Sialylated O-glycans and sulfated tyrosines in the NH2-terminal domain of CC chemokine receptor 5 contribute to high affinity binding of chemokines

Abstract

The chemokine receptor CCR5 plays an important role in leukocyte chemotaxis and activation, and also acts as a coreceptor for human and simian immunodeficiency viruses (HIV-1, HIV-2, and SIV). We provide evidence that CCR5 is O-glycosylated on serine 6 in the NH2 terminus. The O-linked glycans, particularly sialic acid moieties, significantly contribute to binding of the chemokine ligands. By contrast, removal of O-linked oligosaccharide exerted little effect on HIV-1 infection. Sulfation of specific tyrosine residues in the CCR5 NH2 terminus was important for efficient beta-chemokine binding. Thus, as has been observed for the binding of selectins and their ligands, O-linked carbohydrates and tyrosine sulfates play major roles in promoting the interaction of chemokines with CCR5. The resulting flexible arrays of negative charges on the CCR5 surface may allow specific, high-affinity interactions with diverse chemokine ligands. Although this is the first example of O-linked oligosaccharides and tyrosine sulfates playing a role in chemokine binding, the high density of serines, threonines and tyrosines in the N-termini of many CC chemokine receptors suggests that these posttranslational modifications may commonly contribute to chemokine binding.

Figure 1.

Identification of the major site of CCR5 O-glycosylation in Cf2Th cells. (A) Primary sequence of the CCR5 extracellular domains that contain potential O-linked glycosylation sites (indicated by boxes). (B) Cf2Th cells expressing wild-type CCR5 or the CCR5Δ2–17 mutant, which lacks residues 2–17, were labeled with [³⁵S]cysteine/methionine or [³H]galactose, as indicated. Cell lysates were precipitated with the 1D4 antibody that recognizes a COOH-terminal epitope tag. Precipitates were analyzed by SDS-PAGE. (C) Cf2Th cells stably expressing CCR5 variants were labeled overnight with [³⁵S]cysteine/methionine, lysed and incubated with the 1D4 antibody. Precipitates were divided into equal parts, which were treated overnight with a glycosidase mixture to remove O-linked carbohydrate chains or with enzyme buffer alone. The CCR5 variants are named according to the residues at positions 6, 7, 16, and 17, with SSTS being the wild-type CCR5. (D) Cf2Th cells stably expressing CCR5 variants were labeled with [³⁵S]cysteine/methionine, lysed and incubated with the 1D4 antibody. Precipitates were analyzed by SDS-PAGE. (E) [³H]galactose-labeled Cf2Th cells expressing similar surface levels of wild-type CCR5 (SSTS) and the indicated mutants were used for immunoprecipitation with the 1D4 antibody. Precipitates were analyzed by SDS-PAGE. The experiments were performed at least three times with similar results.

Figure 1.

Identification of the major site of CCR5 O-glycosylation in Cf2Th cells. (A) Primary sequence of the CCR5 extracellular domains that contain potential O-linked glycosylation sites (indicated by boxes). (B) Cf2Th cells expressing wild-type CCR5 or the CCR5Δ2–17 mutant, which lacks residues 2–17, were labeled with [³⁵S]cysteine/methionine or [³H]galactose, as indicated. Cell lysates were precipitated with the 1D4 antibody that recognizes a COOH-terminal epitope tag. Precipitates were analyzed by SDS-PAGE. (C) Cf2Th cells stably expressing CCR5 variants were labeled overnight with [³⁵S]cysteine/methionine, lysed and incubated with the 1D4 antibody. Precipitates were divided into equal parts, which were treated overnight with a glycosidase mixture to remove O-linked carbohydrate chains or with enzyme buffer alone. The CCR5 variants are named according to the residues at positions 6, 7, 16, and 17, with SSTS being the wild-type CCR5. (D) Cf2Th cells stably expressing CCR5 variants were labeled with [³⁵S]cysteine/methionine, lysed and incubated with the 1D4 antibody. Precipitates were analyzed by SDS-PAGE. (E) [³H]galactose-labeled Cf2Th cells expressing similar surface levels of wild-type CCR5 (SSTS) and the indicated mutants were used for immunoprecipitation with the 1D4 antibody. Precipitates were analyzed by SDS-PAGE. The experiments were performed at least three times with similar results.

Figure 1.

