We show here using a transgenic model that human C-reactive protein (CRP) protects against experimental allergic encephalomyelitis (EAE) in C57BL/6 mice. In transgenic compared with wild-type females, the duration of the human CRP acute phase response that accompanies the inductive phase of active EAE correlates with a delay in disease onset. In transgenic males, which have higher human CRP expression than females do, EAE is delayed, and its severity is reduced relative to same-sex controls. Furthermore, in male transgenics, there is little or no infiltration of the spinal cord by CD3+ T cells and CD11b+ monocytes and macrophages, and EAE is sometimes prevented altogether. CRP transgenics also resist EAE induced passively by transfer of encephalitogenic T cells from wild-type donors. Human CRP has three effects on cultured encephalitogenic cells that could contribute to the protective effect observed in vivo: 1) CRP inhibits encephalitogenic peptide-induced proliferation of T cells; 2) CRP inhibits production of inflammatory cytokines (TNF-α, IFN-γ) and chemokines (macrophage-inflammatory protein-1α, RANTES, monocyte chemoattractant protein-1); and 3) CRP increases IL-10 production. All three of these actions are realized in vitro only in the presence of high concentrations of human CRP. The combined data suggest that during the acute phase of inflammation accompanying EAE, the high level of circulating human CRP that is achieved in CRP-transgenic mice inhibits the damaging action of inflammatory cells and/or T cells that otherwise support onset and development of EAE.
The prototypic acute phase protein was termed C-reactive protein (CRP)3 because of its ability to bind C-polysaccharide from the cell wall of pneumococci (1). It is now recognized that human CRP plays a significant role in host defense against varied microbial pathogens (2, 3). The available data indicate that CRP contributes to pathogen elimination mainly by activating the complement system and enhancing complement-mediated opsonization (4, 5, 6, 7). However, more recent evidence indicates that CRP might express additional host-protective activity by binding to FcγRI (CD64) and FcγRII (CD32) expressed on monocytes, neutrophils, and lymphocytes (8, 9, 10, 11) and by binding to and opsonizing apoptotic cells (12). In addition to its elevation in the blood in response to infection, clinical studies have established that CRP blood levels also rise substantially (up to 1000-fold) in response to noninfectious inflammatory conditions including trauma, cancer, autoimmune diseases, and atherosclerosis (13, 14, 15). Regardless of the underlying mechanisms involved, the beneficial effects of CRP likely are not limited to defense against microbes.
Unlike its human counterpart, mouse CRP is synthesized only in trace amounts (16). Taking advantage of this species difference, Ciliberto et al. (17) constructed human CRP-transgenic (CRPtg) mice in which human CRP is expressed as an acute phase protein. We used C57BL/6-congenic descendants of these constructs to verify the antimicrobial activity of human CRP in an in vivo setting (2, 5, 6). Herein to determine whether human CRP might also have a beneficial role against aseptic inflammatory disease, we induced experimental allergic encephalomyelitis (EAE) in CRPtg mice. EAE serves as an animal model of multiple sclerosis (MS) and is characterized by demyelination and infiltration of the CNS by Th1, neuroantigen-specific T cells, monocytes, and macrophages. The disease can be induced actively in C57BL/6 mice by immunization with myelin oligodendrocyte protein (MOG) or immunodominant peptides derived from MOG, and a passive form of EAE can be induced in susceptible recipients by adoptive transfer of encephalitogenic T cells (18). We found that in CRPtg females onset of active EAE was delayed compared with sex-matched C57BL/6 controls. The delay in disease onset was coincident with acute phase expression of the CRP transgene. In male CRPtg, which express significantly higher amounts of human CRP than do females, EAE severity is further attenuated, and CNS infiltration by T cells and monocytes/macrophages is prevented. Like its actively induced counterpart, EAE induced passively by transfer of encephalitogenic T cells is delayed in CRPtg recipients. In vitro experiments demonstrated that human CRP reduces the proliferation of MOG peptide-stimulated encephalitogenic T cells and inhibits the production of proinflammatory Th1 cytokines (TNF-α and IFN-γ) and chemokines (macrophage-inflammatory protein-1α (MIP-1α), RANTES, and monocyte chemoattractant protein-1 (MCP-1)). In contrast, human CRP increases expression of the anti-inflammatory Th2 cytokine IL-10. On the basis of these observations, we speculate that human CRP exerts its protective effect in EAE in the fluid phase and during the acute phase response by inhibiting the development of a Th1-mediated immune response and by reducing the capacity of encephalitogenic T cells to recruit phagocytes to the CNS. These data suggest a role for CRP per se in authentic MS and indicate that in addition to its recognized antimicrobial activity, CRP might have an immunoregulatory function.
