Glycobiology Advance Access originally published online on October 30, 2002
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Glycobiology, 2003, Vol. 13, No. 1 11-21
© 2003 Oxford University Press
Carbohydrate-binding properties of human neo-CRP and its relationship to phosphorylcholine-binding site
Department of Biology, Johns Hopkins University, Baltimore, MD 21218, USA
Received on June 26, 2002; revised on August 20, 2002; accepted on August 23, 2002
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Binding characteristics of two types of ligands for human neo-C-reactive protein (neo-CRP), which is a conformationally altered but physiologically relevant form of CRP, were studied fluorometrically by probing CRP immobilized on a polystyrene surface with europium-labeled ligands. Two Eu-ligands used were bovine serum albumin derivatives that contain on average 40 residues of ligand structures, one derivative containing phosphorylcholine (PC) and the other lactosyl residues. The PC-containing ligands required the presence of calcium for binding, whereas galactose-containing derivatives bound in the absence of calcium. The optimal pH for the PC-dependent binding was broad (pH 6–8), whereas the best binding pH for the galactose-dependent binding was around 6. The carbohydrate-mediated binding is rather nonspecific: the binding site prefers galactose configuration, but other hexoses can be accommodated. The two best monosaccharide inhibitors at this site were galactose-6-phosphate and galacturonic acid, suggesting the importance of having a negatively charged group at C-6 position of galactose. In fact, the phosphate-binding site is common to both PC and sugar phosphates, and the choline- and the sugar-binding sites are probably located on either side of the phosphate-binding site. Binding characteristics of Eu-labeled PC-BSA to neo-CRP are quite similar to that found for native CRP in solution phase [Lee et al. (2002) J. Biol. Chem., 277, 225–232], whereas binding of sugar phosphates by neo-CRP shows considerably less stringent requirements compared to native CRP. For instance, galactose-
1-phosphate was not inhibitory at all in the native CRP binding assay, whereas it was a good inhibitor in the neo-CRP assay. Key words: carbohydrate ligand / C-reactive protein / fluorometric / determination / neo-C-reactive protein / phosphoryl choline
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Human C-reactive protein (CRP) is considered to be a prototype acute-phase reactant. Although the basal serum level of CRP in human is very low (0.1 µg/ml), its concentration in acute phase inflammation may increase as much as a thousandfold (Mortensen, 2001
). It is a defense molecule against certain microbes, because on binding microbes it can initiate classical complement cascade, leading to eventual elimination of the microbes (Szalai et al., 1997). However, more important functions of human CRP may be in the realm of housekeeping at the site of inflammation and tissue destruction. CRP has been shown to help eliminate compromised and apoptotic cells by the phagocytic cells (Volanakis, 2001).
CRP belongs to a group of proteins known as pentraxins, which are composed of five identical monomers noncovalently associating in a ring form in such a way so that each subunit makes contact with two neighboring subunits. X-ray structure of hCRP shows that each subunit is a ß-barrel, and that all five subunits are assembled in such a way as to face the same direction, thus presenting two distinct ring-shaped surfaces (Shrive et al., 1996). On one of the surfaces, two calcium ions are bound close to each other, and the phosphate group of phosphorylcholine (PC) is bound directly to the calcium ions (Thompson et al., 1999). In addition to PC, which is the best-studied ligand, there are at least two other well-known ligand groups: polycationic compounds, such as poly-L-lysine (PL) and protamine sulfate, and compounds containing galactose and related structures (Kilpatrick and Volanakis, 1991). All three ligand groups appear to bind on the same side of CRP, which is the side on which calcium ions are present (Lee et al., 2002).
hCRP circulates in blood as stable pentamer in soluble form, but there is another form of CRP, known as neo-CRP or mCRP (monomeric CRP), which is produced under mildly denaturing conditions, such as low pH and surface adsorption (Potempa et al., 1987). This form of CRP appears to be a physiologically relevant molecular species because it is found at the site of inflammation and is also known to exert stimulatory effects on immunologically important cells (Potempa et al., 1988). Neo-CRP expresses a new set of antigenic determinants, indicating that native CRP has undergone considerable conformational change in the transformation (Ying et al., 1989).
Köttgen et al. (1992) studied binding of surface-immobilized CRP and reported that it binds lactosylated human serum albumin (HSA) but not derivatives containing other terminal sugars, such as mannose, glucose, N-acetylglucosamine, and sialic acid. We also used plastic- surface immobilized CRP, presumably expressing neo-CRP epitopes, and studied its PC-dependent and carbohydrate-dependent binding in detail. We found that (1) neo-CRP binds carbohydrate ligands much better in the absence of calcium, whereas the binding of choline-containing compounds strongly prefers the presence of calcium; (2) the specificity of neo-CRP with respect to the sugar structure is broad; and (3) the presence of a negatively charged group, such as a phosphate or a carboxylic acid group, on a sugar residue greatly increases the affinity of the parent sugar.
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Ligand-binding assay for microplate-immobilized CRP
Ligand-binding characteristics of neo-CRP were probed using Eu-labeled bovine serum albumin (BSA) conjugates containing large numbers of determinants, as will be described in detail in the Materials and methods section. The two Eu-labeled ligands used were Eu-PC40-BSA and Eu-Lac40-AD-BSA, each containing on the average 40 residues of prosthetic groups (see Materials and methods for detailed description) for PC- and carbohydrate-mediated binding, respectively. Both Eu-PC40-BSA and Eu-Lac40-AD-BSA bound to the immobilized CRP in the prosthetic group-specific manner; that is, binding of Eu-PC40-BSA and Eu-Lac40-AD-BSA was inhibited almost completely by PC (1 mM) and lactose (0.1 M), respectively. The bound amount of Eu-labeled ligands increased with increasing input, and although the net bound amount of Eu-PC40-BSA leveled off at around 200 nM input, that of Eu-Lac40-AD-BSA did not show leveling off up to 400 nM. We also used Eu-labeled PL of DP 240 (Lee et al., 2002
) in a similar fashion to probe for possible polycation-dependent binding. However, unlike the Eu-labeled BSA derivatives, the bound amount of Eu-PL was very low, about 10% of that of Eu-PC40-BSA, which made the determination rather inaccurate. For this reason, we did not pursue the study of polycation-based binding using Eu-PL. We suspect that the presence of a large number of positively charged amino groups on PL lowers the binding efficiency considerably as compared to protein-based ligands. As shown in Figure 1, the presence of calcium ion had an opposite effect on the binding of Eu-PC40-BSA and Eu-Lac40-AD-BSA. There is a sharp loss of binding of Eu-PC40-BSA below 0.1 mM CaCl2, whereas that of Eu-Lac40-AD-BSA was highest when calcium was not present and gradually decreased with increasing CaCl2 concentration up to 0.4 mM. The dependence of binding on pH of the incubation medium was studied from pH 4 to 9 using the following buffer systems with pH range shown in parenthesis: acetate buffers (4–4.5), MES [2-(N-morpholino) ethanesulfonic acid] buffers (5–6.5), HEPES [N-(2-hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid)] buffers (6.5–8.5), and AMPSO [3-[N,N-dimethyl(2-hydroxyethyl) amino]-2-hydroxylpropanesulfonic acid] buffers (8.5–9), all at 25 mM containing 0.15 M NaCl and 0.1% BSA. Calcium ion was absent for the Eu-Lac40-AD-BSA experiments and at 0.5 mM for the Eu-PC40-BSA experiments. The corresponding wash buffers contained 0.05% Tween 20 in lieu of BSA.
