Cover Page
The handle http://hdl.handle.net/1887/48207 holds various files of this Leiden University dissertation
Author: Kotimaa, Juha
Title: Analysis of systemic complement in experimental renal injury and disease Issue Date: 2017-04-25
8
Chapter 8
—
General discussion
1. COMPLEMENT, A WORK IN PROGRESS
The complement system is anything but a closed chapter in the annals of immunology. The first decades of 21st century have been exceptionally productive in complement research, in part culminating in the arrival of the first approved complement therapeutic, an anti-C5 antibody Eculizumab, for treatment of aHUS and PNH at the clinic [1]. Progress in immunotechnology, proteomics, genomics, and complement specific in vitro and in vivo methodologies have opened new fields of research and expanded established paradigms throughout the past few decades [2]. Major advances have been made not only in the clinic but also in expansion of understanding the complex roles of complement in injury, disease, and homeostasis [3–7].
Figure 1 depicts the multifaceted aspects of complement activation and inhibition which have been characterised and discovered during the past decades. These include B-cell and polymorphonuclear cell activation which both are essential for innate and adaptive immune system function.
Discoveries in extrahepatic complement expression has established that injuries and infection can result in local modulation of complement regulator and factor expression which in some cases are the main drivers of the underlying injury. Last, the complement activation results in soluble and membrane bound complement activation fragments. These form a complex network of intra- and intercellular signalling that orchestrates inflammatory and repair responses [8, 9].
After a century of continuous research, fundamental aspects of the complement system are still discovered and redefined. For example, the role of properdin as a novel pattern recognition molecule (PRM) of complement system was proposed already in the 1950s by Pillemer et al. However, role of properdin has been contested up until recent years [10–12]. Recent in vitro studies have shown that the old paradigm may be true, and that properdin can indeed act as an initiation factor on certain ligands in vitro [7, 13].
8
We investigated the role of properdin in chapter 5, establishing that properdin binds injured glomeruli independent of C3 involvement and that it may act as an initiation factor also in vivo.
One of the most prominent impacts of the 21st century complement research has been the discovery of the lectin pathway (LP) and its four PRM initiators and associated serine proteases. The discovery of LP has opened new fields of research on its role in health and disease [14–18], which we have also addressed in chapter 2 in the context of renal I/RI. The complement effectors, C3a, C5a, and C5b-9 have been characterised decades ago, however the role of anaphylatoxins and their respective receptors in injury, such as described in chapter 4, has become increasingly important aspect of immunological research [19, 20].
2. MEASURING COMPLEMENT
As outlined in chapter 1, the complement system can be initiated through three main pathways, classical (C1 complex), lectin (MBL and ficolin 1-3), and alternative pathway (C3 tick-over, properdin, amplification loop initiated by CP and LP). All pathways converge at the level of C3 and proceed through C5 to activate the terminal pathway. Due to the relative complexity and crosstalk of the pathways during complement activation, understanding the mechanisms of activation warrants a combination of different methods for accurate analysis.
Immunohistochemistry offers invaluable insights into temporal and spatial activation of complement on the affected tissue. We used various complement stains in Chapter 5 with an experimental mouse model of anti-GBM disease to establish that properdin binding can occur independent of C3 and that terminal pathway activation with C9 deposition in that model occurs relatively late despite prominent C3 activation. For clinical and preclinical studies,
analysis of human biopsies is an important tool for understanding the possible mechanisms of complement activation in vivo. For example graft rejection has characteristic activation of CP with C1q and C4 deposition, whereas renal I/RI exhibits acute co-deposition of C1q/C4d and MBL/C4d with later activation of terminal pathway. Clinical evidence with anti-GBM shows prominent presence of alternative pathway factors such as properdin and fB independent of CP, although it is an autoimmune disease [21–23]. Despite the fact that immunohistochemistry allows in-depth interrogation of complement, it requires invasive procedures in the clinic and extensive animal experiments if monitoring of complement is required.
Serum, or systemic complement, is the main source of complement factors. Serum is easily acquired in the clinic and from animals without the need to terminate the animals to acquire samples. Haemolytic assays have traditionally been the industry golden standard for interrogating serum complement. They can be used to determine pathway specific total complement activity (CP and AP) or the concentrations of functionally active complement factors [24, 25]. Haemolytic assays can detect aberrant classical (CH50) and alternative (AP50) pathway activity in serum, suggesting that the sample is either deficient for a specific component in this pathway or that the complement system has been prominently activated leading to consumption [24, 25].
Although haemolytic assays have been adapted for measurements in rodent sera [26–28], it has been reported that haemolytic assays do not accurately measure complement in certain strains such as BuB/BnJ [29]. A second concern of haemolytic assays, especially in the context of preclinical mouse studies, is the relatively high amount of serum required for simultaneous CH50 and AP50 measurement, and the requirement of fresh erythrocytes which poses a batch variability problem [27, 28].
8
Functional ELISAs with immobilised ligands have been shown to activate complement with similar specificity as haemolytic assays [30–32]. As an example, we compared the standard CH50 and AP50 against ELISA-based systems (Fig 2.) described in Chapter 2 and Chapter 3, for the analysis of human, rat, and mouse sera. The comparison shows that classical and alternative pathway activities can be efficiently measured in human serum, however poor results were obtained with mouse AP50 and with rat CH50 (Fig 2.A-C).
Modifications on standard haemolytic assays have been developed to improve species compatibility and assay sensitivity, however especially AP50 remains relatively challenging to perform with mouse sera [26, 33]. In contrast, functional activities measured in ELISA with identical activation surfaces and species specific detection antibodies result in comparable C9 activity in all species, with an additional advantage that also LP can be detected easily (Fig 2. D-F).
An additional advantage of the ELISA-based functional complement assays is the relative ease of measuring different factors along the activation pathways.
We used this possibility in Chapter 3 with detection of complement activation at the level of 3 and C9 activation. Similar approach has been used to quantify C4 and C2 and the functional activities of initiation factors, such as C1q, MBL, and properdin [32, 34, 35]. Therefore ELISA-based methodologies can supplement, or when applicable, replace haemolytic assays in many preclinical studies. In preclinical mouse studies the practical volume of serum that can be expected from an individual mouse is just 100–200 µl upon termination [36]. The available volume is far less if an individual mouse is tapped for blood more than once during the experiment. Therefore the higher sensitivity of functional ELISAs described here allows more determinations from a single sample obtained from single animal. This approach allows a more complete understanding of complement dynamics within an individual animal as was described in Chapter 5, where we used a pre-post analysis around the induction of anti-GBM, and in Chapter 6, where we analysed key complement factors and functional activities at the level of C3 and C9 activation from single mice.
