alphaIIbbeta3 biogenesis is controlled by engagement of alphaIIb in the calnexin cycle via the N15-linked glycan

Abstract

Although much is known about alphaIIbbeta3 structure and function, relatively little is understood about its biogenesis. Thus, we studied the kinetics of pro-alphaIIb production and degradation, focusing on whether proteasomal degradation or the calnexin cycle participates in these processes. In pulse-chase analyses, the time to half-disappearance of pro-alphaIIb (t1/2) was the same in (1) HEK293 cells transfected with (a) alphaIIb plus beta3, (b) alphaIIb alone, (c) mutant V298FalphaIIb plus beta3, or (d) I374TalphaIIb plus beta3; and (2) murine wild-type and beta3-null megakaryocytes. Inhibition of the proteasome prolonged the t1/2 values in both HEK293 cells and murine megakaryocytes. Calnexin coprecipitated with alphaIIb from HEK293 cells transfected with alphaIIb alone, alphaIIb plus beta3, and V298FalphaIIb plus beta3. For proteins in the calnexin cycle, removal of the terminal mannose residue of the middle branch of the core N-linked glycan results in degradation. Inhibition of the enzyme that removes this mannose residue prevented pro-alphaIIb degradation in beta3-null murine megakaryocytes. alphaIIb contains a conserved glycosylation consensus sequence at N15, and an N15Q mutation prevented pro-alphaIIb maturation, complex formation, and degradation. Our findings suggest that pro-alphaIIb engages the calnexin cycle via the N15 glycan and that failure of pro-alphaIIb to complex normally with beta3 results in proteasomal degradation.

Figure 1.

Pro-αIIb containing either of 2 Glanzmann thrombasthenia mutations is degraded at the same rate as normal pro-αIIb. Cells were transiently transfected with cDNA constructs expressing both normal αIIb and β3 (Nl) (A), αIIb alone (B), or the αIIb mutants V298F (VF; C) or I374T (IT; D) in combination with β3. At 36 hours, cells were metabolically labeled with 35S-methionine/cysteine and then harvested at the indicated time points. Radiolabeled protein was immunoprecipitated using a combination of the αIIb-specific mAbs, B1B5 and M-148; the samples were then subjected to SDS-PAGE under reducing conditions, and dried gels were exposed to film. Arrowheads indicate bands representing pro-αIIb, mature αIIb, and β3; the migration of molecular weight standards is shown on the right. (E) Band intensities representing each subunit were measured by densitometry and plotted as a percentage of maximum band density of pro-αIIb (at 2 hours for Nl and at 0 hours for VF and IT). The times to half-maximum appearance of mature αIIb and β3 were both 2 plus or minus 1 hours (n = 5).

Figure 2.

The half-life of pro-αIIb is prolonged in the presence of proteasome inhibitors. (A-C) Pulse-chase analyses were performed as described in Figure 1 on HEK293 cells stably expressing αIIb alone in the presence of DMSO (vehicle control; A) or either of the proteasome inhibitors MG132 (B), or PSI (C). (D) The densities of the pro-αIIb bands from a representative experiment are plotted as a percentage of the density at 3 hours after the chase.

Figure 3.

The t1/2 values of normal pro-αIIb and VFαIIb subunits expressed with β3 are prolonged in the presence of a proteasome inhibitor. Pulse-chase analyses were performed as described in Figure 1 on HEK293 cells transiently transfected with β3 and either normal αIIb or mutant VFαIIb in the presence of either DMSO (A,C), or MG132 (B,D). (E) The density of the pro-αIIb bands of one representative experiment are plotted as a percentage of their 3-hour postchase density. The normal pro-αIIb and pro-VFαIIb subunits disappeared from the cells cultured in DMSO with t1/2 values of 5 ± 2 (n = 3) and 5 ± 3 (n = 5) hours, respectively. Disappearance of both the normal and mutant pro-αIIb subunits was delayed in the presence of MG132 to 11 plus or minus 2 hours (n = 3, P = .03), and 14 plus or minus 10 hours (n = 5, P = .05), respectively.

Figure 4.

The half-lives of murine pro-αIIb are prolonged in the presence of a proteasome inhibitor in both wild-type and β3-null murine megakaryocytes. Megakaryocytes generated from the bone marrow of WT (A-B) and β3-null mice (C-D) were cultured in the presence of DMSO or MG132, and the cells were lysed at the times indicated. Samples were analyzed as described in Figure 1. (E) The densities of the pro-αIIb bands from a representative experiment are plotted as a percentage of the density at 0 hours. The disappearance of pro-αIIb subunits was delayed in WT megakaryocytes and significantly delayed in β3-null megakaryocytes in the presence of MG132 4 plus or minus 2 hours versus 9 plus or minus 5 hours (n = 3), and 3 plus or minus 2 hours versus 7 plus or minus 2 hours (n = 6; P = .01), respectively. This indicates that pro-αIIb is degraded by a proteasomal mechanism in WT and β3-null murine megakaryocytes.

