Serum complement

Briefly, complement is comprised of more than 60 cell membrane-bound, intracellular, and plasma proteins [2, 3]. Most of these circulate as inactive precursors in the blood and interstitial fluid and function to activate one another following an initial trigger in a cascade-like manner similar to the coagulation system [4]. The activation of serum complement occurs via the following 3 pathways: the classical pathway (CP), lectin pathway (LP), and alternative pathway (AP) [3, 5]. Complement can be activated by numerous danger signals, such as immune complexes containing IgG or IgM antibodies (CP), pathogen-derived molecules such as lipopolysaccharide (AP), acetylated surfaces or mannan (LP), and apoptotic cells (CP). A low level of spontaneous activation of the third complement component, termed C3, occurs continuously in a process called “tickover” activation of AP. All 3 serum complement pathways converge first at the level of C3 and share a common terminal pathway, initiated with the cleavage of C5 [6]. With the activation of each pathway an enzymatic cleavage event occurs [i.e., C1s or mannan-binding protein-associated serine protease 2 (MASP-2) cleave C4 and C2 to form C4bC2b, C3 convertase of CP, or LP] [7], while factor D cleaves factor B to Bb and Ba, with Bb forming a complex with either C3(H₂O) or existing C3b. The C3(H₂O)Bb complex is termed proconvertase, while the C3bBb complex is the C3 convertase of AP [6] (Figure 2). C3 convertases and multiple proteases such as cathepsin L, thrombin, or factor Xa can cleave C3 to biologically active fragments C3b (large fragment) and C3a (small fragment) [8–10]. Additionally, C3b generated by any pathway can, in turn, lead to the activation of AP via the amplification loop. It has been estimated in vitro that up to 80% of complement activity of LP and CP comes from AP amplification [11–13].

Display full size

Figure 2. 

Schematic of serum complement activation cascades. The three activation pathways—alternative pathway (AP), classical pathway (CP), and lectin pathway (LP) are shown. Their activation leads to C3 activation and generation of C3 activation products C3b, iC3b, and C3dg via the action of complement regulators such as complement factor H (CFH), cluster of differentiation 46 (CD46), and CR1 which act as cofactors for the protease complement factor I (CFI). These C3 fragments can be recognized by a number of receptors on immune and non-immune cells and lead to phagocytosis of opsonized particles via complement receptor 3 (CR3), complement receptor 4 (CR4) or adaptive immune activation via CR2, C3a receptor (C3aR) or C5a receptor 1 (C5aR1). Furthermore, C3a and C5a act as anaphylatoxins and recruit immune cells such as monocytes, neutrophils, or T cells to sites of complement activation. C3b association with C3 convertases of AP (C3bBb) or CP/LP (C4bC2b) leads to generation of C5 convertases that cleave C5 and activate the terminal pathway of complement leading to generation of C5b-9 or membrane attack complex (MAC). MAC can lyse erythrocytes or Gram-negative bacteria or in case of sublytic MAC activate some cells. Complement proteins which have been targeted in the clinical studies are highlighted in yellow. GPCR: G-protein coupled receptors; Clu: clusterin; PAR1/4: protease-activated receptor 1 or 4; MASP: mannan-binding lectin serine protease; C4BP: C4 binding protein; C1INH: C1 inhibitor; CR1: complement receptor 1; CR2: complement receptor 2

C5 convertases are formed on surfaces covered with large amounts of deposited C3b and serve to cleave C5 to a large cleavage fragment termed C5b and a low-molecular-weight anaphylatoxin called C5a. These powerful effector molecules are generated during complement activation—C3a and C5a exert their function through cognate G protein-coupled receptors (GPCRs), C3aR, C5aR1, and C5a receptor 2 (C5aR2, formerly known as C5L2)—and drive a diverse set of processes such as inflammation, immune cell modulation and chemotaxis, cell survival, tissue repair and regeneration, vasodilation, and smooth muscle contraction [4–6, 14]. C5b generation gives rise to MAC, a 1,500 kDa complex that forms a pore in the membrane of target cell or a microorganism. It consists of C5b, C6, C7, C8, and multiple copies of C9 [15]. The MAC-induced lysis is important in the defence against invading Gram-negative bacteria, enveloped viruses, and parasites [15]. There are many excellent reviews that cover the intricacies of extra- and intracellular complement, so we will focus on the therapeutic implications in this review [1, 4, 6, 14, 16, 17].

Key contributions to our understanding of complement as a system

Since there are many important contributions that are beyond the scope of this review, we will focus on a few key points from its discovery until today (Figure 1) [1]. Some of the most important discoveries were made possible by an advancement in cloning technology that enabled the cloning of complement receptors such as complement receptor 1 (CR1), complement receptor 2 (CR2), and CR4, as well as the characterization of their specific patterns of cellular expression [18–20]. The discovery of the membrane complement regulators CD46, CD55, and CD59 happened in the mid to late 1980s [21–23]. The “liver dogma” that ruled the complement world, and stipulating that the liver is the primary organ of complement protein production, was also challenged in the 1980s [24, 25], thus establishing the prerequisite paradigm shift in thinking for the discovery of local complement in late the 2000s [26, 27] and then of intracellular complement in early the 2010s [9].

