HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 36, 32830-32836, September 6, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/36/32830    most recent M202493200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Semel, A. C. Articles by Bellis, S. L. Search for Related Content PubMed PubMed Citation Articles by Semel, A. C. Articles by Bellis, S. L. Social Bookmarking                      What's this?

Hyposialylation of Integrins Stimulates the Activity of Myeloid Fibronectin Receptors^*

Alexis C. Semel, Eric C. Seales, Anuj Singhal, Elizabeth A. Eklund^§, Karen J. Colley^¶, and Susan L. Bellis

From the  Department of Physiology and Biophysics, University of Alabama at Birmingham, Birmingham, Alabama, 35294, the ^§ Lakeside Veterans Administration Hospital, Northwestern University Medical School and the Robert H. Lurie Comprehensive Cancer Center, Chicago, Illinois, 60611, and the ^¶ Department of Biochemistry and Molecular Biology, University of Illinois College of Medicine, Chicago, Illinois, 60612

Received for publication, March 14, 2002, and in revised form, June 24, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Despite numerous reports suggesting that ₁ integrin receptors undergo differential glycosylation, the potential role of N-linked carbohydrates in modulating integrin function has been largely ignored. In the present study, we find that ₁ integrins are differentially glycosylated during phorbol ester (PMA)-stimulated differentiation of myeloid cells along the monocyte/macrophage lineage. PMA treatment of two myeloid cell lines, U937 and THP-1, induces a down-regulation in expression of the ST6Gal I sialyltransferase. Correspondingly, the ₁ integrin subunit becomes hyposialylated, suggesting that the ₁ integrin is a substrate for this enzyme. The expression of hyposialylated ₁ integrin isoforms is temporally correlated with enhanced binding of myeloid cells to fibronectin, and, importantly, fibronectin binding is inhibited when the Golgi disrupter, brefeldin A, is used to block the expression of the hyposialylated form. Consistent with the observation that cells with hyposialylated integrins are more adhesive to fibronectin, we demonstrate that the enzymatic removal of sialic acid residues from purified ₅₁ integrins stimulates fibronectin binding by these integrins. These data support the hypothesis that unsialylated ₁ integrins are more adhesive to fibronectin, although desialylation of ₅ subunits could also contribute to increased fibronectin binding. Collectively our results suggest a novel mechanism for regulation of the ₁ integrin family of cell adhesion receptors.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The U937 and THP-1 myeloid cell lines have been widely used to study myeloid differentiation along the monocyte/macrophage lineage. Following treatment with the phorbol ester, phorbol myristate acetate (PMA),¹ U937 and THP-1 cells express proteins characteristic of terminally differentiated cells, including components of the respiratory burst oxidase (gp91^phox, p67^phox, and p47^phox), granule proteins (myeloperoxidase and elastase) and cell adhesion receptors (CD11, CD18, and CD49e) (1, 2). Consistent with the expression of these proteins, PMA-treated cells acquire functions characteristic of mature phagocytes, including respiratory burst activity, phagocytosis, and increased adhesiveness to endothelial cells as well as to extracellular matrix components such as fibronectin. In vivo, the increased adhesiveness of monocyte/macrophage cells contributes to the extravasation of cells from the vasculature and may further tether cells at sites of inflammation within the tissues.

The PMA-induced adhesion of myeloid cells to extracellular matrix components is mediated by the integrin family of cell adhesion receptors. Integrins are heterodimeric glycoproteins composed of one  and one  subunit (3, 4). The specificity of integrin heterodimers is dictated by the pairing of various  and  subunits; for example, the ₅₁ integrin (VLA5) associates with fibronectin, whereas the ₂₁ integrin (VLA2) binds to either collagen or laminin. The binding of integrins to ligand, followed by receptor clustering, initiates signal transduction events that ultimately regulate many fundamental cellular processes, including initiation of gene transcription, cell survival, cell motility/invasiveness, and cytoskeletal reorganization.

The mechanisms that underlie the PMA-induced adhesion of differentiated myeloid cells to fibronectin have not been well-defined. Several groups have reported that sustained PMA treatment of U937 cells stimulates both the transcription and cell surface expression of the ₅ and ₁ integrin subunits (2, 5-7). However, untreated U937 cells express an abundant amount of the ₅₁ integrin heterodimer, and yet these cells do not bind to fibronectin. This finding suggests that these receptors are in an inactive state. Increased expression may therefore contribute to enhanced binding but is unlikely to fully account for the robust cell adhesion that occurs following PMA treatment. Intriguingly, it has been shown that PMA treatment of U937 and THP-1 cells induces the synthesis of a ₁ integrin subunit with altered N-glycosylation (8). Hence, PMA not only increases the expression of ₁ integrins but also directs the synthesis of ₁ integrins that are structurally different from the ₁ integrins expressed by untreated cells.

In the present study we provide evidence implicating differentially glycosylated ₁ integrins as mediators of the PMA-dependent fibronectin binding of U937 and THP-1 cells. Our studies show that PMA treatment causes a down-regulation in the ST6Gal I sialyltransferase, a trans-Golgi enzyme that adds the negatively charged sugar, 2-6-linked sialic acid, to glycoproteins. In turn, ₁ integrins that are synthesized following PMA treatment are hyposialylated and demonstrate enhanced fibronectin-binding capability. Collectively, our results suggest a novel mechanism for the regulation of ₁ integrin function during myeloid differentiation.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Culture

A U937 cell subclone selected for sensitivity to granulocyte macrophage-colony stimulating factor was obtained from Dr. Andrew Kraft (University of Colorado, Denver, CO). THP-1 cells were obtained from the American Type Culture Collection (ATCC). U937 cells were maintained in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and gentamicin, whereas THP-1 cells were maintained in RPMI 1640 media with 10% fetal bovine serum and gentamicin. The PMA-resistant cell line, PRU, was derived from U937 cell cultures by treating U937 cells with 100 ng/ml PMA for 36 h and then by selecting cells that did not adhere to the tissue culture dish. The PMA-containing media was replaced with normal media, and cells were grown for 72 h. This population of cells was then challenged again with 100 ng/ml PMA for 36 h, and nonadherent cells were again collected. The nonadherent cells were diluted and seeded into 96-well dishes to obtain single-cell clones. Clones were subsequently expanded and evaluated for resistance to PMA-induced cell adhesion.

