Polymorphism R62W results in resistance of CD23 to enzymatic cleavage in cultured cells

Article metrics


A nonsynonymous single nucleotide polymorphism (SNP) of the low-affinity IgE receptor (FcɛRII/CD23) gene resulting in an arginine to tryptophan exchange at amino-acid position 62 (R62W) has been associated with enhanced T-cell responses to antigen in allergic subjects. To explore the mechanism, a CD23(a) cDNA was cloned into the plasmid pCMVScript-CD23a-C with a C allele (R62). The pCMVScript-CD23a-T with T (W62) was produced using a site-directed mutagenesis approach. The pCMVScript-CD23a-C only (CC), mixture of pCMVScript-CD23a-T and pCMVSCript-CD23a-C (CT) and pCMVScript-CD23a-T only (TT) plasmids were transfected in Cos-7 cells at equivalence in transfection efficiency. No soluble CD23 was released from TT transfectants whereas a higher level of soluble CD23 was detected in CC than in CT transfectants. Human leukocyte elastase (HLE), cathepsin G, the dust mite allergen Der p I and ADAM 33 (A disintegrin and metalloproteinase) were found to cleave membrane CD23 in CC but not in TT transfectants, implying the resistance of CD23 to enzymatic cleavage associated with T mutant. Addition of tunicamycin resulted in the resistance of CD23 to Der p I mediated cleavage in CC but no change in TT transfectants. These results indicate that R62W influences the stability of membrane CD23 molecules due to possibly diminished N-glycosylation.


IgE has long been considered to be the key molecule in allergic diseases. Binding of IgE antigen complexes to its two receptors, high-affinity IgE receptor (FcɛRI) and low-affinity IgE receptor (FcɛRII/CD23), initiates immune, inflammatory and regulatory reactions in allergic diseases. Activation of FcɛRI seems to be involved in acute allergic responses to allergens mainly through mast cells and basophils. CD23 appears to mediate various pathogenic functions in IgE-mediated immunity including antigen focusing and presentation, regulating IgE synthesis, promoting T- and B-cell differentiation and growth and mediating the production of proinflammatory cytokines.1, 2, 3, 4, 5, 6, 7, 8

CD23 is a single-chain glycoprotein with an apparent molecular weight of 45 kDa. It is widely distributed on a variety of human cells9 and can be cleaved into soluble CD23 fragments, which contribute to its diverse immune activities.10, 11 Two subtypes, FcɛRIIa/CD23a and FcɛRIIb/CD23b, have been identified. They differ in six to seven different amino acids in the N-terminal cytoplasmic side, but share the common C-terminal extracellular region. FcɛRIIa/CD23a is believed to constitutively express only in normal B cells and B-cell lines, whereas FcɛRIIa/CD23b is induced in various cell types. The levels of CD23 expression on blood lymphocytes, alveolar macrophages and airway smooth muscle cells, as well as soluble CD23 levels in serum have been shown to be elevated in allergic diseases. Immunotherapy of allergic diseases and systemic treatment significantly reduce its expression.12, 13, 14, 15, 16, 17 In CD23-deficient transgenic mice, allergic airway hyper-responsiveness was significantly reduced in comparison to wild-type mice. Anti-CD23 antibodies also can reduce allergic airway inflammation including IgE levels and eosinophilia, and normalize allergen-stimulated airway responsiveness in wild-type sensitized mice.18, 19

Recent studies also indicate that genetic polymorphisms of CD23 as well as its expression levels are important factors for the occurrence, severity and even treatment in allergic diseases. A significant allele and haplotype association of CD23 with the phenotypes in allergic diseases has been observed in a genetic epidemiological study.20 CD23 single nucleotide polymorphisms (SNPs) have recently been associated with a significantly increased severe exacerbation in children with asthma by Tantisira et al.21 Given that CD23 can mediate allergen presentation through IgE, it was hypothesized that the SNPs of CD23 may impact allergen-induced T-cell responses, possibly accounting for some associations of CD23 genetic polymorphisms with allergic diseases. In an epidemiological study, we recruited 42 human subjects, including 22 with allergy or asthma and 20 as healthy controls. Using a PCR-linked restriction fragment length polymorphism (RFLP) method, a coding region SNP (R62W) was identified at allele 3750 in exon 4 of CD23 in which C is replaced by T (rs228137), which is expected to result in an amino-acid change from Arg to Trp at position 62 (NP_001993) close to the root of the N-linked carbohydrate chain on CD23 (Figure 1). T-cell proliferation in response to tetanus toxoid antigen was significantly higher in allergic subjects with CT heterozygous polymorphism allergic subjects than those with CC wild type.21 Recently, Tantisira et al.21 found the T allele at this SNP is associated with an increased risk of asthma exacerbation (odd ratio (OR)=1.8). The T allele seems not to be a rare event (allele frequency=0.15 in our study population) and the proportion of patients with the heterozygous polymorphism was 26.2%. Owing to the possible variations among different populations, the frequency of T allele can even reach up to 0.25 in a report by WIAF-CSNP: WIAF-CSNP-MITOGPOP5. However, no direct evidence has been obtained concerning the functional consequences derived from this genetic polymorphism. In order to understand the mechanisms involved in the polymorphism-associated enhancement of T-cell response to antigen, we explored the functional variation caused by this cSNP in CD23 in vitro. Using in vitro transient transfection systems and site-directed mutagenesis techniques, we observed that the polymorphism R62W could result in resistance of CD23 to enzymatic cleavage.

Figure 1

Location of polymorphism R62W (rs2228137). Gene CD23 (IgE low-affinity receptor) is located on chromosome 19p13.3. Mutation is found at loci 3750 in genome (NC_000019). Mutation from C to T at 362 in mRNA coding sequence (NM_002002) results in amino-acid change from arginine to tryptophan at position 62 (NP_001993).


We cloned CD23(a) cDNA into pCMVScript vectors forming pCMVScript-CD23a-C with allele C at position 362 as described in the Materials and methods section. A site-directed mutagenesis system was then utilized to produce plasmid pCMVScript-CD23a-T with a T allele instead of C. The insertion of CD23 in the right direction with content C or T allele was confirmed through endonuclease enzyme digestion and DNA sequencing approaches (data not shown).

