Over hundred genes encoding allergens have been isolated so far and re
combinant (r)allergens have been produced. To evaluate the usefulness
for their in vivo or in vitro application a thorough analysis of these
molecules has to be performed by comparing the function of (r)allerge
ns in vitro and in vivo to the natural (n) counterpart. We have perfor
med such an analysis using the main bee venom allergen phospholipase A
2 (PLA) as a model system. Detailed knowledge about the structure, the
enzymatic activity as well as the antigenicity/allergenicity of this
molecule is available. We thus produced rPLA with or without enzymatic
activity in its correctly folded or denatured form and compared these
products to the natural counterpart. In addition, we have produced en
zymatically inactive but correctly folded PLA using molecular biologic
al techniques. We show that enzymatic active or inactive correctly ref
olded rPLA has very similar characteristics compared to purified natur
al PLA when assayed in vitro by serological or enzymatic assays and in
vivo as determined by skin test activity. In contrast, incorrectly fo
lded rPLA had no in vitro or in vivo activity. In particular incorrect
ly folded rPLA did not bind human monoclonal or polyclonal antibody of
IgG or IgE and consequently did not induce positive type I skin react
ivity. The data show the importance of comparatively analysing the nat
ural and (r)allergens for the intrinsic nature of allergens such as en
zymatic activity or binding to their natural substrate. Both natural o
r rPLA bound equally well to synthetic phospholipid bilayers indicatin
g their similar binding capacity to the natural substrate the cell sur
face of eukaryotic or prokaryotic cells. Hence, a thorough in vitro an
alysis comparing the natural to the (r)allergen has to precede the in
vivo use. The prospects and limitations of (r)allergens for in vitro o
r in vivo diagnostics and potentially for immunotherapy is discussed.