impact of these observations were realized. A shift in the pH to 5.3 of a crude butanol extract of erythrocyte stromata obtained at a pH of 8.2 affected neither the LPS-receptor nor IH inhibitor activities. This suggested the possibility that the LPSreceptor activity of the material extracted at a pH of 5.3 was not destroyed but was probably redistributed into another phase. Stromal extractions were carried out at pH 5.3 employing smaller volumes of erythrocytes stromata, using Springer's procedure, in an attempt to localize the LPS-receptor activity. As shown in Table V, some IH inhibitor activity was observed in the crude aqueous butanol phase when 10.0 ml of packed stromata were extracted. There was no detectable LPS-receptor in this butanol layer. The lipid phase mixed with 10 volumes of PBS was further subjected to an ether extraction resulting in four layers: a pellet, an aqueous layer, lipid interphase, and an organic layer. All of the LPS-receptor activity was recovered in the organic ether layer

    PAGE 51

    43 c c/^ E M— 5 o o zs 4_> (J •1— O > i_ •rD+J O CD ct ^ i_ O CT. -Q V) •r^ C CO uo LlJ ^ J 1 ^ CQ Pi "O 1 1 o SX 1 LjJ CU ro _I 4-> no tE o o s_ S1/1 OJ > QJ o +-> o >, O) (J q: o +-> >1 s> c •rLD -(-> rn O 1-1 CM CM o Dc c cr. o CU _J CL Ln c 4-> CD • •r— LO CM o cr o > CO t/) •ryf — *-> m o >^ •a: 1 — ra "3 (13 > CM +-> c o C o CM •rO a> +-> ht r— o (O EA 4OO o _1 C +-> o c +-> +J •1— so JU cu (L> +-> (J CM +-> x: +-> c (_) CO CM •rc ro Coi O LO o Q • O 3 1— imu c > O •r+-> X t. cr. ro -l-> E D. •n-M E E CD cn o <4C c O (tS c o CM c un c •r+-> > 3 to o leve o to (U x: CD 3 CD •rx: T3 o he -t-> +-> M CU 3 3 •a o o (U E to ro T3 "o o ro (O CU c c s^ So c (O ro Ol Q. OJ to o •r— c: +-> 4-> x: S_ SI CU iE o 3 3 +J 0) +-> Q. J'rXI X QJ +-> 0) CU +-> c ro o +-> CU x: +-> > XI +-> o ro ro o o CU ro > CU CD o > o •a CU 3 o ro C_3 -a

    PAGE 52

    46 with IH inhibitor activity distributed in both the aqueous and lipid interphases. A comparison of the purification tables (IV and VI) of crude extracts obtained at pH 8.2 and pH 5.3 indicates that, as was observed at pH 8.2, extracts obtained at pH 5.3 and subjected to DEAESephadex chromatography resulted in a substantially greater loss of total mass as estimated by the adsorbance at 220 nm and the yield of IH inhibitor activity. Treatment of Springer's crude butanol extracts with sheep erythrocytes To further establish that the two activities are distinctly different, equal volumes of sheep erythrocytes at either 10^ or 10^ cells/ml were mixed with equal volumes of a butanol high speed top layer extract prepared according to the procedure of Springer. This extract had an initial LPS-receptor titer of 128 and a IH50 inhibitor titer greater than 80. Control tubes consisting of equal volumes of buffer and the extracts were also prepared. All tubes were mixed at 30C for 30 minutes with shaking. The cells were pelleted by centrifugation and the supernatant fluids along with the buffer controls were diluted and assayed for LPS-receptor and IH inhibitor activities. As can be seen in Figure 8, the IH inhibitor activity was reduced substantially when extracts were treated with 10^ cells/ml. In contrast, the LPSreceptor activity remained constant when the cell treated and buffer control supernatants fluids were compared. The biological consequence of the IH inhibitor and LPS-receptor on the erythrocyte membrane The data in the previous section indicated two essential points. First, the membrane of the human erythrocyte.

    PAGE 53

    Treatment of partial purified extracts of human erythrocyte stromata prepared by the method of Springer. Either 10^ sheep erythrocytes or buffer were mixed with an equal volumes of the extracted material. Following a 30 minute incubation period at 37C, the cell suspension was pelleted and all supernatant fluids were assayed for LPS-receptor and IH inhibitor activities. The hatched bars represent the sheep erythrocyte treated extracts, and the opened bars the buffer treated extracts.

    PAGE 54

    48

    PAGE 55

    49 o >, O CL to O tl (U SSQ_ CD +-> o > •1 — +-> E +J c u >i sca: j2 -c ;_) Sro o -l-> LO 0) c o OJ o CLJ= _J (1) d +J |_lJ -a O Q c rn to so < +-> *-> 1— t. X o LlJ "O +-> (O 4-> +-> o ra QJ O XI 1 — < s. ^ +-> oo H-H 00 o QJ C MM , o cu sen > o o QJ LjJ 00 q; o o c o LT) o • LT) +J >—
    0£. 00 1 D. CM 1 a O Ln •1+-> 4T1 — 1 •1> O -10) +-> g: CTi G. U 00 CM OO CC Q. _i O O o O O o O >• — in 00 +-> 1 — t LT) fO •>Cvl +-> > O r1— +-> q; o oo o o <: Q•n CM _J 1 o E fO o O LO CO OJ cn o o CNJ CO 1— o <: o o o LT) 1—) > la q; 00 U o. _j c o •1— !-> nj 'i icn -!-> o 00 c CM QJ C\J ro O <^ c o o QJ E o LO E O > c X X o OJ QJ •r•Q o +J (O 3 -C SCL CL (O QJ 4-> QJ Q. 00 Z3 t/0 OJ 1 Q. iLU c LU a. < •r< LlJ O LU O cn J — 4-1 Its o oo Qo o J3 E •f— X ra CP. c > CD o oo CM "vjO c o o uo cn > •i — cr. c o -M ui QJ jC CD QJ O O CO QJ QJ o o CL (J QJ SQJ (J c o O o rT3 S+-> c QJ U c o o +-> o !-> i. o to QJ (t3 > QJ 4J )-> O .13 4-> O so <+QJ to > QJ x: +j CD c > >, a QJ la 3 o 03 O o U 1 1 •f— u > QJ •rCL 4-> (/) O 4-* "r" C +j r— QJ QJ je: x: 4-* So o CU CU — CU >, >i -Q _Q t/) 4-* C •r> (J •r— 1. CJ O. > (U .c 1 \ 1 t % T— CJ o -M o O 03 +-> ) r— X) r— 03 > o > QJ OJ CU x: +-> CL CI 4-> c T3 •1— > ft > 4J o "O gj >> >> (-> fO QJ -a QJ S_ CU !-> +-> 4J (O 03 QJ 3 3 O o O
    PAGE 56

    50 which is fairly resistant to complement mediated lysis, possesses at least two distinctly different molecules with biologically apposing properties: 1 ipoglycoproteins with a high affinity for lipopolysaccharides which are potent activators of the complement system, and a class of molecules shown to be potent inactivators of complement. Second, sheep cells which are normally extremely sensitive to immune lysis have been shown to be devoid of the IH inhibitor but possess molecules with an affinity for LPS which are confined to the lipid moiety of the membrane. As shown in Figure 9, the interaction of free LPS (extracted by both the Boivin and Viestphal procedures) with guinea pig serum which had been absorbed with sheep E coated with LPS (E-LPS), resulted in a substantial consumption of complement. Additionally, it can be seen that LPS (Boivin) appeared to be a much more efficient activator of the alternative complement pathway compared to LPS extracted by the Westphal procedure. LPS (extracted by both procedures) coated onto the surfaces of sheep erythrocytes showed a similar profile (Table VII), except erythrocytes coated with LPS extracted by the Westphal procedure were far more efficient activators of complement in the absence of natural antibodies to LPS. It was of interest, therefore, to determine if LPS on the surface of sheep erythrocytes, in the presence of the IH inhibitor and E-LPS absorbed guinea pig serum, would alter the complement consumption profile of E-LPS. To explore this possibility, erythrocytes were coated with LPS and IH inhibitor then reacted with guinea pig serum (absorbed

    PAGE 57

    51 TABLE VII Consumption of Total Compleinent in Either E Absorbed or E-LPS Absorbed GP Serum by Untreated and LPS Coated Sheep Erythrocytes Cell Suspension Percent Consumed E absorbed E-LPS absorbed serum serum E ^LPS Westphal 21 .3 16.89 r^LPS Boivin 47.9 6.38