|
Figure 2 shows the pH curves obtained in such experiments, where the amounts of Eu-PC40-BSA and Eu-Lac40-AD-BSA bound at pH 7 were set as 100%. For Eu-PC40-BSA, the bound amounts showed a broad plateau from pH 6 to 8, beyond which there was a slight decrease. This decrease on the alkaline side appears to be the result of lowered affinity for calcium ion at higher pHs, because the bound amounts at pH 8.5 and 9 became 100% and 90%, respectively, when the calcium concentration was raised to 5 mM. On the acidic side, there is a sharper decrease below pH 5.5, showing essentially no binding at pH 4. Raising calcium concentration on the acidic side even up to 50 mM did not restore the activity. For Eu-Lac40-AD-BSA, the pH range for maximum binding was narrower, with the midpoint of the highest bound amount occurring around pH 6. Again there is a sharp decrease in the bound amount below pH 5.5 and a more gradual decrease above 6.5.
|
Inhibition assays using Eu-Lac40-AD-BSA
Inhibition assays for Eu-Lac40-AD-BSA, which were carried out under the optimal binding conditions, that is, incubation at pH 6 without added CaCl2, showed that the binding of Eu-Lac40-AD-BSA was inhibited by a wide variety of compounds. The results are compiled in several tables according to the category of compounds. Monosaccharides, without aglycon, were very poor inhibitors, IC50 of Gal, Man, and GalNAc being >>60, 96, and 160 mM, respectively. As shown in Table I, addition of a ß-aglycon as small as methyl increased the inhibitory potency of monosaccharides to a measurable level. For instance, whereas IC50 of Gal could not be determined, that of Me-ß-Gal was 22 mM. Replacement of 2-OH of methyl glycosides of Gal and Glc with 2-acetamido group increased the inhibitory potency by 20-fold and 5-fold, respectively. For both galactoside and glucoside, elongation of the ß-aglycon increased the inhibitory potency further by four- to fivefold (compare 6-trifuloroacetamidohexyl [TFA-ah] with Me). However, for GalNAc, the elongation of aglycon actually decreased the affinity, suggesting that a longer aglycon may be physically interfering at the binding site. This might mean that GalNAc-terminated glycans in general are not good inhibitors.
|
Methyl ß glycosides of L-arabinopyranose and D-xylopyranose, which are homomorphous with galactopyranoside and glucopyranoside, respectively but without the exocyclic hydroxymethyl group, had decreased affinity as compared to the hexose counterpart, suggesting that the hydroxymethyl substituent on the ring augmented affinity. Not listed in the table are two other ß-galactosides, allyl (IC50 = 3 mM) and p-nitrophenyl (PNP) (IC50>>30 mM). It appears that only a small increase in the length of aliphatic chain from one to three carbon atoms (methyl to allyl) is sufficient to increase the affinity to the level afforded by a much longer TFA-ah group. In contrast to many other carbohydrate-binding systems, such as legume lectins (Goldstein and Poretz, 1986
) and FimH adhesin of Escherichia coli (Firon et al., 1987) which prefer strongly the presence of an aromatic aglycon, a PNP aglycon, whether or ß, was actually detrimental to the inhibitory potency in the neo-CRP system. We also tested some ß-thiogalactosides, which turned out to be much poorer inhibitors than Me-ß-Gal. Aglycons of these ß-thiogalactosides were methoxycarbonylmethyl (90 mM), cyanomethyl (102 mM), and isopropyl (>>105 mM) with IC50 shown in parentheses.
The following general conclusions can be drawn from Table I. For the sake of discussion, we define a good inhibitor as a compound with IC50 in low mM range; those with higher IC50 are considered poor inhibitors. (1) Monosaccharide glycosides are poor inhibitors in the absence of a long aliphatic aglycon, and (2) there appears to be no single strongly favored monosaccharide structure; perhaps ß-galactoside and /ß-mannoside being somewhat favored over others.
Table II shows inhibition data of a number of disaccharides arranged in the order of increasing IC50 values. Again there seems to be no uniquely preferred structure. Some conclusions that can be made are: (1) both Gal-Glyc and Man-Glyc disaccharides can be good inhibitors; (2) Galß(1,3)-linked disaccharides were good inhibitors, which agrees with the inhibition data of sugars reported by Köttgen et al. (1992); (3) a 2-acetamido group on the reducing-end sugar slightly improves the binding affinity (compare Galß(1,6)GlcNAc with Galß(1,6)Glc and Galß(1,4)ManNAc with Galß(1,4)Man); and (4) some Galß-linked disaccharides, including lactose, were poor inhibitors.
|
Table III shows the inhibitory potencies of phosphate-containing compounds. The phosphate group seems to be a dominant factor for the binding, because most of the sugar phosphates had IC50 values that are quite similar to that of Pi and are much lower than that of the parent sugar structures, indicating that the sugar moiety did not contribute significantly to the binding. However, the galactosyl configuration was still the most preferred structure, because Gal-6-P was definitely a better inhibitor than phosphate itself. Interestingly, PC was at least a 10-fold worse inhibitor than inorganic phosphate, and L-glycerophosphorylcholine was a very poor inhibitor, indicating that the presence of choline group is unfavorable. We tested two other galactose derivatives substituted with a negatively charged group at 6-position: Gal-6-S was not inhibitory at 23 mM, but galacturonic acid (GalA) was a good inhibitor. Because of the latter observation, we tested other carboxylic acid-containing compounds. As shown in Table IV, the shortest carboxylic acid (formic acid) was a very poor inhibitor, but attachment of an aliphatic chain with one or more hydroxy groups improved the affinity dramatically. The effect of sugar residue was more prominently manifested with the carboxylic acid derivatives than the phosphate-containing compounds. For instance, GalA was a 20-fold better inhibitor than acetic acid, whereas Gal-6-P was only twice as good as Pi. Also, the preference of the Gal configuration over the Glc configuration is more pronounced, because GalA was a nearly 10-fold better inhibitor than GlcA, but Gal-6-P was only 3-fold better than Glc-6-P.
|
|
Because negatively charged groups in general may give some degree of inhibition, a number of anions were tested. Aside from phosphate (predominant form being H2P
at pH 6), acetate (IC50 = 65 mM) and sulfate (IC50 = 70 mM) were the only two that gave measurable IC50. Nitrate and HEPES were totally noninhibitory at the highest concentration tested (about 100 mM), and the IC50 of MES and bicarbonate may be around 140 mM. Chloride, which is present at 0.15 M in the assay buffer, was not inhibitory; in fact the bound amount of ligand increased with further increase in NaCl concentration, with 0.3 M being optimal for the binding.
