Immunoprophylaxis against bacterial sepsis

Sepsis 1999;3:225–234 © 2000 Kluwer Academic Publishers. Manufactured in The Netherlands.

Immunoprophylax Op al et al. is

Immunoprophylaxis Against Bacterial Sepsis Steven M. Opal,1 Alan S. Cross,2 Apurba K. Bhattacharjee,3 Kumar Visvanathan4 and John B. Zabriskie4 1

Brown Univ. School of Medicine, Providence RI; Univ. of Maryland School of Medicine, Baltimore MD; 3Walter Reed Army Institute of Research, Washington D.C.; and 4Rockefeller Institute, New York, NY 2

Abstract. Sepsis can be viewed as toxigenic illness resulting from the release of excess quantities of microbial-derived in_ammatory mediators into the systemic circulation. Principal among these microbial mediators is bacterial endotoxin. Endotoxin is an essential component of the outer membrane of gram-negative bacteria. Humans are exquisitely susceptible to endotoxin-induced systemic in_ammatory reactions that may prove to be rapidly fatal. Many gram-positive bacteria express speci~c exotoxins that have the capacity to function as superantigens. These superantigens mediate deleterious systemic immune responses that may result in a toxic shock-like syndrome. These microbial toxins provide potential targets for vaccine development as a means of immunoprophylaxis and/or immunotherapy against septic shock. It may be possible to immunize susceptible populations at risk for sepsis and generate protective antibodies toward the common, injurious bacterial toxins. The feasibility of immune protection against microbial mediators of sepsis is the focus of this brief review. Key words. vaccines, endotoxin, lipopolysaccharide, sepsis, superantigens, septic shock

Introduction The development of protective vaccines against serious bacterial infections has been one of the major goals of medical research since the germ theory of disease was ~rst developed over 100 years ago. While some vaccines against bacterial pathogens and/or their toxins have been remarkably successful (e.g., the conjugate H. in_uenzae type B vaccine, the tetanus toxoid vaccine) [1], vaccines that protect patients from the adverse sequelae of septic shock have not yet been developed. The challenges presented in an attempt to develop a successful anti-sepsis vaccine are numerous but ~nite and increasingly understood. A limited number of microbial mediators have the capacity to initiate a generalized septic response in clinically relevant concentrations. It may be possible to develop an immunoprophylactic strategy to limit the deleterious effects the microbial mediators implicated in the molecular pathogenesis of sepsis [2,3]. The current status of some of the active immunization approaches in the

prevention and treatment of bacterial sepsis will be reviewed in this paper.

The rationale for immunoprophylaxis against sepsis Despite the fact that sepsis is an unpredictable and heterogeneous clinical syndrome, an active immunization strategy continues to be an attractive (preventative) and therapeutic approach. The ever-increasing incidence of sepsis [4], progressive development of antimicrobial resistance among common bacterial pathogens [5], and the limited therapeutic options currently available to treat sepsis [3,6] provide the stimulus for the continued anti-sepsis vaccine efforts. The intrinsic complexities of human sepsis make it inevitable that effective immunoprophylaxis will necessitate a multi-component vaccine strategy. A wide variety of invasive microbial pathogens and potent microbial mediators are known to cause sepsis. Microbial mediators that have been identi~ed as likely inducers of human bacterial sepsis are all potential targets for vaccine development. Candidate vaccine targets for an active immunization program against sepsis include endotoxin [3,6–18], bacterial superantigens [19–22], peptidoglycan [23–25], lipoteichoic acid [24,26], bacterial DNA [27,28] and other microbial mediators [3,6,8,24]. A combination strategy consisting of an active immunization program supplemented with passive immunotheraphy with high titer protective antibodies against microbial mediators may ultimately prove to be the most ef~cacious approach in septic patients. There are several potential advantages of a vaccine approach against bacterial sepsis. First, this vaccine strategy is directed against the microbial mediators of sepsis and not the host-derived endogenous in_ammatory mediators. This is a safer approach than immunoprophylactic measures directed against host derived immune responses. There is ample evidence that the

Address correspondence to: Steven M. Opal, M.D., Infectious Disease Division, Memorial Hospital of Rhode Island, 111 Brewster Street, Pawtucket, RI 02860 USA. Tel.: 401-729-2545; Fax: 401-729-2795; E-mail: [email protected] 225

226

Opal et al.

administration of anti-in_ammatory agents may prove to be detrimental in the presence of ongoing systemic infection [2,14]. Inhibitors directed against microbial toxins and microbial virulence factors will not disrupt the host response to invasive microbial pathogens as may occur with inhibitors of the host immune response (i.e., anti-IL-1 and anti-TNF strategies). Second, the generation of protective antibodies present in the circulation prior to the onset of sepsis would be most likely to bene~t patients in the early phases of the systemic in_ammatory response. The pathogenetic mechanisms responsible for septic shock are activated within hours and sometimes minutes after the precipitating microbial challenge. It is often dif~cult to recognize patients at this early phase of sepsis when corrective actions are most likely to be effective. Prophylactic vaccines provide an opportunity to protect patients at this critical time when they are most vulnerable and yet most amenable to treatment intervention. If immunoprophylaxis can disrupt the septic process before irreversible damage to vital organs has occurred, then this preventative approach may succeed where other strategies have failed [3]. Third, vaccine approaches may complement other immunotherapeutic measures and standard treatment regimens that are routinely instituted in the management of the severely septic patient. Surgical interventions to drain abscesses or relieve obstruction may cause transient worsening of the systemic release of microbial elements as the infected tissues are manipulated. In addition, bactericidal antimicrobial agents may facilitate the release of bacterial endotoxin and other microbial constituents that induce a generalized host in_ammatory response [29]. The availability of vaccine-induced blocking antibodies in the circulation at the time of the initial medical or surgical intervention should limit the potential deleterious effects associated with the transient release of microbial mediators.

