Two groups of bacteriophage clones displaying the antigenic properties of serotype 6B pneumococcal capsular polysaccharide (PS) were obtained from different phage libraries expressing random heptameric peptides. One group, biopanned with a mouse mAb (Hyp6BM1), is comprised of 17 phage clones expressing 10 unique sequences of linear peptides. The other group, selected with another mAb (Hyp6BM8), contained six clones, all of which expressed the identical circular peptide. Phage clones expressing the linear peptides (e.g., PhaM1L3) bound only to Hyp6BM1, but not other 6B PS-specific mAb, and their binding could be inhibited with pneumococcal capsular type 6B PS only. In contrast, a phage clone expressing the circular peptide (PhaM8C1) cross-reacted with several other 6B PS-specific mAbs, and their binding could be inhibited with pneumococcal capsular PS of 6A and 6B serotypes. Two short peptides, PepM1L3 and PepM8C1, reflecting the peptide inserts of the corresponding phage clones, could inhibit the binding of the two clones to their respective mAb. Interestingly, the peptide insert in PhaM8C1 was identical to that in PhaB3C4, a previously reported mimotope of α(2→8) polysialic acid, Neisseria meningitidis group B PS. Indeed, PhaM8C1 bound to HmenB3 (a meningococcal Ab), and their association could be inhibited with α(2–8) polysialic acid, but not with 6B PS. Conversely, α(2–8) polysialic acid could not inhibit the binding of PhaM8C1 to Hyp6BM8. The two-dimensional nuclear magnetic resonance studies indicate that PepM8C1 peptide can assume several conformations in solution. The ability of this peptide to assume multiple conformations might account for its ability to mimic more than one Ag type.

As different peptides can assume diverse conformations, a large peptide library expressing random sequences should contain peptides that can conformationally mimic virtually any ligand or Ag. This idea has been shown to have practical application, with the recent demonstration that peptide mimics of antigenic specificities can be efficiently identified using large peptide libraries. For instance, random peptides can be displayed on a coat protein of a bacteriophage, and biopanning can be used to isolate phage clones expressing peptides of a desired binding specificity (1, 2, 3). This approach has been successfully used to identify various antigenic mimics (mimotopes). The mimotopes have been used to examine Ab fine specificity (4), to examine the requirement for B cell stimulation by Ags (5), and to elicit desirable Abs (6).

Peptides mimicking polysaccharide (PS)3 Ags can be readily produced and used various ways. The peptide mimics of PS Ags should be useful for studying differences in immune mechanisms used by PS and protein Ags. In addition, the peptide mimics should, unlike PS, elicit T cell help, immune memory, and strong Ab responses. Also, the peptides should be easier to manufacture and modify than PS Ags (7, 8). Thus, the mimotopes may be useful as vaccine components. However, despite reports of successful mimotope vaccines (4, 9, 10, 11), mimotopes often either are poorly immunogenic, elicit ineffective Abs, or induce B cell memory for ineffective Abs. Consequently, there is a need for further studies of peptide mimics of PS Ags before peptide mimic vaccines can reach their full potential.

Streptococcus pneumoniae is a well-known pathogen, causing several serious diseases in young children and the elderly (12). As a result, Ab responses to pneumococcal PS Ags have been extensively studied in the past. The peptide mimics can be easily compared with other pneumococcal Ags for their immunogenicity and for inducing protective Abs. Also, there is a need for more effective pneumococcal vaccines. The widely available pneumococcal vaccine containing capsular PS from 23 common serotypes (13) is not immunogenic in young children, and its effectiveness is reduced among the elderly (14). A new conjugate vaccine, although effective in young children (15), is not effective among the elderly (16) and is expensive to manufacture. We have studied the peptide mimics of pneumococcal capsular PS and have identified a peptide that can mimic pneumococcal PS as well as meningococcal PS.

HmenB1 and HmenB3 are IgM mAb specific for Neisseria meningitidis group B capsular PS (17). The 101.4.1 is a mAb to N-acetyl-β-D-glucosamine from M. Cunningham (University of Oklahoma, Oklahoma City, OK) (18). Hyp6A1 is a mouse IgM mAb specific for pneumococcal 6A PS. Dob1 is a human IgG2 mAb binding to 6B PS (19). Hyp6BM1, Hyp6BM7, Hyp6BM8, and Hyp6BM10 are mouse IgM mAb specific for pneumococcal 6B PS (20). Dob1, Hyp6BM7, Hyp6BM8, and Hyp6BM10 cross-react with 6A PS, an isopolymer of 6B PS. Hyp6BM1 does not cross-react with 6A PS, and Hyp6A1 does not cross-react with 6B PS (20). Hyp6BM1 and Hyp6BM8 were used for biopanning, and the mAb were purified from the mouse ascites by (NH4)2SO4 precipitation and by chromatography over a column of Sephacryl S-300HR (Pharmacia Biotech, Uppsala, Sweden).

