Heart failure (HF) is a chronic medical condition characterized by the heart's inability
to efficiently pump blood, often resulting from recurrent myocardial infarctions (MIs).[1 ]
[2 ] Despite improvements in risk management and interventional strategies, HF presents
significant health care challenges and leads to increased health care expenditure
due to gaps in HF therapy targets.[3 ] Therefore, it is essential to identify treatable traits.[4 ] Cardiac repair is a highly regulated process consisting of inflammatory, proliferation,
and remodeling phases overlapping each other. This follows necrotic loss of cardiomyocytes
after MI in an attempt to repair the damage and restore heart function. Mechanisms
in the repair process following MI include activation of the complement system as
part of the innate immune response. The complement anaphylatoxin C5a interacts primarily
with its two receptors, the classical proinflammatory C5a receptor 1 (C5aR1, CD88)
and C5a receptor 2 (C5aR2, C5L2 [C5a receptor like 2]). Previous studies have implicated
C5aR1 in cardiac regeneration after left ventricular apical resection[5 ] and cardiac inflammation.[6 ]
[7 ] However, its role in cardiac repair processes following MI is poorly defined and
the role of C5aR2 in MI is largely unknown. Our current study focused on late phases
after MI up to 4 weeks after the infarction insult, a time window that specifically
reflects the outcome of the repair process and that so far has not been investigated.
Using C5ar1 - and C5ar2 -deficient mice, we studied the role of both C5aR1 and C5aR2 in cardiac function and
repair processes following MI. We combined functional analyses and mechanistic studies
in vivo and in vitro to demonstrate that targeting C5aR1 may provide a potent lever
to improve cardiac repair after MI.
To systematically study the role of C5a receptors in cardiac repair processes following
MI, wild-type C57Bl/6J, C5ar1−/−
and C5ar2−/−
mice were investigated after chronically ligating the left anterior descending artery
(LAD). Histological analysis revealed a significantly reduced MI size ([Fig. 1A, B ]) as well as lower collagen content in the infarcted area ([Fig. 1C ] and [Supplementary Fig. S1A ], available in online version the online version) in C5ar1−/−
mice compared with wild-type mice 4 weeks after MI. Infarct size and collagen content
were not reduced in C5ar2−/−
mice. This indicated a protective effect of C5ar1 deficiency in MI. Consistent with the reduced MI size, C5ar1−/−
mice showed a significantly increased ejection fraction and reduced end-diastolic
volume when compared with wild-type mice 4 weeks after MI, indicating an improved
ventricular function and contractility ([Supplementary Table S1 ], available in the online version). Cardiac repair after MI relies on an intense
inflammatory response that provides molecular cues for activation of reparative cells.[8 ] As recruitment of inflammatory cells and the proliferation of tissue-resident macrophages
in later stages are pivotal in this process, we analyzed the content of monocytes/macrophages
and neutrophils in the infarcted areas. One week after induction of MI, the content
of Mac3-positive monocytes/macrophages in the infarcted area peaked in wild-type mice
and was significantly reduced in C5ar1−/−
mice by 43% ([Fig. 1D, E ] and [Supplementary Fig. S1B ], available in the online version). This peak of monocyte/macrophage accumulation
drastically lowered at 4 weeks after MI without significant differences between groups
([Fig. 1E ]). In contrast, the content of MPO+ cardiac neutrophils transiently increased 24 hour
after MI in wild-type mice and this increase was significantly reduced in both C5ar1−/−
and C5ar2−/−
mice ([Supplementary Fig. S2 ], available in the online version). The reduced neutrophil accumulation in the infarcted
heart could provide some cardioprotective effects in both C5ar1−/−
and C5ar2−/−
mice. However, the observation that infarction size was only reduced in C5ar1−/−
mice suggested a predominant involvement of another C5ar1-mediated mechanism in cardiac
repair after MI, urging us to also examine effects on cell proliferation versus apoptosis
and the expression of genes implicated in the late phase of cardiac repair. At 1 week,
cell proliferation in the infarcted heart increased and the number of proliferating
cells was significantly further increased in both C5ar1−/−
and C5ar2−/−
mice compared with control wild-type mice ([Supplementary Fig. S3A, B ], available in the online version). Myocardial necrosis resulting from improper blood
perfusion to the cardiac tissue after MI is a detrimental event, and cardiac repair
mechanisms aim at removing necrotic tissue while inducing neovascularization. We therefore
also examined the effect of C5ar1 deficiency on myocardial necrosis in vivo and found reduced necrosis in C5ar1- deficient hearts 24 hour after MI ([Supplementary Fig. S3C–E ], available in the online version). Likewise, C5ar1 deficiency reduced apoptotic cells in the infarcted hearts ([Fig. 1F, G ] and [Supplementary Fig. S3F, G ], available in the online version).