    PAGE 58

    E cn OI D 3 ro s0) 'r— +J 3 *-> i~ 4/1 C7> ro 3 QJ E CU E l43 S-io •r+j r > — M o o Q. QJ E Q. OJ +-> QJ c: O t/! 3 CO >: l/l •r+-> sQ. +-> o a; o QJ _J SC D OJ QJ OJ QJ +-> LU I— C 1— ro T3 a; -M "r— QJ QJ •O Oi sp: s+-> c c ro ro o c 3 C_3 Q) ro Q) E +-> 0 4-> Sc st/l QJ QJ >-> (O to OJ s_ CO -o l/l > E 1/1 QJ t/l +-> OJ XJ o ro C OJ "O ro i. Q) CO j= XJ CL' +J Qro OJ S2 3 Q) 1/) c ^ 4-> 3 -l-> ^ QJ •1Q. -o > sOl to 4OJ C -r•r— +J QJ t-J O QJ 3 o C •ro QJ *J E rcc o O) 4QJ 0) •r— (/) o SQJ c >,-c: c LT) HSE -O 4J ro o E O 3 O (/) o 3 SU E >, *-> •f— Q. •rro QJ O Mc 1. QJ o cu -o ro ro 4-! ro j:: l/l QJ Qu +-> QJ c: 4-> QJ -C c C7> O t S(J to o •rro 1— +-> -t-J Q. S_ C£5 ro c c +-> LU QJ QJ E ro X E to 3 Ql QJ _J S3 Hi +-> CO c < q; •r— S•1OO tQ' +-> 1/5 o 3 Qo x: 4QJ QJ C 4-> ro C i. ro QJ

    PAGE 59

    53 % COMPLEMENT CONSUMED 30 20 10 UNTREATED SERUM E-LPS TREATED SERUM BUFFER EGTA EDTA % COMPLEMENT CONSUMED 30 20 10 BUFFER EGTA EDTA

    PAGE 60

    54 with E-LPS V/estphal). Five grouDS of cells were prepared: one group was coated with LPS only (E-LPS); a second group was coated with LPS first, then was treated with the IH inhibitor (E-LPSIH); a third group was treated with IH inhibitor first, followed by the LPS-receptor (E-IHLPS); a fourth group was treated with IH inhibitor only (E-IH); and, a fifth group of cells (E) was treated with PBS under the same conditions and served as the control. The efficiency of the LPS coating procedure in the presence and absence of the IH inhibitor was evaluated by assaying the five groups of cells and their respective supernatant fluids for LPS activity, employing the hemagglutination assay (described in Materials and Methods) as determined with antiserum to Salmonella typhimurium group B. The results of these assays indicated that all of the LPS treated cells adsorbed equal quantities of LPS in the presence and absence of the IH inhibitor. The complement consumption profiles of the five groups of erythrocytes treated with E-LPS Westphal absorbed guinea pig serum are presented in Figure 10. It can be seen that the presence of IH inhibitor on the cell caused about 50% reduction in the LPS mediated hemolytic action of complement. Of particular interest was the fate of the cells when complement was treated with IH LPS coated erythrocytes; significantly less cells were lysed. This was in contrast to the case where complement was mixed with E-LPS and the cells were completely lysed. These results suggested the possibility that complement activation had taken place, but that cells were protected from > -' lysis by the presence of IH inhibitor on the membrane. .; <

    PAGE 61

    Figure 10. Consumption of complement in E-LPS westphal absorbed guinea pig serum by LPS and/or IH inhibitor coated sheep erythrocytes. One tenth ml containing 10^ sheep erythrocytes (E) either untreated or treated with LPS (E-LPS), IH inhibitor (E-IH) or LPS and IH inhibitor (E-LPS-IH) were incubated with 0.9 ml of guinea pig serum at 37C for 1 hr. Residual complement hemolytic activities were assayed and the % of the available complement consumed was calculated.

    PAGE 63

    DISCUSSION The experiments reported here have demonstrated that extracts from human erythrocyte membranes possessing LPS-receptor activity obtained by the method of Springer et al. (15) were also capable of inhibiting complement mediated lysis. The anticomplementary activity of these extracts was demonstrated to share many of the properties of the IH inhibitor previously described by Hoffmann (50). Data are presented which strongly suggest that the two biologic activities are closely associated, but separable. Evidence for this was provided by the results of five different experimental approaches in the analysis of Springer human erythrocyte stromal extracts. The first was based on the chromatographic properties on sepharose 4B and DEAE-Sephadex where slight differences between the elution patterns and recoveries of the two activities were observed. The second piece of evidence came from the electrophoretic profile of the crude and partially purified extracts on 7.5% polyacryl amide gels under nonreducing conditions. The IH inhibitor activity was shown to cover a fairly large area at the top third of the gel with the LPS-receptor activity being localized in a narrow, single band with a peak of activity near the top of the gel. A third item of evidence was based on the redistribution and separation of the two activities into different phases when the pH during the crude stromal butanol extraction procedure 57

    PAGE 64

    58 shifted from 8.2 to 5.3. The fourth approach, based on the high affinity of the IH inhibitor for the membranes of sheep erythrocytes, demonstrated that the IH inhibitor activity was partially removed leaving the LPSreceptcr activity unchanged by the treatment of the stromal extracts with sheep erythrocytes. Finally, evidence was presented indicating that the binding specificity of sheep erythrocytes for LPS of gram negative bacteria is localized in a lipid moiety of the crude stromal extracts and is free of all detectable IH inhibitor activity. It should be emphasized, however, that these experiments cannot exclude the possibility that both activities may be associated on the same macromolecule with the differences reported here being a consequence of experimental manipulation. That the two activities may be a function of a single macromolecule is certainly a major possibility. Springer et al (16), in assessing the chemical and physical properties of a homogenous preparation of the LPS-receptor, observed that both citraconylation and dissociating polyacryamide gel electrophoresis under standard conditions yielded two fragments, one of which absorbed significantly only at 230 nm. Decitraconylation of the citraconylated fragment restored high LPS-receptor activity to only one of the fragments. These studies are only suggestive and do not permit a decision on whether the activities are on the same molecule however. In contrast, the data obtained from sheep erythrocytes which completely lack IH inhibitor, but which possess LPS-receptor activity, would support the finding that the two macromolecules may be distinctly different. However, the evidence would suggest that the LPS-receptors

    PAGE 65

    59 on sheep cells differ from those observed on human cells, since they are confined to the lipid moiety. Good purification of the LPS-receptor and IH inhibitor activities following DEAE-Sephadex chromatography is indicated by the quantitative data tables. These results would suggest, however, that as a preparative purification step it should be modified to encourage higher yields of the two activities. The observation that sheep erythrocyte membranes possess molecules with receptor specificity for LPS is not surprising for it has long been known that sheep cells could be modified by the presence of LPS of gram negative bacteria and that these modified cells are readily lysed in the presence of homologous antiserum to LPS and complement (44). In contrast, hemolysis was not observed when LPS treated human erythrocytes were treated under the same conditions. This raises the possibility that human erythrocytes, a natural source of the IH inhibitor even when treated with LPS, are extremely resistant to LPS mediated lysis because of the presence of inhibitor molecules. The findings reported here generally agree with those of Phillips et al. (46) indicating that LPS treated sheep erythrocytes can activate the complement system in the presence of natural antibodies to the LPS. In contrast to their results, however, erythrocytes coated with a preparation of LPS extracted by the procedure of Westphal were shown to be capable of activating the complement system in the absence of natural antibodies to LPS. Additionally, fluid phase LPS extracted by the Boivin method (LPS-boivin) was shown to be a more effective activator of the

    PAGE 66

    T 60 complement system than LPS extracted by the Westphal procedure (LPSwestphal) in the presence and absence of natural antibodies to LPS. However, once LPS extracted by the Boivin becomes cell associated, its capacity to activate the complement system in the absence of natural antibodies is greatly dininished. This is significant because LPS activation of the complement system by an antibody independent mechanism requires either an exposed lipid A moiety or polysaccharide core (8,19). This would suggest then that the orientation of LPS-boivin on the membrane may be different from that of LPS-westphal resulting in the masking of active sites necessary for the activation of complement. The fact that different preparations of LPS from the same species, when coated onto the surface of sheep erythrocytes, activated the complement system to different degrees and by different pathways, depending on the presence or absence of natural antibodies to LPS, introduces the possibility that the LPS activation of complement may require substances other than the LPS molecule alone. This especially appears to be true since LPS extracted by the Westphal procedure, which was shown to activate complement in the presence and absence of natural antibody to LPS, is known to contain less protein and lipoproteins than LPS extracted by the Boivin procedure. Placing the IH inhibitor on LPS treated sheep erythrocytes reduced the ability of erythrocytes to consume complement. The fact that the cell is protected even when complement is activated suggests that the IH inhibitor may occupy specific sites on the red cell membrane rendering it resistant to immune lysis, but also leaving it available

