In vitro human antibody production to the Haemophilus influenzae type b capsular polysaccharide

In vitro production of human antibody to the Haemophilus influenzae type b capsular polysaccharide (PRP) and to tetanus toxoid (TT) and diphtheria toxoid was measured in culture supernatants of peripheral blood mononuclear cells and by enumeration of antibody secreting cells (AbSC) in an enzyme-linked immunosorbent-plaquing assay. Normal adult peripheral blood mononuclear cells stimulated with Epstein-Barr virus secreted anti-PRP antibody with a frequency of 1/552 to 1/1190 relative to total Ig secreting cells; the frequency of AbSC to tetanus toxoid (TT) was 7.5 times higher (p less than 0.05). These frequencies did not change significantly after in vivo immunization, although the isotype distribution shifted toward increased IgG for TT and increased IgG and IgA for PRP. At 8 days postimmunization, spontaneous AbSC to PRP and TT were detected; frequencies for total anti-TT AbSC again being higher than anti-PRP, but there were significantly more IgA plaques among anti-PRP AbSC. Spontaneous AbSC were suppressed in culture by pokeweed mitogen and enhanced by cyclosporine. Three wk after in vivo immunization with PRP and TT, in vitro stimulation with pokeweed mitogen, Staphylococcus aureus Cowan 1 bacteria, or antigen induced anti-TT but not anti-PRP in vitro antibody secretion, although Epstein-Barr virus induced both. These data suggest that PRP, a polysaccharide, and TT, a protein, differ in their requirements for in vitro activation with antigen and mitogens.     

Subclass distribution of human antibodies to Haemophilus influenzae type b capsular polysaccharide

The polysaccharide (PS) capsule of Haemophilus influenzae type b (Hib) is a "simple" antigen, polyribosylribitolphosphate. Although similar carbohydrate antigens have been reported to elicit IgG antibodies relatively restricted to the IgG2 subclass in man, we report here that Hib PS elicits substantial quantities of both IgG1 and IgG2 serum antibodies in most individuals. Because the determination of IgG subclass distribution can be technically difficult, we used four different approaches to establish our finding. First, we used an IgG subclass-specific, antigen-specific "sandwich assay." Second, we measured IgG subclasses of purified antibodies to Hib PS. Third, we showed that significant amounts of IgG anti-PS can be absorbed with a monoclonal anti-IgG1 affinity column. Fourth, we showed that IgG1 and IgG2 fractions of immune sera have clonally restricted anti-Hib PS antibodies that are easily distinguishable by their isoelectric points. The data indicate that both IgG1 and IgG2 contribute substantially to the IgG antibody response of most adults to immunization with Hib PS.     

Isoelectric focusing of human antibody to the Haemophilus influenzae b capsular polysaccharide: restricted and identical spectrotypes in adults

Serum antibody to the capsular polysaccharide of Haemophilus influenzae b of human adults was analyzed by isoelectric focusing. Restricted antibody spectrotype patterns were commonly observed with as few as one spectrotype in some subjects after immunization with the isolated capsular polysaccharide. Some patterns were as restricted as human hybridoma antibody. There was no correlation of antibody titer and heterogeneity of patterns. The dominant spectrotype persisted unchanged for over 2 yr after immunization, and the pattern detected in preimmunization serum samples persisted unchanged after immunization. Indistinguishable patterns were commonly observed in genetically unrelated adults. Adults immunized with conjugate vaccines, which were composed of oligosaccharides prepared from the capsular polysaccharide that were covalently linked to protein carriers, also produced restricted serum antibody spectrotype patterns. Immunization with the cross-reactive polysaccharide of E. coli K100 induced a spectrotype pattern that was restricted but different from that induced by the H. influenzae b capsular polysaccharide.     

Evidence for covalent attachment of phospholipid to the capsular polysaccharide of Haemophilus influenzae type b.