The generation of the first complement deficient C3 and factor B mice allowed for a rapidly increased understanding of the physiological roles of complement in healthy and disease states [28]. This led to the generation of complement disease mouse models, such as membranoproliferative glomerulonephritis [29] and other rare AP-driven diseases that are a result of mutations in key complement activators and regulators [30]. Another key step in increasing our understanding of complement system dysregulation and contribution to disease came from human genetic studies, unveiling the gain or loss of function mutations in key complement genes and complement deficiencies [31], which was complemented by work done in mouse complement deficient models allowing for better understanding of molecular underpinnings of human disease [32]. Classic examples of this are the deficiencies in CP complement proteins C1, C2, and C4, which are often associated with an increased risk of developing systemic lupus erythematosus (SLE) [31]. MASP-2 deficiency is known to increase susceptibility to recurrent infections [33]. Deficiencies in complement regulators have also been described. For example, complement factor I deficiency leads to an increased risk of atypical haemolytic uremic syndrome (aHUS) and age-related macular degeneration (AMD) [34]. Co-factors for the enzyme factor I are key complement regulators such as CD46, CR1, and complement factor H (CFH) [6]. Mutations in factor H, a key complement regulator in the fluid phase, lead to complement activation and C3 consumption which can lead to aHUS [35]. Deficiencies in key complement proteins such as C3 and C5 have been associated with an increased risk of infections [36, 37]. Interestingly, this increased infection risk did not translate to the clinical use of C3 and C5 inhibitors, most likely due to the requirements for vaccination prior to treatment and most patients being adults with fully developed immune system [38, 39].

Another key advancement in our understanding of complement was the crystallization of key complement proteins such as factor B in 2000 [40], C1s and C1q in 2003 [41, 42], and C3 in 2005 [43]. Important advances in our understanding of the molecular mechanisms of complement activation were also made shortly thereafter, particularly on key complement-activating enzymes such as the convertases of the AP [44–48]. C5 and the terminal pathway deserve their own mention. While structural information on C5a has been available since at least the late 1980s [49], it took until 2008 for the C5 crystal structure to be available [50]. Since 2016, the MAC structure has also been available [51]. The availability of crystal structures helps with developing better therapeutics targeting those proteins.

Along with those preclinical advances, an increased understanding of the aetiology of new, potentially complement-driven diseases has been achieved in recent decades. For example, Dr. Stevens and colleagues [52] led a mini revolution uncovering the role of complement in normal brain development and during neurodegeneration. In 2007, Stevens et al. [52] showed that mice deficient in complement protein C1q or C3 have defects in central nervous system synapse elimination. Furthermore, the authors show that this complement-mediated synapse elimination mechanism can become hijacked in neurodegenerative disease. C1q, C3, and CR3 expressed on microglia were implicated as key players in tagging and removing synapses during Alzheimer’s disease (AD), based on data from mouse AD models [53]. Thus, it is not surprising that a recent study by Veteleanu and colleagues [54] showed that C1q levels are significantly increased in the plasma of patients with AD. Furthermore, single nucleotide polymorphisms (SNP) in other complement genes such as CR1, C1S, and CFH were found to influence their plasma levels, which suggests a deeper involvement of complement in the pathogenesis of AD. The MAC inhibitor clusterin and CR1 were the first complement genes to be associated with AD in the initial large-scale, genome-wide association studies [55]. Currently, there are no complement inhibitors being tested in AD, but this is probably going to change in the future due to the large and growing body of preclinical evidence in this disease.

Genetic studies have also been very useful in deciphering the contribution of the complement system to disease. For example, SNPs in C3, complement factor B (CFB), and CFH genes were long known to be associated with an increased risk of developing AMD [56]. The genetic data along with animal model data [57] and the discovery of C3 and MAC deposits in the retinal tissue of patients with geographic atrophy (GA) [58], have served as a platform for multiple clinical trials for drug candidates such as for pegcetacoplan (a C3/C3b inhibitor), GT005 [a recombinant adeno-associated virus 2 (rAAV2) based gene therapy delivering FI in the eye], iptacopan (a factor B inhibitor) and avacincaptad pegol (a C5 inhibitor) [59].

A great example of the successful clinical translation of mechanistic complement knowledge is that in paroxysmal nocturnal hemoglobinuria (PNH). The sensitivity of PNH patients’ erythrocytes to complement lysis has been postulated since the 1960s [60, 61]; however, it was not until the mid-1980s that researchers started to attribute this sensitivity to a lack of complement regulatory proteins on the erythrocyte surface [62, 63]. The mounting knowledge resulted in the recognition of PNH as a complement-mediated disease in late the 1980s [64]. The next decade marked further the understanding of the molecular causes of PNH as a somatic deficiency of the GPI-anchored proteins CD55 and CD59, which inhibit complement activity on the surface of red blood cells [65]. The add-back of CD59 to CD59-deficient erythrocytes resulted in protection from complement-mediated lysis [66, 67]. The accumulation of preclinical evidence eventually led to clinical trials testing the anti-C5 monoclonal antibody (mAb) eculizumab in PNH [68]. Eculizumab was shown to be both safe and efficacious in the treatment of PNH and was approved by the Food and Drug Administration (FDA) in 2007, marking the first approval of a complement inhibitor in the clinic [69]. Since then, anti-C5 therapeutics have established themselves as a dominant class of complement inhibitors in multiple diseases, with C3 inhibitors more recently emerging as a class of differentiated complement therapeutics [3].

