First published online 2 November 2004
doi: 10.1242/jcs.01516
Journal of Cell Science 117, 5887-5895 (2004)
Published by The Company of Biologists 2004
PG-M/versican binds to P-selectin glycoprotein ligand-1 and mediates leukocyte aggregation
Peng-Sheng Zheng1,3,
Dana Vais1,3,
David LaPierre1,3,
Yao-Yun Liang2,
Vivian Lee1,3,
Bing L. Yang1,3 and
Burton B. Yang1,3,*
1 Sunnybrook and Women's College Health Sciences Centre, 2075 Bayview Avenue, Toronto, ON M4N 3M5, Canada
2 Michael E. DeBakey Department of Surgery, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA
3 Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, ON M5G 1L5, Canada
View larger version (32K):
[in a new window]
|
Fig. 1. Identification of the PSGL-1 fragment that interacts with versican G3 domain. (A) Amino acid structure of PSGL-1 showing motifs of the glycoprotein. (B) Recombinant constructs containing fragments of PSGL-1. (C) A filter membrane binding assay of recombinant PSGL-1 constructs with versican G3 domain was performed. Constructs containing amino acids 279-318 of PSGL bound to the G3 domain. The large-T antigen was used as a negative control. (D) Liquid binding activity assay using the same constructs as in C. Values shown are the mean±s.d. for the relative binding activity (n=3). **P<0.01.
|
|
View larger version (41K):
[in a new window]
|
Fig. 2. Identification of motifs in the G3 domain interacting with PSGL-1. (A) Constructs containing different combinations of versican G3 domain motifs used for yeast two-hybrid assays. EGF, epidermal growth factor domain; CRD, carbohydrate recognition domain; CBP, complement binding protein domain. (B) Filter membrane binding assay of recombinant G3 constructs with PSGL-1. EGF-containing constructs interacted with PSGL-1.
|
|
View larger version (24K):
[in a new window]
|
Fig. 3. Confirmation of the interaction of PSGL-1 with versican. (A) G3, CRDCBP and CBP were incubated with glutathione S-transferase (GST), GST fused with PSGL-1 or fused with amino acids 279-402 of PSGL, GST(279-402). The mixture was purified using GST affinity columns, and purified products were analyzed on a western blot probed with 4B6 that recognizes an epitope engineered in each construct. Only G3 product interacted with both fusion proteins, GST-PSGL and GST(279-402). (B) HL60 cells that express endogenous PSGL-1 were incubated with medium containing G3, CRDCBP or CBP. HL60 cell lysate prepared by sonication was incubated with protein G beads pre-incubated with bovine serum to avoid non-specific binding or protein G saturated with 4B6. After washing, proteins binding to the beads were subjected to western blotting with anti-PSGL-1 antibody. Untreated lysate served as a control. Only G3 interacted with endogenous PSGL-1. (C) Human leukocytes were lysed and centrifuged. The supernatant was subjected to immunoprecipitation using 2B1. After washing, the bound proteins were analyzed on a western blot probed with anti-PSGL-1 antibody to detect the interaction of endogenous versican and PSGL-1.
|
|
View larger version (64K):
[in a new window]
|
Fig. 4. PSGL-1-mediated cell aggregation in the presence of versican G3. (A) Two recombinant PSGL-1 constructs, PSGL-1 and PSGL (279-402), were stably expressed in A2058 and U87 cells. Expression of PSGL(1-402) and PSGL(279-402) was assayed by western blotting. Controls were lysates from HL60, U937 and K562 cells. (B) Flow cytometry assays confirmed cell surface localization of these proteins. Blue, negative control; pink, PSGL-1 expression; green, PSGL(279-402) expression. (C) Versican G3-containing medium was incubated with U87 and A2058 cells stably transfected with PSGL(42-402), PSGL(279-402) and a control vector respectively. Ten minutes after cell inoculation, cells expressing PSGL(42-402) or PSGL(279-402) started to aggregate. (D) The G3-containing medium was added to HL60, U937 and K562 cells. Only G3 induced aggregation of HL60 cells (n=3; **P<0.01). (E) G3 was purified through a Ni-NTA affinity column and was analyzed by Coomassie Blue staining and western blotting with 4B6. The medium from vector-transfected cells was used as a control without purification. The strong band in the vector-transfected culture medium stained with Coomassie Blue was BSA. (F) Purified G3 and total proteins from vector-transfected cells were added to HL60 cells and human leukocytes. The purified G3 induced cell aggregation (n=3; **P<0.01).
|
|
View larger version (42K):
[in a new window]
|
Fig. 5. The effects of mini-versican and the G3 motifs on cell aggregation. (A) Mammalian expression constructs of mini-versican, G3, CRDCBP and CBP were generated as shown. (B) Expression of these constructs in U87 cells was confirmed by western blotting with 4B6 using culture medium from transfected cells. (C) Culture medium containing mini-versican or G3-induced aggregation of cells transfected with PSGL-1 or PSGL(279-402) construct (n=3; **P<0.01). (D) The A2058 cells stably transfected with PSGL-1 were incubated with culture medium collected from U87 cells transfected with the G3 construct at different dilutions for different time periods. The aggregation effects were reduced as G3 was diluted.
|
|
View larger version (56K):
[in a new window]
|
Fig. 7. Removal of G3-containing fragments from human plasma or blocking with PSGL fusion proteins reduces leukocyte aggregation. (A) Human plasma was subjected to western blot analysis probed with 2B1 to detect endogenous G3-containing fragments on 4% and 10% SDS gels. (B) Human plasma was incubated with protein G-bound 2B1 or serum-treated protein G alone. After centrifugation, treated plasma and the protein G beads were analyzed on a western blot probed with 2B1. Treatment with 2B1 reduced the amount of versican in the plasma (left panel, immunodeprivation). The binding of versican to the protein G bead-2B1 complex was confirmed (right panel). (C) The effects of anti-G3 antibody treated (immunodepleted) and untreated human plasma on human leukocyte aggregation were examined. Freshly isolated human leukocytes did not aggregate when maintained in RPMI medium containing 10% FBS. The addition of human plasma or protein G bead-treated plasma induced leukocyte aggregation. This effect diminished greatly when the plasma was immunodeprived by 2B1 treatment (n=3; **P<0.01). (D) Human plasma was added to human leukocytes in the presence or absence of purified GST, GST-PSGL, GST(279-402). Both PSGL fusion proteins blocked human plasma-induced leukocyte aggregation (n=3; **P<0.01). (E) Human plasma and plasma mixed with polyclonal antiserum against versican G3 domain were injected into the mouse peritoneal cavity. Two days after the injection, ascites was recovered and monocyte aggregation was examined immediately. Plasma-induced monocyte aggregation was neutralized in the presence of anti-G3 antibody.
|
|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati 
Twitter What's this?
© The Company of Biologists Ltd 2004
) and cell aggregation mediated by G3 binding to PSGL-1 (
). (B) Culture medium was analyzed on western blots in reducing and non-reducing conditions probed with 4B6. Without DTT treatment, G3 migrated as multiple bands. In the presence of DTT, G3 migrated as a major single band. (C) U87 cells stably transfected with PSGL-1 or PSGL(279-402) were subjected to different treatments as follows: incubation with G3-containing medium; treatment with DTT before addition of G3-containing medium [+DTT (cells)] or incubation with G3-containing medium that had been treated with DTT [+DTT (G3 medium)]. When the cells were treated with DTT, a partial reduction in cell aggregation was observed. When the medium was treated with DTT, cell aggregation was completely inhibited (n=3; **P<0.01).