Based on the observation that terminal Galß- and Man/ß- may be reasonably good targets for binding by neo-CRP, a number of natural glycoproteins were tested. We think that IC50 values higher than 1.3 mg/ml mean that the protein mass is nonspecifically influencing the binding of Eu-Lac40-AD-BSA, because their inhibition curves were rather flat and BSA itself had an IC50 in this range (about 1.7 mg/ml, 26 µM). None of the Man-terminated glycoproteins (soybean agglutinin, invertase, and hen ovalbumin) showed inhibition; their IC50 values were above 1.3 mg/ml. Ovalbumin had the highest IC50 of 2.8 mg/ml (62 µM). Other noninhibitory glycoproteins include 1-acid glycoprotein, human transferrin, and human IgG. Those with IC50 in the range of 0.1–1 mg/ml were designated poor inhibitors; these include bovine submaxillary mucin (IC50 = 0.15 mg/ml), bovine asialotransferrin (IC50 = 0.28 mg/ml, 3.7 µM), bovine transferrin (IC50 = 0.40 mg/ml, 5.3 µM), and desialylated 1-acid glycoprotein (IC50 = 0.58 mg/ml, 15 µM) in the order of decreasing affinity. The two best glycoprotein inhibitors were bovine fetuin with or without desialylation, IC50 values being in the range of 30–55 µg/ml (0.7–1.1 µM). This means that these proteins are more potent inhibitors on the basis of molar concentration than highly sugar-substituted neoglycoproteins (see later discussion). It may be that fetuin contains some protein determinant or carbohydrate determinant of unknown nature that will augment the affinity, and this point will be the subject of future studies. In general, however, Man-terminated natural glycoproteins are not inhibitory, and Gal-terminated glycoproteins are poor inhibitors even when there are many terminal Gal residues.
Other Eu-labeled neoglycoproteins as reference ligand
Inhibition data obtained with Eu-Lac40-AD-BSA suggested that neo-CRP exhibits very lax specificity with respect to sugar structure. It appears to accommodate all the hexoses tested, with a small but significant preference for Gal and Man structures. To test whether Gal and Man are indeed occupying the same space in the binding area, three other neoglycoprotein, Gal42-AD-BSA, Man32-AD-BSA, and Galß(1,4)GlcNAc (LacNAc)33-AD-BSA, were labeled with Eu–diethylenetriaminepentaacetic acid (DTPA). When these new Eu-labeled neoglycoproteins were used in addition to Eu-Lac40-AD-BSA to test the inhibitory potencies of Me-ß-Gal and Me--Man, the IC50 values obtained varied little for each inhibitor regardless of the labeled reference ligand used. With Eu-labeled Gal42-, Man32-, LacNAc33-, and Lac40-AD-BSA as reference, the IC50 of Me-ß-Gal was 20, 18.5, 20, and 22 mM and that of Me--Man was 13, 11.5, 20, and 15.5 mM, respectively. This result strongly suggests that galactoside and mannoside are occupying the same space in the binding pocket.
O-glycoside-containing neoglycoproteins
Because Gal-6-P and GalA are much better inhibitors than galactose or Me-ß-Gal, we wanted to test the inhibitory potencies of neoglycoproteins containing Gal-6-P and GalA. Most of the neoglycoproteins on hand in our laboratory are of AI [-SCH2C(=NH)-] or AD [-SCH2CONH(CH2)2-] linker type, both types containing thioglycosides linked via BSA amino groups (see Figure 3). However, we found that both AI- and AD-type neoglycoproteins behaved erratically as inhibitors. For example, the IC50 of Gal33-AD-BSA and Gal33-AI-BSA were 7.2 µM and 
30 µM, respectively, and that of Man29-AI-BSA and Man35-AD-BSA were 7 µM and >30 µM, respectively. We suspect that this unpredictable behavior stems partly from the fact these neoglycoproteins contain thiosugars, because we found that ß-thiogalactosides are very poor inhibitors compare to O-glycosides. Therefore, we developed a new type (ASA-type) of neoglycoprotein that contains O-glycosides linked to protein amino groups via a linking arm of -O(CH2)6NHCO(CH2)2CONH(CH2)2- (see Figure 3 and Materials and methods). Several ASA-type neoglycoproteins, including those containing Gal-6-P and GalA, were prepared and tested as inhibitors. Results are shown in Table V. Unlike thiosugar-containing neoglycoproteins, all the ASA-neoglycoproteins had measurable IC50 values in the µM range. The inhibitory potency increased only slightly with sugar density, and above 
25 mol sugar/mol BSA there may not be any further increase in inhibitory potency. When IC50 values at similar sugar substitution levels were compared, Gal-6-P-ASA-BSA had the highest affinity, followed by GalA-, Gal-, and Man-ASA-BSA in that order, which more or less parallels the IC50 values of sugars and glycosides.
|
|
Inhibition assays using Eu-PC40-BSA
Based on the binding properties of Eu-PC40-BSA, the inhibition assays for Eu-PC40-BSA were carried out at pH 6 in the presence of 0.5 mM CaCl2. We have shown earlier (Lee et al., 2002
) that native CRP binds PC very tightly (IC50 = 40 µM) in the presence of calcium and that Gal-6-P binds in the same general binding area, but with 50-fold lower affinity, with galactosyl group occupying the opposite side of phosphate from the choline-binding area (Figure 4). To test if neo-CRP also possesses a similar binding-site arrangement as that of native CRP, PC and related compounds as well as sugar phosphates and others were tested as inhibitors of Eu-PC40-BSA binding to neo-CRP. Table VI shows these results together with IC50 determined for native CRP (Lee et al., 2002) and IC50 values determined for neo-CRP in the absence of calcium (i.e., assayed with Eu-Lac40-AD-BSA). In the presence of calcium ions, both native and neo-CRP had similar IC50 values for PC and Gal-6-P. However, as already noted, the affinity of PC for neo-CRP decreased dramatically in the absence of calcium.