The limitations of the active vaccine approach to sepsis The primary limitation to vaccine approaches to sepsis prevention is need to provide protective antibody levels against most if not all the major microbial components implicated in the pathogenesis of sepsis. A multitude of bacterial pathogens can precipitate sepsis and combinations of microbial toxins and virulence factors contribute in concert to induce sepsis [23–27]. Despite the diffuse array of potentially pathogenic bacterial organisms, a small number of microbial mediators result in systemic immune activation. It may be possible to immunize patients against the two of the principal toxins responsible for the generation of gram-negative sepsis and staphylococcal or streptococcal toxic shock. These are bacterial endotoxin and superantigen peptides respectively. An additional drawback to preventive vaccine ap-

proaches against sepsis is uncertainly of duration of immune protection and the level of antibody response necessary to protect against sepsis. Experiments in susceptible animal populations are beginning to address this question in the laboratory but the real answers to these questions will require careful, long term, clinical investigations in selected patient populations [3]. Another potential limitation to immunprophylaxis is the requirement to immunize patients who are traditionally viewed to be poor vaccine responders. These would include immunocompromised patients (e.g., bone marrow transplants, neutropenic cancer patients, and asplenic patients) and the multi-trauma patient. It may be dif~cult to immunize a severe trauma patient in time to prevent late nosocomial infections that continues to complicate the recovery of trauma patients [15]. A ~nal challenge to preventive vaccine development is to de~ne the patient population that is most likely to bene~t from an antisepsis vaccine. This requires a population with an expected incidence of sepsis in the relatively near future in which to test the ef~cacy of the vaccine. An unpredictable, low frequency, sporadic event such as sepsis will make it dif~cult to test this preventive vaccine approach. The risk/bene~t estimates of vaccine ef~cacy mandates that the vaccine will need to carry little or no risk to the vaccine recipient. It would be dif~cult to justify anything other than a exceedingly safe vaccine designed against a low frequency event (sepsis) that may occur some time in the future. An inconvenient and/or poorly tolerated vaccine with frequent side effects will clearly not be clinically acceptable.

The anti-endotoxin vaccine approach Most of the pathophysiologic consequences of gramnegative sepsis can be reproduced by administration of puri~ed LPS or cell wall fragments for the outer membrane of non-viable gram-negative bacteria [11,14,18]. The widely held view that endotoxin is the principal mediator of gram-negative bacterial sepsis in humans remains a highly plausible but as yet unproven hypothesis [30]. It will be essential to ultimately demonstrate that a speci~c anti-endotoxin agent provides a signi~cant survival bene~t in patients with gram-negative bacterial sepsis to prove that endotoxin is a key mediator in human sepsis. There are a number of agents currently under development that should be able to conclusively demonstrate the central role of endotoxin in the pathophysiology of human sepsis [14]. There has been an explosion of new information over the last several years regarding the interactions of LPS with proteins and immune effector cells [18,30–33]. The recent discovery that the Toll-like receptor family functions as a transmembrane activator for endotoxin signaling opens up a whole new area of potential LPS modifying agents that may be utilized in the management of human sepsis [33]. Blocking the

Immunoprophylaxis

septic process at the bacterial toxin level is appealing because it is a precipitating upstream event, so that the deleterious secondary host-derived in_ammatory networks of sepsis will not be activated. A variety of anti-endotoxin strategies have been proposed. A vaccine strategy is only one of a number of novel anti-sepsis agents under consideration at the present time [14].

The Structure and Function of Endotoxin The biochemistry of bacterial endotoxin reveals that it is a complex, polar macromolecule consisting of three basic components: (1) a highly conserved lipid A structure that is principally responsible for its endotoxic properties; (2) a relatively conserved yet variable core oligosaccharide region covalently linked to lipid A; and (3) repeating units of distinct polysaccharide molecules (the O side chain). The lipid A portion of the molecule is a highly conserved structure that consists of a b1-6 linked diglucosamine backbone structure that is diphosphorylated and has six acyl groups linked by ester or amide bonds. Most pathogenic gram-negative organisms have 12–14 carbon atom fatty acids (lauric or myristic acid) linked to the diglucosamine nucleus [13,18].

The biochemistry and immunology of lipid A Lipid A biosynthesis is a highly complex synthetic process by gram-negative bacteria. Lipid A is an essential component to outer membrane structure and physiology. Inhibition of lipid A biosynthesis is lethal to gram-negative bacteria [18]. Lipid A is an amphophilic molecule that forms supramolecular aggregates (micells) in aqueous solution. Endotoxin aggregates may be taken up by immune cells and internalized without cellular activation [18]. Recent evidence suggests that a unique 3-dimensional “endotoxic conformation” needs to be maintained in order to express endotoxin activity of the lipid A component of bacterial endotoxin [13,18]. Monomeric, diphosphorylated lipid A presented in a hexagonal three-dimensional structure is the most biologically active. Monophosphorylated or membrane bound lipid A is often found in a lamellar three-dimensional structure that is intrinsically less biologically active. While lipid A is the toxic moiety of bacterial endotoxin, it is immunorecessive with respect to the antibody responses to bacterial endotoxin. Lipid A is located deep into the outer membrane of bacteria and is not readily available to circulating immunoglobulins. This is a major limitation towards vaccine development against this common and highly conserved toxic component of LPS [3,9,10,30]. Anti-lipid monoclonal antibodies (HA-1A and E5) have been studied in considerable detail recently [34–37]. Neither antibody bound to endotoxin with high

227

af~nity and neither antibody was convincingly effective in large, multi-center phase 3 trials. Although HA-1A seems to offer some protection against lethality in a recent report of meningococcal sepsis in children (18% mortality [HA-1A] vs. 28% [placebo]), the differences were not statistically signi~cant [38]. It was observed that one antibody [HA-1A] could bind to autoantigens on human red blood cells and B cells [39] and actually worsened the mortality rate in a canine sepsis model [40]. Immunoprophylactic strategies against lipid A itself may not be feasible and is not the target of current experimental vaccine efforts.

