The ex vivo gene therapy approach for Duchenne muscular dystrophy is promis
ing since myoblast transplantation in primates is now very efficient. One o
bstacle to this treatment is the low transfection efficiency of large DNA c
onstructs in human primary myoblasts. Small plasmids can be easily transfec
ted with the new phosphonolipid described in this study, However, a dramati
c drop in transfection efficiency is observed with plasmids of 12 kb or mor
e containing EGFP minidystrophin and EGFP dystrophin fusion genes. The tran
sfection of human primary myoblasts with such large plasmids could only be
achieved when the DNA was linked to an adenovirus with the use of polyethyl
enimine (PEI), with efficiencies ranging between 3 and 5% of transitory tra
nsfection. Branched 2 kDa PEI was less toxic in PEI adenofection than branc
hed 25 kDa PEI or linear 22 kDa PEI. The adenovirus was an absolute necessi
ty for an efficient transfection. An integrin-binding peptide, a nuclear lo
calization signal peptide, chloroquine, glycerol or cell cycle synchronizat
ion using aphidicolin did not enhance PEI adenofection. Following PEI adeno
fection, the adenoviral proteins were detected using a polyclonal antibody.
The detected antigens fell below the detectable level after 12 days in cul
ture. We thus provide in this study an efficient and reproducible method to
permit efficient delivery of large plasmids to human primary myoblasts for
the ex vivo gene therapy of Duchenne muscular dystrophy.