Download with Protein Kinases Associate and the Transmembrane Form of

The Glycosylphosphatidylinositol-Anchored Form
and the Transmembrane Form of CD58 Associate
with Protein Kinases
This information is current as
of August 3, 2017.
Subscription
Permissions
Email Alerts
J Immunol 1998; 160:4361-4366; ;
http://www.jimmunol.org/content/160/9/4361
This article cites 44 articles, 29 of which you can access for free at:
http://www.jimmunol.org/content/160/9/4361.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 1998 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
References
Dganit Itzhaky, Nava Raz and Nurit Hollander
The Glycosylphosphatidylinositol-Anchored Form and the
Transmembrane Form of CD58 Associate with Protein
Kinases1
Dganit Itzhaky, Nava Raz, and Nurit Hollander2
S
ome cell surface proteins lack a transmembrane-spanning
domain and are anchored to the outer leaflet of the cell
membrane by a glycosylphosphatidylinositol (GPI)3 moiety (1–3). The physiologic significance of the novel GPI anchor is
as yet unknown. GPI-anchored proteins can mediate signaling. In
T cells, for example, cross-linking of any GPI-anchored protein
causes a rise in intracellular calcium, cytokine production, and
mitogenesis in the presence of phorbol esters (4, 5). This recurrent
relationship between triggering capability and membrane attachment has led to the suggestion that the GPI anchor may be involved in signal transduction. Since GPI-anchored proteins are restricted to the outer leaflet of the membrane lipid bilayer and lack
cytoplasmic domains, it is not clear how the activation signal is
transmitted. GPI-anchored proteins associate with tyrosine kinases
(6, 7). They cocluster in detergent-resistant complexes that contain
cholesterol, glycolipids, protein tyrosine kinases, and some of their
substrates (8 –13). The existence of these large complexes in specialized membrane domains may explain the surprising phenomenon of signaling through GPI-anchored receptors and support a
functional role for the GPI moiety as a signal transducing element.
However, since many transmembrane proteins can associate with
tyrosine kinases and mediate signaling, the division of the signal-
Department of Human Microbiology, Sackler School of Medicine, Tel-Aviv University, Tel-Aviv, Israel
Received for publication September 8, 1997. Accepted for publication January
8, 1998.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was supported by a grant from the Israel Academy of Sciences and
Humanities.
2
Address correspondence and reprint requests to Dr. Nurit Hollander, Department of
Human Microbiology, Sackler School of Medicine, Tel-Aviv University, Tel-Aviv
69978, Israel.
3
Abbreviations used in this paper: GPI, glycosylphosphatidylinositol; HRP, horseradish peroxidase; KOH, potassium hydroxide.
Copyright © 1998 by The American Association of Immunologists
ing function between transmembrane and GPI-anchored proteins is
unclear. Comparative studies of distinct membrane-anchored
forms of the same polypeptide may resolve this issue.
The adhesion molecule CD58 (LFA-3) plays an important role
in immunoregulation. In addition to mediating intercellular adhesion by binding to its ligand, CD2 (14 –16), it is also involved in
signal transduction (17–19). CD58 is expressed on the cell surface
of nucleated cells in both a transmembrane and a GPI-anchored
form (20 –22) and hence provides a model for the study of signaling by two distinct membrane-anchored forms of the same protein.
In this report, we demonstrate that both the transmembrane form
and the GPI-anchored form of CD58 can associate with protein
kinases. Thus, the mere capacity of a particular protein to interact
with kinases is not necessarily directed by the type of membrane
anchor.
Materials and Methods
Cells
The human B lymphoblastoid cell line JY and the variant cells derived
from it were maintained in RPMI 1640 medium supplemented with 10%
FCS. The mouse L929 cell line was maintained in Dulbecco’s modified
Eagle’s medium supplemented with 10% FCS.
Antibodies
The hybridoma TS2/9 secreting mouse anti-human CD58 (23), the hybridoma TS2/18 secreting mouse anti-human CD2 (23), and the hybridoma
W6/32 secreting mouse anti-HLA class I (24), were purchased from the
American Type Culture Collection (Rockville, MD). The hybridoma 1452C11 secreting hamster anti-murine CD3 e (25), was kindly provided by
Dr. J. A. Bluestone (University of Chicago, Chicago, IL). mAbs secreted
by the hybridoma cell lines were purified from ascites by protein G-Sepharose (Pharmacia, Uppsala, Sweden). The biotin-conjugated antiphosphotyrosine Ab 4G10 was obtained from Upstate Biotechnology
(Lake Placid, NY). F(ab9)2 of goat anti-human Ig was obtained from
Zymed (South San Francisco, CA). Anti-z Abs were raised in rabbits immunized with the synthetic peptide RRRGKGHDGLYQG from the carboxyl-terminal region of the protein.