Two phage libraries from New England Biolabs (Beverly, MA) were used for our study. One contained linear peptides composed of 7 random aa, and the other contained circular peptides of 9 aa; circularization was achieved by a covalent bond between the cysteines at positions 1 and 9, and the 7 aa between the two cysteines are randomly chosen. Biopanning the phage libraries was performed as described (17). Briefly, 60-mm petri dishes (Nunc, Roskilde, Denmark) were coated with mAb at a concentration of 100 mg/L in 0.1 M NaHCO3 (pH 8.6). The mAb-coated petri dishes were blocked with 0.5% BSA in 0.1 M NaHCO3 and washed with 0.1% TBST (50 mM Tris-Cl (pH 7.5), 150 mM NaCl). For each biopanning cycle, ∼2 × 1011 PFU phage were placed in the petri dish and incubated for 30 min. Following the removal of unbound phage particles, bound phage particles were eluted from the dish with 0.2 M glycine-HCl (pH 2.2), and the recovered phage solution was neutralized with 1 M Tris-Cl (pH 9.1). The recovered phage was expanded in number by growing with Escherichia coli for a new cycle of biopanning. The biopanning was performed for three or four cycles before individual phage clones were isolated.

To determine the DNA sequence of the mimotope peptide, phage clones were expanded by growing them in 1 ml E. coli cultures for 4–5 h at 37°C. The bacteria present in the culture were removed by a brief centrifugation (10,000 × g for 30 s). A total of 500 μl bacteria-free culture supernatant was then mixed with 200 μl 20% PEG/2.5 M NaCl solution, and the mixture was centrifuged (10,000 × g) for 10 min to precipitate the phage as a pellet. The pellet was isolated and resuspended with 100 μl Tris buffer with iodide (10 mM Tris-Cl (pH 8), 1 mM EDTA, 4 M NaI) and 250 μl absolute ethanol. The phage DNA was then washed with 70% ethanol, dried, and resuspended in 30 μl TE buffer (50 mm Tris and 10 mm EDTA, pH8). A total of 5 μl DNA suspension was subjected to dideoxy termination reaction using DNA sequencing kit (PerkinElmer, Norwalk, CT). The sequencing primer are position −96 from New England Biolabs, and AmpliTaq DNA polymerase FS. The sequence of the mimotope was obtained by running the above reaction products through an automated DNA sequencer from PE Applied Biosystems (Foster City, CA).

A sandwich type ELISA was performed, as described below. Hyp6BM1 or Hyp6BM8 mAb was absorbed on 96-well microtiter plates (Nunc) at a concentration of 10 mg/L in carbonate-bicarbonate buffer (pH 9.6). The plates coated with Ab were blocked with 2% skim milk in PBS after washing, and serially diluted phage clones (ranging from 6 × 1010 to 3.7 × 106 PFU/well) were added to the wells and incubated for 1.5 h at room temperature. After the unbound phage particles were washed, peroxidase-labeled anti-phage mAb (Pharmacia Biotech) was added. Bound peroxidase was quantitated with tetramethyl benzidine (Kirkegaard & Perry Laboratories, Gaithersburg, MD) substrate.

The sandwich type assay was modified in some cases, as below. To evaluate cross-reactive binding of the phage clones to other mAbs, the plates were coated with HmenB3, Hyp6A1, Hyp6BM7, Hyp6BM10, or 101.4.1. To determine the specificity of the binding of the phage particles to mAb, phage particles (1 × 109 PFU/well) were added to the wells in the presence of varying concentrations of inhibitors. Used as inhibitors were synthetic peptides or PS. The PS used as inhibitors include E. coli K1 PS, pneumococcal cell wall polysaccharide, capsular PS of Hemophilus influenzae type b, and capsular PS of S. pneumoniae serotypes 1, 3, 7F, and 23F. E. coli K1 PS is α(2, 3, 4, 5, 6, 7, 8)-linked polysialic acid and identical in structure to N. meningitidis group B capsular PS (21, 22).

The peptides used for this study were NLpeptide6 (HSACTTPGPWFC), NLpeptide10 (YHSNIKFNPPG), NLpeptide9 (HSACTGPGSWFCG), NLpeptide11 (CHSHYHKFG), and NLpeptide12 (YSACTTPGPWFC). NLpeptide6, NLpeptide9, and NLpeptide12 were circularized at two cysteine sites. NLpeptide10 and NLpeptide12 were also referred to as PepM1L3 and PepM8C1, respectively. The core sequence of the peptides was identified with an underline and derived from the peptide inserts of PhaB3C1 (17), PhaM1L3, PhaM1L9, or PhaM8C1. The amino acid Y or C at the N-terminal was added for the purpose of radiolabeling or conjugation with carrier protein. The remaining amino acids in either sides of the core peptide were derived from the sequence of the phage protein pIII flanking the peptide inserts. These peptides were synthesized by Biosynthesis (Lewisville, TX) or Biomolecules Midwest (Waterloo, IL).