8
Figure 1. Impact of complement activation at the local and cellular level. Panel A) Healthy host cells have various regulators of complement activation (RCAs) which inhibit the constant probing of the complement system. Panel B) Unprotected cells or pathogens without complement inhibitors can be bound at will by complement pattern recognition molecules (PRMs). Without RCAs the complement activation proceeds unimpeded with generation of opsonins, lytic C5b-9, and anaphylatoxins that recruit inflammatory cells. Panel C) Macrophages are important phagocytic cells and they are responsible for the detection and clearance of opsonised cells and debris. Surface deposited C3b is inhibited through sequential cleavage by fI, resulting in various C3 activation fragments that are essential opsonins used in clearance and inflammatory cell activation. Macrophages can be activated by C3a and C5a, which chemotactically attract macrophages to the site of activation. Phagocytosis is facilitated via complement receptors CR1, CR3, and CR4 and macrophage complement receptor CRIg. Macrophages can also detect opsonised cells via C1qR and FCγR which bind C1q and antibodies respectively. Panel D) Complement activation is essential for the development of adaptive immune responses. B-cell activation is achieved when B-cell receptor forms a co- receptor complex with CR2 which is specific for C3dg, as well as with CD19 and CD81. Panel E) Neutrophils and other PMNs are chemotactically attracted by C3a and C5a, recruiting them to the site of complement activation. Local inflammatory conditions induce PMN degranulation that aggravates local inflammation, aids in clearance and destruction of pathogens, and can also secrete complement factors which amplify local complement activation. Panel F) Injured cells enter apoptosis, which results in downregulation of RCAs and changes in the composition of the phospholipid membrane. Together these allow complement activation on host cells via PRMs, specific ligands, and complement tickover. Activation of the terminal pathway results in C5a generation and in formation of sublytic amounts of C5b-9 which attract inflammatory cells and promote apoptosis, respectively. Panel G) Local complement activation results in a complex network of cascading and interacting effects. Initial complement activation promotes local inflammation resulting in vascular leakage allowing systemic complement to penetrate the tissue and inflammatory cells to migrate towards the anaphylatoxin gradient. Anaphylatoxins have three known receptors, C3aR, C5aR, and C5L2, which have been shown to promote cellular inflammation. The inflammatory signalling can result in cellular injury and in upregulation of local expression and secretion of complement factors that amplifies the local complement activation. Chemotaxis of inflammatory cells to the site of injury results in PMNs activation and secretion of cytotoxic compounds and complement factors which promotes further killing of pathogens and potentially injuring host cells. Phagocytic cells also activate in the local inflammatory milieu, facilitating the clearance of pathogens and injured cells via opsonin specific complement receptors [adapted in part from 8, 9].
Figure 2. Comparison of haemolytic and functional complement ELISAs. A) Efficiency of haemolytic assays optimised for human use were used to compare cross-species compatibility. Results of mouse and rat serum CH50 and AP50 were calculated relative to human NHS activity set to 100%. Dose dependent activity of B) CH50 and C) AP50 activities were determined with reciprocal dilutions of sera from the three species and performed in duplicate. Dose dependent activity of the three species determined on functional complement ELISAs at the level of C5b-9 activation with (D) CP determined on IgM plates (E) LP as determined on mannan coated plates and (F) AP on LPS coated plates.
Both haemolytic and ELISA based assays require understanding of the specificity and potential cross-reactivity issues. We addressed those issues in Chapter 2 and Chapter 3, with complement deficient sera and with pathway specific chemical inhibition. In contrast to analogous human assays, we did not observe interference from antibodies against mannan, which would have required further modifications such as the use of sodium polyanethole sulfonate (SPS) or anti-C1q antibody [35, 37]. However, this should be provisionally tested when mice or rats are expected to have been exposed to mannan.
8
Analysis of circulating complement factors and complement activation biomarkers are an important addition to the complement toolkit. They have been used extensively in clinical and preclinical research to distinguish complement deficiencies from total complement consumption and to detect acute complement activation [24]. They are invaluable in determining whether low or absent complement activity is a result of deficiency or an acute consumption of complement [25]. Biomarker ELISAs have been extensively used to measure complement activation systemically in blood and in other compartments such as urine, synovial fluid, and cerebrospinal fluid [38–40].
The main concern with measuring complement activation fragments is their relatively low concentrations in serum and the autoactivation of complement during sampling and handling. Serum or lepirudin plasma is generally considered the preferred sample type for functional complement analysis [24, 35, 41]. Our results in Chapter 3 showed that C3 autoactivates during serum and lepirudin plasma preparation even if samples are kept on ice at all times. We speculate that this is due to coagulation induced C3 cleavage reported previously, which can be inhibited by EDTA, whereas lepirudin inhibits only the thrombin mediated clot formation [42]. Interestingly the coagulation-dependent autoactivation of C3 did not result in activation of terminal pathway as a similar effect was not observed for rat SC5b-9 in Chapter 2. Our results suggest that supplementation of blood with 10 mM EDTA directly after sampling is sufficient to inhibit C3b/C3c/iC3b generation during sample preparation, avoiding the issue of measurement artifacts.
To avoid the necessity of preparing both serum and plasma, we tested the impact of EDTA supplementation to functional activities. Our results in both Chapter 2 and Chapter 3 suggest that there was only a marginal difference in complement pathway activities in serum and EDTA-plasma when activated in Mg2+ and Ca2+ supplemented veronal buffer. This appears in contrast with existing literature on human assays (Wielisa) which have shown lower
complement activity in EDTA-plasma [35]. The buffer composition of Wielisa has not been disclosed, therefore it is not possible to compare whether veronal buffer or other details could explain the better compatibility shown in Chapter 2 and Chapter 3. We did not observe any acute issues with handling of either rat or mouse sera or plasmas. Our results were in agreement with human and mouse studies where it was shown that complement remains stable if stored at -70°C for longer periods, is handled at 4°C for shorter times, and extensive freeze-thawing is avoided [27, 28, 43]. Together our results show that in rodent studies a single sample can be used for extensive analysis of complement to better understand complement activation mechanisms and to supplement results of immunohistological determination of complement activation.
