| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* MOE Key Laboratory of Arid and Grassland Ecology, Institute of Biophysics,
Instrumental Analysis Research Center, and
Institute of Cell Biology, Lanzhou University, Lanzhou, China; and
Immtech Pharmaceuticals Inc., Vernon Hills, Illinois, USA
1Correspondence: Life Science Bldg., Rm. 227, Lanzhou University, Lanzhou 730000, China. E-mail: wuy{at}lzu.edu.cn
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: inflammation • cardiovascular disease • protein-membrane interaction
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. CRP has two naturally occurring and conformationally distinct isoforms, i.e., pentameric CRP (pCRP) and monomeric CRP (mCRP) (3)
. pCRP consists of five identical subunits arranged as a cyclic pentamer. If the subunits in pCRP are separated into monomers, they undergo a conformational change that leads to formation of mCRP exhibiting very distinct bioactivities (3
4
5)
. mCRP is preferentially expressed in tissues (6
, 7)
, in contrast to pCRP, which is primarily found in plasma protein. The origin of mCRP in tissues and its relationship to pCRP remains unclear.
An important bioactivity of pCRP is to activate the classical complement pathway when bound to multivalent ligands (e.g., damaged cell membrane) and regulate the alternative complement pathway via recruitment of factor H (8)
. The complement activation/regulation activity of pCRP is proposed to be of potential significance to its in vivo function, which facilitates reactions of innate immunity and limits tissue damage in autoimmune reactions (1
, 9)
. As the prototypical acute-phase reactant, the plasma concentration of pCRP can increase from basal level of 1 µg/ml to more than 500 µg/ml within 24–48 h in response to inflammation or tissue injury . But since the biological function(s) of CRP have not yet been defined, it is regarded in clinical practice simply as a nonspecific marker of inflammation for long periods.
In the past decade, as more is learned about the magnitude of changes in pCRP plasma levels in the defined populations and in its bioactivities, the notion has evolved that pCRP is not only an independent predictor of cardiovascular disease (CVD) risk, but also a direct participator in the underlying process of CVD (1
2
3
, 10
, 11)
. However, it is still a topic of debate whether pCRP is a real participant or simply a nonspecific inflammatory marker (12)
. One major criticism is how a plasma protein with a dynamic range of 1000-fold increases within hours could function like a fine modulator of sophisticated cellular or physiological systems (1)
. In addition, the stimulatory effect of pCRP on cells such as endothelial cell (EC) usually require long incubation (>6–12 h) and relatively high concentrations (>25 µg/ml). These concentrations are much higher than the 10 µg/ml cutoff recommended for CVD risk evaluation (13)
and thus have less relevance to the subclinical CVD risk conditions. The above enigmas, however, may be better explained by the controlled transition/interplay and the differential contribution of distinct CRP isoforms.
In careful studies directly comparing pCRP and mCRP, mCRP is more active in exerting biological effects. For example, we have recently showed that mCRP is the primary CRP isoform that binds native and modified low-density lipoproteins (14)
and can regulate complement activation in a more effective and flexible fashion (15)
. Likewise, mCRP is a significantly more potent activator of EC (16)
and neutrophils (17)
(e.g., it can stimulate EC at concentrations as low as 1 µg/ml following 4 h incubation) than pCRP, which suggests that a structural rearrangement leading to formation of mCRP or mCRP-like isomer is a prerequisite for "pCRP" to exert stimulatory effects on EC and inflammatory cells. However, pCRP is very stable under physiological conditions and is not expected to lose the pentameric structure without denaturation (1)
.