C5 inhibitors

The development of anti-C5 antibodies started in the 1990s. The first publication from Alexion Pharmaceuticals (now part of AstraZeneca Rare Disease) was released in 1995 when Evans and colleagues [71] reported the reformat of an anti-C5 mAb into a single-chain variable fragment (scFv). This fragment was shown to bind to C5 and inhibit complement in vitro and ex vivo, similar to the parental clone. The scFv was based on the N19-8 clone, which was a murine anti-human C5 antibody first described by and Würzner and colleagues [72] in 1991 and later by Rinder and colleagues [73] in 1995. By the following year, Alexion reported the successful humanization of the antibody (now termed h5G1.1), while retaining the complement-inhibiting properties of the original clone and continuing to work on the scFv version (Figure 3A) [74]. The monoclonal version of the antibody (i.e., eculizumab) was targeted for multiple indications, such as membranous nephritis, lupus nephritis (LN), rheumatoid arthritis (RA), and dermatomyositis, while the scFV version (i.e., pexelizumab) was to be used for cardiovascular indications [75]. Indeed, Alexion started phase 1 clinical development in dermatomyositis and RA a few years later (Figure 3A) [75, 76]. It was until 2007, however, that Alexion had its first clinical approval in PNH after successful phase 2 and 3 clinical trials [68, 77]. Multiple clinical trials followed, and eculizumab was eventually FDA approved in aHUS in 2011 [78], generalized myasthenia gravis (gMG) in 2017 [79], and neuromyelitis optica spectrum disorder (NSMOD) in 2019 (Figure 3B) [80]. To date, there are 127 studies listed on ClinicalTrials.gov that have eculizumab as part of the intervention or treatment agent (ClinicalTrials.gov, accessed March 21, 2024). Apart from marketing authorized indications, eculizumab has been used off-label in several other diseases with varying success, including C3 glomerulopathy (C3G), antibody-mediated rejection (AMR) of kidney transplantation, and hematopoietic stem cell transplantation-associated thrombotic microangiopathy (HSCT-TMA) [81]. This shows the immense success of the first FDA-approved complement treatment. Eculizumab, while very effective, requires a relatively high weekly dose of 600 mg for PNH and 900 mg for aHUS/gMG in the first month [81]. The weekly or bi-weekly administration is not very characteristic for most therapeutic monoclonal antibodies, so Alexion eventually addressed this by the introduction of a long-acting version of eculizumab (i.e., ALXN1201 or ravulizumab) [82]. Apart from mutations in the Fc domain (M428L, N434S) that enable more efficient neonatal Fc receptor (FcRn) recycling, ravulizumab also has histidine insertions to reduce binding to C5 at pH 6.0 and target mediated drug disposition, both of which result in improved half-life in mice [82] and humans [83]. Ravulizumab eventually gained FDA approvals in all diseases in which eculizumab was previously shown to be efficacious i.e., PNH, aHUS, gMG, and NSMOD (Figure 3B) and [84–86].

Display full size

Figure 3. 

Timeline of key events in C5 antibody inhibitors development. A. Preclinical and early clinical milestones in eculizumab development; B. timeline of FDA approvals of eculizumab and ravulizumab

The clear clinical and commercial success of targeting C5 spurred a lot of interest in both preclinical and clinical development in the C5 inhibitor space. Currently, C5, C5a, and C5aR are the most clinically targeted complement proteins (Table 1). While detailed descriptions of each of these therapeutics are beyond the scope of this review, recent reviews are available [1, 70, 87]. Notable mentions in this category are avacopan, a C5a receptor antagonist that was approved in 2021 for the treatment of anti-neutrophile cytoplasmic autoantibody (ANCA)-associated vasculitis (AAV) [88] and avacincaptad pegol (also known as Zimura), which was approved in 2023 for the treatment of GA, an advanced form of AMD [89]. Vilobelimab (also known as IFX-1) is an anti-C5a mAb developed by InflaRx that was granted an Emergency Use Authorization by the FDA in 2023 for use in patients with severe coronavirus disease 2019 (COVID-19) [90]. This antibody is also in phase 3 clinical testing for pyoderma gangrenosum with data expected to be released in 2025.

