Ultrasensitive Confocal Fluorescence Microscopy of C-Reactive Protein Interacting With FcγRIIa
Background— C-Reactive protein (CRP) is an acute phase protein with a suggested pathogenic role in cardiovascular disease. Previous reports proposed that the low-affinity IgG receptor FcγRIIa is the major receptor for CRP. However, these reports were met with criticism because the use of anti-CRP antibodies in the detection of CRP binding to FcγRIIa may have caused false-positive results.
Methods and Results— To resolve this controversy, we used ultrasensitive fluorescence microscopy to study the association, dissociation, and equilibrium of CRP binding to FcγRIIa. CRP indeed binds to FcγRIIa, with low association rates and dissociation rates. Anti-CRP antibodies markedly enhance binding, as is evident from the decrease of the equilibrium dissociation coefficient by 2 orders of magnitude.
Conclusions— Our study demonstrates the virtues of single fluorophore labeling and highlights the pitfalls of immunolabeling in investigating CRP/Fc receptor interactions. Importantly, this article provides the first quantitative characterization of CRP binding to FcγRIIa and explains and reconciles the diverse and conflicting data presented in the literature.
C-Reactive protein (CRP) is the prototype human acute-phase protein.1,2 It also is a powerful cardiovascular risk marker.3 Recent investigations suggested a pathogenic role of CRP in cardiovascular disease,4,5 which has spawned widespread interest in studies of its biological function.6–12 Importantly, it has been demonstrated that human CRP transgene expression causes accelerated aortic atherosclerosis in apolipoprotein E–deficient mice, providing first in vivo evidence of a direct involvement of CRP in atherogenesis.12
To date, ligand binding, opsonization of bioparticles,13–15 and complement activation16 are rigorously defined pathobiological CRP functions. CRP interactions with nucleated cells have gained increasing interest,6–11 and CRP binding to cellular receptors has been intensely investigated with conflicting results. Whereas some reports provided evidence of specific CRP receptors,17 other experiments demonstrated interaction with Fc receptors.18,19 The low-affinity IgG receptor FcγRIIa was proposed to be the major CRP receptor.18,19 Several observations supported this concept. When coincubated with low-density lipoprotein (LDL), CRP colocalizes with clusters of FcγRIIa on monocyte membranes.8 Furthermore, CRP was reported to induce FcγRIIa-signaling in human promyelocytic cell line HL-60,20 and finally, experiments in FcγRII- and γ-chain–deficient mice showed lacking CRP-mediated biological responses compared with wild-type mice.21 To demonstrate CRP binding to FcγRIIa, anti-CRP antibodies were used in the initial reports18,19 because direct labeling of CRP with fluorescein isothiocyanate (FITC) or 125-I may damage the structure of the molecule and lead to ambiguous results. It was also suggested that CRP binding to FcγRIIa is allele specific.19 High-affinity binding was reported for FcγRIIa R/R-131, intermediate affinity binding for FcγRIIa R/H-131, and low-affinity binding for FcγRIIa H/H-131. Subsequently, several authors proposed that CRP may not interact with FcγRIIa at all and that the observed binding of CRP to FcγRIIa results from an interaction of the Fc portion of the anti-CRP antibody with FcγRIIa itself.2,22,23 Indeed, using F(ab′)2 fragments of anti-CRP antibodies, fluorescence-activated cell sorter (FACS) analysis revealed no CRP binding to FcγRIIa-R131 on polymorphonuclear leukocytes and FcγRIIa-transfected IIA.6 cells.22 Other authors have suggested that the observed binding of CRP to FcγRIIa might be attributable to IgG contamination of the CRP reagent,23 and a recent review claims that CRP does not interact at all with cellular Fc receptors.2
Here we applied the novel technology of ultrasensitive confocal fluorescence microscopy to study CRP interactions with FcγRIIa.24–26 Our results visually demonstrate and quantitatively show that (1) use of anti-CRP antibodies indeed affects CRP binding and leads to false-positive results; and (2) CRP, however, does bind to FcγRIIa, although with lower affinity than anti-CRP antibody/CRP complexes.
COS-7 cells were obtained from DSMZ-German Collection of Microorganisms and Cell Cultures (Braunschweig, Germany) and maintained in DMEM/10% FCS with 1% penicillin/streptomycin/l-glutamine. This cell line does not express Fc receptors.
