Introduction
The metabolic syndrome is a cluster of cardiovascular risk factors that is characterized by central obesity, dyslipidemia, hypertension, and insulin resistance. Obesity is one of the initial events in the pathological processes that define the metabolic syndrome [
1]. Dysfunctional adipose tissue is considered an important driver in the development of the adverse metabolic profiles in people with obesity [
1].
The complement system is an essential part of the innate immune system. It is widely expressed in adipose tissue, and its expression is increased in adipose tissue of obese individuals. The complement system has three main activation pathways: the alternative, the classical, and the lectin pathway (as reviewed in [
2], Supplementary Fig.
S1). Previous research has shown that the expression of some complement components, especially of the alternative and classical pathways, is higher in adipose tissue of people with obesity (as reviewed in [
3]). Several, mainly cross-sectional, studies have addressed the associations of individual complement factors with adverse metabolic profiles (as reviewed in [
4]). However, these studies often focused on one or a few complement components and on a specific aspect of metabolism. At the same time, studies on the associations of complement with metabolic diseases mostly focused on advanced disease states such as CVD (as reviewed in [
5]) or T2DM [
6,
7]. As it stands, the association of complement with the metabolic syndrome, which is the main comorbidity of obesity and predisposes to both CVD and T2DM [
8], has received much less attention.
Most available studies on complement and the metabolic syndrome are cross-sectional and mainly focused on C3 [
9‐
11], the central component of the alternative pathway, and its degradation product C3a-desarg (also known as acylation-stimulating protein, ASP) [
12,
13]. To the best of our knowledge, the prospective association of C3 with the metabolic syndrome was only reported in two human observational studies. One demonstrated that Caucasian individuals in highest C3 quartile had a higher risk to develop the metabolic syndrome during 3 years of follow-up [
14]. The other study was performed in Chinese men and reported a similar positive association for C3. In this latter study, C4, the downstream component of the classical (and lectin) pathway, was also reported to be positively associated with the development of the metabolic syndrome [
15]. The information on the relationships, especially the prospective ones, of other complement proteins, regulators, and activated products with the metabolic syndrome is even more limited, particularly in humans. Factor D [FD], also known as adipsin, is the rate-limiting enzyme of alternative pathway and is mainly produced by adipocytes. Although previous data revealed a possible role for FD in lipid metabolism and β cell function [
16,
17], little is known about its association with the metabolic syndrome. Factor B [FB], factor H [FH], and properdin are important components and regulators of the alternative complement pathway. Thus far, only one cross-sectional study in Caucasian men reported positive associations of FH and FB with several aspects of metabolism [
18]. Although a possible role of properdin in lipid metabolism was recently demonstrated in a properdin-deficient mouse model [
19], human data are still lacking. In addition, Hillian et al. revealed a protective effect of C1q, the initiator of classical pathway activation, on high-fat diet-induced hepatic insulin resistance and impaired glucose homeostasis in a C1q-knockout mouse model. And apolipoprotein E-deficient mice on an atherogenic diet that were treated with C1-inhibitor [C1-INH] showed a decrease in serum triglyceride (TG) [
20]. Yet again, except for some information on C4, human data on the association of the classical pathway with the metabolic syndrome are largely lacking.
In the present study, we hypothesized that activation of the alternative and the classical complement pathways, represented by higher circulating levels of their components, is associated with the metabolic syndrome. Therefore, we investigated in a Caucasian cohort with moderately increased risk of cardiometabolic disease, the associations of these components, especially those of the alternative (i.e., C3, C3a, Bb, FD, FH, and properdin) and classical (i.e., C1q, C1-INH, and C4) pathways with the prevalence of the metabolic syndrome, as well as its incidence during a 7-year follow-up period.
Discussion
We here show that at baseline, C3 and C4, but also systemic concentrations of C3a, FH, and properdin, were higher in individuals with metabolic syndrome compared to those without. We also show that baseline concentrations of complement C3 and C4 were positively and significantly associated with development of the metabolic syndrome during the 7-year follow-up period. In contrast, this was not the case for the other complement components (i.e., C3a, Bb, FD, FH, properdin, C1q, and C1-INH).
Cross-sectional associations of complement with components of the metabolic syndrome, such as obesity [
33] and insulin resistance [
34], were described more than 10 years ago. Cross-sectional associations of complement components, mainly C3 and its downstream product C3adesarg/ASP, with the metabolic syndrome were also reported previously [
9‐
13]. In line with this, we confirm the positive associations of C3 and C3a (the precursor of ASP) with prevalence of the metabolic syndrome. Also, we extend previous findings by reporting the associations of factor Bb, FD, FH, and properdin (components and regulators of the alternative pathway), with presence of the metabolic syndrome (significant only for FH and properdin). In our current evaluations, the associations of the proximal factors of the classical pathway, (C1q and C1-INH), with the metabolic syndrome were modest. Only C4 was significantly associated with the presence of the metabolic syndrome. Positive associations between C4 and prevalence of the metabolic syndrome were reported previously [
12,
15], although not consistently [
11]. Thus, our cross-sectional observations indeed suggested that activation of the complement system, especially the alternative pathway, is related to the metabolic syndrome.