Identification of the major site of CCR5 O-glycosylation in Cf2Th cells. (A) Primary sequence of the CCR5 extracellular domains that contain potential O-linked glycosylation sites (indicated by boxes). (B) Cf2Th cells expressing wild-type CCR5 or the CCR5Δ2–17 mutant, which lacks residues 2–17, were labeled with [³⁵S]cysteine/methionine or [³H]galactose, as indicated. Cell lysates were precipitated with the 1D4 antibody that recognizes a COOH-terminal epitope tag. Precipitates were analyzed by SDS-PAGE. (C) Cf2Th cells stably expressing CCR5 variants were labeled overnight with [³⁵S]cysteine/methionine, lysed and incubated with the 1D4 antibody. Precipitates were divided into equal parts, which were treated overnight with a glycosidase mixture to remove O-linked carbohydrate chains or with enzyme buffer alone. The CCR5 variants are named according to the residues at positions 6, 7, 16, and 17, with SSTS being the wild-type CCR5. (D) Cf2Th cells stably expressing CCR5 variants were labeled with [³⁵S]cysteine/methionine, lysed and incubated with the 1D4 antibody. Precipitates were analyzed by SDS-PAGE. (E) [³H]galactose-labeled Cf2Th cells expressing similar surface levels of wild-type CCR5 (SSTS) and the indicated mutants were used for immunoprecipitation with the 1D4 antibody. Precipitates were analyzed by SDS-PAGE. The experiments were performed at least three times with similar results.

Figure 1.

Identification of the major site of CCR5 O-glycosylation in Cf2Th cells. (A) Primary sequence of the CCR5 extracellular domains that contain potential O-linked glycosylation sites (indicated by boxes). (B) Cf2Th cells expressing wild-type CCR5 or the CCR5Δ2–17 mutant, which lacks residues 2–17, were labeled with [³⁵S]cysteine/methionine or [³H]galactose, as indicated. Cell lysates were precipitated with the 1D4 antibody that recognizes a COOH-terminal epitope tag. Precipitates were analyzed by SDS-PAGE. (C) Cf2Th cells stably expressing CCR5 variants were labeled overnight with [³⁵S]cysteine/methionine, lysed and incubated with the 1D4 antibody. Precipitates were divided into equal parts, which were treated overnight with a glycosidase mixture to remove O-linked carbohydrate chains or with enzyme buffer alone. The CCR5 variants are named according to the residues at positions 6, 7, 16, and 17, with SSTS being the wild-type CCR5. (D) Cf2Th cells stably expressing CCR5 variants were labeled with [³⁵S]cysteine/methionine, lysed and incubated with the 1D4 antibody. Precipitates were analyzed by SDS-PAGE. (E) [³H]galactose-labeled Cf2Th cells expressing similar surface levels of wild-type CCR5 (SSTS) and the indicated mutants were used for immunoprecipitation with the 1D4 antibody. Precipitates were analyzed by SDS-PAGE. The experiments were performed at least three times with similar results.

Figure 1.

Identification of the major site of CCR5 O-glycosylation in Cf2Th cells. (A) Primary sequence of the CCR5 extracellular domains that contain potential O-linked glycosylation sites (indicated by boxes). (B) Cf2Th cells expressing wild-type CCR5 or the CCR5Δ2–17 mutant, which lacks residues 2–17, were labeled with [³⁵S]cysteine/methionine or [³H]galactose, as indicated. Cell lysates were precipitated with the 1D4 antibody that recognizes a COOH-terminal epitope tag. Precipitates were analyzed by SDS-PAGE. (C) Cf2Th cells stably expressing CCR5 variants were labeled overnight with [³⁵S]cysteine/methionine, lysed and incubated with the 1D4 antibody. Precipitates were divided into equal parts, which were treated overnight with a glycosidase mixture to remove O-linked carbohydrate chains or with enzyme buffer alone. The CCR5 variants are named according to the residues at positions 6, 7, 16, and 17, with SSTS being the wild-type CCR5. (D) Cf2Th cells stably expressing CCR5 variants were labeled with [³⁵S]cysteine/methionine, lysed and incubated with the 1D4 antibody. Precipitates were analyzed by SDS-PAGE. (E) [³H]galactose-labeled Cf2Th cells expressing similar surface levels of wild-type CCR5 (SSTS) and the indicated mutants were used for immunoprecipitation with the 1D4 antibody. Precipitates were analyzed by SDS-PAGE. The experiments were performed at least three times with similar results.

Figure 2.