Western Blotting

U937 cells were treated with or without 50 ng/ml PMA for 15 h. Cells were then lysed in 50 mM Tris-HCl buffer (pH 7.4) containing 1% Triton X-100, 0.5 mM phenylmethylsulfonyl fluoride, 20 µg/ml leupeptin, 4 mM sodium fluoride, and 200 µM sodium pervanadate ("lysis buffer"). Protein concentrations were determined using a modified Bradford assay (Sigma Chemical Co.). One to two hundred micrograms of cell lysate were resolved by reducing SDS-PAGE, and proteins were subsequently transferred to polyvinylidene difluoride membrane. Membranes were blocked for 1-2 h with 3% nonfat dry milk. After blocking, blots were incubated for 2-4 h with a monoclonal antibody specific for the ₁ integrin (Transduction Labs) or with a polyclonal antibody specific for ST6Gal I (provided by Dr. Karen Colley). Blots were then washed several times with Tris-buffered saline containing 0.05% Tween-20 (TBST). Following the washes, blots were incubated for 1 h with a horseradish-peroxidase (HRP)-coupled secondary antibody (Amersham Biosciences), then washed again with TBST. Blots were incubated with a luminescent HRP substrate (Amersham Biosciences) for 1 min, and positive signals were detected using the enhanced chemiluminescence method.

PNGase F Treatment-- Cell lysates were boiled in buffer containing 1% SDS to denature cellular proteins. Cell lysates were then diluted 5-fold in PBS, and Triton X-100 was added to a final concentration of 1%. The deglycosylating enzyme, PNGase F (Roche Molecular Biochemicals), was added to a final concentration of 10 units/ml, and lysates were incubated overnight at 37 °C.

Sialidase Treatment-- Cell lysates were treated for 3 h at 37 °C in the presence or absence of 10 milliunits/ml protease-free Clostridium perfringens sialidase (Roche Molecular Biochemicals). Desialylated or control lysates were then boiled in SDS-PAGE sample buffer and subsequently subjected to Western blot analysis.

Lectin Affinity Analyses

Six hundred micrograms of cell lysate protein was incubated for 3 h at 4 °C with 4 µg of either biotinylated SNA or biotinylated MAA (Vector Laboratories). 20 µl of streptavidin-agarose (Sigma Chemical Co.) was then added, and samples were incubated for an additional 2 h at 4 °C with rotation. Lectin/glycoprotein complexes were collected by brief centrifugation and washed three times with lysis buffer, followed by one wash with PBS. Glycoproteins were released from the complexes by boiling in SDS-PAGE sample buffer. The glycoproteins were resolved by SDS-PAGE, then immunoblotted to detect the ₁ integrin.

Cell Attachment Assays

Cells were treated with or without 50 ng/ml PMA then were seeded onto tissue culture dishes that had been precoated overnight at 4 °C with 20 µg/ml fibronectin. Cells were allowed to adhere for 15 h at 37 °C. Following this incubation, nonadherent cells were removed from the dishes by washing the dishes with PBS. Adherent cells were fixed for 40 min in 3.7% formaldehyde then stained for 40 min with 0.1% crystal violet. Stained cells were subsequently solubilized in 10% acetic acid, and absorbance spectrophotometry (540 nm) was used to quantify the amount of dye in each sample.

Integrin Function Blocking Studies-- Cells were incubated for 1 h at 37 °C with a function blocking antibody specific for either the ₁ integrin subunit (Invitrogen) or the ₂ or ₃ integrin subunits (Chemicon International). A nonspecific, isotype-matched mouse IgG (Chemicon International) was included as a control. Following incubation with the function-blocking antibody (final concentration = 10 µg/ml), 50 ng/ml PMA was added to the samples, and cells were seeded onto fibronectin-coated dishes. Cell adhesion was quantified as described above.

Cell Adhesion Time Course-- Cells were resuspended in media containing 50 ng/ml PMA and were then placed in low attachment tissue culture dishes (BD Transduction Laboratories) to eliminate cell adhesion to the dishes. At the designated time points, aliquots of cells were removed from the nonattaching dishes and seeded onto standard tissue culture dishes that had been precoated with 20 µg/ml fibronectin. Cells were allowed to adhere for 40 min, and then adhesion was quantified as described above.

Brefeldin A-- At selected time points during the adhesion time course experiment, brefeldin A (Sigma) was added to the media at a final concentration of 20 µg/ml.

Sialidase Treatment of Purified Integrins and Modified ELISA Integrin Binding Assay

