Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Jan 3;2(1):14-24.
doi: 10.1182/bloodadvances.2017013094. eCollection 2018 Jan 9.

Glycophorin-C Sialylation Regulates Lu/BCAM Adhesive Capacity During Erythrocyte Aging

Affiliations
Free PMC article

Glycophorin-C Sialylation Regulates Lu/BCAM Adhesive Capacity During Erythrocyte Aging

T R L Klei et al. Blood Adv. .
Free PMC article

Abstract

Lutheran/basal cell adhesion molecule (Lu/BCAM) is a transmembrane adhesion molecule expressed by erythrocytes and endothelial cells that can interact with the extracellular matrix protein laminin-α5. In sickle cell disease, Lu/BCAM is thought to contribute to adhesion of sickle erythrocytes to the vascular wall, especially during vaso-occlusive crises. On healthy erythrocytes however, its function is unclear. Here we report that Lu/BCAM is activated during erythrocyte aging. We show that Lu/BCAM-mediated binding to laminin-α5 is restricted by interacting, in cis, with glycophorin-C-derived sialic acid residues. Following loss of sialic acid during erythrocyte aging, Lu/BCAM is released from glycophorin-C and allowed to interact with sialic acid residues on laminin-α5. Decreased glycophorin-C sialylation, as observed in individuals lacking exon 3 of glycophorin-C, the so-called Gerbich phenotype, was found to correlate with increased Lu/BCAM-dependent binding to laminin-α5. In addition, we identified the sialic acid-binding site within the third immunoglobulin-like domain within Lu/BCAM that accounts for the interaction with glycophorin-C and laminin-α5. Last, we present evidence that neuraminidase-expressing pathogens, such as Streptococcus pneumoniae, can similarly induce Lu/BCAM-mediated binding to laminin-α5, by cleaving terminal sialic acid residues from the erythrocyte membrane. These results shed new light on the mechanisms contributing to increased adhesiveness of erythrocytes at the end of their lifespan, possibly facilitating their clearance. Furthermore, this work may contribute to understanding the pathology induced by neuraminidase-positive bacteria, because they are especially harmful to patients suffering from sickle cell disease and are associated with the occurrence of vaso-occlusive crises.