C3 inhibitors

The recent FDA approvals of pegcetacoplan for PNH and GA further galvanized interest in the field of C3 complement therapeutics. Pegcetacoplan contains 2 C3-, C3b-, and C3c-inhibiting cyclic peptides that belong to the compstatin family. Those peptides are linked by a 40 kDa polyethylene gycol (PEG) for improved serum half-life [3]. The first compstatin, a 13-mer synthetic C3 inhibitor, was first published in mid-1990s by Sahu and colleagues [91]. It was discovered after screening a phage display library for C3 binders [91]. Since 1996, the compstatin family of C3 inhibitors has grown with at least 7 new members [92]. The newer members of the family are characterized by better stability in serum, increased half-life, and higher potency in inhibiting complement activation [93]. The development of the compstatin family of inhibitors split in 2007 when Potentia Pharmaceuticals licensed the rights to compstatin derivative under the name POT-4. POT-4 was tested in a phase 1 study in patients with AMD [94]. From that initial trial in AMD, it took until 2015 for the next clinical study in PNH to be initiated (NCT02264639) [95]. Like eculizumab, pegcetacoplan was shown to be safe and efficacious in patients with PNH, which led to its first FDA approval in 2021 (Figure 4) [96]. In 2023, pegcetacoplan intravitreal injection was also approved as the first ever treatment for patients with GA secondary to AMD based on positive data from 2 phase 3 studies [97, 98]. It was joined a few months later by avacincaptad pegol, a pegylated RNA aptamer targeting complement at the level of C5 as the second approved treatment for GA [89]. Its proposed mechanism of action and comparison with C5 inhibition has been covered extensively elsewhere [3].

Display full size

Figure 4. 

Timeline of key events in development of C3 inhibitors from compstatin family. Top: preclinical and clinical milestones in pegcetacoplan development; bottom: timeline of milestones of AMY-101 development. AMD: age-related macular degeneration; PNH: paroxysmal nocturnal haemoglobinuria; GA: geographic atrophy; AA: amino acid; ARDS: acute respiratory distress syndrome; COVID-19: coronavirus disease 2019

Amyndas Pharmaceuticals took a different approach and instead started clinical development of their lead asset AMY-101, a cp40-based, fifth generation compstatin, in periodontitis (Figure 4) [99] and NCT03694444 after completing a successful phase 1 safety study in healthy volunteers (NCT03316521). AMY-101 was also tested in patients with acute respiratory distress syndrome (ARDS) due to COVID-19 infection in 2021 (NCT04395456). The preliminary results published in 2022 showed a trend towards a reduction in the use of supplemental oxygen at day 14 and a reduction of C-reactive protein (CRP), ferritin, and sustained C3 inhibition [100]. Phases 3 and 2 studies on AMY-101 are planned in periodontitis and in C3G, PNH, and kidney transplantation, respectively (Table 2), www.amyndas.com.

The clinical and regulatory success of pegcetacoplan and, the wealth of preclinical data highlighting the importance of C3 as a complement target have sparked an increased interest in C3 by a number of companies (Table 2). The publicly disclosed C3-targeting pipeline seems to be dominated by silencing ribonucleic acid (siRNA) as a modality compared with C5 pipeline (Table 1). This could be reflective of the specific mechanisms related to the target since C5 knockdown alone might not be sufficient to reduce AP to therapeutically meaningful levels. An example of this is the inability of cemdisiran (C5 targeting with GalNAc-siRNA) alone to control lactate dehydrogenase (LDH) levels in patients with PNH [101]. An alternative explanation could be that this is due to the fact that therapeutic pipelines that target C3 are newer compared with C5 for the historic reasons outlined above.

Other complement inhibitors

C3 and C5 have emerged as the most clinically attractive complement targets over the past 2 decades; however, many other complement-targeting therapeutics are being tested in clinical studies or have recently gained regulatory approval. A recent example is the December 2023 FDA approval of the small molecule factor B inhibitor iptacopan (also known as LNP023) for the treatment of PNH in adult patients. This compound, which was developed by Novartis Pharmaceuticals, is the first small molecule therapeutic targeting the complement system to be approved. Currently, there are 13 active clinical studies with iptacopan on ClinicalTrials.gov in multiple complement mediated diseases such as PNH, C3G, aHUS, immune complex membranoproliferative glomerulonephritis (IC-MPGN), IgA nephropathy (IgAN), GA, and LN. Another small molecule, factor D inhibitor, danicopan (also known as ALXN2040) developed by AstraZeneca was recommended for marketing authorization in Europe for the treatment of patients with PNH who were previously treated with C5 inhibitors but still present with significant extravascular hemolysis. It was FDA-approved on the basis of positive phase 3 clinical trial data [102]. In addition to PNH, danicopan is currently being tested in a phase 2 study in patients with GA. Sutimlimab (also known as SAR445088 and developed by Sanofi, is a mAb inhibiting C1s and thus the CP [103]. It was approved in 2022 for the treatment of cold agglutinin disease (CAD), making it the first FDA-approved CP inhibitor [104]. Sanofi is also testing sutimlimab in other disease indications, such as AMR and chronic inflammatory demyelinating polyneuropathy (CIDP). Sanofi is also developing an anti-factor Bb mAb for the treatment of rare renal disease. For further information on the discussed complement therapies, readers are referred to a recent review by Bortolotti et al. [105].