Different combinations of plasmids pCMVScript-CD23a-C and pCMVScript-CD23a-T were transiently transfected into Cos-7 cells to simulate three distinct genotypes observed in vivo. These include 2 μg of pCMVScript-CD23a-C for wild type (CC), 1 μg of each pCMVScript-CD23a-C and pCMVScript-CD23a-T for heterozygous polymorphism (CT) and 2 μg of pCMVScript-CD23a-T for homozygous polymorphism (TT). The pCMVScript vector was used as a negative control and Epstein–Bar virus (EBV) transformed B cells were used as a positive control. The equivalence in transcription efficiency and gene expression levels among transfection groups was examined using real-time PCR and co-transfection with plasmid pAAV-lacZ for β-galactosidase (provided by Dr Hongbin Su's lab). No significant differences for β-galactosidase DNA copies ( × 106/15 ng total RNA) were observed among the negative control (136.6±32.0), CC (132.2±38.3), CT (152.7±53.2) and TT (120.1±41.4) groups 48 h after transfection, suggesting equivalence in transfection efficiency. There was also no significant difference in gene transcription as evidenced by the observation that CD23 DNA copies ( × 106/15 ng total RNA) did not exhibit any significant difference between CC (96.6±23), CT (71.0±29.4) and TT (75.1±20.1) groups (ANOVA, P>0.05; Tukey test, P>0.05). Similarly, there were no significant differences in terms of mean fluorescence intensity (MFI) among CC (4.81±0.44), CT (4.62±0.83) and TT (4.64±0.79) groups (ANOVA, P>0.05; Tukey test, P>0.05) using flowcytometry, indicating that different combinations of transfection did not result in any significant differences in CD23 expression in our transient transfection system.

Reduction of soluble CD23 release after C to T mutation

Given that soluble CD23 could be released from membrane CD23 by various catalytic enzymes, its concentrations in the supernatant were examined 48 h after transfection. Surprisingly, there was a significantly higher level of soluble CD23 in CC (10.30±2.31 ng/ml) than in CT transfected cells (6.36+2.73 ng/ml) (Student's t-test, P=0.0393<0.05) (see Figure 2), whereas soluble CD23 is undetectable in TT-transfected cells. No soluble CD23 was detectable in the supernatant of vector control cells, although about 210 ng/ml of soluble CD23 (210.26±12.93 ng/ml) was detected in EBV transformed B cells as anticipated. These results suggest the possibility that the cSNP R62W may reduce the release of soluble CD23 from the transfected cells and induce the resistance of membrane CD23 to enzymatic cleavage.

Figure 2

Differences in soluble CD23 released from transiently transfected Cos-7 cells with differently simulated genotypes. After 48 h of transient transfection with pCMVScript vector (control), pCMVScript-CD23a-C only (CC), 1:1 mixture of pCMVScript-CD23a-T and pCMVScript-CD23a-C (CT) and pCMVSCript-CD23a-T only (TT), soluble CD23 in the supernatants were measured by ELISA kit. Bars represent mean and standard deviation of soluble CD23 concentration in four independent replicate experiments. Student's t-test between CC and CT (P=0.039*).

Resistance to cleavage of CD23 by Der p I, human leukocyte elastase and cathepsin G after C to T mutation

To evaluate the aforementioned possibility, the transiently transfected Cos-7 cells were treated with several enzymes that have been reported to cleave CD23, and their capacities to cleave membrane CD23 in CC, CT and TT groups were compared.

Der p I, a cysteine protease and a major allergen of the house dust mite Dermatophagoides pteronyssinus, has been reported to cleave CD23 from the surface of cultured human B cells.22 At 48 h after transfection, Cos-7 cells were cultured in the presence or absence of Der p I (50 μg/ml, 1 h). Der p I treatment significantly reduced CD23 MFI in CC and CT transfected Cos-7 cells (Figure 3A and Table 1) although it was less efficient in comparison to EBV B cells (about 90% cleaved). The MFI for CD23 was reduced about 30% by Der p I for CC (71.28±4.32% left) in contrast to 8% for TT (92.73±6.12% left) (see Table 1; ANOVA, P<0.0001; Tukey test, P=0.0011<0.05). There were also differences among the CC, CT and TT groups in terms of Der p I catalytic potential. A higher level of MFI for CD23 was observed in TT than in CC and CT after treatment with Der p I (Figure 3A; ANOVA, P=0.0009<0.001; Tukey test, TT vs CC, P=0.0034<0.05 and TT vs CT, P=0.038<0.05), although no significant difference in membrane CD23 among the CC, CT and TT groups was shown in the absence of Der p I (P=0.4456>0.05). These results support the possibility that the cSNP R62W within CD23 results in the resistance of CD23 to Der p I in terms of cleavage capacity and has a stabilized membrane CD23.

Figure 3

Differences on the capacity of Der p I to cleave membrane CD23 among different simulated genotypes. (A) After 48 h of transient transfection in Cos-7 cells with pCMVScript vector (control), pCMVScript-CD23a-C only (CC), 1:1 mixture of pCMVScript-CD23a-T and pCMVScript-CD23a-C (CT) and pCMVSCript-CD23a-T only (TT), cells were incubated with Der p I at 50 μg/ml for CC (CC+Der p 1), CT (CT+Der p I), TT (TT+Der p I) and EBV B cells (EBV+Der p I) or Der p I free media (CC, CT or TT) for 1 h at 37°C. Bars represent mean and standard deviation of membrane CD23 MFI in four replicate experiments. ANOVA test among CC, CT and TT: P=0.4456>0.05. ANOVA test among CC+Der p I, CT+Der p I and TT+Der p I: P=0.0009<0.001. Tukey's adjusted post hoc multiple comparisons: TT+Der p I vs either CT+Der p I (P=0.038 <0.05) or CC+Der p 1 (P=0.0034<0.01**). (B) Similar treatment scheme was applied to Jurkat cells. After 48 h of transient transfection with pCMVScript vector (a), pCMVScript-CD23a-C only (b), 1:1 mixture of pCMVScript-CD23a-T and pCMVScript-CD23a-C (c), and pCMVSCript-CD23a-T only (d), cells were incubated without (- - - dashed line) or with Der p I at 50 μg/ml (—— solid line) for 1 h at 37°C as displayed in histogram.