3. INHIBITING COMPLEMENT
Functional and standard complement analysis, whether haemolytic or ELISA-based, is essential to establish the specificity, duration, and efficacy of complement inhibition.
Human functional complement ELISAs, analogous to the rodent assays described in Chapter 2 and Chapter 3, are routinely used to monitor and optimise complement therapy in the preclinical and clinical studies [23, 44–
46]. For example, the efficacy of Eculizumab (anti-C5) therapy of aHUS and PNH is routinely monitored with tandem determination of CP or CH50 and sC5b-9 [25].
High doses of intravenous immunoglobulins (IVIG) have been shown to interact with systemic complement, either inhibiting or, in case of aggregated immunoglobulins, activating complement in humans and experimental models [47–49]. In our studies, validation of complement inhibition was central in Chapter 2, where we described the use of therapeutic anti-MBL in the context of rat renal I/RI. Our results showed that the inhibition specifically blocked LP activity for of 24h, while leaving both CP and AP intact, which is in
8
agreement with our previous results [50]. In Chapter 5 we used the functional complement assays and C3 activation products to validate antibody mediated experimental anti-GBM. In agreement with literature, our results revealed an acute activation of complement with an increase in circulating C3-activation fragments and significant loss of both CP and AP [51–53]. A possible IVIG- like activation of systemic complement was excluded with the control group that received identical dose of non-specific antibody preparation.
4. COMPLEMENT IN RENAL INJURY
The complement system plays a multifaceted role in renal autoimmune diseases and in acute and chronic kidney injuries as described in Chapter 1.
Experimental animal models are used extensively to investigate which pathways of complement are involved and whether different therapeutic intervention strategies would attenuate or ameliorate injury. Our work on establishing methodology to measure systemic complement in experimental rat and mouse models in Chapter 2 and Chapter 3 allowed us to analyse the temporal changes in systemic complement following renal I/RI in rat and mouse. In Chapter 2, the experimental rat model of kidney I/RI model resulted in relatively minor impact on systemic complement. Indeed, only AP showed consumption at 24h, which was not reflected in generation of terminal pathway activation biomarker SC5b-9, suggesting that the complement activation following the reperfusion did not result in acute activation of terminal pathway. The relatively minor impact of reperfusion on complement is in line with clinical evidence where it is shown that reperfusion results in acute but transient release of SC5b-9 following reperfusion after transplantation [54]. During the development of renal I/RI we observed a steady increase in CP activity towards 72h, possibly as a reflection of the acute phase response to renal failure. Interestingly, the levels of SC5b-9 did show a steady increase from 48h to 72h, which in the light of past studies would suggest that it reflects the renal activation of terminal
pathway reported earlier [50]. In a follow-up study, we investigated long term impact of I/RI on inflammatory parameters. Interestingly, circulating SC5b-9 increased up until day 7, whereas previous results had shown that intrarenal C5b-9 deposition was mainly cleared by 96h [50]. We found that a similar increase in CD11b (Ox42) positive macrophages infiltrates, suggesting that the serum SC5b-9 has similar dynamics as the cellular inflammatory response rather than the acute injury (Fig 3. unpublished observation). Together these results would suggest that the serum SC5b-9 might not always reflect the local complement activation, but could also be the consequence of an inflammatory reaction. Further determinations of inflammatory markers such as serum amyloid protein (SAP) and complement activation fragments, such as C3d, and pathway specific components, such as Ba, MBL and, C1q, could reveal whether this is an indication of pathway specific complement activation.
Based on our findings it may be important to determine if the injury or disease results in an acute phase response as many complement factors are acute phase proteins and the reaction may directly impact various complement determinations through increased synthesis and turnover [55].
Previous studies have convincingly shown that an MBL-dependent, complement- independent mechanism dominates this model of rat renal I/RI. Indeed, with the systemic neutralisation of MBL in Chapter 2, the renal injury was attenuated and the neutralisation verified as a transient loss of LP functional activity in line with our results [50].
8
Figure 3. Time course observation of rat serum SC5b-9 and renal infiltration of macrophages following experimental renal I/RI. A) SC5b-9 measurements from rats with experimental renal I/RI or sham operation B) histological analysis of infiltrating macrophages in the corticomedullary junction was performed with CD11b (OX42) C) linear correlation of matched measurements of SC5b-9 and OX42 histolog y.
In contrast to the rat model, the two mouse renal I/RI models described in Chapters 3–4 revealed minor and acute consumption of complement. Our results show that early after reperfusion there is minor activation of C3 that results in increased circulatory C3 fragments (C3b, C3c, iC3b). However, this is not reflected as significant loss of pathway specific functional activities at the level of C3. Interestingly, early after I/RI there was a reduction in the complement activity of all three pathways, when measured at the level of C9 deposition. This suggests that there might be prominent consumption of terminal pathway components following reperfusion. Our observation in the mouse renal I/RI is in line with experimental mouse models, which have convincingly shown that terminal pathway effector molecules like C5a [56, 57] and C5b-9 [58] all have a dominant role in mouse renal I/RI and that the experimental injury can result in early systemic loss of haemolytic CH50 activity [59]. Furthermore, intrarenal expression of C3 alone is sufficient to initiate and maintain complement activation following renal I/RI, which may in part explain our findings [60].
In Chapter 4, we studied the role of two known receptors for C5a, C5aR, and C5L2 to renal I/RI using genetically engineered KO mice. Previous studies by van Werkhoven et al. have shown that within the kidney C5aR and C5L2 do not co-localise, and therefore it is unlikely that C5L2 would mitigate the proinflammatory impact of C5aR as a scavenger receptor [61]. In line with previous experimental renal I/RI studies in mice, we observed significant attenuation of renal injury in C5aR KO mice. Interestingly, similar reduction of injury was observed also with C5L2 KO mice. Upon further investigation we found that the C5L2 receptor has a distinct role in pro-inflammatory signalling of both PMNs and kidneys. In contrast to the well characterized C5aR, C5L2 does not impact chemotaxis of PMNs.
Instead, the loss of C5L2 reduces the inflammatory response of PMNs. Similar protective modulation of pro-inflammatory signalling was observed within C5L2 KO kidneys with evidence of diminished IL-6 signalling after renal I/
RI.
8
Together our results on mouse renal I/RI would suggest that serum terminal pathway components are prominently consumed following reperfusion and that intrarenal formation of C5a has a deleterious impact on the reperfused kidneys. Although specific injury mechanism could not yet be defined, our results clearly demonstrate that C5L2 promotes injury independent of C5aR.