Solutions (2.5 mM) of NLpeptide6 and NLpeptide9 were prepared in PBS. The 1-H TOCSY spectra of both peptides were obtained on a 500-MHz Varian INOVA nuclear magnetic resonance (NMR) spectrometer (Varian, Palo Alto, CA) using a standard pulse sequence (23, 24). Data were processed and analyzed using FELIX 97 software (Accelrys, San Diego, CA).

Biopanning of the phage library displaying the linear peptides with Hyp6BM1-coated petri dishes readily enriched for the Ab-specific phage clones and 17 phage clones were chosen randomly after three or four cycles of biopanning. Biopanning the same phage library with Hyp6BM8-coated petri dish did not enrich for the Ab-binding clones. When a second phage library expressing the circular peptide was biopanned, 6 clones binding to the plates were selected at random. In addition to these 23 clones selected by panning, a phage clone was chosen from each library (PhaM8-CNTL and PhaM1-CNTL) without any selection to use as negative controls. PhaM1-CNTL expresses the linear peptide and PhaM8-CNTL, a circular peptide. Binding patterns of select phage clones were shown in Fig. 1.

FIGURE 1.

The amount of bacteriophage bound to the ELISA wells (y-axis) vs the number of phage particles in each well (x-axis). PhaM1L3 (▪), PhaM1L9 (•), or PhaM1-CNTL (□) phage clones were placed in Hyp6BM1-coated wells. PhaM8C1 (▴) or PhaM8-CNTL (▵) phage clones were placed in Hyp6BM8-coated wells. PhaM1-CNTL and PhaM8-CNTL do not bind to either Ab and were used as negative controls. The amount of phage particles bound to the wells was determined with HRP-conjugated anti-phage Ab, as described in Materials and Methods.

FIGURE 1.

The amount of bacteriophage bound to the ELISA wells (y-axis) vs the number of phage particles in each well (x-axis). PhaM1L3 (▪), PhaM1L9 (•), or PhaM1-CNTL (□) phage clones were placed in Hyp6BM1-coated wells. PhaM8C1 (▴) or PhaM8-CNTL (▵) phage clones were placed in Hyp6BM8-coated wells. PhaM1-CNTL and PhaM8-CNTL do not bind to either Ab and were used as negative controls. The amount of phage particles bound to the wells was determined with HRP-conjugated anti-phage Ab, as described in Materials and Methods.

Close modal

All 23 biopanned clones bound to the selection Abs in a dose-dependent manner, and they did not bind to the Abs in the presence of 6B PS in solution (data not shown). The data for two biopanned clones (PhaM1L3 and PhaM8C1) are shown in Fig. 2. Fig. 2,A shows the data for PhaM1L3; Fig. 2 B shows the data for PhaM8C1. The inhibition of binding was almost complete in the presence of 0.5 mg/L 6B PS, but it was not demonstrable even in the presence of 100-fold more unrelated PS. Interestingly, PhaM8C1, but not PhaM1L3, was partially inhibited with 6A PS, which is an isopolymer of 6B PS. These findings indicate that all 23 biopanned clones bound to the Ab in an Ag-specific manner.

FIGURE 2.

The amount of phage bound to the microwells in the presence of 50 mg/L (filled bars) or 0.5 mg/L (hatched bars) pneumococcal capsular PS of different serotypes. The serotypes were indicated for each group at the x-axis. Pneumococcal cell wall polysaccharide and H. influenzae type b PS were used as irrelevant PS controls. A, Microwells were coated with Hyp6BM1 and had 1010 PFU PhaM1L3 phage particles. B, Microwells were coated with Hyp6BM8 and contained 109 PFU PhaM8C1.

FIGURE 2.

The amount of phage bound to the microwells in the presence of 50 mg/L (filled bars) or 0.5 mg/L (hatched bars) pneumococcal capsular PS of different serotypes. The serotypes were indicated for each group at the x-axis. Pneumococcal cell wall polysaccharide and H. influenzae type b PS were used as irrelevant PS controls. A, Microwells were coated with Hyp6BM1 and had 1010 PFU PhaM1L3 phage particles. B, Microwells were coated with Hyp6BM8 and contained 109 PFU PhaM8C1.

Close modal

The binding specificity of clones PhaM1L3 and PhaM8C1 was examined in greater detail by testing their binding to other pneumococcal Abs. PhaM1L3 demonstrably bound only to Hyp6BM1, the mAb used for the selection of PhaM1L3, and did not bind several other mAbs binding 6B PS. Thus, PhaM1L3 appears to be specific for a unique (idiotypic) determinant of the Ab used to select the clone. In contrast, PhaM8C1 bound to various anti-6B Abs. Although PhaM8C1 was selected with Hyp6BM8, it actually bound better to the anti-6B mouse mAb Hyp6BM7 and Hyp6BM1 than to Hyp6BM8. It did not bind to Dob1 (data not shown), which is a human mAb to 6B PS (19). PhaM8C1 also failed to bind to an unrelated Ab, 101.4.1. Thus, PhaM8C1 is broadly reactive with several mouse anti-6B Abs, but its binding is selective.