Materials and Methods
The CRP transgene (17) and its human-like pattern of expression in CRPtg mice has been fully described (4, 5, 6, 17, 19). CRPtg mice carry a 31-kb human DNA fragment containing the CRP gene; the gene per se is flanked by regions that include all the known cis-acting regulatory elements (i.e., the entire human promoter) and the CRP pseudogene. Human CRP is expressed by CRPtg on endotoxin treatment or infection and reaches blood levels comparable with those observed in human diseases (up to 1 mg/ml). cis-acting regulatory elements included in the human CRP transgene are responsible for both tissue specificity and acute phase inducibility of its expression in CRPtg mice, and the trans-acting factors required for its correct regulation are conserved from mice to humans. We discovered a sexually dimorphic pattern of expression of the CRPtg (reviewed in Ref. 6), viz baseline and acute phase expression of human CRP, is a magnitude higher in CRPtg males than in females.
We backcrossed CRPtg to wild-type C57BL/6J (The Jackson Laboratory, Bar Harbor, ME) for 10 generations to produce C57BL/6-congenic CRPtg and wild-type littermates for our studies. Mice were screened for inheritance of the CRP transgene using a human CRP-specific PCR (19) and were used in experiments when 8–12 wk old. All animal studies were approved by the Institutional Animal Care and Use Committee of the University of Alabama (Birmingham, AL).
Induction of active EAE
CRPtg and littermate non-tg mice were immunized with MOG peptide 35–55 as described (18). MOG peptide was synthesized by standard 9-fluorenylmethoxycarbonyl chemistry and was >95% pure as determined by reversed phase HPLC (Research Genetics, Huntsville, AL). On days 0 and 7, mice received s.c. an injection of 150 μg MOG peptide emulsified in CFA containing 500 μg heat-killed Mycobacterium tuberculosis (Difco, Detroit, MI). On days 0 and 2, mice received i.p. an injection of pertussis toxin (500 ng; List Biological Laboratories, Campbell, CA). Development of EAE symptoms was monitored twice daily using a standard clinical scale ranging from 0 to 6 as follows: 0, asymptomatic; 1, loss of tail tone; 2, flaccid tail; 3, incomplete paralysis of one or two hind limbs; 4, complete hind limb paralysis; 5, moribund (animals were humanely euthanized); 6, dead. Mice were observed for at least 30 days, and only mice with a score of at least 2 for >2 consecutive days were judged to have fully developed EAE. The maximum clinical score achieved by each animal during the 30-day observation period was used to calculate average maximum clinical scores for each experimental group. To study the time course of disease development, average clinical scores were calculated daily for each group of mice and plotted. When determining the average day of onset of EAE, animals that did not develop any symptoms of EAE during the 30-day survey period were assigned a day of onset of 31.
Induction of passive EAE
Two weeks after induction of active EAE, the spleens of non-tg donors were removed and used to make a single-cell suspension in RPMI 1640 (Life Technologies, Gaithersburg, MD). After treatment with ammonium chloride to lyse erythrocytes, encephalitogenic T cells were enriched by passage through nylon wool. These MOG peptide-sensitized cells were transferred to RPMI supplemented with 10% FCS, 5 × 10−5 2-ME, and penicillin/streptomycin (100 μg/ml) before seeding into six-well tissue culture plates. The cells (4 × 106) were then restimulated in vitro for 24 h with MOG peptide (20 μg/ml) in the presence of 2 × 106 freshly irradiated (naive) splenic APCs. IL-2 (20 U/ml) was added, and after an additional 24 h of culture, encephalitogenic T cells were purified by gradient centrifugation using Ficoll-Hypaque (Pharmacia, Peapack, NJ). Passive EAE was induced in non-tg and CRPtg littermate recipients by injecting them i.v. with 5 × 106 purified T cells.
