Teratology 2002 Jan;65(1):26-37
Spence S, Vetter C, Hagmann WK, Van Riper G, Williams H, Mumford RA, Lanza TJ, Lin LS, Schmidt JA.
Developmental and Reproductive Toxicology, Merck Research Labs, West Point, Pennsylvania 19486.
Pharmacological antagonism of VLA-4 (Very Late Antigen 4, [alpha](4)[beta](1) integrin) has become an attractive target for the treatment of predominantly eosinophil mediated disease states such as asthma, allergic rhinitis, multiple sclerosis, rheumatoid arthritis, diabetes, and inflammatory bowel disease. Gene knockouts of the [alpha](4)-integrin subunit of VLA-4 or its cell surface ligand, VCAM-1, however, have been shown to result in embryo-lethality in homozygous null mice due to defects in chorio-allantoic or epi-myocardial fusion. Although gene knockout phenotypes are not always manifested by pharmacological antagonism, those studies suggested that VLA-4 antagonists might cause embryo-lethality or drug-induced malformations.
To test these concepts, early neurulating rat embryos were cultured by the methods of New ('78) after intra-coelomic microinjection of a VLA-4 blocking antibody or in the presence of small molecule VLA-4 antagonists.
Defects in chorio-allantoic fusion were induced after microinjection of VLA4 blocking antibody and after continuous exposure to small molecule antagonists. In a minority of affected embryos chorio-allantoic fusion was completely blocked whereas the majority of affected embryos had only superficial chorio-allantoic fusion and the allantois was enlarged and edematous. Although the allantoic mesoderm covered the trophoblasts of the chorionic plate and contained blood vessels there was only minimal invasion of the trophoblasts by the allantoic mesoderm. The lowest observed effect level generally correlated with the IC([similar]95), as determined in 90% plasma.
Based on these data, VLA-4 antagonism
might represent a significant risk to the developing embryo/fetus. In vitro
exposure, however, is "constant" and does not take into account the elimination
phase of these xenobiotics in vivo. Given the high concentrations required
to elicit an effect, therapeutic blood levels in vivo may be several fold
lower than those that affect the conceptus, depending on the tissue penetration
of the compound and the route of administration. VLA-4 also exists in a
range of conformations and activation states in vivo and the gene KOs and
present studies do not define whether these developmental processes are
dependent upon a particular activation state of VLA-4. Therefore, state-selective
antagonists may have an improved embryonic safety profile. Additional studies
will be required to determine potential effects of VLA-4 antagonists on
embryo/fetal development in vivo.
Copyright 2002 Wiley-Liss, Inc