Table 1 Resistance of CD23 to enzymatic cleavage on Cos-7 cells which were transiently transfected with different plasmids

The resistance of the cSNP R62W within CD23 to enzymatic cleavage was also observed in Jurkat cells. After Jurkat cells were transfected and incubated with Der p I in the same experimental procedures as described above, a lower level of MFI for CD23 in Der p I-treated CD23 positive cells was shown (Figure 3B). A greater decrease in MFI for CD23 was noticed in CC (b) than in CT (c) and TT (d). Thus, the results from the experiments utilizing Jurkat cells further support the possibility that the cSNP R62W enhanced the stability of surface-expressed CD23.

Human leukocyte elastase (HLE) and cathepsin G are polymorphonuclear neutrophil-derived serine proteases. They have been reported to cleave membrane CD23 and account for the release of soluble CD23 fragments.23 Thus, we compared their capacity to cleave membrane CD23 in transiently transfected Cos-7 cells among different genotypes. At 48 h after transfection, Cos-7 cells transfected with CC, CT and TT as well as EBV transformed B cells were cultured in the presence or absence of HLE (200 μg/ml, 1 h) and cathepsin G (25 μg/ml, 1 h). Similar to the results seen with Der P I, the MFI for CD23 was diminished significantly for EBV cells and for CC and CT transfected Cos-7 cells by HLE and cathepsin G (Table 1). TT-transfected cells seem to be more resistant to HLE and cathepsin G than CC, which is evidenced by no significant differences for MFI compared to the control (Table 1). These results are consistent with the hypothesis that the cSNP R62W renders CD23 more resistant to cleavage by HLE, cathepsin G and Der p I and results in a more stabilized membrane CD23.

Capacity of ADAM33 to cleave CD23 and the resistance to cleavage after C to T mutation

ADAM33 is a gene encoding a metalloproteinase enzyme whose genetic polymorphisms have been associated with asthma and respiratory hyper-responsiveness. To determine whether ADAM33 exhibits catalytic capacity towards CD23, different mixtures of pScript-CD23CC with pcDNA3.1-ADAM33 or vector pcDNA3.1 were transiently transfected into FreeStyle 293-F suspension cells. From Figure 4a, MFI for CD23 was significantly lower in CC plus ADAM33 than CC plus pcDNA3.1 transfectants, indicating that the membrane CD23 level is reduced by ADAM33. The reduction of CD23 MFI was notably correlated with an amount of ADAM33 cotransfected, when 0.5 μg of plasmid CD23 CC was cotransfected with different amounts of pcDNA3.1 or pcDNA3.1-ADAM33 ranging from 0.25 to 2.0 μg. The proportions of CD23 MFI for the different amounts of ADAM33 vector over the same quantity of pcDNA3.1 vector showed a dose-dependent decrease with higher levels of ADAM33 transfected (Spearman correlation coefficient r=−0.9417, P<0.0001, see Figure 4b). Meanwhile, the proportions of soluble CD23 in the supernatants of the ADAM33 group over vector control group were increased with higher levels of ADAM33 groups seen in a dose response manner (Spearman correlation coefficient r=−0.8605, P<0.0001, see Figure 4c). These results clearly suggest that ADAM33 exhibits catalytic activity towards CD23 molecules.

Figure 4

Capacity of ADAM33 to cleave CD23 CC. (a) Reduction of membrane CD23 CC MFI by ADAM33. One microgram of pScript-CD23 CC was transfected into FreeStyle 293-F cells alone (CC), or together with 1 μg of vector control (CC+pcDNA3.1), or 1 μg of pcDNA3.1-ADAM33 (CC+ADAM33) as well as no transfection (control), 1 μg of pcDNA3.1 only (pcDNA), or pcDNA3.1-ADAM33 (ADAM33). Bars represent mean and standard deviation of membrane CD23 MFI at 48 h in four independent replicate experiments. ANOVA test among CC, CC+pcDNA and CC+ADAM33 (P<0.0001). Post hoc Tukey adjusted multiple comparisons: **CC+pcDNA vs CC+ADAM33 (P=0.0011<0.01). (b) Reduction of CD23 CC MFI by ADAM33 in a dose–response manner. The pcScript-CD23 CC (0.5 μg) was either co-transfected with pcDNA3.1 or pcDNA3.1-ADAM33 (0, 0.25, 0.5, 1.0 or 2.0 μg). MFI was measured at 48 h and the ratio of CD23 MFI for ADAM33 over same quantity of pcDNA3.1 vector only was displayed. Spearman correlation coefficient between ratio and amounts of ADAM33, r=−0.94171, P<0.0001. (c) Increase of soluble CD23 by ADAM33 in a dose–response manner. Using similar transfection scheme in FreeStyle 293-F cells, the ratio of soluble CD23 in supernatant of CD23 CC plus ADAM33 over CD23 plus same quantity of pcDNA3.1 vector was displayed. Spearman correlation coefficient between ratio and amounts of ADAM33, r=0.8605, P<0.0001. The statistical analysis was based on four replicate experiments.

To examine whether there is a difference in the effect of ADAM33 on CD23 CC and TT, transfection of pScript-CD23 TT in Cos-7 cells was performed. There was a significantly higher level of soluble CD23 in the supernatant of the CD23 CC plus ADAM33 group than in that of CD23 CC plus vector control group, and a similar observation was seen in transfected FreeStyle 293-F cells (21.95±0.09% higher, P=0.0192<0.05). No soluble CD23 could be detected in either CD23 TT plus ADAM33 or CD23 TT plus vector control. The addition of ADAM33 did not enhance the release of CD23 from membrane in TT-transfected cultures, suggesting that the cSNP R62W within CD23 leads to a resistance of CD23 to the enzymatic effects of ADAM33.