Characterization of O-glycosylation of CCR5 expressed in HEK 293T, HeLa, and primary macrophages. Wild-type (wt) and mutant CCR5 proteins were expressed transiently in HEK 293T cells (A) and HeLa cells (B). To express CCR5 variants in primary rhesus monkey macrophages (C), cells were cultured for 7 d and then transduced with 30,000 RT units of VSV G-pseudotyped SHIV vectors containing the CCR5 genes in place of the nef gene. All cells were labeled with [³⁵S]cysteine/methionine overnight, lysed, and used for precipitation by the 1D4 antibody. Precipitated proteins were analyzed by SDS-PAGE.

Figure 2.

Characterization of O-glycosylation of CCR5 expressed in HEK 293T, HeLa, and primary macrophages. Wild-type (wt) and mutant CCR5 proteins were expressed transiently in HEK 293T cells (A) and HeLa cells (B). To express CCR5 variants in primary rhesus monkey macrophages (C), cells were cultured for 7 d and then transduced with 30,000 RT units of VSV G-pseudotyped SHIV vectors containing the CCR5 genes in place of the nef gene. All cells were labeled with [³⁵S]cysteine/methionine overnight, lysed, and used for precipitation by the 1D4 antibody. Precipitated proteins were analyzed by SDS-PAGE.

Figure 2.

Characterization of O-glycosylation of CCR5 expressed in HEK 293T, HeLa, and primary macrophages. Wild-type (wt) and mutant CCR5 proteins were expressed transiently in HEK 293T cells (A) and HeLa cells (B). To express CCR5 variants in primary rhesus monkey macrophages (C), cells were cultured for 7 d and then transduced with 30,000 RT units of VSV G-pseudotyped SHIV vectors containing the CCR5 genes in place of the nef gene. All cells were labeled with [³⁵S]cysteine/methionine overnight, lysed, and used for precipitation by the 1D4 antibody. Precipitated proteins were analyzed by SDS-PAGE.

Figure 3.

Contribution of O-glycosylation to HIV-1 and SIV entry. Cf2Th cells stably expressing CD4 were transfected with plasmids encoding wild-type CCR5 or the indicated mutants. 2 d after transfection, the CCR5 expression level was assessed by FACS^® analysis of a cell sample stained with phycoerythrin-conjugated 2D7 antibody. Cells expressing similar levels of the CCR5 variants were infected with recombinant CAT-expressing viruses pseudotyped with different HIV-1 and SIV envelope glycoproteins. 3 d later, the CAT activity in cell lysates was measured. The results represent means and standard deviations of triplicate samples. One representative experiment of four performed is shown.

Figure 4.

Effects of changes in CCR5 O-glycosylation sites on binding and signaling of MIP-1α and MIP-1β. Cf2Th cells stably expressing CCR5 variants were incubated with 0.1 nM ¹²⁵I-MIP-1α (A) or ¹²⁵I-MIP-1β (B) and increasing amounts of unlabeled competitor. Cells were washed and bound labeled chemokine was quantitated in a β-counter. The binding data were analyzed with Prism software (GraphPad) (reference 27). FACS^® measurements were done in parallel on the same cells using phycoerythrin-conjugated 2D7 antibody. Mean fluorescence for cells expressing wild-type (wt) CCR5 was 469 (100%); for SSAA, 434 (92.5%), for AATS, 375 (80%), and for AAAA, 399 (85.1%). The experiments shown are representative of four binding assays performed with similar results. The same cells expressing wild-type CCR5 or the SSAA or AATS mutants, as well as CCR5-negative parental Cf2Th cells, were transfected with a plasmid encoding the G protein subunit G_α16, loaded with the indicator dye Fura-2, and used to measure changes in intracellular Ca²+ concentrations after stimulation (arrowhead) with 10 nM MIP-1α (C) or 10 nM MIP-1β (D). Ca²+ influx is shown as the ratio of the fluorescence signals observed at 510 nm after excitation at 340 and 380 nm. The curves were superimposed for comparison and are representative of the results obtained in the two experiments performed. (E) Homologous competition experiment with MIP-1β similar to B except that Cf2Th cells transiently expressing CCR5 variants were used. The mean fluorescence intensities of the cells stained with the 2D7 antibody were: wild-type, 174 (100%); AATS, 174 (100%); SATS, 139 (79.9%); ASTS, 152 (87.4%). A representative experiment of three, all with similar results, is shown.

Figure 4.