Since membrane is a known location for pCRP function and since many proteins undergo conformational/functional alterations on interaction with membrane, we investigated the possible influence of membrane on pCRP structure and function in the present study and provided evidence that membrane could be a physiological regulator for CRP structure and bioactivity.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. All pCRP preparations were stored in TBS (10 mM Tris, 140 mM NaCl, pH 7.4) containing 2 mM CaCl2 (TBSCa) to prevent spontaneous dissociation (19
, 20)
. The functional integrity of pCRP was verified by calcium-dependent binding to PC-BSA (2 mol PC/mol BSA, Biosearch Technologies, Novato, CA, USA). Protein batches were dialyzed to remove NaN3 and assayed for endotoxin contaminant by Limulus assay (Sigma, St. Louis, MO, USA) before use. Additional purification step through Detoxi-Gel Columns (Pierce, Rockford, IL, USA) was performed to remove endotoxin when necessary. The final endotoxin level of all protein solutions was below 0.06 EU/ml. Standard curve of pCRP was constructed using PC-BSA captured pCRP, and standard curve of CRP subunit was constructed by directly immobilized mCRP.
Mouse anti-human mCRP monoclonal antibody (mAb) (3H12, 6B7, 9C9, 8C10) and pCRP mAb (2C10, 1D6, 8D8) was generated as described (18)
. Sheep anti-human CRP polyclonal antibody (pAb) was purchased from the Binding Site. mAb IgG was purified by immobilized protein A column (GE Healthcare Life Sciences, Little Chalfont, Buckinghamshire, UK). Fab'2 fragment of mAb IgG was prepared using immobilized pepsin (Pierce) according to the manufacturer’s instruction. FITC and Rhodamine B (RHB, Sigma) labeling of pCRP and mAb IgG was performed according to standard procedure.
Liposome preparation
Lipids dissolved in chloroform/methanol (3/1, v/v) were dried under N2, then evaporated by vacuum and finally hydrated with TBS for 2 h with occasional shaking in a water bath at 60°C. The lipid mixture was extruded through two pieces of polycarbonate membrane (0.1 µM hole size, Whatman, Clifton, NJ, USA) with a Mini-Extruder (Avanti, Polar Lipids, Inc., Alabaster, AL) to prepare large unilaminar liposomes according to the manufacturer’s instruction. Liposomes were stored at 4°C and used within one week.
Fluorescence spectroscopy
The intrinsic tryptophane fluorescence of pCRP was measured using a LS-55 fluorometer (PerkinElmer, Wellesley, MA, USA) to monitor the change in its tertiary structure. pCRP (5 µg/ml) was incubated either alone or with 200 µg/ml eggPC/lysoPC (4/1, mol ratio) (lipid/protein
5000/1, mol ratio) in TBSCa. Samples were excited at 295 nm, and emitted fluorescence signals were collected at 345 nm every 10 min. The excitation and emission slits were set to be 8 and 5 nm, respectively. The measured fluorescence was expressed as "Ratio of residue intrinsic fluorescence": (fluorescence intensity at time T)/(fluorescence intensity at time zero).
pCRP binding and dissociation
Neoepitope expression of pCRP on membrane was examined by ELISA on fused eggPC/lysoPC (4/1, mol ratio) liposomes. For efficient liposome fusion, 20 µg/ml liposome was incubated in microtiter wells at 60°C for 4 h. Other ligands such as PC-BSA, polylysine, or Pneumococcal C-polysaccharide (PnC; 20 µg/ml) were immobilized at 37°C for 4 h. After blocking with 1% BSA overnight at 4°C, pCRP was added in TBSCa or TBSEDTA (TBS containing 5 mM EDTA) with 1% BSA for indicated time at 37°C. All the following operations were performed in the presence of 2 mM calcium. Washing buffer was TBSCa plus 0.02% Nonidet P-40. For assays on fused liposome, washing was performed without detergent (0.02% Nonidet P-40) to minimize the disturbance to membrane. pCRP binding was reported by mAb 2C10 (1:100), and neoepitope expression was probed by mAb 3H12 (1:100). HRP-conjugated anti-mouse antibodies (Sigma, 1:20,000) were then added and wells were developed by 0.4 mg/ml OPD. Optical density (OD)492nm was measured after addition of 1M H2SO4. Experiments substituted pCRP with BSA was served as background control (optical density492nm <0.05).