|
|
The major difference between native CRP and neo-CRP is a stringent specificity for sugar phosphates shown by native CRP in comparison to neo-CRP. As seen in Table VI, native CRP in the presence of calcium ions has a strong preference for Gal-6-P among all the sugar phosphates and is not capable of binding sugar-
1-phosphates and GalA (Lee et al., 2002). In contrast, neo-CRP binds all the sugar phosphates and GalA with similar affinity whether calcium is present or not. To make sure that this broad specificity of neo-CRP is not due to a slightly lower pH and CaCl2 concentration used in neo-CRP assay (i.e., pH 6 and 0.5 mM CaCl2) as compared to the conditions used in the native CRP assay (pH 7 and 5 mM CaCl2), a few inhibitors were tested in the neo-CRP assay at pH 7 with 5 mM CaCl2. Gal-6-P, Man-6-P, Gal-1-P, and GalA had IC50 values of 0.53, 1.5, 5.5, and 35 mM, respectively, as compared to 1.1, 3.0, 3.3, and 8.7 mM at pH 6 with 0.5 mM CaCl2, indicating that the relaxed specificity was more or less maintained at pH 7 and higher calcium concentration. Neutral sugars, such as lactose and TFA-ah-ß-Gal, do not possess a structural element that will compete directly at the PC-binding site. Nevertheless, TFA-ah-ß-Gal does inhibit the binding of Eu-PC40-BSA, albeit at much higher concentrations than obtained for the inhibition of Eu-Lac40-AD-BSA (Table VI). This would reinforce the notion that the sugar-binding area is located very close to the PC-binding area.
PC presented in the form of BSA conjugates had highly enhanced affinity; the IC50 values of PC40- and PC21-BSA were 7 and 27 nM, respectively. However, of the three ASA-type neoglycoproteins tested, which were Gal30-ASA-BSA, GalA27-ASA-BSA, and Gal-6-P30-ASA-BSA, only the last BSA derivative showed measurable IC50 against Eu-PC40-BSA, which was 36 µM and about five-fold higher than the value obtained in the Eu-Lac40-AD-BSA inhibition (7.6 µM, Table V). The reason for the low affinity of Gal-6-P30-ASA-BSA and very poor inhibition by GalA27-ASA-BSA against Eu-PC40-BSA is not clear.
|
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
CRP is a covalently associated pentamer in the native state, with each subunit of pentamer making contact with two neighboring subunits in a circular fashion to form a disc-shaped molecule (Shrive et al., 1996
). However, when CRP is treated under mildly denaturing conditions, such as acidic pH (<3), 8 M urea, or immobilization on plastic surface, it appears to dissociate into monomers, which is accompanied by conformational changes and presentation of a new set of antigenic epitopes (Potempa et al., 1987; Ying et al., 1989). This form of CRP, known as mCRP or neo-CRP, has poor solubility in aqueous media and spontaneously forms large aggregates. Neo-CRP is a physiologically relevant molecular species in that it is found in normal vascular tissues (Diehl et al., 2000) and also found attached to tissues at the site of inflammation (Rees et al., 1988), and it also exerts stimulatory effect on monocytes, neutrophils, and platelets (Potempa et al., 1988).
A report by Köttgen et al. (1992) has indicated that surface-adsorbed CRP (presumably manifesting neo-CRP epitopes) binds lactosylated HSA, whereas HSA neoglycoproteins with terminal Man, Glc, and GlcNAc did not show measurable binding. Based on this observation, we developed a fluorometric assay using Eu-Lac40-AD-BSA and probed the sugar-binding characteristics of neo-CRP in detail. In addition, we probed the PC-dependent ligand-binding characteristics using Eu-PC40-BSA, which had been used previously to probe binding characteristics of native CRP (Lee et al., 2002).
As shown in Figure 1, the binding of Eu-PC40-BSA and Eu-Lac40-AD-BSA exhibited reciprocal requirements for calcium ions. The sugar-mediated binding was favored in the absence of calcium, whereas the PC-mediated binding required calcium. This suggests that the surface-adsorbed CRP presents two distinct conformations depending on the presence or absence of calcium ions. Native CRP in the solution phase also requires calcium for binding of Eu-PC40-BSA, and we showed that Gal-6-P competed well (IC50 = 1.9 mM) against Eu-PC40-BSA. From this and other observations, we proposed that the PC-binding site and Gal-6-P-binding site share the same phosphate-binding area with choline and galactose occupying the opposite side of phosphate from each other, as shown in Figure 4 (Lee et al., 2002).
In the present inhibition studies of neo-CRP, we found that sugar phosphates and glycerol phosphate inhibited the binding of both Eu-PC40-BSA and Eu-Lac40-AD-BSA with similar affinity (Table VI), suggesting that sugar phosphates are binding next to PC as observed for the native CRP. The fact that sugar and glycerol phosphates exhibited similar affinity with or without calcium ions suggests that the binding subsite for sugar phosphates does not undergo the calcium-induced conformational change. However, all the choline-containing compounds were much poorer inhibitors in the absence of calcium than in the presence of calcium. For instance, glycerol PC was about a 40-fold worse inhibitor than glycerol phosphate, and PC was at least a 1000-fold worse inhibitor in the absence of calcium (assayed against Eu-Lac40-AD-BSA) than in the presence of calcium (assayed against Eu-PC40-BSA) (Table VI). Obviously, the presence of choline residue is refractory to the binding in the absence of calcium ions. These facts plus the inability of Eu-PC40-BSA to bind to neo-CRP in the absence of calcium ions suggest that the choline-binding subsite collapses by some sort of conformational change when the calcium ion is taken away.
Although the location of Gal-6-P binding site is similar between the native CRP and neo-CRP, there appear to exist subtle differences in the ligand-binding characteristics at this subsite. In essence, native CRP (in the presence of calcium) has much more stringent structural requirement for the phosphate substitution as compared to neo-CRP (with or without calcium). This is most dramatically shown by the fact that sugar 1-phosphates were either noninhibitory or very poor inhibitors for native CRP, whereas neo-CRP bound all the sugar phosphates, including sugar 1-P, with similar IC50 values (in the mM range) with or without calcium present (i.e., inhibition against Eu-PC40-BSA or Eu-Lac40-AD-BSA) (Table VI). Similarly, GalA, which is totally noninhibitory for native CRP, showed an IC50 in the mM range for neo-CRP. This relaxed specificity of neo-CRP prevailed whether assays were done at pH 6 or pH 7 with a low calcium concentration (0.5 mM) or a much higher concentration of 5 mM, suggesting that the immobilization of CRP on a plastic surface caused a structural change responsible for the expression of the more relaxed sugar specificity.