Effects of changes in CCR5 O-glycosylation sites on binding and signaling of MIP-1α and MIP-1β. Cf2Th cells stably expressing CCR5 variants were incubated with 0.1 nM ¹²⁵I-MIP-1α (A) or ¹²⁵I-MIP-1β (B) and increasing amounts of unlabeled competitor. Cells were washed and bound labeled chemokine was quantitated in a β-counter. The binding data were analyzed with Prism software (GraphPad) (reference 27). FACS^® measurements were done in parallel on the same cells using phycoerythrin-conjugated 2D7 antibody. Mean fluorescence for cells expressing wild-type (wt) CCR5 was 469 (100%); for SSAA, 434 (92.5%), for AATS, 375 (80%), and for AAAA, 399 (85.1%). The experiments shown are representative of four binding assays performed with similar results. The same cells expressing wild-type CCR5 or the SSAA or AATS mutants, as well as CCR5-negative parental Cf2Th cells, were transfected with a plasmid encoding the G protein subunit G_α16, loaded with the indicator dye Fura-2, and used to measure changes in intracellular Ca²+ concentrations after stimulation (arrowhead) with 10 nM MIP-1α (C) or 10 nM MIP-1β (D). Ca²+ influx is shown as the ratio of the fluorescence signals observed at 510 nm after excitation at 340 and 380 nm. The curves were superimposed for comparison and are representative of the results obtained in the two experiments performed. (E) Homologous competition experiment with MIP-1β similar to B except that Cf2Th cells transiently expressing CCR5 variants were used. The mean fluorescence intensities of the cells stained with the 2D7 antibody were: wild-type, 174 (100%); AATS, 174 (100%); SATS, 139 (79.9%); ASTS, 152 (87.4%). A representative experiment of three, all with similar results, is shown.

Figure 4.

Effects of changes in CCR5 O-glycosylation sites on binding and signaling of MIP-1α and MIP-1β. Cf2Th cells stably expressing CCR5 variants were incubated with 0.1 nM ¹²⁵I-MIP-1α (A) or ¹²⁵I-MIP-1β (B) and increasing amounts of unlabeled competitor. Cells were washed and bound labeled chemokine was quantitated in a β-counter. The binding data were analyzed with Prism software (GraphPad) (reference 27). FACS^® measurements were done in parallel on the same cells using phycoerythrin-conjugated 2D7 antibody. Mean fluorescence for cells expressing wild-type (wt) CCR5 was 469 (100%); for SSAA, 434 (92.5%), for AATS, 375 (80%), and for AAAA, 399 (85.1%). The experiments shown are representative of four binding assays performed with similar results. The same cells expressing wild-type CCR5 or the SSAA or AATS mutants, as well as CCR5-negative parental Cf2Th cells, were transfected with a plasmid encoding the G protein subunit G_α16, loaded with the indicator dye Fura-2, and used to measure changes in intracellular Ca²+ concentrations after stimulation (arrowhead) with 10 nM MIP-1α (C) or 10 nM MIP-1β (D). Ca²+ influx is shown as the ratio of the fluorescence signals observed at 510 nm after excitation at 340 and 380 nm. The curves were superimposed for comparison and are representative of the results obtained in the two experiments performed. (E) Homologous competition experiment with MIP-1β similar to B except that Cf2Th cells transiently expressing CCR5 variants were used. The mean fluorescence intensities of the cells stained with the 2D7 antibody were: wild-type, 174 (100%); AATS, 174 (100%); SATS, 139 (79.9%); ASTS, 152 (87.4%). A representative experiment of three, all with similar results, is shown.

Figure 4.

Effects of changes in CCR5 O-glycosylation sites on binding and signaling of MIP-1α and MIP-1β. Cf2Th cells stably expressing CCR5 variants were incubated with 0.1 nM ¹²⁵I-MIP-1α (A) or ¹²⁵I-MIP-1β (B) and increasing amounts of unlabeled competitor. Cells were washed and bound labeled chemokine was quantitated in a β-counter. The binding data were analyzed with Prism software (GraphPad) (reference 27). FACS^® measurements were done in parallel on the same cells using phycoerythrin-conjugated 2D7 antibody. Mean fluorescence for cells expressing wild-type (wt) CCR5 was 469 (100%); for SSAA, 434 (92.5%), for AATS, 375 (80%), and for AAAA, 399 (85.1%). The experiments shown are representative of four binding assays performed with similar results. The same cells expressing wild-type CCR5 or the SSAA or AATS mutants, as well as CCR5-negative parental Cf2Th cells, were transfected with a plasmid encoding the G protein subunit G_α16, loaded with the indicator dye Fura-2, and used to measure changes in intracellular Ca²+ concentrations after stimulation (arrowhead) with 10 nM MIP-1α (C) or 10 nM MIP-1β (D). Ca²+ influx is shown as the ratio of the fluorescence signals observed at 510 nm after excitation at 340 and 380 nm. The curves were superimposed for comparison and are representative of the results obtained in the two experiments performed. (E) Homologous competition experiment with MIP-1β similar to B except that Cf2Th cells transiently expressing CCR5 variants were used. The mean fluorescence intensities of the cells stained with the 2D7 antibody were: wild-type, 174 (100%); AATS, 174 (100%); SATS, 139 (79.9%); ASTS, 152 (87.4%). A representative experiment of three, all with similar results, is shown.