N-glycosylation may account for the induced resistance

As C to T mutation at allele 362 is assumed to cause the amino-acid change at residue 62 (R62W) close to the N-glycosylation site, it is hypothesized that this amino-acid change from arginine to tryptophan may affect N-glycosylation and thereby mediate the enhanced resistance. The resistance of CD23 to enzymatic cleavage was tested in the presence of an N-glycosylation inhibitor. Tunicamycin is an inhibitor of the N-glycosylation of glycoproteins that blocks the first step in the biosynthesis of the lipid-linked oligosaccharide precursor.24 Transiently transfected Cos-7 cells were incubated with tunicamycin at 0.8 μg/ml for 48 h to inhibit N-glycosylation. The cells were subsequently incubated with Der p I (50 μg/ml, 1 h). The approximately 30% (71.28±4.32% left) cleavage by Der p I in the CC group was decreased to about 10% (90.4±3.98% left) as a result of pretreatment with tunicamycin, which was a statistically significant difference (P=0.0039<0.01) (see Figure 5). The reduction of soluble CD23 in the supernatant is concordant with this finding (data not shown). These results indicate that inhibition of N-glycosylation by tunicamycin results in resistance to enzymatic cleavage in CC transfected cultures. At the same time, tunicamycin did not change the resistance of TT to Der p I cleavage significantly. The difference in catalytic potential of Der p I toward membrane CD23 CC and TT was diminished after pretreatment with tunicamycin (see Figure 5). This suggests that the amino-acid change from arginine to tryptophan, as a result of mutation from C to T at allele 362, may influence the N-glycosylation reaction accounting for the observed resistance to enzymatic cleavage.

Figure 5

Potential role of N-glycosylation in the resistance of CD23 to Der pI cleavage. Cos-7 cells were transiently transfected with pCMVScript vector (control), pCMVScript-CD23a-C only (CC), 1:1 mixture of pCMVScript-CD23a-T and pCMVScript-CD23a-C (CT), and pCMVSCript-CD23a-T only (TT) and treated with (+tunicamycin) or without tunicamycin for 48 h, and then incubated with Der p I (50 μg/ml) or Der p I free media for 1 h at 37°C. Bar and error bar represent mean and standard deviation of percentage of membrane CD23 left by Der p I over Der p I free media. ANOVA Test for all eight groups (P<0.0001). Tukey post hoc adjusted multiple comparisons: minus vs plus tunicamycin (P=0.0039<0.01**), or CC vs TT in the absence of tunicamycin (P=0.0013<0.01**), CC vs TT in the presence of tunicamycin (P=0.9381>0.05).


The fact that CD23 mediates IgE-specific antigen presentation has been extensively studied. CD23 could enhance IgE antigen complex presentation 100-fold more than either unfocused or IgG bound antigen in animal studies.25, 26, 27 The proposed roles for CD23 in antigen presentation include the binding of antigen-IgE–antibody complexes, internalization of the complexes, transport to compartments of the endosomal network containing proteolytic enzymes, interaction with major histocompatibility complex (MHC) class II antigens and interaction with CD21 at the points of contact on the B- and T-cell surfaces.28, 29 An integrated CD23 is believed to be important for the IgE-mediated antigen presentation. A stabilized CD23 on the antigen presenting cells may result in enhanced T-cell proliferation in response to antigen. Our present results indicate that the cSNP R62W within CD23 may mediate the resistance of CD23 to enzymatic cleavage, and therefore stabilizes CD23 on cell membranes. This may result in higher T-cell proliferation in response to antigen as observed in our genetic epidemiological study.

Our epidemiological study also showed that the significant difference in antigen presentation between CT and CC occurs in allergic subjects rather than in healthy control subjects although the increased trend was observed for CT in the healthy controls too. Allergic individuals may have higher levels of IgE that could mediate allergen reactions related to CD23 genetic variations. Our present study clearly indicated that various endogenous enzymes can destabilize CD23 molecules and the T variant induces the resistance of CD23 to these enzymes. Although the resistance of CD23 to each individual enzyme by the T variant is small, the combined effects of resistance to multiple enzymes may result in a meaningful change in cell activation and allergic risk. Allergic inflammation is a complicated immunological process that requires diverse cells to aggregate and be activated. Numerous endogenous enzymes such as HLE, cathespin G and perhaps others could be produced from mononuclear cells and could affect this process.23 Therefore, the cSNP R62W may contribute to immune mediated dysfunction in allergic reactions.

The ADAM (a disintegrin and metalloproteinase) gene family is a subgroup of the zinc-dependent metalloproteinase superfamily, which mediates extracellular matrix remodeling during discrete physiological processes.31 A major function of active ADAMs is the proteolytic release of cell-surface membrane proteins such as cytokines, growth factors and receptors.31 Several membrane-bound matrix metalloproteases such as ADAM8, ADAM15 and ADAM28 have been found to catalyze ectodomain shedding of CD23.30 ADAM33 is a newly discovered gene that has been associated with asthma and bronchial hyper-responsiveness in a genetic-linkage analysis.31 However, it is still unknown what the physiological functions and the roles of ADAM33 are in the pathophysiology of asthma. The full-length ADAM33 alternatively signalized message may affect lung cell development. ADAM33 is primarily expressed in lung fibroblasts and bronchial smooth muscle cells,30 and is hypothesized to be involved in bronchial tissue remodeling in asthmatics.32 ADAM33 has been reported to cleave several substrates and mediate the production of growth factors and cytokines from the cell surface for stimulating lung cells. The proteolytic reactions may involve part of the substrate and the catalytic domain of the ADAMs, or between the juxtamembrane region of the substrate and both the active site and the disintegrin domain (or with the cysteine-rich domain) of the ADAM. The substrate and the disintegrin domain (or with cysteine-rich domain) of ADAM on the cytoplasmic membrane33, 34, 35 may also be involved. Our results suggest that ADAM33 could catalyze the shedding of soluble CD23, implying that the interaction of CD23 with ADAM33 may be important in allergic asthmatic pathogenesis. The cSNP R62W within CD23 may impact allergic reactions in the airway of asthmatic patients due to its resistance to ADAM33. The consequences may be further complicated owing to the polymorphisms of ADAM33, which cause the loss of its catalytic activity.

In addition to the above-discussed endogenous enzymes, many allergens themselves have an enzymatic activity that influences the shedding of surface molecules. As an example, Der p I, a cysteine protease representing a major allergen of the house dust mite D. pteronyssinus, has been shown to cleave CD23 from the membrane of cultured human B cells.22, 36, 37 It is suggested that Der p I might enhance its potent allergenicity by disrupting the IgE negative feedback network via selectively cleaving CD23 and CD25.22 The proteolytic activity of Der p I toward CD23 was confirmed in transfected Cos-7 cells and mutation to T from C caused the resistance of CD23 to Der p I in this study. These suggest that patients with the T allele at R62W might be less susceptible to Der p I in terms of the IgE-mediated reaction.

