Conjugation of bacterial polysaccharides (PS) to protein carriers confers the ability to elicit protective serum Ab in infants, who respond poorly to plain PS. The serum Ab of young children immunized with Haemophilus influenzae type b (Hib) PS conjugate vaccine varies with age and Ag formulation. To understand these age-related changes in human anti-Hib PS immune responses we determined the variable region gene sequences encoding anti-Hib PS mAbs of infants immunized with Hib oligosaccharide-diphtheria toxin vaccine. The anti-Hib PS repertoire of children differs from that of adults. A smaller proportion of mAbs from children have high affinity for Hib PS, and the overall variable region gene repertoire of infants is more diverse than that in adults. Variable region genes encoding high affinity mAbs of infants are similar to the restricted repertoire described in adults. Low affinity anti-Hib PS mAbs of infants are encoded by a heterogeneous group of genes that are uncommonly observed in the adult repertoire. Abs with high affinity for Hib PS from infants, like most mAbs from adults, react only with Hib PS and the structurally similar PS of Escherichia coli K100, whereas low affinity mAbs of infants are polyreactive. The low affinity anti-Hib PS mAbs of infants immunized with Hib oligosaccharide-diphtheria toxin vaccine vaccine are not reflected in serum Ab. However, the differences between the variable region gene repertoires of adults and infants may account for the distinct immunologic characteristics of the anti-Hib PS responses in young children immunized with other vaccine formulations.

Polysaccharide (PS)3-encapsulated bacteria are major causes of serious infection in humans. Ab directed against the PS capsule is the principal means of defense against invasive disease (1). Bacterial PS, however, are prototypical T-independent (TI) Ags (2). In contrast to T-dependent (TD) protein Ags, immunization with plain PS vaccine typically elicits Ab with relatively low affinity for Ag, and repeated immunization does not produce memory or booster responses (2, 3, 4). Humans and other higher organisms exhibit a highly characteristic age-dependent response to TI Ags (5, 6). Young children typically do not begin to produce serum Ab against PS Ags until about 18 mo of age and, therefore, cannot be readily protected against life-threatening infections by PS immunization (3). The basis for this developmental delay in specific immunity is not known. Conjugation of bacterial PS or oligosaccharide to protein carriers confers many TD characteristics on anti-PS immune responses, including the ability to elicit protective serum Ab in infants as young as 4 to 6 mo of age (4).

The adult immune repertoire to Haemophilus influenzae type b capsular polysaccharide (Hib PS) is dominated by a limited number of heavy chain (IgVH) and light chain (IgVL) variable region genes (7, 8, 9, 10). Studies of serum anti-Hib PS Ab suggest that this repertoire varies with age and Ag formulation (11, 12, 13). Identification of the molecular basis of this diversity may shed light on the age-dependent development of the immune response to TI Ags. To better understand age-related changes in the human anti-Hib PS repertoire, we determined variable region gene sequences encoding anti-Hib PS mAbs of young children immunized with a TD Hib PS vaccine. Our data confirm that quantitative and qualitative differences exist between the anti-Hib PS repertoire of young children and the typical repertoire of adults following immunization.

A total of 19 young children were studied. Seventeen children were immunized at 2, 4, 6, and 15 to 18 mo of age with Hib oligosaccharide-CRM197 mutant Corynebacterium diphtheriae protein conjugate vaccine (HbOC; Lederle Laboratories, Wayne, NJ). Blood was obtained 5 to 10 days following immunization at 15 to 18 mo for Ab determination and to obtain immune PBL. Similar studies were performed on sera and lymphocytes of two infants (aged 12 and 18 mo) who had been immunized 7 days earlier with varicella vaccine (Varivax, Merck, West Point, PA) but who had not received Hib vaccine within 3 mo. Adult subjects received a single immunization with HbOC (six subjects), Hib-diphtheria toxoid conjugate vaccine (Connaught Laboratories, Swiftwater, PA; four subjects), or unconjugated Hib PS vaccine (Praxis, Rochester, NY; one subject). By 6 yr of age the vast majority of individuals have acquired natural immunity to Hib, and all adult subjects from whom preimmune sera were available (10 of 11) were found to have detectable anti-Hib PS serum Ab (7, 14). Cell lines obtained from both adults and infants after immunization, therefore, represent secondary immune responses.

Serum total anti-Hib PS Ab was measured by RIA with 125I-labeled Ag (15). IgG and IgM anti-Hib PS Ab concentrations were measured by ELISA using multiwell plates (model 3590, Costar, Cambridge, MA) coated with Hib PS coupled to poly-l-lysine. Ab was detected by incubation with biotinylated goat anti-human IgG or IgM Abs (BioSource, Camarillo, CA) followed by detection with avidan-coupled alkaline phosphatase and p-nitrophenylphosphate enzyme substrate as previously described (16). The level of IgG-anti-Hib PS Ab was measured quantitatively by comparison with a reference serum from the Center for Biologics Evaluation and Research (Bethesda, MD) that contained 66 μg/ml of IgG anti-Hib PS Ab. The results of the IgM assays were expressed as units relative to a serum pool (A144) that we prepared from adults immunized with Hib PS. The amount of Hib PS required for 50% inhibition of serum Ab or mAb binding to Hib PS (ID50) was measured by inhibition ELISA as previously described (16, 17). Low affinity anti-Hib PS mAbs were defined as those that were not inhibitable by 10 μg/ml of soluble Hib PS in this ELISA assay. Inhibition of binding by Escherichia coli K100 PS was measured using formalin-fixed pelleted bacteria containing K100 PS (18). The expression of HibId-1 and HibId-2 Ids (associated with the use of the VκII A2 and VλVII DPL19 light chain variable segments, respectively) was measured by ELISA as previously described (12). Binding to type 3 Streptococcus pneumoniae PS and to tetanus toxoid was measured by ELISA as described for Hib PS. Streptococcal membranes were prepared from M type 5 Streptococcus pyogenes as described by Cunningham (19). Type 3 S. pneumoniae PS was provided by Dr. Philip Vella, Merck. The type 3 PS was coupled to poly-l-lysine as previously described (15). Tetanus toxoid was obtained from Wyeth (Marletta, PA). Binding to streptococcal membranes, ssDNA, dsDNA (Calbiochem-Behring, La Jolla, CA), myosin, cardiolipin, and smRNP (Sigma, St. Louis, MO) was measured by ELISA. For these assays, the Ags were diluted 10 μg/ml in carbonate-bicarbonate buffer, pH 9.6, and were coated directly onto microtiter plates (Immulon 4, Dynatech, Chantilly, VA). Bound Ab was detected with alkaline phosphatase-conjugated goat anti-human Ig Ab (Sigma) as previously described (8). Binding of mAbs to other bacterial and self Ags was defined as present, or positive, when the absorbance was greater than twice the background.