Electron microscopy (EM)
EM visualization of lipid monolayer was performed as described (21
, 22)
. Briefly, droplets (
75 µl) of TBSCa buffer were first placed in TeflonTM wells and the surface was coated with 0.5 µl eggPC/lysoPC (4/1, mol ratio) dissolved in chloroform/methanol (3/1, v/v). The organic solvent was allowed to evaporate for 30 min at room temperature followed by injection of 0.1 mg/ml pCRP (final concentration) from the side hole of the TeflonTM well. Samples were incubated in a sealed humidified chamber for indicated time at 37°C. Then the lipid monolayer was transferred to grids coated with aged (hydrophobic) carbon film. After washing with incubation buffer, the grids were blotted up and negatively stained with uranyl acetate solution (1%, w/v) for 1 min. The grids were observed in a Tecnai G2 200 kV EM.
Complement deposition
After pCRP binding to fused liposome or immobilized PC-BSA, microtiter wells were blocked with 1% BSA and assayed for complement deposition. For C1q binding, C1q was added and the binding was detected with goat pAb against human C1q (10 µg/ml) (Calbiochem) and HRP-conjugated rabbit anti-goat antibody (Ab; 10 µg/ml) (Sigma). For C3d deposition, 1% normal human serum (NHS) diluted in VBSCa+Mg veranol buffered saline (VBS, pH 7.4) containing 0.15 mM calcium and 0.5 mM magnesium), VBSMg+EGTA (VBS containing 0.5 mM magnesium and 5 mM EGTA) or VBSEDTA (VBS containing 5 mM EDTA) with 1% BSA were added at 37°C for 30 min. C3d deposition was detected by sheep anti-human C3d pAb (1 µg/ml) (Binding Site) and HRP-conjugated donkey anti-sheep Ab (1:5000) (Sigma). The measured C1q or C3d deposition was normalized to the bound pCRP obtained in parallel experiments.
Cellular binding assay
Jurkat T and U937 cells were cultured as described (14
, 23)
. Apoptosis of Jurkat (23)
and U937 was induced by incubation with 0.5 µg/ml staurosporine (Sigma) for 4 h and 0.1 µg/ml staurosporine for 1–2 h, respectively. More than 80% cells underwent apoptosis as accessed by PE-Annexin V and PI staining. More than 95% FITC-pCRP positive Jurkat cells was also PE-Annexin V positive. In some experiments, cells were pretreated with trypsin (0.5 mg/ml, 30 min, 37°C), Type I (snake venom) sPLA2 or PLD (1U, 37°C, 30 min) (Sigma). For treatment with iPLA2 inhibitor, live cells were induced apoptosis in staurosporine together with bromoeonol lactone (BEL, 10 µM; Sigma). Apoptotic cells (106) were incubated with 20 µg/ml pCRP in medium with 1 mg/ml BSA (100 µl reaction vol) for 30 min at room temperature (23)
. pCRP binding and dissociation were probed by 2C10 or 3H12 under 4°C, respectively, followed by addition of FITC-anti-mouse Ab (Sigma, 1:1000). Relative fluorescence intensity was measured after lysis and normalized according to the cell protein (protein amount of 106 cell was designated as 1). Cell protein was determined by bicinchoninic acid (BCA) kit (Pierce).
Endothelial stimulation
Human artery endothelial cells (HAECs, passage 3, from Cascade Biologics, Portland, OR, USA) were cultured in Medium 200 with low-serum growth supplements (Cascade Biologics). Confluent HAEC (passages 4–6) monolayer in 24-well were challenged with CRP isoforms. At the indicated time, culture supernatants were collected, and MCP-1 and interleukin (IL)-8 concentrations were determined in duplicate by commercial ELISA kits (BD Pharmingen).
Fluorescence microscopy
For visualization of pCRP dissociation on apoptotic cells, FITC-labeled pCRP was allowed for binding as described followed by staining with RHB- 3H12. In some experiments, after incubation with pCRP, 5% NHS was added to CRP-coated cells for another 30 min at 37°C (23)
. pCRP dissociation and C3 deposition were probed by RHB-3H12 or sheep anti-human C3d pAb under 4°C. Stained samples were examined by a Leica DM5000B fluorescence microscopy with a 100x oil objective. The images were captured by a Leica DC300F CCD.
