Abstract
Phenoloxidase (PO) is a crucial component of the insect immune response against microbial infection. In the tobacco hornworm Manduca sexta, PO is generated from its precursor proPO by prophenoloxidase activating proteases (PAPs) in the presence of two noncatalytic serine protease homologs (SPHs). cDNA cloning and genome analysis indicate that SPH1a (formerly known as SPH1), SPH1b, SPH4, SPH101, and SPH2 contain a clip domain, a linker, and a protease-like domain (PLD). The first 22 residues of the SPH1b, SPH4, and SPH101 PLDs are identical, and differ from SPH1a only at position 4, Thr(154) substituted with Asn(154) in SPH1a. While the sequence from Edman degradation was used to establish PAP cofactor as a high M(r) complex of SPH1a and SPH2, this assignment needed further validation, especially because SPH1b mRNA levels are much higher than SPH1a's and better correlate with SPH2 transcription. Thus, here we determined expression profiles of these SPH genes in different tissues from various developmental stages using highly specific primers. High levels of SPH1b and SPH2 proteins, low SPH4, and no SPH1a or SPH101 were detected in hemolymph from larvae in the feeding, wandering and bar stages, pupae, and adults by targeted LC-MS/MS analysis, based on unique peptides from the trypsin-treated SPHs. We expressed the five proSPHs in baculovirus-infected Sf9 cells for use as standards to identify and quantify their counterparts in plasma samples. Moreover, we tested their cleavage by PAP3 and efficacy of the SPH1a, 1b, 4, and 101 as SPH2 partners in PAP3-mediated proPO activation. PAP3 processed proSPH1b and 101 more readily than proSPH1a and 4; PAP3 activated proPO more efficiently in the presence of SPH2 with SPH101 or SPH1b than with SPH1a or SPH4. These results generally agree with their order of appearance or sequence similarity: SPH101 > SPH1b (98%) > SPH1a (90%) > SPH4 (83%). In other words, likely due to positive selection, products of the newly duplicated genes (SPH1b and SPH101) are more favorable substrates of PAP3 and better SPH2 partners in forming a high M(r) cofactor than SPH1a or SPH4 is. Electrophoresis on native gel and immunoblot analysis further indicated that SPH101 or 1b form high M(r) complexes more readily than SPH1a or 4 does. In comparison, SPH2 showed a small mobility decrease and then increase on native gel after PAP3 cleavage at the first site. Since the natural cofactor in bar-stage hemolymph is complexes of SPH1 and 2 with an average M(r) of 790 kDa, PAP3-activated SPH2 may associate with the higher M(r) SPH1b scaffolds to form super-complexes. Their structures and formation in relation to cleavage of SPH1b at different sites await further exploration.