Fig. 1 C5aR1 controls cardiac repair mechanisms following myocardial infarction (MI). (A –E ) MI was induced by chronically ligating the left anterior descending artery (LAD).
(A, B ) Histomorphometrical analysis of the infarcted myocardium 4 weeks after MI in wild-type
(WT), C5ar1−/−
and C5ar2−/−
mice (n = 7–9 per group) as measured by planimetry. Shown are representative Gomori's one-step
trichrome-stained sections (A ) and the quantification of infarcted areas (B ). (C ) Quantification of collagen content in infarcted area. (D ) Representative immunostaining of MAC3+ monocytes/macrophages in infarcted myocardium analyzed 1 week after MI. (E ) Quantification of myocardial infiltration of Mac3+ monocytes/macrophages analyzed 1 day, 1 week, and 4 weeks after MI. Cells were visualized
by immunostaining and the quantification is presented as the percentage of positively
stained cells of total cell count in infarcted area in field of view (FOV). n = 4–6 per group. (F, G ) Apoptosis rate in infarcted myocardium 24 hours after MI induction in WT and C5ar1
−/− mice detected by TUNEL staining. Shown is representative immunostaining (F ). The quantification of cells is presented as the percentage of positively stained
cells per FOV-infarcted area (G ). A complete set of these data are presented in [Supplementary Fig. S3C, D ] (available in the online version). (H –J ) MI was performed in WT and C5ar1−/−
mice, thereafter, infarcted heart areas were isolated 4 weeks after infarction. Representative
immunoblot (H ) and corresponding quantification of Tgf-β1 (I ) and Vegf-A (J ) normalized to actin. Representative staining (K ) and quantification (L ) of SMA+ /CD31− myofibroblasts in the infarcted myocardium. White arrows indicate SMA+ myofibroblasts (green fluorescence). Scale bars 100 µm. (M ) Cardiac fibroblasts isolated from WT and C5ar1−/−
mice were stimulated with TGF-β1 alone or together with C5a. Quantification of protein
expression of Vegf-A normalized to actin. (N, O ) Supernatant from TGF-β1-stimulated WT or C5ar1−/−
cardiac fibroblasts were incubated on WT-ECs or C5ar1−/−
-ECs respectively, in a coculture, and tube formation was determined in WT-ECs and
C5ar1−/−
-ECs. Shown are representative images (N ) and quantification of tube formation (O ). (P, Q ) Neoangiogenesis was assessed by CD31+ /α-SMA− staining of WT, and C5ar1−/−
infarcted heart areas 4 weeks after infarction. (P ) Representative CD31 staining of infarcted heart areas and (Q ) corresponding quantification. N = 7 mice per group. Data are presented as mean ± standard error of mean.
To further assess the mechanisms underlying the overall improved heart function upon
C5ar1 deficiency, we determined gene expression of key repair cytokines in infarcted hearts
from wild-type and C5ar1−/−
mice that underwent MI. Four weeks after MI (complete healing and maturation of the
scar), we found significantly increased expression of Tgf-β1 and Vegf-A both on protein
and mRNA levels, in C5ar1−/−
infarcted hearts, when compared with corresponding control wild-type hearts, whereas
the levels 1 week after MI were not significantly altered ([Fig. 1H–J ] and [Supplementary Fig. S4A–D ], available in the online version) . We further observed an increased expression of Vegf-A in TGF-β1-stimulated cardiac fibroblasts ([Supplementary Fig. S4E ], available in the online version). TGF-β1 was described as having a transient role
in macrophage polarization and myofibroblasts differentiation during healing after
MI.[9 ] Given the increased expression of Tgf-β1 and Vegf-A, a highly potent angiogenic
agent, during the maturation of the scar, we reasoned that C5ar1 deficiency may promote a transient myofibroblast differentiation response resulting
in improved cardiac repair. Surprisingly though, while the number of cardiac myofibroblasts
remained unchanged in both genotypes 1 week after MI ([Supplementary Fig. S4F, G ], available in the online version), the number of α-SMA+ cells in the infarcted myocardium was significantly reduced in C5ar1−/−
mice compared with wild-type controls 4 weeks after MI ([Fig. 1K, L ]). To further scrutinize the effect of the increased expression of Tgf-β1 in C5ar1−/−
mice, we isolated cardiac fibroblasts from wild-type mice and quantified the expression
of C5ar1. We found considerable expression of C5ar1 in cardiac fibroblasts, which
was not altered upon TGF-β1 stimulation ([Supplementary Fig. S5A-D ], available in the online version). We then asked whether C5a/C5aR1 signaling affects