We also evaluated to what extent the components of the complement system were associated with incidence of the metabolic syndrome over a 7-year follow-up period. Positive associations were observed for C3 and C3a, but only reached statistical significance for C3. Notably, the associations of all the other components of the alternative pathway, including FH and properdin, were non-significant with ORs close to 1. This strongly suggests that, despite the observed strong cross-sectional associations, FH and properdin are not risk factors for the development of metabolic syndrome. Two previous human studies have reported on the role of C3 in the development of the metabolic syndrome [
14,
15], and their results are in line with our current observation. We additionally observed a very robust association of C4, but not the proximal components of the classical pathway, with development of the metabolic syndrome. To the best of our knowledge, only one previous study reported on the association between C4 and incident metabolic syndrome [
15]. In that study, a positive association with incidence of the metabolic syndrome was shown but no adjustments were made for components of the metabolic syndrome. So taken together, in our prospective cohort, systemic concentration of the major complement components C3 and C4, but not their activators or activated products, were associated with incident metabolic syndrome.
The association with incidence of the metabolic syndrome was stronger for C4 than C3. Moreover, for C4, but not for C3, this association was independent of baseline levels of the individual components of the metabolic syndrome and its main underlying metabolic aberrancies. This suggests that the effects of C3 and C4 on the development of the metabolic syndrome are distinct. This is for instance illustrated by the observation that the association of complement C3, but not of C4, with incident metabolic syndrome was partly explained by liver function. Also, activation of the C3–C3a–C3adesArg/ASP axis may provide a partial explanation for the observation that C3 is a risk factor for the development of the metabolic syndrome. Experimental data showed that C3adesArg/ASP stimulates TG synthesis and glucose uptake and inhibits hormone-sensitive lipase in several cell types (as reviewed in [
4]). These known effects of C3 and its activation products may affect lipid metabolism and thus contribute to the development of the metabolic syndrome. An exciting novel mechanism of complement activation is the so-called intracellular complement system. Intracellular complement activation was first identified in T cells and has been implicated in the regulation of increased glycolysis and oxidative phosphorylation in Th1 cells [
35]. In the circulation, C3 can be spontaneously activated and transformed into C3(H
2O), the hydrolytic product of C3, via the so-called tick-over mechanism. And recently it was established that many cell types can take up C3(H
2O) from plasma, and part of this intracellular C3(H
2O) provides an intracellular source of C3a via a process that is independent of C3 convertase [
36]. This latter finding is particularly interesting given our current observation that C3, but not FH or properdin, was associated with the development of the metabolic syndrome. The effects of intracellular C3 activation have not yet been established. However, given the effects of intracellular complement activation on TH1 metabolism, intracellular C3 activation will likely affect metabolism in many cell types and, as such, potentially affect the metabolic dysregulation that characterizes the metabolic syndrome. Notably, C4 cannot be spontaneously activated via the tick-over mechanism, which illustrates that the above-mentioned intracellular route may be particularly relevant for C3. This may add to the possibility that C3 and C4 contribute to the metabolic syndrome via distinct mechanisms.
The available information on C4 in relation to metabolism is limited. Genetic variation in the C4 binding protein (C4BP), which is an inhibitor of the classical and the lectin pathway, has been related to higher blood pressure and higher fasting blood glucose [
37]. C4BP may also have a protective effect on β cell function [
38]. In addition, recent developments showed that C4a, which is generated upon activation of C4, can bind to and activate the G-coupled protein receptor protease-activated receptor (PAR)1 and PAR4 [
39]. Notably, this receptor cannot be activated by C3a [
39]. Activation of PAR1 and 4 by C4a leads to cellular activation and enhanced endothelial permeability [
39], and PAR1 has been previously implicated in, e.g., cardiac remodeling [
40] and hepatic injury [
43]. Notably, C4a cannot signal via the C3aR [
41]. These data, again, support the notion that in addition to their well-known conjunction on the terminal complement pathway, activation of C3 and C4 may indeed have distinct effects on metabolism. This also opens the path toward the investigation of a possible relation of C4, activated via the classical or the lectin pathway, with platelet activation and/or endothelial dysfunction, irrespective of the extent of terminal pathway activation.
The major strength of our study is that it provides information on several components and regulators of the alternative and the classical complement pathways, measured within one cohort, in the relation to the metabolic syndrome. Another key strength is that it provides prospective information on the development of the metabolic syndrome. Nevertheless, some limitations need to be acknowledged. The number of cases in the prospective analyses was relatively small, which may have limited our power to detect associations when effect sizes were smaller than those of C3 and C4. This may have been the case for C3a and perhaps also C1-INH. With respect to the interpretation of the data, we are limited by the fact that all complement measures were obtained in plasma which does not provide insight in local activation of these complement components relevant metabolic organs. Also, our cohort consists of individuals with a moderately increased risk to develop cardiometabolic diseases, which limits the generalizability of our findings. In addition, our study may be limited by the missing of information for several participants in the cross-sectional (6%) and the prospective (10%) analyses. However, our current observations are substantiated by their consistency with previous reports. Moreover, given our observational study design, we cannot draw causal conclusions on the relationship between complement and the metabolic syndrome. However, our findings are consistent with previous reports and are also supported by suggestive biological evidence, such as the altered metabolic profiles in several complement gene-deficient mouse models [
17,
19,
42], which suggests that complement may play a role in the development of metabolic syndrome.
In conclusion, we herein report that complement C3 and C4 are positively associated with incidence of the metabolic syndrome. For C3, but not for C4, these effects appear to be related to the metabolic profile at baseline. Importantly, properdin and FH, which showed strong associations in the cross-sectional analyses, were not related to incident metabolic syndrome. In the discussion, we refer to novel developments in our understanding of activation and signaling of complement in order to provide some explanation for these divergent observations (see graphical summary in Fig.
S3), but further work is needed to better understand the etiological role of complement activation in the cellular and metabolic pathways that underlie the development of the metabolic syndrome and related cardiometabolic diseases. Future work should include direct metabolic effects of C3 and C4 as well as effects of their activation products, C3a/C3a-desarg and C4a.
Electronic supplementary material