0022-1767/98/$02.00
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
The significance of the glycosylphosphatidylinositol (GPI) anchor is unknown. Since GPI-anchored proteins mediate signaling, it
has been suggested that the GPI structure serves as a signal-transducing element. However, the division of signaling functions
between transmembrane and GPI-anchored proteins is unclear. Studies of distinct membrane-anchored forms of the same protein
may resolve this issue. The adhesion molecule CD58 is expressed on the cell surface in both a transmembrane and a GPI-anchored
form and hence provides a useful model. We studied CD58 in the human B lymphoblastoid cell line JY. In addition to mediating
adhesion, CD58 is involved in signal transduction. Incubation of JY cells with immobilized anti-CD58 Abs results in extensive
tyrosine phosphorylation and in secretion of TNF-a. We demonstrate that CD58 is associated with protein kinase(s) and with
several kinase substrates. We further demonstrate that both CD58 isoforms are involved. CD58 in JY variant cells, which express
only the transmembrane form, as well as CD58 in JY variant cells, which express only the GPI-anchored form, are associated with
kinase activity. This association results in a phosphorylation pattern that is common to the variant and to wild-type JY cells. Thus,
these findings suggest that the capacity of GPI-anchored proteins to interact with kinases is not always dependent on the GPI
anchor itself. The Journal of Immunology, 1998, 160: 4361– 4366.
4362
ASSOCIATION OF CD58 WITH PROTEIN KINASES
Reagents
Detergents, protease inhibitors, and phosphatase inhibitors were purchased
from Sigma (St. Louis, MO). [g-32P]ATP and [32P]orthophosphate were
purchased from Amersham (Buckinghamshire, England). Recombinant human TNF-a and N-glycanase were purchased from Genzyme (Cambridge,
MA). Horseradish peroxidase (HRP)-conjugated streptavidin was obtained
from Jackson ImmunoResearch (West Grove, PA).
Ab cross-linking of CD58
Flat-bottom 24-well culture plates were incubated with 10 mg/ml of TS2/9
mAb in PBS for 20 h at 4°C. The plates were rinsed with PBS, incubated
for 1 h at 37°C with PBS containing 1% BSA, and rinsed with medium. JY
cells were incubated in Ab-coated wells at 5 3 105 cells/well in 1 ml of
medium.
Assay for TNF-a
Cell biotinylation
For surface labeling with biotin, washed cells were resuspended at 5 3 106
cells/ml in PBS, pH 8.0, containing 0.1 mg/ml sulfosuccinimidobiotin
(Pierce Chemical Co., Rockford, IL). After 30 min at 4°C, cells were
washed three times.
Cell radiolabeling
Cells were washed three times in phosphate-free RPMI 1640 and resuspended in this medium supplemented with 10% dialyzed FCS at a cell
density of 4 3 107/ml. [32P]Orthophosphate (0.4 mCi/ml) was then added,
and cells were incubated for 2 h. Labeled cells were stimulated for 30 min
as described below, then washed in PBS containing phosphatase inhibitors
(400 mM sodium orthovanadate, 5 mM EDTA, 10 mM sodium fluoride,
and 10 mM sodium pyrophosphate). Washed cells were then solubilized.
Cell lysis, immunoprecipitation, kinase assay, and
electrophoresis
Cells were solubilized in 1% Nonidet P-40, except where described differently, for 30 min at 4°C. The lysis buffer contained 150 mM NaCl, 25
mM Tris-HCl, pH 8.0, 1 mM PMSF, 10 mg/ml aprotinin, 10 mM iodoacetamide, 1 mM sodium orthovanadate, 10 mM sodium fluoride, and 10 mM
sodium pyrophosphate. Lysates were cleared by centrifugation at 12,000 3
g for 15 min at 4°C. CD58 or HLA were immunoprecipitated from cell
lysates by the addition of 50 ml slurry of TS2/9-Sepharose or W6/32Sepharose, respectively, and incubated for 2 h at 4°C with constant shaking. TCR components were immunoprecipitated by rabbit anti-z Abs
preadsorbed to protein A-Sepharose. Sepharose beads were then washed
three times in lysis buffer and once in kinase buffer containing 100 mM
NaCl, 50 mM HEPES, pH 7.3, and 5 mM MnCl2. The beads were then
resuspended in 50 ml of kinase buffer, 20 mCi of [g-32P]ATP were added,
and the kinase reaction was conducted at 4°C for 15 min. The reaction was
stopped by washing three times with lysis buffer containing 25 mM EDTA.
