The development of effective receptor-targeted nonviral vectors for use in
vivo is complicated by a number of technical problems. One of these is the
low efficiency of the conjugation procedures used to couple protein ligands
to the DNA condensing carrier molecules. We have made and characterized a
multi-domain protein (SPKR)(4)inv, that is designed to target plasmid DNA t
o beta(1) integrins in remodeling tissue. If contains a nonspecific DNA-bin
ding domain (SPKR)(4), a rigid or-helical linker, and the C-terminal beta(1
) integrin binding domain (aa 793-987) of the Yersinia pseudotuberculosis i
nvasin protein. (SPKR)(4)inv could be purified at high yields using a bacte
rial;expression system. We show that (SPKR),inv binds with high affinity to
both plasmid DNA and beta(1) integrins. in a cell attachment assay, the ap
parent affinity of (SPKR)(4)inv for beta(1) integrins is three orders of ma
gnitude higher than that of the synthetic peptide integrin ligand RGDS. (SP
KR)(4)inv-plasmid complexes are not active in an in vitro transfection assa
y. However, transfection efficiencies of plasmid complexes with a cationic
lipid micelle (DOTAP/Tween-20) or a cationic polymer (polyethylenimine), ar
e significantly increased in combination with (SPKR),inv. (SPKR),inv-mediat
ed transfection can be inhibited by a soluble form of beta(1) integrin, whi
ch is evidence for ifs receptor specificity. In conclusion, (SPKR),inv allo
ws PI integrin-specific targeting of plasmid-carrier complexes, while avoid
ing inefficient and cumbersome coupling chemistry. The modular design of th
e expression vector allows production of similar multi-domain proteins with
a different affinity. The further development of such complexes for use in
vivo is discussed.