Conflict of interest statement

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Figures

None
Figure 1.
Figure 1.
Loss of sialic acids on the erythrocyte membrane activates Lu/BCAM. (A) Representative micrograph of adhesion of healthy erythrocytes (top panel; original magnification ×10) and sickle erythrocytes (lower panel) to a laminin-α5–coated ibidi chamber at 0.2 dyn/cm. (B) Quantification of adhesion frequency of old and young erythrocytes to laminin-α5 at 0.2 dyn/cm2 (n = 3). Percoll dilutions were stacked in a 15-mL tube; erythrocytes isolated from the fraction denser than 1.096 g/mL Percoll were defined as dense and old erythrocytes (roughly 3% of total red blood cell [RBC]), whereas erythrocytes lighter than 1.060 g/mL Percoll are here defined as light and young erythrocytes (roughly 0.75% of total RBC). (C) Biotinylated Maackia amurensis lectin type II (MA) was used to quantify α2,3-linked sialic acid (SIA) by flow cytometry (Data shown as mean fluorescence intensity [MFI] and normalized [Norm.] to control erythrocytes [aaRBC].). (D) Membrane sialic acid was removed from control erythrocytes by V cholerae neuraminidase (N'ase RBC), and adhesion frequency was assessed at 0.2 dyn/cm2 (n = 5). Specificity was addressed using an Lu/BCAM blocking polyclonal antibody. (E) Sickle erythrocyte (ssRBC) adhesion frequency to laminin-α5 at 0.2 dyn/cm2, either treated or not treated with neuraminidase for 30 minutes at 37°C (n = 5). *P < .05; ***P < .001.
Figure 2.
Figure 2.
Lu/BCAM amino acid residue R338 is critical for laminin-α5 binding. (A) Laminin-α5–coated ibidi chambers and erythrocytes were treated with neuraminidase and compared with nontreated erythrocytes (n = 4), and adhesion frequency to laminin-α5 was assessed at 0.2 dyn/cm2. (B) Various Lu/BCAM domains were aligned to SIGLEC sequences that are centered around the arginine residue critical for interaction with sialic acid residues. (C) Flow cytometry histogram comparing expression of Lu-WT (red) and Lu-R338A (blue) in transfected K562 cells. (D) Adhesion of K562 cells transfected with Lu-WT and Lu-R338A to laminin-α5–coated ibidi chambers (n = 3). K562 cell adhesion strength was measured to correct for variation between controls. Adhesion strength was defined as the percentage of cells that remain attached after gradually increasing flow shear from a static to 0.2 dyn/cm2 up to a maximum of 2.5 dyn/cm2. (E) Flow cytometric comparison of control (light gray), Lu-WT-Fc (red), and Lu-R338A-Fc (blue) coated protein-G beads. (F) Adhesion strength of Lu-WT-Fc and Lu-R338A-Fc–coated protein-G beads (n = 3). *P < .05; **P < .01; ***P < .001.
Figure 3.
Figure 3.
GpC restricts Lu/BCAM activity. (A) BRIC4, 10, and 100 were used to immunoprecipitate (IP) GpC from an erythrocyte lysate. Only BRIC10 was able to co-IP Lu/BCAM, suggesting that Lu/BCAM may mask the sialylated GpC epitope recognized by BRIC4 and possibly BRIC100 as well. Anti-CD235 was used to IP glycophorin-A (GpA). Western blotting for Lu/BCAM shows that BRIC10 co-IPs Lu/BCAM (top arrow). The anti-goat horseradish peroxidase–linked antibody directed against primary Lu/BCAM antibody was found to cross-react with the mouse light chain of the BRICS used to IP GpC (lower arrow). (B) Confocal micrograph showing Lu/BCAM clusters (fluorescein isothiocyanate, green; original magnification ×40) partially colocalizing (yellow, indicated with white arrowheads) with GpC (BRIC10, PE, red). (C) Western blot of GpC immunoprecipitation from Lu-WT and Lu-R338A transfected cells. GpC was found to co-IP only with Lu-WT. (D) GpC-ex3 (Gerbich phenotype GPC) sialylation was quantified by flow cytometry using BRIC4 and is expressed as a percentage compared with control erythrocytes (x-axis). GpC sialylation was then plotted against adhesion frequency to laminin-α5 (n = 5), which was also normalized to control (y-axis). The squared correlation coefficient (R2) is indicated. (E) Gerbich phenotype erythrocyte adhesion negatively correlates with Lu/BCAM expression. (F) Flow cytometric comparison, using BRIC4 and goat anti-human Lu/BCAM (R&D Systems), of LuGpC (orange), Lu+GpC (red), and Lu+GpC+ (blue) transfected HEK293T cells. (G) Adhesion frequency of Lu+GpC (normalized to 1) and Lu+GpC+ HEK293T cells to laminin-α5. (H) Of the HEK293T transfected with both Lu and GpC (blue population), 5.2% failed to express GpC but did express Lu (blue population in red gate, right lower quadrant), as was assessed by flow cytometry. Upon flowing the total HEK293T cell population over laminin-coated ibidi chambers, the 5.2% GpC negative fraction was significantly enriched for on laminin-α5–coated ibidi chambers as determined by fluorescence microscopy (n = 4). *P < .05.
Figure 4.
Figure 4.
Lu/BCAM membrane localization. (A) FRAP of biotin-labeled erythrocytes (top row; original magnification ×40) and Lu/BCAM-labeled, neuraminidase-treated, erythrocytes (bottom row). Alexa Fluor 488 was bleached at 70% output power for 3 iterations for a duration of 1.29 seconds. (B) FRAP quantification of biotin-labeled erythrocytes (dotted line), untreated Lu/BCAM-labeled erythrocytes (thin line), and neuraminidase-treated Lu/BCAM-labeled erythrocytes (thick line). (C) Lu/BCAM western blot of DSM and DRM either treated with neuraminidase (+) or not treated (−). Membrane was separated using 1% Triton X-100. Flotillin was used as a positive control to show successful isolation of lipid rafts (DRM). (D) Depicted is how GpC-derived sialic acid residues interact with arginine 338 on the third domain of Lu/BCAM, inhibiting the interaction with laminin-α5. Upon loss of sialic acid residues, through either erythrocyte aging or removal of sialic acid by neuraminidase, this interaction is lost, leading exposure of the sialic acid binding domain of Lu/BCAM, facilitating an interaction with laminin-α5–derived sialic acid residues.
Figure 5.
Figure 5.
S pneumoniae activates Lu/BCAM through desialylation of the erythrocyte membrane. (A) Effect of S pneumoniae, S aureus, and purified neuraminidase from V cholerae on α2,3-linked sialic acid content of erythrocytes using M amurensis lectin and flow cytometry (n = 3-6). (B) Effect of erythrocyte incubation with S pneumoniae, S aureus, and purified neuraminidase from V cholerae on Lu/BCAM-mediated erythrocyte adhesion frequency to laminin-α5 (n = 5-7). *P < .05; **P < .01; ***P < .001. ns, not significant.

Similar articles

See all similar articles

Cited by 1 article

  • Red Blood Cells: Chasing Interactions.
    Pretini V, Koenen MH, Kaestner L, Fens MHAM, Schiffelers RM, Bartels M, Van Wijk R. Pretini V, et al. Front Physiol. 2019 Jul 31;10:945. doi: 10.3389/fphys.2019.00945. eCollection 2019. Front Physiol. 2019. PMID: 31417415 Free PMC article. Review.

Publication types

Feedback