Hybridoma cell lines secreting human anti-Hib PS mAb were produced, and variable region genes encoding heavy (IgVH) and light (IgVL) chains amplified from mRNA were cloned and sequenced as described previously (7, 16). Briefly, 107 to 108 immune PBLs were fused to 5 × 107 murine myeloma SP2/O-Ag14 cells, and heterohybridoma cells were selected by resistance to 1 × 10−4 M hypoxanthine, 4 × 10−5 M thymidine, and 4 × 10−7 M aminopterin as described by Carroll (20). Hybridoma cell lines secreting anti-Hib PS binding Ab were initially detected by ELISA, using multiwell plates (model 3590, Costar) coated with Hib PS coupled to poly-l-lysine (16). Hib PS-binding Ab was detected by incubation with biotinylated goat anti-human κ or λ Abs (BioSource, Camarillo, CA) followed by detection with avidan-coupled alkaline phosphatase and p-nitrophenylphosphate enzyme substrate as previously described. All hybridomas secreting anti-Hib PS binding mAb in a sufficient quantity of absorbance of >1.0 units by ELISA on culture supernatant were cloned by limiting dilution.

The geometric mean serum Ab concentration of serum IgG anti-Hib PS Ab from infants postimmunization was 31 μg/ml (range, 3–588.0), and that of adult subjects was 24 μg/ml (range, 7–60). The geometric mean concentration of IgM anti-Hib PS Abs in infants after immunization was 27 ELISA units (range, 2–262), and that in adult subjects was 124 (range, 24–611; p = 0.03). The mean ID50 of serum anti-Hib PS IgG in infants was 95 ng/ml, compared with 156 ng/ml in adults (p = 0.03). The mean ID50 of serum anti-Hib PS IgM in infants was 38 ng/ml, compared with 54 ng/ml in adults.

A total of 2094 mAb cell lines were generated from 16 infants immunized with HbOC vaccine. Of these, 78 (3.7%) reacted with Hib PS. Individually, however, there was great variation in the frequency of Hib PS-reactive cell lines (0–11%). To estimate the frequency of naturally occurring Hib PS-reactive cells, 391 hybridomas were produced from two infants of comparable age who had not been recently immunized with Hib vaccine. Four of 307 cell lines (1.4%) from one child and none of 84 cell lines from the other bound Hib PS. A similar analysis was performed on four adult subjects, all immunized with HbOC vaccine. Of 237 heterohybridoma cell lines obtained 7 to 9 days after immunization, 72 (30.4%) were reactive with Hib PS.

Most anti-Hib PS binding mAbs were not completely evaluable because these cell lines often secrete Ab only transiently. Of a total of 41 stable heterohybridomas from infants, 14 (34%) were not inhibitable by Hib PS, suggesting a low affinity interaction. All of the cell lines secreting anti-Hib PS mAbs that were inhibitable by Hib PS were recovered from three children, whereas noninhibitable mAbs were obtained from five children. Stable cell lines were recovered from three of four adult subjects immunized with HbOC vaccine, and all were inhibitable by Hib PS (n = 116). In this and previous studies, a total of 337 (78%) of 432 cell lines obtained from immunized adult subjects were inhibitable by Hib PS, including 89 of 102 (87%) from subjects immunized with Hib-diphtheria toxoid conjugate and 122 of 204 (60%) of those immunized with plain Hib PS.

Twenty-seven percent of Hib PS binding mAbs from infants were of the IgG isotype, 35% were IgA, and 38% were IgM. Of cell lines from infants that were inhibitable by Hib PS, 46% were IgG, 46% were IgA, and 8% were IgM. Cell lines from adults immunized with HbOC, which were all inhibitable by Hib PS, were 71% IgG, 26% IgA, and 4% IgM.

Of the 27 Hib PS-inhibitable mAbs from infants that could be further evaluated, 8 of 10 κ mAbs expressed the HibId-1 Id (associated with the VκII A2 germline Vκ segment), and 16 of 17 λ mAbs expressed the HibId-2 Id (associated with the VλVII DPL19 variable segment; Table I) (7, 12). Each of these Ids is frequently expressed in the adult anti-Hib PS repertoire. No mAb that was not inhibitable by Hib PS expressed the HibId-1 or HibId-2 Ids. All evaluable anti-Hib PS mAbs from adults immunized with HbOC vaccine used κ light chains; 24 of 44 (54.5%) were HibId-1 positive.

Table I.