    PAGE 67

    61 to partially inhibit immune lysis. This may be due either to the masking of LPS-receptors resulting in less LPS uptake (demonstrated not to be the case here), or the prevention of C3b fixation to the cell membrane thus blocking the initiation of the membrane attack system of serum complement. An analogous site on human erythrocytes is already occupied by the IH inhibitor rendering this cell naturally immune to LPS mediated lysis. The findings presented here have led us to hypothesize that a necessary criterion for resistance to LPS induced complement mediated lysis would be the localization of LPS-receptor and IH inhibitor molecules on the same membrane. This would imply that any red cell devoid of IH inhibitor molecules would be far more susceptible to the cytolytic action of LPS activated complement. LPS of gram negative bacteria are potent activators of the complement system. It is of great clinical interest, therefore, that substances on the surface of erythrocytes which bind LPS are found closely associated with materials capable of inhibiting complement mediated lysis The symptoms of several infectious diseases, such as typhoid fever, have been observed to include very intensive erythrophagocytic activity by macrophages of lymph nodes. The consequence of this observation could have great clinical importance. Erythrocytes coated with LPS, in contact with serum complement, would naturally lead to activation of the complement system followed by increased phagocytosis with minimal cytolysis. This would lead naturally to an amplification of the activation of the complement cascade resulting in either clearance or a heightened inflammatory response. •

    PAGE 68

    REFERENCES 1. Buston, A. 1959. The in-vivo sensitization of avian erythrocytes with Salmonella gallinarum polysaccharide. Immunology. 2.:203. 2. Leive, L., V. K. Shovlin and W. E. Mergenhagen. 1968. Physical, chemical and immunological properties of LPS released from Escherichia coli by ethylene diaminetetra acetate. J. Biol. Chem. 243:6384. 3. Voll, M. J. and L. Leive. 1970. Release of LPS in Escherichia coli resistant to be the permeability increase induced by ethylenediamine tetraacetate. J. Biol. Chem. 245:1 108. 4. Rapin, A. M. C. and H. M. Kalckar. 1971. Microbial Toxins Weinbaum, G., S. Kadis, and S. J. Ajl eds. Academic Press, New York. Vol 4, pp 267-303. 5. Morrison, D. C. and L. Leive. 1975. Fractions of lipopolysaccharide from Escherichia coli QUI: prepared by two extraction procedures. J. of Biol. Chem. 250 : 291 1 6. Osborn, M. J. 1969. Structure and biosynthesis of the bacterial cell wall. Ann. Rev. Biochem. 38:501. 7. Luderitz, 0., V. L. Galanos, M. Nurminen, E. T. Rietschel G. Rosenfelder, M. Simon, and 0. Westphal. 1973. 'Lipid A.' Chemical structure and biological activity. J. Infect. Dis. 128:817. 8. Morrison, D. C. and L. F. Kline. 1977. Activation of the classical and properdin pathways of complement by bacterial lipopolysaccharides (LPS). J. Immunol. 118 :362. 9. Springer, George F. and J. C. Adye. 1975. Endotoxin-binding substances from human leukocytes and platelets. Infect, and Immun. 12:978. 10. Hawiger, J., A. Hawrger, and S. Timmons. 1975. Endotoxin sensitive membrane components of human platelets. Nature. 256:125. n. Kane, M. A., J. E. May, and M. M. Frank. 1973. Interaction 62

    PAGE 69

    63 of the classical and alternate complement pathway with endotoxin 1 ipopolysaccharide effect on platelets and blood coagulation. J. Clin.. Invest. 52^:370. 12. Hill, G. J. and D. W. Weiss. 1964. Relationships between susceptibility of mice to heat-killed Salmonel lae and endotoxin and the the affinity of their red blood cells for killed organisms. In M. Landy and W. Braun, eds. Bacterial Endotoxins Institute of Microbiology, Rutgers, the State University, New Brunswick, N.J. p422-427. 13. Springer, G. F., E. T. Wang, J. H. Nichols, J. M. Shear. 1966. Relations between bacterial 1 ipopolysaccharide structures and those of human cells. An.. N.Y. Sci. 133:566. 14. Springer, G. F., V. Shankar, S. V. Huprikar, and E. Neter. 1970. Specific inhibition of endotoxin coating of red cells by a human erythrocyte membrane component. Infec. Immun. U9S. 15. Springer, G. F., J. C. Adye, A. Bezkorovainy, and J. R. Murthy. 1973. Functional aspects and nature of the 1 ipopolysaccharidereceptor of human erythrocytes. J. Infect. Dis. SI 28 : 5202. 16. Springer, G F., J. C. Adye, A. Bezkorovainy, and B. Jirgensons. 1974. Properties and activity of the 1 ipopolysaccharidereceptor from human erythrocytes. Biochem. 1_3:1379. 17. Loos, M., D. Bitter Suermann, and M. Dierich. 1974. Interaction of the first (CI) and second (C2) and fourth (C4) component of complement with different preparations of bacterial 1 ipopolysaccharide and with lipid A. J. Immunol. 112:935. 18. Herring, W. B. J. C. Herion, R. I. Walker, and J. G. Palmer. 1963. Distribution and clearance of circulating endotoxin. J. Clin. Invest. 42:79. 19. Mergenhagen, S. E., R. Snyderman, and J. K. Phillips. 1973. Activation of complement by endotoxin. J. Infect. Dis. 128:S86. 20. Mueller-Eberhard, H. J. 1975. Complement. An.. Rev. Biochem. 44:697. 21. Gotze, 0. and H. J. Mueller-Eberhard. 1976. The alternative pathway of complement activation. Advances in Immunol. 24:1. 22. McConnell, Ian and P. J. Lachmann. 1977. Complement receptors and cell associated complement components. Immunol. Comm. 6:111.

    PAGE 70

    64 23. Pillemer, L., M. D. Schoenberg, L. Blum, and L. Wurz. 1955. Properdin system and immunity. II. Interaction of the properdin system with polysaccharides. Science. 122:545. 24. Muschel L. H. K. Schmoker, and P. M. Webb. 1964. Anticomplementary action of endotoxin. Proc. Soc. Exp. Biol. Med. 117:639. 25. Dierich, M. P., D. Bitter Suermann, W. Konig, U. Hadding, C. Galanos, and E. J. Rietschel 1973. Analysis of bypass activation of C3 by endotoxin LPS and loss of this potency. Immunology. 24:721. 26. Marcus, R. I., H. S. Shin, and M. M. Mayer. 1971. An alternative complement pathway. C3-cleaving activity not due to C4, 2a on endotoxic 1 ipopolysaccharide after treatment with guinea pig serum; relation to properdin. Proc. Natl. Acad. Sc. 68:1351. 27. Galanos, C. E., T. Rietschel, 0. Liidertiz, and 0. Westphal. 1971. Interaction of 1 ipopolysaccharides and lipid A with complement. Eur. J. Biochem. 1 9:143. 28. Mbller, G. and G. Michael. 1971. Frequency of antigensensitive cells to thymus-independent antigens. Cell. Immunol. 2:309. 29. Cooper, N. R. and D. C. Morrison. 1978. Binding and activation of the first component of human complement by the lipid A region of 1 ipopolysaccharides. J. of Immunol. 120 :1862. 30. Gbtze, 0. and H. J. Mijller-Eberhard. 1971. The C3 activator system: An alternate pathway of complement activation. J. Exp. Med. 134:905. 31. Lay, W. H. and V. Nussenzwerg. 1968. Receptors for complement on leukocytes. J. Exp, Med. 128 :991 32. Snyderman, R. and M. C. Pike. 1975. Interaction of complex polysaccharides with the complement system: Effect of calcium depletion on terminal component consumption. Infect. Immun. 11:273. 33. Fearon, D. T., K. F. Austen, and S. Ruddy. 1973. Formation of a hemolytical ly active cellular intermediate by the interaction between properdin factors B and D and the activated third component of complement. J. Exp. Med. 138:1305.

