Dense deposit disease (DDD) is a severe renal disease characterized by accumulation of electron-dense material in the mesangium and glomerular basement membrane. Previously, DDD has been associated with deficiency of factor H (fH), a plasma regulator of the alternative pathway (AP) of complement activation, and studies in animal models have linked pathogenesis to the massive complement factor 3 (C3) activation caused by this deficiency. Here, we identified a unique DDD pedigree that associates disease with a mutation in the C3 gene. Mutant C3 923ΔDG , which lacks 2 amino acids, could not be cleaved to C3b by the AP C3-convertase and was therefore the predominant circulating C3 protein in the patients. However, upon activation to C3b by proteases, or to C3(H 2 O) by spontaneous thioester hydrolysis, C3 923ΔDG generated an active AP C3-convertase that was regulated normally by decay accelerating factor (DAF) but was resistant to decay by fH. Moreover, activated C3b 923ΔDG and C3(H 2 O) 923ΔDG were resistant to proteolysis by factor I (fI) in the presence of fH, but were efficiently inactivated in the presence of membrane cofactor protein (MCP). These characteristics cause a fluid phase-restricted AP dysregulation in the patients that continuously activated and consumed C3 produced by the normal C3 allele. These findings expose structural requirements in C3 that are critical for recognition of the substrate C3 by the AP C3-convertase and for the regulatory activities of fH, DAF, and MCP, all of which have implications for therapeutic developments. IntroductionComplement is a major component of innate immunity, with crucial roles in microbial killing, apoptotic cell clearance, immune complex handling, and modulation of adaptive immune responses. Complement is activated by 3 independent activation pathways: the classical pathway (CP), the lectin pathway (LP), and the alternative pathway (AP). The critical steps in complement activation are the formation of unstable protease complexes, named complement factor 3-convertases (C3-convertases; specifically, C3bBb for AP and C4b2a for CP and LP), and the cleavage of C3 by the convertases to generate C3b. Convertase-generated C3b can form more AP C3-convertase, providing exponential amplification to the initial activation. Binding of C3b to the C3-convertases generates the C5-convertases with the capacity to bind and cleave C5, initiating formation of the lytic membrane attack complex (MAC). In contrast to the CP and the LP, whose activation is triggered by immune complexes and bacterial mannose groups, respectively, the AP is intrinsically activated. Spontaneous activation of C3 in plasma occurs through the tick-over mecha-

Multipoint covalent immobilization of enzymes (through very short spacer arms) on support surfaces promotes a very interesting 'rigidification' of protein molecules. In this case, the relative positions of each residue of the enzyme involved in the immobilization process have to be preserved unchanged during any conformational change induced on the immobilized enzyme by any distorting agent (heat, organic solvents etc.). In this way, multipoint covalent immobilization should induce a very strong stabilization of immobilized enzymes. Epoxy-activated supports are able to chemically react with all nucleophile groups placed on the protein surface: lysine, histidine, cysteine, tyrosine etc. Besides, epoxy groups are very stable. This allows the performance of very long enzyme-support reactions, enabling us to get very intense multipoint covalent attachment. In this way, these epoxy supports seem to be very suitable to stabilize industrial enzymes by multipoint covalent attachment. However, epoxy groups exhibit a low intermolecular reactivity towards nucleophiles and hence the enzymes are not able to directly react with the epoxy supports. Thus a rapid physical adsorption of enzymes on the supports becomes a first step, followed by an additional rapid 'intramolecular' reaction between the already adsorbed enzyme and the activated support. In this situation, a suitable first orientation of the enzyme on the support (e.g. through regions that are very rich in nucleophiles) is obviously necessary to get a very intense additional multipoint covalent immobilization. The preparation of different 'generations' of epoxy supports and the design of different protocols to fully control the first interaction between enzymes and epoxy supports will be reviewed in this paper. Finally, the possibilities of a directed immobilization of mutated enzymes (change of an amino acid by cysteine on specific points of the protein surface) on tailor-made disulfide-epoxy supports will be discussed as an almost-ideal procedure to achieve very intense and very efficient rigidification of a desired region of industrial enzymes.

Mutations and polymorphisms in the gene encoding factor H (CFH) have been associated with atypical haemolytic uraemic syndrome, dense deposit disease and age-related macular degeneration. The disease-predisposing CFH variants show a differential association with pathology that has been very useful to unravel critical events in the pathogenesis of one or other disease. In contrast, the factor H (fH)-Ile(62) polymorphism confers strong protection to all three diseases. Using ELISA-based methods and surface plasmon resonance analyses, we show here that the protective fH-Ile(62) variant binds more efficiently to C3b than fH-Val(62) and competes better with factor B in proconvertase formation. Functional analyses demonstrate an increased cofactor activity for fH-Ile(62) in the factor I-mediated cleavage of fluid phase and surface-bound C3b; however, the two fH variants show no differences in decay accelerating activity. From these data, we conclude that the protective effect of the fH-Ile(62) variant is due to its better capacity to bind C3b, inhibit proconvertase formation and catalyze inactivation of fluid-phase and surface-bound C3b. This demonstration of the functional consequences of the fH-Ile(62) polymorphism provides relevant insights into the complement regulatory activities of fH that will be useful in disease prediction and future development of effective therapeutics for disorders caused by complement dysregulation.