Statistical analysis
All experiments were performed in triplicate and repeated for 3–5 times. Data analysis was performed by Student’s t test or one-way ANOVA followed by the Bonferroni multiple range tests. A value of P < 0.05 was considered significant. Data are presented as mean ± SEM.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Coincubation of pCRP with liposome did not alter the secondary structure of pCRP as detected by Circular Dichroism and Fourier Transform Infrared Spectroscopy (not shown). Intrinsic tryptophane fluorescence was used to monitor the tertiary structure of pCRP. Addition of liposome resulted in mild but evident decrease in pCRP’s tryptophane fluorescence, indicating alteration of the tertiary structure of pCRP (Fig. 1
A). In contrast, pCRP alone in TBSCa exhibited nearly unvaried time profile over the same time period. As positive controls, when pCRP was dissociated by 8M Urea to produce mCRP (6)
, a sharp decrease in tryptophane fluorescence was noted (Fig. 1A
).
|
Whether membrane could affect the quarternary or oligomeric structure of pCRP was next examined. Because pCRP dissociation is associated with exposure of neoepitopes, we used a specific mAb (3H12) against aa.199–206 (25)
, a predominant neoepitope exposed in mCRP only (18)
, to probe changes in the oligomeric structure of pCRP. A neoepitope expression ratio (NER, see Fig. 1
) was defined to evaluate the extent of dissociation. pCRP calcium-dependently bound to liposome fused onto microtiter wells (Fig. 1B
). Over 5% of bound pCRP expressed neoepitope on membrane within 30 min (Fig. 1C
), and a maximum of NER of 30–40% was observed after 1–2 h (Fig. 1C and D
), suggesting that pCRP dissociated rapidly on binding to membrane. Parallel size-exclusion chromatography (SEC) and dot-blot controls showed that pCRP alone in calcium-containing buffer did not dissociate to expose neoepitope. Additional experiments indicated that membrane binding is the prerequisite for pCRP dissociation (Supplemental Fig. 1
).
The specificity of different ligands in inducing pCRP dissociation was identified further (Fig. 1E
). Except for the exclusive native antigenicity shown by pCRP attached to PC-BSA, neoepitope expression could be observed after pCRP bound to fused liposome, PnC, and polylysine, in order of decreased efficiency. These results suggest that a hydrophobic microenvironment, ligand mobility (fluidity), and multipoint attachment are necessary elements for efficient dissociation of pCRP. Notably, the combination of PC-BSA and fused liposome provides an ideal model for comparing the possible functional consequence of pCRP dissociation.
EM observation provided further support for pCRP dissociation on membrane (Fig. 2
). In the presence of calcium, a large number of pCRP molecules attached to lysoPC containing monolayer within 30 min and primarily exhibited their cyclic, pentameric structure (Fig. 2A
). However, loss in cyclic appearance became evident after 2 h (Fig. 2B
). Prolonged incubation for 24–48 h resulted in predominant distribution of separated subunits (Fig. 2C
) and sometimes formation of CRP subunit arrays. These results confirmed the previous findings that the two-dimensional crystals formed by pCRP on monolayer were composed of separated subunits but not annular pentamers (26)
. Figure 2D
quantitively summarized the time course of pCRP dissociation on monolayer based on EM observation.
|
To avoid batch-to-batch variations, two purchased purified or recombinant pCRP reagents were tested in aforementioned experiments and similar results were obtained. The dissociation on membrane was not due to lipid oxidation, since inclusion of antioxidants (1 mM ascorbic acid) did not affect the results. Taken together, although calcium is sufficient to keep the pentameric structure of pCRP in fluid phase (6
, 27)
, membrane attachment can overcome the stabilization effect of calcium and drive a rapid dissociation of pCRP.