An overall picture emerging for the carbohydrate-binding site of neo-CRP depicts a rather amorphous binding site, where the binding at the subsite for a negatively charged group, for example, phosphate and carboxylic acid, is a dominant factor and a site where the presence of hydroxyl group(s) and hydrophobic interactions are somehow helpful. The positive effect of a hydroxyl group can be seen in the comparison of hexose glycosides versus pentose glycosides (Table I), and glycolic acid versus acetic acid (Table IV). The positive effect of hydrophobic interactions can be seen in the comparison of methyl glycosides versus free sugars: the positive effect of having N-acetyl group on the penultimate sugar (Table II) and the ability of a longer aliphatic aglycon to increase affinity dramatically (Table I). In this regard, it is interesting to note that GalNAc glycosides showed an opposite effect in that the attachment of a longer aglycon lowered the inhibitory effect (Table I). It may be that the energetically most favored orientation of Me-ß-GalNAc in the binding site is different from that of neutral hexose derivatives (such as Me-ß-Gal), and the presence of a long aglycon somehow interferes in the binding in this alternate orientation. Though sugar phosphates are good inhibitors as such, neutral sugars can become good inhibitors, if a proper combination of structural elements are present, for example, sugar structure being galactose and having a long aliphatic aglycon or a ß-1,3-linked GalNAc/GlcNAc as the penultimate sugar (Table II).
Multiple presentation of a determinant, such as PC, on BSA conjugates increased binding affinity tremendously. The IC50 of PC40-BSA in neo-CRP assay was 7 nM, which represents 70,000-fold increase in affinity compared to glycerol-PC (IC50 = 0.5 mM), which resembles most closely to the PC structure in the BSA conjugates (Figure 3). The same PC40-BSA had an IC50 of 100 nM in the native CRP assay (Lee et al., 2002). The 14-fold higher affinity of PC40-BSA for neo-CRP than for native CRP most probably results from the presence of many neo-CRP-binding sites in close proximity to each other on the microplate surface, as compared to a single CRP pentamer interacting in the native CRP assay. In contrast, neoglycoproteins showed a no more than 500-fold increase in affinity compared with the component monomer. For instance, Gal30-ASA-BSA had an IC50 of 10 µM compared to that of TFA-ah Gal being 4 mM. The reasons for this large discrepancy in the affinity enhancement factor between PC-BSAs and sugar-BSAs are not clear at this time. One obvious structural difference between the two types of BSA conjugates is a long, flexible linking arm between sugars and BSA, as opposed to a very short linking arm connecting PC to BSA (Figure 3). A short linking arm should result in more rigid orientation of PC residues relative to each other and to the BSA surface, and therefore it is expected to have lower entropic energy loss in the binding event than the binding of more flexible sugar residues on ASA-type neoglycoproteins.
The level of CRP in serum has been a powerful clinical index in assessing the presence of infection and inflammation. It was shown recently that the increased level of CRP in serum indicates increased future risk of coronary events (Pepys and Berger, 2001). CRP is found attached at the site of inflammation, including atherosclerotic lesion (Reynolds and Vance, 1987), where neo-CRP antigenic epitopes are often expressed. However, exact mechanisms of such CRP accumulation at the inflammation sites and the subsequent events and consequences of such accumulation are not well understood. Because the binding of CRP and recruitment of its binding partners are likely to be the first sequences of event at the inflammation sites, it is hoped that a detailed knowledge on the binding characteristics and partners of both native and neo-CRP, such as described in this study, would be instrumental in understanding and interpretation of in vivo events.
|
Materials and methods |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Materials
Unless otherwise noted, chemicals and proteins were obtained from Sigma (St. Louis, MO). BSA was from Pentex (Kankakee, IL); organic buffers, MES, HEPES, and AMPSO were from Research Organics (Cleveland, OH). Recombinant human CRP was a gift of Dr. I. Takagahara (Oriental Bio-Service Kanto, Tsukuba, Japan). Preparations of 6-aminohexyl (ah) glycosides of ß-Gal,
-Man, and ß-lactose and the preparation of DTPA labeling reagent, mono(2,2-dimethoxyethyl)amino derivative of DTPA, have been reported (Lee and Lee, 2001; Weigel et al., 1979). The AI-type (Kranz et al., 1976) and AD-type (Lee and Lee, 1980) neoglycoproteins contain thiosugars and have linkages to protein amino groups as shown in Figure 3. PC-BSAs were prepared as described (Stults et al., 1987), and the linkage between PC and the protein amino group is also shown in Figure 3. The lower case number after the sugar or PC means the average number of substitution. For instance, PC40-BSA means that it contains on average 40 residues of PC per molecule of BSA. Fluorescence enhancing solution for europium fluorescence determination was prepared as described (Hemmilä et al., 1984).
General methods
TLC was carried out using silica gel 60 (F254) layers precoated on aluminum sheets (Merck, Darmstadt, Germany). After chromatography, compounds were visualized by UV absorption (aromatic groups) and by spraying with 15% sulfuric acid in 50% ethanol and heating on a hot plate (sugar and 2-aminoacetaldehyde dimethyl acetal group). Elemental analyses were performed by Galbraith Labs (Knoxville, TN), and 1H-nuclear magnetic resonance (NMR) was done using a Varian Unity 400 MHz FT-NMR spectrometer. For quantitative determination of various groups, the following methods were used: neutral sugars, phenol–sulfuric acid method (McKelvy and Lee, 1969); uronic acids, carbazole method (Bitter and Muir, 1962); organic phosphates, pyrolysis-molybdate method (Ames and Dubin, 1960); and acetal group, acid hydrolysis–neocuproine method (Dygert et al., 1965; Lee and Lee, 1980).
Determination of Eu fluorescence has been described elsewhere (Lee and Lee, 2001). Briefly, solutions containing either free or complexed (to DTPA derivative) europium (<1.5 pmol, preferably in <40 µl) were placed in the wells of a 96-well microplate, and to this was added 0.2 ml of enhancing solution. The content of wells was mixed well by tapping and swirling, and the mixture was let stand for 20 min. The fluorescence was read in a Wallac microplate fluorometer (Victor 1420 multilabel counter) using the time-resolved mode set for Eu fluorescence. The enhancing solution, which contains strong chelators of Eu as well as detergent, competes Eu off from DTPA-protein conjugate and places it in a milieu that makes Eu highly fluorescent. The fluorescence is proportional to the amount of Eu placed in the well up to 1.5 pmol, at which level the fluorescence reading was about 7 million cps (counts per second).
Labeling of PC40-BSA and Lac40-AD-BSA withEu-DTPA
Labeling of proteins with mono(2,2-dimethoxylethyl) amino derivative of DTPA has been described (Lee and Lee, 2001). Briefly, the acetal group of the reagent is converted to aldehyde by heating in 0.05 M HCl for 15 min. After cooling and neutralization, the aldehyde was conjugated to protein amino groups via reductive amination at pH 7 (0.2 M sodium phosphate buffer) for 48 h using pyridine-borane as reducing agent. The amount of DTPA incorporated is measured as Eu fluorescence after addition of Eu(NO3)3 and separation of the protein by gel filtration. The incorporation of DTPA depended on the ratio of the aldehyde reagent to protein as well as on the availability of protein amino groups. To minimize the undesirable effect of the DTPA-Eu substitution, the level of incorporated DTPA was kept below 1 mol/mol of protein. For neoglycoproteins and PC-BSAs, the use of 10-fold molar excess of reagent generally incorporated 0.25 to 0.5 mol/mol of Eu-DTPA.
