Standardized antigen preparation to achieve comparability of anti-beta2-glycoprotein I assays.

INTRODUCTION: The Sydney classification for diagnosis of the antiphospholipid syndrome (APS) first introduced the determination of anti-beta2-glycoprotein I (anti-beta2-GPI)-antibodies in serum as laboratory criteria. In this context, widely differing results of anti-beta2-GPI assays are a concerning issue. Considerable efforts have been made to optimize ELISAs, however little attention was hitherto spent to the antigen preparation. We evaluated the influence of different beta2-GPI preparations on the ability to separate ill and healthy patients and on the comparability of anti-beta2-GPI-assays. MATERIALS AND METHODS: Microplates were coated with various beta2-GPI preparations and anti-beta2-GPI IgG- and IgM-ELISAs were performed for 21 APS patients and 21 controls using the monoclonal calibrators HCAL and EY2C9. Subsequently, by use of a surface plasmon resonance (SPR) biosensor, affinity constants for the HCAL- and EY2C9-interaction with each beta2-GPI preparation were determined and antigen binding of sera of APS patients and controls was studied. RESULTS: All ss2-GPI preparations showed good discrimination ability ill vs. healthy but poor inter-assay comparability in the ELISAs. Affinity constants for HCAL and EY2C9 were independent of the beta2-GPI variant (K(A) 0.105 - 0.200 and 0.449 - 1.04 x 10(10)M(-1); K(D) 50.0 - 95.5 and 9.61 - 22.3 x 10(-11)M, respectively). In the biosensor, reactivity to the different beta2-GPIs was negligible for the controls and varied considerably for patients. CONCLUSION: Inter-assay comparability of anti-beta2-GPI ELISAs is highly dependent upon the beta2-GPI preparation. Only agreement on one common beta2-GPI preparation will improve the requested inter-assay comparability.

Role of pathogenicity island-associated integrases in the genome plasticity of uropathogenic Escherichia coli strain 536.

The genome of uropathogenic Escherichia coli isolate 536 contains five well-characterized pathogenicity islands (PAIs) encoding key virulence factors of this strain. Except PAI IV(536), the four other PAIs of strain 536 are flanked by direct repeats (DRs), carry intact integrase genes and are able to excise site-specifically from the chromosome. Genome screening of strain 536 identified a sixth putative asnW-associated PAI. Despite the presence of DRs and an intact integrase gene, excision of this island was not detected. To investigate the role of PAI-encoded integrases for the recombination process the int genes of each unstable island of strain 536 were inactivated. For PAI I(536) and PAI II(536), their respective P4-like integrase was required for their excision. PAI III(536) carries two integrase genes, intA, encoding an SfX-like integrase, and intB, coding for an integrase with weak similarity to P4-like integrases. Only intB was required for site-specific excision of this island. For PAI V(536), excision could not be abolished after deleting its P4-like integrase gene but additional deletion of the PAI II(536)-specific integrase gene was required. Therefore, although all mediated by P4-like integrases, the activity of the PAI excision machinery is most often restricted to its cognate island. This work also demonstrates for the first time the existence of a cross-talk between integrases of different PAIs and shows that this cross-talk is unidirectional.