CRP subunits dissociated on membrane partly retain pCRP’s conformation
Membrane attached pCRP continuously dissociated into subunits (Fig. 2)
; however, the maximal NER was nearly invariant with the extension of incubation (Fig. 1C, D
). This finding led us to hypothesize that the subunits converted on membrane may partly retain pCRP’s native conformation and/or antigenicity. In support of this view, after being bound to fused liposome, both membrane-associated native and neoantigen were sensitive to EDTA and haptenic PC competition (Fig. 3
A). In addition, despite the accumulation of separated CRP subunits on membrane during prolonged incubation, the sensitivity of these antigens to EDTA or PC was not altered. EM observation further confirmed that EDTA wash almost eliminated the presence of both pentameric and monomeric CRP on the monolayer (not shown). Therefore, the neoantigen also uses an EDTA- and PC-sensitive mechanism to associate with membrane, indicating that this CRP isoform is a hybrid that partly preserves pCRP’s native conformation, such as determinants for mAb recognition (2C10) and calcium-dependent PC binding.
|
The above data strongly suggest that CRP subunits converted on membrane (denoted as mCRPm) represent a stable conformation that is distinguishable from mCRPs, the long-recognized mCRP that converted in solution, which interact with membrane through hydrophobic insertion (14)
and is insensitive to EDTA and PC treatments (Fig. 3A
). Thus, on one hand, membrane can facilitate pCRP dissociation; while on the other hand, the membrane helps the dissociated CRP subunits to maintain their native conformation when the subunit is in the pentameric structure, possibly due to the combined effects of calcium stabilization and ligand binding. Indeed, pCRP underwent just a moderate conformational change on membrane as compared with the drastic decrease in urea denaturation for fluid phase mCRP preparation (Fig. 1A
). These results also support the deduction that membrane-converted CRP subunits do not change much in structure from their pentameric counterpart.
Formation of mCRPm facilitates complement activation
Complement activation is a major activity of pCRP that link this protein with inflammation and atherosclerosis. However, why multivalent ligand binding is required for pCRP to activate complement remains to be elucidated. According to aforementioned results, pCRP dissociation on membrane to form mCRPm may reflect a mechanism for efficient complement activation by overcoming certain negative factors (see Discussion). We therefore examined the possible correlation between pCRP dissociation and complement activation by comparing C1q binding and C3d deposition on pCRP attached to PC-BSA (without contribution from mCRPm; see Fig. 1E
) or liposome (inducing mCRPm formation; see Fig. 1E
). Significantly more C1q bound to liposome-pCRP complex compared with PC-BSA-pCRP complex, especially at low pCRP concentrations (Fig. 3B
). Accordingly, liposome-pCRP complex was obviously more potent than PC-BSA-pCRP complex in C3d deposition (3 fold, P<0.05) through the classical pathway (Fig. 3C
). These data indicate that formation of mCRPm conformer is associated with significantly increased efficiency in complement activation. A Fab’2 fragment of 3H12 did not exhibit significant competition for C1q binding to mCRPm, suggesting that the neoepitope (i.e., aa.199–206) does not directly contribute to the enhanced CRP-C1q interaction. Rather, the structural changes associated with mCRPm attenuate the steric hindrance and overcome the relative size limitation in pCRP for C1q recognition and activation (28)
(see Discussion).
Endothelial stimulation by mCRPs derived from detached mCRPm
Further analysis revealed that mCRPm could detach from membrane (Fig. 4
A). After incubation of fused liposome with attached pCRP for 12–24 h in TBSCa, the supernatant was collected, concentrated, and subjected to SEC. The major peak eluted at 15.7 ml corresponding to MW 23,000 was purified, concentrated, and assayed in capturing ELISA and dot-blot (18)
. In contrast to the dual antigenicity of mCRPm, this component exhibited exclusive mCRP antigenicity, indicating the total expression of the mCRPs conformation. Other components in the elution profile between pentameric and monomeric peak should represent a spectrum of dissociation/reassociation intermediates. Similar results were obtained when using supernatant separated from fluid phase liposome and pCRP interaction design (as in Fig. 1A
). The purified MW 23,000 component (i.e., detached mCRPm) exhibited potent stimulatory effect on human artery endothelial cells (HAEC) (Fig. 4B
), similar to that reported for mCRPs (16)
. After 4 h challenge, comparable release of IL-8 and MCP-1 was induced by the purified component and mCRPs at 1–5 µg/ml, whereas no change was noted by treatment with 20 µg/ml pCRP.