The immune complex was eluted from the beads by boiling in Laemmli
sample buffer, and proteins were resolved by SDS-PAGE. Labeled proteins
were detected by autoradiography. For alkaline hydrolysis of proteins, gels
were treated with 1 M KOH for 2 h at 55°C as described (27).
Western blot analysis
For determination of tyrosine phosphorylation, lysate samples (30 mg
protein) were subjected to SDS-PAGE and transferred to nitrocellulose.
Membranes were probed sequentially with biotin-conjugated antiphosphotyrosine Abs and with HRP-conjugated streptavidin. The
probed proteins were visualized by enhanced chemiluminescence. For
detection of biotinylated cell surface CD58, immunoprecipitates were
subjected to SDS-PAGE and transferred to nitrocellulose. Proteins were
probed with HRP-conjugated streptavidin and visualized by enhanced
chemiluminescence.
FIGURE 1. Up-regulated release of TNF-a from JY cells following
cross-linking of CD58. A, JY cells were incubated for 24 h on uncoated
(medium) plates or on TS2/9 (anti-CD58)-coated plates in the absence or
presence of polymyxin B (5 mg/ml). B, JY cells were incubated for 24 h on
uncoated (medium) plates, TS2/9 (anti-CD58)-coated plates, or TS2/18
(anti-CD2)-coated plates. Culture supernatants were then tested for TNF-a
activity (see Materials and Methods).
N-Glycanase treatment
CD58 eluted from Sepharose beads was incubated for 20 h with N-glycanase according to the supplier’s instructions (Genzyme). The proteins
were precipitated with acetone in the presence of 10 mg tRNA as carrier
and were dissolved in sample buffer.
Purification of CD58
CD58 was purified from Triton X-100 cell lysates by immunoaffinity chromatography on TS2/9-Sepharose as described (16). After intensive washing of the immunoaffinity column, CD58 was eluted with 50 mM glycine
HCl, pH 3.0, 0.15 M NaCl, 1% octyl b-glucoside. Eluted fractions were
immediately neutralized with 1 M Tris-HCl, pH 9.0, and analyzed by SDSPAGE and silver staining. Selected fractions were pooled and passed
through Centricon-30 filters (Amicon, Danvers, MA) for concentration and
removal of detergent. Ultrafiltration was repeated three times, adding 2 ml
of PBS and reducing the volume to 100 ml with each cycle.
Incorporation of CD58
Cells were washed three times, resuspended at 10 3 106 cells/ml, and
incubated at 37°C for 1 h with 10 mg/ml purified CD58. Cells were then
washed three times.
Results
It has been previously reported that CD58 can transmit the necessary signals for the release of cytokines by human monocytes (18).
To determine whether CD58 can also transmit signals in JY cells
(B lymphoblastoid cells), immobilized mAb was used to cross-link
surface CD58. Culture supernatants of JY cells, incubated for 24 h
in plastic plates coated with TS2/9, were assessed for TNF-a activity. JY cells constitutively release low amounts of TNF-a. However, a three- to fivefold increase over baseline production of
TNF-a was observed following CD58 cross-linking (Fig. 1). The
up-regulation of TNF-a release by JY cells was not caused by
endotoxin contaminating TS2/9 preparations. Thus, incubation of
JY cells for 24 h with concentrations of LPS reaching 10 mg/ml did
not trigger release of TNF-a (data not shown). Polymyxin B,
which is an inhibitor of endotoxin activity, had no effect on the
up-regulation of TNF-a release by JY cells (Fig. 1A). In addition,
immobilized control Abs such as anti-CD2, which were purified in
a sequence with TS2/9 by the same protein G-Sepharose column,
did not trigger TNF-a release (Fig. 1B).