Reagents and Antibodies
Partially purified CRP was obtained from Sigma. Highly purified CRP was kindly provided by Dr T.W. Du Clos (University of New Mexico, Albuquerque). Recombinant CRP (rCRP) was obtained from Calbiochem. Western blot analysis revealed <0.1% IgG for partially purified and no detectable IgG for highly purified and rCRP. Monoclonal anti-CRP antibody, clone 2C10, was generously provided by Dr Du Clos with kind permission of Dr Larry Potempa (ImmTech, Evanston, Il).27 Anti-CD32–FITC, clone FLI8.26(2003), and unlabeled and phycoerythrin (PE)-labeled monoclonal mouse IgG1 isotype were purchased from BD Biosciences. Anti-CD32, clone KB61, was purchased from DAKO, affinity-isolated F(ab′)2 PE-goat anti-mouse (GAM) from Caltag Labs, and human serum from the blood transfusion service of the University of Ulm.
FcγRIIa Cloning and Transfection
Human FcγRIIa cDNA (G/G genotype, coding for FcγRIIa R/R-131) was generated by RT-PCR and cloned into pcDNA3.1 using the Directional TOPO Expression Kit (Invitrogen). The cDNA FcγRIIa A/A genotype (coding for FcγRIIa H/H-131) was generated from the FcγRIIa G/G genotype using site-specific mutagenesis.28,29 Vectors were sequenced. Six-well plates were seeded at 2.5×105 cells/well. Approximately 70% to 80% confluent cells were transfected using Polyfect Reagent (Qiagen). Cells expressing heterozygous FcγRIIa R/H-131 were obtained by cotransfection with FcγRIIa G/G and A/A. Mock-transfected cells were treated with transfectant reagent only. Anti-CD32–FITC staining revealed similar transfection efficiency for the alleles.
Binding Assays With Nonlabeled CRP and Anti-CRP Antibodies Using FACS
CRP-binding assays were performed with COS-7 cells 48 hours after transfection as described.18 Detached cells were incubated with different preparations of CRP at a concentration of 1.74 μmol/L, which corresponds to 200 μg/mL. This concentration was used by Bharadway et al18 and found to induce optimal signaling by Chi et al20 in ice-cold PBS containing 0.05% azide and 0.1% BSA (PAB) for 1 hour on ice. After washing with PAB, cells were stained with anti-CRP (2C10) for 0.5 hours on ice. Cells were washed and stained with PE-GAM–F(ab′)2 for 0.5 hours on ice in dark. Cells were washed and analyzed using FACS with CellQuest software (BD Biosciences). A total of 30 000 cells were gated by fluorescence-1 (green) and fluorescence-2 (red). A total of 95% mock-transfected cells stained in the absence of CRP were assessed as background. Cell viability assays (trypan blue) revealed 99% viable transfected cells. Staining with IgG1-PE isotype antibody was used as a control. Because of identical results for the different CRP preparations, rCRP was used for all subsequent analyses. Results are expressed as mean±SD. Scores were compared using Student paired t test (Microsoft Excel 2000). A P<0.05 was considered statistically significant.
Confocal Imaging and Analysis
Confocal images were collected using a laser scanning confocal fluorescence microscope with single-fluorophore sensitivity.24–26 An Ar+/Kr+ ion laser (Spectra Physics 164) and an HeNe laser (Polytec) were used for fluorescence excitation at 514.5 nm and 632.8 nm, respectively. The excitation light was focused into the sample, and the resulting fluorescence emission was collected by a water immersion objective (C-Apochromat 63×/1.2 W; Zeiss). Highly efficient detection in 2 spectral channels (green 557 to 607 nm; red 665 to 850 nm) was accomplished by splitting the fluorescence light using custom-made bandpass filters in conjunction with dichroic mirrors (AHF) and subsequent detection with single-photon counting detectors (AQR-14; Perkin–Elmer). The whole instrument is controlled by our homemade software.
Confocal fluorescence images consisting of 128×128 pixels were acquired in a field of 90×90 μm2 with a depth resolution of ≈2 μm for both excitation wavelengths with 0.5-μW laser power incident on an area of 0.3 μm2.
For quantitative analysis, the fluorescence emitted by membranes of selected cells was examined as a function of time after incubation with fluorescently labeled CRP or anti-CD32 antibodies or after equilibration with these proteins at different concentrations. Within 1 series of measurements, the same number (typically 100) of the brightest pixels from the membrane of a chosen cell was analyzed.