Figure 4.

Effects of changes in CCR5 O-glycosylation sites on binding and signaling of MIP-1α and MIP-1β. Cf2Th cells stably expressing CCR5 variants were incubated with 0.1 nM ¹²⁵I-MIP-1α (A) or ¹²⁵I-MIP-1β (B) and increasing amounts of unlabeled competitor. Cells were washed and bound labeled chemokine was quantitated in a β-counter. The binding data were analyzed with Prism software (GraphPad) (reference 27). FACS^® measurements were done in parallel on the same cells using phycoerythrin-conjugated 2D7 antibody. Mean fluorescence for cells expressing wild-type (wt) CCR5 was 469 (100%); for SSAA, 434 (92.5%), for AATS, 375 (80%), and for AAAA, 399 (85.1%). The experiments shown are representative of four binding assays performed with similar results. The same cells expressing wild-type CCR5 or the SSAA or AATS mutants, as well as CCR5-negative parental Cf2Th cells, were transfected with a plasmid encoding the G protein subunit G_α16, loaded with the indicator dye Fura-2, and used to measure changes in intracellular Ca²+ concentrations after stimulation (arrowhead) with 10 nM MIP-1α (C) or 10 nM MIP-1β (D). Ca²+ influx is shown as the ratio of the fluorescence signals observed at 510 nm after excitation at 340 and 380 nm. The curves were superimposed for comparison and are representative of the results obtained in the two experiments performed. (E) Homologous competition experiment with MIP-1β similar to B except that Cf2Th cells transiently expressing CCR5 variants were used. The mean fluorescence intensities of the cells stained with the 2D7 antibody were: wild-type, 174 (100%); AATS, 174 (100%); SATS, 139 (79.9%); ASTS, 152 (87.4%). A representative experiment of three, all with similar results, is shown.

Figure 5.

Effect of O-glycosylation of CCR5 expressed in ldlD cells on chemokine association. The addition of O-glycans in ldlD cells depends on the presence of GalNAc and Gal in the culture media. Parental CHO cells expressing CCR5 were grown in normal F12 medium, as described in Materials and Methods. ldlD cells stably expressing CCR5 were grown for 48 h in F12 medium supplemented with 2% dialyzed FCS and ITS, either with or without 500 μM GalNAc and 50 μM Gal. (A) Cells were labeled for 8 h with [³⁵S]cysteine/methionine. CCR5 was precipitated with the 1D4 antibody from cell lysates and subjected to SDS-PAGE. Homologous competitive binding assays with MIP-1α (B) and MIP-1β (C) were performed on CCR5-expressing ldlD-CCR5 cells and, to assess nonspecific binding, on CCR5-negative ldlD cells. The mean fluorescence values determined with phycoerythrin-conjugated 2D7 antibody in parallel to the binding assays were 3.6 for ldlD cells, 176.1 (100%) and 134 (76.1%) for ldlD-CCR5 cells grown with or without GalNAc and Gal, respectively. Experiments are representatives of two (B) or three (C) assays performed with similar results.

Figure 5.

Effect of O-glycosylation of CCR5 expressed in ldlD cells on chemokine association. The addition of O-glycans in ldlD cells depends on the presence of GalNAc and Gal in the culture media. Parental CHO cells expressing CCR5 were grown in normal F12 medium, as described in Materials and Methods. ldlD cells stably expressing CCR5 were grown for 48 h in F12 medium supplemented with 2% dialyzed FCS and ITS, either with or without 500 μM GalNAc and 50 μM Gal. (A) Cells were labeled for 8 h with [³⁵S]cysteine/methionine. CCR5 was precipitated with the 1D4 antibody from cell lysates and subjected to SDS-PAGE. Homologous competitive binding assays with MIP-1α (B) and MIP-1β (C) were performed on CCR5-expressing ldlD-CCR5 cells and, to assess nonspecific binding, on CCR5-negative ldlD cells. The mean fluorescence values determined with phycoerythrin-conjugated 2D7 antibody in parallel to the binding assays were 3.6 for ldlD cells, 176.1 (100%) and 134 (76.1%) for ldlD-CCR5 cells grown with or without GalNAc and Gal, respectively. Experiments are representatives of two (B) or three (C) assays performed with similar results.