The activation of JY cells by anti-CD58 Abs involves tyrosine
phosphorylation. As shown in Figure 2, incubation of JY cells with
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
Culture supernatants of JY cells incubated for 24 h with immobilized Abs
were assayed for TNF-a activity as previously described (26). Briefly,
L929 cells were plated at 2.5 3 104 cells/well in flat-bottom 96-well microplates. After 24 h, culture supernatants or rTNF-a were added at different dilutions to the L929 monolayers in the presence of 1.5 mg/ml actinomycin D. Plates were incubated for an additional 24 h and then fixed
and stained with Hemacolor (26). SDS (0.5%) was added to each well, and
the staining intensity was then determined by a microplate reader spectrophotometer at 630 nm. TNF-a concentration in culture supernatants was
extrapolated from standard curves with rTNF-a.
The Journal of Immunology
4363
immobilized TS2/9 resulted in extensive protein tyrosine phosphorylation. The pattern of phosphorylated proteins was different
from the pattern observed in JY cells activated by anti-Ig Abs. We
further examined whether CD58 is associated with kinase activity.
CD58 was immunoprecipitated from lysates of nonactivated (untreated) JY cells. Kinase reactions performed on CD58 immunoprecipitates resulted in phosphorylation of several proteins (Fig.
3). Such an activity did not coimmunoprecipitate with several
other cell surface proteins such as HLA class I. The kinase activity
associated with CD58 persisted in different detergents, including
Nonidet P-40, digitonin, and CHAPS (Fig. 3A). Therefore, in subsequent experiments we used Nonidet P-40 cell lysates.
To identify the type of in vitro kinase activity, gels containing
the phosphoproteins were treated with alkali. Comparing labeled
proteins before and after alkaline hydrolysis takes advantage of the
fact that phosphotyrosine is relativley stable to base (27). We demonstrated a similar phosphoprotein pattern before and after alkali
treatment (Fig. 3B), indicating that the kinase activity associated
with CD58 was predominantly tyrosine kinase activity. To ascertain the efficacy of alkali treatment, we used as a positive control
the phosphorylation of TCR components (Fig. 3C). Activation of
murine T cells with the anti-CD3 mAb 145-2C11 results in tyrosine phosphorylation of the z polypeptide, whereas activation by
PMA results in extensive serine phosphorylation of the g polypeptide. The two phosphorylated chains have an apparent m.w. of 21
kDa in SDS-PAGE (28). TCR components, immunoprecipitated
with anti-z Abs from lysates of 32P-labeled activated splenocytes,
were treated with alkali (Fig. 3C). It was demonstrated that the
radiolabel was removed from the 21-kDa component, which was
phosphorylated following PMA activation (phosphoserine of the g
chain). In contrast, alkali treatment had no significant effect on the
radiolabel of the 21-kDa component, which was phosphorylated
following anti-CD3 activation (phosphotyrosine of the z chain).
We further considered which of the CD58 isoforms (the transmembrane or the GPI-anchored form) is associated with kinase
activity. We have previously reported the isolation of JY variant
cells, which lack expression of GPI-anchored proteins (29). These
variant cells are deficient in PIG-A, a component in the early steps
of GPI anchor biosynthesis (30). Therefore, they express only the
transmembrane form of CD58 (29). The ability of CD58 from two
variant clones (clone 5 and clone 33) to associate with kinase was
FIGURE 3. coprecipitation of kinase activity with CD58. A, JY cells
were lysed in lysis buffers containing 1% CHAPS (lanes 1 and 4), 1%
digitonin (lanes 2 and 5), or 1% Nonidet P-40 (lanes 3 and 6). Postnuclear
supernatants were immunoprecipitated with Abs to CD58 (lanes 1–3) or to
HLA class I (lanes 4 – 6). The immunoprecipitated complexes were subjected to an in vitro kinase assay. They were then eluted from Sepharose
beads, separated by SDS-PAGE, and analyzed by autoradiography. B, In
vitro kinase assay was performed on CD58 immunoprecipitates from Nonidet P-40 JY cell lysates. Following protein separation by SDS-PAGE, the
gel was treated with 1 M KOH for 2 h at 55°C. Autoradiograms taken
before (lane 1) and after (lane 2) alkali treatment are shown. C, Murine
splenocytes were labeled with [32P]orthophosphate and activated for 30
min with medium alone (lanes 1 and 5), 1:10 culture supernatant of 1452C11 mAb (lanes 2 and 6), 2.5 mg/ml Con A (lanes 3 and 7), or 100 ng/ml
PMA (lanes 4 and 8). Following cell lysis, immunoprecipitation with antiz Abs, and protein separation by SDS-PAGE, the gel was treated with 1 M
KOH for 2 h at 55°C. Autoradiograms taken before (lanes 1– 4) and after
(lanes 5– 8) alkali treatment are shown.