Although the 23 phage clones were isolated from separate phage plaques, it is expected that many clones are duplicate clones sharing the clonal origin and expressing the identical peptide sequence. To identify the duplicate clones, the nucleotide sequence of the inserted DNA of all the cloned phages was determined and the DNA sequences were translated into amino acid sequences (Table I). As expected, many clones were found to have the identical sequences. Among the phage clones obtained with Hyp6BM1, five clones (29%) expressed the sequence SHHKFSP, and PhaM1L1 clone was chosen as the representative of the five clones for additional studies. Two more linear sequences, represented by PhaM1L2 and PhaM1L3 clones, were also found within multiple clones. All six phage clones obtained with Hyp6BM8 expressed the identical sequence CTTPGPWFC, and PhaM8C1 clone was chosen as the representative clone.

Table I.

Mimotope peptide sequences of Hyp6BM1 and Hyp6BM8

mAbPhageaPeptide SequencebNo. of Clonesc
Hyp6BM1 PhaM1L1 hsSHHKFSPggg 5 /17 
 PhaM1L2 hsWNPKFHWggg 3 /17 
 PhaM1L3 hsNIKFNPPggg 2 /17 
 PhaM1L4 hsNPKFSPPggg 1 /17 
 PhaM1L5 hsNWKFAPPggg 1 /17 
 PhaM1L6 hsNWRFHDTggg 1 /17 
 PhaM1L7 hsHFHYFSPggg 1 /17 
 PhaM1L8 hsHSHYHKFggg 1 /17 
 PhaM1L9 hsHYHKFggg 1 /17 
 PhaM1L10 hsSMDPRFYggg 1 /17 
Hyp6BM8 PhaM8C1 hsacTTPGPWFcggg 6 /6 
mAbPhageaPeptide SequencebNo. of Clonesc
Hyp6BM1 PhaM1L1 hsSHHKFSPggg 5 /17 
 PhaM1L2 hsWNPKFHWggg 3 /17 
 PhaM1L3 hsNIKFNPPggg 2 /17 
 PhaM1L4 hsNPKFSPPggg 1 /17 
 PhaM1L5 hsNWKFAPPggg 1 /17 
 PhaM1L6 hsNWRFHDTggg 1 /17 
 PhaM1L7 hsHFHYFSPggg 1 /17 
 PhaM1L8 hsHSHYHKFggg 1 /17 
 PhaM1L9 hsHYHKFggg 1 /17 
 PhaM1L10 hsSMDPRFYggg 1 /17 
Hyp6BM8 PhaM8C1 hsacTTPGPWFcggg 6 /6 
a

The names of phage clones with linear (L) and circular (C) random peptides.

b

Sequences of clones from heptapeptide insert phage library after third biopanning by Hyp6BM1 and Hyp6BM8 are shown. Flanking sequences are shown in small letters and inserted sequences are shown in capital letters.

c

Number of identical peptide sequences vs number of isolated clones.

A totally unexpected observation was that the PhaM8C1 insert sequence is identical to the insert sequence found for the phage clone PhaB3C4, which was obtained previously with a mAb against NMGB PS (17). Consistent with the sequence data, the PhaM8C1 clone was observed to bind to Abs to meningococcal group B PS as well as Abs to pneumococcal 6B PS (see Fig. 4).

FIGURE 4.

The amount of phage bound to the microwells in the presence of varying concentrations of 6B PS (•) and E. coli K1 PS (○) in solution. PhaM8C1 phage was added to Hyp6BM8 (A)- and HmenB3 (B)-coated microwells.

FIGURE 4.

The amount of phage bound to the microwells in the presence of varying concentrations of 6B PS (•) and E. coli K1 PS (○) in solution. PhaM8C1 phage was added to Hyp6BM8 (A)- and HmenB3 (B)-coated microwells.

Close modal

In addition to the identical sequences of some of the clones, there are also shared amino acid sequence motifs among different clones. NXKF motif was found in seven nonidentical phage clones (41%). Dimeric motif KF was very common and was found in 14 clones. In two additional clones, K, of the KF motif, was replaced with R, another cationic amino acid. HS motif is also present in the viral protein III adjoining the amino end of the peptide insert (Table I). This HS in the viral protein may participate in the formation of the epitope of PhaM1L9 because it is selected by PhaM1L8 (HSHYHKF) as a part of the peptide insert. Phenylalanine was in all phage clones. Proline is very common, and two proline residues are found in several mimotopes, including PhaM8C1.

Although the heptameric peptide inserts are most likely responsible for the binding of the phage particles to the pneumococcal Abs, it is possible that other parts of the phage protein may also be critical to the binding. To directly investigate the role of the peptide inserts in binding, three peptides were chemically synthesized and were examined for their ability to inhibit the binding of the Abs to their Ags. Peptides did not inhibit the binding of Abs to 6B PS Ags (data not shown) perhaps because the PS bound to the Ab too strongly for the peptides to displace it from the Ab. However, all three peptides (NLpeptide10, NLpeptide11, NLpeptide12) that were examined inhibited the phage clones from binding to the mAb used for their biopanning in a dose-dependent manner (Fig. 3). These data, along with the fact that 6B PS can inhibit the binding of the phage clones, suggest that the peptide inserts are directly involved in the binding.