Anti-Hib PS mAbs obtained from infantsa

Cell LineIsotypeHibid1Hibid2K100IIStreptococcal MembranesPn3ssDNAdsDNAsmRNPCardiolipinMyosin
Hib PS inhibitable mAbs             
DUN3 A1λ − ND ND ND ND ND ND ND ND 
DUN8 A1λ − ND ND ND ND ND ND ND ND 
DUN9 A1λ − − − − − − − − − 
MM1.1 G1κ − − − − − − − − − 
MM2 G1κ − ND ND ND ND ND ND ND ND 
MM4.1 G1κ − − − − − − − − − 
Non-Hib PS inhibitable mABS             
MAX1 Mκ − − ND ND 
MAX4 Mκ − − − ND ND − 
MAX5 Mκ − − − ND ND 
MAX6 Mλ − − − ND − ND − − − − − 
MAX8 Mλ − − − ND ND − − − 
MAX10 Mκ − − ND ND ND ND ND ND ND ND ND 
HW1 Mκ − − − ND − ND − − − − − 
RL1 Mλ − − ND ND − 
EF16 Mλ − − ND ND − 
Cell LineIsotypeHibid1Hibid2K100IIStreptococcal MembranesPn3ssDNAdsDNAsmRNPCardiolipinMyosin
Hib PS inhibitable mAbs             
DUN3 A1λ − ND ND ND ND ND ND ND ND 
DUN8 A1λ − ND ND ND ND ND ND ND ND 
DUN9 A1λ − − − − − − − − − 
MM1.1 G1κ − − − − − − − − − 
MM2 G1κ − ND ND ND ND ND ND ND ND 
MM4.1 G1κ − − − − − − − − − 
Non-Hib PS inhibitable mABS             
MAX1 Mκ − − ND ND 
MAX4 Mκ − − − ND ND − 
MAX5 Mκ − − − ND ND 
MAX6 Mλ − − − ND − ND − − − − − 
MAX8 Mλ − − − ND ND − − − 
MAX10 Mκ − − ND ND ND ND ND ND ND ND ND 
HW1 Mκ − − − ND − ND − − − − − 
RL1 Mλ − − ND ND − 
EF16 Mλ − − ND ND − 
a

Shown are the monoclonal cell lines, isotype of secreted Ab, expression of Hibid1 and Hibid2 idiotypes, inhibition of binding to Hib PS by 1.25 mg/ml of soluble E. coli K100 PS (K100), and binding (+, present; −, absent) of mAb to tetanus toxoid (TT), group A streptococcal membranes, type 3 S. pneumoniae capsular PS (PN3), single-stranded DNA (ssDNA), double-stranded DNA (dsDNA), ribonuclear protein (smRNP), cardiolipin, and myosin.

A subset of anti-Hib PS Abs cross-reacts with the structurally similar capsular PS of the K100 serotype of E. coli (7, 21). All six Hib PS-inhibitable mAbs from infants cross-reacted with K100 PS. Three of these mAbs were tested for cross-reactivity with streptococcal membrane Ags, pneumococcal type 3 PS, and tetanus toxoid, and no reactivity was seen. Three of nine mAbs that were not inhibitable by Hib PS cross-reacted with K100 PS, and six of eight tested cross-reacted with streptococcal membranes (Table I).

We previously demonstrated that a small proportion of anti-Hib PS mAb obtained from adults cross-react with ss- and dsDNA (8). Like mAbs obtained from adults, high affinity mAbs from children had little or no cross-reactivity with self-Ags (Table I). In contrast, Hib PS-reactive mAbs from infants that were not inhibitable by soluble Hib PS usually cross-reacted with one or more self Ags, including cardiac myosin, smRNP, DNA, and cardiolipin.

To identify the molecular basis of the anti-Hib PS repertoire of infants we cloned and sequenced IgVH and IgVL encoding the most stable of the mAb cell lines (Tables 2 and 3). Two mAbs that were inhibitable by Hib PS were studied. The mAbs MM1.5 and MM4.1 were obtained from the same subject. These cell lines have identical IgVH genes, encoded by heavy chain variable (VH) segments 100% homologous to the V3–23/DP-47 germline gene, DN1 segments, and germline JH4 genes. IgVL encoding these hybridomas use the VκII A2 germline gene and germline Jκ3 segments and express the Hibid1 Id. Like mAbs from adults, both IgVκ have an arginine residue at the VJ junction. Five nucleotide differences from the A2 germline sequence are shared between MM1.5 and MM4.1, and the former has an additional mutation resulting in a coding change in FR3. The shared mutations suggest that these two hybridomas are clonally related, although we were unable to identify heavy chain or light chain rearrangements by Southern blotting to confirm this (data not shown).

Low affinity, non-Hib PS-inhibitable mAbs of infants are encoded by at least five members of the VH3a gene family: V3–7, V3–21, V3–23, V3–30, and DP-51, and a VH4 family member, V4–34. D segments encoding these IgVH are, in general, long and can be readily compared with known germline genes (DXP1, DIR1/2, and DN1; Table II). In one instance (MAX4) the D is generated by a D-D fusion consisting of a DN1 segment and inverted DIR1 gene (22, 23). In each of these IgVH, nongermline-encoded nucleotides are present at V-D and d-J joints, which may represent N or P addition (24, 25).

Table II.

