Recombinant laminin B1 chains exhibit intact short-arm domains but do not form oligomeric molecules

The human laminin B1 chain has been produced in the baculovirus expression system in sufficient amounts for biochemical and functional studies. A full-length cDNA, which was constructed of four partially overlapping clones and verified by in vitro transcription and translation to be functional, was cloned into the transfer vector pVL1392 behind the polyhedrin promoter. The recombinant construct was incorporated by in vivo homologous recombination into the genome of the wild-type baculovirus, Autographa californica nuclear polyhedrosis virus. Infection of Spodoptera frugiperda cells (Sf9) with the recombinant virus resulted in the expression of the recombinant B1 chain (recB1) in these insect cells. The recB1 was found to be synthesized in two forms with apparent molecular masses of 220 kDa and 200 kDa. The 220-kDa form is an N-glycosylated form of recB1, because it was not present in cultures containing tunicamycin, an inhibitor of N-linked glycosylation. The recB1 accumulated inside the cell and only a small portion of it was secreted into the culture medium. Thus purification had to be started from the cell extract in order to obtain reasonable amounts of the protein. About 500 micrograms was obtained from a 500-ml culture with three steps of chromatography, concavalin A, DEAE-Sepharose and Mono Q anion-exchange chromatography. Only the glycosylated form was purified. The recB1 was found to be sensitive to degradation during the purification, because two proteolytic forms of about 180 kDa were present in every preparation. The accumulation of recB1 inside the cell was possibly due to the lack of correct assembly. Electron microscopy studies showed that the short arm part had a native or near-native structure, but the C-terminal heptad repeat domain had not folded correctly and did not exist in an alpha-helical structure, as it does in native laminin. Electron microscopy and cross-linking studies further revealed that recB1 was a monomeric protein. It was also shown to be unable to oligomerize in vitro, suggesting that the B1 chain is not designed to form homo-oligomers. Finally, cell attachment assays were carried out, but the native recB1 appeared to be inactive in these assays.