Purified ₅₁ integrins (Chemicon International) were resuspended to a final concentration of 4.4 µg/ml in 50 mM Tris buffer containing 150 mM NaCl, 2 mM MgCl₂, 0.1 mM CaCl₂, and 0.1% Triton X-100 ("ELISA buffer"), adjusted to pH 6.5. Two hundred milliunits of agarose-conjugated Vibrio cholerae neuraminidase (Calbiochem) was added to the integrin solution, and samples were incubated for 3 h at 37 °C with rotation. As a control, an equivalent amount of ₅₁ integrin was incubated in pH 6.5 ELISA buffer for 3 h at 37 °C in the absence of neuraminidase. Following the 3-h incubation, the buffer was adjusted to pH 7.4 by dilution with pH 8.0 ELISA buffer. Samples were centrifuged to precipitate the agarose-conjugated neuraminidase, and the integrin-containing supernatants were loaded onto 12-well tissue culture dishes that had been precoated with varying amounts of fibronectin. A final amount of 350 ng of neuraminidase-treated ₅₁ integrin (or control integrin) was added to each fibronectin-coated well. Treated or untreated samples were also loaded onto wells that had been precoated with denatured BSA, rather than fibronectin, to control for nonspecific binding. Purified integrins were allowed to adhere to BSA or fibronectin for 1 h at 37 °C. The wells were then washed three times with ELISA buffer. The anti-₁ integrin monoclonal antibody, MAB2000, was subsequently added to the wells (1:1000 dilution, Chemicon International), and the samples were incubated for an additional hour at 37 °C. The MAB2000 antibody recognizes both native and denatured ₁ integrins and will precipitate ₅₁ from both control and PMA-treated U937 and THP-1 cells (data not shown); therefore, this antibody is insensitive to changes in glycosylation. Following the incubation with MAB2000, wells were washed three times with ELISA buffer. An HRP-conjugated anti-mouse IgG was added, and samples were incubated for 1 h at 37 °C. Samples were washed three times and then incubated for 30 min at 37 °C with the colorimetric HRP substrate, Chromogen (BIOSOURCE International). The amount of integrin bound to matrix-coated wells was quantitated by absorbance spectroscopy using a wavelength of 450 nm. Values for specific binding were obtained by subtracting the BSA-dependent binding ("nonspecific") from the total fibronectin binding.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

₁ Integrins from PMA-treated U937 and THP-1 Cells but Not PMA-resistant U937 cells (PRU cells) Exhibit Altered N-Glycosylation, as Evidenced by Increased Electrophoretic Mobility-- U937 or THP-1 myeloid cells were incubated for 15 h in the presence of 50 ng/ml PMA, a treatment that is known to induce the differentiation of myeloid cells along the monocyte/macrophage lineage. Following this incubation, PMA-treated and untreated cells (control) were lysed, and lysates were subsequently resolved by SDS-PAGE. ₁ integrins were detected by Western blot analysis. As shown in Fig. 1A, the mature ₁ integrins harvested from PMA-treated U937 and THP-1 cells migrated more rapidly during SDS-PAGE than ₁ integrins from untreated U937 and THP-1 cells. The electrophoretic mobility of the precursor ₁ integrin isoform was unaffected by PMA treatment. To determine whether the change in the mobility of mature ₁ integrins was due to altered N-glycosylation, lysates were treated with PNGase F, an enzyme that removes N-linked carbohydrates from glycoproteins. Subsequent Western blot analysis of PNGase F-treated lysates revealed that deglycosylated ₁ integrins from control and PMA-treated cells had identical mobility (Fig. 1B), indicating that the differential mobility of glycosylated integrins was due to an alteration in the composition of N-linked carbohydrates. Most probably, altered glycosylation occurred at the level of the Golgi, in that the partially glycosylated precursor ₁ integrin, an endoplasmic reticulum-localized isoform (9, 10), did not exhibit differential electrophoretic mobility in response to PMA.

View larger version (23K): [in this window] [in a new window]   Fig. 1.   The 1 integrins harvested from PMA-treated U937 and THP-1 cells, but not PRU cells, demonstrate more rapid electrophoretic mobility due to altered N-glycosylation. A, U937, THP-1, or PRU cells were incubated in the presence or absence of 50 ng/ml PMA for 15 h. Cells were then lysed in 50 mM Tris buffer containing 1% Triton X-100 and protease inhibitors. Lysates were resolved by SDS-PAGE then Western-blotted to detect the 1 integrin. The mature 1 integrins of PMA-treated U937 and THP-1 cells exhibited more rapid electrophoretic mobility as compared with untreated U937 and THP-1 cells (control). The precursor 1 integrin species observed in U937 and PRU cells is a partially glycosylated, endoplasmic reticulum-resident integrin isoform (9, 10). B, cell lysates were treated with the deglycosylating enzyme, PNGase F (see "Experimental Procedures") and were then subjected to Western blot analysis to detect the 1 integrin. No differences were observed in the electrophoretic mobility of 1 integrins from PNGase F-treated lysates, suggesting that the previously observed differences in mobility (A) were due to alterations in N-glycosylation.

One of the hallmarks of myeloid differentiation is increased cell adhesiveness to a variety of extracellular matrix components. To determine whether the expression of an altered ₁ integrin glycoform was associated with myeloid differentiation, we evaluated the effects of PMA on the electrophoretic mobility of ₁ integrins harvested from the PRU cell line, a U937 subclone that does not exhibit increased cell adhesiveness in response to PMA (see "Experimental Procedures"). In contrast to PMA-sensitive U937 and THP-1 cells, PMA did not cause altered electrophoretic mobility of mature ₁ integrins from PRU cells (Fig. 1A).