    PAGE 71

    65 34. Marcus, R. L., H. S. Shin, and M. M. Mayer. 1971. An alternate complement pathway: C3 cleaving activity not due to C4, 2a, on endotoxic 1 ipopolysaccharide after treatment with guinea pig serum; relation to properdin (complement components). Proc. Natl. Acad. Sci. USA. 68:1 351 35. Fine, D. P., S. R. Marney, D. G. Colley, J. S. Sergent, and R. M. Des Prez. 1972. C3 shunt activation in human serum chelated with EGTA. J. Immunol. 109:807. 36. Mayer, M. M. 1972. Mechanism of cytolysis by complement. Proc. Nat. Acad. Sci. USA. 69:2954. 37. Jensen, J. 1967. Anaphylatoxin and its relation to the complement system. Science. 1 55 :1 122. 38. Shin, H. R. Snyderman, E. Friedman, A. Mellors and M. Mayer. 1968. Chemotactic and anaphylatoxic fragment cleaved from the fifth component of guinea pig complement. Science. 162:361. 39. May, J. E., M. A. Kane, and M. M. Frank. 1972. Immune adherence by the alternative complement pathway. Proc. Soc. Exp. Biol. Med. 141 :287. 40. Rother, K. 1972. Leukocyte mobilizing factor: A new biological activity derived from the third component of complement. Eur. J. Immunol. 2^:550. 41. May, J. E. and M. M. Frank. 1972. Complement mediated tissue damage: Contribution of the classical and alternate complement pathways in the Forssmann reaction. J. Immunol. 108:1517. 42. Mlieller-Eberhard, H. J. and I. H. Lepow. 1965. CI esterase effect on activity and physicochemical properties of the fourth component of complement. J. Exp. Med. 121 :819. 43. Ruddy, S. and K. F. Austen. 1971. C3b inactivator of man. II. Fragments of cell bound or fluid phase C3b. J. Immunol. 107:742. 44. Neter, E. 1956. Bacterial hemagglutination and hemolysis. Bacteriol Rev. 20:166. 45. Westphal 0. and J. Kann. 1965. Bacterial lipopolysaccharides. Extraction with phenol water and further applications of the procedure. In Methods in Carbohy• drate Chemistry R. L. Whistler, ed. Academic Press, New York. Vol 5, p83.

    PAGE 72

    66 46. Phillips, J. K., R. Snyderman, and S. E. Mergenhagen. 1972. Activation of complement by endotoxin: A role for yZ globulin, CI, C4 and C2 in the consumption of terminal complement components by endotoxin-coated erythrocytes. J. of Immunol. 109:334. 47. May, J. E., M. A. Kane, and M. M. "Frank. 1972. Host defense against bacterial endotoxemia Contribution of the early and late components of complement to detoxification. J. Immunol. 109:893. 48. Gewurz, H., H. S. Shin, and S. E. Mergenhagen. 1968. Interactions of the complement system with endotoxic lipopolysaccharide: Consumption of each of the six terminal complement components. J. Exp. Med. 123:1049. 49. May, J. E., M. A. Kane, and M. M. Frank. 1972. Immune adherence by the alternate complement pathway. Proc. Soc. Exp. Biol. Med. 141 : 287. 50. Hoffmann, Edward M. 1969. Inhibition of complement by a substance isolated from human erythrocytes extraction from human erythrocyte stromata. Immunochem. 6_:391. 51. Hoffmann, E. M., W. C. Cheng, E. J. Tomeu, and C. M. Renk. 1974. Resistance of sheep erythrocytes to immune lysis by treatment of the cells with a human erythrocyte extract: Studies on the site of inhibition. J. of Immunol. 113:1501 52. Hoffmann, E. M. 1969. Inhibition of complement by a substance isolated from human erythrocytes. II. Studies on the site and mechanism of action. Immunochem. 5^:405. 53. Hoffmann, E. M. and H. M. Etlinger. 1973. Extraction of complement inhibitory factors from the erythrocytes of nonhuman species. J. of Immunol. Ill : 946. 54. Rosse, W. F. and J. V. Dacie. 1966. Immune lysis of normal and PNH red cells. I. Sensitivity of PNH cells. J. Clin. Invest. 45:736. 55. Kabat, E. A. and M. M. Mayer. 1961. Complement and complement fixation. In Experimental Immunochemistry Charles C. Thomas, Springfield, 111. pl49. 56. Springer, G. F., Y. Nagai and H. Tegtmeyer. 1966. Isolation and properties of human blood-group NN and meconium-Vg antigens. Biochem. 5^:3254.

    PAGE 73

    67 57. Maurer, H. R. 1971. Disc Electrophoresis and Related Techniques of Polyacryl amide Ge1-Electrophores1s 2nd revised ed. Walter de Gruyter, Berlin, N.Y. 58. Boivin, A. I. 1933. Extraction of bacterial 0-antigens (endotoxins). Basic Exercises in Immunochemistry A. Nowotny, ed. Springer-'/erl ag, New York. 1969. p25. 59. Nelson, R. A., J. Jensen, I. Gigli, and N. Tamura. 1966. Methods for the separation, purification, and measurement of nine components of hemolytic complement in guinea pig serum. Immunochem. _3-^^' • 60. Ruddy, S. and K. F. Austen. 1967. A stoichiometric assay for the fourth component of complement in whole human serum using EACSP and functionally pure human second component. J. Immunol. 99:1162. 61. Ruddy, S. and K. F. Austen. 1969. C3 inactivator of man. I. Hemolytic measurement by the inactivation of cell bound C3. J. Immunol. 102:533. 62. Boros, T. and H. J. Rapp. 1967. Inmune hemolysis: A simplified method for preparation of EAC4 with guinea pig or with human complement. J. Immunol. 99 : 263.

    PAGE 74

    BIOGRAPHY I, Gloria Jean Jackson, born on October 4, 1939 to Charles and Sarah McKnight, was the second of four children. The only girl, I attended and was graduated May 27, 1957, valedictorian of my class at Boylan Haven, a private school for girls located at that time in Jacksonville, Florida. I later attended Bennett College, majoring in premedicine with a minor in psychology, graduating June of 1961. Lack of funds prevented entry into medical school, therefore, I was employed as a technician at University of Florida for one year. I later entered the University of Kansas at Lawrence, Kansas to pursue a master's degree in Microbiology specializing in microbial physiology. After a year and a half I completed the investigative requirements for the master's degree and was employed as a research technician in the Department of Microbiology, where I served as an assistant to a Microbial Geneticist. In October 1965 I was married to Virgil Lawrence Jackson, following which we relocated to Chicago, Illinois. During my three years in Chicago, I worked as a research technician at the American Medical Association Biomedical Research Institute and later as assistant supervisor of the Microbiology Research Department at the Metropolitan Sanitary District. On May 16, 1959 I gave birth to a son, Jacques Duval 1, following which we relocated to Parsippany, New Jersey.

    PAGE 75

    69 During my stay in New Jersey I was employed at the Warner-Lambert Research Institute where I assisted in studies in oral microbiology and virology. Later relocating to Westen, Connecticut and finally Dover, Massachusetts where my husband was employed as vice president of a printing company. Following a legal separation, I returned with my son, Jacques, to Gainesville to pursue a doctorate at the University of Florida, majoring in Microbiology and specializing in Immunology. I am presently employed at Abbott Labs of North Chicago as a product manager in research and development. In addition to immunology, I enjoy tennis, chess, bridge, skiing, and sailing.

    PAGE 76

    I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. /J] 2 / / '^> Edward M. Hoffman, Chairman, Professor of Microbiology and Cell Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Arnold S. Bleiweis Professor of Microbiology and Cell Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissert^'on for the degree of Doctor of Philosophy. Lester W. Clem Professor of Immunology and Medical Microbiology

    PAGE 77

    I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. :A : ^ Lonnie 0. Ingram Associate Professor of '/^J Microbiology and Cell Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Associate Professor of Biochemistry and Molecular Biology This dissertation was submitted to the Graduate Faculty of the College of Liberal Arts and bciences and the Graduate Council and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December 1978 Dean, College of Liberal Arts and Sciences Dean, Graduate School


    n
    is magnesium dependent and unlike classical pathway activation it is
    inhibited by high concentrations of calcium (34). Under proper experi
    mental conditions these cation requirements make serum chelated with
    either ethylendiamine tetraacetic acid (EDTA) or ethyleneglycol tetra-
    acetic acid (EGTA) useful reagents for distinguishing between the two
    pathways (35). The former, being an effective chelator of both calcium
    and magnesium blocks the activation of the two pathways, while the
    latter, a less effective chelator of magnesium, preferentially blocks
    the classical pathway.
    The last and final stage of complement activation is the same for
    both pathways and is initiated with the cleavage of C5 into two frag
    ments, C5a and C5b by the C5 convertase. The larger C5b fragment then
    reacts sequentially with C6 and C7 to form either a cell bound or fluid
    phase trimolecular complex C5b67. The cell bound complex has the capa
    city to bind C8 and C9. If the cell to which this C5-C9 complex is
    associated is sensitive to complement mediated cytolysis, lysis ensues (36).
    Fluid phase C5b67, C3a, and C5a are potent anaphylatoxins and chem-
    otactins (37,38). Once leukocytes such as polymorphonuclear leukocytes
    (PMN) and macrophages have migrated to the site of complement activa
    tion, phagocytoses is initiated. As previously stated, the phagocytic
    process is enhanced by the fixation of complement components, especially
    C3b, onto the surface of particulate antigens or target tissues which
    promote adherence, thus facilitating ingestion (39). Release of lipo
    somal hydrolases, either as a direct consequence of ingestion or expul
    sion of an indigestible target into the surrounding tissues, results
    in the generation of additional chemotactic factors (40,41). This


    Figure 10. Consumption of complement in E-LPS westphal
    absorbed guinea pig serum by LPS and/or IH
    inhibitor coated sheep erythrocytes. One
    tenth ml containing 10^ sheep erythrocytes
    (E) either untreated or treated with LPS
    (E-LPS), IH inhibitor (E-IH) or LPS and IH
    inhibitor (E-LPS-IH) were incubated with
    0.9 ml of guinea pig serum at 37C for 1 hr.
    Residual complement hemolytic activities
    were assayed and the % of the available com
    plement consumed was calculated.