Microplate binding assay
Based on the observation that CRP bound to polystyrene surface presented neo-CRP epitope (Potempa et al., 1987), the binding assay for neo-CRP was carried out on 96-well microtiter plates. The principle for this assay is as follows: CRP was immobilized onto microplate wells and then incubated with one of the Eu-labeled ligands. For the probing of PC-binding site and carbohydrate-binding site, Eu-PC40-BSA and Eu-Lac40-AD-BSA were used as labeled ligands, respectively. After incubation and washing, the bound ligand was determined by measuring the Eu fluorescence, as described.
CRP was adsorbed on the microplate wells according to the procedure of Köttgen et al. (1992) with some modifications. Briefly, CRP solution (0.3 µg in 0.1 ml) in plating buffer (25 mM HEPES buffer, pH 8.2, containing 0.15 M NaCl and 0.5 mM CaCl2) was placed in each well, which was then covered and incubated at 37°C for 72 h. To each well was added 10 µl of 10% BSA, and incubation was continued for further 2 h at room temperature. The liquid was removed by aspiration, and wells were washed twice with 0.15 ml of plate-washing buffer (25 mM MES buffer, pH 6.0, containing 0.15 M NaCl and 0.05% Tween 20). The Eu-ligand solution (100–400 nM, 0.1 ml) in incubation buffer (same as the plate-washing buffer, except containing 0.1% BSA instead of 0.05% Tween 20) was added, and the plate was incubated for 1 h at 37°C. The liquid in wells was removed one row at a time and washed twice with the plate-washing buffer (0.15 ml). The enhancing solution (0.2 ml) was added to all the wells; the plate was gently tapped a few times, covered, and let stand at room temperature for 20 min. Fluorescence was read as described previously. For the assay of Eu-PC40-BSA, calcium was present at 0.5 mM in both the incubation and plate-washing buffers, whereas no CaCl2 was added for the assay of Eu-Lac40-BSA.
For the immobilization of CRP, we found that immobilization at pH 8.2 gave higher amounts of bound ligand than immobilization at pH 6, and the presence of 0.5 mM CaCl2 was better than either without added CaCl2 or in the presence of 0.25 mM EDTA. Input of CRP higher than 0.2 µg per well did not increase the amount of ligand bound. For the incubation of the immobilized CRP with the labeled ligand, incubation time of 1 h at 37°C was sufficient. After the incubation, wells were washed up to six times, which showed that washing twice was sufficient, and thereafter the bound amount of Eu-Lac40-AD-BSA decreased gradually, about 6% per wash.
Inhibition assay
Inhibition assays were carried out essentially like the binding assay, except incubation mixtures contained various amounts of potential inhibitors together with Eu-labeled ligand. Inhibitors were assayed at five or six different concentration levels with the range of concentration chosen so as to achieve, if possible, 100% and 0% inhibition at the highest and the lowest inhibitor concentrations, respectively. At least two independent assays were carried out for each inhibitor.
For the inhibition of Eu-PC40-BSA binding to immobilized CRP, the incubation buffer contained 0.5 mM CaCl2 and the labeled ligand was at 50 nM. For the inhibition of Eu-Lac40-AD-BSA, the buffer did not contain CaCl2 and the labeled ligand was at 200 nM. For the inhibition assay using Eu-PC40-BSA, blank fluorescence value (=100% inhibition) was obtained either by substituting CaCl2 with EDTA or by adding PC at 1 mM in the incubation mixture. For the Eu-Lac40-AD-BSA assay, the blank fluorescence was obtained in the presence of 10 mM Gal-6-P instead of lactose, because the former turned out to be a much stronger inhibitor than the latter (see Results). Inhibition percentage calculated at each inhibitor concentration was plotted against the inhibitor concentration in logarithmic scale, and the inhibitor concentration that causes 50% reduction in the bound fluorescence was determined (IC50). At least two independent assays were carried out for each inhibitor. Typical inhibition curves are presented in Figure 5.
|
Preparation of BSA neoglycoproteins with -(CH2)6NHCO(CH2)2CONH(CH2)2- (ASA) linking arm
Preparation of ASA-type neoglycoproteins is shown in Scheme 1. It consists of synthesis of (a) ah glycoside; (b) linking arm, (CH3O)2CHCH2NHCO(CH2)2COOH (1); (c) conjugation of the two; and (d) conversion of acetal to aldehyde, followed by conjugation to BSA by reductive amination. Ah glycosides of neutral sugars were prepared as described (Weigel et al., 1979
). The preparation of ah ß-glycosides of galacturonic acid and Gal-6-P will be described elsewhere.
|
Synthesis of the linking arm (1)
To a stirred solution of 2-aminoacetaldehyde dimethylacetal (2.2 ml, 20 mmol) in 20 ml dry methanol was added batchwise succinic anhydride (2.5 g, 25 mmol) and equimolar amount of triethylamine. The reaction was monitored by TLC using ethyl acetate/acetic acid/water (8:2:1) as solvent. RF of the product 1 was 0.63 and that of the starting acetal reagent 0.23. When the acetal reagent was consumed, the mixture was evaporated to syrup and then left in a vacuum desiccator over NaOH pellets and 98% sulfuric acid overnight. The product was purified by silica gel chromatography using ethyl acetate as solvent. The fractions containing 1 were combined and evaporated to produce syrupy 1 in about 55% yield. This product was stored in the cold. When ethyl acetate/acetic acid/water (8/2/1) was used as eluant for the silica gel column, the yield would improve. However, it is imperative not to introduce acetic acid during the entire synthetic process, because complete removal of acetic acid from syrupy product is difficult and any residual acetic acid would interfere in the subsequent reaction step. 1H-NMR of 1 in CD3OD showed correct ratios of -OCH3 signal (
3.37, s, 6H) to methylene signals [ 3.286 (d, -CH2COOH, 2H); 2.54 (td, -CH2CONH-, 2H); 2.33 (td, -NHCH2CH-, 2H)] to methine signal ( 4.395, t, -CH-, 1H).