Together, the targeting both systemic C5, and the C5a receptors C5aR and C5L2 could be a promising avenue for treating renal I/RI.
Targeting C5 is already possible in humans, through approval of anti-human C5 mAb, Eculizumab, for treatment of aHUS and PNH[62, 63]. Recent and ongoing clinical trials with Eculizumab have investigated whether systemic inhibition could improve acute rejection [64, 65] and potential amelioration of renal I/RI (ClinicalTrials.gov Identifiers: NCT01756508, NCT01403389).
Our results in Chapter 3 showed that experimental renal I/RI in mice resulted in consumption of serum terminal pathway (C5 through C9). These results are in line with previous studies, which show that systemic inhibition of complement at the level of C5 with Eculizumab analogue BB5.1 results in protection [59]. The results in Chapter 4 showed clearly that C5a can promote renal inflammation independent of PMN infiltrations through C5aR and C5L2. These results are in line with previous studies showing that although C5aR agonists do reduce PMN infiltration mediated injury, there is also renal specific impact and protection [58, 59]. However, this cannot be achieved with systemic neutralisation of C5a alone [58]. Together the results from mouse studies strongly suggest that complement-mediated injury in a mouse model of renal I/RI can be ameliorated through systemic inhibition of complement activation or complement effectors at the level of C5 activation. These results are in contrast to rats, where MBL alone seems to convey the majority of injury independent of C5a and C5b-9 [50]. During mouse renal I/RI, AP seems to be prominently involved in complement activation and C5a signalling, and C5b-9 contributing to the injury [56].
Interestingly, recent studies have shown that MBL-MASP1/2 dependent, terminal pathway independent injury may also be present also in mouse as C4 mice are not protected, but MBL-Masp1/2 mice show partial protection [66, 67].
5. ALTERNATIVE PATHWAY AND ANTI-GBM RENAL DISEASE
Anti-GBM, or Goodpasture disease, is an autoimmune disease which targets glomerular basement membrane and results in renal failure [68]. A recent study with an experimental mouse model of anti-GBM established that classical and alternative pathways, and Fc-mediated neutrophil activation contribute to the renal injury [53]. However the exact mechanisms by which the alternative pathway activates on the glomeruli and contributes to the injury were not addressed, leaving the possibilities open for AP autoactivation and properdin- enhanced or -directed AP activation. To better understand the mechanism of AP activation and the role of properdin we established a single-step anti-GBM model with novel properdin KO mice in Chapter 5.
Our primary aim there was not only to determine the impact of properdin knockout to renal injury, but also to determine kinetics of the different complement pathways within the injured glomeruli. To achieve this, we analysed the pathway specific complement consumption in serum, followed by time-course staining of key complement factors.
Although our results from renal injury suggested that both properdin- and C3-knockout mice had lower degree of renal injury, the variation in each group, wild type mice included, did not reach significant levels of protection against experimental anti-GBM. It has been previously shown that Fc-mediated neutrophil activation is a major contributor in experimental anti-GBM. We found that neutrophils infiltrate the glomeruli to a similar degree in WT and properdin KO mice. Based on the serum MPO analysis the activation occurs 2–6h after anti-GBM injection.
8
These results suggest that in this anti-GBM model and at this timeframe, the neutrophils convey majority of the injury, independent of CP or AP activation. The time course activation of complement in the affected glomeruli revealed that both CP and AP were activated already at 2h post induction, which was also observed as systemic consumption of CP and AP, and generation of C3 activation fragments. These results are also in line with previous studies on different anti-GBM models showing CP and AP involvement [21, 52, 53].
The deficiency of properdin resulted in reduced systemic CP consumption and severely reduced the generation of C3 fragments. These results were expected as properdin stabilises the AP C3 convertase formed independently or as part of the AP amplification loop as a result of CP activation [69].
To better understand the complement activation dynamics in this model, we performed a series of histological stains with different complement factors. First, in line with previous studies with anti-GBM, it was clear that CP drives the acute C3 activation in the glomerular vasculature [51, 53]. However, we could not detect C9 deposition before 48h, suggesting that the terminal pathway activation was attenuated and complement mediated injury through C5a and C5b-9 was not prominent until 48–
72h. Interestingly, the staining of C9 increased in tandem with properdin increase at 48–72h, exhibiting similar pattern. Further studies are required to study the dependence of terminal pathway activation and properdin, as our results indicate that properdin mediated AP activation may be the driver of terminal pathway activation and complement mediated injury in anti-GBM.
Our results were clear that in addition to rabbit IgG and C3, also properdin was present within the glomeruli at 2h, however there was hardly any co- localisation with C3, anti-GBM rabbit IgG, or neutrophils, suggesting that properdin may be binding independently to the injured glomeruli. To verify
this, we performed an experiment where we investigated properdin deposition in C3 KO mice, showing that properdin was still present within the glomeruli and did not co-localise with the rabbit IgG. To our knowledge our results show for the first time that properdin binding can occur independent of C3 or IgG in vivo, verifying the previous in vitro studies suggesting a properdin dependent AP activation [7, 10, 70].
6. LOCAL AND SYSTEMIC
COMPLEMENT: KIDNEY VERSUS LIVER
Several studies with renal I/RI have reported that mice show a clear gender dimorphism where female mice exhibit clear protection against ischemia/
reperfusion injury. For example, in case or renal I/RI, female BALB/c mice must have up to 30 minutes longer ischemia time to reach similar degree of renal injury as male mice [71]. Previous studies into this interesting phenomenon have revealed a number of mechanisms which contribute to the protection, including hormonal regulation of vascular endothelin [72], differential expression of Na+/K+-ATPases [73], and differential expression endothelial cell nitric oxide synthase [74]. The complement system is also influenced by gender, resulting in low or absent CH50 activity in female mice, and has been traditionally attributed to gender specific expression of complement components such as sex-limiting protein (Slp, C4) and post-translational modifications of complement factors such as Hc (C5) [26, 75]. However, the impact of the gender dimorphism on functionality of the three main pathways of complement has not been studied in detail previously. In Chapter 6 we used the novel methodology developed in Chapter 3 and Chapter 5 to investigate the gender differences in mice with functional analysis at the level of C3 and C9 activation, followed by detailed analysis of individual complement factors.