    64
    23. Pillemer, L., M. D. Schoenberg, L. Blum, and L. Wurz. 1955.
    Properdin system and immunity. II. Interaction of the
    properdin system with polysaccharides. Science. 122:545.
    24. Muschel, L. H., K. Schmoker, and P. M. Webb. 1964. Anti
    complementary action of endotoxin. Proc. Soc. Exp. Biol.
    Med. 117:639.
    25. Dierich, M. P., D. Bitter Suermann, W. Konig, U. Hadding, C.
    Galanos, and E. J. Rietschel. 1973. Analysis of bypass
    activation of C3 by endotoxin LPS and loss of this potency.
    Immunology. 24:721.
    26. Marcus, R. I., H. S. Shin, and M. M. Mayer. 1971. An alter
    native complement pathway. C3-c1eaving activity not due to
    C4, 2a on endotoxic 1ipopolysaccharide after treatment with
    guinea pig serum; relation to properdin. Proc. Natl. Acad.
    Sc. 68:1351 .
    27. Galanos, C. E., T. Rietschel, 0. Liidertiz, and 0. Westphal.
    1971. Interaction of 1ipopolysaccharides and lipid A
    with complement. Eur. J. Biochem. 19:143.
    28. Moller, G. and G. Michael. 1971. Frequency of antigen-
    sensitive cells to thymus-independent antigens. Cell.
    Immunol. j?:309.
    29. Cooper, N. R. and D. C. Morrison. 1978. Binding and activa
    tion of the first component of human complement by the lipid
    A region of 1ipopolysaccharides. J. of Immunol. 120:1862.
    30. Gotze, 0. and H. J. M1ler-Eberhard. 1971. The C3 activator
    system: An alternate pathway of complement activation. J.
    Exp. Med. 134:905.
    31. Lay, W. H. and V. Nussenzwerg. 1968. Receptors for com
    plement on leukocytes. J. Exp. Med. 128:991.
    32. Snyderman, R. and M. C. Pike. 1975. Interaction of complex
    polysaccharides with the complement system: Effect of calcium
    depletion on terminal component consumption. Infect. Immun.
    U:273.
    33. Fearon, D. T., K. F. Austen, and S. Ruddy. 1973. Formation
    of a hemolytically active cellular intermediate by the
    interaction between properdin factors B and D and the
    activated third component of complement. J. Exp. Med.
    138:1305.


    40
    0/
    /o
    COM PLEMENT
    CONSUMED
    E.LPS.IH E.IH.LPS E.IH
    en
    en


    I certify that I have read this study and that in my opinion it
    conforms to acceptable standards of scholarly presentation and is fully
    adequate, in scope and quality, as a dissertation for the degree of
    Doctor of Philosophy.
    u/ X;
    y //
    Lonnie 0. Ingram
    Associate Professor of
    Microbiology and Cell Science
    I certify that I have read this study and that in my opinion it
    conforms to acceptable standards of scholarly presentation and is fully
    adequate, in scope and quality, as a dissertation for the degree of
    Doctor of Philosophy.
    Associate Professor of
    Biochemistry and Molecular Biology
    This dissertation was submitted to the Graduate Faculty of the College of Lib
    eral Arts and Sciences and the Graduate Council and was accepted as partial
    fulfillment of the requirements for the degree of Doctor of Philosophy.
    December 1978
    Dean, College of Liberal Arts and Sciences
    Dean, Graduate School


    24
    the same buffer. The mixture was incubated with shaking for 30 minutes
    at 37C, and then at 0C for either 30 minutes or overnight. The cells
    were washed twice and standardized to the desired concentration before
    use.
    Complement (GPC). Fresh frozen guinea pig complement was obtained
    from Pel Freeze Laboratories (Rogers, AR). The serum was shipped in
    dry ice and was stored at -70C until use. In some studies, aliquots
    of GP serum were absorbed three times at 0C with either untreated or
    LPS treated sheep or human erythrocytes before use.
    Complement Components. Guinea pig Cl and C2 were prepared by
    methods described by Nelson et al. (59) and Ruddy and Austin (60,61).
    Individual lyophilized functionally pure guinea pig complement compo
    nents C3, C5, C6, C7, C8 and C9 were purchased from Cordis Laboratories
    (Miami, FL).
    Complement component intermediates. For IH inhibitor assays and
    complement consumption studies, sheep erythrocytes in the intermediate
    state EAC1, EAC14, and EAC142 were prepared by the methods of Borsos
    and Rapp (62).
    Determination of Tmax of EAC142. The kinetics of the generation
    of EAC142 was determined by the Tmax procedure described by Borsos
    et al. (35).
    IH inhibitor preparation. Crude human erythrocyte stromata ex
    tracts, high in IH inhibitor activity, were isolated by a procedure
    described by Hoffmann (50). The method is outlined in Figure 2.
    Briefly, the essential differences in Hoffmann's preparation of butanol


    Figure 7. A DEAE-Sephadex A-25 chromatography profile of the
    butanol extracted human erythrocyte stromata pre
    pared at pH 5.3 by the method of Springer. 40.0 ml
    of the input material with an optical density at
    220 nm of 9.9 were applied to a 22 X 2.5 cm column.
    4.0 ml fractions were eluted with a linear sodium
    chloride gradient of 0.05-0.75!M, Tris buffer at
    pH 7.0. Optical densities of the fractions at 220
    nm are shown by the closed circles. The closed
    triangles indicate the elution position of the
    fractions capable of inhibiting lysis of sheep
    EAC142 and the shaded bars represent the fractions
    having LPS-receptor activity.


    Figure 6. Correlation between the distribution of LPS-receptor
    and IH inhibitor activities from extracts of human
    erythrocyte stromata when subjected to polyacryla
    mide disc gel electrophoresis. Extracts were ap
    plied to duplicate 7.5% gels and were electro-
    phoresed in a non-reducing Tris-glycine buffer,
    pH 8.6 for 45 minutes at 4C. After electro
    phoresis, one gel was stained for protein with co-
    omassie blue and the other was cut into suitable
    segments which were eluted and analyzed for IH in
    hibitor and LPS-receptor activities. The closed
    triangles indicate the elution position of the IH
    inhibitor and the shaded bars represent the elution
    profile of the LPS-receptor activity.


    DISCUSSION
    The experiments reported here have demonstrated that extracts
    from human erythrocyte membranes possessing LPS-receptor activity ob
    tained by the method of Springer et al. (15) were also capable of inhi
    biting complement mediated lysis. The anticomplementary activity of
    these extracts was demonstrated to share many of the properties of the
    IH inhibitor previously described by Hoffmann (50).
    Data are presented which strongly suggest that the two biologic
    activities are closely associated, but separable. Evidence for this
    was provided by the results of five different experimental approaches
    in the analysis of Springer human erythrocyte stromal extracts. The
    first was based on the chromatographic properties on sepharose 4B and
    DEAE-Sephadex where slight differences between the elution patterns
    and recoveries of the two activities were observed. The second piece
    of evidence came from the electrophoretic profile of the crude and
    partially purified extracts on 7.5% polyacrylamide gels under non
    reducing conditions. The IH inhibitor activity was shown to cover
    a fairly large area at the top third of the gel with the LPS-receptor
    activity being localized in a narrow, single band with a peak of
    activity near the top of the gel. A third item of evidence was based on
    the redistribution and separation of the two activities into different
    phases when the pH during the crude stromal butanol extraction procedure
    57