Figure 5.

Effect of O-glycosylation of CCR5 expressed in ldlD cells on chemokine association. The addition of O-glycans in ldlD cells depends on the presence of GalNAc and Gal in the culture media. Parental CHO cells expressing CCR5 were grown in normal F12 medium, as described in Materials and Methods. ldlD cells stably expressing CCR5 were grown for 48 h in F12 medium supplemented with 2% dialyzed FCS and ITS, either with or without 500 μM GalNAc and 50 μM Gal. (A) Cells were labeled for 8 h with [³⁵S]cysteine/methionine. CCR5 was precipitated with the 1D4 antibody from cell lysates and subjected to SDS-PAGE. Homologous competitive binding assays with MIP-1α (B) and MIP-1β (C) were performed on CCR5-expressing ldlD-CCR5 cells and, to assess nonspecific binding, on CCR5-negative ldlD cells. The mean fluorescence values determined with phycoerythrin-conjugated 2D7 antibody in parallel to the binding assays were 3.6 for ldlD cells, 176.1 (100%) and 134 (76.1%) for ldlD-CCR5 cells grown with or without GalNAc and Gal, respectively. Experiments are representatives of two (B) or three (C) assays performed with similar results.

Figure 6.

Effect of sialidase treatment of Cf2Th cells expressing CCR5 on chemokine binding. CCR5-expressing Cf2Th cells were divided into two groups and incubated with either sialidase or buffer without enzyme. (A) Sialidase treatment increased CCR5 mobility on a reducing SDS-polyacrylamide gel. Cells were labeled with ³⁵S-cysteine/methionine overnight before incubation with buffer alone (−) or sialidase (+). Subsequently, cells were lysed and the lysates precipitated with the 1D4 antibody. Homologous competitive binding assays with MIP-1α (B) and MIP-1β (C) were performed on Cf2Th cells expressing wild-type CCR5 after sialidase treatment. Cells expressing the deletion mutant CCR5Δ2–17 were included in (B) to determine nonspecific chemokine binding. Sialidase treatment did not change the mean fluorescence of cells stained with the phycoerythrin-conjugated 2D7 antibody (data not shown). wt, wild-type.

Figure 6.

Effect of sialidase treatment of Cf2Th cells expressing CCR5 on chemokine binding. CCR5-expressing Cf2Th cells were divided into two groups and incubated with either sialidase or buffer without enzyme. (A) Sialidase treatment increased CCR5 mobility on a reducing SDS-polyacrylamide gel. Cells were labeled with ³⁵S-cysteine/methionine overnight before incubation with buffer alone (−) or sialidase (+). Subsequently, cells were lysed and the lysates precipitated with the 1D4 antibody. Homologous competitive binding assays with MIP-1α (B) and MIP-1β (C) were performed on Cf2Th cells expressing wild-type CCR5 after sialidase treatment. Cells expressing the deletion mutant CCR5Δ2–17 were included in (B) to determine nonspecific chemokine binding. Sialidase treatment did not change the mean fluorescence of cells stained with the phycoerythrin-conjugated 2D7 antibody (data not shown). wt, wild-type.

Figure 6.

Effect of sialidase treatment of Cf2Th cells expressing CCR5 on chemokine binding. CCR5-expressing Cf2Th cells were divided into two groups and incubated with either sialidase or buffer without enzyme. (A) Sialidase treatment increased CCR5 mobility on a reducing SDS-polyacrylamide gel. Cells were labeled with ³⁵S-cysteine/methionine overnight before incubation with buffer alone (−) or sialidase (+). Subsequently, cells were lysed and the lysates precipitated with the 1D4 antibody. Homologous competitive binding assays with MIP-1α (B) and MIP-1β (C) were performed on Cf2Th cells expressing wild-type CCR5 after sialidase treatment. Cells expressing the deletion mutant CCR5Δ2–17 were included in (B) to determine nonspecific chemokine binding. Sialidase treatment did not change the mean fluorescence of cells stained with the phycoerythrin-conjugated 2D7 antibody (data not shown). wt, wild-type.