|
pCRP dissociation on cell surface
Since the binding of pCRP to the blebs of apoptotic cells is thought to be mediated by lysoPC (23)
, we investigated whether binding of pCRP to apoptotic cells membrane also induced pCRP structural changes and whether such changes correlated with lysoPC generation. pCRP primarily bound to apoptotic Jurkat T cells (Fc
R negative) (Fig. 5
A) and >95% FITC-pCRP positive cells was also PE-Annexin V positive as accessed by flow cytometry. After incubation of pCRP with apoptotic cells for 30 min, appreciable expression of 3H12 reactivity was noted (Fig. 5A
), indicating pCRP dissociation did occur. Further evidence was provided by the colocalization of immunofluorescence of pCRP and CRP subunit on cell surface (Fig. 5B
). In addition, treatment with exogenous sPLA2 (catalyzes the generation of lyso-phospholipids) significantly promoted, while PLD (cleaves the head groups from phospholipids that destroys ligand for pCRP binding) markedly decreased pCRP binding and neoepitope expression (P<0.01 vs. control) (Fig. 5C
). Furthermore, inhibition of endogenous PLA2 activity by BEL attenuated both native and neoantigen on apoptotic cell surfaces (P<0.01 vs. control). These results indicate that lysoPC containing plasma membrane domain is the major location for pCRP binding on apoptotic cell surface, and such interaction results in the partial structural change of pCRP into mCRPm. Further experiments were conducted with the presence of complement. Consistent with the previous findings (23)
, complement deposition was obviously enhanced on mCRPm-coated apoptotic cells. However, we failed to demonstrate the colocalization of the neoantigen with C3d, possibly caused by the steric hindrance introduced by the sequential deposition of complement components.
|
Since pCRP can also bind live cells via FcRs (CD32 and CD64) (29
30
31
32)
, we examined whether pCRP structural change to mCRPm could occur on HAECs that expresses these receptors. mCRPm was consistently detected on binding of pCRP to HAECs, albeit higher pCRP concentration (>50 µg/ml) was required and the signal was less evident than that detected on apoptotic cell surface (Fig. 5D
). The structural change was not likely due to the calcium-independent pCRP-FcRs interaction, because the effects were determined to be calcium-dependent (Fig. 5D
) and could not be blocked with antibodies against CD32 and CD64 (not shown). These observations suggest membrane, possibly lipid raft (see Discussion), are involved in the binding of pCRP to cell membranes. Indeed, both the previous findings (2
, 28)
and the present results suggest the conformational rearrangement is mediated by multipoint ligand-binding to the recognition face of pCRP, which opposes to the effector face that mediates pCRP binding to FcRs (31)
.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Membrane enriched in PC ligands is widely expressed throughout the body and thus appears to be an important site for pCRP function. In the current study, we showed that membrane binding could induce sequential changes in pCRP structure. On membrane binding, pCRP first dissociates into subunits with partly retained native conformation, i.e., mCRPm (Figs. 1
, 2
, 3A)
. Because the affinity of pCRP with PC is relative low [Kd40 µM (34)
], and because the multipoint attachment to membrane is lost during dissociation, mCRPm can detach from membrane to form the widely recognized mCRPs (Fig. 4)
. Such a stepped structural transition (pCRP
mCRPmmCRPs) provides unique advantage for regulating CRP function (Fig. 6
and discussed below).
|
Structural transition of pCRP on membrane
The expression of neoepitope in pCRP on membrane is not necessarily associated with concomitant physical separation of subunits. The dissociation process may consist of following steps: i) pCRP binds to membrane; ii) pCRP undergoes conformational changes on membrane, such as slight but appreciable tertiary structure alteration and neoepitope expression (Fig. 1A, C
), which may begin from certain lysoPC-bound subunits and then propagate to adjacent subunits within the parent cyclic pentamer; iii) the aforementioned changes further result in relaxation of pCRP oligomeric structure, which may not be readily distinguishable under negatively stained EM; iv) physical separation of subunits due to the lateral movement (or fluidity) of the subunit bound ligands (e.g., lysoPC). It is likely that the mobility of ligand can also provide additional energy that contributes to weaken the subunits association. Of note, pCRP subunit contains six tryptophanes, one of which is located at amino acid position 205. Hence, the noted partial change in intrinsic tryptophan fluorescence of pCRP on membrane binding (Fig. 1A
) may reflect the movement of Trp205 within the key neoepitope aa.199–206 out of the hydrophobic intersubunit contact region (25)
to the protein interface where the sequence can be recognized by specific mAb (3H12).
