Molecular Immunology 59 (2014) 142–153

Contents lists available at ScienceDirect

Molecular Immunology journal homepage: www.elsevier.com/locate/molimm

Characterization of Escherichia coli K1 colominic acid-specific murine antibodies that are cross-protective against Neisseria meningitidis groups B, C, and Y In Ho Park a , Jisheng Lin b , Ji Eun Choi c , Jeon-Soo Shin d,e,∗ a

Ewha Center for Vaccine Evaluation and Study, Medical Research Institute, School of Medicine, Ewha Womans University, Seoul 158-710, Republic of Korea Department of Pathology, School of Medicine, Division of Laboratory Medicine, University of Alabama at Birmingham, Birmingham, AL 35294, United States c Department of Pediatrics, Seoul National University Boramae Hospital, Seoul National University College of Medicine, Seoul 156-707, Republic of Korea d Department of Microbiology, Yonsei University College of Medicine, Seoul 120-752, Republic of Korea e Severance Biomedical Science Institute and Institute for Immunology and Immunological Diseases, Yonsei University College of Medicine, Seoul 120-752, Republic of Korea b

a r t i c l e

i n f o

Article history: Received 15 January 2014 Received in revised form 24 January 2014 Accepted 27 January 2014 Available online 4 March 2014 Keywords: Neisseria meningitidis, Escherichia coli K1 Capsular polysaccharide N-acetylneuraminic acid O-acetylation, Monoclonal antibody Variable germline gene Structure modeling

a b s t r a c t The capsular polysaccharide (PS) of Neisseria meningitidis serogroup B (NMGB) is ␣(2-8)-linked Nacetylneuraminic acid (Neu5Ac), which is almost identical to the O-acetylated colominic acid (CA) of Escherichia coli K1 Although E. coli K1 has long been known to elicit cross-protective antibodies against NMGB, limited information on these highly cross-reactive antibodies is available. In the present study, six new monoclonal antibodies (mAbs) specific to both E. coli K1 CA and NMGB PS were produced by immunizing Balb/c mice with E. coli K1, and their serological and molecular properties were characterized, together with 12 previously reported hybridoma mAbs. Among the bactericidal mAbs against NMGB, both HmenB5 and HmenB18, which are genetically identical though of different mouse origins, were able to kill serogroup C and Y meningococci. Based on SPR sensograms, the binding affinity of HmenB18 for PS was suggested to be associated with at least two different binding forces: the polyanionicity of Neu5Ac and an interaction with the O-acetyl groups of Neu5Ac. Molecular analysis showed that similar to most mAbs presenting a few restricted V region germline genes, the V region genes of HmenB18 were 979% and 986% identical to the closest IGHV1-1401 and IGLV15-10301 germline gene alleles, respectively, and V-D-J editing in this mAb generated an unusually long VH-CDR3 sequence (17 amino acid residues), containing one basic arginine, two hydrophobic isoleucine residues and a ‘YAMDY’ motif. Models of the mAb combining sites demonstrate that most of the mAbs exhibited a wide, shallow groove with a high overall positive charge, as seen in mAb735, which is specific for a polyanionic helical epitope. In contrast, the combining site of HmenB18 was shown to be wide but to present a relatively weak positive charge, consistent with the extensive recognition by HmenB18 of the various structural epitopes formed with the Neu5Ac residue and its O-acetylation. © 2014 Elsevier Ltd. All rights reserved.