compared with that manifested by clone 25 (a wild-type clone
derived from the JY cell line) and by parental JY cells. Figure 4
demonstrates a similar association of CD58 with kinase activity
and with a set of kinase substrates in wild-type and in GPI-deficient JY cells. Different autoradiogram intensities were examined
in each experiment to verify the identity of phosphorylation patterns in wild-type and variant cells. These results indicate that
transmembrane CD58 associates with protein kinase activity.
Moreover, transmembrane CD58 can mediate signaling. As shown
in Figure 5, incubation of GPI-deficient JY cells (clone 5) with
immobilized anti-CD58 Abs resulted in up-regulated TNF-a
release.
The interpretation of these results may imply either that the
GPI-anchored CD58 is not associated with kinase activity or that
the two isoforms are indistinguishable in this regard. A more conclusive interpretation may be provided by kinase assays in cells
that express only the GPI-anchored CD58. Despite extensive efforts, we were unable to select a stable variant that expressed only
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
FIGURE 2. Phosphorylation of proteins on tyrosine following crosslinking of CD58. JY cells were incubated for 30 min in the absence (lane
1) or presence of immobilized TS2/9 (lane 2), F(ab9)2 fragments of antihuman Ig (lane 3), or immobilized TS2/18 (lane 4). Cells were then lysed.
Cell lysates were analyzed for tyrosine phosphorylated proteins by SDSPAGE and Western blot analysis with anti-phosphotyrosine Abs. Molecular weight standards are indicated in kilodaltons.
4364
ASSOCIATION OF CD58 WITH PROTEIN KINASES
FIGURE 6. Expression of GPI-anchored CD58 following incorporation
into JY cells. GPI-anchored CD58 was purified from human RBC. CD58negative JY cells were incubated for 1 h with 10 mg/ml purified CD58,
washed three times, and labeled with biotin. Control wild-type cells and
GPI-deficient clone 5 cells were similarly labeled. Following biotinylation,
cells were incubated for 4 h and lysed. CD58 was immunoprecipitated,
treated with N-glycanase, subjected to SDS-PAGE, and analyzed by Western blot analysis with HRP-conjugated streptavidin. Arrowheads indicate
bands corresponding to transmembrane (TM) CD58 and GPI-anchored
CD58. GPI-deficient cells (lane 1), CD58-deficient cells (lane 2), CD58reconstituted cells (lane 3), and wild-type cells (lane 4) are presented.
the GPI-anchored form of CD58. However, we obtained a cell line
that did not express CD58 in any form. We used this CD58-negative cell line to derive cells that transiently express GPI-linked
CD58. We took advantage of the fact that GPI-anchored proteins
are able to spontaneously incorporate into the lipid bilayer of cell
membranes (31–33). CD58 was purified from human erythrocytes,
which express only its GPI-anchored form. Purified GPI-CD58
was added to CD58-deficient JY cells. Following incubation, cells
were washed, labeled with biotin, and further incubated in culture
medium. CD58 expression was then determined (Fig. 6). CD58
FIGURE 5. Up-regulated release of TNF-a from GPI-deficient JY cells
following cross-linking of CD58. Variant clone 5 cells were incubated for
24 h on uncoated (medium) plates, TS2/9 (anti-CD58)-coated plates, or
TS2/18 (anti-CD2)-coated plates. Culture supernatants were then tested for
TNF-a activity.
Discussion
In this report, we demonstrate that the transmembrane form and the
GPI-anchored form of the adhesion molecule CD58 are associated
with kinase activity. Incubation of JY cells with immobilized antiCD58 mAbs resulted in extensive tyrosine phosphorylation and
up-regulation of the release of TNF-a. Up-regulation of TNF-a
release by JY cells is similarly mediated by cross-linking of HLA
class II Ags (34). JY cells express the 75-kDa TNF-a receptor. It
has been demonstrated that exogenous as well as endogenous
TNF-a mediated the enhancement of NF-kB-like activity through
the binding of the TNF-a receptor (34). Thus, the secreted TNF-a
may function as an autocrine regulator of cell activation.