Heavy chain variable region genes encoding anti-Hib PS mAbs from infantsa

mAbVHTotalFRCDRDJHCDR-3
NAAANAAANAAA
Hib PS inhibitable mAbs           
MM1.5 V3-23/DP-47 100.0 100.0 100.0 100.0 100.0 100.0 DN1 JH4 39 
MM4.1 V3-23/DP-47 100.0 100.0 100.0 100.0 100.0 100.0 DN1 JH4 39 
Noninhibitable mAbs           
MAX1 V3-30/DP-49 96.3 94.9 99.1 100.0 91.3 78.3 DXP-1 JH6 60 
MAX4. V3-21/DP-77 100.0 100.0 100.0 100.0 100.0 100.0 iDIR/DXP1 JH6 99 
MAX5 V3-23/DP-47 99.7 99.0 99.6 98.7 100.0 100.0 DXP-1 JH6 69 
MAX6 DP-51 100.0 100.0 100.0 100.0 100.0 100.0 DN1 JH6 69 
MAX8 ND          
MAX10 ND          
HW1 ND          
RL1 V3-7/DP-54 95.9 91.8 96.9 96.0 92.8 78.3 Dir1/2 JH4 36 
EF16 V4-34/DP-63 100.0 100.0 100.0 100.0 100.0 100.0 JH6 54 
mAbVHTotalFRCDRDJHCDR-3
NAAANAAANAAA
Hib PS inhibitable mAbs           
MM1.5 V3-23/DP-47 100.0 100.0 100.0 100.0 100.0 100.0 DN1 JH4 39 
MM4.1 V3-23/DP-47 100.0 100.0 100.0 100.0 100.0 100.0 DN1 JH4 39 
Noninhibitable mAbs           
MAX1 V3-30/DP-49 96.3 94.9 99.1 100.0 91.3 78.3 DXP-1 JH6 60 
MAX4. V3-21/DP-77 100.0 100.0 100.0 100.0 100.0 100.0 iDIR/DXP1 JH6 99 
MAX5 V3-23/DP-47 99.7 99.0 99.6 98.7 100.0 100.0 DXP-1 JH6 69 
MAX6 DP-51 100.0 100.0 100.0 100.0 100.0 100.0 DN1 JH6 69 
MAX8 ND          
MAX10 ND          
HW1 ND          
RL1 V3-7/DP-54 95.9 91.8 96.9 96.0 92.8 78.3 Dir1/2 JH4 36 
EF16 V4-34/DP-63 100.0 100.0 100.0 100.0 100.0 100.0 JH6 54 
a

Shown are the most closely homologous germline VH segments, percent homology of total genes, framework, and CDR regions to candidate germline genes and their translated amino acid sequences, most closely homologous germline D and JH segments, and length of the CDR-3 region (nucleotides).

The noninhibitable anti-Hib PS mAbs from infants are encoded by eight different Vκ or Vλ segments. Two mAbs from unrelated subjects are encoded by VκI L1/Jκ3 combinations, and two are encoded by VλVII DPL2/Jλ2/3 combinations. Although one mAb uses the VκII A2 gene that dominates the response of adults, this is paired with a V3–21-encoded heavy chain, rather than the V3–23/VκII A2 combination observed in adults. Interestingly, this light chain also lacks a VJ arginine residue and does not express the HibId-1 Id, whereas, to our knowledge, all other VκIIA2-encoded light chains with this residue are HibId-1 positive. The invariant VJ arginine residue observed in anti-Hib PS Ab of adults is, in fact, absent in all non-Hib PS inhibitable mAbs from infants (7). The noninhibitable mAb MAX5 that is encoded by the adult VH segment, V3–23, is paired with a VκIII L6-encoded IgVL.

The homology of VH segments encoding infants’ mAbs to germline genes ranges from 95.9 to 100.0% (Table II). Half are unmutated. The homology of Vκ and Vλ segments to candidate germline segments is also high, ranging from 98.0 to 100.0% (Table III). In both heavy and light chains, however, there is little evidence of Ag-driven somatic mutation (Table IV). Only the DUN1, DUN2, and RL1 IgVH have R:S ratios that exceed the expected ratio of 2.9:1 (26). No IgVL gene has an elevated R:S ratio.

Table III.

Light chain variable region genes encoding anti-Hib PS mAbs from infantsa

mAbVLTotalFRCDRJL
NAAANAAANAAANAAAR
Hib PS inhibitable mAbs            
MM1.5 VκII A2 98.0 96.0 98.6 96.7 96.7 95.7 Jκ3 100.0 100.0 
MM4.1 VκII A2 98.3 97.0 99.0 96.7 96.7 97.1 Jκ3 100.0 100.0 
Noninhibitable mAbs            
MAX1 VκI L1 100.0 100.0 100.0 100.0 100.0 100.0 Jκ3 100.0 100.0 − 
MAX4. VκII A2 97.7 96.0 98.6 97.1 96.7 90.0 Jκ4 90.0 81.8 − 
MAX5 VκIII L6 100.0 100.0 100.0 100.0 100.0 100.0 Jκ1 97.4 100.0 − 
MAX6 VλVII DPL18 100.0 100.0 100.0 100.0 100.0 100.0 Jλ1 100.0 100.0 − 
MAX8 VλI DPL2 98.0 98.0 98.6 97.1 100.0 100.0 Jλ2/3 100.0 100.0 − 
MAX10 VκI L9 99.6 100.0 100.0 100.0 98.7 100.0 Jκ2 100.0 100.0 − 
HW1 VκI L1 99.3 97.9 99.5 98.6 98.7 96.0 Jκ3 97.4 100.0 − 
RL1 VλI DPL2 100.0 100.0 100.0 100.0 100.0 100.0 Jλ2/3 100.0 100.0 − 
EF16 VλI DPL5 99.3 99.0 99.5 100.0 98.9 96.6 Jλ2/3 94.6 91.7 − 
mAbVLTotalFRCDRJL
NAAANAAANAAANAAAR
Hib PS inhibitable mAbs            
MM1.5 VκII A2 98.0 96.0 98.6 96.7 96.7 95.7 Jκ3 100.0 100.0 
MM4.1 VκII A2 98.3 97.0 99.0 96.7 96.7 97.1 Jκ3 100.0 100.0 
Noninhibitable mAbs            
MAX1 VκI L1 100.0 100.0 100.0 100.0 100.0 100.0 Jκ3 100.0 100.0 − 
MAX4. VκII A2 97.7 96.0 98.6 97.1 96.7 90.0 Jκ4 90.0 81.8 − 
MAX5 VκIII L6 100.0 100.0 100.0 100.0 100.0 100.0 Jκ1 97.4 100.0 − 
MAX6 VλVII DPL18 100.0 100.0 100.0 100.0 100.0 100.0 Jλ1 100.0 100.0 − 
MAX8 VλI DPL2 98.0 98.0 98.6 97.1 100.0 100.0 Jλ2/3 100.0 100.0 − 
MAX10 VκI L9 99.6 100.0 100.0 100.0 98.7 100.0 Jκ2 100.0 100.0 − 
HW1 VκI L1 99.3 97.9 99.5 98.6 98.7 96.0 Jκ3 97.4 100.0 − 
RL1 VλI DPL2 100.0 100.0 100.0 100.0 100.0 100.0 Jλ2/3 100.0 100.0 − 
EF16 VλI DPL5 99.3 99.0 99.5 100.0 98.9 96.6 Jλ2/3 94.6 91.7 − 
a