Beside membrane, other ligands such as PnC and polylysine can also induce pCRP dissociation (Fig. 1E
). In contrast, although comparable pCRP binding can be observed on immobilized PC-BSA, essentially no dissociation occurred (Fig. 1E
). These results implicate the importance of multipoint attachment for pCRP dissociation: fused membrane, immobilized PnC, and polylysine all presumably provide more closely arrayed ligand groups compared with PC-BSA, and such multipoint attachment may facilitate the conformational change in pCRP (28)
. In addition to multipoint attachment, a hydrophobic microenvironment exists as another key determinant. As reported before (6)
, strong hydrophobic interaction, such as immobilization onto plastic surface (e.g., microtiter well), initiates a prompt and thorough dissociation of pCRP with total loss of native antigenicity. A mild hydrophobic surface, e.g., membrane, however, leads to a rapid dissociation of pCRP subunits with partly retained native conformation (Fig. 3A
). Thus, together with the lateral fluidity of PC ligands in membrane, these factors can account for the different efficiency of various substrates in inducing pCRP dissociation.
In addition to lysoPC in apoptotic cell membranes, lipid raft, a specialized membrane domain with high cholesterol and sphingomyelin contents, may also provide the context for pCRP binding and dissociation. Cholesterol (35)
and the PC headgroup of sphingomyelin both are pCRP ligands. The interdigitating assembly pattern of cholesterol and sphingomyelin in raft (36)
may resemble the conditions of disturbed membrane (2
, 24)
that expose sites for pCRP binding. Furthermore, by using a raft composition liposome, comparable results as shown above could be obtained (unpublished findings). Indeed, HAEC is enriched of lipid raft and anti-FcR inhibition studies suggested that there were FcRs-independent binding sites on HAEC (32)
. Because the affinity of pCRP with FcRs [Kd88.2 nM (32)
] is much higher (450-fold) than the affinity of pCRP with PC [Kd40 µM (34)
], it may explain why higher concentration is required for detection of pCRP dissociation on HAEC.
Membrane regulates pCRP function
According to the crystal structure and mutational analysis, Agrawal et al. (28)
proposed that: i) multipoint ligand binding would result in conformational change in the CRP subunits that is necessary for exposure of the C1q binding site to the surface of the molecule; ii) the binding of C1q to one CRP subunit was hindered by Lys (114) in the neighboring subunit; iii) although pCRP possesses 5 C1q binding sites in its pentameric structure, one pCRP molecule could bind only one C1q subunit because of the relative size limitation, and thus closing-packing of pCRP is required for C1q activation. However, while pCRP is linked as a marker for inflammation, significant accumulation of pCRP in inflammatory or injured tissues was not noted (37)
. Given the trace level of pCRP under normal conditions, the large number of apoptotic or injured cells and the huge volume of single-cell or pathogen (µm scale) compared with the tiny pCRP molecule (10 nm diameter), it is unlikely that the efficient complement-dependent clearance of these self-materials (23
, 38)
can be achieved simply by close-packing of pCRP on surface.
Based on the current results, we proposed that the close-packed pCRP and the membrane-converted CRP hybrid conformer (i.e., mCRPm) can act in concert to promote the complement-dependent clearance of self-materials and pathogens. When bound to the plasma membrane of apoptotic or injured cell, pCRP undergoes a partial structural rearrangement to form mCRPm (Figs. 1
, 2
, 3A)
. On one hand, the conformational change can transform pCRP into a priming state that is more potent for C1q recognition, which may involve the exposure of C1q binding site to the surface (2
, 28)
. On the other hand, pCRP dissociation can i) attenuate the steric hindrance from Lys (114) of the neighboring subunit, and ii) overrun the aforementioned size limitation for pCRP-C1q interaction and activation. Thus, it is possible that even the dissociation of a single pCRP molecule can provide multipoint binding site for hexameric C1q to initiate classical complement pathway. Taken together, the formation of mCRPm may represent a rapid and reliable amplification mechanism for pCRP function such as complement activation.