Measurement of T cell proliferation and cytokine and chemokine production
All of these in vitro assays were performed using 96-well flat-bottom microtiter plates and a 200-μl reaction volume. Nylon wool-enriched T cells were isolated from non-tg mice undergoing active EAE as described in Induction of passive EAE, and 3 × 105 cells were cocultured (duplicate wells) for 72 h with 5 × 105 irradiated non-tg APCs, MOG peptide (5 μg/ml), and 0 to 100 μg/ml purified human CRP (Sigma, St. Louis, MO). The human CRP used for these studies was >99% pure as assessed by SDS-PAGE, it was recognized by the human CRP-specific mAb HD2–4 (20), it was endotoxin free as judged by Limulus amebocyte lysate assay, and in direct and competitive binding assays it bound to its natural ligand phosphorylcholine but not to MOG peptide (data not shown). Importantly, trypan blue dye exclusion tests confirmed that T cells cultured for 72 h in the presence of human CRP or in its absence had equal viability (data not shown). After addition of the stimulants, the cells were pulsed with 0.5 μCi [ 3H]thymidine/well, harvested and lysed 6 h later, and assessed for radioisotope incorporation (cpm). For comparison, the proliferation assay was also performed using naive T cells isolated from healthy non-tg mice. Proliferation of T cells is reported as the fold-increase in cpm relative to background incorporation by T cells from the same donors achieved in the presence of APCs but without addition of MOG peptide and human CRP (viz 2208 ± 742 cpm). The data are pooled from four separate experiments.
The in vitro cytokine and chemokine assays were performed essentially as described for the proliferation assay, except that both non-tg and CRPtg T cells were used. Duplicate cultures were either left untreated, stimulated with MOG peptide alone (5 μg/ml) or human CRP alone (50 μg/ml), or costimulated with MOG peptide plus human CRP. Culture supernatants were routinely collected at 6, 16, 24, and 48 h for use in cytokine and chemokine ELISAs, but only the 48-h data are shown. Cytokine and chemokine production by cultures of CRPtg and non-tg cells is reported as the percent of their respective maximal MOG peptide-induced response. The data are pooled from three separate experiments.
Acute phase protein, cytokine, and chemokine ELISAs
Human CRP was measured by ELISA as previously described (19), using sheep anti-CRP antiserum (ICN Pharmaceuticals, Costa Mesa, CA) as the capture Ab and mAb HD2–4 as the reporter. The assay does not detect mouse CRP and has a lower limit of detection of 20 ng human CRP/ml mouse serum. The mouse acute phase protein serum amyloid P component (SAP) also was measured by ELISA (19). The lower limit of detection of mouse SAP is 40 ng SAP/ml serum. ELISA kits for murine cytokines (TNF-α, IFN-γ, IL-4, IL-2, IL-12, IL-10, TGF-β) and chemokines (MIP-1α, RANTES) were purchased from R&D Systems (Minneapolis, MN), and an ELISA for the mouse chemokine MCP-1 was purchased from BioSource (Hopkinton, MA). Each assay was performed according to the manufacturer’s instructions.
CRPtg males and their non-tg littermates (n = 3 each) were sacrificed 4–9 wk after induction of active EAE, and their lumbothoracic spinal cords were removed and snap-frozen in liquid nitrogen. Immunohistochemistry was performed using the Vectastain ABC kit (Vector Laboratories, Burlingame, CA) on representative 10-μm-thick frozen transverse sections of tissue using mouse anti-CD11b mAb (BD PharMingen, San Diego, CA) to detect monocytes/macrophages and rat anti-mouse CD3 mAb (Serotec, Kidlington, U.K.) to detect T cells.
Differences in the day of onset of EAE (mean ± SD), maximum clinical scores (mean ± SD), [ 3H]thymidine incorporation during proliferation (cpm; mean ± SEM), and relative production of cytokines and chemokines (mean ± SEM) between CRPtg and control groups were evaluated with Student’s t tests. A value of p < 0.05 was considered significant.