Validation of complement inhibition in non-clinically validated indications

One obvious answer to increasing competition is the validation of existing complement therapeutics in diseases for which there are no established complement therapies. This presents a challenge, as complement involvement in the pathogenesis of most of those diseases is not always clear. Some of the diseases where there is preclinical and anecdotal clinical information are listed in Table 3 and Figure 5. We will not discuss those in detail, however, we will use kidney disease and transplantation as an example since there has been a significant accumulation of preclinical and clinical data. There is active clinical development in these diseases suggesting that the FDA approval of complement therapy is just a matter of time. Kidney diseases with suspected complement involvement include aHUS, IgAN, C3G, and IC-MPGN. Atypical HUS is classified as thrombotic microangiopathy (TMA) that is predominantly renal in nature, unlike thrombotic thrombocytopenic purpura (TTP) and is not caused by Shiga-like toxin-producing Escherichia coli HUS (STEC-HUS) [111]. Atypical HUS can arise due to mutations in complement regulatory genes such as factor H [112], CD46 [113], factor I [112], and others. Autoantibodies targeting factor H have also been reported [114]. Mechanistically, the mutations in complement regulatory proteins and anti-factor H antibodies reduce complement regulation and lead to uncontrolled systemic complement activation.

Complement has been long implicated in IgAN as part of the pathology since most patients with IgAN have C3 deposition in the kidney [115] and potential systemic complement activation exemplified by the presence of circulating C3 opsonized IgA-containing immune complexes [116]. Both the LP and the AP have been implicated [117]. Despite the anecdotal mixed efficacy data for eculizumab in case reports [118–120] today, there are multiple clinical trials testing complement inhibition in patients with IgAN and it seems that we are on the verge of FDA approval in near future [121].

Organ transplantation is another area where there is ample evidence for complement involvement in the rejection. In transplantation, complement activation can occur in both the donor and the recipient [122]. On the donor side, shock and inflammation after death can cause complement activation in multiple organs including the kidney [8, 13]. For example, during reperfusion in human transplanted kidneys, C5b-9 serum levels were shown to be noticeably higher in deceased-donor kidney transplantation compared with living ones [123]. Ischemia has been shown to result in complement activation mostly via the AP, although the LP and the CP were also implicated [13, 124]. The data on ischemia comes predominantly from mouse models of ischemia reperfusion injury (IRI) whereby C3, C3aR, or CFB knock-out mice were shown to be protected from injury [125, 126]. Consistent with those data, factor B-blocking antibody was shown to prevent C3b deposition in mouse kidneys [127]. Probably not surprisingly, complement inhibition pre-transplantation results in the reduction of delayed graft function. A recent study by Danobeitia et al. [128] showed that the complement recombinant C1 inhibitor (blocks CP at C1 complex level) pre-treatment of donor kidney resulted in a reduction of complement CP activation, a reduction in proinflammatory cytokines [i.e., TNF-α and monocyte chemoattractant protein-1 (MCP-1)], and improvement in delayed graft function in a primate transplantation model. The authors found that complement inhibition was associated with superior renal function and a reduction in urinary neutrophil gelatinase-lipocalin (NGAL) [128]. In a pig model of autotransplantation, pre-reperfusion treatment with recombinant C1 inhibitor therapy resulted in a faster recovery of glomerular function and was associated with improved long-term kidney function and reduced fibrosis [129]. On the recipient side, it is known that recipients can develop donor-specific antibodies such as anti-ABO or anti-HLA. These antibodies bind to epithelial and other cells in the transplanted organ and trigger complement activation via the CP in some patients, again pointing to the role of complement in rejection [130]. While the preclinical data are positive so far, the clinical use of eculizumab and C1 inhibitors has been more or less disappointing in that it failed to prevent chronic AMR [131].

While there is mounting evidence for the role of the complement system in cancer and cancer immunotherapy this is beyond the scope of this review, and readers are directed to our recent review on the topic [2].

Combination therapy

Another way to improve the efficacy of complement targeted therapy in diseases with more complex aetiologies is to use combination therapy with a complement inhibitor plus standard-of-care medicine. This has long been an established strategy in oncology. For example, combinational therapy with both anti-PD-L1 and anti-vascular endothelial growth factor (VEGF) antibodies, has shown improved efficacy in patients with hepatocellular carcinoma [178]. In addition, the combination of danicopan plus eculizumab or ravulizumab is a complement-specific example of improved control of extravascular hemolysis in patients with PNH who do not respond well to anti-C5 therapy alone [102]. Based on the reduction of C5 in serum after treatment with Cemdisiran, (C5 targeting GalNAc siRNA) one can infer that this will allow for lower doses of eculizumab or ravulizumab to be used for the treatment of PNH or other indications where there is clinical validation of C5 inhibition. A clinical stage example marrying both complement inhibition and oncology is the ongoing phase 2 clinical trial of pegcetacoplan in combination with either pembrolizumab [anti-programmed cell death protein 1 (PD-1) antibody] or pembrolizumab and bevacizumab (anti-VEGF antibody) (NCT04919629). It is expected that complement inhibition at the level of C3 in these patients will reduce C5a generation in the tumor microenvironment and thus aid the efficacy of pembrolizumab. This notion is based on data published by Markiewski et al. [179] showing that C5a can enhance tumor growth via the recruitment of myeloid-derived suppressor cells (MDSC) into tumors and block the function of antitumor CD8⁺ T cells thus complement inhibition is expected to reverse these processes. In our opinion, combinational approaches that include complement inhibitors plus a non-complement medication might be the best way to approach oncology [180] and autoimmune indications where complement activation is probably just one of the facets of a complex disease network.