1. Introduction Neisseria meningitidis is one of the most common causes of bacterial meningitis and sepsis (Rosenstein et al., 2001). Based on the capsular polysaccharide (PS) of N. meningitidis, which is one of its major essential virulence factors, the bacterium can be classified

∗ Corresponding author at: Department of Microbiology, Yonsei University College of Medicine, 50-1 Yonsei-ro Seodaemoon-gu, Seoul 120-752, Korea. Tel.: +82 2 2228 1816; fax: +82 2 392 7088. E-mail address: [email protected] (J.-S. Shin). 0161-5890/$ – see front matter © 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.molimm.2014.01.016

into 13 distinct serogroups. Of these 13 serogroups, six (serogroups A, B, C, W-135, X and Y) are responsible for most meningococcal disease around the world (Xie et al., 2013). Because antibodies specific for meningococcal PS are protective against invasive meningococcal diseases (IMD) (Gill et al., 2011; Pace, 2009), meningococcal PS-based vaccines targeting serogroups A, C, W-135, and Y have been developed as a form of meningococcal capsule conjugated to carrier proteins such as tetanus toxoid and CRM197 (Pichichero, 2013). However, meningococcal PS-based vaccines for serogroups B and X are still not available. N. meningitidis serogroup B (NMGB) PS is a homopolymer of ␣(2-8)-linked N-acetylneuraminic acid (Neu5Ac), which is almost

I.H. Park et al. / Molecular Immunology 59 (2014) 142–153

identical to the O-acetylated colominic acid (CA) of Escherichia coli K1 (E. coli K1 CA), a major cause of neonatal meningitis, urinary tract infections and systemic infections (Glode et al., 1977). ␣(28)-linked Neu5NAc is also found as a long chain of polysialic acid (PSA) linked to the neural cell adhesion molecule (N-CAM) in the human fetal stage (Finne et al., 1983). PSA is recognized as a selfantigen by host immune cells, and this self-recognition results in the low immunogenicity of NMGB PS (Wyle et al., 1972). Despite the low immunogenicity of NMGB PS, the fear of any potential immunopathology caused by autoreactive anti-NMGB PS antibodies has hindered the development of PS-based NMGB vaccines (Edelman, 1983). In an attempt to overcome both the low immunogenicity and potential autoimmunity caused by NMGB PS, various alternative antigens have been introduced (Granoff et al., 1998; Shin et al., 2001). An N-propionylated (N-Pr) NMGB PS-conjugated vaccine has been widely studied as an alternative PS antigen for eliciting protective antibodies that are bactericidal against NMGB, but not cross-reactive with human PSA (Flitter et al., 2010; Granoff et al., 1998; Jennings et al., 1987; Moe et al., 2009; Pon et al., 1997). In a previous study, we produced a panel of 12 hybridoma cell lines expressing mAbs specific for NMGB by immunizing mice with E. coli K1 bacteria (Shin et al., 2001). Most of the mAbs were bactericidal against NMGB in vitro, while showing no or decreased cross-reactivity with human PSA expressed on CHP-134 neuroblastoma cells. In particular, it was interesting to find that one mAbs (HmenB5) was cross-reactive with serogroups B, C and Y and killed them through activating rabbit complement. Although NMGB PS is not O-acetylated, other cross-reactive bacteria commonly present high numbers of O-acetylated Neu5Ac residues within their capsular PSs (Bhattacharjee et al., 1976; Claus et al., 2004; Jann and Jann, 1983; Lemercinier and Jones, 1996; Longworth et al., 2002; Orskov et al., 1979) (Fig. 1). Thus, we hypothesized that the extensive cross-reactivity of HmenB5 with various capsular PSs may be dependent on the degree of O-acetylation in the capsular PS. This hypothesis was supported by experimental evidence showing that HmenB5 did not kill meningococcal serogroup W-135, whose capsular PS is less O-acetylated (data not shown) (Claus et al., 2004; Trotter et al., 2012). Anti-PSA antibodies elicited by NMGB PS have been studied through serology and molecular immunology. Among these antibodies, mAb735 is the best characterized in terms of its antigenic specificity and through molecular analyses of its variable (V) region genes, structural crystallography, and three-dimensional modeling (Berry et al., 2005; Frosch et al., 1985; Hayrinen et al., 1989, 2002; Moe et al., 2006; Nagae et al., 2013). Despite of the detailed molecular knowledge of the cross-reactive antibodies elicited by NMGB PS, the molecular properties of the cross-reactive antibodies elicited by E. coli K1 CA are not fully understood. Although the development of a PS-based NMGB PS vaccine has long been hindered by a fear of ␣(2-8)-Neu5Ac as a cause of autoimmune diseases attributed to autoantibodies, this safety issue is controversial due to the lack of clinical evidence of pathology or autoimmune diseases induced by ␣(2-8)-Neu5Ac antibodies, even in neonates and infant mothers who have recovered from IMD (Robbins et al., 2011). In addition to NMGB and E. coli K1, other bacterial species, such as Pasteurella haemolytica A2 and Moraxella nonliquefaciens, have been found to produce ␣(2-8)-Neu5Ac as a virulence factor (Devi et al., 1991; Puente-Polledo et al., 1998). Thus, research on the anti-PSA antibodies elicited by bacterial ␣(2-8)-Neu5Ac must be expanded. In the present study, we produced six new hybridoma cell lines expressing anti-E. coli K1 CA mAbs that are cross-reactive with NMGB PS by immunizing mice with E. coli K1 and characterized them through bactericidal assays, surface plasmon resonance analysis, and a V region germline gene analysis using twelve previously described mAbs (Shin et al., 2001). The results were compared to previously reported data on the anti-PSA antibodies elicited by