    60
    complement system than LPS extracted by the Westphal procedure (LPS-
    westphal) in the presence and absence of natural antibodies to LPS.
    However, once LPS extracted by the Boivin becomes cell associated, its
    capacity to activate the complement system in the absence of natural
    antibodies is greatly diminished. This is significant because LPS
    activation of the complement system by an antibody independent mech
    anism requires either an exposed lipid A moiety or polysaccharide core
    (8,19). This would suggest then that the orientation of LPS-boivin on
    the membrane may be different from that of LPS-westphal, resulting in
    the masking of active sites necessary for the activation of complement.
    The fact that different preparations of LPS from the same species,
    when coated onto the surface of sheep erythrocytes, activated the com
    plement system to different degrees and by different pathways, depending
    on the presence or absence of natural antibodies to LPS, introduces
    the possibility that the LPS activation of complement may require sub
    stances other than the LPS molecule alone. This especially appears to
    be true since LPS extracted by the Westphal procedure, which was shown
    to activate complement in the presence and absence of natural antibody
    to LPS, is known to contain less protein and lipoproteins than LPS
    extracted by the Boivin procedure.
    Placing the IH inhibitor on LPS treated sheep erythrocytes
    reduced the ability of erythrocytes to consume complement. The fact
    that the cell is protected even when complement is activated suggests
    that the IH inhibitor may occupy specific sites on the red cell membrane
    rendering it resistant to immune lysis, but also leaving it available


    xml version 1.0 encoding UTF-8
    REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
    INGEST IEID EDOJZU074_6NDN75 INGEST_TIME 2014-12-19T00:19:10Z PACKAGE AA00026422_00001
    AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
    FILES


    Figure 8. Treatment of partial purified extracts of human
    erythrocyte stromata prepared by the method of
    Springer. Either lo" sheep erythrocytes or buf
    fer were mixed with an equal volumes of the ex
    tracted material. Following a 30 minute incu
    bation period at 37C, the cell suspension was
    pelleted and all supernatant fluids were assayed
    for LPS-receptor and IH inhibitor activities.
    The hatched bars represent the sheep erythrocyte
    treated extracts, and the opened bars the buffer
    treated extracts.


    FRACTION NUMBER (4.0ml)
    A 220
    co 4* en
    o
    % Inhibition


    BIOGRAPHY
    I, Gloria Jean Jackson, born on October 4, 1939 to Charles and
    Sarah Mcknight, was the second of four children. The only girl, I
    attended and was graduated May 27, 1957, valedictorian of my class
    at Boylan Haven, a private school for girls located at that time in
    Jacksonville, Florida.
    I later attended Bennett College, majoring in premedicine with a
    minor in psychology, graduating June of 1961. Lack of funds prevented
    entry into medical school, therefore, I was employed as a technician at
    University of Florida for one year. I later entered the University of
    Kansas at Lawrence, Kansas to pursue a master's degree in Microbiology
    specializing in microbial physiology. After a year and a half I com
    pleted the investigative requirements for the master's degree and was
    employed as a research technician in the Department of Microbiology,
    where I served as an assistant to a Microbial Geneticist. In October
    1965 I was married to Virgil Lawrence Jackson, following which we re
    located to Chicago, Illinois.
    During my three years in Chicago, I worked as a research technician
    at the American Medical Association Biomedical Research Institute and
    later as assistant supervisor of the Microbiology Research Department
    at the Metropolitan Sanitary District. On May 16, 1969 I gave birth
    to a son, Jacques Duvall, following which we relocated to Parsippany,
    New Jersey.
    68


    14
    yielded products having different binding affinities for the membranes
    of sheep erythrocytes. Material prepared at an ionic strength of 0.15
    was shown to be capable of binding to sheep erythrocytes and protecting
    them from immune lysis, and has been designated IH inhibitor.
    An extract prepared under the same conditions but at a lower
    ionicity was incapable of binding to sheep erythrocytes but was capable
    of accelerating the decay of the complement component intermediate
    EAC142 to EAC14. This material was designated DAF for decay accelerating
    factor (51).
    Preliminary studies indicate that the IH inhibitor is a large
    molecule with a molecular weight greater than 250,000. It appears to
    contain at least two sugar moieties, glucose and galactose and it is
    about 4% protein by weight (50).
    Existing data suggest that the IH inhibitor probably acts at the
    C3 convertase step. This was suggested by the finding that IH coated
    sheep erythrocytes in the intermediate state EAC142 consumed less C3
    than untreated controls and that the inhibitory effects of IH ceased
    once activated C3 became fixed to the cell bound C3 convertase (51,52).
    Attempts to more fully define the biochemical and biological pro
    perties of these macromolecules have been hampered by the inability
    to obtain them in a highly purified state.
    Erythrocytes of different species differ in their susceptibilities
    as target cells in immune hemolysis, with sheep and chicken erythro
    cytes being far more sensitive than human and guinea pig erythrocytes.
    Differences within the same species have also been observed (53). For


    6
    bacteria (13-15). This material, designated as an LPS-receptor, has
    now been purified to homogeneity and characterized. Springer has re
    ported that the LPS-receptor is a lipoglycoprotein, rich in N-acetyl-
    neuraminic acid (NANA), galactose, hexosamine and contains about 61%
    protein (16). It appears to be a pentameric molecule with a molecular
    weight of about 228,000 daltons. The LPS-receptor functions by direct
    interaction with groups on the LPS molecule which provide an attachment
    site for tissue components (16). Strong evidence has accumulated sug
    gesting that this attachment site is the lipid A moiety of LPS (17).
    This high affinity of the LPS-receptor for endotoxins is quite remark
    able because both macromolecules are highly negatively charged: the
    receptor, because of its high NANA content and LPS because of its
    phosphoric acid radicals.
    Because the immunological specificity of LPS bound to erythro
    cytes remains unchanged, Springer has suggested that the lipid A of
    LPS binds to the specific receptor via clusters of hydrophobic amino
    acids which makeup about 40% of the total peptide content of the
    receptor leaving the polysaccharide available for the reaction with
    antibodies (16). A complete understanding of the orientation of LPS
    on tissues, bound either by specific receptors or by non-specific
    mechanisms, maycome from studies involving the interaction of cell
    bound LPS with serum complement.
    The anti complementary effects of LPS have long been established.
    For some time, evidence seemed to suggest that the single most impor
    tant factor in the development of a noxious response to endotoxins
    was the direct interaction of the lipid A region with biological


    21
    DEAE-Sephadex A25 (Pharmacia Fine Chemicals, Piscataway, N.J.) column.
    After extensive washing of the column with the starting buffer, a
    linear sodium chloride gradient was initiated with 150 ml of 0.05 M pH
    7.0 Tris-HCl buffer and 150 ml of 0.75 M, pH 7.5 Tris-HCl buffer. Three
    milliliter fractions were collected and assayed for the two activities.
    Ether extraction. Sometimes LPS-receptor or binding activity could
    not be localized in the aqueous butanol phase after shift in the pH
    during extraction. The intermediate lipid phase was then further ex
    tracted with equal volumes of ether (Mallanchrodt, St. Louis, M0).
    Equal volumes of the lipid phase, suspended in 0.05 M Tris-Hcl buffer
    at a pH of 7.0 (1:2) and ether were vigorously mixed in a separatory
    funnel for 5-15 minutes. After phase separation the ether was removed
    from the aqueous and syrupy interphase layers by dialysis against phos
    phate buffered saline (pH 7.4) and from the organic phase by evapora
    tion using a stream of nitrogen at 4C. Following evaporation, the
    residue remaining from the organic phase was reconstituted to its
    original volume with PBS and all phases were tested for LPS-receptor
    and IH inhibitor activities.
    Polyacrylamide gel electrophoresis (PAGE). Disc polyacrylamide
    electrophoresis was performed using a modification of the method of
    Maurer (57). Extracts were applied to 7.5% acrylamide gels and were
    electrophoresed in a non-reducing Tris-glycine buffer system, pH 8.6,
    for 45 minutes at 4C. The gels were stained with 0.02% Coomassie blue
    containing 12.5% trichoracetic acid. For some studies, duplicate gels
    were run. One gel was stained as above'with the remaining gel being
    sliced for analysis for LPS-receptor and IH inhibitor activities.


    22
    Lipopolysaccharide (LPS). Lyophilized preparations of Salmonella
    typhimurium LPS extracted by the Boivin (58) and Westphal (45) proce
    dures were purchased from Difco Laboratories (Detroit, MC). Heated
    (100C for 3 hours) and unheated LPS stock solutions (1.0 mg/ml PBS
    at pH 7.4) were stored at -22C until used.
    Antisera. Appropriate dilutions of Salmonella group B 0-antiserum
    obtained from Baltimore Biological Laboratories (Cockeysville, MD)
    were made in 0.01 M EDTA GVB^. Hemagglutination (HA) titers of the sera
    ranged from 64 to 256.
    For hemolytic assays, rabit 19S antibodies to sheep erythrocyte
    stromata were obtained from Cordis Laboratories (Miami, FL). Stock
    solutions at a dilution of 1:100 in PBS (pH 7.4) were maintained until
    use at -22C.
    Treatment of erythrocytes with LPS. Freshly acquired erythrocytes
    from a group 0, Rh positive adult were obtained and used in the coating
    and coating inhibition assays which were carried out as described by
    Springer et al. (14). For speed and economy, screening assays were as
    sessed using a microtiter hemagglutination system. Briefly, the prodedure
    consisted of mixing equal volumes of either human or sheep erythrocytes
    at 2 X 10 cells/ml and dilutions of LPS for 45 minutes with shaking at
    37C. After extensive washing, the smallest amount of LPS which afford
    ed maximal hemagglutination by subsequently added antiserum was determined.
    This dilution, defined as the optimal coating dose, was used in all sub
    sequent hemogglutination-inhibition assays.