Cells of Haemophilus influenzae type b were grown in a liquid medium containing [3H]palmitate or [14C]ribose or both for two generations of exponential growth. Radiolabeled type-specific capsular polysaccharide, polyribosyl ribitol phosphate (PRP), was purified from the culture supernatant by Cetavlon precipitation, ethanol fractionation, and hydroxylapatite and Sepharose 4B chromatography. The doubly labeled ( [3H]palmitate and [14C]ribose) PRP preparation was found to coelute in a single peak from a Sepharose 4B column, suggesting that both precursors were incorporated into the purified PRP. A singly labeled ( [3H]palmitate) purified PRP preparation was found to be quantitatively immune precipitated by human serum containing antibody against PRP. The radioactivity of this preparation could not be dissociated from PRP by treatment with chloroform-methanol, 6 M urea, sodium dodecyl sulfate, or Zwittergent. Only after acid, alkaline, or phospholipase A2 treatment of PRP labeled with [3H]palmitate or [3H]palmitate and [14C]ribose followed by chloroform-methanol extraction could most of the 3H-radioactivity be recovered in the organic phase. The chloroform-soluble acid-hydrolyzed or phospholipase A2-treated product was identified as palmitic acid after thin-layer chromatography. These results strongly suggest that a phospholipid moiety is covalently associated with the H. influenzae type b polysaccharide PRP.     

Functional activity of different IgG subclass antibodies against type b capsular polysaccharide of Haemophilus influenzae

The biologic activity of different human IgG subclass antibodies directed against the Haemophilus influenzae type b (Hib) capsular polysaccharide (PRP) was compared by using an in vitro complement-mediated bactericidal assay and an in vivo passive protection assay in infant rats. An IgG pool was made by Sephacryl S-300 chromatography of sera from adults immunized with PRP vaccine. An IgG2 subclass fraction was prepared by column immunoabsorption of the IgG pool with anti-IgG1 monoclonal antibody. An IgG1 subclass fraction was eluted from the affinity matrix. IgG1, IgG2, IgG3, and IgG4 concentrations in the fractions were measured by solid-phase competitive radioimmunoassays, and anti-PRP antibody was measured by a modified Farr assay. Each fraction was greater than 90% pure IgG2 or IgG1, respectively. There were no significant differences in the minimal anti-PRP antibody concentrations required to kill 50% of Hib cells in vitro (IgG, 0.22; IgG1, 0.21; and IgG2, 0.42 microgram/ml). Similarly, equivalent amounts of anti-PRP antibody of the IgG1 or IgG2 fractions protected against bacteremia (IgG1, 0.12; IgG2, 0.24 microgram per rat). IgG absorbed to remove anti-PRP antibody was neither bactericidal nor protective. Thus IgG1 and IgG2 anti-PRP antibody have equivalent functional activities against Hib as determined by these biologic assays.     

Human antibody response to outer membrane proteins and fimbriae of Haemophilus influenzae type b

We investigated the prevalence of antibodies in childrens' sera directed against outer membrane proteins (OMP) and fimbriae of Haemophilus influenzae type b. Invasive isolates of H. influenzae type b were enriched for fimbriae production; OMP and fimbriae were resolved by SDS-PAGE. After blotting to nitrocellulose, the proteins were incubated with homologous patient sera or with sera from healthy children. IgG antibodies bound to OMP were detected by immunoperoxidase staining. Immunoblotting was also performed using purified, nondenatured fimbriae as antigen. Nine of the 10 patients studied had antibodies in the acute serum directed against one or more of the OMP. Neither the acute nor the convalescent serum of the remaining patient contained antibodies against OMP. Antibodies against a greater number of OMP were present in the convalescent serum, in comparison to the acute serum, in 4 of the 10 patients. Five of 10 patients had antibodies against the purified fimbriae of an unrelated invasive isolate in either the acute or the convalescent serum. Acute sera from patients more frequently contained antibodies directed against OMP 60K (p less than or equal to 0.01) and OMP 51K (p less than or equal to 0.003) compared with the sera of healthy controls. In contrast, the sera of healthy children more frequently contained antibodies directed against OMP 40K (p less than or equal to 0.04). Sera from both patients and controls contained antibodies against commensal Haemophilus. We conclude that although antibodies against OMP are commonly present in healthy children, antibodies against certain OMP may be markers for susceptibility or protection.     