FIGURE 3.

The amount of phage bound to Ab-coated microwells in the presence of varying concentrations of synthetic peptides in solution. The synthetic peptides, identified for each experimental group at the x-axis, were present at 1 mM (open bars), 0.1 mM (gray bars), or 0 mM (filled bars). PhaM1L3, PhaM1L9, or PhaM8C1 phage clones were used respectively for the experimental group, using PepM1L3 (NLpeptide10), NLpeptide11, or PepM8C1 (NLpeptide12) peptides. PepM1L3 and NLpeptide11 peptides were added to Hyp6BM1-coated microwells, and PepM8C1 peptides were added to Hyp6BM8-coated wells.

FIGURE 3.

The amount of phage bound to Ab-coated microwells in the presence of varying concentrations of synthetic peptides in solution. The synthetic peptides, identified for each experimental group at the x-axis, were present at 1 mM (open bars), 0.1 mM (gray bars), or 0 mM (filled bars). PhaM1L3, PhaM1L9, or PhaM8C1 phage clones were used respectively for the experimental group, using PepM1L3 (NLpeptide10), NLpeptide11, or PepM8C1 (NLpeptide12) peptides. PepM1L3 and NLpeptide11 peptides were added to Hyp6BM1-coated microwells, and PepM8C1 peptides were added to Hyp6BM8-coated wells.

Close modal

The finding that the peptide insert of PhaM8C1 is identical to a meningococcal mimotope identified with HmenB3, a mAb against N. meningitidis group B PS, was unexpected. One explanation for the above finding is that HmenB3 and Hyp6BM8 may have the identical or very similar V region structures. To examine the possibility, we determined the DNA sequence of the V regions of HmenB3 and Hyp6BM8, and their sequences were deposited in GenBank. The accession numbers are: AF486641 for VH of Hyp6BM8, AF486642 for VL of Hyp6BM8, AF486643 for VH of HmenB3, AF486644 for VL of HmenB3. Hyp6BM8 VH region is derived from a J606 family VH gene and JH1 gene, whereas HmenB3 VH region is formed with a J558 VH gene and JH4 gene.

The VL region is formed with a Vκ1 family gene for Hyp6BM8 and a Vκ2 family gene for HmenB3. In addition, there are differences in DH regions and somatic mutations between the two Abs. Thus, the V regions of the two mAb display large differences and show little similarity.

We next examined the binding of PhaM8C1 to HmenB3 and Hyp6BM8. PhaM8C1 did bind to HmenB3-coated plates readily (Fig. 4), but it did not bind to another meningococcal Ab, HmenB1 (data not shown). Its binding to HmenB3 was inhibitable with E. coli K1 PS, but not with an excess amount of 6B PS (Fig. 4,A). E. coli K1 PS is chemically identical to N. meningitidis group B PS (25). Conversely, PhaM8C1 clone bound to Hyp6BM8, and its binding was efficiently inhibited with 6B PS, but not even with a large amount of E. coli K1 capsular PS (Fig. 4 B). This observation showed that PhaM8C1 clone does bind to pneumococcal as well as meningococcal Abs in an Ag-specific manner.

The inhibition studies described above suggest that distinct structures are binding the different mAb reactive with PhaM8C1. One way this could happen would be if PhaM8C1 can express two or more conformations, with each expressing a different epitope. To examine this possibility, we examined two circular peptides, NLpeptide6 and NLpeptide9, with two-dimensional (2-D) NMR spectroscopy. NLpeptide6 is based on PhaM8C1 insert, and NLpeptide9 is a control peptide that is also circular. To simplify the identification of the major conformations of the two peptides, we examined the proton NMR TOCSY patterns of the single tryptophan found in both peptides. TOCSY cross-peaks between the ring HN (chemical shift range, ∼10–10.3 ppm) and Hd (chemical shift range, ∼7.1–7.3 ppm) are shown in Fig. 5. For NLpeptide9, three Trp ring HN-Hd cross-peaks are observed with the intensity ratio of 4:1:0.25 (Fig. 5, upper panel). One cross-peak is significantly more intense than the others, indicating that one conformation is favored for NLpeptide9. In contrast, NLpeptide6 shows three strong Trp HN-Hd cross-peaks (Fig. 5, lower panel) with the intensity ratio of 2:1:1, indicating the existence of three well-populated forms. TOCSY peak patterns for other residues in both peptides also indicate the presence of multiple conformations present in similar relative amounts. The existence of multiple conformers was shown not to be a result of peptide aggregation; the relative intensities of resonances assigned to different conformers were found to be invariant over a range of peptide concentrations (0.15–3.4 mM) for both NLpeptide6 and NLpeptide9. Rather, multiple conformers may arise from proline cis-trans isomerization: the proline resonances show the largest chemical shift differences among conformers, and the analysis of 2-D ROESY data detects both NMR patterns of cis- and trans-forms of proline (L. Zhong and K. L. Bren, unpublished observations). Taken together, we conclude that cis-trans-isomerization of proline residues gives rise to more well-populated conformers for NLpeptide6, which has two proline residues, than for NLpeptide9, which has only one proline.