Onset of active EAE was delayed by at least 1 wk in female CRPtg compared with age- and sex-matched wild-type (non-tg) littermates (Fig. 1 and Table I). However, once disease was established in CRPtg, its progression paralleled that in non-tg (Fig. 1), and its severity was not lessened (Table I). Human CRP was up-regulated in CRPtg during the inductive phase of disease (Fig. 2), achieving peak levels ∼10-fold above baseline. Surprisingly, elevation of human CRP was only transient. Importantly, the duration of the human CRP acute phase response approximated the delay in disease development between CRPtg and non-tg females (Fig. 1 and Table I), and the onset of active EAE in CRPtg females coincided with the recovery of approximately baseline expression of human CRP (compare Figs. 1 and 2). Like human CRP, the mouse SAP acute phase response was rapid, and peak levels of mouse SAP were ∼10-fold above initial values (Fig. 2). However, unlike human CRP, mouse SAP acute phase levels were maintained throughout the 30-day observation period (Fig. 2). Thus, despite the transient nature of the human CRP acute phase response, the prolonged elevation of mouse SAP suggests maintenance of an inflammatory state throughout the 30-day symptomatic phase.
|Group .||n .||Maximum Clinical Score .||Day of Onset .|
|Female non-tg||20||4.3 (1.1)||13.5 (2.2)|
|Female CRPtg||13||4.1 (0.9)||20.2 (3.1)|
|NSb||p < 0.0001b|
|Male non-tg||10||4.4 (1.0)||19.4 (4.6)|
|Male CRPtg||12||3.0 (2.4)||23.8 (4.9)|
|p = 0.05c||p < 0.05c|
|Group .||n .||Maximum Clinical Score .||Day of Onset .|
|Female non-tg||20||4.3 (1.1)||13.5 (2.2)|
|Female CRPtg||13||4.1 (0.9)||20.2 (3.1)|
|NSb||p < 0.0001b|
|Male non-tg||10||4.4 (1.0)||19.4 (4.6)|
|Male CRPtg||12||3.0 (2.4)||23.8 (4.9)|
|p = 0.05c||p < 0.05c|
Sex-matched CRPtg and non-tg were injected with MOG peptide, and the day of onset of EAE and its maximum severity, described in Materials and Methods, were noted. The scale used to score EAE symptoms is described in Materials and Methods. Mice were observed for 30 days. Results are mean (SD).
Values of p for Student’s t tests; CRPtg vs non-tg females.
Values of p for Student’s t test; CRPtg vs non-tg males.
To investigate further whether the duration of the human CRP acute phase response was related to the delay in disease onset, we induced active EAE in males. We showed previously that CRPtg males have significantly higher basal and acute phase expression of human CRP than do females (19). During active EAE in males, the kinetics of the human CRP response paralleled that in females, but as expected the human CRP levels in males were an order of magnitude higher (Fig. 3,A). In male CRPtg, onset of disease was also delayed compared with same-sex controls (Fig. 3,B and Table I). Unlike observations in females, once EAE was established in male CRPtg, average and maximum disease severity was significantly reduced compared with controls (Fig. 3,B and Table I, respectively). Although we did not exhaustively analyze the spinal cords of these mice, sections obtained from CRPtg males sacrificed 4–8 wk after induction of active EAE revealed little infiltration of the CNS by CD3+ T cells and CD11b+ monocytes/macrophages (data not shown). Furthermore, of 12 CRPtg males injected with MOG peptide, 3 remained asymptomatic up to 70 days afterward. Fig. 4 shows the results of experiments in which encephalitogenic T cells obtained from non-tg mice were injected into non-tg vs CRPtg recipients. Similar to the outcome of actively induced disease (Figs. 1 and 3), the onset of passively induced EAE was reproducibly delayed in CRPtg compared with non-tg recipients (Fig. 4).
The combined data from male and female CRPtg with active EAE strongly suggest that a cause-and-effect relationship exists between acute phase elevation of human CRP during the inductive phase of disease and its delayed onset, and the results from passive EAE confirm that the CRP-protective effect we observed is not due to an intrinsic T cell defect in CRPtg mice. How then can we explain the protective effect of CRP in EAE? One scenario is reduction of the encephalitogenic potential of T cells by circulating human CRP. We investigated this possibility in vitro and found that in the presence of human CRP, at concentrations equivalent to those seen in female and male CRPtg during the active EAE-induced acute-phase (i.e., 25 and 100 μg/ml, respectively; see Fig. 3,A), proliferation of encephalitogenic T cells elicited by stimulation with MOG peptide is in fact inhibited or completely abolished (Fig. 5). Because development of EAE is dependent on a number of cytokines and chemokines (21, 22), modification of their expression could also contribute to the beneficial effects of human CRP that we observed. To investigate this possibility, we tested whether or not human CRP alters the MOG peptide-induced production of encephalitogenic cytokines and chemokines by cultures of non-tg T cells. Human CRP significantly inhibited the production by MOG peptide specific T cells of the cytokines TNF-α, IFN-γ, and TGF-β (Fig. 6) and also the chemokines MIP-1α, RANTES, and MCP-1 (Fig. 7). In comparison, human CRP significantly enhanced IL-2 and IL-10 (Fig. 6) but had no effect on the production of IL-4 and IL-12.