Emerging complement biology

Another potential opportunity in expanding the scope of complement therapies is targeting the more recently discovered roles of local and intracellular complement [3, 87]. Currently, the external facing drug pipelines of pharmaceutical and biotechnology companies are, for the most part, focused on targeting systemic complement activation. This is starting to change with the advent of ocular-delivered therapeutics such as pegcetacoplan and avacincaptad pegol as well as the emergence in the clinic of C3 tissue targeting bi-specific molecules such as Kanaph Therapeutics’ KNP-301, a fusion protein of C3b inhibitor and anti-VEGF antibody [181] or Q32’s ADX-097 (C3d targeted FH domains 1–5). These approaches, which have been first established by multiple academic labs [182–184] have the potential to reduce not only the amount of drug required while also reducing systemic complement inhibition and thus improving safety. This also results in improved efficacy via targeting local complement interaction with immune and non-immune cells. Locally produced and activated complement proteins can interact with complement receptors via both paracrine and autocrine signaling pathways which modulates the functions of several immune cell types [185]. For example, Strainic et al. [27] have shown that the endogenous production of C3a and C5a by antigen presenting cells and CD4⁺ T cells boost T cell proliferation and survival through signaling via C3a and C5a receptors. Another study by the same group revealed a role for the C5a-C5aR1 signaling axis in T cell survival and expansion [26]. Intracellular complement is another recent discovery that is yet to be explored in the clinic. Briefly, Liszewski et al. [9] first identified the presence of intracellular C3 in the endosomal and lysosomal compartments of CD4⁺ T cells. In resting T cells, C3 is constantly intracellularly activated by the cysteine protease cathepsin L, akin to the tickover mechanism described for serum C3. The generated C3a then engages C3aR on lysosomes to sustain the mammalian target of rapamycin (mTOR) activation required for cell survival [9]. During T-cell activation, T-cell receptor signaling drives activated C3a and C3b fragments to translocate to the cell surface and signal via surface located C3aR and CD46, which, in turn, drives a differentiation into T helper 1 (T_H1) phenotype response [186]. Importantly, dysregulated C3 activation and CD46 signaling leads to hyperactive T_H1 responses in some autoimmune disease such as RA [9, 187]. In addition to C3, intracellular C5 signaling induces the expression of NOD-, LRR- and pyrin domain-containing protein 3 (NLRP3) inflammasome and IL-1β, further driving T_H1 differentiation of CD4⁺ T cells [188]. The autocrine activation of CD46 is required to induce metabolic rearrangements and mount a T_H1 response and allow the production of interferon γ (IFN-γ) [186]. CD46 was also shown to be an important costimulatory molecule for CD8⁺ T-cell function [189]. The field of intracellular complement has been rapidly expanding with more evidence accumulating in recent years [17, 190–196]. While it has not even been proven that it has relevance in vivo yet, the potential for controlling both innate and adaptive immunity with a single complement therapeutic seems like an enticing proposition.

Finally, emerging therapeutic modalities such as CRISPR-Cas9 editing, base editing, prime editing, mRNA editing, and rAAV-based gene therapy might provide differentiation from existing marketed products which are mostly protein or small molecules in nature. These technologies, along with specific challenges, and opportunities associated with using gene therapy approaches for targeting the complement system, will be discussed at length in the next section.

Gene augmentation therapy

Gene augmentation involves the delivery of genetic constructs encoding a healthy gene to replace a dysfunctional one [206]. In the context of complement dysregulation, complement regulators that can be overexpressed to restore complement homeostasis and reduce inflammation associated with complement activation are most often targeted. Gene augmentation typically uses a virus as a carrier to enable the delivery of the gene of interest to the appropriate cells. Different types of viruses, such as adenovirus, adeno-associated virus, herpes simplex virus, and lentivirus, can be used to deliver gene therapies [174, 207]. Among these, rAAV vectors are currently the most common, and they constitute approximately 70% of ongoing clinical trials for one-time gene therapies [197]. Recombinant AAV vectors are preferred because they can transduce dividing and non-dividing cells, ensuring sustained transgene expression through circular concatemers that persist in the nucleus [208]. Recent breakthroughs have enabled the engineering of proprietary rAAVs with improved transduction efficiency and tissue specificity, further strengthening their utility as gene therapy vectors [209, 210]. Despite the advantages of rAAV-based gene therapies, some challenges remain. The major challenges are innate and adaptive immune responses against rAAV, determining the optimal dosage to balance therapeutic efficacy and minimize toxicity, and limited cargo capacity [208, 211]. Thus, a lot of companies and researchers have focused on utilizing this technology for specific tissues such as the eyes to minimize systemic exposure and limit side effects [211].