Shown are the most closely homologous germline VL genes, percent homology of total genes, framework and CDR regions to candidate germline genes and their translated amino acid sequences, most closely homologous J segments, and presence of an arginine residue at the VJ joint.

Table IV.

R:S Ratios of anti-Hib PS mAbsa

Cell LineIGVHIGVL
TotalFRCDRTotalFRCDR
MM1.5 0:0 0:0 0:0 4:2 3:0 1:2 
MM4.1 0:0 0:0 0:0 3:2 2:0 1:2 
MAX1.2 5:6 0:2 5:4 0:0 0:0 0:0 
MAX4 0:0 0:0 0:0 4:3 2:2 2:1 
MAX5 1:0 1:0 0:0 0:0 0:0 0:0 
MAX6 0:0 0:0 0:0 0:0 0:0 0:0 
MAX8    0:0 0:0 0:0 
MAX10    0:1 0:1 0:0 
HW1.3    2:0 1:0 1:0 
RL1 9:3 4:3 5:0 2:1 2:1 0:0 
EF16 0:0 0:0 0:0 1:1 0:1 1:0 
Cell LineIGVHIGVL
TotalFRCDRTotalFRCDR
MM1.5 0:0 0:0 0:0 4:2 3:0 1:2 
MM4.1 0:0 0:0 0:0 3:2 2:0 1:2 
MAX1.2 5:6 0:2 5:4 0:0 0:0 0:0 
MAX4 0:0 0:0 0:0 4:3 2:2 2:1 
MAX5 1:0 1:0 0:0 0:0 0:0 0:0 
MAX6 0:0 0:0 0:0 0:0 0:0 0:0 
MAX8    0:0 0:0 0:0 
MAX10    0:1 0:1 0:0 
HW1.3    2:0 1:0 1:0 
RL1 9:3 4:3 5:0 2:1 2:1 0:0 
EF16 0:0 0:0 0:0 1:1 0:1 1:0 
a

Shown are the ratios of replacement and substitution mutations in total genes, framework, and CDR regions.

Serum Abs measured on day 7 following immunization differed between children and adults only in higher IgM responses in adults and a lower affinity of IgG adult serum Abs. Both of these differences probably reflect the fact that children’s sera were obtained following their fourth dose of a TD form of Hib PS, whereas the adults had received only one dose of vaccine. Adult subjects had presumably been primed by exposure to unconjugated Hib or other bacterial PS on bacterial surfaces; thus, their anamnestic responses may be limited.

Attempts to generate hybridomas from infants aged 2 to 4 mo immunized with one or two doses of HbOC vaccine were unsuccessful. Although fusions generated large numbers of immortalized cell lines, few yielded mAbs specific for Hib PS, and none was sufficiently stable for additional characterization (data not shown). Larger numbers of Hib-PS-specific cell lines were obtained from older children aged 15 to 18 mo after four doses of HbOC vaccine, but this was a significantly lower proportion than was obtained from adults. Differences in the frequency of anti-Hib PS-specific mAbs may be an artifact of the fusion process; however, an identical protocol was followed for all age groups, suggesting an intrinsic difference in the numbers of Hib PS-specific circulating lymphocytes. Using a plaque assay to quantitate specific Ab-secreting lymphocytes, Munoz and Insel found between 2.0 and 16.8% of Ig-secreting cells in adults 8 days after immunization with plain Hib PS vaccine were specific for this Ag (27). Barington et al. described two adults immunized with Hib-TT in whom 52.5 and 82.9% of affinity-purified PBL were Hib PS specific (28). In this study 15 to 61% of cell lines from four adults immunized with HbOC bound Hib PS. Although each of these studies used different methodologies, collectively these data suggest that the frequency of Hib PS-specific lymphocytes is higher in adults than in children and higher in response to TD vaccines than to TI formulations.

The variable region gene repertoire of adults immunized with different Hib vaccine formulations has been extensively characterized (7, 8, 29, 30). Most anti-Hib PS mAbs and most serum Ab of adults are encoded by V3–15/DP-38 and V3–23/DP-47 VH segments (8, 10, 31). These IgVH are further characterized by short (mean, 23 bases; range, 9–48 bases) and highly conserved CDR-3 regions (8). Light chain variable region gene use in the adult anti-Hib PS response is more diverse than IgVH use; however, on the average, >50% of serum Ab of adults expresses the HibId-1 Id, suggesting that the VκII A2 variable segment encodes a large proportion of expressed anti-Hib PS Ab (32). VκII A2-encoded IgVL are almost exclusively paired with V3–23/DP-47 encoded IgVH. Almost all anti-Hib PS mAbs obtained from adults have an arginine residue at the VJ joint (7, 8, 33, 34).

