Targeting complement amplification at the level of C3
As aptly put by the Lambris lab, C3 serves as the ”swiss army-knife” of
the complement proteins. C3b generated from C3 in response to CP, LP or
alternative pathway (AP) activation can amplify the initial complement
activation by either pathway, when it forms the C3bB complex that can be
cleaved by the serine protease factor D (FD) resulting in the
self-amplifying C3bBb convertase (Ricklin, Reis, Mastellos, Gros &
Lambris, 2016). This amplification loop will feed the cleavage of many
molecules of C3 when not appropriately controlled by complement
regulator proteins (Figures 1 and 3). The available data suggest that
such control by complement inhibitors of the regulator of complement
activation (RCA) family is disturbed in patients developing ARDS and TMA
following infection with highly pathogenic coronaviruses, in particular
in African Americans (Figure 3). All of these RCA proteins harbor
complement control protein domains. Some of them are membrane-bound such
as complement receptor 1 (CR1/CD35) membrane cofactor protein
(MCP/CD46), decay-accelerating factor (DAF/CD55), whereas other are
found in the circulation (FH, C4 binding protein (C4BP)).
Mechanistically, the RCA proteins either destabilize the C3 convertases
or serve as co-factors for Factor I (FI)-mediated degradation of C3b to
iC3b and C3dg, which no longer contribute to the formation of the
amplification loop. As implied by the name, CD55 accelerates the decay
of the convertase, whereas CD46 mediates degradation of C3b. CR1, FH and
C4BP exert both functions.
Several compounds have been developed that target the C3 convertase
either by targeting molecules that are critical for assembly (FB, FD,
MASP-3) or the destabilization of the convertase complex and degradation
of C3b (CR1, FH). These compound have recently been discussed in detail
in two excellent reviews (Mastellos, Ricklin & Lambris, 2019; Ricklin,
Mastellos, Reis & Lambris, 2018). Prima vista, FD is an attractive
target in COVID-19 infection, given that the proteolytic cleavage of FB
by this serine protease is a crucial step to ignite the amplification
loop. Also, plasma levels are relatively low, although high plasma
turnover might pose a challenge. Thus, it is not surprising that small
molecule inhibitors (Novartis, Achillion) and FD-antibodies (Genentech)
have been generated and tested in several clinical trials. Also, a
MASP-3-specific antibody (OMS906) has been developed by Omeros to block
the conversion of por-FD to FD (Dobo et al., 2016; Hayashi et al.,
2019). However, several other serine proteases can cleave C3 including
elastase from neutrophils or proteases of the coagulation, the kinin and
the fibrinolysis system. As the substrate specificity of these proteases
for C3 is much lower than for their cognate substrates, the impact of
such proteases under homeostatic conditions is probably minor. However,
under pro-thrombotic conditions such as TMA, when control systems are
exhausted and intravascular protease inhibitor concentration is low,
such proteases are likely to cleave C3 and drive AP activation (Ekdahl
et al., 2019). In light of these consideration, blocking of systemic AP
activation by specific targeting of FD or MASP-3 seems difficult in
patients suffering from ALI/ARDS with TMA or multiorgan failure.
As an alternative to FD, FB-targeting intervention has been developed by
Ionis with Roche as a partner. They use a ligand-conjugated antisense
drug to reduce the production of FB, which is now in phase II trials for
IgA nephropathy and age-related macular degeneration (AMD). At this
point it is unclear, how efficient this drug would attenuate FB
production in a severe systemic inflammatory disease state. We will only
briefly touch molecules that destabilize the C3 convertase as clinical
development of some of these molecules has either been discontinued
(TP10 or TT30, extracellular variants of CR1)(Lazar et al., 2007;
Risitano et al., 2012) or are still in pre-clinical development
(mini-FH, Amyndas)(Schmidt et al., 2013).
As an alternative approach to target C3, conversion to C3 convertase has
been selected by the Lambris lab. They selected a peptide from a phage
library, compstatin, that prevents the binding of C3 to the assembled
convertase independent of its origin (Sahu, Kay & Lambris, 1996).
Through several round of iteration, the affinity of this compound for C3
has been increased by more than 3 orders of magnitude as compared with
the original compound resulting in CP40, which served as drug candidate
for AMY-101 (Amyndas), which is now in phase II trials for periodontal
disease and C3G (Mastellos, Ricklin & Lambris, 2019). This approach of
direct C3 targeting is attractive, as it is supposed to block
virus-induced LP and CP activation as well as LP/CP-driven activation of
the amplification loop at the bottleneck of all pathways (Figure 1).
However, C3 is one of the most abundant plasms proteins with a
concentration in the range of 1.5 mg/ml. Thus, high amounts of inhibitor
would be required to efficiently reduce circulating C3. The high
turnover of C3 under strong inflammatory conditions would add to this
problem. Finally, it remains to be determined whether compstatin
derivatives would also prevent the cleavage of C3 by all circulating or
cell derived serine proteases as outlined above. Despite these
challenges, direct C3 targeting appears an attractive target in severe
infection with highly pathogenic coronaviruses.
The list of compounds that specifically target the interaction of the
small cleavage product of C3, the C3a anaphylatoxin with its C3aR is
short. The selective nonpeptide C3aR antagonist SB 290157 has been
generated almost 20 years ago(Ames et al., 2001). It has been used in
with different success in several preclinical models to target C3aR.
However, no clinical development has been pursued. During the past 10
years, the Fairlie lab has developed sophisticated approaches to design
small molecule agonist and antagonists from proteins including C3a (Reid
et al., 2013). In this context, they have recently reported on the new
compound JR14a, a very potent C3aR antagonist that is 100-fold more
potent than SB290157 (Rowley et al., 2020). This molecule awaits broad
preclinical testing in animal models of inflammation.