Located at asparagine 63 of CD23 is an N-glycosylation site for N-linked sugar chains. N-glycosylation has been suggested to have the potential to mediate a variety of functions for CD23. However, reports about whether N-glycosylation determines the resistance of CD23 to enzymatic cleavage are contradictory. Sarfati et al.38 reported that tunicamycin did not affect any IgE binding factors' production, which were assumed to be soluble CD23 in RPMI 8866 cells, but the production of soluble CD23 was observed to be increased by tunicamycin from the same group suggesting that N-glycosylation may protect CD23 from degradation into soluble CD23.39 In our present study, we observed that the inhibition of N-glycosylation actually protects CD23 from the degradation by Der p I. The differences in experimental conditions or cell systems adopted may have contributed to the contradictory results, because we noticed that CD23 on EBV B cell surface is more susceptible to enzymatic cleavage than on Cos-7 cells in our experiments. With respect to the potential influences of nonsynonymous mutation on protein structure and function, the substitution of arginine and tryptophan may cause a change in structure or function possibly due to their dissimilar physical and chemical properties. The association of R/W SNP with various genetic related diseases has been well documented in the literature.40 As R62W occurs at the root of N-glycosylation in CD23, we hypothesized that the functional consequences derived from R62W are associated with a change in N-glycosylation. Our results using tunicamycin suggest that the resistance of CD23 TT to enzymes may be an outcome of the inhibition of N-glycosylation.

Carbohydrate residues related to N-glycosylation have been suggested to be involved in IgE binding, cell interactions and the formation of α-helical coiled-coil stalks that mediate the formation of trimers of CD23 as well as the interaction of CD23 with MHCII class. Fucose-1-phosphate has been described as a competitive inhibitor of IgE binding to CD23 suggesting a possible role of glycosylation in the binding of IgE to CD23.9, 39 Leucine zipper sequences, located between the N-glycosylation site and the lectin domain, have been suggested as one of the structural features that make the α-helical coiled-coil stalk. This α-helical coiled-coil stalk has similar functions as an S–S bridge, and an even more convenient means of polymerization. Clustered receptors are expected to exhibit higher affinity to the carbohydrate chains and may enhance cell–cell interactions.41 Another important structural feature of CD23 is the RGD-binding inhibitory peptide spanning from amino-acid position 58 to 69 around the N-glycosylation site.42, 43 This inhibitory sequence has been suggested to prevent the reverse RGD sequence near the extracellular terminus of the CD23 molecule from binding to MHC class II molecules.43 As the binding of CD23 to MHC II has been observed to be involved in CD23-mediated antigen presentation, it is quite possible that loss of N-glycosylation may lead to functional changes in the binding affinity of CD23 to MHC II and result in enhanced IgE-antigen complex presentation.28 The genetic polymorphism R62W is expected to affect these various functions via the involvement of N-glycosylation. A further clarification of association of the genetic polymorphism R62W with IgE binding affinity and its interaction with MHCII will offer other potential mechanisms to understand the significance of R62W in specific immunological and allergic reactions.

Materials and methods


The pCMV-Script XR predigested vector for cloning CD23 was purchased from Stratagene (LaJolla, CA, USA). Cloned full ADAM33 cDNA fragment in pcDNA3.1 plasmid was kindly provided by Dr Shelby Umland (Schering Plough Research Institute, Kenilworth, NJ, USA). HLE and cathepsin G were purchased from Sigma Aldrich (St Louis, MO, USA). Der p I was ordered from Indoor Biotechnologies, Ltd (Charlottesville, VA, USA). All primers for CD23 real-time PCR and reverse transcriptase PCR were made by National Jewish Molecular Resource Center. The culture medium was obtained from Life Technologies-GIBCOL (Grand Island, NY, USA). The sources of various kit, reagents and facilities in the methods section are listed below.


Cloning of CD23a

Total RNA was isolated from the EBV-transformed B cells of a subject with wild genotype CC at 362 allele (provided by Dr James Jones) using RNeasy mini kit according to the instruction of the manufacturer (Qiagen Inc., Valencia, CA, USA). Total RNA collected was examined for purity, quantified by spectrophotometer (DU-640) (Beckman, Fullerton, CA, USA) and used for the subsequent reverse transcriptase PCR.

CD23a cDNA at region 148–1260 (NM_002002) was derived from the above RNA using reverse transcription system from Promega (Madison, WI, USA) and then amplified by PCR. The primers' sequences are as follows: forward primer – IndexTermCGGAATTCCGGTGAGTGCTCCATCATCGGG and reverse primer – IndexTermCCGCTCGAGCGGCGTTTGGGTGGCAGAAAATG. The produced PCR fragment of CD23a with 1113 bp length was purified using the QIAquick PCR purification kit (QIAGEN inc., Valencia, CA, USA).

The purified CD23a PCR fragment was digested by enzymes EcoR I and Xho I producing 3′ and 5′ protruding termini. Then this restriction fragment was ligated with pCMV-Script XR predigested vector, which owns 5′ EcoR I and 3′ Xho I restriction sites (Stratagene, Lajolla, CA, USA) overnight at 4°C under T4 ligase. The ligated vector was screened based on kanamycin resistance and amplified in Escherichia coli, and then purified using QIAprep Miniprep kit (Qiagen Inc., Valencia, CA, USA). The purified plasmids were digested by molecular endonucleases Ase I and Bgl II to verify the insertion of CD23a PCR fragment into pCMVScript vector correctly and in the right direction based on the expected DNA sequence in the constructed plasmid pCMVScript-CD23a-C. The plasmid that produced two DNA fragments with length 1.1 and 4.26 kb was considered as pCMVScript-CD23a-C (data not shown). DNA sequencing with forward primer in the T3 promoter and reverse primer in the T7 promoter regions was performed using a DNA sequencing kit with the Big Dye Terminator Cycle sequencing ready reaction (Applied Biosystems, Foster City, CA, USA) to verify that the DNA sequence at position 362 of CD23a in the plasmid is C, and that the insertion of CD23a cDNA fragment was complete in the right direction.