Recognition of riboflavin and the capsular polysaccharide of Haemophilus influenzae type b by antibodies generated to the haptenic epitope D-ribitol

D-Ribitol, a five–carbon sugar alcohol, is an important metabolite in the pentose phosphate pathway; it is an integral part of riboflavin (vitamin B2) and cell wall polysaccharides in most Gram-positive and a few Gram-negative bacteria. Antibodies specific to D-ribitol were generated in New Zealand white rabbits by using reductively aminated D-ribose-BSA conjugate as the immunogen. MALDI-TOF and amino group analyses of ribitol-BSA conjugate following 120 h reaction showed ~27–30 mol of ribitol conjugated per mole BSA. The presence of sugar alcohol in the conjugates was also confirmed by an increase in molecular mass and a positive periodic acid–Schiff staining in SDS-PAGE. Caprylic acid precipitation of rabbit serum followed by hapten affinity chromatography on ribitol–KLH–Sepharose CL-6B resulted in pure ribitol–specific antibodies (~45–50 μg/mL). The affinity constant of ribitol antibodies was found to be 2.9?×?10⁷ M^?1 by non-competitive ELISA. Ribitol antibodies showed 100 % specificity towards ribitol, ~800 % cross–reactivity towards riboflavin, 10–15 % cross–reactivity with sorbitol, xylitol and mannitol, and 5–7 % cross–reactivity with L-arabinitol and meso-erythritol. The specificity of antibody to ribitol was further confirmed by its low cross-reactivity (0.4 %) with lumichrome. Antibodies to D-ribitol recognized the purified capsular polysaccharide of Haemophilus influenzae type b, which could be specifically inhibited by ribitol. In conclusion, antibodies specific to D-ribitol have been generated and characterized, which have potential applications in the detection of free riboflavin and ribitol in biological samples, as well as identification of cell-surface macromolecules containing ribitol.     

An NMR spectroscopic identity test for the control of the capsular polysaccharide from Haemophilus influenzae type b.

We describe the use of Nuclear Magnetic Resonance (NMR) spectroscopy to control the identity of purified bulk capsular polysaccharide [called poly(ribosylribitolphosphate) or PRP] from Haemophilus influenzae type b (Hib), and derivatised forms, used in the production of Hib polysaccharide-protein conjugate vaccines. We describe the approaches we have developed to validate this test.     

Clonal characterization of the human IgG antibody repertoire to Haemophilus influenzae type B polysaccharide. Demonstration of three types of V regions and their association with H and L chain isotypes

The antibody response to Haemophilus influenzae type b polysaccharide (Hib-PS) is pauciclonal but can vary between different individuals. To estimate the size of this antibody repertoire we examined the constant and V regions of human IgG anti-Hib antibodies from 14 individuals at the clonal level using various serologic and IEF methods. Examination of H chains showed that 11 of 14 individuals produced both IgG1 and IgG2 antibodies, two individuals produced only IgG2 and one individual produced only IgG1 antibody. All 14 individuals produced kappa-containing antibody clones and three persons also produced significant lambda-containing antibody clones. V region heterogeneity was examined by comparing cross-reactivity of anti-Hib-PS antibodies to the capsular polysaccharide of Escherichia coli K100 carbohydrate (K100 CHO). These studies showed that clones of IgG anti-Hib-PS antibodies cross-reactive with K100 CHO were present in 5 of 14 (36%) individuals and also revealed at least three types of V regions among these antibodies. The first type has no cross-reaction with K100 CHO and was found in 13 of the 14 individuals. The second type, found in three of 14 individuals, cross-reacts with K100 CHO and uses a lambda L chain V region. The third type, found in 2 of 14 individuals, cross-reacts with K100 CHO and uses a kappa L chain V region. Although the lambda type V region was found only in association with IgG2, the other two V region types associate with both IgG1 and IgG2. Thus, five IgG antibody clones are serologically discernable. An individual generally responds to Hib-PS by expressing several clones selected from these discernable antibody clones. Indeed, we can observe six individual response patterns among these 14 individuals and conclude that considerable variability in individual responses to Hib-PS can be achieved with very few V regions.     

IGH V3-23*01 and its allele V3-23*03 differ in their capacity to form the canonical human antibody combining site specific for the capsular polysaccharide of Haemophilus influenzae type b

The IGH V3-23*01 gene is used in the formation of the canonical combining site which dominates the human antibody repertoire to the Haemophilus influenzae type b (Hib) polysaccharide (PS). An allele of the human IGH V3-23*01 gene, known as V3-23*03, differs from V3-23*01 in nine bases, eight of which are located in the second complementarity determining region. These eight differences encode five amino acid substitutions. In this study we investigated whether the V3-23*03 sequence polymorphism affected Hib PS binding. We constructed two Fab fragments that had the canonical Hib PS combining site VH-VL configuration but that had either V3-23*01 or V3-23*03. Radioantigen binding assay showed that on a concentration basis the V3-23*03 Fab was 20-fold more effective in binding Hib PS than the V3-23*01 Fab. The V3-23*03 Fab was 4-fold more effective than the V3-23*01 Fab in mediating facilitated bactericidal activity against Hib organisms. These findings identify a functional consequence of V3-23 allelism, and suggest that utilization of the V3-23*03 gene in the human Hib PS repertoire would generate canonical antibodies with higher affinity and protective efficacy than canonical antibodies utilizing V3-23*01. Thus, IGH V gene allelic variation has the potential to impact the generation of protective immunity to Hib.     