Whereas anti-Hib PS mAbs obtained from adults immunized with HbOC were all inhibitable by Ag, almost 30% of mAbs obtained from young children were not, suggesting that Ab secreted by these cell lines has low affinity for Ag. The greater frequency of high affinity Abs noted in adults may be attributable to the rapid proliferation of a pool of memory cells that have undergone affinity maturation as a result of repeated Ag exposure. Variable region genes encoding secondary anti-Hib PS immune responses of both adults and children, however, exhibit little evidence of affinity maturation. A more likely explanation may be that different vaccine formulations elicit distinct, but overlapping, Ab repertoires. Consistent with this is the observation that in adults, the proportion of high affinity mAbs is dependent on vaccine formulation, with low affinity mAbs most prevalent in response to TI plain PS vaccine (our unpublished data). The avidity of serum Ab of adults elicited by different Hib PS conjugate vaccines also varies (13, 35). Low affinity anti-Hib PS mAbs from adults are more likely to be encoded by infrequently expressed variable region genes, suggesting a greater fraction of serum Ab elicited by TI vaccine formulations may be encoded by an unconventional repertoire (our unpublished observation). In young children, the serum expression of certain light chain variable region families and Ids also varies with vaccine formulation, implying that variability in gene usage may be responsible for the functional differences observed in response to alternative TD Ag formulations (12). Our current studies confirm that the Ab repertoire of adults and infants elicited by one of these vaccines, HbOC, is fundamentally different when examined at the level of B cell responses.

The overall anti-Hib PS repertoire of infants is more diverse than that of older subjects. Variable region genes encoding low affinity, noninhibitable anti-Hib PS Ab of infants are encoded by at least six different VH segments. Two of these genes, V3–23/DP-47 and V3–21/DP-77, have been described in the adult anti-Hib PS repertoire (8, 29). In anti-Hib PS mAbs of adults, V3–23/DP-47 is almost invariably paired with light chains encoded by the predominant VκII A2 germline gene, whereas in the noninhibitable mAb MAX5, it is paired with a VκIII L6-encoded light chain. Likewise, the only VκII A2-encoded noninhibitable anti-Hib PS Ab is not paired with V3–23/DP-47, but with V3–21/DP-77. Whereas the IgVH CDR-3 of adult anti-Hib PS mAbs is typically short and highly conserved, the IgVH CDR-3 region of noninhibitable anti-Hib PS mAbs from infants is longer (mean, 64 nucleotides; range, 36–84 bases) and diverse. Notably, the invariant arginine residue at the light chain VJ joint of adult anti-Hib PS IgVL is not observed in any noninhibitable anti-Hib PS mAbs obtained from an infant. Although not formally proven, these data suggest that three molecular features are important for high affinity interaction with Hib PS: heavy and light chain pairing, short IgVH CDR-3 regions, and the presence of an arginine residue at IgVL VJ joints.

The higher affinity anti-Hib PS mAbs of young children immunized with HbOC may also be encoded by a subtly different repertoire from that of adults. Two of these mAbs are encoded by the canonical V3–23/DP-47/VκII A2 combination. The IgVH D segment, however, uses a DN1 gene in germline configuration and is relatively long (39 nucleotides). The limited number of cell lines studied here does not allow us to determine whether the minor differences between these variable region genes and those encoding higher affinity adult anti-Hib PS Abs are a usual feature.

Immunization with HbOC may selectively stimulate a population of B cells that is distinct from those elicited by other Ag formulations and that will remain stable, or over time and repeated Ag exposure it is possible that very small increases in affinity may shape the anti-Hib PS repertoire of these children to more closely resemble that of adults. Consistent with the former hypothesis is the observation of phenotypically similar, but functionally different, memory B cell subpopulations that may differ in responsiveness to secondary challenge with TI or TD Ag formulations in some murine immune responses (36). To date, the anti-Hib PS repertoire of adults has been studied in individuals who were primarily immunized with Hib PS as a result of natural exposure to Hib itself or to cross-reactive bacterial Ags, whereas both primary and subsequent exposures of infants were to a TD formulation (9, 10, 12, 13, 16, 31, 37). Further studies will be required to ascertain whether the anti-Hib PS repertoire of infants primarily immunized with TD Ag formulations evolves to resemble that of subjects with natural immunity or if the early repertoire persists.

Our data suggest that there may be considerable individual differences in response to Hib PS immunization. Other investigators have noted differences in serum Ab concentration and light chain expression elicited by different lots of the same vaccine (13, 37). These differences may result from variability in Ag preparations, diversity in the immunologic maturity of subjects, differing environmental Ag exposure, or intrinsic genetic differences in the Ig repertoire (38).

Compared with that in adult mice, a high frequency of unselected mAbs obtained from neonates expresses poly- and autoreactive Ab (39). The noninhibitable anti-Hib PS Ab of infants may be similar primordial polyreactive Abs (40). The low affinity (noninhibitable) mAb secreted by cell lines from infants are not reflected in serum Ab responses. We previously noted that serum Ab responses may not reflect contributions from all cell lines recovered by fusion of peripheral lymphocytes following immunization. This may occur because cells that contribute the majority of serum Ab may reside in large numbers in lymphoid tissues, whereas the cell lines secreting low affinity, cross-reacting Ab may secrete relatively small amounts of Ab in vivo. Alternatively, or additionally, the binding of low affinity Abs may be inhibited by high affinity Abs in the ELISA.

The relationship between polyreactive low affinity anti-Hib PS Abs and high affinity protective Ig is not clear. It is unlikely that such Igs contribute significantly to protective immunity. It is possible, however, that somatic mutation may increase the affinity for Ag and result in the selection of unconventional anti-Hib PS Abs from low affinity precursors. Although all the adult subjects studied by Silverman and Lucas expressed anti-Hib PS IgVH encoded by VH3–23/DP-47 and/or VH3–15/DP-77, a minority also had small amounts of serum Ab encoded by VH4 and/or VH1 family members; these may be examples of this phenomenon (10).