Gene editing

Gene editing is a groundbreaking technology that enables precise modifications to the genome, including insertions, deletions, and base substitutions [212]. It holds the potential for controlling genetic diseases by addressing mutations in single genes that lead to changes in gene expression in vivo. Over the years, gene-editing technology has evolved through 3 primary generations of development.

Other gene therapies

The DNA editing technologies have many advantages, with a key advantage being highly specific and permanent DNA changes that can enable the use of one-and-done therapies in multiple diseases [229]. As discussed above, there are also challenges associated with off-target editing and the delivery to target cells. Recognizing the potential risks of a permanent alteration of DNA sequences, researchers have turned to epigenetic and RNA editing as alternative approaches.

Delivery challenges and opportunities associated with gene therapies

The clinical translation of CRISPR-based gene therapy presents a significant hurdle in the safe and efficient delivery of CRISPR components to target cells or organs [242]. While rAAVs are the primary vectors for in vivo gene therapy due to their high transduction efficiency and relatively low immunogenicity, their small packaging size poses a challenge in delivering large genome-editing tools such as base editors and prime editors [243]. This limitation has led to the development of dual rAAV vector approaches, where the gene of interest is divided into 2 parts and is packaged into separate rAAV vectors. However, these dual rAAV systems face several challenges, including low reconstitution efficiency, the production of unnatural proteins, and safety concerns related to immune responses and adverse effects [243, 244].

Due to the limitations of viral vectors, significant attention is being directed toward the development of non-viral delivery technologies [222]. Non-viral vectors such as LNP offer promising advantages such as a larger packaging capacity, simplified manufacturing processes, and the potential for re-dosing due to lower immunogenicity compared with rAAV [245]. LNPs have gained popularity for in vivo CRISPR gene editing due to their ability to deliver Cas9 mRNA and gRNA to the liver, enabling the transient expression of gene editing proteins and minimizing off-target effects [246]. Moreover, LNPs degrade within a couple of weeks, supporting long-term safety considerations [247]. LNPs also offer the potential for split-dosing, allowing for repeat administration due to their lower immunogenicity compared with rAAV [245]. An example of a successful clinical use of LNPs is patisiran, an FDA-approved LNP-encapsulated siRNA for the treatment of hereditary transthyretin (TTR)-mediated amyloidosis, which demonstrates the feasibility of chronic administration without safety and efficacy concerns [248]. Another example of a successful clinical use of LNPs include the Pfizer/BioNTech COVID-19 vaccine (alternative name, BNT162b2) which has been widely used across the world [249]. While less immunogenic than rAAVs, LNPs still activate the innate immune system and even complement, and this can lead to complications in some patients, such as low transduction efficiency and difficulty with extrahepatic delivery [250, 251]. Safety concerns related to transient liver injury have underscored the need for thorough evaluation and optimization. In conclusion, LNPs have emerged as a clinically mature non-viral platform for the safe and effective delivery of genetic medicines. LNPs played a pivotal role in enabling the FDA approval of groundbreaking therapies such as BNT162b2 and patisiran [252, 253]. However, intravenously administered LNPs predominantly accumulate in the liver and are internalized by hepatocytes, constraining their therapeutic utility in other organs [246, 254]. Targeted delivery is a critical area of current LNP research, aiming to improve delivery outside of the liver, mitigate toxicity and off-target effects, and enhance efficacy in challenging-to-transfect targets.

Gene therapy approach to target the complement system dysregulation

In terms of using gene therapies to target complement dysregulation, the gene augmentation therapies utilizing rAAV in AMD are the most clinically advanced and the only disclosed approach so far. The basic idea is to block complement activation by using rAAV to increase complement regulator expression for the treatment of AMD. This is perhaps unsurprising since rAAV-based gene therapy is currently the most widely clinically tested form and is particularly well suited for local rather than systemic delivery, and the recent FDA approvals of pegcetacoplan and avacincaptad pegol provided the clinical validation of complement as target. Table 4 summarizes some of the known gene therapies targeting complement dysregulation.

Apart from those examples, there is not a lot of publicly available information on targeting the complement system using advanced forms of gene therapy. Here, we want to share our thoughts on how utilizing base, prime, or RNA editing might have specific advantages and disadvantages compared with peptides, antibodies, and small-molecule-based therapies which are currently the clinically established modalities for complement inhibition.

One key benefit of gene therapy such as base or prime editing is dosing convenience. Unlike conventional therapies that often require frequent and long-term intravenous or subcutaneous administration, gene therapy typically involves a single dose [256]. This eliminates the need for frequent clinic visits and reduces the burden of treatment on patients and caregivers. Most complement therapies, such as pegcetacoplan, eculizumab, and iptacopan require either twice daily or weekly administration to achieve complement inhibition due to an abundance of the targets in serum [81, 257, 258]. A notable exception of this is ravulizumab, which has been specifically engineered to have a longer half-life [82]. Thus, a gene therapy involving complement inhibition might have an advantage over current standard of care complement inhibitors by requiring a single dose for the long-lasting control of the complement system.