Formation of Gal-O(CH2)6NHCO(CH2)2CONHCH2- CH(OCH3)2 (asa-ß-Gal, 2) and asa-glycosides of -Man, ß-Lac, ß-GalA, and ß-Gal-6-P
The formation of amide bond between an ah glycoside and 1 was effected by activating the carboxylic acid group of 1 by N,N,N',N'-tetramethyl-O-(N-succinimidyl)uronium tetrafluoroborate (TSTU), and the reaction mixture was then immediately reacted with the amino compound (Bannwarth and Knorr, 1991). An example using ah ß-galactoside (ah-ß-Gal) is given later in this paragraph. Compound 1 (100 mg, 0.5 mmol) was dissolved in 1.25 ml solvent mixture consisting of dimethylformamide (DMF)/dioxane/water in 2:2:1 ratio, and diisopropylethylamine (210 µl, 1.2 mmol) and TSTU (152 mg, 0.5 mmol) were added. After 10 min at room temperature, ah-ß-Gal (80 mg, 0.29 mmol) was added and the mixture was stirred for 2 h. The mixture was evaporated to a syrup, dissolved in 4 ml of 0.1 M acetic acid, and fractionated on a column (2.5x140 cm) of Sephadex G-15 using 0.1 M acetic acid as eluant and collecting 6 ml per fraction. Fractions were monitored by TLC using ethyl acetate/acetic acid/water (3 : 2 : 1) as solvent. The product 2 (RF = 0.32), which was estimated to be 90% of the total char-positive material on the TLC plate, eluted ahead of a small amount of remaining ah-ß-Gal and an unidentified by-product. Fractions containing 2 were combined and evaporated to yield solid residue, which was crystallized by dissolving it in either ethanol or isopropanol and then adding ether, yielding 100 mg (0.23 mmol, 79% yield), mp 104–105°C. Elemental analyses: Calc for C20H38N2O10: C, 51.49; H, 8.21; N, 6.01. Found: C, 50.75; H, 8.58; N, 5.88. In a similar fashion, 
-acetal glycosides of mannose and lactose were synthesized. For GalA and Gal-6-P glycosides, the activation reaction was carried out in a solvent mixture containing DMF/water in 2:1 ratio without dioxane, and the aminohexyl glycoside was dissolved in water (0.4 ml) to which 0.2 ml DMF was added. The Gal-6-P glycoside was crystallized from 95% ethanol-ether. Of the five acetal derivatives, only the mannose derivative was not obtained in crystalline form.
The identity of products was confirmed by 1H-NMR and various colorimetric determinations on the crystalline products (except the mannose derivative). 1H-NMR data of D2O-exchanged 2 in D2O are described in detail later. The spectrum showed the presence of aglycon and galactose in equimolar amount: 4.517 (t, 5.6 Hz, methine H, 1H); 4.376 (d, 8 Hz, H1, 1H); 3.946–3.887 (m, H5, 1H); 3.911 (dd, 0.8 and 3.6 Hz, H4, 1H); 3.776 (dd, 8 and 11.2 Hz, H6, 1H); 3.732 (dd, 4.8 and 11.6 Hz, H6, 1H); 3.686–3.594 (m, CH2O-, 2H); 3.627 (dd, 3.2 and 10 Hz, H3, 1H); 3.487 (dd, 8 and 10 Hz, H2, 1H); and 3.417 (s, -OCH3, 6H). Other methylene signals are 3.338 (d, 5.2 Hz, 2H), 3.159 (t, 6.8 Hz, 2H), 2.517 (m, 4H), 1.62 (m, 2H), 1.49 (m, 2H), and 1.346 (m, 4H). Similarly, the -acetal glycosides of Man, Lac, GalA, and Gal-6-P all had the expected characteristic signals (H1, methine H, OCH3) in correct proportions. Phenol–sulfuric acid determination of neutral sugars indicated, on the weight base, 2, 6-N-(2,2-dimethoxyethylaminosuccinyl)aminohexyl (asa)--Man and asa-ß-Lac had 99.5%, 90%, and 96% of expected amount of the sugar component, respectively. Similarly, uronic acid and organic phosphate contents of asa-ß-GalA and asa-ß-Gal-6-P were 98% and 95%, respectively. The acetal content was determined using 2 as standard. The acetal contents of asa-glycosides of Man, Lac, GalA, and Gal-6-P were 91%, 96.5%, 97.3%, and 102%, respectively. Thus all the asa-glycosides are correct compounds in satisfactory purity.
Conjugation of -aldehydo-glycosides to BSA
An appropriate amount of 2 (or other glycosides) was heated at 100°C in 0.05 M HCl for 15 min. After cooling, the acid was neutralized with NaOH, and the mixture was evaporated to a syrup. The syrupy residue was dissolved in 0.2 M sodium phosphate buffer, pH 7, and added to solid BSA (5–10 mg), and more buffer was added, if needed, to make 10 mg/ml in the BSA concentration. Pyridine-borane was added in twofold molar excess of 2. Due to limited solubility of pyridine-borane in the buffer, the mixture contained small oily droplets that dissipated slowly overnight, at which point the same amount of pyridine-borane was added again, and the reaction was left to proceed for a total of 48 h. The mixture was dialyzed in a 12-kDa mwco dialysis bag sequentially against phosphate-buffered saline (PBS), 0.15 M NaCl (twice), and water (twice) in the cold. For neutral sugars, dialysis against PBS was omitted. Protein solutions were freeze-dried and stored in the cold. Incorporations of sugars were determined on the protein-weight basis, using appropriate colorimetric method already described. The ratio of the aldehydo reagent to BSA depended on the desired conjugation level. For example, to achieve incorporation level (mol sugar/mol BSA) of 15 and 25, the aldehydo reagent was used in 200- and 500-fold molar excess, respectively, over BSA.
|
Acknowledgements |
|---|
We thank Dr. I. Takagahara for a generous supply of human rCRP and Ms. M. Hughes for recording 1H-NMR spectra.
|
Footnotes |
|---|
1 To whom correspondence should be addressed; e-mail: reiko.lee{at}jhu.edu 
|
Abbreviations |
|---|
ah, 6-aminohexyl; AMPSO, 3-[N,N-dimethyl(2-hydroxyethyl)amino]-2-hydroxypropanesulfonic acid; asa, 6-N-(2,2-dimethoxyethylaminosuccinyl)aminohexyl; BSA, bovine serum albumin; CRP, C-reactive protein; DMF, dimethylformamide; DTPA, diethylenetriaminepentaacetic acid; HEPES, N-(2-hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid); HSA, human serum albumin; MES, 2-(N-morpholino)ethanesulfonic acid; NMR, nuclear magnetic resonance; PBS, phosphate buffered saline; PC, phosphoryl choline; PL, poly-L-lysine; PNP, p-nitrophenyl; TFA-ah, 6-trifuloroacetamidohexyl; TSTU, N,N,N',N'-tetramethyl-O-(N-succinimidyl)uronium tetrafluoroborate.
|
References |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsReferences |
|---|
Ames, B.N. and Dubin, D.T. (1960) The role of polyamines in the neutralization of bacteriophage deoxyribonucleic acid. J. Biol. Chem., 235, 769–775.