Site-directed mutagenesis

To produce plasmid pCMV-Script-CD23a-T, which contains T instead of C at position 362 (NM_002002), the QuikChange XL Directed Mutagenesis kit (Stratagene, LaJolla, CA, USA) was used for a point mutation on pCMV-Script-CD23a-C. As instructed by the manufacturer, opposite primers were designed as Oligo primer 1: IndexTermGAAGAGAGGGCTGCCTGGAACGTCTCTCAAG and Oligo primer 2: IndexTermCTTGAGAGACGTTCCAGGCAGCCCTCTCTTC. The primers and pCMVScript-CD23a-C were added together with Turbo DNA polymerase for the new DNA strand synthesis. After digestion by Dpn I to get rid of methylated or hemimethylated plasmid DNA, the mutated clone was screened for kanamycin resistance. Verification for plasmids containing CD23a cDNA was performed by molecular endonucleases Ase I and Bgl II digestion. DNA sequencing with forward primer in the T3 promoter as described above was also used to confirm the mutation to T from C at position 362.

Transfection of plasmids

The transfection of plasmids in Cos-7 cells was performed using LIPOFECTAMINE PLUS Reagent from Life Technologies Inc. (Rockville, MD, USA). Cos-7 cells (3 × 105) were cultured in six-well culture plates 1 day before transfection. Specific amounts of plasmid DNA were mixed with Plus reagent, and then incubated with lipofectamine for complex formation. Subsequently, the DNA-plus-lipofectamine reagent complexes were added to Cos-7 cell culture. After 48-h incubation, Cos-7 cells were harvested by PBS/EDTA (0.1 mM) and used for subsequent analysis of CD23 expression.

An electroporation protocol was utilized to transfect plasmid DNA into Jurkat cells. In brief, Jurkat cells at late logphase were harvested and washed with PBS, and then resuspended in antibiotic-free RPMI 1640 medium, transferred to electroporation cuvettes and mixed with plasmid DNA. The cells in cuvettes were shocked two times at voltage 800 μF, low resistance, fast charge, 240 V in cell porator (GIBCO-BRL Life Technologies, Gaitherburg, MD, USA). Finally, the transfected cells were diluted 20-fold in antibiotic-free complete RPMI 1640 medium and cultured with PHA (1 μg/ml) and PMA (50 ng/ml) until expression evaluation was performed.

FreeStyle 293-F suspension cells were cultured in FreeStyle 293 expression medium and transfection was carried out according to the protocol of Invitrogen Inc. (Carlsbad, CA, USA). DNA–lipid complexes were prepared by mixing plasmid DNA with 293 fectin in Opti-MEM I medium and incubating for 20–30 min. Then the DNA–lipid complexes were added into cell suspension cultures. The transfected cells were incubated at 37°C on an orbital shaker until the determination of cell-surface CD23 and soluble CD23 in cell-culture supernatant was performed.


Flowcytometry was used to evaluate membrane CD23 expression on diverse cell types. Target cells were harvested in PBS/EDTA (0.1 mM) solution, washed with Hank's buffer solution and then resuspended in FACS buffer (1 × PBS, 2.0% calf serum, 0.1% BSA), which contains PE (phycoerythrin)-conjugated CD23 mouse anti-human monoclonal antibody (BD Pharmingen Bioscience, San Diego, CA, USA). After 30 min in the dark, cells were washed and resuspended in FACS buffer. The cell harvesting and data acquisition were accomplished using BD FACScan flowcytometers linked with CellQuest software (BD, San Jose, CA, USA). CD23-PE fluorescence intensity of 1 × 104 cells was counted for each sample as the expression level on the cell membrane. A geometric value of MFI was used due to less variation on repeated experiments.

Real-time PCR

Real-time PCR was developed to detect the mRNA level of CD23 and β-galactosidase by applying TaqMan universal PCR mix (AB Applied Biosystems, Foster City, CA, USA). Initially, total RNA was extracted using the RNeasy mini kit (Qiagen Inc., Valencia, CA, USA) and cDNA was synthesized with the reverse transcription system (Promega, Madison, WI, USA) as described above. For CD23, the forward primer (IndexTermGGTGACCAGATG GCGCA) and reverse primer (IndexTermTTCAGCTCGAAGTTCCTCCAGT) at 300 nM and the TaqMan probe (IndexTermATCCCAGTCCACGCAGATTTCACAGG) at 250 nM were added in a 50 μl total reaction volume. For β-galactosidase, the forward primer (IndexTermAACCCG AAACTGTGGAGCG) and the reverse primer (IndexTermCGTCGGCGGTGTGCA) at 300 nM and the TaqMan probe (IndexTermATCCCGAATCTCTATCGTGCGGTGGT) at 250 nM were added in a 50 μl total reaction volume. The fluorescent reporter dye FAM (6-carboxyfluoresceine) linked to 5′ end of the probe was quenched by Black hole quencher (Biosearch Technologies Inc., Novato, CA, USA) linked to 3′ end. Real-time PCR for CD23 and β-galactosidase was performed separately for each sample using ABI PRISM 7700 sequence detection system (AB Applied Biosystems, Foster City, CA, USA) and threshold cycle (CT) was captured. Standard curves of threshold cycles with the amounts of plasmid pCMVScript-CD23a-C at 0, 0.005, 0.05 and 0.5 ng or with pAAV-lacZ at 0.5, 1.0 and 5.0 ng were produced for each experiment so that DNA quantization was normalized between experiments. The exact DNA copies were calculated from the following formula: DNA copies=(DNA amounts (ng) × 6.02)/(0.66 × number of base pairs for plasmid) ( × 1011).

ELISA for soluble CD23

The culture supernatant was collected and evaluated for soluble CD23 concentration using the BD OptEIA human CD23 Elisa kit (BD Pharmingen Bioscience, San Diego, CA, USA). As instructed, standard soluble CD23 and samples were incubated in coated wells. The working detector solution containing CD23 antibody linked to avidin–horseradish peroxidase was added to each well and incubated for 1 h. After repeatedly washing, tetramethylbenzidine (TMB) substrate solution was added and incubated for 30 min in the dark for the color reaction. Finally, the stop solution was added to each well before the absorbance at 450 nm was read using a Microplate Autoreader EL311 (Bio-Tek instruments Inc., Winooski, VT, USA). The soluble CD23 concentration for each sample was derived using a standard curve for soluble CD23.