    TABLE V
    Recovery of l.PS-receptor and IH Inhibitor Activity of Human
    Erythrocyte Stromata Extracted at pH 5.3 Using the Procedure of Springer
    Preparation
    Quantity
    Total Concentration
    (A220/ml)
    Total Activity
    LPSRa IH5Qb
    Specific
    LPSR
    Activityd
    ih50
    Aqueous butanol
    16.0 ml
    15.0
    0
    32.0
    0
    2.1
    Organic butanol
    23.0 ml
    NDC
    0
    0
    Lipid
    1.2 g
    NDC
    Aqueous ether
    5.0
    6.32
    0
    160
    0
    25.3
    Lipid interphase
    4.5
    ND
    0
    144
    0
    ND
    Organic ether
    12.0
    313.2
    192
    0
    0.6
    0
    aThe smallest amount (dilution) giving maximum inhibition of LPS activity.
    bThe reciprocal of the highest dilution giving 50% inhibition of EAC124 lysis
    cNot determined, due to high levels of particulate matter.
    Calculated by dividing the value for total activity by the value for total concentration.


    42
    A significant difference in the two activities was evident when
    it was serendipitously observed that a shift in the pH from 8.2 to 5.3
    in the butanol extraction step of the crude erythrocyte stromata resulted
    in a preparation consisting of little or no detectable LPS-receptor activ
    ity, but high in IH inhibitor activity. The elution profile of the IH
    inhibitor activity following ion exchange chromatography on DEAE-Sephadex
    of the partially purified material, as shown in Figure 7, was similar
    to that observed for material extracted at pH 8.2. LPS-receptor activity
    assays were not carried out on the remaining butanol phases because these
    materials had been discarded before the impact of these observations were
    realized.
    A shift in the pH to 5.3 of a crude butanol extract of erythrocyte
    stromata obtained at a pH of 8.2 affected neither the LPS-receptor nor
    IH inhibitor activities. This suggested the possibility that the LPS-
    receptor activity of the material extracted at a pH of 5.3 was not de
    stroyed but was probably redistributed into another phase. Stromal ex
    tractions were carried out at pH 5.3 employing smaller volumes of erythro
    cytes stromata, using Springer's procedure, in an attempt to localize
    the LPS-receptor activity. As shown in Table V, some IH inhibitor activ
    ity was observed in the crude aqueous butanol phase when 10.0 ml of
    packed stromata were extracted. There was no detectable LPS-receptor
    in this butanol layer. The lipid phase mixed with 10 volumes of PBS
    was further subjected to an ether extraction resulting in four layers:
    a pellet, an aqueous layer, lipid interphase, and an organic layer.
    All of the LPS-receptor activity was recovered in the organic ether layer


    33
    in Figure 2. Figure 4 is a gel filtration elution profile on sepharose
    4B, of the crude high speed top layer obtained from the crude butanol
    human erythrocyte stromal preparation. Fractions were monitored at 220
    and 280 nm and were assayed for LPS-receptor and IH inhibitor activities
    as previously described. Two peaks were observed with both activities
    eluting in the peak following the void volume. Close examination of
    the sepharose 4B profile indicates that there is a slight displacement
    of the IH inhibitor activity to the left of the LPS-receptor activity.
    This would suggest that perhaps the two activities may be different.
    Further attempts to separate and purify the two activities were
    accomplished using ion exchange chromatography. The sepharose 4B active
    peaks were pooled, dialyzed against the starting Tris-HCl buffer, and
    applied to a DEAE-Sephadex column. Fractionation was accomplished with
    a linear NaCl gradient. A typical chromatogram of the partially puri
    fied material(s) is shown in Figure 5. LPS-receptor and IH inhibitor
    activities eluted in a relatively narrow peak at about 0.3MNaCl, again
    with the IH inhibitor slightly preceding the LPS-receptor activity.
    The recovery of the LPS-receptor and IH inhibitor activities
    following sepharose 4B and DEAE-Sephadex chromatography is presented
    in Table IV. It should be noted that gel filtration on sepharose 4B
    yielded only about onefold increase in the purity of both activities
    with recovery of only 52% of the LPS-receptor activity and 65% of the
    IH inhibitor activity. DEAE-Sephadex was shown to result in as much as
    a 19-fold increase inactivity, but resulted in a recovery of 50%
    of the specific LPS-receptor activity but only one third or 31% of the


    23
    LPS-receptor activity assay. LPS-receptor activity was determined
    by measuring the ability of a material to inhibit LPS fixation to
    erythrocytes. The procedure in the coating HA-inhibition assay differed
    from that in the coating test in that dilutions of LPS binding material
    were added to equal volumes of an optimal coating dose of LPS and in
    cubated with shaking for 30 minutes at 37C. Erythrocytes were added,
    and HA titers determined as previously described. In each assay, a
    control consisting of LPS and erythrocytes but no LPS-receptor material
    followed by the subsequent addition of antiserum was included.
    Isotonic buffer solutions employed in complement assays. The basic
    diluent for most hemolytic assays was the isotonic gelatin veronal
    buffer (6VB) described by Kabat and Mayer (55) which contained 0.00015
    M CaCl^* 0.0005 M MgCl2> and 0.1% gelatin at pH 7.5. In some cases,
    gelatin-veronal without CaCl2 or MgCl2, containing enough isotonic
    ethylenediaminetetra acetate (EDTA, pH 7.4) to bring the final concen
    tration to either 0.01 M or 0.04 M, was employed. These buffers were
    designated as 0.01 M EDTA-GVB and 0.04 M EDTA-GVB respectively. In order
    to achieve maximum sensitivity, hemolytic assays involving individual
    complement components, and IH inhibitor assays were performed using a
    low ionic strength gelatin-veronal prepared by mixing equal volumes of
    5.0% glucose with gelatin-veronal buffer containing twice the standard
    amount of CaCl2 and MgCl2 (DGVB).
    Sensitized sheep erythrocytes (EA). Sheep erythrocytes at a con
    centration of 109 per ml in 0.01 M EDTA-GVB were mixed with an equal
    volume of antibody to sheep stromata at-a final dilution of 1 :500 in


    69
    During my stay in New Jersey I was employed at the Warner-Lambert
    Research Institute where I assisted in studies in oral microbiology
    and virology. Later relocating to Westen, Connecticut and finally
    Dover, Massachusetts where my husband was employed as vice president
    of a printing company.
    Following a legal separation, I returned with my son, Jacques, to
    Gainesville to pursue a doctorate at the University of Florida, major
    ing in Microbiology and specializing in Immunology. I am presently
    employed at Abbott Labs of North Chicago as a product manager in re
    search and development. In addition to immunology, I enjoy tennis,
    chess, bridge, skiing, and sailing.


    26
    after which, three volumes of guinea pig complement diluted 1:25 in
    0.04 M EDTA-GVB were added. The tubes were then incubated for 60 min
    utes at 37C with shaking. At the end of the incubation period, 10
    volumes of ice-cold PBS was added to each reaction mixture. The cells
    were pelleted at 500xg for 10 minutes at 4C and the optical densities
    of the supernatant fluids were determined at a wave length of 414 nm.
    Chelators. Stock solutions of disodium ethylenediametetra acetate
    (EDTA, Fisher Scientific Co., Fair Lawn, NJ) and ethyleneglycolbis
    (beta-amino-ethyl ether) N^,N tetraacetic acid (EGTA, Sigma Chemical
    Co., St. Louis, M0) were prepared as described by Fine et al. (35).
    The stock solutions were stored at 4C and diluted to a final concen
    tration of 200 mM before use. Magnesium EGTA was prepared as described
    by Fine et al. (35).
    Complement Consumption. The ability of erythrocytes treated
    with LPS and/or IH inhibitor to consume complement was determined in
    reaction mixtures containing 0.1 ml of the treated cells (1 X 10^ cells)
    or LPS and 0.9 ml of normal or absorbed guinea pig serum chelated with
    either EGTA or EDTA. The mixtures were incubated with shaking 60
    minutes at 37C. Following the incubation period, the cells were pel
    leted at 500xg at 4C for 10 minutes. The supernatant fluids were
    reconstituted with magnesium and/or calcium and were analyzed for
    residual whole complement activity using a modification of the pro
    cedure as outlined by Kabat and Mayer (55).