PMA Induces a Loss in 2-6 Sialic Acid Residues from the ₁ Integrin-- Previous studies of ₁ integrin glycosylation in other cell types have suggested that ₁ integrins can undergo changes in the content of sialic acids (11-14). To examine whether PMA caused a difference in ₁ integrin sialylation, we used a lectin affinity approach. Briefly, cell lysates containing an equal amount of ₁ integrin (see Fig. 1A) were incubated with a biotinylated lectin, followed by the addition of streptavidin coupled to agarose beads. Lectin/glycoprotein complexes were precipitated by centrifugation, washed, and resolved by SDS-PAGE. The ₁ integrin was then detected by Western blot analysis. Two lectins were examined; SNA, a lectin that recognizes 2-6-linked sialic acids, and MAA, a lectin that binds to 2-3-linked sialic acids. As shown in Fig. 2A, ₁ integrins from untreated U937 and THP-1 cells were precipitated by SNA, suggesting that the ₁ integrins of undifferentiated myeloid cells typically carry 2-6-linked sialic acid residues. However, SNA failed to precipitate ₁ integrins from the lysates of PMA-treated U937 and THP-1 cells. These data suggest that PMA treatment of U937 and THP-1 cells stimulated the expression of ₁ integrins that lack in 2-6 sialic acids. Unlike U937 and THP-1 cells, PMA-resistant PRU cells did not express hyposialylated ₁ integrins following PMA treatment (Fig. 2A), consistent with previous data suggesting that the ₁ integrins of PRU cells do not undergo altered glycosylation in response to PMA (Fig. 1A). In contrast to the differential recognition of ₁ integrins by the SNA lectin, the 2-3-specific lectin MAA failed to precipitate ₁ integrins from any of the cell lysates tested.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   The 1 integrins of PMA-treated U937 and THP-1 cells do not carry 2-6-linked sialic acid residues, as evidenced by the lack of recognition of these integrins by SNA lectin. A, cell lysates from U937, THP-1, and PRU cells were incubated for 4 h with 4 µg of biotinylated SNA, a lectin that specifically recognizes 2-6-linked sialic acid residues. Streptavidin-coupled agarose was subsequently added, and lysates were incubated for another 2 h at 4 °C. The SNA-glycoprotein complexes were collected by brief centrifugation and then washed extensively with lysis buffer. The complexes were boiled in SDS-PAGE sample buffer, resolved by SDS-PAGE, then Western-blotted to detect the 1 integrin. The 1 integrins of PMA-treated U937 and THP-1 cells were not precipitated by SNA, suggesting that PMA treatment induces the expression of 1 integrins that are lacking in 2-6 sialic acids. B, lysates from untreated U937 cells (control lysates) were incubated with C. perfringens sialidase to cleave sialic acids from lysate glycoproteins. These lysates, along with samples of control and PMA-treated U937 cell lysates, were resolved by SDS-PAGE and Western-blotted to detect the 1 integrin. The electrophoretic mobility of mature 1 integrins from sialidase-treated lysates was essentially identical to that of hyposialylated 1 integrins expressed by PMA-treated cells.

The SNA lectin analyses suggested that the more rapid electrophoretic mobility of ₁ integrins from PMA-treated U937 and THP-1 cells (Fig. 1A) resulted from a PMA-induced loss in 2-6 sialic acids. To further confirm that altered sialylation represented the major PMA-induced change in glycosylation, we compared the electrophoretic mobility of ₁ integrins harvested from PMA-treated cells with that of ₁ integrins from control cell lysates that had been desialylated by the enzyme, C. perfringens sialidase (Roche Molecular Biochemicals). Briefly, cell lysates were incubated for 3 h at 37 °C in the presence or absence of sialidase. Following this incubation, the lysates were resolved by SDS-PAGE then Western blot analysis was done for the ₁ integrin. As shown in Fig. 2B, the electrophoretic mobility of control cell ₁ integrins that had been enzymatically desialylated was essentially identical to that of the hyposialylated ₁ integrins expressed by PMA-treated cells.

PMA Treatment Causes a Down-regulation in ST6Gal I, the Golgi Enzyme That Adds 2-6-linked Sialic Acids-- Given that PMA treatment of U937 and THP-1 cells induced the expression of ₁ integrins that were lacking in 2-6 sialic acids, we hypothesized that PMA caused a down-regulation in the enzyme that directs 2-6 sialylation, the ST6Gal 1 sialyltransferase. Accordingly, we used Western blot analysis to examine the protein levels of ST6Gal 1 in treated and untreated cell lysates. As shown in Fig. 3, PMA caused a down-regulation of ST6Gal 1 in U937 and THP-1 cell lines but not in the PMA-resistant PRU cell line. Duplicate samples were also blotted for -actin as a loading control.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   PMA induces a down-regulation in the enzyme that adds 2-6-linked sialic acid residues, the ST6Gal I sialyltransferase. Lysates from control and PMA-treated U937, THP-1, and PRU cells were Western-blotted to detect the ST6Gal I sialyltransferase. In addition, duplicate samples were blotted for -actin as a loading/lysis control. As shown, PMA caused a down-regulation in ST6Gal I in U937 and THP-1 cells but not in PRU cells.

Expression of a Hyposialylated ₁ Integrin Glycoform Is Associated with Enhanced Cell Binding to Fibronectin-- To determine whether the expression of a hyposialylated ₁ integrin glycoform was correlated with an alteration in integrin function, we examined cell adhesion to fibronectin. U937, THP-1, and PRU cells were incubated in the presence or absence of PMA for 15 h and were then examined for adhesiveness to fibronectin using a standard colorimetric assay (see "Experimental Procedures"). As shown, U937 and THP-1 demonstrated enhanced binding to fibronectin following PMA treatment, whereas the binding of PRU cells was unaffected by PMA (Fig. 4A). The PMA-dependent binding of U937 cells to fibronectin resulted from some change in the activity of ₁ integrins, because binding was inhibited by antibodies that block the function of ₁ but not ₂ or ₃ integrins (Fig. 4B).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   PMA-treated U937 and THP-1 cells, but not PRU cells, demonstrate enhanced binding to fibronectin. A, U937, THP-1, and PRU cells were incubated for 15 h in the presence or absence of 50 ng/ml PMA and then evaluated for cell adhesion to fibronectin using a standard colorimetric assay. Briefly, cells were fixed in 3.7% formaldehyde then stained with crystal violet. Stained cells were subsequently solubilizing in 10% acetic acid, and adhesion was quantified by measuring the absorbance of the samples at 540 nm. Data shown represent the mean ± S.E. for three experiments performed in duplicate. B, antibodies that block the function of either 1, 2, or 3 integrins were added to cell suspensions and incubated for 1 h at 37 °C. A nonspecific, isotype-matched IgG was also included as a control. Following the incubation with blocking antibody, PMA was added to the antibody-containing solutions, and cells were incubated for an additional 15 h. Cell adhesion to fibronectin was then evaluated as described above. Data shown represent the mean and S.E. of three experiments performed in duplicate.