8
Our results clearly show that the gender dimorphism does not have a major impact on functional complement activities at the level of C3 activation, confirmed with individual factor analysis which showed no major differences with initiation factors or C3 levels in mouse serum. In line with the functional analysis, determination of terminal pathway factors C6 and C9 were up to eight-fold lower in female mice. Together these results show that the previously observed low CH50 activity in female mice is due to impaired ability of female sera to form C5b-9. As shown in Chapter 4, activation of complement at the level of C5 and C5a generation results in aggravation of renal I/RI, and is central in many other experimental models such as sepsis [76]. Although our results in Chapter 6 do not directly show that C5 levels are significantly lower in female mice, there is convincing evidence that also C5 is androgen inducible with lower level in females [77–80]. Furthermore, in the context of the mouse renal I/RI, systemic inhibition of C5, congenital deficiency of C5 and C6 all convey protection against injury [58, 59, 81]. Our results in Chapter 6 as well as literature suggest that the female specific protection in experimental mice models can be attributed to lower level of terminal pathway activity and impaired potential to generate C5a.
Importantly for experimental models, similar female-specific protection has been reported in rats, with evidence that female rats have protection against different complement driven injuries, such as renal I/RI and rejection [72, 82, 83]. Future studies could be performed not only to determine C5 activity in mice with functional assays, but also to use and adapt the assays described in Chapter 2 to analyse gender differences in rat complement.
Systematic analysis of adult human complement has not revealed similar extensive differences in complement in analogy to mice. Similar but more subtle differences were reported by Seelen et al., showing lower AP and LP activity and lower C3 levels in females. However, no terminal pathway specific impact, in analogy to the mice, was observed [84]. Although it is clear that human and mouse
complement do not have similar gender differences, it is essential to include gender when interpreting results both in experimental and clinical settings. It is clear that female mice are virtually deficient in terminal pathway, and therefore complement mediated models in female mice are bound to underestimate the impact of complement, whereas models performed in male mice with intact complement pathways probably reflect the human setting better.
Previous studies have shown that renal and hepatic I/RI is attenuated through upregulation of cytoprotective proteins and amelioration of oxidative stress by dietary restriction [85]. However, the impact of dietary interventions on the predominantly liver derived serum complement [55] has not been evaluated previously, although it is known that fasting reduces liver synthesis of serum proteins [86]. Results in Chapter 7 revealed that fasting results in significant modulation of complement. Interestingly, this modulation resulted in a phenotype of the complement system similar to female mice described in Chapter 6, with severely limited terminal pathway activity. The pathway activities or hepatic expression of complement factors at the level of C3 were not prominently affected, except for MBL-A and MBL-C which could convey some protective effect as has been reported earlier [66, 87]. In contrast, fasting results in specific downregulation of terminal pathway specific factors in liver, which results in decreased terminal pathway activity together with decrease in serum levels of C6 and C9.
Together, our results in Chapter 6 and Chapter 7 highlight the importance of understanding the impact of physiology and environment to a seemingly constant system such as complement. In addition to gender and diet, also age of mice affects the activity of complement system [26]. These variables should be taken into account when comparing the results across different studies in literature and in drawing conclusions from preclinical studies.
8
7. CONCLUDING REMARKS AND FUTURE DIRECTIONS
The coming years in complement research will be exciting for both clinicians and fundamental research. Complement researchers now have the access to comprehensive complement toolkits, consisting of analysis tools and genetically modified rodents, for investigating complement in vitro and in vivo.
We contributed to this toolkit with novel validated functional complement assays, which are ideal for verification complement inhibition and acute consumption in vivo and in vitro and with complement biomarker assays that prove to be useful in monitoring complement activation dynamics in vivo [2, 24, 25]. Experimental rodent models and complement knockout mice have been essential within this thesis and will continue to be the central tool for preclinical and fundamental studies on complement activation in health and disease. However, precautions should be taken in experimental design and in interpretation of results, as we showed that in mouse the gender, strain, and diet all have a major impact on the complement system. Furthermore, our results show that local and circulating complement can be significantly affected by inflammation, which in turn can result in changes of complement turnover and synthesis. These factors should be taken into account when designing experimental studies, choosing analysis tools, and interpreting results.
The ever increasing understanding of local and systemic complement requires considerable efforts to differentiate the main contributors of the injury. The complement in local injury can in theory be derived from three sources: 1) serum complement, which is predominantly liver derived, 2) locally expressed complement which has both basal and inducible characteristics, and 3) inflammatory cell derived complement which can be locally secreted upon stimulation [55, 88–90]. Our results are in line with others, who have shown that both serum and local complement is subject to considerable modulation
following injury, inflammation, and nutritional status [60, 91–94]. Together these factors are important aspects to consider during experiment planning and interpreting the results on complement activation in vivo.
The expanding understanding of the complement activation mechanisms and means to investigate them in vivo and in vitro has laid the groundwork for advances in better treatment of renal disease and injury. Our results with therapeutic anti-MBL treatment in context of renal I/RI continues to fortify the concept of MBL/MASP dependent, complement independent mode of injury with the experimental rat model, allowing researchers to move towards clinical trials [23, 50, 67, 87]. However, the discrepancy between the mouse and rat models is clearly evident in our studies, as the role of AP and C5a in mouse model is clear and in line with literature, suggesting that further studies are required in both species with further clinical studies. The role of properdin in complement system has undergone a recent revision with first solid evidence of its role as a PRM in vitro and in vivo [7, 13, 95]. This shift in paradigm, with concurrent development of key properdin research tools, will allow researchers to address its role in health and disease, with potential therapeutic applications in conditions such as anti-GBM and renal I/RI [56, 96]. Furthermore, the role of properdin in other diseases such as ANCA- vasculitis may now be addressed in detail [13, 97].
In addition to the activation mechanisms of complement, also the effector mechanisms have received noticeable interest in recent years. Recent studies have shown that anaphylatoxin receptors exhibit differential expression and localization within kidney, allowing researchers to identify clues to novel effector mechanisms of complement in vivo [61, 98]. Indeed, our work suggests that C5aR and C5L2 have independent contribution to renal I/RI, however further work is still required to understand the molecular mechanisms by which C5L2 mediated injury is exerted.