The immunology and biochemistry of the core oligosaccharide Lipid A is covalently linked to an oligosaccharide core structure that consists of both inner and outer core components. This oligosaccharide component combined with lipid A is known as the core glycolipid structure [18]. The oligosaccharide epitopes found in this region of the LPS molecule are now the major focus of vaccine development. This LPS region can be rendered immunogenic and contains highly conserved regions that are common among many gram-negative bacterial pathogens [3,30,41]. The inner core region consists of 2-keto-3-deoxyoctonic acid (KDO) linked to a small series of unique heptose sugars. The inner core is attached a more heterogeneous series of hexose sugars (the outer core structure). There are ~ve different core types among E. coli strains and one major core type in Salmonella species [18]. Monoclonal antibodies have been de~ned that preferentially bind to each distinct core types (chemotype) among common gram-negative bacteria [9–11,18]. Since the initial discovery of common antigenic components among the core LPS structure of different gram-negative bacterial species, it was assumed that antibodies to this common glycolipid structure would protect against endotoxin-mediated pathophysiologic events from a variety of different gram-negative pathogens [42–46]. This seemingly straightforward and readily testable hypothesis has proven to be exceedingly dif~cult to either con~rm or refute [3,9,30,47]. Evidence that such antibodies can, in fact, prevent lethal septic shock and other deleterious consequences of endotoxin is found in a number of laboratory and clinical studies. A large literature review of animal studies using anticore glycolipid antibodies as a preventative measure against endotoxin-mediated lethality has been published [9]. Elevated serum antibody titers against the core glycolipid of endotoxin has been associated with a lower incidence of mortality from P. aeruginosa sepsis [48], and morbidity following major surgery [49,50]. Despite these ~ndings, the protective ef~cacy of core glycolipid antibodies as a prevention or treatment of sepsis remains unproven [30]. Earlier clinical studies with polyclonal antisera directed against the common core structures of endotoxin have been plagued with

228

Opal et al.

inconsistent data and continued controversy [30,51,52]. Although uncertainty remains in the interpretation of the previous studies, monoclonal antibodies have recently been developed that convincingly cross-react with the core LPS structures of multiple different species and genera of gram-negative bacteria. These monoclonal antibodies support the hypothesis that core glycolipid structures among gram-negative bacteria may serve as a potential vaccine to generate a cross protective antibody response [14,18,41]. The best-studied core glycolipid structure is found in the galactose epimerase (gal E) mutant of E. coli 0111:B4. This mutant, known as E. coli J5, expresses its inner core oligosaccharide structure on its cell surface as the absence of galactose prevents completion of the outer core structure and binding to the O-speci~c side chains of complete LPS. The inner core region is now surface exposed and rendered immunogenic [9,17,51]. This mutant expresses a core glycolipid structure that is functionally Rc mutant form of rough LPS. The core structure consists only of three unusual heptose molecules and a terminal a 1-7 linked D-glucosamine and a a 1-3 linked glucose moiety on the second heptose molecule [16,18]. Another well-described oligosaccharide core structure is the Re mutant of Salmonella minnesota [45,46]. This molecule consists only of lipid A and KDO. Unfortunately, most of the antibodies that react to this core glycolipid structure are directed against the KDO molecule itself [18]. These KDO epitopes are sterically inhibited from antibody binding by smooth LPS structures that contain complete core structure and O speci~c side chains. Since most invasive gram-negative express smooth LPS, such antibodies will not bind to pathogenic gram-negative bacteria and predictably would not be effective in actual septic states [18,30]. Interestingly, monoclonal antibodies have been de~ned (e.g., mAb SDZ 219-800) that will bind to the core oligosaccharide structure of bacterial endotoxin of multiple species and genera of gram-negative bacteria [41]. This IgG antibody is able to bind to smooth LPS with O speci~c side chains in place. This indicates that at least some anti-core antibodies exist that are not subject to steric inhibition by the O side chains of LPS. These antibodies bind to conserved epitopes found in the core glycolipid of natural LPS structures found in most pathogenic gram-negative bacteria. This particular antibody binds to the lateral segment of the core region at the junction of the inner and outer core. The minimal structure that this antibody will bind is the core structure presented by E. coli J5 (Rc) [18,41]. Additional support for the potential ef~cacy of the protective anti-core glycolipid hypothesis comes from a recent series of studies using puri~ed LPS from the rough mutant bacteria, E. coli J5. Our group has been able to prepare af~nity column puri~ed, J5 LPS speci~c IgG from lapine antisera following immunization with E. coli J5. This was accomplished using a solid phase matrix to which E. coli J5 LPS had been coupled. The IgG prepa-

ration protected neutropenic rats from an otherwise lethal systemic infectious challenge with P. aeruginosa [53]. It appears that there are critical conformational epitopes at the core region of the glycolipid structure of LPS that may cross react to other LPS serotypes of other gram-negative bacteria. The major challenge in vaccine development is to present these common core glycolipid epitopes in a manner that facilitates high titer protective anti-core antibodies [3,11,16,54]. Since immunoprophylaxis is intended for asymptomatic patients who may be at risk for sepsis in the future, an essential prerequisite for any candidate vaccine is that it be safe and well tolerated. It is mandatory to remove the reactogenicity of the vaccine antigens while maintaining their immunogenicity. This requires that the endotoxic properties of bacterial LPS be detoxi~ed. Severe vaccine reactions related to the endotoxin components of the vaccine would be clearly unacceptable as a preventative vaccine approach. To accomplish this goal, the endotoxin molecule has been de-O-acylated by mild alkali hydrolysis to cleave off ester linked C-12 and C-14 fatty acids. This detoxi~cation step removes the endotoxicity of LPS preserves the critical cross protective epitopes of the core region of the molecule. This detoxi~ed LPS structure was then complexed with the outer membrane protein (OMP) of group B Neisseria meningitidis. This step was taken to increase the immune response to the essential conformational epitopes of the core region of detoxi~ed LPS [3,16]. Our group had been previously demonstrated that a vaccine consisting of detoxi~ed J5 LPS alone, or conjugated with a protein carrier elicited only a 2-10 fold rise in anti-core glycolipid antibody responses. Moreover the vaccine failed to provide signi~cant protection in experimental sepsis models [3]. Linkage of the detoxi~ed J5 LPS with the OMP of group B N. meningitidis resulted in a 100-1000-fold increase in antibody response to the core glycolipid structure [54]. This noncovalent complex with the OMP and detoxi~ed J5 LPS was highly soluble and remarkably stable in aqueous solutions. The complex expresses a critical conformational epitope of the core oligosaccharide of J5 LPS that remains cryptic and unavailable for antibody binding in covalently linked protein vaccine constructs. The OMP/de-acylated J5 LPS vaccine induces the formation of anti-core glycolipid antibodies that cross react with a wide variety of pathogenic gram-negative bacteria, and the vaccine provides protection from lethal gram-negative sepsis in animal models [17]. This vaccine predictably fails to induce antibodies that can recognize cell wall antigens found on gram-positive bacterial pathogens [53]. The mechanism by which anti-core glycolipid antibodies protect the host has not been clearly de~ned. Such antibodies fail to promote opsonophagocytosis in vitro and have limited ability to block endotoxin activity in bioassays such as the Limulus amebocyte lysate assay. The anti-core antibodies may promote clearance of endotoxin but the precise molecu-

Immunoprophylaxis

lar explanation for its bene~cial action needs further clari~cation. This candidate vaccine will soon undergo phase I clinical testing in human volunteers.