Expression of a Hyposialylated ₁ Integrin Is Temporally Correlated with Enhanced Cell Adhesion to Fibronectin-- We next examined whether the expression of a hyposialylated ₁ integrin isoform was temporally correlated with PMA-induced cell binding to fibronectin. To this end, we established a time course for PMA-stimulated cell adhesion to fibronectin. Briefly, cells were treated with PMA and then held in low attachment tissue culture dishes for varying time points. Cells were subsequently seeded onto fibronectin-coated standard tissue culture dishes, and adhesion was quantitated as described previously. As shown, PMA rapidly stimulated the binding of cells to fibronectin (Fig. 5A). However, after this initial increase, cell adhesiveness appeared to diminish between 3 and 5 h and then rise again from 7 to 15 h. We anticipated that the expression of a hyposialylated integrin would be associated with the later, more prolonged phase of cell adhesion, rather than the early, rapid phase, given that trafficking of integrins through the endoplasmic reticulum and Golgi typically requires several hours in most cell types (15-18). To determine whether the expression of a hyposialylated integrin was temporally correlated with the delayed phase of cell adhesion, we performed a Western blot analysis of ₁ integrins expressed at various time points following PMA treatment (Fig. 5B). At 4 h, there was no apparent alteration in glycosylation, as evidenced by a lack of change in ₁ integrin electrophoretic mobility. At 7 h following PMA treatment, the more rapidly migrating ₁ isoform (the hyposialylated form) was observed, although clearly some wild type integrin was still present. Nearly all of the ₁ integrin appeared to be hyposialylated at 12 following PMA treatment (Fig. 5B), and expression of the hyposialylated form was sustained at 15 h (Fig. 1A). These data suggest that the expression of a hyposialylated integrin begins between 4 and 7 h following PMA treatment and reaches a maximum by 12-15 h. Thus, the expression of the hyposialylated glycoform is temporally correlated with the delayed phase of PMA-dependent cell adhesion to fibronectin.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   The expression of a hyposialylated 1 integrin isoform is temporally correlated with the delayed phase of PMA-dependent cell adhesion to fibronectin. A, U937 cells were treated with 50 ng/ml PMA and then seeded into low attachment tissue culture dishes. At selected time points following the addition of PMA, cells were removed from the dishes and reseeded onto standard tissue culture dishes that had been precoated with 20 µg/ml fibronectin. Cell adhesion to fibronectin was then quantified as described previously. As shown, PMA induced both a rapid, but transient (0.5-3 h), and delayed (7-15 h) increase in cell adhesion to fibronectin. A representative time course experiment is depicted. Similar results were observed in five additional experiments performed in either duplicate or triplicate. B, U397 cells were treated in the presence or absence of 50 ng/ml PMA for selected time points. The cells were then lysed and Western blotting was performed to detect the 1 integrin. Note that the more rapidly migrating 1 integrin isoform (the hyposialylated form) was observed at 7 and 12 h, but not at 4 h, following PMA treatment, suggesting that the expression of this isoform is temporally correlated with the delayed phase of PMA-dependent cell adhesion to fibronectin.

Inhibiting the Expression of Hyposialylated ₁ Integrins Blocks the Delayed Phase of PMA-induced Cell Adhesion to Fibronectin-- To further address the hypothesis that the expression of a hyposialylated ₁ integrin glycoform is associated with altered integrin function, we treated cells with brefeldin A, a reagent that disrupts the Golgi apparatus, and then performed cell adhesion assays. We anticipated that brefeldin A treatment would block the expression of the hyposialylated ₁ integrin glycoform, and thus inhibit the delayed, but not the early, phase of cell adhesion to fibronectin. Accordingly, we examined PMA-dependent cell binding to fibronectin at two time points following brefeldin A treatments. First, we pretreated cells for 2 h with 20 µg/ml brefeldin A. PMA was then added to the brefeldin A-containing media, and cells were incubated with both PMA and brefeldin A for an additional hour. As shown in Fig. 6, brefeldin A had no effect on cell adhesion to fibronectin following a 1-h treatment with PMA. These data suggest that PMA-induced cell adhesion at this time point does not require the synthesis of a new integrin species. In addition, these data indicate that the amount of ₁ integrin turnover that occurs within the 3-h brefeldin A treatment does not substantially affect cell adhesion to fibronectin. We next assessed the effect of brefeldin A on cell adhesion at 8 h following PMA treatment. Brefeldin A was added at 5 h following the initiation of PMA treatment, a time point when binding is diminished, and then cell adhesion was examined 3 h later. As shown, brefeldin A significantly inhibited the enhanced cell adhesion that occurs at 8 h following PMA treatment. These data suggest that the expression of a new integrin species, presumably the hyposialylated ₁ integrin isoform, was required for the delayed phase of PMA-induced cell adhesiveness.

View larger version (13K): [in this window] [in a new window]   Fig. 6.   The Golgi-disrupting agent, brefeldin A, blocks the delayed, but not the early, phase of PMA-dependent cell adhesion to fibronectin. To determine whether the expression of a hyposialylated 1 integrin was required for PMA-dependent cell adhesion to fibronectin, U937 cells were preincubated for 3 h with brefeldin A to disrupt the Golgi apparatus and were then evaluated for cell adhesion to fibronectin. As shown, brefeldin A did not block the binding of cells exposed to PMA for 1 h, suggesting that synthesis of a new integrin species was not required for the early phase of PMA-dependent cell adhesion. However, as indicated (asterisk), fibronectin binding was significantly inhibited by brefeldin A when binding was assayed at 8 h following PMA treatment (p < 0.05, as evaluated by a Student's t test). Data shown represent the mean and S.E. for four independent experiments performed in duplicate or triplicate.