Figure 5.

Normal and mutant αIIb subunits associate with calnexin. (A) Whole-cell lysates of cells either transiently or stably expressing normal αIIbβ3, VFαIIb plus β3, or αIIb alone were immunoprecipitated with antibodies against αIIb, subjected to SDS-PAGE, and then immunoblotted with mAbs against αIIb (PMI-1, top left panel, or H-160, top right panel) or calnexin (AF8, bottom left panel, or SPA-865, bottom right panel). Calnexin was coprecipitated from cells expressing normal αIIbβ3, either transiently or stably, VFαIIb plus β3, and cells expressing αIIb alone, but not from mock (vector-only)–transfected cells. As a loading control, equivalent amounts of whole-cell lysate from each cell type were subjected to SDS-PAGE and immunoblotted with anticalnexin, revealing that equal amounts of total calnexin had been loaded (data not shown). (B,C) Pulse-chase analysis demonstrated a markedly prolonged t1/2 for pro-N15QαIIb (17 ± 2 hours) as compared with normal pro-αIIb (5 ± 2 hours, n = 3, P < .001), and failure of pro-N15QαIIb to undergo normal processing to mature N15QαIIb or to form the N15QαIIb-β3 complex normally. (D) The top panel shows the sequence alignment of the first 60 amino acids of human αIIb with αIIb from mouse, rat, pig, horse, rabbit, and dog. The bottom panel shows the sequence alignment of the corresponding residues of human αV, α8, α5, α3, α7, and α4. N-linked glycosylation consensus sequences at position 15 are highlighted in yellow, and those at other positions are in cyan. Residues comprising β strands are indicated in magenta. (E) Location of N15 in the αIIbβ propeller, and the relative location of N45 in αV. Propeller blades are numbered 1 through 7 and the Cap domain is indicated. In the crystal structures of both subunits, residues at position 15 are located at the top outer corner of the first propeller blade, whereas residues at approximately position 45 lie at the apex of the next upward-facing loop of the same blade, thus occupying a position adjacent to residue 15 on the propeller's surface.,

Figure 6.

Inhibition of N-linked core oligosaccharide processing impairs degradation of normal pro-αIIb in murine megakaryocytes. (A) The N-linked core oligosaccharide that is added to nascent polypeptide chains in the ER contains 14 saccharide units in a branched configuration. Enzymes that are involved in the modification of this glycan are shown, along with inhibitors of these enzymes. represents the mannose that is removed by ER mannosidase I. The checkered diamond represents the single glucose in the oligosaccharide structure that is recognized by calnexin. (B) Pulse-chase analyses were performed as described in Figure 1 on bone marrow megakaryocytes from β3-null mice in the presence or absence of the inhibitors 1, deoxy-mannojirimycin (DMJ), deoxynojirimycin (DNJ), or kifunensine (KIF). Whole cell lysates were immunoprecipitated with a combination of mAbs to αIIb (B1B5 and M-148). ER Man I indicates ER mannosidase I; ER Man II, ER mannosidase II; UGGT, UDP-glucose:glycoprotein glucosyltransferase.

Figure 7.

Model of pro-αIIb interactions with the calnexin cycle. (A) The core N-linked glycan, Glc₃Man₉GlcNAc₂, is attached to the Asn at position 15 (N15) of pro-αIIb during translocation into the ER.(B) The glucosidases I and II (Gluc I + II) cleave off the glucose moieties, forming Glc₁Man₉GlcNAc₂, which is the glycan form recognized by calnexin (CNX). (C) Pro-αIIb remains bound to calnexin until the third glucose moiety is cleaved off by glucosidase II. (D) The glycan formed by removal of the third glucose, Man₉GlcNAc₂, is a substrate for UDP-glucose:glycoprotein glucosyltransferase (UGGT), which then reattaches a glucose moiety, recreating the Glc₁Man₉GlcNAc₂. Because incompletely folded glycoproteins (shown as squiggle) are substrates for UGGT, partially folded pro-αIIb subunits cycle between calnexin and UGGT, which provides additional time to achieve its native fold. There are 2 exits from the cycle: (E) Pro-αIIb may achieve its native fold (shown as spiral), at which point it is no longer a substrate for UGGT, and thus continues on the biogenesis pathway to complex formation with β3, or (F) the slowly active mannosidase I (Man I) may cleave the terminal mannose from the middle branch of the glycan, creating the structural signal that targets pro-αIIb to the proteasome for degradation via binding to the mannosidase-like protein EDEM and retrotranslocation out of the ER (not shown). Of note is the finding that proteasomal degradation of pro-αIIb continues at the normal rate in the presence of glucosidase inhibitors; thus pro-αIIb may proceed directly from step A to step F, without interaction with calnexin.