CRP and monoclonal anti-CRP antibodies were labeled with Cy3-N-hydroxy-Succinimidyl ester (NHS) (Amersham) by coupling the succinimidyl ester derivative of the dye to amine groups in phosphate buffer at pH 8.2. Anti-CD32 (KB61) antibodies were conjugated with Alexa Fluor 647 dye (Molecular Probes) using maleimide coupling to free thiol groups. Unreacted dye was removed by gel filtration. The degree of labeling was kept low (1 to 2 fluorescent labels per protein molecule, as quantified by optical absorption spectroscopy) to minimize dye interactions. Pseudonative SDS-PAGE revealed identical bands for nonlabeled and Cy3-NHS–labeled CRP.
Solutions containing FcγRIIa-transfected COS-7 cells were transferred to a sandwich chamber consisting of 2 glass cover slips separated by mylar spacers (thickness 200 μm). After 15 minutes, cells were exposed to solutions of CRP–Cy3 at 0.87 μM (100 μg/mL, shown to induce optimal signaling20) or anti-CRP antibody/CRP–Cy3 complexes at 6.7 nM (1 μg/mL, for anti-CRP antibody 2C10). After washing, receptor staining was performed using solutions with 0.5 μM Alexa 647-labeled anti-CD32 antibodies, and subsequently, confocal images were taken.
Association Kinetics and Equilibrium Binding
The kinetics of association of Cy3-labeled CRP to FcγRIIa-transfected COS-7 cells was studied by acquiring confocal images as a function of time. For the kinetics of the anti-CRP antibody/CRP complex, a constant concentration of 0.87 μM unlabeled CRP was used in combination with different concentrations of Cy3-labeled anti-CRP antibodies.
For studies of equilibrium binding, incubation times were adjusted in accordance with the kinetic data. COS-7 cells were exposed to different concentrations of Cy3-labeled CRP and CRP/anti-CRP–Cy3 complexes. Confocal images were analyzed to assess the degree of saturation of the receptors.
FACS Analysis Using Nonlabeled CRP and Anti-CRP Antibodies
FACS analysis of FcγRIIa- and mock-transfected COS-7 cells was performed in the presence and absence of CRP. FcγRIIa–131R/R-transfected cells showed 54.2% positivity (PE staining) after incubation with 1.74 μM CRP (Figure 1a). This experiment confirmed the original data,18,19 which led to the interpretation of high-affinity binding of CRP to FcγRIIa. Three different CRP preparations (partially purified, highly purified, and rCRP) yielded identical results (data not shown).
Binding assays with CRP (1.74 μM) were performed in cells transfected with the FcγRIIa alleles (131R/R, 131R/H, and 131H/H). A decrease in staining was seen in the order 131R/R→131R/H→131H/H in the presence (Figure 1b, black) (RR:RH:HH=1.6:1.2:1.0, RR/RH[P=0.0015], RH/HH[P=0.000056], RR/HH[P=0.041]); and absence (Figure 1b, dotted) of CRP, (RR:RH:HH=2.8:2.0:1.0, RR/RH[P=0.0052], RR/HH[P=0.0017], RH/HH[P=0.0006]; and also for a mouse IgG1 isotype control (Figure 1b, hatched; (RR:RH:HH=2.5:1.8:1.0, RR/RH[P=0.0027], RR/HH[P=0.0002], RH/HH[P=0.0079]). The differences in staining reflect the differences in binding of IgG1 to the “high” (131R/R) and “low responder” (131H/H) forms of FcγRIIa.29 Treatment of FcγRIIa–131R/R-transfected cells with preformed anti-CRP antibody/CRP complex led to the same positive results (Figure 1c, right) as obtained by addition of CRP to cells and subsequent incubation with anti-CRP antibodies (Figure 1c, left).
After incubation of FcγRIIa-transfected COS-7 cells with CRP–Cy3 and subsequent washing with PBS, green fluorescence from the cell membrane was observed by confocal imaging. As an example, Figure 2a shows CRP–Cy3 binding to a cell that strongly expresses FcγRIIa. The fluorescence shows a focal pattern. The signal was much stronger when the cells were coincubated with unlabeled anti-CRP antibodies (Figure 2c). During incubation of cells that had bound CRP–Cy3 (Figure 2a) with anti-CD32–alexa 647 (Figure 2b), strict colocalization of binding sites for CRP–Cy3 and anti-CD32–alexa-647 was observed. Incubation of CRP–Cy3/ anti-CRP antibody complex binding cells (Figure 2c) with anti-CD32–alexa 647 (Figure 2b) also showed strict colocalization of binding sites for CRP–Cy3 and anti-CD32–alexa-647. Competitive binding was evident from the ratio of red to green membrane fluorescence compared with the ratio obtained when incubating anti-CD32 antibodies before the immune complexes (data not shown). Figure 2e and 2f displays control cells from a different area of the sample shown in Figure 2a and 2b that do not express FcγRIIa and do not bind CRP.
