the recombinant E. coli protein. For heme detection experiments, samples were prepared using a non-reducing sample buffer and were not heated prior to SDSPAGE, which was performed under denaturing conditions. Bacterial lysates of the wild type, the msp22 gene deletion mutant and the complemented strains all served 
as additional heme controls. Positive signals were obtained for hemoglobin, purified Msp22 and the cell lysates expressing Msp22, indicating the presence of heme-dependent peroxidase activity. No signal was detected for the negative control protein BSA. The 9 Protective Moraxella catarrhalis Antigens ID MCR_0063 MCR_0076 MCR_0136 MCR_0186 MCR_0196 MCR_0439 MCR_0560 MCR_0681 MCR_0686 MCR_0691 MCR_0692 MCR_0739 MCR_0918 MCR_0996 MCR_1003 MCR_1010 MCR_1228 MCR_1303 MCR_1357 MCR_1416 MCR_1690 MCR_1742 MCR_1761 Annotation hypothetical protein TonB-dependent receptor hypothetical protein outer membrane lipoprotein LolB MltB; lytic murein transglycosylase Pbp1A; penicillin-binding protein 1A hypothetical protein putative lytic transglycosylase peptide methionine sulfoxide reductase MsrA/MsrB hypothetical protein hypothetical protein hemoglobin utilization protein MhuA M16-like peptidase hypothetical protein LysM BMS-833923 domain protein DacC; D-alanyl-D-alanine carboxypeptidase D15 surface antigen family protein OppA; oligopeptide ABC transport system substrate binding protein Cyt1; cytochrome c1 family protein cytochrome c class II Msp22 extracellular solute-binding protein family 3 outer membrane protein OlpA; OPA-like protein A aa 232 913 278 190 473 786 355 303 558 105 503 954 470 146 819 386 907 679 241 152 262 1975694 111 235 GD 47/47 44/47 47/47 46/47 47/47 47/47 44/47 45/47 47/47 46/47 45/47 46/47 46/47 47/47 47/47 47/47 47/47 47/47 47/47 47/47 44/47 46/47 47/47 Hits 8 13 2 6 12 5 5 7 3 4 7 15 5 3 9 48 4 6 22 2 6 24 7 1 + 2 + 3 4 PP OM PP 5 2 2 + 2 + + 2 + + + + + + + + 2 2 + 2 + + + 2 + + IM IM IM + + + + + + + + + + + + CP + + OM PP OM PP + + + + + OM PP + PP + OM aa, amino acids; GD, gene distribution; 1, Proteins detected in the whole membrane preparation; 2, Proteins detected in outer membrane vesicles; 3, Proteins detected in outer membrane vesicles; 4, Bioinformatic analysis, predicted localization using PSORTb3.0.3, OM = outer membrane, PP = periplasmic, IM = inner membrane, CP = cytoplasmic, = unknown; 5, Peptide ELISA; selected for in vivo studies. doi:10.1371/journal.pone.0064422.t004 absence of the respective protein band at 17 kDa in the msp22 gene deletion mutant, the presence of a strong signal in the complemented strain, and a weak signal in 17496168 the wild type strain suggested that the 17 kDa hemoprotein was indeed Msp22. Discussion Over the last three decades, M. catarrhalis has become recognized as an important pathogen of the human respiratory tract. However, even though M. catarrhalis is the third most frequent bacterial pathogen to be associated with otitis media and is a major cause of exacerbations of COPD in adults, none of the currently available bacterial vaccines developed to prevent these diseases include M. catarrhalis antigens. Therefore, the aim of this study was to comprehensively identify potential vaccine targets of M. catarrhalis by applying the ANTIGENome technology that had previously been developed by Intercell AG, and which had been previously successfully used for vaccine discovery for several other bacterial pathogens. Genomic libraries displaying multiple epitopes of all potential antigens of isolate BBH18