143

NMGB PS and used for three-dimensional (3D) structural modeling of mAb combining sites to provide insight into the structural differences among the E. coli K1 CA-induced antibodies specific for serogroup B, C, and Y meningococci. 2. Materials and methods 2.1. Ethics statement All animal handling works were performed in accordance with the Korean Food and Drug Administration (KFDA) guidelines. Protocols were reviewed and approved by the animal IRB at Yonsei University (approval No. 10-082-1). All efforts were made to minimize suffering. 2.2. Bacterial strains and reagents We used various strains of N. meningitidis in this study, from serogroups A (M239), B (ATCC13090), C (BB-305), and W-135 (NCCP15745) as well as three Y strains (S-1975, HF13, and HF74) (Park et al., 2005; Shin et al., 2001). HF13 and HF74 were provided by Dr. Carl E. Frasch of CBER, USFDA (Caldwell et al., 2003). All meningococcal strains were grown on chocolate agar plates in a candle jar. The E. coli K1 RS218 (serotype O18, K1, H7) and K1 EV11 strains were grown in Luria-Bertani (LB) broth or on LB agar plates. All bacteria were aliquoted in Hank’s balanced salt solution (HBSS) containing 20% glycerol and stored at −70 ◦ C before use. Purified capsules of serogroups C and Y and E. coli K1 CA were obtained from Merrell-National Laboratories (Cincinnati, OH (lot No.1964)), CBER of the USFDA (Dr. Carl E. Frasch) and the University of Rochester, NY (Dr. W. Vann), respectively. Colominic acid (10 mg) was treated with 0.2N NaOH at 37 ◦ C for 2 h and then neutralized with 6N HCl for de-O-acetylation. 2.3. Mouse hybridoma cell lines Six new hybridoma cell lines expressing mAbs specific for NMGB PS were generated according to a previously described protocol (Shin et al., 2001). Briefly, BALB/cByJ mice from Jackson Laboratory (Bar Harbor, ME) were intraperitoneally immunized with the E. coli K1 RS218 strain four times: twice with 1 × 107 CFU on days 0 and 4 and twice with 1 × 108 CFU on days 7 and 10. Their spleens were then harvested on day 13 for fusion with Sp2/0-Ag14 myeloma cells. The hybridoma clones were primarily screened via ELISA with E. coli K1 CA, and the cross-reactivity of mAbs with NMGB PS was confirmed through ELISA with heat-killed NMGB (ATCC13090). Of the six new hybridoma clones, five presented immunoglobulin (Ig) M kappa [IgM(␬)] and one presented IgM lambda [IgM(␭)], as shown in Table 1 All mAbs were concentrated with ammonium sulfate, dialyzed in PBS, and fractionated via gelfiltration chromatography with Sephacryl S-300HR (Pharmacia, Uppsala, Sweden) (Shin et al., 2001). 2.4. Complement-mediated bactericidal assay Bactericidal assays were performed in 96-well microtiter plates (Corning, NY) as described previously (Shin et al., 2001). Briefly, 30 ␮L of a bacterial suspension containing 2500 CFU, 50 ␮L of an appropriately diluted antibiotic-free antibody, and 20 ␮L of baby rabbit complement were added to each well. The concentration of complement was 3–20%, depending on the susceptibility of the bacteria to this complement, and an isotype-matched irrelevant mAb was used as a negative control. Serogroups A, B, C, W-135 and Y were employed as target bacteria. After a 1h incubation with shaking, 10 ␮L of the reaction mixture was