8
In conclusion, the complement system has a major and multifaceted role in renal injury and disease. Novel complement analysis tools, first in class complement therapeutics, sophisticated experimental models, and detailed understanding of complement contribution to diseases and injuries will allow future research to shift complement specific therapeutic applications towards the clinic.
references
1. J. a. McCaughan, D. M. O'Rourke, a. E. Courtney, Recurrent dense deposit disease after renal transplantation: An emerging role for complementary therapies. Am. J. Transplant. 12, 1046–1051 (2012).
2. R. B. Sim, W. Schwaeble, T. Fujita, Complement research in the 18th-21st centuries: progress comes with new technolog y. Immunobiolog y (2016), doi:10.1016/j.imbio.2016.06.011.
3. R. Roozendaal, M. C. Carroll, Complement receptors CD21 and CD35 in humoral immunity. Immunol. Rev.
219, 157–166 (2007).
4. K. M. Cunnion, D. K. Benjamin, C. G. Hester, M. M. Frank, Role of complement receptors 1 and 2 (CD35 and CD21), C3, C4, and C5 in survival by mice of Staphylococcus aureus bacteremia. J. Lab. Clin. Med. 143, 358–365 (2004).
5. Y. Wu, B. Brookshire, R. Verani, F. Arnett, C. Yu, Clinical presentations and molecular basis of complement C1r deficiency in a male African-American patient with systemic lupus erythematosus. Lupus. 20, 1126–1134 (2011).
6. L. a. Trouw, a. M. Blom, P. Gasque, Role of complement and complement regulators in the removal of apoptotic cells. Mol. Immunol. 45, 1199–1207 (2008).
7. D. Spitzer, L. M. Mitchell, J. P. Atkinson, D. E. Hourcade, Properdin can initiate complement activation by binding specific target surfaces and providing a platform for de novo convertase assembly. J. Immunol. 179, 2600–2608 (2007).
8. D. Ricklin, G. Hajishengallis, K. Yang, J. D. Lambris, Complement: a key system for immune surveillance and homeostasis. Nat. Immunol. 11, 785–797 (2010).
9. D. Ricklin, E. S. Reis, J. D. Lambris, Complement in disease: a defence system turning offensive. Nat. Rev.
Nephrol. 12, 383–401 (2016).
10. C. Kemper, J. P. Atkinson, D. E. Hourcade, Properdin: emerging roles of a pattern-recognition molecule. Annu.
Rev. Immunol. 28, 131–155 (2010).
11. M. Harboe et al., The role of properdin in z ymosan- and Escherichia coli-induced complement activation. J.
Immunol. 189, 2606–13 (2012).
12. L. Pillemer et al., The properdin system and immunity. I. Demonstration and isolation of a new serum protein, properdin, and its role in immune phenomena. Science. 120, 279–85 (1954).
13. J. O’Flynn, K. O. Dixon, M. C. Faber Krol, M. R. Daha, C. Van Kooten, Myeloperoxidase directs properdin- mediated complement activation. J. Innate Immun. 6, 417–425 (2014).
14. Y. Liu et al., Human M-ficolin is a secretory protein that activates the lectin complement pathway. J. Immunol.
175, 3150–3156 (2005).
15. Y. Aoyagi et al., L-ficolin/mannose-binding lectin-associated serine protease complexes bind to group B streptococci primarily through N-acetylneuraminic acid of capsular polysaccharide and activate the complement pathway. Infect.
Immun. 76, 179–188 (2008).
16. R. M. Zacho, L. Jensen, R. Terp, J. C. Jensenius, S. Thiel, Studies of the pattern recognition molecule H-ficolin:
Specificity and purification. J. Biol. Chem. 287, 8071–8081 (2012).
17. J. S. Presanis, M. Kojima, R. B. Sim, Biochemistry and genetics of mannan-binding lectin (MBL). Biochem. Soc.
Trans. 31, 748–752 (2003).
18. K. Hajela et al., The biological functions of MBL-associated serine proteases (MASPs). Immunobiolog y. 205, 467–475 (2002).
19. S. H. Sacks, Complement fragments C3a and C5a: The salt and pepper of the immune response. Eur. J. Immunol.
40, 668–670 (2010).
20. A. Klos et al., The role of the anaphylatoxins in health and disease. Mol. Immunol. 46, 2753–2766 (2009).
21. R. Ma, Z. Cui, Y. H. Liao, M. H. Zhao, Complement activation contributes to the injury and outcome of kidney in human anti-glomerular basement membrane disease. J. Clin. Immunol. 33, 172–178 (2013).
22. R. Vargha et al., C4d in pediatric renal allograft biopsies: a marker for negative outcome in steroid-resistant rejection. Pediatr. Transplant. 10, 449–53 (2006).
23. G. Castellano et al., Therapeutic targeting of classical and lectin pathways of complement protects from ischemia- reperfusion-induced renal damage. Am. J. Pathol. 176, 1648–1659 (2010).
24. T. E. Mollnes et al., Complement analysis in the 21st century. Mol. Immunol. 44, 3838–3849 (2007).
25. B. Nilsson, K. N. Ekdahl, Complement diagnostics: concepts, indications, and practical guidelines. Clin. Dev.
Immunol. 2012, 962702 (2012).
26. G. L. Ong, M. J. Mattes, Mouse strains with typical mammalian levels of complement activity. J. Immunol.
Methods. 125, 147–158 (1989).
27. van Dijk, P. M. Rademaker, J. M. N. Willers, Estimation of classical pathway of mouse complement activity by use of sensitized rabbit erythrocytes. J. Immunol. Methods. 39, 257–268 (1980).
28. van Dijk, P. M. Rademaker, J. M. N. Willers, Determination of alternative pathway of complement activity in mouse serum using rabbit erythrocytes. J. Immunol. Methods. 36, 29–39 (1980).
29. I. Osmers, A. J. Szalai, A. J. Tenner, S. R. Barnum, Complement in BuB/BnJ mice revisited: Serum C3 levels and complement opsonic activity are not elevated. Mol. Immunol. 43, 1722–1725 (2006).
30. J. Zwirner, E. Felber, C. Reiter, G. Riethmüller, H. E. Feucht, Deposition of complement activation products on plastic-adsorbed immunoglobulins. A simple ELISA technique for the detection of defined complement deficiencies.
J. Immunol. Methods. 124, 121–129 (1989).
31. A. Roos et al., Functional characterization of the lectin pathway of complement in human serum. Mol. Immunol.
39, 655–668 (2003).
32. G. N. Fredrikson, L. Truedsson, a G. Sjöholm, New procedure for the detection of complement deficiency by ELISA. Analysis of activation pathways and circumvention of rheumatoid factor influence. J. Immunol. Methods.