FIGURE 5.

Proton TOCSY 2-D NMR cross-peaks between the ring HN (x-axis) and δH (y-axis) of the single tryptophan present in the two peptides. Contour lines indicate the magnitude of the cross-peaks.

FIGURE 5.

Proton TOCSY 2-D NMR cross-peaks between the ring HN (x-axis) and δH (y-axis) of the single tryptophan present in the two peptides. Contour lines indicate the magnitude of the cross-peaks.

Close modal

By biopanning phage libraries with two IgM mAbs against S. pneumoniae type 6B PS (20), we have produced two groups of phage clones binding to anti-6B Abs. Their binding is specific because the phage clones examined did not bind to irrelevant IgM Abs and their binding was inhibited by pneumococcal capsular PS of 6B serotype, but not by pneumococcal PS of unrelated serotypes. Furthermore, the peptide insert was directly involved in the binding because synthetic peptide can inhibit the binding of the phage clones to the two mAbs (Fig. 3). Thus, peptide inserts expressed by our phage clones are the mimotopes for 6B PS.

Although both groups of phage clones bind to the Abs in an Ag-specific manner, the two groups are clearly different in their binding characteristics. PhaM8C1 clone binds to Hyp6BM8 as well as other anti-6B Abs, whereas the linear PhaM1L3 (the linear mimotopes) binds only to Hyp6BM1, the mAb used for the selection. Perhaps the linear mimotope can assume multiple conformations (NMR data not shown), and only a very small fraction of the linear mimotope can bind to the Ab. Consequently, the binding can be demonstrated only with the Ab used for the biopanning. In contrast, PhaM8C1 is a circular peptide that can assume only a limited number of conformations and may behave like a hapten expressing the 6B PS epitope. Because Hyp6BM8 binds to both 6A and 6B PS, PhaM8C1 may express the epitope shared by both 6A and 6B PS. This appeared to be the case because both 6A and 6B PS could also inhibit the binding of the PhaM8C1 to Hyp6BM8.

When the sequences of the individual mimotopes of Hyp6BM1 were determined, areas of consensus were readily recognizable. KF was expressed on 82.4% of the linear mimotopes. In some cases, K is replaced with R, another positively charged amino acid. KF often appears in association with NX, suggesting NXKF is a broader consensus sequence for the linear peptide mimotopes. These consensus sequences were not previously known, and the consensus sequences already associated with PS mimotopes, such as WXY, were not observed. When single amino acids are examined, proline is found to be common: all but three mimotopes have proline, and five mimotopes have two prolines. In addition, our sequence contained a large number of aromatic amino acids. F could be found in all phage clones. PhaM8C1 clone has two, and PhaM1L2 and PhaM1L7 clones have three aromatic amino acids. Aromatic amino acids have been noted for various PS mimotopes of group C meningococcal PS (26), group B streptococcal type III capsular PS (27), and other PS (26, 28). Taken together, these results indicate that our mimotope sequences are distinct from other PS mimotopes.

An unexpected finding was that PhaM8C1 is identical in sequence to a meningococcal mimotope, PhaB3C4 (17). PhaB3C4 phage clone was obtained in our prior studies with a mAb against NMGB capsular PS, α(2, 3, 4, 5, 6, 7, 8)-linked polysialic acid (17). PhaM8C1 clone was found to bind to both mAb in an Ag-specific manner, indicating that the sequence identity was not based on trivial technical accidents (e.g., contamination in the phage clone). Also, this observation is not based on the similarity in the two PS molecules or two mAbs. No serological relationship has been noted between the two PS, even though these two PS have been extensively studied serologically for their pathogenic importance. The two mAbs were formed with totally different V region gene families, and their complementarity-determining regions were found to be quite distinct when the DNA sequences of the VL and VH regions of the two Abs were determined.

The explanation is most likely based on the peptide mimotope itself. For instance, PhaB3C4 may express multiple epitopes at all times. The circular mimotope is relatively small and would behave like a hapten expressing one epitope. Nevertheless, it may be possible that the top side of the circular mimotope may bind one Ab, but the bottom side of the mimotope may bind another. Alternatively, the mimotope may express multiple conformations over time by rapidly switching from one conformation to another. Although the circular mimotope forms a small molecular ring with little conformational freedom, PhaM8C1 mimotope has two proline residues that can slowly (in 10–100 s) switch between cis and trans conformations. This is consistent with the fact that proline is often important in the overall conformation of a protein molecule (29). PhaM8C1 is shown to have multiple conformations in our NMR studies, although one conformation is preferred over the other. We propose that one conformation seen in the NMR pattern mimics pneumococcal PS, and the other mimics meningococcal PS. While we believe that the two major conformations are involved in the reactions, it is possible the third conformation may be stabilized by the Ab, becomes the dominant conformation in the presence of Ab, and participates in the binding. Further studies are necessary to examine these possibilities.