The immunology and biochemistry of the polysaccharide portion of LPS The O speci~c polysaccharide antigens of bacterial endotoxin consists of repeating units of distinct high molecular weight polymers formed by multiple repeating oligosaccharide structures. These are serotype-speci~c polysaccharide structures that de~ne the serotype of the somatic O antigens of gram-negative bacteria. Antibodies to these polysaccharide structures generate highly protective immune response to subsequent challenge in experimental animals [9,55]. Unfortunately, these are serotype-speci~c antibody responses and no cross protection is generated against other serotypes. Since there are several hundred serotypes of pathogenic gram-negative bacteria, this has limited the use of polysaccharide antigens as a potential vaccine strategy. Nonetheless, careful analysis has demonstrated that a relatively limited number of serotypes are primarily responsible for gram-negative bacteremia in humans [7,14,15]. Thus, an attempt has been made to develop a polyclonal anti-O-speci~c antibody approach to bacterial sepsis. These O speci~c antibodies are opsonophagocytic, but not anti-endotoxic. They promote rapid phagocytosis and clearance of gramnegative bacterial pathogens but do not intrinsically inhibit the ability of endotoxin to activate host-derived immune mediators [55]. A hyperimmune serum consisting of the 24 most common capsular polysaccharides in clinical isolates of Klebsiella, eight O serotypes of Pseudomonas aeruginosa and 12 O serotypes of E. coli has been tested in a large multicenter clinical trial [7]. The study failed to demonstrate an overall improvement in survival from bacterial sepsis, but it did demonstrate an apparent reduction in the incidence of serotype-speci~c gramnegative infections. It was found that a population of severe trauma patients could be successfully immunized with a vaccine consisting of 24 capsular antigens of Klebsiella and 8 O polysaccharide antigens of Pseudomonas aeruginosa. It was determined that vaccine-induced serum antibody titers were as elevated in trauma patients as antibody levels measured in normal volunteer vaccine recipients [15]. While antigen processing by macrophages may be brie_y attenuated by severe trauma, the Th2-type cytokine responses induced by trauma (i.e., IL-4, IL-10, and IL-13) actually promote a humoral antibody response [15]. Trauma patients remain viable candidates for active immunization programs with an anti-endotoxin vaccine. The sepsis-related mortality observed in severe trauma often occurs several weeks after the initial injury [2,3,6]. There may be suf~cient time to initiate an active immunization strategy against endotoxin in trauma patients.

229

Vaccine Approaches Directed Against Bacterial Superantigens The streptococcal pyrogenic exotoxins (SPE’s) and the staphylococcal enterotoxins (SE’s) belong to a family of structurally similar protein toxins that share some unique immunologic properties [56,57]. They have the capacity to simultaneously stimulate large numbers of CD4, CD8 and cd T cells by a superantigen-mediated process. These toxins possess the capacity to bind to a speci~c region in the variable part selected b chains (Vb of the T cell receptor (TCR) adjacent to the usual epitope recognition site. They also bind to the lateral surface of the class II major histocompatibility complex (MHC) of antigen presenting cells. This binding site is adjacent to the epitope groove that makes up the normal antigen-presenting motif of the MHC class II molecule. This precipitates the aberrant activation and proliferation of an unusually large subset of T-cells [57] (see Figure 1). The crosslinking of TCR and MHC class II molecules by superantigens result in the concomitant release of pro-in_ammatory cytokines from both the T lymphocyte and monocyte populations. CD 4 ⫹ T lymphocytes stimulated by bacterial superantigens produce the Th1 cytokines TNF, IL-2 and IFN-c along with the anti-in_ammatory cytokine IL-10 [58,59] while monocytes generate a wide range of in_ammatory cytokines including: TNF a, IL-1, IL-6, IL-12 and possibly IL-18 in the presence of activated T cells [57,60]. Costimulatory molecules that participate in conventional immune responses play a signi~cant role in the response of immune cells to superantigens as well. The co-stimulatory T-cell antigen, CD28, and its corresponding ligand on MHC-II bearing cells, B7, contribute to superantigen mitogenicity [61,62]. Other costimulatory molecules such as LFA-1/ICAM-1 and VLA-4/VCAM-1 also contribute to the activation of immune cells by superantigens [63]. These immunostimulatory activities of superantigens are central to their pathophysiologic effects in humans. This excessive activation of the immune system may culminate in such severe consequences that a toxic shock-like state results [56,57,60]. The systemic immune response to superantigens is similar in many respects to that observed by bacterial endotoxin from invasive Gram-negative bacterial organisms. In fact, LPS and superantigens can work synergistically to produce lethal toxic shock [57,60,64] (Figure 1).

Structure-function relationships of the superantigens Staphylococcal exotoxins that function as superantigens cause a variety of syndromes in humans. They were ~rst recognized as a cause of staphylococcal food poisoning [57,58]. It was subsequently shown that the same toxins induced proin_ammatory cytokines and produced toxic shock-like syndromes [56,60]. At least

230

Opal et al.

