The mannan-binding lectin (MBL) pathway of complement activation is part of the innate immune defense. The binding of MBL to microbial carbohydrates activates the MBL-associated serine proteases (MASPs), which recruit the complement factors, C4 and C2, to generate the C3 convertase or directly activate C3. We present a phylogenetically highly conserved member of the MBL complex, MASP-3, which is generated through alternative splicing of the MASP-1/3 gene. The designation of MASP-3 as a protease is based on homology to known MASPs. Different MBL oligomers were found to have distinct MASP composition and biological activities. MASP-1, MAp19, and direct C3-cleaving activity are associated with smaller oligomers whereas MASP-3 is found together with MASP-2 on larger oligomers. MASP-3 downregulate the C4 and C2 cleaving activity of MASP-2.

No abstract

Mannan-binding lectin (MBL)-associated serine proteases (MASP-1, -2, and -3) are homologous modular proteases that each associate with MBL and L- and H-ficolins, which are oligomeric serum lectins involved in innate immunity. To investigate its physicochemical, interaction, and enzymatic properties, human MASP-3 was expressed in insect cells. Ultracentrifugation analysis indicated that rMASP-3 sedimented as a homodimer (s20,w = 6.2 ± 0.1 S) in the presence of Ca2+, and as a monomer (s20,w = 4.6 ± 0.1 S) in EDTA. As shown by surface plasmon resonance spectroscopy, it associated with both MBL (KD = 2.6 nM) and L-ficolin (KD = 7.2 nM). The protease was produced in a single-chain, proenzyme form, but underwent slow activation upon prolonged storage at 4°C, resulting from cleavage at the Arg430-Ile431 activation site. Activation was prevented in the presence of protease inhibitors iodoacetamide and 1,10-phenanthroline but was not abolished upon substitution of Ala for the active site Ser645 of MASP-3, indicating extrinsic proteolysis. In contrast, the corresponding mutations Ser627→Ala in MASP-1 and Ser618→Ala in MASP-2 stabilized the latter in their proenzyme form. Likewise, the MASP-1 and MASP-2 mutants were each activated by their active counterparts, but MASP-3 S645A was not. Activated MASP-3 did not react with C1 inhibitor; had no activity on complement proteins C2, C4, and C3; and only cleaved the N-carboxybenzyloxyglycine-l-arginine thiobenzyl ester substrate to a significant extent. Based on these observations, it is postulated that MASP-3 activation and control involve mechanisms that are different from those of MASP-1 and -2.

No abstract