We believe that peptides with proline residues may not be so desirable as mimotope vaccines because proline can assume multiple semistable conformations. This belief is reinforced by the observation that a short peptide containing several proline residues (PPPGMRPP) can elicit autoantibodies binding Sm and nRNP (30, 31). Immediately following the immunization with the peptide, the animals initially produce Abs binding the peptide. In the later phase of immunization, the animals produce Abs to multiple epitopes found in Sm and nRNP molecules. If the animals have the proper genetic background, the immunized animals may develop systemic lupus erythematosus disease. Although it is unclear how this Ag-spreading phenomenon occurs, the existence of distinct, semistable conformations associated with the presence of multiple proline residues in the peptide may contribute to this phenomenon.

Over the last several years, peptide mimotopes have been studied extensively for their usefulness as a vaccine. While there are successful examples reported in the literature, there are also studies showing limitations in this approach. For instance, mimotope vaccines may produce Abs binding the native Ag, but without protective function. These peptides could lead to the induction of inappropriate memory B cells. Our study now suggests that the mimotope vaccines should be tested for the induction of unintended Abs. Although peptide mimotopes are a promising new approach for vaccines, their immunological properties should be carefully investigated before their use as vaccines.

We thank Dr. D. Briles for a careful reading and W. Bartlett for secretarial assistance.

1

This work is supported by National Institutes of Health Grant AI-31473 (to M.H.N.) and a Howard Hughes Biomedical Research Support Program grant (to K.L.B.).

2

Address correspondence and reprint requests to Dr. Moon H. Nahm, Department of Pathology, School of Medicine, Division of Laboratory Medicine, University of Alabama, 845 19th Street South, BBRB 614, Birmingham, AL 35294. E-mail address: nahm@uab.edu

3

Abbreviations used in this paper: PS, polysaccharide; 2-D, two-dimensional; NMR, nuclear magnetic resonance.