Desialylated, Purified ₅₁ Integrins Demonstrate Enhanced Fibronectin Binding-- Previous studies have shown that the enzymatic removal of sialic acids from the cell surface can modulate cell adhesion to extracellular matrix ligands (19, 20). However, because multiple cell surface receptors are likely to be affected by the desialylating enzyme, it is not currently clear to what extent desialylation modifies ₁ integrin function directly. To address the hypothesis that unsialylated ₁ integrins exhibit better binding to fibronectin, we developed a novel method to desialylate purified ₁ integrins and then monitor fibronectin binding using a modified ELISA assay (see "Experimental Procedures"). Briefly, purified ₅₁ integrins (that were confirmed to carry 2-6 sialic acid residues, see Fig. 7C) were treated with or without sialidase and were then added to fibronectin-coated dishes and allowed to adhere for 1 h at 37 °C. Bound integrins were detected by ELISA assay, using the anti-₁ integrin antibody, MAB2000 (Chemicon International). As shown in Fig. 7A, ₅₁ integrins that had been desialylated demonstrated significantly enhanced attachment to fibronectin, relative to ₅₁ integrins with 2-6-linked sialic acid sugars. These data imply that unsialylated ₁ integrins are more adhesive to fibronectin, although desialylation of ₅ subunits may also contribute to increased ligand binding. Importantly, the results garnered from binding assays with purified integrins suggest that sialic acid residues play a direct role in the ligand/receptor interaction.

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Sialidase-treated, purified 51 integrins demonstrate enhanced binding to fibronectin. A, purified 51 integrins were incubated in the presence or absence of agarose-conjugated Vibrio cholerae neuraminidase (sialidase) for 3 h at 37 °C. Samples were centrifuged to remove the sialidase, and the integrin-containing supernatants were loaded onto tissue culture dishes that had been precoated with varying amounts of fibronectin. Integrin binding to the dishes was quantitated using a modified ELISA assay (see "Experimental Procedures"). Where indicated (asterisk), the amount of fibronectin binding by sialidase-treated integrins was significantly greater than binding by control integrins (p < 0.05, evaluated by a Student's t test). Values represent the mean and S.E. for three experiments performed in duplicate. B, untreated and sialidase-treated purified 51 integrins were resolved by SDS-PAGE, then Western-blotted to detect the 1 integrin. As shown, sialidase-treated integrins demonstrated a more rapid electrophoretic mobility, consistent with the loss of sialic acids. C, untreated and sialidase-treated purified 51 integrins were precipitated by SNA lectin, resolved by SDS-PAGE, then Western-blotted for the 1 integrin. A significantly lower level of 1 integrin was observed in SNA precipitates of sialidase-treated samples, confirming that the sialidase treatment protocol was highly efficient.

To verify that 2-6 sialic acid sugars were cleaved from ₁ integrins by the sialidase enzyme, ₁ integrins were Western blotted to assay for changes in electrophoretic mobility. As shown in Fig. 7B, sialidase-treated ₁ integrins had a more rapid electrophoretic mobility than untreated ₁ integrins, consistent with the loss of sialic acid residues. In addition, a marked reduction was noted in the amount of ₁ integrins precipitated by the SNA lectin (Fig. 7C), further confirming that sialidase treatment was highly efficient in removing 2-6 sialic acid sugars.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In the present study we have investigated the mechanisms that underlie the PMA-induced adhesion of myeloid cells to fibronectin. In particular, we have shown that PMA treatment has a multiphasic effect on cell adhesiveness. PMA induces a rapid increase in adhesion, followed by a diminution at 3-5 h, and then a second phase of enhanced adhesion that begins by about 7 h, and is sustained for at least 15 h. Our results suggest that this second phase of enhanced cell adhesion is due to the PMA-dependent expression of a ₁ integrin glycoform that is lacking in the negatively charged sugar, 2-6 sialic acid. The expression of this hyposialylated ₁ integrin glycoform is highly correlated temporally with the delayed phase of cell adhesion. Expression of the hyposialylated form begins at some time between 4 and 7 h following PMA treatment, is expressed with increasingly greater abundance until ~12 h, and is sustained for at least 15 h. Importantly, the delayed phase of PMA-dependent cell adhesion is inhibited when the Golgi disrupter, brefeldin A, is used to block the expression of the hyposialylated form.

In contrast to the delayed phase, the rapid phase of PMA-dependent fibronectin binding does not appear to be dependent upon the expression of a variant integrin glycoform, because brefeldin A had no significant effect when adhesion is assayed at early time points following PMA treatment (Fig. 6). Moreover, the hyposialylated integrin glycoform is not expressed within the first 4 h following PMA treatment (Fig. 5B). It is likely that the rapid increase in cell adhesiveness is due to the PMA-dependent activation of pre-existing integrin receptors. It has been well-established, particularly in hematopoietic cells, that integrins are activated by "inside-out" signaling events (21, 22). Alternately, it has been suggested that PMA may stimulate cell adhesion by modulating events that lie downstream of the integrin-ligand interaction (23, 24). Clearly, further experiments will be required to elucidate the mechanisms that underlie the early phase of PMA-dependent myeloid cell adhesion.

Our results suggest that the PMA-dependent expression of a hyposialylated ₁ integrin glycoform is due, at least in part, to the down-regulation of the ST6Gal I sialyltransferase. Expression of the ST6Gal I sialyltransferase is known to be developmentally regulated (25, 26), and the levels of this enzyme vary in response to cell differentiation status (27-29). In our study, two myeloid cell lines that normally undergo differentiation in response to PMA (U937 and THP-1 cells) were shown to have decreased levels of ST6Gal I following PMA treatment. Consistent with these results, PMA has been previously shown to down-regulate ST6Gal I in HL-60 myeloid cells (27). Unlike U937 and THP-1 cells, myeloid cells that have been selected for PMA resistance (PRU cells) do not down-regulate ST6Gal I in response to PMA, do not express hyposialylated ₁ integrins, and, finally, do not bind to fibronectin following PMA treatment. Collectively, these results suggest that the ₁ integrin is an important substrate for ST6Gal I and that integrin function may be regulated by the level of 2-6 sialylation.