Statistical analyses were performed using SAS 9.13. Mean and standard deviation of values in replicate experiments were calculated in PROC TTEST and PROC GLM. To compare the difference between two groups, the equality of variance was first examined by folded F-test and then the difference was tested by Student's t-test given equal variance existed. Otherwise, the Satterthwaite method was applied. To compare the differences among more than two groups, the homogeneity of variance was examined by Levene's test and ANOVA was applied for the significance test. Tukey's adjustment was used for multiple comparisons to avoid inflated type I error. Significance level was set at 0.05. Spearman correlation coefficients were estimated for the correlation of two independent variables to evaluate a dose-dependent response using PROC CORR.


  1. 1

    Pirron U, Schlunck T, Prinz JC, Rieber EP . IgE-dependent antigen focusing by human B lymphocytes is mediated by the low-affinity receptor for IgE. Eur J Immunol 1990; 20: 1547–1551.

  2. 2

    Flores-Romo L, Johnson GD, Ghaderi AA, Stanworth DR, Veronesi A, Gordon J . Functional implication for the topographical relationship between MHC class II and the low-affinity IgE receptor: occupancy of CD23 prevents B lymphocytes from stimulating allogeneic mixed lymphocyte responses. Eur J Immunol 1990; 20: 2465–2469.

  3. 3

    Sarfati M, Delespesse G . Possible role of human lymphocyte receptor for IgE (CD23) or its soluble fragments in the in vitro synthesis of human IgE. J Immunol 1988; 141: 2195–2199.

  4. 4

    Delespesse G, Hofstetter H, Sarfati M . Low-affinity receptor for IgE (FcɛRII, CD23) and its soluble fragments. Int Arch Allergy Appl Immunol 1989; 90: 41–44.

  5. 5

    Gordon J . B-cell signalling via the C-type lectins CD23 and CD72. Immunol Today 1994; 15: 411–417.

  6. 6

    Gordon J, Cairns JA, Millsum MJ, Gillis S, Guy GR . Interleukin 4 and soluble CD23 as progression factors for human B lymphocytes: analysis of their interactions with agonists of the phosphoinositide ‘dual pathway’ of signalling. Eur J Immunol 1988; 18: 1561–1565.

  7. 7

    Dugas B, Mossalayi MD, Damais C, Kolb JP . Nitric oxide production by human monocytes: evidence for a role of CD23. Immunol Today 1995; 16: 574–580.

  8. 8

    Lecoanet-Henchoz S, Gauchat JF, Aubry JP, Graber P, Life P, Paul-Eugene N et al. CD23 regulates monocyte activation through a novel interaction with the adhesion molecules CD11b-CD18 and CD11c-CD18. Immunity 1995; 3: 119–125.

  9. 9

    Delespesse G, Sarfati M, Wu CY, Fournier S, Letellier M . The low-affinity receptor for IgE. Immunol Rev 1992; 125: 77–97.

  10. 10

    Nakajima T, Sarfati M, Delespesse G . Relationship between human IgE-binding factors (IgE-BF) and lymphocyte receptors for IgE. J Immunol 1987; 139: 848–854.

  11. 11

    Sarfati M, Nakajima T, Frost H, Kilccherr E, Delespesse G . Purification and partial biochemical characterization of IgE-binding factors secreted by a human B lymphoblastoid cell line. Immunology 1987; 60: 539–545.

  12. 12

    Prinz JC, Endres N, Rank G, Ring J, Rieber EP . Expression of Fc epsilon receptors on activated human T lymphocytes. Eur J Immunol 1987; 17: 757–761.

  13. 13

    Prinz JC, Rieber EP, Ring J . Loss of FceR2/CD23-expression on T and B lymphocytes during rush hyposensitization. In: Ring J, Braun-Falco O, Przybilla B (eds). New Trends in Allergy. Springer: Berlin, 1991, pp 105–108.

  14. 14

    Jung CM, Prinz JC, Rieber EP, Ring J . A reduction in allergen-induced Fc epsilon R2/CD23 expression on peripheral B cells correlates with successful hyposensitization in grass pollinosis. J Allergy Clin Immunol 1995; 95: 77–87.

  15. 15

    Roever AC, Henz BM, Worm M . Wasp venom rush immunotherapy induces transient downregulation of B cell surface molecule expression. Int Arch Allergy Immunol 2002; 127: 226–233.

  16. 16

    Khan S, McClellan JS, Knutsen AP . Increased sensitivity to IL-4 in patients with allergic bronchopulmonary aspergillosis. Int Arch Allergy Immunol 2000; 123: 319–326.

  17. 17

    Hakonarson H, Carter C, Kim C, Grunstein MM . Altered expression and action of the low-affinity IgE receptor FcepsilonRII (CD23) in asthmatic airway smooth muscle. J Allergy Clin Immunol 1999; 104: 575–584.

  18. 18

    Dasic G, Juillard P, Graber P, Herren S, Angell T, Knowles R et al. Critical role of CD23 in allergen-induced bronchoconstriction in a murine model of allergic asthma. Eur J Immunol 1999; 29: 2957–2967.

  19. 19

    Haczku A, Takeda K, Hamelmann E, Loader J, Joetham A, Redai I et al. CD23 exhibits negative regulatory effects on allergic sensitization and airway hyperresponsiveness. Am J Respir Crit Care Med 2000; 161: 952–960.

  20. 20

    Laitinen T, Ollikainen V, Lazaro C, Kauppi P, de Cid R, Anto JM et al. Association study of the chromosomal region containing the FCER2 gene suggests it has a regulatory role in atopic disorders. Am J Respir Crit Care Med 2000; 161: 700–706.

  21. 21

    Tantisira KG, Siverman ES, Mariani T, Xu J, Richter BG, Klanderman BJ et al. FCER2: a pharmacogenetic basis for severe exacerbations in children with asthma (submitted for publication).

  22. 22

    Hewitt CR, Brown AP, Hart BJ, Pritchard DI . A major house dust mite allergen disrupts the immunoglobulin E network by selectively cleaving CD23: innate protection by antiproteases. J Exp Med 1995; 182: 1537–1544.