Bannwarth, W. and Knorr, R. (1991) Formation of carboxamides with N,N,N',N'-tetramethyl(succinimido) uronium tetrafluoroborate in aqueous/organic solvent systems. Tetrahedron Lett., 32, 1157–1160.[CrossRef]
Bitter, T. and Muir, H.M. (1962) A modified uronic acid carbazole reaction. Anal. Biochem., 4, 330–334.[CrossRef][Web of Science][Medline]
Diehl, E.E., Haines, G.K., Radosevich, J.A., and Potempa, L.A. (2000) Immunohistochemical localization of modified C-reactive protein antigen in normal vascular tissue. Am. J. Med. Sci., 319, 79–83.[CrossRef][Web of Science][Medline]
Dygert, S., Li, L.H., Florida, D., and Thoma, J.A. (1965) Determination of reducing sugar with improved precision. Anal. Biochem., 13, 367–374.[CrossRef][Web of Science][Medline]
Firon, N., Ashkenazi, S., Mirelman, D., Ofek, I., and Sharon, N. (1987) Aromatic alpha-glycosides of mannose are powerful inhibitors of the adherence of type 1 fimbriated Escherichia coli to yeast and intestinal epithelial cells. Infect. Immun., 55, 472–476.
Goldstein, I.J. and Poretz, R.D. (1986) Isolation, physicochemical characterization, and carbohydrate-binding specificity of lectins. In Liener, I.E., Sharon, N., and Goldstein, I.J. (eds.), The lectins. Academic Press, Orlando, FL, pp. 33–247.
Hemmilä, I.A., Dakubu, S., Mukkala, V.-M., Siitarl, H., and Lövgren, T. (1984) Europium as a label in time-resolved immunoflurometric assays. Anal. Biochem., 137, 335–343.[CrossRef][Web of Science][Medline]
Kilpatrick, J.M. and Volanakis, J.E. (1991) Molecular genetics, structure, and function of C-reactive protein. Immunol. Res., 10, 43–53.[Web of Science][Medline]
Köttgen, E., Hell, B., Kage, A., and Tauber, R. (1992) Lectin specificity and binding characteristics of human C-reactive protein. J. Immunol., 149, 445–453.[Abstract]
Kranz, M.J., Holtzman, N.A., Stowell, C.P., and Lee, Y.C. (1976) Attachment of thioglycosides to proteins: enhancement of liver membrane binding. Biochemistry, 15, 3963–3968.[CrossRef][Medline]
Lee, R.T. and Lee, Y.C. (1980) Preparation and some biochemical properties of neoglycoproteins produced by reductive amination of thioglycosides containing an -aldehydoaglycon. Biochemistry, 19, 156–163.[CrossRef][Medline]
Lee, R.T. and Lee, Y.C. (2001) A derivative of diethylenetriaminepentaacetic acid for europium labeling of proteins. Bioconjug. Chem., 12, 845–849.[CrossRef][Web of Science][Medline]
Lee, R.T., Takagahara, I., and Lee, Y.C. (2002) Mapping the binding areas of human C-reactive protein for phosphorylcholine and polycationic compounds. J. Biol. Chem., 277, 225–232.
McKelvy, J.F. and Lee, Y.C. (1969) Microheterogeneity of the carbohydrate group of Aspergillus oryzae -amylase. Arch. Biochem. Biophys., 132, 99–110.[CrossRef][Web of Science][Medline]
Mortensen, R.F. (2001) C-reactive protein, inflammation, and innate immunity. Immunol. Res., 24, 163–176.[CrossRef][Web of Science][Medline]
Pepys, M.B. and Berger, A. (2001) The renaissance of C reactive protein. Br. Med. J., 322, 4–5.
Potempa, L.A., Siegel, J.N., Fiedel, B.A., Potempa, R.T., and Gewurz, H. (1987) Expression, detection and assay of a neoantigen (neo-CRP) associated with a free, human C-reactive protein subunit. Mol. Immun., 24, 531–541.
Potempa, L.A., Zeller, J.M., Fiedel, B.A., Kinoshita, C.M., and Gewurz, H. (1988) Stimulation of human neutrophils, monocytes, and platelets by modified C-reactive protein (CRP) expressing a neoantigenic specificity. Inflamm., 12, 391–405.
Rees, R.F., Gewurz, H., Siegel, J.N., Coon, J., and Potempa, L.A. (1988) Expression of a C-reactive protein neoantigen (neo-CRP) in inflamed rabbit liver and muscle. Clin. Immunol. Immunopathol., 48, 95–107.[CrossRef][Web of Science][Medline]
Reynolds, G.D. and Vance, R.P. (1987) C-reactive protein immunohistochemical localization in normal and atherosclerotic human aortas. Arch. Pathol. Lab. Med., 111, 265–269.[Web of Science][Medline]
Shrive, A.K., Cheetham, G.M.T., Holden, D., Myles, D.A.A., Turnell, W.G., Volanakis, J.E., Pepys, M.B., Bloomer, A.C., and Greenhough, T.J. (1996) Three dimensional structure of human C-reactive protein. Nat. Struct. Biol., 3, 346–354.[CrossRef][Web of Science][Medline]
Stults, N., Lee, Y.C., Hoppe, C.A., Kawaguchi, K., Kohda, S., Takagahara, I., Koishi, T., and Liu, T.Y. (1987) Preparation of phosphorylcholine derivatives of bovine serum albumin and their application to the affinity chromatography of C-reactive protein. Anal. Biochem., 161, 567–573.[CrossRef][Web of Science][Medline]
Szalai, A.J., Agrawal, A., Greenbough, T.J., and Volanakis, J.E. (1997) C-reactive protein—structural biology, gene expression, and host defense function. Immunol. Res., 16, 127–136.[Web of Science][Medline]
Thompson, D., Pepys, M.B., and Wood, S.P. (1999) The physiological structure of human C-reactive protein and its complex with phosphocholine. Structure, 7, 169–177.[Medline]
Volanakis, J.E. (2001) Human C-reactive protein: expression, structure, and function. Mol. Immun., 38, 189–197.
Weigel, P.H., Naoi, M., Roseman, S., and Lee, Y.C. (1979) Preparation of 6-aminohexyl D-aldopyranosides. Carbohydr. Res., 70, 83–91.[CrossRef]
Ying, S.-C., Gewurz, H., Kinoshita, C.M., Potempa, L.A., and Siegel, J.N. (1989) Identification and partial characterization of multiple native and neoantigenic epitopes of human C-reactive protein by using monoclonal antibodies. J. Immun., 143, 221–228.[Abstract]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||