Several studies support a role for integrin carbohydrate groups in regulating the association between integrins and ligand. Akiyama et al. (30) reported that the treatment of human foreskin fibroblasts with glycosylation inhibitors blocked cell adhesion to fibronectin. Similarly, Zheng et al. (20) demonstrated that ligand binding was altered when N-linked carbohydrates were enzymatically cleaved from cell surface ₅₁ integrins. In our study, purified ₅₁ integrins were treated with sialidase enzyme to recapitulate the hyposialylated integrins expressed by PMA-treated myeloid cells. Importantly, enzymatically desialylated ₅₁ integrins demonstrated enhanced binding to fibronectin, consistent with the observation that myeloid cells expressing hyposialylated ₁ integrins adhere better to fibronectin. The enhanced binding of desialylated, purified integrins supports the hypothesis that the presence of sialic acids can directly modulate the ligand/receptor interaction.

Accumulating evidence suggests that, in hematopoietic cells, sialylation of cell surface receptors is inversely correlated with cell adhesion to extracellular matrix ligands (13, 19, 31), although this has not been universally observed (14). In one study, the enzymatic removal of sialic acid residues from the surface of HL60 cells stimulated cell adhesion to fibronectin (19). This enhanced binding was thought to be due to altered activity of the ₁ integrin, because ₁ integrins from sialidase-treated cells expressed elevated levels of a ₁-specific activation epitope. The mechanism by which sialylation of ₁ integrins alters integrin function is currently unclear. It has been suggested that sialic acid residues can mask important functional domains on membrane receptors (32). Alternately, sialic acids, due to their negative charge, could affect protein conformation.

Clearly, ₁ integrin function can be modulated by pharmacologic or enzymatic reagents that alter the composition of integrin carbohydrates. However, there is also substantial evidence that, in vivo, ₁ integrins undergo changes in carbohydrate composition in response to physiologic stimuli. Variant ₁ integrin glycoforms have been observed in multiple cell types, and the expression of a variant glycoform is typically associated with a profound change in cell phenotype (11-14, 18, 33-40). For example, changes in ₁ integrin sialylation have been observed in differentiating thymocytes (13). Modifications in ₁ integrin glycosylation are also correlated with alterations in cell migratory or invasive capability. The highly invasive cytotrophoblasts harvested from early stage human placentas have hyposialylated ₁ integrins, as compared with ₁ integrins expressed by the less invasive cytotrophoblasts of later stage placentas (12). Similarly, the ₁ integrins of activated keratinocytes, which demonstrate enhanced migration on fibronectin, have a different glycosylation pattern than naive keratinocytes (34). Finally, variant ₁ glycoforms have been observed in numerous transformed and metastatic cells (11, 35-40). The finding that variant ₁ integrins are expressed by multiple and diverse cell types, under conditions that promote long term changes in cell adhesiveness and motility, strongly suggests that variant ₁ glycoforms are functionally important.

Although our results implicate differentially 2-6-sialylated ₁ integrin subunits in regulating fibronectin receptor function, the potential role of differentially glycosylated integrin  subunits has not yet been clarified. It was previously reported that the ₅ integrin subunits expressed by G361 melanoma cells are modified by 2-8-linked oligosialic acids (41). The enzymatic removal of these polysialic acid moieties inhibited G361 cell adhesion to fibronectin, implicating ₅ carbohydrates in ligand binding (41). N-Linked glycosylation of ₅ may also indirectly modulate fibronectin binding by mediating the association of ₅ with other membrane components. Wang et al. (42) reported that epithelial cell adhesion to fibronectin was inhibited by the binding of GT1b gangliosides to the carbohydrate domain of integrin ₅. Finally, N-linked carbohydrates may regulate the localization of ₅ subunits to a glycolipid-rich plasma membrane microdomain (43). Collectively, these data suggest that glycosylation of ₅ integrin subunits may be important for fibronectin binding; however, it remains to be determined whether ₅ integrins undergo differential glycosylation in response to signal transduction cascades, as we have observed for ₁ integrins.

In light of evidence implicating glycosylation, and particularly sialylation, as a modulator of cell adhesion, we propose that inducible changes in carbohydrate composition represent an important and novel mechanism for regulation of the ₁ integrin receptor. Our studies suggest that activation of a PMA-stimulated signaling cascade causes the down-regulation of the ST6Gal I sialyltransferase, which, in turn, leads to the expression of hyposialylated ₁ integrins that bind fibronectin more actively. The expression of hyposialylated ₁ integrin glycoforms likely plays a key role in mediating the prolonged fibronectin-binding capability of differentiated myeloid cells.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants RO1 CA84248 (to S. L. B.), 5 P60 AR 20614-23 (to S. L. B.), Hl5400 (to E. A. E.), and RO1 GM48134 (to K. J. C.), by a Veterans Administration Merit Review (to E. A. E.), and by a grant from the University of Alabama Cell Adhesion and Matrix Research Center (to S. L. B.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Dept. of Physiology and Biophysics, Rm. 982A McCallum Building, 1918 University Blvd., Birmingham, AL 35294. Tel.: 205-934-3441; Fax: 205-975-9028; E-mail: bellis@ physiology.uab.edu.

Published, JBC Papers in Press, June 28, 2002, DOI 10.1074/jbc.M202493200

	ABBREVIATIONS

The abbreviations used are: PMA, phorbol myristate acetate; HRP, horseradish peroxidase; MAA, Maackia amurensis; PBS, phosphate-buffered saline; SNA, Sambucus nigra; PNGase F, N-glycosidase F; ELISA, enzyme-linked immunosorbent assay; BSA, bovine serum albumin.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				S. Kitazume, R. Oka, K. Ogawa, S. Futakawa, Y. Hagiwara, H. Takikawa, M. Kato, A. Kasahara, E. Miyoshi, N. Taniguchi, et al. Molecular insights into {beta}-galactoside {alpha}2,6-sialyltransferase secretion in vivo Glycobiology, May 1, 2009; 19(5): 479 - 487. [Abstract] [Full Text] [PDF]

				T. Isaji, Y. Sato, T. Fukuda, and J. Gu N-Glycosylation of the I-like Domain of {beta}1 Integrin Is Essential for {beta}1 Integrin Expression and Biological Function: IDENTIFICATION OF THE MINIMAL N-GLYCOSYLATION REQUIREMENT FOR {alpha}5{beta}1 J. Biol. Chem., May 1, 2009; 284(18): 12207 - 12216. [Abstract] [Full Text] [PDF]