To further confirm the interaction of CRP with FcγRIIa in the absence of anti-CRP antibody, we imaged cells before and during equilibration with Cy3–CRP (Figure 3a and 3b). Incubation of Cy3–CRP (4.2 μM) cells with high concentrations of unlabeled CRP (100 μM) revealed a moderate decrease in fluorescence on the hour time scale, suggesting competitive inhibition (data not shown). Figure 3c and 3d shows consecutive incubation of Cy3–CRP-incubated cells with anti-CRP antibody, which caused a pronounced increase of membrane fluorescence (Figure 3d). Apparently, residual CRP–Cy3 diffusing freely in solution was trapped by FcγRIIa on the membranes mediated by anti-CRP antibody. Experiments with isotype-matched control antibodies at identical concentrations did not show any enhancement of CRP binding to the cells (data not shown). After addition of 10% human AB serum, a slight increase in CRP binding (at ≈50% receptor saturation) was observed, possibly attributable to affinity enhancement by ligand (eg, lipoprotein) binding to CRP.
The association rate of CRP and anti-CRP antibody/CRP with FcγRIIa was determined from an analysis of the membrane-located fluorescence as a function of incubation time (Figure 4). The data in Figure 4a show that equilibration with Cy3-labeled CRP at a concentration of 0.87 μM takes >1 hour. The kinetics are observed to speed up in proportion to the CRP concentration, and consequently, at 5.2 μM CRP, equilibration takes only a few minutes. Assuming that the dissociation rate coefficient is much smaller than the association rate coefficient, an exponential fit of the kinetic data yields a second-order association coefficient of (370±100) M−1s−1. The assumption is justified because we did not observe significant CRP dissociation from FcγRIIa on the hour time scale. Kinetic experiments on the anti-CRP antibody/CRP complex were performed at fixed CRP concentration of 0.87 μM and varying the concentration of Cy3-labeled anti-CRP antibodies (0.67 nM, 2 nM, 67 nM; Figure 4b). A linear increase of the association rate with anti-CRP antibody concentration was observed in the low concentration range, yielding a second-order association rate coefficient of (1.1±0.3)×106 M−1s−1 (at 0.87 μM CRP).
Equilibrium Binding Studies
The affinities of CRP and anti-CRP antibody/CRP to FcγRIIa were examined quantitatively by confocal imaging. The saturation of receptors with ligands was determined from the membrane-located fluorescence as a function of the free ligand concentration. Figure 5 shows the data as symbols; lines are best-fit model calculations assuming simple receptor/ligand equilibria. For CRP–Cy3, the fit yields an equilibrium dissociation coefficient KD=3.7±1 μM for the interaction with FcγRIIa. For affinity studies of the anti-CRP antibody/CRP complex, unlabeled CRP was present at a concentration of 0.87 μM, and the concentration of Cy3-labeled anti-CRP antibodies was varied to obtain the saturation curve. From the data in Figure 5, an almost 2 orders of magnitude higher affinity of the complexes is apparent. Quantitative analysis of the data yields KD=45±20 nM.
In this study, we investigated binding of CRP to FcγRIIa in transfected COS-7 cells. FACS analysis of COS-7 cells transfected with FcγRIIa alleles in the presence and absence of CRP (Figure 1b) strongly suggested a critical involvement of anti-CRP antibody/FcγRIIa interactions in the detection of CRP binding. Thus, we applied ultrasensitive confocal fluorescence microscopy to clarify CRP interactions with FcγRIIa. This novel technique enables us to apply a very gentle labeling (1 to 2 fluorescent labels per protein molecule), which ensures that the protein is still in its functionally competent state.30 Despite the low emission level, the single-molecule sensitivity of the microscope still allows direct visual interpretation of the images. Two major observations were made: (1) CRP indeed binds to FcγRIIa, and (2) addition of anti-CRP antibodies leads to anti-CRP antibody/CRP complex formation and clustering of the ligand. This is obvious already from qualitative inspection of the confocal images: Cy3-labeled CRP colocalizes with FcγRIIa on the membrane surface. This fluorescence is absent for cells not expressing FcγRIIa and shows a focal pattern. Addition of excess unlabeled CRP results in a moderate decrease in the fluorescence emission on the hour time scale, suggesting competitive inhibition. Addition of unlabeled anti-CRP antibodies significantly increases the fluorescence emission (Figures 2 and 3⇑).