In summary, we demonstrate that the anti-Hib PS Ab repertoire of infants immunized with HbOC vaccine differs from that of adults. A significant proportion of B cells secrete Ab that binds to Hib PS with low avidity. These Abs often are polyreactive and are encoded by a variable region gene repertoire distinct from that of adults. The variable region gene repertoire of higher affinity anti-Hib mAbs of infants is more similar to the conventional repertoire of adults, but also has subtle differences that may be responsible for the distinct immunologic characteristics of each TD conjugate vaccine. Further studies will be required to determine whether primary exposure to Hib PS in a TD Ag formulation results in the modification of the mature anti-Hib PS repertoire.

We thank the parents and infants who participated in this study, and we are grateful to Drs. Paul Simons and Edward Clark for their enthusiastic support.

1

This work was supported by Grants AI01251 and AI19350 from the National Institute of Allergy and Infectious Diseases, a grant from the Primary Children’s Research Foundation, and U.S. Public Health Service Grant RR00036 to the General Clinical Research Center of Washington University School of Medicine.

3

Abbreviations used in this paper: PS, polysaccharide; TI, T independent; TD, T dependent; Hib PS, Haemophilus influenzae type b polysaccharide; HbOC, Haemophilus influenzae type b oligosaccharide conjugate.

1
Robbins, J. B., J. C. Parke, R. Schneerson, J. C. Whisnant.
1973
. Quantitative measurement of “natural” and immunization-induced Haemophilus influenzae type b capsular polysaccharide antibodies.
Pediatr. Res.
7
:
103
2
Stein, K. E..
1992
. Thymus-independent and thymus-dependent responses to polysaccharide antigens.
J. Infect. Dis.
165
:
S49
3
H., Kayhty, K. V., H. Peltola, P. H. Makela.
1984
. Serum antibodies after vaccination with Haemophilus influenzae type b capsular polysaccharide and response to reimmunization: no evidence of immunologic tolerance or memory.
Pediatrics
74
:
857
4
Insel, R. A., P. W. Anderson.
1986
. Oligosaccharide-protein conjugate vaccines induce and prime for oligoclonal IgG antibody responses to the Haemophilus influenzae b capsular polysaccharide in human infants.
J. Exp. Med.
163
:
262
5
Pabst, H. F..
1980
. Ontogeny of the immune response as a basis of childhood disease.
J. Pediatr.
97
:
519
6
Teale, J. M..
1985
. B cell immune repertoire diversifies in a predictable temporal order in vitro.
J. Immunol.
135
:
954
7
Adderson, E. E., P. G. Shackelford, R. A. Insel, A. Quinn, P. M. Wilson, W. L. Carroll.
1992
. Immunoglobulin light chain variable region gene sequences for human antibodies to Haemophilus influenzae type b capsular polysaccharide are dominated by a limited number of Vκ and Vλ segments and VJ combinations.
J. Clin. Invest.
89
:
729
8
Adderson, E. E., P. G. Shackelford, A. Quinn, P. M. Wilson, M. W. Cunningham, R. A. Insel, W. L. Carroll.
1993
. Restricted immunoglobulin VH usage and VDJ combinations in the human response to Haemophilus influenzae type b capsular polysaccharide-nucleotide sequences of monospecific anti-Haemophilus antibodies and polyspecific antibodies cross-reacting with self-antigens.
J. Clin. Invest.
91
:
2734
9
Scott, M. G., D. L. Crimmins, D. W. McCourt, I. Zocher, R. Thiebe, H. G. Zachau, M. H. Nahm.
1989
. Clonal characterization of the human IgG antibody repertoire to Haemophilus influenzae type b polysaccharide. III. A single VκII gene and one of several JK genes are joined by an invariant arginine to form the most common L chain V region.
J. Immunol.
143
:
4110
10
Silverman, G. J., A. H. Lucas.
1991
. Variable region diversity in human circulating antibodies specific for the capsular polysaccharide of Haemophilus influenzae type b.
J. Clin. Invest.
88
:
911
11
Granoff, D. M., P. G. Shackelford, S. J. Holmes, T. C. V. S. Group, A. H. Lucas.
1993
. Variable region expression in the antibody responses of infants vaccinated with Haemophilus influenzae type b polysaccharide-protein conjugates: description of a new light chain associated idiotype, and the relation between idiotype expression, avidity, and vaccine formulation.
J. Clin. Invest.
91
:
786
12
Lucas, A. H., F. H. Azmi, C. M. Mink, D. M. Granoff.
1993
. Age-dependent V region expression in the human antibody response to the Haemophilus influenzae type b polysaccharide.
J. Immunol.
150
:
2056
13
Chung, G. H., K. H. Kim, R. S. Daum, R. A. Insel, G. R. Siber, S. Sood, R. K. Gupta, C. Marchant, M. H. Nahm.
1995
. The V-region repertoire of Haemophilus influenzae type b polysaccharide antibodies induced by immunization of infants.
Infect. Immun.
63
:
4219
14
Peltola, H. J., H. Kayhty, M. Virtanen.
1984
. Prevention of Haemophilus influenzae type b infections with the capsular polysaccharide vaccine.
N. Engl. J. Med.
310
:
1566
15
Granoff, D. M., P. G. Shackelford, J. P. Pandey, E. G. Boies.
1986
. Antibody responses to Haemophilus influenzae type b polysaccharide vaccine in relation to Km(1) and G2 m(23) immunoglobulin allotypes.
J. Infect. Dis.
154
:
257
16
Adderson, E. E., P. G. Shackelford, A. Quinn, W. L. Carroll.
1991
. Restricted IgH chain V gene usage in the human antibody response to Haemophilus influenzae type b capsular polysaccharide.