Fig. 1. The synergistic actions of bacterial endotoxin and superantigens in the pathogenesis of sepsis. Ab—antibody; LBP—lipopolysaccharide binding protein; LPS—lipopolysaccharide; TLR—Toll-like receptor; MU—macrophage; ICE—interleukin1b converting enzyme; TNF—tumor necrosis factor; IL—interleukin; IFN—interferon; NK—natural killer cell; TCR—T cell receptor; SAg—superantigen; MHC—major histocompatibility complex

seven staphylococcal enterotoxins have been described thus far [57,65]. The streptococcal pyrogenic exotoxins (SPE’s) are responsible for the pathogenesis of scarlet fever and streptococcal toxic shock-like syndrome (i.e. toxic strep syndrome) [57,58]. The sequences the SPE’s have many structural similarities to the staphylococcal enterotoxins [56,66]. Toxic shock syndrome toxin (TSST-1) from S. aureus is a potent superantigen; however, amino acid sequences of this toxin differ signi~cantly from SE’s and SPE’s [67]. Despite the differences in primary amino acid sequence, the overall topology of TSST-1 and the SE/SPE family of exotoxins is similar [19,20]. All the bacterial superantigens are soluble proteins of approximately 230 amino acids and have a central disul~de loop with the single exception of TSST-1. TSST-1 has only 194 amino acids and lacks a disul~de bridge [68]. Mutational analysis reveals that mutations at various positions throughout the SPEA and SEB molecules were suf~cient to inactivate biological activity [69,70]. Mutated protein toxins have different effects, indicating that functional activities are not limited to any one region of the toxin [71]. The basic functional tertiary structure must be maintained. Chemical modi~cations of highly conserved histidine residues inactivated biologic activity [72].

The central disul~de loop is required for mitogenic activity of staphylococcal Enterotoxins SEA and SEB. Reduction of the disul~de loop inactivated T cell stimulatory activity, but did not effect MHC-II binding and stimulation of monocytes [22,73]. Peptide cleavages within the loop had no effect on T-cell mitogenicity; however, cleavage of conserved sequences outside the loop of SEA resulted in loss of mitogenic activity [73]. Residues determining TCR Vb speci~city appear to be located within the carboxyl-terminus of the SE/SPE toxins [22,74], while residues critical for MHC-II binding are located in the amino-terminal region, and the central portion of the molecule near the disul~de loop [22,68]. The disul~de loop and adjacent highly conserved sequences contribute to the structural integrity of the toxins, and serve to bring the TCR and MHC binding regions in functional proximity to each other [20,68]. The variable sequences in the TCR binding region and the MHC-II binding regions account for different speci~city for speci~c Vb molecules and MHCII types for each unique superantigen [76].

The immune response to superantigens Two distinct regions of the SE/SPE toxins have been identi~ed which share highly conserved amino acid homology. The ~rst consensus region begins immediately after the carboxyl side of the disul~de loop. Sequence homology in this region has been identi~ed in all of the

Immunoprophylaxis

staphylococcal enterotoxins and streptococcal pyrogenic exotoxins, but not in TSST-1 [77]. The second consensus region sequence (K-x(2)(LIV)-x(4)-(LIV)-D-x(3)-R-x(2)-L-x(5)-(LIV)-Y) had been identi~ed in all of the staphylococcal enterotoxins, streptococcal pyrogenic exotoxins, and TSST-1. The two conserved regions are separated by roughly thirty residues. The high degree of sequence homology in these regions contributes to the immunologic cross reactivity shared by many members of the SE/SPE family [78,79]. These highly conserved may serve as a substrate for the development of therapeutic measures against these superantigens [77,80]. Recent laboratory investigations suggest that speci~c immunoglobulins directed against the conserved regions of the SE/SPE toxins might prove an effective therapeutic approach [77]. Peptides constructed of the combined consensus regions of the toxins have been used to raise antibodies in mice. These antibodies inhibited blastogenesis of human mononuclear cells in response to various superantigens. Passive protection of immune serum in rabbits to SEB and SPEA challenge was also demonstrated. There is considerable experimental evidence to support the use of intravenous immunoglobulin G (IVIG) in superantigenic shock [81–83]. The deleterious cytokine responses stimulated by superantigens can be down regulated by the use of IVIG. Human plasma from patients treated with IVIG can inactivate the mitogenic effects of SE/SPE’s [84]. There are anecdotal reports of the bene~cial effects of this therapy with human patients. A recent comparative clinical trial in North America suggests that IVIG may be a valuable adjunctive therapy in streptococcal toxic shock syndrome [85]. The practical value of IVIG therapy in human sepsis and superantigen-induced toxic shock patients is still debated [86]. Well controlled, doubleblind, multicenter clinical studies are needed to fully evaluate the ef~cacy of this therapy. The use of peptide derived vaccines to develop antitoxin antibodies against critical epitopes found within superantigen structures is the focus of current research efforts. Whether such a strategy will prove to be bene~cial in the clinical setting remains to be seen. It should be noted that in one experimental model of streptococcal fasciitis, an active vaccine against one streptococcal superantigen (SPEA) actually increased the mortality rate [87]. These authors speculate that this superantigen-based vaccine resulted in clonal deletion of the relevant Vb speci~c T cells. This may induce T cell anergy to subsequent a microbial challenge with an SPEA-producing strain of S. pyogenes [87]. Unanticipated immune-mediated toxicity of this nature will need to be fully investigated before candidate anti-sepsis vaccines are tested in human subjects. It may be possible to present epitopes that are conserved structural elements of the superantigens as vaccines. These epitopes will not directly interact with

231

Vb binding sites on T cells. This will limit the possibility of vaccine-associated immune-mediated toxicity.

Summary Septic shock is the result of a deleterious, generalized host immune response to microbial mediators. Viewed in this context, septic shock should be treated as a toxigenic illness. Antibody strategies designed to block the injurious properties of microbial toxins upon release into the systemic circulation remains a logical preventative and therapeutic approach to sepsis. While major challenges remain in the development of an anti-sepsis vaccine, recent advances in the molecular biology microbial toxins provide new opportunities for vaccine development. Immunoprophylaxis remains a viable option in the prevention of sepsis. The dramatic rapidity with which sepsis develops and the disappointingly few useful interventions available to clinicians in the care of severe sepsis mandates that work on immune prevention continue. Whether recent basic science advances in the immunopathogenesis of sepsis will lead to clinically meaningful bene~ts in the prevention of sepsis awaits carefully conceived, future clinical investigation.