  23. 23

    Brignone C, Munoz O, Batoz M, Rouquette-Jazdanian A, Cousin JL . Proteases produced by activated neutrophils are able to release soluble CD23 fragments endowed with proinflammatory effects. FASEB J 2001; 15: 2027–2029.

  24. 24

    Elbein AD, Sanford BA, Ramsay MA, Pan YT . Effect of inhibitors on glycoprotein biosynthesis and bacterial adhesion. Ciba Found Symp 1981; 80: 270–287.

  25. 25

    Kehry MR, Yamashita LC . Low-affinity IgE receptor (CD23) function on mouse B cells: role in IgE-dependent antigen focusing. Proc Natl Acad Sci USA 1989; 86: 7556–7560.

  26. 26

    Heyman B, Tianmin L, Gustavsson S . In vivo enhancement of the specific antibody response via the low-affinity receptor for IgE. Eur J Immunol 1993; 23: 1739–1742.

  27. 27

    Gustavsson S, Hjulstrom S, Liu T, Heyman B . CD23/IgE-mediated regulation of the specific antibody response in vivo. J Immunol 1994; 152: 4793–4800.

  28. 28

    Karagiannis SN, Warrack JK, Jennings KH, Murdock PR, Christie G, Moulder K et al. Endocytosis and recycling of the complex between CD23 and HLA-DR in human B cells. Immunology 2001; 103: 319–331.

  29. 29

    Boltz-Nitulescu G, Wiltschke C, Langer K, Nemet H, Holzinger C, Gessl A et al. Augmentation of IgE receptor expression and IgE receptor-mediated phagocytosis of rat bone marrow-derived macrophages by murine interferons. Immunology 1988; 63: 529–535.

  30. 30

    Fourie AM, Coles F, Moreno V, Karlsson L . Catalytic activity of ADAM8, ADAM15, and MDC-L (ADAM28) on synthetic peptide substrates and in ectodomain cleavage of CD23. J Biol Chem 2003; 278: 30469–30477.

  31. 31

    Van Eerdewegh P, Little RD, Dupuis J, Del Mastro RG, Falls K, Simon J et al. Association of the ADAM33 gene with asthma and bronchial hyperresponsiveness (comment). Nature 2002; 418: 426–430.

  32. 32

    Shapiro SD, Owen CA . ADAM-33 surfaces as an asthma gene. New Engl J Med 2002; 347: 936–938.

  33. 33

    Black RA, White JM . ADAMs: focus on the protease domain. Curr Opin Cell Biol 1998; 10: 654–659.

  34. 34

    Zou J, Zhu F, Liu J, Wang W, Zhang R, Garlisi CG et al. Catalytic activity of human ADAM33. J Biol Chem 2004; 279: 9818–9830.

  35. 35

    Powell RM, Wicks J, Holloway JW, Holgate ST, Davies DE . The splicing and fate of ADAM33 transcripts in primary human airways fibroblasts. Am J Respir Cell Mol Biol 2004; 31: 13–21.

  36. 36

    Schulz O, Sutton BJ, Beavil RL, Shi J, Sewell HF, Gould HJ et al. Cleavage of the low-affinity receptor for human IgE (CD23) by a mite cysteine protease: nature of the cleaved fragment in relation to the structure and function of CD23. Eur J Immunol 1997; 27: 584–588.

  37. 37

    Shakib F, Schulz O, Sewell H . A mite subversive: cleavage of CD23 and CD25 by Der p 1 enhances allergenicity. Immunol Today 1998; 19: 313–316.

  38. 38

    Sarfati M, Rector E, Sehon AH, Delespesse G . In vitro synthesis of IgE by human lymphocytes. IV. Suppression of the spontaneous IgE synthesis by IgE-binding factors secreted by tunicamycin-treated RPMI 8866 cells. Immunology 1984; 53: 783–790.

  39. 39

    Letellier M, Nakajima T, Delespesse G . IgE receptor on human lymphocytes. IV. Further analysis of its structure and of the role of N-linked carbohydrates. J Immunol 1988; 141: 2374–2381.

  40. 40

    Begovich AB, Carlton VE, Honigberg LA, Schrodi SJ, Chokkalingam AP, Alexander HC et al. A missense single-nucleotide polymorphism in a gene encoding a protein tyrosine phosphatase (PTPN22) is associated with rheumatoid arthritis. Am J Hum Genet 2004; 75: 330–337.

  41. 41

    Dierks SE, Bartlett WC, Edmeades RL, Gould HJ, Rao M, Conrad DH . The oligomeric nature of the murine Fc epsilon RII/CD23. Implications for function. J Immunol 1993; 150: 2372–2382.

  42. 42

    Moulder K . The role of RGD in CD23-mediated cell adhesion. Immunol Today 1996; 17: 198–199.

  43. 43

    Kijimoto-Ochiai S, Noguchi A . Two peptides from CD23, including the inverse RGD sequence and its related peptide, interact with the MHC class II molecule. Biochem Biophys Res Commun 2000; 267: 686–691.

Download references


We appreciate the assistance of Dr James Jones' lab (NJMRC) which provided EBV transformed B cells, of Dr Hongbin Shu's lab (NJMRC) which provided pAAV-lacZ plasmid and of Dr Shelby Umland (Schering-Plough Research Institute) which provided the pcDNA 3.1 ADAM33 plasmid. Dr Dennis Voelker's (NJMRC) advice regarding transfection experiments and the technical contribution of Ms Lin Cao are also appreciated. We are grateful to Dr Edward K Hill (PRA International) for professional editorial assistance and to Dr John Q Wang (UMKC School of Medicine) for reviewing this paper. This work was supported by an American Academy of Allergy, Asthma and Immunology Grant, 2004 AAAAI/Fujisawa Healthcare Allergic Skin Diseases Research Award and National Institute of Health Grant 1U01-GM/HL-61376.

Author information

Correspondence to J-F Meng.

Rights and permissions

Reprints and Permissions

About this article


  • single nucleotide polymorphism (SNP)
  • CD23
  • Der p I antigen
  • leukocyte elastase
  • cathepsins
  • ADAM33

Further reading