A second potential benefit is with regard to improving safety. The frequent dosing of traditional therapies can increase the risk of adverse effects, including infusion reactions and complications associated with long-term drug exposure [259]. By contrast, a single dose of gene therapy minimizes the cumulative exposure to potentially harmful agents, reducing the risk of adverse events. Furthermore, one-and-done gene therapy has the potential to improve efficacy by providing sustained therapeutic benefits over an extended period. This has been shown by Musunuru et al. [260], who successfully used base editing to reduce PCSK9 and LDL in the blood of non-human primates for at least 8 months after a single treatment. This was an improvement over ALN-PCSsc, an RNA interference agent that reduced PCSK9 for up to 6 months after 2 administrations [261]. This continuous efficacy after a single dose, can slow disease progression, as it eliminates the disruptions in treatment adherence that ca be associated with conventional therapies [262]. The lessons learned from PCSK9 and ATTR inhibition can be applied in the future to complement inhibition.

Base and prime editing both present a highly promising avenue for tackling complement-mediated diseases, offering diverse strategies such as correcting point mutations that lead to protein dysfunction, silencing genes through splice disruption, creating stop codons, or directly disrupting proteins [263]. Diving deeper into the first point, multiple complement activators and regulators are known to have pathogenic variants associated with the disease of complement dysregulation, particularly of the AP [264]. Since AP lacks the specificity of activation and amplification, polymorphisms in complement regulators (CFH, CD46, CFI) or activators [C3, CFB, complement factor D (CFD)] can lead to an overactive AP [264]. In diseases such as AMD, where approximately 50% of the total genetic risk comes from a point mutation in CFH [265, 266], one can imagine that, by using gene therapy to correct the Y402H (rs1061170) variant, for example, long-lasting complement homeostasis can be achieved. Other examples include C3 polymorphisms such as C3_102G, which is suspected to lead to an overactivation of the AP [266] and is associated with diseases such as AMD, IgAN, transplant dysfunction, and systemic vasculitis [267–269]. Editing C3_102G (also known as C3 fast or C3F) to C3_102R (also known as C3 slow or C3S) can improve regulation by factor H and reduce the overall activity of AP [266] thus reducing the risk of developing disease. Another advantage of this potential approach is that C3S is a naturally occurring variant and thus presents a low risk.

An alternative approach would be the inhibition of key components of the complement cascade that are responsible for its initiation and amplification, similar to the use of siRNAs to target C3 or C5 today but with the benefit of a single administration. This approach could be particularly beneficial in chronic systemic conditions where the complement system is constantly overactivated, such as in PNH or C3G. The main advantage here is that a single dose of gene therapy can achieve long-lasting control of the complement system. While infection risk with permanent complement inhibition will be one of the key issues to monitor for, to date, therapies such as C3 inhibitors have been well tolerated [38]. This emphasizes how important it is to choose the right target to avoid those safety concerns. The first probable application of gene therapy targeting complement inhibition could be as a maintenance therapy whereby 50% to 70% inhibition is sufficient to control disease progression in most patients, while minimizing bacterial infection concerns. Limitations such as the permanent nature of the DNA editing, off-target editing, and unknown long-term safety means that it will only be applicable to certain targets and mutations in a limited set of life-long and life-threatening disease until the wider clinical adoption of the technology.

The permanent nature of the base and prime editing might not be suited for all diseases and for the needs of all patients. This is where epigenetic and RNA editing might take center stage due to their unique advantages. Epigenetic editing technologies offer precise control over gene expression by altering epigenetic marks such as DNA methylation, histone modifications, and chromatin structure [235]. One strategy involves targeting specific epigenetic regulators or chromatin-modifying enzymes to increase the expression levels of complement regulators, similar to the rAAV approach discussed previously, with the view of restoring homeostasis. Since epigenetic editing is, in theory, reversible and could allow for the fine-tuning of protein expression rather than acting as an on/off switch it is expected to be safer than DNA editing [234, 235, 270]. One can imagine that reversible complement activator knock-down will be possible in the future using this technology. Chroma Medicine has already demonstrated the efficiency of epigenetic editing of its platform in reducing and restoring PCSK9 levels [271].

Silencing RNA targets specific mRNA molecules for degradation, reducing the expression levels of targeted proteins, and has been used to target C3, C5 and CFB protein expression in the clinic (Tables 1 and 2). RNA editing offers unique advantages, as it allows for precise modifications to be made to RNA molecules [241], thereby directly altering the protein-coding sequences without affecting the underlying genomic DNA [240]. This approach enables transient and reversible changes to gene expression, offering flexibility in modulating the cellular pathways involved in serum and intracellular complement activation and regulation for example. Furthermore, RNA editing can be used to modulate alternative splicing events or RNA stability, thereby influencing the expression levels of specific complement isoforms or regulatory factors similar to the strategy outlined above for DNA and epigenetic editing. The advantage of RNA editing over the other discussed gene therapy strategies is that it is easier to deploy since it does not require additional enzymes [223, 237].