References 1. Hayden GF, Sato PA, Wright PF, et al. Progress in worldwide control and elimination of disease through immunization. J Pediatrics 1989;114:520–527. 2. Zeni F, Freeman B, Natanson C. Anti-in_ammatory therapies to treat sepsis and septic shock: A reassessment. Crit Care Med 1997;25:1059–1100. 3. Bhattacharjee AK, Cross AS. Vaccines and antibodies in the prevention and treatment of sepsis. Infect Dis Clin North America 1999; (in press). 4. Brun-Buisson C, Doyon F, Carlet J, et al. Incidence, risk factors, and outcome of severe sepsis and septic shock in adults: A multicenter, prospective study in intensive care units. JAMA 1995;274:968–974. 5. Flaherty JP, Weinstein RA. Nosocomial infections caused by antibiotic-resistant organisms in the intensive care unit. Infect Control Hosp Epidemiol 1996;17:236–248. 6. Opal SM. Sepsis. Sci Am Med 1998;30:1–17. 7. Donta ST, Paduzzi P, Cross AS, et al. Immunoprophylaxis against Klebsiella and Pseudomonas aeruginosa infections. J Infect Dis 1996;174:537–543. 8. Goldie AS, Fearon KCH, Ross JA, et al. Natural cytokine antagonists and endogenous antiendotoxin core antibodies in sepsis syndrome. JAMA 1995;274:172–177. 9. Hustinx WNM, Kraaijeveld K, Hoepelman AIM, et al. Cross protection by anti-core glycolipid antibodies: Evidence from animal experiments. J Antimicrob Chemother 1997;40:475– 486. 10. Lugowski C, Niedziela TN, Jachymek W. Anti-endotoxin antibodies directed against Escherichia coli R1 oligosaccharide core-tetanus toxoid conjugate bind to smooth, live bacteria and smooth lipopolysaccharides and attenuate their tumor necrosis factor stimulating activity. FEMS Immunol Med Microbiol 1996;16:31–39.

232

Opal et al.

11. Muller-Loennies S, Holst O, Brade H. Chemical structure of the core region of Escherichia coli J5 lipopolysaccharide. Eur J Biochem 1994;224:751–759. 12. Bailat S, Heummann D, LeRoy D, et al. Similarities and disparities between core-speci~c and O-side chain-speci~c antilipopolysaccharide monoclonal antibodies in models of endotoxemia and bacteremia in mice. Infect Immun 1997;65: 811–814. 13. Rietschel ET, Schletter J, Weidemann B, et al. Lipoploysaccaride and peptidoglycan: CD14-dependent bacterial inducers of in_ammation. Microbial Drug Resistance 1998;1: 37–44. 14. Opal SM, Yu RL Jr. Antiendotoxin strategies for the prevention and treatment of septic shock. Drugs 1998;55:497– 508. 15. Campbell WN, Hendrix, E, Cryz S Jr, Cross AS. Immunogenicity of a 24-valent Klebsiella capsular polysaccharide vaccine and an eight-valent Pseudomonas O-polysaccharide conjugate vaccine administered to victims of severe trauma. Clin Infect Dis 1996;23:179–181. 16. Bhattacharjee AK, Cross AS Opal, et al. New trends in Escherichia coli O111: B4 J5 mutant (Rc chemotype) vaccine development for use in gram-negative bacillary sepsis. In: Morrison DC, Ryan JL, eds. Novel Therapeutic Strategies in the Treatment of Sepsis. New York: Marcel Dekker, Inc., 1996:31–41. 17. Opal SM, Cross AS, Bhattacharjee A, et al. Active immunization with an E. coli J5 LPS vaccine in an experimental model of gram-negative sepsis. Program abstracts of the 36th ICAAC Meeting 1996; (Abstract# G2). 18. Rietschel ET, Brade H, Holst O, et al. Bacterial endotoxin; chemical composition, biological recognition, host response, and immunological detoxi~cation. Curr Topics Microbiol Immunol 1996;216:39–81. 19. Swaminathan S, Furey W, Pletcher, et al. Crystal structure of staphylococcal enterotoxin B, a superantigen. Nature 1992;359:801–805. 20. Jardetzky TS, Brown JH, Gorga JC, et al. Three-dimensional structure of a human class II histocompatibility molecule complexed with superantigen. Nature 1994;368:711– 718. 21. Wahlsten JL, Ramakrishman S. Separation of function between the domains of toxic shock syndrome-1. J Immunol 1998;160:854–859. 22. Woody MA, Krakauer T, Ulrich RG, Stiles BJ. Differential immune responses to staphylococcal enterotoxin B mutation in a hydrophobic loop dominating the interface with major histocompatibility complex class II receptors. J Infect Dis 1998;177:1013–1022. 23. Kusunoki T, Wright SD. Chemical characteristics of Staphylococcus aureus molecules that have CD14-dependent cellstimulating activity. J Immunol 1996;157:5112–5118. 24. Weidemann B, Schletter J, Dziarski R, et al. Speci~c binding of soluble peptidoglycan and muramyl dipeptide to CD14 on human monocytes. Infect Immun 1997;65:858–863. 25. Dziarski R, Tapping RI, Tobias PS. Binding of bacterial peptidoglycan to CD14. J Biol Chem 1998;273:8680–8686. 26. van Langevelde P, van Dissel JT, Ravensbergen E, et al. Antibiotic-induced release of lipoteichoic acid and peptidoglycan from Staphylococcus aureus: Quantitative measurements and biological reactivities. Antimicrob Agents Chemother 1998;3073–3078. 27. Krieg AM, Love-Homan L, Ae-Kyung Y, et al. CpG DNA induces sustained IL-12 expression in vivo and resistance to

28.

29.