Our data showing that human CRP protects CRPtg mice from MOG peptide-induced EAE demonstrate that human CRP can play a beneficial role in aseptic inflammatory disease. Furthermore, our observations of EAE in CRPtg mice indicate that human CRP might impact development of authentic MS and perhaps other T cell-mediated diseases. The pattern of human CRP acute phase expression that we observed during active EAE in CRPtg mice is similar to the cyclic rise and fall in blood CRP reported to occur in rabbits in the only other study of CRP in EAE of which we are aware (23). In that study, as in ours, CRP levels fluctuated before onset of active EAE (23). In our experiments, mouse SAP was fully elevated in CRPtg mice throughout the 30-day study period; therefore, we know that the inflammatory response per se was not transient. Thus, during EAE, the short-lived elevation of human CRP in CRPtg mice, and perhaps also endogenous CRP in rabbits, probably reflects disease-associated regulation of the CRP genes in question.
Despite the known contribution of complement in EAE (24) and the ability of human CRP to modulate complement activation (7), a fully functional complement system is not required for CRP-mediated protection in EAE. We know this because we crossed CRPtg mice with C3- and factor B-deficient mutants (24), thus producing CRPtg mice with impaired complement systems, and found that the protective effect of human CRP against MOG peptide-induced EAE was still fully expressed in the CRPtg/complement-deficient hybrids (data not shown). Furthermore, as seen in rabbits with slowly progressing EAE (23), we did not detect local deposition of human CRP in the CNS (Ref. 25 and data not shown). Thus, the in vivo data suggest that human CRP exerts its protective effect during the inductive phase of the disease and likely before destabilization of the blood-brain barrier. We speculate that the major mechanism by which human CRP promotes protection in EAE is via its direct inhibitory action on encephalitogenic cells. On one hand, human CRP protects CRPtg mice from active EAE by delaying or preventing infiltration of inflammatory cells into the CNS; on the other hand, it protects them from passive EAE by reducing the disease-inducing capacity of transferred T cells. The strength of the inhibitory effect in CRPtg mice depends intimately on the level of human CRP in circulation; as long as sufficiently high levels of the protein are maintained a detrimental outcome is averted. Consequently, for CRPtg females with low baseline expression of the transgene, the human CRP acute phase response is protective and onset of EAE is delayed until its resolution. CRPtg males that express a magnitude higher level of human CRP than females do exhibit delayed disease with reduced severity, and sometimes EAE is prevented altogether. Given that MS is at least 2-fold more common in women than in men (26), the fact that human CRP exerts a sex-dependent influence on the outcome of EAE in CRPtg mice may prove to be clinically relevant.
On the basis of our findings and in light of previously reported observations, we propose that the ability of human CRP to lessen production of the C-C chemokine MIP-1α is the crucial protective step in EAE. According to our model (Fig. 8), reduction of MIP-1α by acute phase human CRP is the initiating step of an inhibitory cascade that leads to reduced CNS infiltration and impaired T cell activation. MIP-1α has varied proinflammatory effects (27); thus, any reduction in its level would clearly benefit the host. Furthermore, MIP-1α is a major stimulus for production of proinflammatory cytokines including TNF-α (28), and thus reduction of MIP-1α also leads secondarily to reduced levels of additional potent proinflammatory molecules. Presumably, this weakens the Th1 response, which explains the lowered levels of the proinflammatory Th1 signature cytokine, IFN-γ. Expression of RANTES and MCP-1, both also proinflammatory C-C chemokines, is driven by TNF-α and IFN-γ (29). Thus, their levels also are reduced. Lowered amounts of these detrimental cytokines and chemokines, resulting either directly or indirectly from the actions of human CRP on encephalitogenic cells, culminates in the same protective effect, viz reduced T cell activation, chemotaxis, and parenchymal infiltration. Also it might be highly relevant that human CRP elevates IL-10 production. IL-10 further contributes to inhibition of TNF-α (30), and IL-10 is a Th2 signature cytokine in the mouse (31); by increasing IL-10 production human CRP might contribute to immune deviation from a Th1 phenotype to a less pathogenic Th2. Our model is in accordance with a variety of observations made about EAE using different cytokine- and chemokine-deficient mouse mutants. Like the expression of human CRP in CRPtg, lack of expression of TNF-α in TNF-α-deficient mice delays onset of active EAE and prevents inflammatory cell infiltration into the parenchyma but has no effect on disease progression (32, 33). Likewise, MCP-1-deficient mice resist EAE (34).
