Six napin large (L) chains (as well as six napin small chains) were resolved from the seeds of kohlrabi ( Brassica napus var. rapifera) by a procedure involving extraction, batchwise elution from carboxymethylcellulose (CM52) and reversed-phase HPLC after treatment with guanidine hydrochloride and 2-mercaptoethanol. The precise average molecular masses of the circa 4.5 kDa small subunits and the circa 10 kDa large subunits were detern ined by electrospray ionisation mass spectrometry (ESMS). Of six large subunits resolved (LIA, L1B, L1C, L2A, L2B and L2C), the complete amino acid sequences of four (LIA, L2A, L2B and L2C) and the near-complete sequences of two (LIB and LIC) were deduced from the ESMS-based masses of tryptic fragments, Edman sequencing and previously published data. The deduced structures are precisely consistent with this data and with the ESMS-based average molecular masses of these polypeptides. ESMS analysis of unreduced napin extract revealed only seven circa 14.5 kDa complexes, the observed masses being in close agreement with those calculated for 1:1 complexes of particular small and large subunits assuming four disulfides in each napin complex. The structures of the napin large subunits (86–91 residues) are very similar and all amino acid differences observed are confined to only 25 positions. The L2A, L2B and L2C large chains (but not the L1 A, LIB and LIC large chains) are phosphorylated well by plant Ca 2+-dependent protein kinase (CDPK). The CDPK-catalyzed phosphorylation site on the large chain L2A is inferred to be S 57 within the sequence LQQVIS17RIYQT (the site being S 60 within the same sequence in L2B and L2C). The napin-containing basic protein fraction from B. napus seeds largely abolishes the Ca 2+-dependent fluorescence enhancement of dansyl-calmodulin and also inhibits calmodulin (CaM)-dependent myosin light chain kinase (MLCK). The resolved napin long chains also inhibit MLCK. Each kohlrabi large chain contains 2 sequences (corresponding to L1 10-Q 20 and Q 51-L 64 of LIA) which have the potential to form amphipathic α-helices. Each large chain also contains a Q-rich 19 amino acid sequence (corresponding to L 30-Q 48 of LIA) which has the potential to form a ‘2-sided’ α-helix with basic residues confined to one side. These structural elements may be involved in the inferred interaction of these proteins with CaM and may be relevant to the biological activity of antifungal proteins of this kind.
Read full abstract