From the quantitative analysis of the membrane fluorescence (Figures 4 and 5⇑), the following results were obtained: (1) Dissociation of Cy3-CRP and anti-CRP antibody/CRP from FcγRIIa receptors was not observed after flushing the samples with buffer solution, which implies that both ligands are tightly bound, with dissociation times on the hour time scale. (2) In contrast, binding of labeled anti-CRP antibody alone to FcγRIIa receptors can only be observed as long as the antibody is present in solution. After purging with buffer solution, the enhanced fluorescence from the cell surface vanishes immediately, which is in agreement with off-rates of ≈1 s−1 reported for the interaction of mouse IgGs with FcγRII receptors.31 (3) The binding equilibrium of CRP–Cy3 with FcγRIIa is established very slowly (minutes to hours at micromolar concentrations) and can be quantified by a second-order association rate coefficient of 370 M−1s−1. This low value strongly suggests that persistent association to cell membranes can only be achieved by interaction with multiple FcγRIIa receptors. CRP is known to be pentameric,1,2 and a low probability of interacting with >1 receptor may explain the small association rate. Competitive inhibition by unlabeled CRP and the focal pattern of fluorescence also suggest receptor clustering in response to Cy3-labeled CRP. CRP interaction with multiple FcγRIIa receptors may be a prerequisite for kinase activity and FcγRIIa signaling.20,32 (4) The association of CRP is observed to speed up markedly in the presence of anti-CRP antibodies. This observation implies that the additional interaction of the anti-CRP antibody with FcγRIIa greatly assists in forming a persistent bond between CRP and the receptors. The bimolecular association rate coefficient of ≈106 M−1s−1 determined for the anti-CRP antibody/CRP complexes (at fixed CRP concentration) is close to the value of 0.4×106 M−1s−1 reported for the association of mouse IgGs with FcγRIIa.31 Interestingly, properties of the 2 components are combined in the immune complex, which shows an association rate coefficient typical of an antibody and a very small dissociation rate coefficient, as with CRP alone. (5) The KD for CRP dissociation from FcγRIIa is 3.7 μM, whereas it is ≈80-fold smaller for the anti-CRP antibody/CRP complex. Qualitatively, this decrease is expected from the behavior of the association rates (Figure 4). We note that our KD of 45 nM for the anti-CRP antibody/CRP complex is identical within the experimental error to the previously reported value of 66 nM for CRP binding, which was determined by antibody-dependent assays on transfected COS-7 cells.18
The visual demonstration that use of anti-CRP antibodies indeed affects CRP binding and leads to false-positive results supports the criticism22,23 directed toward the original observations.18,19 In view of the data presented in this report, some conclusions drawn on CRP interactions with Fc receptors may have to be reconsidered. It is of utmost importance to apply antibody-independent methods to the study of CRP interactions with Fc receptors. In view of the high sensitivity of the assay resulting from selective analysis of receptor-located fluorescence, the gentleness of single-fluorophore labeling of CRP, and the possibility of quantitative analysis, ultra-sensitive confocal fluorescence microscopy may be the method of choice to answer some of the most intriguing questions concerning CRP interactions with cellular receptors, such as: (1) What is the affinity of CRP to FcγRI compared with FcγRIIa?33 (2) Does ligand (for example, LDL) binding to CRP increase its affinity to Fc receptors? (3) Is clustering of Fc receptors with other cell surface molecules involved in CRP binding to leukocytes?
To conclude, ultrasensitive confocal fluorescence microscopy may significantly contribute to the understanding of CRP binding to nucleated cells, with the potential aim of developing CRP receptor blockers for the treatment of atherosclerosis and its sequelae.
This work was supported by Deutsche Forschungsgemeinschaft (SFB 451 and 569).
D.E.M. and C.R. contributed equally to this work.
- Received July 6, 2004.
- Accepted September 24, 2004.
Torzewski J, Torzewski M, Bowyer DE, Frohlich M, Koenig W, Waltenberger J, Fitzsimmons C, Hombach V. C-Reactive protein frequently colocalizes with the terminal complement complex in the intima of early atherosclerotic lesions of human coronary arteries. Arterioscler Thromb Vasc Biol. 1998; 18: 1386–1392.
Pasceri V, Willerson JT, Yeh ET. Direct proinflammatory effect of C-reactive protein on human endothelial cells. Circulation. 2000; 102: 2165–2168.