Intrinsic tritium labeling of the capsular polysaccharide antigen of Haemophilus influenzae type B.

The capsular polysaccharide of Haemophilus influenzae type b was intrinsically labeled with tritium by a microculture technique with 6-3H-D-glucose and was isolated in radioantigenically pure form by a combination of selective precipitation and molecular sieve chromatography. Labeling with tritated sugar residues approached one-fourth maximum and produced a specific activity 10-fold that previously described for extrinsic labeling methods. In radioantigen-binding assays for antibody, sensitivity depended on the size of the antigen; preparations were readily made that could detect 0.01 microgram Ab/ml in serum samples of 25 microliter. Stability of the labeled antigen appears limited only by the primary radiodecomposition of tritium.     

Recognition of flavin mononucleotide,Haemophilus influenzae type b and its capsular polysaccharide vaccines by antibodies specific to D-ribitol-5-phosphate

D-Ribitol-5-phosphate (Rbt-5-P) is an important metabolite in the pentose phosphate pathway and an integral part of bacterial cell wall polysaccharides, specifically as polyribosyl ribitol phosphate (PRP) in Haemophilus influenzae type b (Hib). The major objective of this study was to investigate whether an antibody specific to Rbt-5-P can recognize the PRP of Hib. D-Ribose-5-phosphate was reacted with proteins in the presence of sodium cyanoborohydride to obtain Rbt-5-P epitopes; 120 h reaction resulted in conjugation of ~30 and ~17 moles of Rbt-5-P/mole of BSA and OVA, respectively, based on decrease in amino groups, MALDI–TOF analyses, an increase in apparent molecular weight (SDS–PAGE) and glycoprotein staining. Immunization of rabbits with Rbt-5-P–BSA conjugate generated antibodies to Rbt-5-P as demonstrated by dot immunoblot and non-competitive ELISA. Homogeneous Rbt-5-P-specific antibody was purified from Rbt-5-P–BSA antiserum subjected to caprylic acid precipitation followed by hapten–affinity chromatography; its affinity constant is 7.1?×?10⁸ M^?1. Rbt-5-P antibody showed 100 % specificity to Rbt-5-P, ~230 %, 10 % and 3.4 % cross-reactivity to FMN, riboflavin and FAD, respectively; the antibody showed ~4 % cross-reactivity to D-ribitol and <3 % to other sugars/sugar alcohols. Rbt-5-P–specific antibody recognized Hib conjugate vaccines containing PRP which was inhibited specifically by Rbt-5-P, and also detected Hib cell-surface capsular polysaccharides by immunofluorescence. In conclusion, Rbt-5-P–protein conjugate used as an immunogen elicited antibodies binding to an epitope also present in PRP and Hib bacteria. Rbt-5-P–specific antibody has potential applications in the detection and quantification of free/bound Rbt-5-P and FMN as well as immunological recognition of Hib bacteria and its capsular polysaccharide.     

Detection of mucosal and serum antibodies specific for the capsular polysaccharide of Haemophilus influenzae type b by enzyme-linked immunosorbent assay

A sensitive enzyme-linked immunosorbent assay (ELISA) with rough-surfaced glass beads used as the solid phase was developed for detection of IgG, IgM, and IgA antibodies to Haemophilus influenzae type b capsular polysaccharide (HITB-CP). A successful method for indirect coating of glass bead surfaces with HITB-CP, and parameters affecting the specificity and sensitivity of the assay are described. This ELISA system proved to be 100 times as sensitive as the standard indirect fluorescent-antibody assay. The assay was applied to the measurement of antibodies to HITB-CP in serum and nasal secretions and proved to be a useful tool in the evaluation of immunological response to HITB infection.     