				Y. Sato, T. Isaji, M. Tajiri, S. Yoshida-Yamamoto, T. Yoshinaka, T. Somehara, T. Fukuda, Y. Wada, and J. Gu An N-Glycosylation Site on the{beta}-Propeller Domain of the Integrin {alpha}5 Subunit Plays Key Roles in Both Its Function and Site-specific Modification by{beta}1,4-N-Acetylglucosaminyltransferase III J. Biol. Chem., May 1, 2009; 284(18): 11873 - 11881. [Abstract] [Full Text] [PDF]

				S. Diskin, Z. Cao, H. Leffler, and N. Panjwani The role of integrin glycosylation in galectin-8-mediated trabecular meshwork cell adhesion and spreading Glycobiology, January 1, 2009; 19(1): 29 - 37. [Abstract] [Full Text] [PDF]

				A. V. Woodard-Grice, A. C. McBrayer, J. K. Wakefield, Y. Zhuo, and S. L. Bellis Proteolytic Shedding of ST6Gal-I by BACE1 Regulates the Glycosylation and Function of {alpha}4{beta}1 Integrins J. Biol. Chem., September 26, 2008; 283(39): 26364 - 26373. [Abstract] [Full Text] [PDF]

				Y. Zhuo, R. Chammas, and S. L. Bellis Sialylation of {beta}1 Integrins Blocks Cell Adhesion to Galectin-3 and Protects Cells against Galectin-3-induced Apoptosis J. Biol. Chem., August 8, 2008; 283(32): 22177 - 22185. [Abstract] [Full Text] [PDF]

				M. Lee, H.-J. Lee, S. Bae, and Y.-S. Lee Protein Sialylation by Sialyltransferase Involves Radiation Resistance Mol. Cancer Res., August 1, 2008; 6(8): 1316 - 1325. [Abstract] [Full Text] [PDF]

				L. Bei, Y. Lu, S. L. Bellis, W. Zhou, E. Horvath, and E. A. Eklund Identification of a HoxA10 Activation Domain Necessary for Transcription of the Gene Encoding beta3 Integrin during Myeloid Differentiation J. Biol. Chem., June 8, 2007; 282(23): 16846 - 16859. [Abstract] [Full Text] [PDF]

				T. Isaji, Y. Sato, Y. Zhao, E. Miyoshi, Y. Wada, N. Taniguchi, and J. Gu N-Glycosylation of the beta-Propeller Domain of the Integrin {alpha}5 Subunit Is Essential for {alpha}5beta1 Heterodimerization, Expression on the Cell Surface, and Its Biological Function J. Biol. Chem., November 3, 2006; 281(44): 33258 - 33267. [Abstract] [Full Text] [PDF]

				M. Chiricolo, N. Malagolini, S. Bonfiglioli, and F. Dall'Olio Phenotypic changes induced by expression of {beta}-galactoside {alpha}2,6 sialyltransferase I in the human colon cancer cell line SW948 Glycobiology, February 1, 2006; 16(2): 146 - 154. [Abstract] [Full Text] [PDF]

				E. C. Seales, F. M. Shaikh, A. V. Woodard-Grice, P. Aggarwal, A. C. McBrayer, K. M. Hennessy, and S. L. Bellis A Protein Kinase C/Ras/ERK Signaling Pathway Activates Myeloid Fibronectin Receptors by Altering {beta}1 Integrin Sialylation J. Biol. Chem., November 11, 2005; 280(45): 37610 - 37615. [Abstract] [Full Text] [PDF]

				E. C. Seales, G. A. Jurado, B. A. Brunson, J. K. Wakefield, A. R. Frost, and S. L. Bellis Hypersialylation of {beta}1 Integrins, Observed in Colon Adenocarcinoma, May Contribute to Cancer Progression by Up-regulating Cell Motility Cancer Res., June 1, 2005; 65(11): 4645 - 4652. [Abstract] [Full Text] [PDF]

				W. Hu, R. Xu, G. Zhang, J. Jin, Z. M. Szulc, J. Bielawski, Y. A. Hannun, L. M. Obeid, and C. Mao Golgi Fragmentation Is Associated with Ceramide-induced Cellular Effects Mol. Biol. Cell, March 1, 2005; 16(3): 1555 - 1567. [Abstract] [Full Text] [PDF]

				J. H. Marino, M. Hoffman, M. Meyer, and K. S. Miller Sialyltransferase mRNA abundances in B cells are strictly controlled, correlated with cognate lectin binding, and differentially responsive to immune signaling in vitro Glycobiology, December 1, 2004; 14(12): 1265 - 1274. [Abstract] [Full Text] [PDF]

				M. Clement, J. Rocher, G. Loirand, and J. Le Pendu Expression of sialyl-Tn epitopes on {beta}1 integrin alters epithelial cell phenotype, proliferation and haptotaxis J. Cell Sci., October 1, 2004; 117(21): 5059 - 5069. [Abstract] [Full Text] [PDF]

				L. Moro, E. Perlino, E. Marra, L. R. Languino, and M. Greco Regulation of {beta}1C and {beta}1A Integrin Expression in Prostate Carcinoma Cells J. Biol. Chem., January 16, 2004; 279(3): 1692 - 1702. [Abstract] [Full Text] [PDF]

				M. Caruso, M. Cavaldesi, M. Gentile, O. Sthandier, P. Amati, and M.-I. Garcia Role of sialic acid-containing molecules and the {alpha}4{beta}1 integrin receptor in the early steps of polyomavirus infection J. Gen. Virol., November 1, 2003; 84(11): 2927 - 2936. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/36/32830    most recent M202493200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Semel, A. C. Articles by Bellis, S. L. Search for Related Content PubMed PubMed Citation Articles by Semel, A. C. Articles by Bellis, S. L. Social Bookmarking                      What's this?