Molecular characterization of the fibronectin-binding protein BBK32 of Borrelia burgdorferi sensu lato

BBK32, a fibronectin (Fn)-binding protein of Borrelia (B.)
burgdorferi sensu lato (s.l.) which is encoded by the bbk32 gene
located on the 36kb linear plasmid (lp36) of isolate B31, is
playing an important role in serological diagnosis of Lyme
borreliosis. Firstly, we were interested in the genomic
localization of bbk32 regarding different B. burgdorferi s.l.
species as well as between strains of the same species. Southern
blot analyses based on 23 strains of the species B. burgdorferi
sensu stricto (s.s.), B. afzelii, B. garinii and B. spielmanii
revealed that position of bbk32 is rather variable between the
species but also within a given species. bbk32 could be located on
different linear plasmids (lp), mainly on lp23kb, lp24kb, lp25kb,
lp31kb and lp36kb. The meaning of this finding remains unclear so
far. Secondly, a mumber of thirteen chimeric polypeptides
representing different parts of the N-terminal regions of BBK32
proteins of both B. burgdorferi s.s. isolate B31 and B. garinii
isolate PHei were generated. Fn-binding capabilities of those
generated polypeptides were evaluated either by Western-ligand
blot-based binding assay or by enzyme-linked immunosorbent assay
(ELISA)-based binding assay. Results showed that BBK32 from PHei
possesses a higher Fn-binding capability than that from B31.
Furthermore, the higher Fn-binding capacity is associated with four
amino acids (Lysine131, Lysine145, Threonine147 and Isoleucine155)
in the 32-amino acid-long segment (from position 131 to 162).
Moreover, both gelatin and collagen could partially inhibit the
binding of BBK32 to Fn. This suggests that BBK32 might also bind to
the collagen-binding domain of Fn (repeat I6-9 and II1, 2) and
partially to its N-terminal fibrin-binding domain (repeat I1-5).
Though the meaning of the different Fn-binding capacities remains
unclear so far, such studies may provide us with markers to define
the different pathogenic potentials of various Borrelia species and
strains. Thirdly, eight recombinantly prepared BBK32 homologues
(either as partial or as whole) were tested in a line assay to
evaluate their contribution for serologic diagnosis of Lyme
borreliosis. Though BBK32 homologues could react with sera from
Lyme borreliosis patients, compared with other Borrelia-antigens
established in the Max von Pettenkofer Institute, these BBK32
homologues could not improve the sensitivity and specificity of the
class-specific IgG or IgM antibody tests. Nevertheless, this study
underlines the fact that the heterogeneity of Lyme disease Borrelia
species must be taken into consideration in the microbiological
diagnosis of Lyme borreliosis in European patients.