A novel approach to study variable heavy chain gene usage in response to the capsular polysaccharide of Neisseria meningitidis serogroup C

The molecular mechanisms involved in the relatively poor immune response in the elderly are not clearly understood. Qualitative aspects of the immune response could be a possible explanation for the differential response to T-independent antigens in young adults and elderly. This study is directed towards elucidating the differential usage of variable heavy chain by young adult and elderly derived sequences in response to the capsular polysaccharide of Neisseria meningitidis serogroup C. We currently report findings of a preliminary study designed to test the feasibility of a novel approach to isolate antigen-specific B cells. Paramagnetic beads coated with an anti-idiotypic antibody, which mimics the capsular polysaccharide of N. meningitidis serogroup C, were used to select B cells. Analysis of the gene usage data indicates some unexpected differences in the use of variable chain heavy chain in the case of young adult versus elderly sequences. The elderly derived sequences use a more diverse array of V_H gene families in contrast to the young adult sequences, where the V_H gene family usage is restricted. Nearly half the young adult sequences utilize V_H3-15 germline sequence while only 25% of the elderly sequences use this germline sequence. There were interesting differences in the types of JH chain and the composition and length of CDR3 utilized by the two groups. Together, these significant differences may contribute towards the poor immune response to T-independent antigens in the elderly. These data validate the techniques used for these studies and suggest that it is pertinent to use this approach towards future investigations to elucidate gene usage in response to an antigen.     

Clonal characterization of the human IgG antibody repertoire to Haemophilus influenzae type b polysaccharide. IV. The less frequently expressed VL are heterogeneous

We previously demonstrated that the human anti-Haemophilus influenzae type b polysaccharide (Hib-PS) VL repertoire is dominated by a product of the V kappa II gene, A2, and that V kappa II-A2 anti-Hib-PS antibodies have little or no somatic mutation in VL. To further study this VL repertoire, we studied non-A2 anti-Hib-PS antibodies that were identified either serologically or by amino-terminal amino acid sequence analysis. Of 15 non-A2 anti-Hib-PS antibodies from 12 vaccinated adults, we found four V lambda, five V kappa I, one non-A2 V kappa II, four V kappa III, and one V kappa IV antibodies. As expected, all but two of these subjects also produced V kappa II-A2 antibodies. Interestingly, one of these subjects lacks the A2 gene in the germ line. However, both subjects who did not produce detectable V kappa II antibody did produce normal amounts of total anti-Hib-PS antibody after vaccination. Candidate V kappa genes for the non-A2 antibodies were identified by comparison of up to 60 VL amino acid residues, including CDR1 and CDR2, with all sequenced V kappa genes. V kappa I antibodies appear to be products of three newly sequenced V kappa I genes, O8, O18, and L11, that are reported here. The O8 and O18 genes encode identical amino acid sequences. The non-A2 V kappa II antibody is a likely product of the A1 or A17 genes, the V kappa III antibodies are likely products of the A27 gene, and the V kappa IV antibody is a product of the single V kappa IV gene, B3. Unlike V kappa II-A2 antibodies, the V kappa I, V kappa III, and V kappa IV antibodies differed by one to five CDR residues from the germ line product of the candidate genes, suggesting the presence of somatic mutations. Thus, anti-Hib-PS antibodies can be divided into two types, the most frequently observed A2 antibodies with little or no somatic mutation and non-A2 antibodies that likely contain somatic mutations.     

Influence of the aminopeptide concentration on the growth of Haemophilus influenzae, type b, and the synthesis of its capsular polysaccharide

The influence of the aminopeptide concentration on the growth of H. influenzae b culture and the synthesis of H. influenzae b capsular polysaccharide was determined. The maximum amount of capsular polysaccharide was accumulated at the concentration of aminopeptide in the culture fluid reaching 50 ml/l. An increase in the aminopeptide concentration led to a decreased amount of synthesized polysaccharide and an increased amount of biomass. The decrease of the aminopeptide concentration to 10 ml/l resulted in decreased amounts of both biomass and synthesized polysaccharide.     

Determination of the structure of the Escherichia coli K100 capsular polysaccharide, cross-reactive with the capsule from type b Haemophilus influenzae

The structure of the Escherichia coli K100 capsular polysaccharide, cross-reactive with that from type b Haemophilus influenzae, was determined by using a combination of chemical and spectroscopic techniques. The structure of the K100 repeating unit was found to be----3)-beta-D-Ribf-(1----2)-D-ribitol-5-(PO4----. The K100 polysaccharide is thus identical in composition to, but different in linkage from, the H. influenzae type b capsular polysaccharide, which has beta-D-Ribf-(1----1)-D-ribitol linkages.