pneumonia

EuroPneumo Special Issue / pneumonia 2015 Oct 21;7:I–72
P2.43
Novel recombinant glycoconjugate vaccines for prevention of
pneumococcal meningitis
Charlie Plumptre 1 , Emily Kay 2 , James Paton 3 , Brendan Wren 2 , Jeremy Brown 1
1
Division of Medicine, University College London, UK; 2 Department of Pathogen Molecular Biology, London School of Hygiene and Tropical Medicine,
UK; 3 Research Centre for Infectious Diseases, University of Adelaide, Australia
The pneumococcus is a major cause of meningitis around the world. Current vaccination strategies suffer from serious
limitations in terms of cost and serotype coverage, and novel approaches are required to overcome these challenges.
In this project, we are aiming to produce vaccines comprised of conserved pneumococcal protein antigens conjugated
to capsular polysaccharide from serotype 4 pneumococci by making use of an unusual oligosaccharyltransferase from
Campylobacter jejuni named PglB. This enzyme is able to conjugate glycans containing an acetamido group in the C2
position of their reducing end sugar to proteins containing a specific amino acid sequence (D/E-Y-N-X-S/T). We have
introduced this ‘glycotag’ sequence into 4 pneumococcal protein vaccine candidates: α-glycerophosphate oxidase
(GlpO), neuraminidase A (NanA), and the ATP-binding cassette transporters PiuA and Sp0148. All 4 of these proteins
have previously shown efficacy as vaccines in animal models of colonisation or meningitis caused by Streptococcus
pneumoniae. Previous work has allowed the expression of pneumococcal serotype 4 capsule in Escherichia coli, and
by co-expressing PglB and the target protein in the same strain, we have been able to demonstrate production and
small scale purification of protein antigens conjugated to the polysaccharide. Our current efforts are directed towards
optimisation of the efficiency of the conjugation process through modifying parameters such as the leader sequence
of the protein, the plasmid used to encode its gene, the background strain of E. coli and the growth conditions used for
expression.
P2.44
Broadly cross-reactive antibodies recognising the proline-rich region
of pneumococcal surface protein A variants show cross-reactivity with
skeletal muscle
Zoltan Magyarics 1 , Harald Rouha 1 , Adriana Badarau 1 , Nels Nielson 2 , Marisa Caccamo 1 , Susanne
Weber 1 , Barbara Maierhofer 1 , Katharina Havlicek 1 , Ivana Dolezilkova 1 , Karin Gross 1 , Eszter Nagy 1
1
Arsanis Biosciences, Vienna, Austria, 2 Adimab LLC, Lebanon, NH, USA
Pneumococcal surface protein A (PspA) is an important surface-expressed virulence factor of Streptococcus pneumoniae.
PspA displays high level of sequence variability in clinical isolates but contains a conserved proline-rich region at the
C-terminus. PspA was shown to be protective in animal models and has been considered as an attractive vaccine
antigen. One of the concerns with PspA is the anecdotal detection of serum antibodies with potential cross-reactivity to
human muscle in vaccines. The antigen responsible for inducing tissue cross-reactive antibodies was assumed to be the
N-terminal coiled coil region of PspA. We selected 4 human IgG antibodies from a yeast-based antibody library utilising
all 5 clade variants of PspA as baits. The resulting mAbs were tested for binding to native PspA by flow cytometry-based
surface staining of 61 S. pneumoniae clinical isolates. Poly-reactivity of the mAbs was tested against cell membrane
extract of Chinese hamster ovary (CHO) cells as well as against porcine myosin. We identified monoclonal antibodies
binding to the majority of tested S. pneumoniae clinical isolates, expressing clades 1 to 5 of PspA. Confirming the
specificity of the mAbs, they do not bind to S. pneumoniae D39 ΔpspA/ΔpspC strain. These antibodies were determined
to react with a PAPAPKP consensus peptide motif of the proline-rich region of PspA. However, we observed significant
binding to CHO cell extract and myosin in ELISA, suggesting that breadth of binding to PspA variants correlated with nonspecific
reactivity towards mammalian cells and specifically, skeletal muscle. Amino acid homology search identified the
presence of the PAPAPKP motif and its variants in several human proteins, including myosin, cell adhesion, and nuclear
proteins. In conclusion, broadly cross-reactive PspA monoclonal antibodies targeting the proline-rich region of PspA can
bind live bacteria, but they raise concern about off-target tissue cross-reactivity related to this region of PspA.
pneumonia 2015 Volume 7
46