1
Cwirla, S. E., E. A. Peters, R. W. Barrett, W. J. Dower.
1990
. Peptides on phage: a vast library of peptides for identifying ligands.
Proc. Natl. Acad. Sci. USA
87
:
6378
2
Scott, J. K., G. P. Smith.
1990
. Searching for peptide ligands with an epitope library.
Science
249
:
386
3
Lane, D. P., C. W. Stephen.
1993
. Epitope mapping using bacteriophage peptide libraries.
Curr. Opin. Immunol.
5
:
268
4
Valadon, P., G. Nussbaum, L. F. Boyd, D. H. Margulies, M. D. Scharff.
1996
. Peptide libraries define the fine specificity of anti-polysaccharide Abs to Cryptococcus neoformans.
J. Mol. Biol.
261
:
11
5
Kouskoff, V., S. Famiglietti, G. Lacaud, P. Lang, J. E. Rider, B. K. Kay, J. C. Cambier, D. Nemazee.
1998
. Antigens varying in affinity for the B cell receptor induce differential B lymphocyte responses.
J. Exp. Med.
188
:
1453
6
Beenhouwer, D. O., P. Valadon, R. May, S. L. Morrison, M. D. Scharff.
2000
. An evolutionary phage display sublibrary reveals a peptide mimotope of the polysaccharide capsule of Cryptococcus neoformans that elicits a high titer protective response in mice.
FASEB J.
14
:
A947
7
Chargelegue, D., O. E. Obeid, S.-C. Hsu, M. D. Shaw, A. N. Denbury, G. Taylor, M. W. Steward.
1998
. A peptide mimic of a protective epitope of respiratory syncytial virus selected from a combinatorial library induces virus-neutralizing Abs and reduces viral load in vivo.
J. Virol.
72
:
2040
8
Stoute, J. A., W. R. Ballou, N. Kolodny, C. D. Deal, R. A. Wirtz, L. E. Lindler.
1995
. Induction of humoral immune response against Plasmodium falciparum sporozoites by immunization with a synthetic peptide mimotope whose sequence was derived from screening a filamentous phage epitope library.
Infect. Immun.
63
:
934
9
Young, A. C., P. Valadon, A. Casadevall, M. D. Scharff, J. C. Sacchettini.
1997
. The three-dimensional structures of a polysaccharide binding Ab to Cryptococcus neoformans and its complex with a peptide from a phage display library: implications for the identification of peptide mimotopes.
J. Mol. Biol.
274
:
622
10
Rudolf, M. P., M. Vogel, F. Kricek, C. Ruf, A. W. Zurcher, R. Reuschel, M. Auer, S. Miescher, B. M. Stadler.
1998
. Epitope-specific Ab response to IgE by mimotope immunization.
J. Immunol.
160
:
3315
11
Meloen, R. H., W. C. Puijk, J. W. Slootstra.
2000
. Mimotopes: realization of an unlikely concept.
J. Mol. Recognit.
13
:
352
12
Fedson, D. S., D. M. Musher.
1994
. Pneumococcal vaccine. S. A. Plotkin, and E. A. Mortimer, eds.
Vaccines
517
Saunders, Philadelphia.
13
Robbins, J. B., R. Austrian, C. J. Lee, S. C. Rastogi, G. Schiffman, J. Henrichsen, P. H. Makela, C. V. Broome, R. R. Facklam, R. H. Tiesjema, J. C. Parke, Jr.
1983
. Considerations for formulating the second-generation pneumococcal capsular polysaccharide vaccine with emphasis on the cross-reactive types within groups.
J. Infect. Dis.
148
:
1136
14
Shapiro, E. D., A. T. Berg, R. Austrian, D. Schroeder, V. Parcells, A. Margolis, R. K. Adair, J. D. Clemens.
1991
. The protective efficacy of polyvalent pneumococcal polysaccharide vaccine.
N. Engl. J. Med.
325
:
1453
15
Shinefield, H. R., S. Black.
2000
. Efficacy of pneumococcal conjugate vaccines in large scale field trials.
Pediatr. Infect. Dis. J.
19
:
394
16
Shelly, M. A., H. Jacoby, G. L. Riley, B. T. Graves, M. Pichichero, J. J. Treanor.
1997
. Comparison of pneumococcal polysaccharide and CRM197 conjugated pneumococcal oligosaccharide vaccines in young and elderly adults.
Infect. Immun.
65
:
242
17
Shin, J. S., J. S. Lin, P. W. Anderson, R. A. Insel, M. H. Nahm.
2001
. Monoclonal Abs specific for Neisseriameningitidis group B polysaccharide and their peptide mimotopes.
Infect. Immun.
69
:
3335
18
Shikhman, A. R., N. S. Greenspan, M. W. Cunningham.
1994
. Cytokeratin peptide SFGSGFGGGY mimics N-acetyl-β-d-glucosamine in reaction with Abs and lectins, and induces in vivo anti-carbohydrate Ab response.
J. Immunol.
153
:
5593
19
Sun, Y., M. K. Park, J. Kim, B. Diamond, A. Solomon, M. H. Nahm.
1999
. Repertoire of human Abs against the polysaccharide capsule of Streptococcus pneumoniae serotype 6B.
Infect. Immun.
67
:
1172
20
Sun, Y., Y. Hwang, M. H. Nahm.
2001
. Avidity, potency, and cross-reactivity of monoclonal Abs to pneumococcal capsular polysaccharide serotype 6B.
Infect. Immun.
69
:
336
21
Kasper, D. L., J. L. Winkelhake, W. D. Zollinger, B. L. Brandt, M. S. Artenstein.
1973
. Immunochemical similarity between polysaccharide antigens of Escherichia coli 07: K1(L):NM and group B Neisseriameningitidis.
J. Immunol.
110
:
262
22
Krambovitis, E., M. B. McIllmurray, P. A. Lock, H. Holzel, M. R. Lifely, C. Moreno.
1987
. Murine monoclonal Abs for detection of antigens and culture identification of Neisseriameningitidis group B and Escherichia coli K-1.
J. Clin. Microbiol.
25
:
1641
23
Bax, A., D. G. Davis.
1985
. MLEV-17-based two dimensional homonuclear magnetization transfer spectroscopy.
J. Magn. Reson.
65
:
355
24
Griesinger, C., G. Otting, K. Wuthrich, R. R. Ernst.
1988
. Clean TOCSY for H-1 spin system identification in macromolecules.
J. Am. Chem. Soc.
110
:
7870
25
Frosch, M., I. Gorgen, G. J. Boulnois, K. N. Timmis, D. Bitter-Suermann.
1985
. NZB mouse system for production of monoclonal Abs to weak bacterial antigens: isolation of an IgG Ab to the polysaccharide capsule of Escherichia coli K1 and group B meningococci.
Proc. Natl. Acad. Sci. USA
82
:
1194
26
Westerink, M. A. J., P. C. Giardina, M. A. Apicella, T. Kieber-Emmons.
1995
. Peptide mimicry of the meningococcal group C capsular polysaccharide.
Proc. Natl. Acad. Sci. USA
92
:
4021
27
Pincus, S. H., M. J. Smith, H. J. Jennings, J. B. Burritt, P. M. Glee.
1998
. Peptides that mimic the group B streptococcal type III capsular polysaccharide antigen.
J. Immunol.
160
:
293
28
Hoess, R., U. Brinkmann, T. Handel, I. Pastan.
1993
. Identification of a peptide which binds to the carbohydrate-specific monoclonal Ab B3.
Gene
128
:
43
29
Simon, I..
1993
. Predictions from the regularities of the primary structure of proteins.
Pept. Res.
6
:
260
30
James, J. A., T. Gross, R. H. Scofield, J. B. Harley.
1995
. Immunoglobulin epitope spreading and autoimmune disease after peptide immunization: Sm B/B′-derived PPPGMRPP and PPPGIRGP induce spliceosome autoimmunity.
J. Exp. Med.
181
:
453
31
James, J. A., J. B. Harley.
1998
. B-cell epitope spreading in autoimmunity.
Immunol. Rev.
164
:
185