J. Immunol.
147
:
1667
17
Granoff, D. M., G. A. Weinberg, P. G. Shackelford.
1988
. IgG subclass response to immunization with Haemophilus influenzae type b polysaccharide outer membrane protein conjugate vaccine.
Pediatr. Res.
24
:
180
18
Tarrand, J. J., M. G. Scott, P. Takes, M. H. Nahm.
1989
. Clonal characterization of the human IgG antibody repertoire to Haemophilus influenzae type b polysaccharide.
J. Immunol.
142
:
2519
19
Cunningham, M. W., K. Krisher, D. C. Graves.
1984
. Murine monoclonal antibodies reactive with human heart and group A streptococcal membrane antigens.
Infect. Immun.
46
:
34
20
Carroll, W. L., K. Theilmans, J. Dilley, R. Levy.
1986
. Mouse × human heterohybridomas as fusion partners with human B cell tumors.
J. Immunol. Methods
89
:
61
21
Insel, R. A., P. W. J. Anderson.
1982
. Cross-reactivity with Escherichia coli K100 in the human serum anticapsular antibody response to Haemophilus influenzae type b.
J. Immunol.
128
:
1267
22
Ichihara, Y., H. Matsuoda, Y. Kurosawa.
1988
. Organization of human immunoglobulin heavy chain diversity gene loci.
EMBO J.
7
:
4141
23
Sanz, I., S.-S. Wang, G. Meneses, M. Fischbach.
1994
. Molecular characterization of human Ig heavy chain DIR genes.
J. Immunol.
152
:
3958
24
Alt, F. W., G. D. Yancopolous, T. K. Blackwell, L. Wood, E. Thomas, M. Boss, R. Coffman, N. Rosenberg, S. Tonegawa, D. Baltimore.
1984
. Ordered rearrangement of immunoglobulin heavy chain variable segments.
EMBO J.
3
:
129
25
Lafaille, J. J., A. DeCloux, M. Bonneville, Y. Takagaki, S. Tonegawa.
1989
. Junctional sequences of T cell receptor ad genes: implications for ad T cell lineages and for a novel intermediate of V-(D)-J joining.
Cell
59
:
859
26
Berek, C., C. Milstein.
1987
. Mutation drift and repertoire shift in the maturation of the immune response.
Immunol. Rev.
96
:
23
27
Munoz, J. L., R. A. Insel.
1987
. In vitro human antibody production to the Haemophilus influenzae type b capsular polysaccharide.
J. Immunol.
139
:
2026
28
Barington, T., L. Hougs, L. Juul, H. O. Madsen, L. P. Ryder, C. Heilmann, A. Svejgaard.
1996
. The progeny of a single virgin B cell predominates the human recall B cell response to the capsular polysaccharide of Haemophilus influenzae type b.
J. Immunol.
157
:
4016
29
Lucas, A. H., J. W. Larrick, D. C. Reason.
1994
. Variable region sequences of a protective human monoclonal antibody specific for the Haemophilus influenzae type b capsular polysaccharide.
Infect. Immun.
80
:
3873
30
Nahm, M. H., K. H. Kim, P. Anderson, S. V. Heatherington, M. K. Park.
1995
. Functional capacities of clonal antibodies to Haemophilus influenzae type b polysaccharide.
Infect. Immun.
63
:
2989
31
Scott, M. G., J. J. Tarrand, D. L. Crimmins, D. W. McCourt, N. R. Siegel, C. E. Smith, M. H. Nahm.
1989
. Clonal characterization of the human IgG antibody repertoire to Haemophilus influenzae type b polysaccharide. II. IgG antibodies contain VH genes from a single VH family and VL genes from at least four VL families.
J. Immunol.
143
:
293
32
Lucas, A. H., D. M. Granoff.
1990
. A major crossreactive idiotype associated with human antibodies to the Haemophilus influenzae b polysaccharide: expression in relation to age and immunoglobulin G subclass.
J. Clin. Invest.
85
:
1158
33
Scott, M. G., M. D. W. C. D. L., R. Theibe, H. G. Zachau, M. H. Nahm.
1989
. Clonal characterization of the human IgG antibody repertoire to Haemophilus influenzae type b polysaccharide. III. A single VκII gene and one of several Jκ genes are joined by an invariant arginine to form the most common L chain V region.
J. Immunol.
143
:
4110
34
Lucas, A. H., R. J. Langley, D. M. Granoff, M. H. Nahm, M. Y. Kitamura, M. G. Scott.
1991
. An idiotypic marker associated with a germline encoded κ light chain variable region that predominates the vaccine-induced human antibody response to the Haemophilus influenzae b polysaccharide.
J. Clin. Invest.
88
:
1811
35
Schlesinger, Y., D. M. Granoff, T. V. S. Group.
1992
. Avidity and bactericidal activity of antibody elicited by different Haemophilus influenzae type b conjugate vaccines.
JAMA
267
:
1489
36
Rennick, D. M., P. R. Morrow, E. Benjamini.
1983
. Two stages of B cell memory development expressing differential sensitivity to stimulation with thymus-dependent and thymus-independent antigenic forms.
J. Immunol.
131
:
567
37
Granoff, D. M., E. L. Anderson, M. T. Osterholm, S. J. Holmes, J. E. McHugh, R. B. Belshe, F. Medley, T. V. Murphy.
1992
. Differences in the immunogenicity of three Haemophilus influenzae type b conjugate vaccines in infants.
J. Pediatr.
121
:
187
38
Kohsaka, H., D. A. Carson, L. Z. Rassenti, W. E. R. Ollier, P. P. Chen, T. J. Kipps, N. Miyasaka.
1996
. The human immunoglobulin VH gene repertoire is genetically controlled and unaltered by chronic autoimmune stimulation.
J. Clin. Invest.
98
:
2794
39
Holmberg, D., S. Forsgren, L. Forni, F. Ivars, A. Coutinho.
1984
. Reactions among IgM antibodies derived from normal neonatal mice.
Eur. J. Immunol.
14
:
435
40
Bouvet, J.-P., G. Dighiero.
1998
. From natural polyreactive autoantibodies for a la carte monoreactive antibodies to infectious agents: is it a small world after all?.
Infect. Immun.
66
:
1