144

I.H. Park et al. / Molecular Immunology 59 (2014) 142–153

Fig. 1. Structure of the repeat units of the capsular polysaccharides of the five pathogenic meningococcal serogroups.

plated on a chocolate agar plate. The number of CFU was determined after overnight incubation at 37 ◦ C in a candle jar. Each assay was performed in triplicate, and the percentage of bacterial killing was calculated using the formula [(CFUno antibody − CFUsample )/CFUno antibody ] × 100. To investigate whether the bactericidal activity of HmenB18 against serogroups C and Y meningococci was inhibited by adding each PS from the corresponding bacteria, bactericidal inhibition assay was performed using the free capsules. Various concentrations of each free-PS were added as an inhibitor (10 ␮L) to wells containing the corresponding bacteria (30 ␮L), rabbit complement (10 ␮L), and mAbs (50 ␮L). Bactericidal activities were measured through the same procedure described above, and the percentage of killing inhibition was obtained from following formula: %inhibition = (CFUwith inhibitor − CFUwithout inhibitor )/(CFUwithout antibody − CFUwithout inhibitor ) × 100.

2.5. Purification of capsular PS To purify PS antigens from meningococci groups B and Y and E. coli K1 EV11, bacteria were first grown in 300 mL of Neisseria chemically defined medium (NCDM) in l-L Erlenmeyer flasks (Kenny et al., 1967). The bacterial PS was harvested via precipitation with hexadecyl trimethyl ammonium bromide (cetavlon) as described previously (Bundle et al., 1974). Any contaminating lipopolysaccharide (LPS) was completely removed from the purified PS antigens as described elsewhere (Gotschlich et al., 1981). Briefly, all culture supernatants were precipitated by adding ethanol to a final concentration of 70%, followed by incubation at 4 ◦ C for 72 h, and the resulting precipitate was dialyzed in distilled water and lyophilized. The capsular PS was refined through a series consisting of precipitation with cetavlon buffer (0.2% cetavlon, 0.9% NaCl, 0.02 M EDTA and 0.10 M Tris-HCl, pH 73), ion exchange

Table 1 Bactericidal characteristics of monoclonal hybridoma antibodies.a No.

mAb ID

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

HmenB1 HmenB2 HmenB3 HmenB4 HmenB5 HmenB6 HmenB7 HmenB8 HmenB9 HmenB10 HmenB13 HmenB14 HmenB15 HmenB16 HmenB17 HmenB18 HmenB19 HmenB20

N. meningitidis group (strain) A (M239)

B (ATCC13090)

C (BB306)

Y (S1975)

W-135 (NCCP15745)

− − − − − − − − − − − − − − − − − −

− +++ +++ +++ +++ +++ +++ +++ +++ +++ +++ +++ + + + +++ +++ −

− − − − +++ − − − − − − − − − − +++ − −

− − − − +++b − − − − − − − − − − +++ − +++

ND c ND ND ND − ND ND ND ND ND ND ND ND ND ND − ND ND

Isotype

Fusion No.