The effects of human CRP on cytokine and chemokine production by encephalitogenic cells that we report here are strikingly similar to earlier observations reported by others. For example, Gewurz’s group showed earlier that CRP decreased proliferation of Ag-stimulated lymphocytes and inhibited production of a lymphokine they referred to as “CTX” (a “lymphocyte-derived chemotactic factor for monocytes”) (35). It is tempting to speculate that CTX and MCP-1 are in fact the same or very similar proteins. More recently, it was established that human CRP alters chemokine-induced neutrophil chemotaxis (36) and the production of MCP-1 by endothelial cells (37). Human CRP reportedly has other anti-inflammatory effects on both neutrophils and leukocytes that might also participate in protection against EAE (36, 38, 39, 40, 41, 42, 43). Most importantly, all of the in vitro effects reported here by us and elsewhere by others consistently required concentrations of human CRP that mimic acute phase levels in patients.
Efforts are now under way to identify the exact encephalitogenic cell(s) with which human CRP presumably interacts to evoke protection in EAE, and the mode of this interaction. Human CRP is known to bind to Ag-induced T cells (44) and also to B cells (45), and it was recently established that human CRP binds with high affinity to the activation receptor FcγRI and the inhibitory receptor FcγRIIb in both mice and humans (8, 9, 10, 11). Notably, there is growing evidence that FcγRIIb plays a major inhibitory role in controlling the emergence of autoimmune disease (46, 47). Despite the fact that MOG peptide-induced EAE can develop fully in the absence of B cells (48), FcγRIIb-inhibitory signals induced by human CRP binding to APCs could possibly contribute to the protective anti-inflammatory response we observed in CRPtg mice.
Regardless of the mechanism of human CRP/encephalitogenic cell interaction and the identity of the effector cell(s) involved, it is clear from our results that human CRP can alter the disease-inducing potential of inflammatory cells contributing to the development of murine EAE, i.e. dendritic cells, monocytes, macrophages, and neutrophils. The relevance of this finding to the MS disease process in humans was made clear in a recent clinical study, wherein it was reported that of seven serum markers of inflammation examined in MS patients, only CRP levels correlated with development of cerebral atrophy and change in cerebral volume (49). Importantly, MS patients with lower CRP levels tended to develop more severe cerebral symptoms (49). Our ongoing animal studies seek to determine whether active EAE in non-tg mice can be blocked by administration of purified human CRP and whether treatment of female CRPtg with androgens (to increase expression of human CRP) induces the highly resistant phenotype we observed in CRPtg males. A positive outcome would raise the possibility that CRP might have therapeutic value in MS and that the gender-bias in multiple sclerosis in humans might be mediated, at least in part, by CRP.
A. J. Szalai thanks Mark McCrory and Ticiane Mello for their assistance.
This work was supported by National Institutes of Health Grant AI42183 (to A.J.S.), National Institutes of Health Grant NS29719 (to S.R.B.), National Multiple Sclerosis Society Grant RG-3216-A-5 (to S.R.B. and A.J.S.), and Advanced Postdoctoral Fellowship FA 1306-A (to S.N.) from the National Multiple Sclerosis Society.
Abbreviations used in this paper: CRP, C-reactive protein; EAE, experimental allergic encephalomyelitis; CRPtg, CRP transgenic; non-tg, nontransgenic; MS, multiple sclerosis; MOG, myelin oligodendrocyte protein; MIP-1α, macrophage-inflammatory protein-1α; MCP-1, monocyte chemoattractant protein 1; SAP, serum amyloid P component.