Furthermore, mCRPm can detach from membrane to form mCRPs which has distinct bioactivities (14
15
16
17
, 39)
that may contribute to the local inflammatory process. This process should represent the secondary regulation for CRP function. Although the expression of characteristics of mCRPm is rapid (within 30 min, Fig. 1A and C
), a longer duration is necessary for the occurrence of separated subunits (over 2 h, Fig. 2
). Furthermore, when considering the relative portion of bound pCRP (usually<15% of the added FITC labeled pCRP) and the dynamic equilibrium of detachment, it will not be surprising that the accumulation of mCRPs to an effective level is a relatively slow and rate-limiting process (>12 h incubation is necessary for detectable mCRPs in our setup). This may readily explain the observed delay of pCRP in inducing several cellular responses and the requirement of high concentration to exert biological effects (16
, 17)
.
Our findings imply that the rapid pCRP mCRPm transition represents a mechanism that could contribute to "acute phase" amplification of the inflammatory response. For example, the formation of mCRPm can facilitate clearance of apoptotic lymphocyte (23)
and necrotic cell debris (38)
, enhance uptake of atherogenic LDL derivatives (40)
and apoptotic macrophages induced by cholesterol-loading in atheroma (41)
, and promote complement-mediated injury during heart attack (33)
. On the other hand, effective mCRPm mCRPs transition should occur mainly in loci with chronic or persistent inflammation (e.g., atherosclerotic lesions), where local pCRP production is significantly up-regulated (42)
and pCRP ligands (e.g., altered cell membrane, modified LDL) are abundant. mCRPs would then effectively trigger cellular responses, regulate complement, and facilitate LDL metabolism at these sites. Of note, the abundance of available pCRP ligands, the pCRP mCRPm mCRPs transformation dynamics, and the heightened proteolytic sensitivity of mCRP compared to pCRP, can serve as a buffering mechanism to enable this protein to finely regulate cell and biochemical processes and avoid unwanted violent stimulation, which is especially important for an acute-phase protein like pCRP.
Based on the current knowledge, we hypothesize that the very stable pCRP isoform represents a rest or quiescent state of CRP, while the monomeric CRP isoforms (mCRPm and mCRPs) represent proinflammatory functional forms of CRP. The apparent distinct bioactivities of the CRP isoforms may unify through finely controlled switches in order to accommodate diverse demands under various pathophysiological circumstances. Thus the understanding of these switches would be important to ascertain out actual biological role of this ancient molecule and the intrinsic merit underlying its survival during evolution. Herein, we described a possibly universal switch (i.e., membrane) that regulates the structure and function of CRP. Also, the local synthesis of mCRPs by cells such as stimulated monocyte (43)
is likely to be controlled by a molecular switch at the stage of posttranslational modification, which may involve processing of the leading peptide in CRP protomer (44)
. Clearly, additional investigations are needed to further elucidate these switches and the spatial and temporal contributions of these different CRP isoforms to any functional system being studied.
|
ACKNOWLEDGMENTS |
|---|
Received for publication June 19, 2006. Accepted for publication July 24, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS |
), TBSCa+eggPC/lysoPC (4/1, mol ratio) liposome (•), TBSEDTA+8M Urea (
). B) Binding of pCRP to fused liposome, immobilized PC-BSA, polylysine, or BSA in the presence of 2 mM calcium (white bar) or 5 mM EDTA (black bar). pCRP (20 µg/ml) in 1% BSA was incubated with immobilized ligands for 1 h at 37°C. pCRP binding was reported by mAb 2C10. Polyclonal CRP antibodies and biotinylated pCRP-avidin detection was also used with similar results. pCRP calcium-dependently bound to fused liposome, or to immobilized PC-BSA, indicating that this setup is feasible for investigating pCRP-membrane interactions. C) Neoepitope expression of pCRP on fused liposome. 1 (
