Open Access Open Badges Research article

Discrimination of multilocus sequence typing-based Campylobacter jejuni subgroups by MALDI-TOF mass spectrometry

Andreas Erich Zautner12*, Wycliffe Omurwa Masanta12, Abdul Malik Tareen1, Michael Weig1, Raimond Lugert1, Uwe Groß1 and Oliver Bader1

Author Affiliations

1 UMG-Labor/Institut für Medizinische Mikrobiologie, Universitätsmedizin Göttingen, Kreuzbergring 57, 37075, Göttingen, Germany

2 UMG-Labor/Institut für Klinische Chemie - Zentrallabor, Universitätsmedizin Göttingen, Robert-Koch-Straße 40, 37075, Göttingen, Germany

For all author emails, please log on.

BMC Microbiology 2013, 13:247  doi:10.1186/1471-2180-13-247

Published: 7 November 2013



Campylobacter jejuni, the most common bacterial pathogen causing gastroenteritis, shows a wide genetic diversity. Previously, we demonstrated by the combination of multi locus sequence typing (MLST)-based UPGMA-clustering and analysis of 16 genetic markers that twelve different C. jejuni subgroups can be distinguished. Among these are two prominent subgroups. The first subgroup contains the majority of hyperinvasive strains and is characterized by a dimeric form of the chemotaxis-receptor Tlp7m+c. The second has an extended amino acid metabolism and is characterized by the presence of a periplasmic asparaginase (ansB) and gamma-glutamyl-transpeptidase (ggt).


Phyloproteomic principal component analysis (PCA) hierarchical clustering of MALDI-TOF based intact cell mass spectrometry (ICMS) spectra was able to group particular C. jejuni subgroups of phylogenetic related isolates in distinct clusters. Especially the aforementioned Tlp7m+c+ and ansB+/ ggt+ subgroups could be discriminated by PCA. Overlay of ICMS spectra of all isolates led to the identification of characteristic biomarker ions for these specific C. jejuni subgroups. Thus, mass peak shifts can be used to identify the C. jejuni subgroup with an extended amino acid metabolism.


Although the PCA hierarchical clustering of ICMS-spectra groups the tested isolates into a different order as compared to MLST-based UPGMA-clustering, the isolates of the indicator-groups form predominantly coherent clusters. These clusters reflect phenotypic aspects better than phylogenetic clustering, indicating that the genes corresponding to the biomarker ions are phylogenetically coupled to the tested marker genes. Thus, PCA clustering could be an additional tool for analyzing the relatedness of bacterial isolates.

MALDI-TOF species identification; Phyloproteomics; Multilocus sequence typing; MLST; Intact cell mass spectrometry; ICMS; Principal component analysis; PCA; Campylobacter jejuni