Reference

IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␬ IgM, ␭ IgM, ␬

1 2 2 2 2 2 2 3 3 3 4 5 6 6 7 8 9 10

(11) (11) (11) (11) (11) (11) (11) (11) (11) (11) (11) (11) This study This study This study This study This study This study

(a) The data for antibodies HmenB1 through HmenB14 came from a previous report (Shin et al., 2001), while the data for HmenB15 through HmenB20 were obtained in this study. (b) The bactericidal activity of HmenB5 against serogroup Y was retested and corrected. (c) ND indicates ‘not done.’ Bactericidal assays were performed in triplicate, and the mean of the triplicate assays was used for the calculations. , +, ++, and +++ indicate killing levels of 70%, respectively. The percent of killing was calculated as follows: % killing of bacteria = [(CFUirrelevant Ab − CFUmAb )/CFUirrelevant Ab ] × 100. The bactericidal assay was performed using a mAb concentration of 0.5 ␮g/mL and performed at least two times to determine the reproducibility of the results. Baby rabbit serum was used at a concentration of 2–20%, depending on the strains. Rabbit serum did not kill the bacteria to any significant degree either before or after the serum was heat inactivated.

I.H. Park et al. / Molecular Immunology 59 (2014) 142–153

chromatography with DEAE-cellulose, and size-exclusion gel filtration with Sephadex G-200, as described in a previous report. 2.6. Surface plasmon resonance (BIAcore) To analyze the binding of various meningococcal PSs to HmenB18, a BIAcore 3000 instrument (BIAcore AB, Uppsala, Sweden) with a CM5 sensor chip was used, as described previously (Park et al., 2005). The HmenB18 mAb (50 ␮g) was coupled to a CM5 sensor chip using the standard amino coupling kit at a flow rate of 5 ␮L/min. Haemophilus influenzae type b PS (PRP) was employed as a control for non-specific binding because of its lack of crossreactivity with carbohydrate antigens, as specified by the supplier (data not shown). To evaluate binding, each PS was diluted in HBS buffer (0.01 M HEPES/0.15 M NaCl/0.03 M EDTA/0.005% Tween 20, pH 74) and analyzed at various concentrations through passage over the sensor chip at a flow rate of 5 ␮L/min. The response curves for all PS samples were recorded on control surfaces. The resulting data were analyzed using BIAevaluation 30 software (BIAcore AB).

145

Modeling was based on the crystallographic structures available in the Protein Data Bank (http://www.rcsb.org/pdb/home/home.do) (Evans et al., 1995; Nagae et al., 2013). Structural analysis, superimposition, and graphical rendering were carried out with PyMol (Delano Scientific, San Carlos, CA). The root-mean-square deviation (RMSD) of the superimposed proteins was obtained using SuperPose Version 10 (http://wishart.biology.ualberta.ca/SuperPose/). Electrostatic surface potentials were calculated using the web-based APBS program on the PDB2PQR Server (http://nbcr-222ucsd.edu/pdb2pqr 18/), and the results were visualized using the PyMOL Molecular Graphics System. The solvent and protein dielectric constants were set at 78 and 2, respectively. The color scale shown in Fig. 5 ranges from −1 (red) to +5 (blue).