the TGF-β1-driven conversion of cardiac fibroblasts to myofibroblasts. Exposure of
wild-type cardiac fibroblasts to TGF-β1 was able to induce transdifferentiation as
determined by α-SMA expression, but this was independent of C5a/C5aR1 signaling ([Supplementary Fig. S5E, F ], available in the online version). As overall myofibroblast transdifferentiation
was not affected upon C5ar1 deficiency, we hypothesized that the regulatory response of these cells was skewed
with consequences for myofibroblast function. To test this notion mechanistically,
cardiac fibroblasts isolated from wild-type and C5ar1−/−
mice were stimulated with TGF-β1. Analysis of mRNA and protein expression revealed
a TGF-β1-induced upregulation of Vegf-A levels in wild-type myofibroblasts, which
was significantly increased in C5ar1−/−
myofibroblasts ([Fig. 1M ] and [Supplementary Fig. S5G ], available in the online version). This is an indication that C5aR1-deficient myofibroblasts
may contribute to the VEGF pool in the myocardium, by increasing the balance toward
a proangiogenic phenotype of cardiac fibroblasts. After MI, neovascularization in
the border zone adjacent to the ischemic region helps to preserve cardiac function
and attenuate adverse left ventricular remodeling.[10 ] Our observation that the expression of Vegf-A is significantly upregulated in C5ar1−/−
myofibroblasts and in infarcted heart areas from C5ar1−/−
mice, potentially in part through effects on intracellular signaling including ERK
and p38 MAPKs,[11 ] led us to reason that upon stimulation, cardiac myofibroblasts may be a source of
Vegf-A for endothelial cells (ECs) to support neoangiogenesis. To investigate this,
we performed matrigel tube formation assays, where wild type-ECs were incubated with
supernatants derived from stimulated wild type-fibroblasts and C5ar1−/−
-ECs were exposed to C5ar1−/−
-myofibroblast supernatant, mimicking the in vivo microenvironment in our model. We
found increased tube formation in C5ar1−/−
-ECs/C5ar1−/−
-myofibroblast cocultures compared with cell responses elicited in wild-type cells
([Fig. 1N, O ]) indicating, to our knowledge, for the first time a C5a/C5aR1-axis-mediated EC–fibroblast
interaction in neovessel formation during the cardiac repair process following MI.
This was consistent with significantly improved neovascularization in infarcted heart
areas from C5ar1−/−
mice as revealed by increased newly formed CD31+ blood vessels 4 weeks after MI ([Fig. 1P, Q ]). Hence, the C5a/C5aR1-axis may mediate EC–fibroblast interactions in cardiac repair
processes and contribute to improved heart function observed in C5ar1−/−
mice 4 weeks after MI. The identified EC–fibroblast interaction requires additional
studies to further scrutinize the role of C5ar1 in cardiac fibroblast activation and
related mechanisms.
Collectively, the results presented here show that C5ar1 deficiency (1) reduces infarct size following MI and enhances overall cardiac function;
(2) attenuates myocardial necrosis; and (3) enhances VEGF production by myofibroblasts
and EC–fibroblast interactions to promote neovascularization in the infarcted heart.
Although the exact contribution of C5aR1 on ECs versus fibroblasts in this EC–fibroblast
interaction remains to be further clarified, our findings overall demonstrate an important
role for the C5a/C5aR1-axis in the late endogenous repair mechanism following MI.
This could complement cardioprotective effects provided by C5ar1 deficiency on other cell types including platelets, where C5ar1 deficiency promotes tissue neovascularization by reducing C5a-triggered secretion
of the anti-angiogenic factor CXCL4.[12 ] Also, other immune cells including dendritic cells and CD4+ T cells have been implicated in left ventricular remodeling and the progression of
cardiac dysfunction following MI.[13 ]
[14 ] As C5aR1 has been shown to regulate the function of both dendritic cells and CD4+ T cells,[15 ]
[16 ] we cannot exclude effects of C5aR1 on these cells in the current study. Considering
the pleiotropic role of C5aR1 in various cell types and signaling pathways, it is
crucial to dissect in future studies the precise cell-specific C5aR1-dependent mechanisms
involved in all phases of healing after MI: the acute inflammatory phase, the intermediate
proliferation phase, and the late fibrosis phase. Unveiling the distinct, cell type-
and time-dependent functions of C5aR1 signaling could further support the design of
personalized therapeutic strategies to improve cardiac repair after MI.