166, 263–270 (1993).
33. F. Beurskens, P. Aerts, D. M. Robins, H. van Dijk, Sex-limited protein: in-vitro and in-vivo functions. Mol.
Immunol. 35, 356 (1998).
34. P. van der Pol, A. Roos, S. P. Berger, M. R. Daha, C. van Kooten, Natural IgM antibodies are involved in the activation of complement by hypoxic human tubular cells. Am. J. Physiol. Renal Physiol. 300, F932–F940 (2011).
35. M. a. Seelen et al., Functional analysis of the classical, alternative, and MBL pathways of the complement system:
Standardization and validation of a simple ELISA. J. Immunol. Methods. 296, 187–198 (2005).
36. S. Parasuraman, R. Raveendran, R. Kesavan, Blood sample collection in small laboratory animals. J. Pharmacol.
Pharmacother. 1, 87–93 (2010).
37. Y. Palarasah et al., Sodium polyanethole sulfonate as an inhibitor of activation of complement function in blood culture systems. J. Clin. Microbiol. 48, 908–914 (2010).
38. B. Woehrl et al., Complement component 5 contributes to poor disease outcome in humans and mice with pneumococcal meningitis. J. Clin. Invest. 121, 3943–3953 (2011).
39. Y. Morita et al., Complement activation products in the urine from proteinuric patients. J. Am. Soc. Nephrol. 11, 700–707 (2000).
40. E. Ballanti et al., Role of the complement system in rheumatoid arthritis and psoriatic arthritis: Relationship with anti-TNF inhibitors. Autoimmun. Rev. 10, 617–623 (2011).
8
41. M. Huber-Lang et al., Generation of C5a in the absence of C3: a new complement activation pathway. Nat. Med.
12, 682–687 (2006).
42. U. Amara et al., Molecular intercommunication between the complement and coagulation systems. J. Immunol.
185, 5628–5636 (2010).
43. G. Bergseth et al., An international serum standard for application in assays to detect human complement activation products. Mol. Immunol. 56, 232–239 (2013).
44. E. B. Volokhina et al., Sensitive, reliable and easy-performed laboratory monitoring of eculizumab therapy in atypical hemolytic uremic syndrome. Clin. Immunol. 160, 237–43 (2015).
45. R. F. Hallstensen et al., Eculizumab treatment during pregnancy does not affect the complement system activity of the newborn. Immunobiolog y. 220, 452–9 (2015).
46. M. Cugno et al., Complement functional tests for monitoring eculizumab treatment in patients with atypical hemolytic uremic syndrome. J. Thromb. Haemost. 12, 1440–8 (2014).
47. R. Rieben et al., Immunoglobulin M-enriched human intravenous immunoglobulin prevents complement activation in vitro and in vivo in a rat model of acute inflammation. Blood. 93, 942–951 (1999).
48. T. E. Mollnes et al., High-dose intravenous immunoglobulin treatment activates complement in vivo. Scand. J.
Immunol. 48, 312–7 (1998).
49. P. J. Späth, G. Granata, F. La Marra, T. W. Kuijpers, I. Quinti, On the dark side of therapies with immunoglobulin concentrates: the adverse events. Front. Immunol. 6, 11 (2015).
50. P. van der Pol et al., Mannan-binding lectin mediates renal ischemia/reperfusion injury independent of complement activation. Am. J. Transplant. 12, 877–887 (2012).
51. R. Ma et al., The alternative pathway of complement activation may be involved in the renal damage of human anti-glomerular basement membrane disease. PLoS One. 9, e91250 (2014).
52. N. S. Sheerin, T. Springall, M. C. Carroll, B. Hartley, S. H. Sacks, Protection against anti-glomerular basement membrane (GBM)-mediated nephritis in C3- and C4-deficient mice. Clin. Exp. Immunol. 110, 403–409 (1997).
53. M. a Otten et al., Both complement and IgG fc receptors are required for development of attenuated antiglomerular basement membrane nephritis in mice. J. Immunol. 183, 3980–3988 (2009).
54. D. K. de Vries et al., Acute but transient release of terminal complement complex after reperfusion in clinical kidney transplantation. Transplantation. 95, 816–20 (2013).
55. B. P. Morgan, P. Gasque, Extrahepatic complement biosynthesis: where, when and why? Clin. Exp. Immunol.
107, 1–7 (1997).
56. T. Miwa, S. Sato, D. Gullipalli, M. Nangaku, W.-C. Song, Blocking Properdin, the Alternative Pathway, and Anaphylatoxin Receptors Ameliorates Renal Ischemia-Reperfusion Injury in Decay-Accelerating Factor and CD59 Double-Knockout Mice. J. Immunol. 190, 3552–3559 (2013).
57. B. de Vries et al., Complement factor C5a mediates renal ischemia-reperfusion injury independent from neutrophils.
J. Immunol. 170, 3883–3889 (2003).
58. W. Zhou et al., Predominant role for C5b-9 in renal ischemia/reperfusion injury. J. Clin. Invest. 105, 1363–1371 (2000).
59. B. de Vries et al., Inhibition of complement factor C5 protects against renal ischemia-reperfusion injury: inhibition of late apoptosis and inflammation. Transplantation. 75, 375–382 (2003).
60. J. Damman et al., Local renal complement C3 induction by donor brain death is associated with reduced renal allograft function after transplantation. Nephrol. Dial. Transplant. 26, 2345–2354 (2011).
61. M. B. van Werkhoven et al., Novel insights in localization and expression levels of C5aR and C5L2 under native and post-transplant conditions in the kidney. Mol. Immunol. 53, 237–245 (2013).
62. C. Loirat, V. Frémeaux-Bacchi, Atypical hemolytic uremic syndrome. Orphanet J. Rare Dis. 6, 60 (2011).
63. R. J. Kelly et al., Long-term treatment with eculizumab in paroxysmal nocturnal hemoglobinuria: Sustained efficacy and improved survival. Blood. 117, 6786–6792 (2011).
64. F. González-Roncero et al., Eculizumab treatment of acute antibody-mediated rejection in renal transplantation:
Case reports. Transplant. Proc. 44, 2690–2694 (2012).
65. A. Mella, Complement cascade and kidney transplantation: The rediscovery of an ancient enemy. World J.
Transplant. 4, 168 (2014).