3. Results 3.1. Hybridoma cell lines expressing cross-reactive mAbs specific for NMGB PS

2.7. Molecular analysis of variable region genes of mAbs To analyze the V region genes of the mAbs, total RNA was extracted from hybridoma cells using the RNeasy Mini Kit (Qiagen, Santa Claris, CA) and cDNA was reverse transcribed from cellular RNA using the Superscript RT kit (Life Technologies, Grand Island, NY). One microliter of first strand cDNA was added to PCR buffer containing 125 U of Taq DNA polymerase and 1–2 ␮M MgCl2 , depending on the template, in a total volume of 50 ␮L. The degenerate primers employed to amplify the V region genes of the mAb heavy and light chains were as follows: VH 5 -primer: AGGTC /G A /C AA /G CTGCAG C /G AGTCA /T GG (Orlandi et al., 1989) and VH 3 -primer: GTGAAGGAAATGGTGCTGGGCAGG (Gene bank No. X03690) (Schreier et al., 1986); VL(␬) 5 -primer: GACG /A TCC /A AGATG /A AT /C CCAGT /A CTC /A CA (Kettleborough et al., 1993) and VL(␬) 3 -primer: TGGTGGGAAGATGG (Krawetz et al., 1989); VL(␭) 5 -primer: CCCAGCCCAGCCCATACT and VL(␭) 3 primer: GGCTGG /A CCTAGGACAGT (Solomon and Weiss, 1987). Touch-down PCR was performed for DNA amplification, which consisted of 15 cycles at 94 ◦ C for 1 min, a one degree change in the annealing temperature from 65 ◦ C to 51 ◦ C for 1 min per cycle and 72 ◦ C for 1 min, followed by another 20 cycles with a constant annealing temperature of 50 ◦ C. The obtained DNA PCR products were cloned into the pCRTM 21 vector (Invitrogen, San Diego, CA) using E. coli TOP10 as host bacteria. Plasmids from transformant bacteria selected on LB plates were isolated using the Wizard Plus Miniprep kit (Promega, Madison, WI), and DNA analysis of the cloned V genes was carried out by Macrogen, Inc. (Seoul, Korea). The DNA sequences were edited and aligned using DNAstar software (Lasergene; DNASTAR Inc., Madison, WI). The V region of Ig genes was analyzed using the web-based IMGT/V-QUEST program (http://www.imgt.org/IMGT vquest/share/textes/). Putative germline genes were determined based on the closest matches between the cloned V gene sequences with germline genes and alleles in the IMGT/V-QUEST database. The complementaritydetermining regions (CDRs) of the mAbs were also analyzed based on the closest amino acid sequence matches in the IMGT/VQUEST database. The nucleotide sequences of the V region genes of the mAb heavy and light chains have been deposited in GenBank (accession numbers AF486643-AF486644 and KF956812KF956845). 2.8. 3D Modeling of mAb combining sites To construct 3D structural models of the mAbs, EazyModeller v.40 (BioInforMatikz, bioinformatikzgooglepages.com) was used.

Eighteen hybridoma mAbs are summarized in Table 1. Of these mAbs, 12 (HmenB1 to HmenB14) have been previously reported (Shin et al., 2001). Six new hybridoma clones (HmenB15 to HmenB20) specific for NMGB PS were obtained from five additional BALB/cByJ mice immunized with E. coli K1 (strain RS218). All of the hybridomas expressed IgM antibodies with a ␮ heavy chain and ␬ light chain, except for HmenB19 [␮ heavy and ␭ light chain].

3.2. Bactericidal activities of hybridoma mAbs against meningococcal groups Despite antibody binding to bacterial PSs, some mAbs may not induce complement-mediated bactericidal activity (Granoff et al., 1998; Shin et al., 2001). To determine whether the hybridoma mAbs could provoke cross-protection against pathogenic meningococcal serogroups A, B, C, and Y, the bactericidal activities of the six new mAbs were tested at a fixed concentration (0.5 ␮g/mL) against serogroups A, B, C, and Y in the presence of baby rabbit complement. Additionally, the bactericidal activities of two mAbs (HmenB5 and HmenB18) were tested against serogroup W-135 The results showed that the six mAbs were not able to kill serogroup A to any detectable degree, whereas all of the mAbs except for HmenB20 killed NMGB, showing various ranges of killing % values at a mAb concentration of 0.5 ␮g/mL. In addition to NMGB, HmenB18 was able to kill three C and Y meningococcal groups. The bactericidal titers (defined as the Ab concentration that kills 50% of bacteria in in vitro complement-mediated bactericidal assays) of HmenB18 were