30. 31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

Listeria monocytogenes challenge. J Immunol 1998;161: 2428–2434. Krieg AM, Wu T, Weeratna R, et al. Sequence motifs in adenoviral DNA block immune activation by stimulatory CpG motifs. Proc Natl Acad Sci USA 1998;95:12631–12636. Horn DL, Opal SM, LoMastro E. Antibiotics, endotoxin, and cytokine release, a complex and dynamic process. Scand J Infect Dis 1996;101:9–13. Cross AS. Antiendotoxin antibodies: A dead end? Ann Intern Med 1994;121:58–60. Ge Y, Ezzell RM, Tompkins RG, et al. Cellular distribution of endotoxin after injection of chemically puri~ed lipopolysaccharide differs from that after injection of live bacteria, J Infect Dis 1994;169:95–104. Wurfel MM, Hailman E, Wright SD. Soluble CD14 acts as a shuttle in the neutralization of lipopolysaccharide (LPS) by LPS-binding protein and reconstituted high density lipoprotein. J Exp Med 1994;181:1743–1754. Yang R-B, Mark MR, Gray A, et al. Toll-like receptor-2 mediates lipopolysaccharide-induced cellular signaling. Nature 1998;395:284–286. Bone RC, Balk RA, Fein AM, et al. A second large controlled clinical study of E5, a monoclonal antibody to endotoxin: Results of a prospective, multicenter, randomized, controlled trial. Crit Care Med 1995;23:994–1005. Greenman RL, Schein RMH, Martin MA, et al. A controlled clinical trial of E5 murine monoclonal IgM antibody to endotoxin in the treatment of gram-negative sepsis. JAMA 1991; 266:1097–1101. McCloskey RV, Straube RC, Sanders C, et al. Treatment of septic shock with human monoclonal antibody HA-1A. A randomized, double-blind, placebo-controlled trial. Ann Intern Med 1994;121:1–5. Ziegler EJ, Fisher CJ Jr, Sprung CL, et al. Treatment of Gram-negative bacteremia and septic shock with HA-1A human monoclonal antibody against endotoxin: A randomized, double-blind, placebo-controlled trial. N Engl J Med 1991;324:429–436. Derkx B, Wittes J, McCloskey R, and the European Pediatric Meningococcal Septic Shock Trial Study Group. Randomized, placobo-controlled trial of HA-1A, a human monoclonal antibody to endotoxin, in children with meningococcal septic shock. Clin Infect Dis 1999;28:770–777. Bhat NM, Bieber MM, Chapman CJ, et al. Human anti-lipid A monoclonal antibodies bind to human B cells and the i antigen on cord red blood cells. J Immunol 1993;151:5011– 5021. Quezado ZMN, Natanson C, Alling DW, et al. A controlled trial of HA-1A in a canine model of gram-negative septic shock. JAMA 1993;269:2221–2227. Di Padova F, Brade H, Barclay GR, et al. A broadly crossprotective monoclonal antibody binding to Escherichia coli and Salmonella lipopolysaccharides. Infect Immun 1993;61: 3863–3872. Braude AI, Douglas H, Davis CE. Treatment and prevention of intravascular coagulation with antiserum to endotoxin. J Infect Dis 1973;128:S157–164. Chedid L, Parant M, Boyer F. A proposed mechanism for natural immunity to enterobacterial pathogens. J Immunol 1968;100:292–306. Greisman SE, Young EJ, DuBuy JB. Mechanisms of endotoxin tolerance. VIII. Speci~city of serum transfer. J Immunol 1973;111:1349–1355. McCabe WR, Kreger BE, Johns M. Type-speci~c and cross-

Immunoprophylaxis

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56. 57. 58.

59.

60.

61.

reactive antibodies in gram-negative bacteremia. N Engl J Med 1972;287:261–267. McCabe WR, DeMaria A, Berberich H, et al. Immunization with rough mutants of Salmonella minnesota: Protective activity of IgM and IgG antibody to the R595 (Re Chemotype) mutant. J Infect Dis 1998;158:291–300. Warren HS, Amato SF, Fitting C, et al. Assessment of ability of murine and human anti-lipid A monoclonal antibodies to bind and neutralize lipopolysaccharide. J Exp Med 1993; 177:89–97. Pollack M, Huang AI, Prescott RW, et al. Enhanced survival in Pseudomonas aeruginosa septicemia associated with high levels of circulating antibody to Escherichia coli endotoxin core. J Clin Invest 1983;72:1874–1881. Nys M, Darnas P, Joassin L, et al. Sequential anti-core glycolipid immunoglobulin antibody activities in patients with and without septic shock and their relation to outcome. Ann Surg 1993;217:300–306. Bennett-Guerrero E, Ayuso L, Hamilton-Davies C, et al. Relationship of pre-operative anti-endotoxin core antibodies and adverse outcomes following cardiac surgery. JAMA 1997;277:646–650. Ziegler EJ, McCutchan JA, Fierer J, et al. Treatment of Gram-negative bacteremia and shock with human antiserum to a mutant Escherichia coli. N Engl J Med 1982;307:1225– 1230. Hellman J, Zanzot EM, Loiselle PM, et al. Antiserum against Escherichia coli J5 contains antibodies reactive with outer membrane proteins of heterologous gram-negative bacteria. J Infect Dis 1997;176:1260–1268. Bhattacharjee AK, Opal SM, Palardy JE, et al. Af~nitypuri~ed Escherichia coli J5 lipopolysaccharide-speci~c IgG protects neutropenic rats against gram-negative bacterial sepsis. J Infect Dis 1994;170:622–629. Bhattacharjee AK, Opal SM, Taylor R, et al. A noncovalent complex vaccine prepared with detoxi~ed Escherichia coli J5 (Rc chemotype) lipopolysaccharide and Neisseria meningitidis Group B outer membrane protein produces protective antibodies against gram-negative bacteremia. J Infect Dis 1996;173:1157–1162. Kim KS, Kang JH, Cross AS, et al. Functional activities of monoclonal antibodies to the O side chain of Escherichia coli lipopolysaccharides in vitro and in vivo. J Infect Dis 1998; 157:47–53. Marrack P, Kappler J. The staphylococcal enterotoxins and their relatives. Science 1990;248:705–711. Schleivert PM. Role of superantigens in human disease. J Infect Dis 1993;167:997–1002. Kotzin BL, Leung DY, Kappler J, et al. Superantigens and their potential role in human disease. Adv Immunol 1993; 54:99–166. Astiz M, Saha D, Lustbader D, et al. Monocyte response to bacterial toxins, expression of cell surface receptors, and release of anti-in_ammatory cytokines during sepsis. J Lab Clin Med 1996;128:594–600. Muller-Alouf H, Alouf JE, Gerlach D, et al. Comparative study of cytokine release by human peripheral blood mononuclear cells stimulated with Streptococcus pyogenes superantigenic erythrogenic toxins, heat killed streptococci, and lipopolysaccharide. Infect Immun 1994;62:4915–4921. Goldbach-Mansky R, King PD, Taylor AP, et al. A co-stimulatory role for CD28 in the activation of CD4⫹ T lymphocytes by staphylococcal enterotoxin B. Int Immunol 1992; 4:1351–1360.