66. M. Møller-Kristensen et al., Mannan-binding lectin recognizes structures on ischaemic reperfused mouse kidneys and is implicated in tissue injury. Scand. J. Immunol. 61, 426–434 (2005).
67. E. Asgari et al., FASEB J., in press, doi:10.1096/fj.13-246306.
68. T. Hellmark, M. Segelmark, Diagnosis and classification of Goodpasture’s disease (anti-GBM). J. Autoimmun.
48–49, 108–112 (2014).
69. L. Camous et al., Complement alternative pathway acts as a positive feedback amplification of neutrophil activation.
Blood. 117, 1340–1349 (2011).
70. W. Xu et al., Properdin binds to late apoptotic and necrotic cells independently of C3b and regulates alternative pathway complement activation. J. Immunol. 180, 7613–7621 (2008).
71. H. Hu, G. Wang, F. Batteux, C. Nicco, Gender differences in the susceptibility to renal ischemia-reperfusion injury in BALB/c mice. Tohoku J. Exp. Med. 218, 325–329 (2009).
72. V. Müller et al., Sexual dimorphism in renal ischemia-reperfusion injury in rats: possible role of endothelin. Kidney Int. 62, 1364–1371 (2002).
73. A. Fekete et al., Sex differences in the alterations of Na(+), K(+)-ATPase following ischaemia-reperfusion injury in the rat kidney. J. Physiol. 555, 471–480 (2004).
74. H. Harada et al., Sexual dimorphism in reduced-size liver ischemia and reperfusion injury in mice: role of endothelial cell nitric oxide synthase. Proc. Natl. Acad. Sci. U. S. A. 100, 739–744 (2003).
75. B. Cinander, S. Dubiski, a C. Wardlaw, Distribution, inheritance, and properties of an antigen, MUB1, and its relation to hemolytic complement. J. Exp. Med. 120, 897–924 (1964).
76. P. a. Ward, The harmful Role of C5a on innate immunity in sepsis. J. Innate Immun. 2, 439–445 (2010).
77. a Baba, T. Fujita, N. Tamura, Sexual dimorphism of the fifth component of mouse complement. J. Exp. Med.
160, 411–419 (1984).
78. Y. Fukuoka, J. Seino, T. Okuda, T. Tachibana, Development of a hemolytic assay for mouse complement components in sera and the variation of their levels with age. Tohoku J. Exp. Med. 137, 79–90 (1982).
79. R. B. Ashman, P. H. Kay, D. M. Lynch, J. M. Papadimitriou, Murine candidiasis: sex differences in the severity of tissue lesions are not associated with levels of serum C3 and C5. Immunol. Cell Biol. 69 ( Pt 1), 7–10 (1991).
80. R. T. Strait et al., MHC class I-specific antibody binding to nonhematopoietic cells drives complement activation to induce transfusion-related acute lung injury in mice. J. Exp. Med. 208, 2525–2544 (2011).
81. J. a. Buras et al., Inhibition of C5 or absence of C6 protects from sepsis mortality. Immunobiolog y. 209, 629–635 (2004).
82. A. Kher et al., Cellular and molecular mechanisms of sex differences in renal ischemia-reperfusion injury. Cardiovasc.
Res. 67, 594–603 (2005).
83. V. Müller et al., Sex hormones and gender-related differences: their influence on chronic renal allograft rejection.
Kidney Int. 55, 2011–2020 (1999).
84. M. Seelen, Complement in health and disease (2005), (available at https://openaccess.leidenuniv.nl/handle/1887/3738).
85. J. R. Mitchell et al., Short-term dietary restriction and fasting precondition against ischemia reperfusion injury in mice. Aging Cell. 9, 40–53 (2010).
86. C. H. Barrows, G. Kokkonen, The effect of age and diet on the cellular protein synthesis of liver of male mice. Age (Omaha). 10, 54–57 (1987).
87. S. Shushimita et al., Mannan-Binding Lectin Is Involved in the Protection against Renal Ischemia/Reperfusion Injury by Dietary Restriction. PLoS One. 10, e0137795 (2015).
88. W. Schwaeble et al., Expression of properdin in human monocytes. Eur. J. Biochem. 219, 759–764 (1994).
89. a K. Høgåsen, R. Würzner, T. G. Abrahamsen, M. P. Dierich, Human polymorphonuclear leukocytes store large amounts of terminal complement components C7 and C6, which may be released on stimulation. J. Immunol.
154, 4734–4740 (1995).
90. M. Botto, D. Lissandrini, C. Sorio, M. J. Walport, Biosynthesis and secretion of complement component (C3) by activated human polymorphonuclear leukocytes. J. Immunol. 149, 1348–1355 (1992).
91. K. Sastry et al., The human mannose-binding protein gene. Exon structure reveals its evolutionary relationship to a human pulmonary surfactant gene and localization to chromosome 10. J. Exp. Med. 170, 1175–1189 (1989).
92. U. Manthei, R. C. Strunk, P. C. Giclas, Acute local inflammation alters synthesis, distribution, and catabolism of the third component of complement (C3) in rabbits. J. Clin. Invest. 74, 424–433 (1984).
8
93. S. E. Degn et al., Biological variations of MASP-3 and MAp44, two splice products of the MASP1 gene involved in regulation of the complement system. J. Immunol. Methods. 361, 37–50 (2010).
94. S. Shushimita et al., Dietary restriction and fasting arrest B and T cell development and increase mature B and T cell numbers in bone marrow. PLoS One. 9, e87772 (2014).
95. C. Kemper, L. M. Mitchell, L. Zhang, D. E. Hourcade, The complement protein properdin binds apoptotic T cells and promotes complement activation and phagocytosis. Proc. Natl. Acad. Sci. U. S. A. 105, 9023–9028 (2008).
96. J. Miao et al., Tissue-specific deletion of Crry from mouse proximal tubular epithelial cells increases susceptibility to renal ischemia–reperfusion injury. Kidney Int. 86, 726–737 (2014).
97. S.-J. Gou, J. Yuan, M. Chen, F. Yu, M.-H. Zhao, Circulating complement activation in patients with anti- neutrophil cytoplasmic antibody–associated vasculitis. Kidney Int. 83, 129–37 (2012).
98. R. Li, L. G. Coulthard, M. C. L. Wu, S. M. Taylor, T. M. Woodruff, C5L2: A controversial receptor of complement anaphylatoxin, C5a. FASEB J. 27, 855–864 (2013).