Torzewski M, Rist C, Mortensen RF, Zwaka TP, Bienek M, Waltenberger J, Koenig W, Schmitz G, Hombach V, Torzewski J. C-Reactive protein in the arterial intima: role of C-reactive protein receptor-dependent monocyte recruitment in atherogenesis. Arterioscler Thromb Vasc Biol. 2000; 20: 2094–2099.
Pasceri V, Cheng JS, Willerson JT, Yeh ET, Chang J. Modulation of C-reactive protein-mediated monocyte chemoattractant protein-1 induction in human endothelial cells by anti-atherosclerosis drugs. Circulation. 2001; 103: 2531–2534.
Zwaka TP, Hombach V, Torzewski J. C-Reactive protein-mediated low density lipoprotein uptake by macrophages: implications for atherosclerosis. Circulation. 2001; 103: 1194–1197.
Wang CH, Li SH, Weisel RD, Fedak PW, Dumont AS, Szmitko P, Li RK, Mickle DA, Verma S. C-Reactive protein upregulates angiotensin type 1 receptors in vascular smooth muscle. Circulation. 2003; 107: 1783–1790.
Yeh ET, Willerson JT. Coming of age of C-reactive protein: using inflammation markers in cardiology. Circulation. 2003; 107: 370–371.
Verma S, Li SH, Badiwala MV, Weisel RD, Fedak PW, Li RK, Dhillon B, Mickle DA. Endothelin antagonism and interleukin-6 inhibition attenuate the proatherogenic effects of C-reactive protein. Circulation. 2002; 105: 1890–1896.
Paul A, Ko KW, Li L, Yechoor V, McCrory MA, Szalai AJ, Chan L. C-Reactive protein accelerates the progression of atherosclerosis in apolipoprotein E-deficient mice. Circulation. 2004; 109: 647–655.
de Beer FC, Soutar AK, Baltz ML, Trayner IM, Feinstein A, Pepys MB. Low density lipoprotein and very low density lipoprotein are selectively bound by aggregated C-reactive protein. J Exp Med. 1982; 156: 230–242.
Chang MK, Binder CJ, Torzewski M, Witztum JL. C-Reactive protein binds to both oxidized LDL and apoptotic cells through recognition of a common ligand: Phosphorylcholine of oxidized phospholipids. Proc Natl Acad Sci U S A. 2002; 99: 13043–13048.
Tebo JM, Mortensen RF. Characterization and isolation of a C-reactive protein receptor from the human monocytic cell line U-937. J Immunol. 1990; 144: 231–238.
Bharadwaj D, Stein MP, Volzer M, Mold C, Du Clos TW. The major receptor for C-reactive protein on leukocytes is fcγ receptor II. J Exp Med. 1999; 190: 585–590.
Chi M, Tridandapani S, Zhong W, Coggeshall KM, Mortensen RF. C-Reactive protein induces signaling through Fc γ RIIa on HL-60 granulocytes. J Immunol. 2002; 168: 1413–1418.
Stein MP, Mold C, Du Clos TW. C-Reactive protein binding to murine leukocytes requires Fc γ receptors. J Immunol. 2000; 164: 1514–1520.
Wiedenmann J, Schenk A, Röcker C, Girod A, Spindler KD, Nienhaus GU. A far-red fluorescent protein with fast maturation and reduced oligomerization tendency from Entacmaea quadricolor (Anthozoa, Actinaria). Proc Natl Acad Sci U S A. 2002; 99: 11646–11651.
Kim HD, Nienhaus GU, Ha T, Orr JW, Williamson JR, Chu S. Mg2+-dependent conformational change of RNA studied by fluorescence correlation and FRET on immobilized single molecules. Proc Natl Acad Sci U S A. 2002; 99: 4284–4289.
Ying SC, Gewurz H, Kinoshita CM, Potempa LA, Siegel JN. Identification and partial characterization of multiple native and neoantigenic epitopes of human C-reactive protein by using monoclonal antibodies. J Immunol. 1989; 143: 221–228.
Sambrook J, Russell DW. Molecular Cloning. Cold Spring Harbor, NY: CSHL Press; 2001.
Warmerdam PA, van de Winkel JG, Gosselin EJ, Capel PJ. Molecular basis for a polymorphism of human Fc γ receptor II (CD32). J Exp Med. 1990; 172: 19–25.
Sako Y, Yanagida T. Single-molecule visualization in cell biology. Nat Rev Mol Cell Biol. 2003; (suppl): SS1–SS5.