233

62. Blankson JN, Morse SS. The CD28/B7 pathway costimulates the response of primary murine T cells to superantigens as well as to conventional antigens. Cell Immunol 1994;157:306–312. 63. Krakauer T. Cell adhesion molecules are co-receptors for staphylococcal enterotoxin B-induced T-cell activation and cytokine production. Immunol Lett 1994;39:121–125. 64. Blank C, Luz A, Bendigs S, et al. Superantigens and endotoxin synergize in the induction of lethal shock. Eur J Immunol 1997;27:825–833. 65. Ren K, Bannan JD, Pancholi V, et al. Characterization and biological properties of a new staphylococcal exotoxin. J Exp Med 1994;180:1675–1683. 66. Reda KB, Kapur V, Mollick JA, et al. Molecular characterization and phylogenetic distribution of the streptococcal superantigen gene (ssa) from Streptococcus pyogenes. Infect Immun 1994;62:1867–1874. 67. Blomster-Hautama DA, Kreiswirth BN, Kornblum JS, et al. The nucleotide and partial amino acid sequence of toxic shock syndrome toxin-1. J Biol Chem 1986;61:15783–15786. 68. Acharya KR, Passalacqua EF, Jones EY, et al. Structural basis of superantigen action inferred from crystal structure of toxic-shock syndrome toxin-1. Nature 1994;367:94–97. 69. Grossman D, Van M, Mollick JA, et al. Mutation of the disul~de loop in staphylococcal enterotoxin A. Consequences for T cell recognition. J Immunol 1991;147:3274–3281. 70. Kappler JW, Herman A, Clements J, et al. Mutations de~ning functional regions of the superantigen staphylococcal enterotoxin B. J Exp Med 1992;175:387–396. 71. Hartwig UF, Fleisher B. Mutations affecting MHC class II binding of the superantigen streptococcal erythrogenic toxin A. Int Immunol 1993;5:869–875. 72. Stelma GN Jr, Bergdoll MS. Inactivation of staphylococcal enterotoxin A by chemical modi~cation. Biochem Biophys Res Comm 1982;105:121–126. 73. Grossman D, Cook RG, Sparrow JT, et al. Dissociation of the stimulatory activities of staphylococcal enterotoxins for T cells and monocytes. J Exp Med 1990;172:1831–1841. 74. Mollick JA, McMasters RL, Grossman D, et al. Localization of a site on bacterial superantigens that determines T cell receptor beta chain speci~city. J Exp Med 1993;177:283–293. 75. Griggs ND, Pontzer CH, Jarpe MA, et al. Mapping of multiple binding domains of the superantigen staphylococcal enterotoxin A for HLA. J Immunol 1992;148:2516–2521. 76. Soos JM, Johnson HM. Multiple binding sites on the superantigen, staphylococcal enterotoxin B, imparts versatility in binding to MHC class II molecules. Biochem Biophys Res Comm 1994;201:596–602. 77. Bannan JD, Mingo F, Viteri A, et al. Neutralization of streptococcal pyrogenic exotoxins and staphylococcal enterotoxins by antisera to synthetic peptides representing conserved amino acid motifs. Adv Exp Med Biol 1997;418:903–907. 78. Bohach GA, Hovde CJ, Handley JP, et al. Cross-neutralization of staphylococcal and streptococcal pyrogenic toxins by monoclonal and polyclonal antibodies. Infect Immun 1988; 56:400–404. 79. Spero L, Morlock B, Metzger J. On the cross-reactivity of staphylococcal enterotoxins A, B, and C. J Immunol 1978;120:86–89. 80. Pontzer CH, Griggs ND, Johnson HM. Agonist properties of a microbial superantigen peptide. Biochem Biophys Res Commun 1993;193:1191–1197. 81. Andersson J, Skansén-Saphir U, Björk L, et al. Lymphokine production induced by streptococcal pyrogenic exotoxin-A

234

Opal et al.

is selectively down-regulated by pooled human IgG. Eur J Immunol 1994;24:916–922. 82. Björk L, Andersson U, Chauvet J-M, et al. Quanti~cation of superantigen induced IFN-c production by computerized image analysis-inhibition of cytokine production and blast transformation by pooled human IgG. J Immunol Methods 1994;175:201–213. 83. Norrby-Teglund A, Basma H, Andersson J, et al. Varying titers of neutralizing antibodies to streptococcal superantigens in different preparations of normal polyspeci~c immunoglobulin G: Implications for therapeutic ef~cacy. Clin Infect Dis 1998;26:631–638. 84. Norrby-Teglund A, Kaul R, Low DE, et al. Plasma from patients with severe invasive group A strepcococcal infections treated with normal polyspeci~c IgG inhibits strepto-

coccal superantigen- induced T cell proliferation and cytokine production. J Immunol 1996;156:3057–3064. 85. Kaul R, McGear A, Norrby-Teglund A, Kotb M, Schwartz B, O’Rourke K, Talbot J, Low DE, and the Canadian Streptococcal Study Group. Intravenous immunoglobulin therapy for streptococcal toxic shock syndrome—A comparative observational study. Clin Infect Dis 1999;28:800–807. 86. Werdan K, Pilz G. Supplemental immune globulins in sepsis: A critical appraisal. Clin Exp Immunol 1996;104(Suppl 1): 83–90 87. Sriskandan S, Moyes D, Buttery LK, Krausz T, Evans TJ, Polak J, Cohen J, Streptococcal pyrogenic exotoxin A release, distribution, and role in a murine model of fasciitis and multiorgan failure due to Streptococcus pyogenes. J Infect Dis 1996;173:1399–1407.