    LIST OF FIGURES
    FIGURE PAGE
    1.Probable structure of LPS of Escherichia coli 4
    2. Schematic representation of purification
    procedures of the LPS-receptor and IH inhibitor.... 19
    3. IH inhibitor activities of crude butanol
    extracted erythrocyte membranes 32
    4. A gel filtration profile of the butanol
    extracted human erythrocyte stromata 36
    5. DEAE-Sephadex chromatography of the
    sepharose 4B active fractions from human stromata.. 38
    6. Polyacrylamide disc gel electrophoresis
    of human stromata extracts 41
    7. DEAE-Sephadex chromatography of
    human erythrocyte stromata extracts
    prepared at pH 5.3 45
    8. Treatment of partial purified extracts
    of human erythrocyte stromata with
    sheep erythrocytes 48
    9. Complement consumption in LPS
    treated guinea pig serum 53
    10. Complement consumption of LPS
    and IH inhibitor treated erythrocytes 56
    v


    GLOSSARY OF ABBREVIATIONS
    C complement
    Cl, C2, C3, C4, C5, C6, C7, C8, C91... complement components
    E erythrocyte
    A antibody
    HA hemagglutination
    HAI hemagglutination inhibition
    EDTA ethylenediamine tetracetate
    ^The nomenclature of complement used conforms to that proposed as a re
    sult of a series of World Health Organizations (Immunochemistry, _7:137,
    1970).
    vi


    u
    20 170 190 210 230
    FRACTION NUMBER (4.0ml)
    CO
    CO
    % Inhibition


    TABLE OF CONTENTS
    PAGE
    ACKNOWLEDGEMENTS
    LIST OF TABLES iv
    LIST OF FIGURES v
    GLOSSARY OF ABBREVIATIONS vi
    ABSTRACT vi i
    INTRODUCTION 1
    MATERIALS AND METHODS 17
    RESULTS 27
    DISCUSSION 57
    REFERENCES 62
    BIOGRAPHY 68
    i i i


    66
    46.Phillips, J. K., R. Snyderman, and S. E. Mergenhagen. 1972.
    Activation of complement by endotoxin: A role for yZ glo
    bulin, Cl, C4 and C2 in the consumption of terminal com
    plement components by endotoxin-coated erythrocytes. J. of
    Immunol. 109:334.
    47.May, J. E., M. A. Kane, and M. M. Frank. 1972. Host defense
    against bacterial endotoxemia Contribution of the early
    and late components of complement to detoxification. J.
    Immunol. 109:393.
    48. Gewurz, H., H. S. Shin, and S. E. Mergenhagen. 1968. Inter
    actions of the complement system with endotoxic lipopoly-
    saccharide: Consumption of each of the six terminal com
    plement components. J. Exp. Med. 128:1049.
    49. May, J. E., M. A. Kane, and M. M. Frank. 1972. Immune
    adherence by the alternate complement pathway. Proc. Soc.
    Exp. Biol. Med. 141:287.
    50. Hoffmann, Edward M. 1969. Inhibition of complement by a
    substance isolated from human erythrocytes extraction
    from human erythrocyte stromata. Immunochem. 6^:391.
    51. Hoffmann, E. M., W. C. Cheng, E. J. Tomeu, and C. M. Renk.
    1974. Resistance of sheep erythrocytes to immune lysis by
    treatment of the cells with a human erythrocyte extract:
    Studies on the site of inhibition. J. of Immunol. 113:1501.
    52. Hoffmann, E. M. 1969. Inhibition of complement by a sub
    stance isolated from human erythrocytes. II. Studies on
    the site and mechanism of action. Immunochem. 6;405.
    53. Hoffmann, E. M. and H. M. Etlinger. 1973. Extraction of
    complement inhibitory factors from the erythrocytes of non
    human species. J. of Irrmunol. Ill:946.
    54. Rosse, W. F. and J. V. Dacie. 1966. Immune lysis of normal
    and PNH red cells. I. Sensitivity of PNH cells. J.
    Clin. Invest. 45:736.
    55. Kabat, E. A. and M. M. Mayer. 1961. Complement and com
    plement fixation. In Experimental Immunochemistry.
    Charles C. Thomas, Springfield, Ill. pi49.
    56. Springer, G. F., Y. Nagai, and H. Tegtmeyer. 1966.
    Isolation and properties of human blood-group NN and
    meconium-Vg antigens. Biochem. 5^:3254.


    29
    erythrocytes. Table III, Column A summarizes the results obtained with
    several crude aqueous butanol preparations of erythrocyte stromata.
    These data indicate that the range of LPS-receptor concentrations or
    dilutions needed to yield optimal inhibition of LPS erythrocyte coating
    varied with the source, concentration and condition (solvent) of the
    erythrocyte stromata extraction procedure.
    IH inhibitor activity of crude butanol extracts of human and
    sheep erythrocyte stromata. EAC142 inactivation by a crude erythrocyte
    stromal extract, assayed by the technique described in the section on
    materials and methods is shown in Figure 3. This procedure was used to
    determine the inhibitory potency of most extracts. Color controls for
    the presence of hemoglobin in the higher concentrations of crude prepara
    tions were necessary. The reciprocal of the dilution of a crude butanol
    preparation yielding greater than 50% inhibition of the lysis of EAC142
    by C-EDTA are also shown in Table III. These results clearly indicate
    that LPS-receptor and IH inhibitor activities are contained in signifi
    cant amounts in crude butanol stromal extracts obtained by either Spring
    er's or Hoffmann's procedures. Of interest also, is that the potency
    of the two activities varied to the same extent.
    A comparison of the chromatographic properties of the IH inhibitor
    and LPS-receptor from erythrocyte stromal extracts. The above data sug
    gests that the LPS-receptor and IH inhibitor are either identical or
    closely related molecules, therefore, additional evidence to resolve
    this issue was sought. Human and sheep erythrocyte stromata were sub
    jected to modifications of Springer's purification procedure as outlined


    30
    TABLE III
    A Comparison of LPS-receptor and IH Inhibitor
    Activities of Several Crude Extracts of Erythrocyte Stromata
    Titers
    Crude Stromata
    Extracts
    A
    LPS-receptora
    IH
    B .
    inhibitorD
    Human cells extracted
    at pH 8.2 (Springer)
    Crude aqueous butanol phase
    40
    40
    Crude aqueous phase
    high speedc top layer
    320
    320
    high speed syrupy interphase
    80
    40
    Human cells extracted
    at pH 7.5 (Hoffmann)
    Crude aqueous butanol phase
    320
    320
    Sheep cells extracted
    at pH 8.2 (Springer)
    Crude aqueous butanol phase
    0
    0
    Crude aqueous ether phase
    0
    0
    Crude ether interphase
    8
    0
    Crude ether organic phase
    0
    0
    aThe smallest amount (dilution) giving complete inhibition of LPS coating.
    ^The reciprocal of the greatest dilution giving 50% inhibition of EAC142
    lysis.
    c
    High speed extracts were obtained by the centrifugation of the crude
    aqueous butanol phase, 151,OOOxy for 2 hours.


    58
    shifted from 8.2 to 5.3. The fourth approach, based on the high affinity
    of the IH inhibitor for the membranes of sheep erythrocytes, demonstrated
    that the IH inhibitor activity was partially removed leaving the LPS-
    receptcr activity unchanged by the treatment of the stromal extracts
    with sheep erythrocytes. Finally, evidence was presented indicating
    that the binding specificity of sheep erythrocytes for LPS of gram nega
    tive bacteria is localized in a lipid moiety of the crude stromal ex
    tracts and is free of all detectable IH inhibitor activity.
    It should be emphasized, however, that these experiments cannot
    exclude the possibility that both activities may be associated on the
    same macromolecule with the differences reported here being a consequence
    of experimental manipulation. That the two activities may be a function
    of a single macromolecule is certainly a major possibility. Springeret al.
    (16), in assessing the chemical and physical properties of a homogenous
    preparation of the LPS-receptor, observed that both citraconylation and
    dissociating polyacryamide gel electrophoresis under standard conditions
    yielded two fragments, one of which absorbed significantly only at 230
    nm. Decitraconylation of the citraconylated fragment restored high
    LPS-receptor activity to only one of the fragments. These studies are
    only suggestive and do not permit a decision on whether the activities
    are on the same molecule however.
    In contrast, the data obtained from sheep erythrocytes which
    completely lack IH inhibitor, but which possess LPS-receptor activity,
    would support the finding that the two macromolecules may be distinctly
    different. However, the evidence would suggest that the LPS-receptors