The GPI anchor has generally been considered required for cell
activation. Studies with decay accelerating factor (DAF or CD55),
Ly-6, and Qa-2 suggested a critical role for the GPI anchor in
signal transduction mediated by these molecules (35–37). However, this hypothesis has recently become controversial, since it
has been reported that the transmembrane form of CD73 can deliver signals (38). This contradiction may be accounted for by the
procedures used for preparation of the transmembrane fusion constructs. For instance, 43 COOH-terminal amino acids of the GPIanchored CD55 were removed to incorporate the membrane and
cytoplasmic domains of a transmembrane protein (37). This
change in the COOH-terminal region of CD55 could have altered
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
FIGURE 4. coprecipitation of kinase activity with transmembrane
CD58. GPI-deficient variant clone 5 (lane 1), GPI-deficient variant clone
33 (lane 2), wild-type clone 25 (lane 3), and the wild-type JY cell line
(lane 4) were lysed in Nonidet P-40 lysis buffer. Postnuclear supernatants
were immunoprecipitated with Abs to CD58. The immunoprecipitated
complexes were subjected to an in vitro kinase assay. They were then
eluted from Sepharose beads, separated by SDS-PAGE, and analyzed by
autoriadiography. A and B show different exposures of the same gel (1 day
and 4 days, respectively).
migrates in SDS-PAGE as a broad band with a mean size of 65
kDa. The heterogeneity of CD58 is caused by N-linked carbohydrates, which cause the two CD58 isoforms to overlap in SDSPAGE. N-Glycanase treatment converts the two isoforms to two
polypeptides of 29 and 25.5 kDa, corresponding to the transmembrane and GPI-anchored species, respectively, and thus allows the
distinction between the two CD58 forms (20, 29). As shown in
Figure 6, CD58-negative JY cells incubated with purified CD58,
expressed the GPI-anchored form but not the transmembrane form
of CD58. The stability of the incorporated protein was limited, and
it was gradually lost from the cells, as has been demonstrated for
other exogenous GPI-anchored proteins inserted into cell membranes (32, 33). We therefore used the reconstituted cells no later
than 4 h after CD58 insertion. Kinase assays on CD58 immunoprecipitates from these transiently reconstituted cells revealed that
the GPI-anchored CD58 was associated with kinase activity (Fig.
7). The phosphorylation patterns observed in CD58 immunoprecipitates from wild-type cells (expressing both forms of CD58),
from GPI-deficient cells (expressing transmembrane CD58), and
from CD58-reconstituted cells (expressing GPI-anchored CD58)
were similar. Such a phosphorylation pattern was not observed in
CD58 immunoprecipitates of CD58-negative cells (Fig. 7).
The Journal of Immunology
its signaling function independently of the membrane anchoring
form. In addition, in the CD55, Ly-6, and Qa-2 systems, transmembrane forms were engineered to include the cytoplasmic portion of the respective transmembrane fusion proteins, whereas in
the CD73 system the cytoplasmic portion of the transmembrane
fusion protein was deleted (38). The advantage of the CD58 model
is that the sequence of the two forms is identical throughout the
entire extracellular domain from the amino terminus to the membrane attachment site (21, 22), and the cytoplasmic portion is expressed in its natural transmembrane form. Hence, it allows comparison of two distinct membrane-anchored forms of the same
polypeptide. Our findings show that the transmembrane CD58 associates with kinase activity and mediates signaling. Thus, the signal transduction capacity of GPI-anchored proteins is not always
dependent on the GPI anchor itself.
Since GPI-anchored proteins are restricted to the outer leaflet of
the membrane lipid bilayer and lack cytoplasmic domains, it is not
clear how they transmit activation signals. GPI-anchored proteins
cluster in membrane microdomains, which are enriched in glycolipids and cholesterol (8 –13). In these microdomains, the GPIanchored proteins are complexed with src-like kinases, suggesting
a mechanism by which these proteins mediate signaling (6, 7).
However, since GPI-anchored proteins and src-like kinases are
restricted to opposite leaflets of the bilayer, they cannot bind each
other directly. A transmembrane linker protein that mediates the
interaction between the two has been postulated. The GPI-anchored protein and the bridging molecule most probably interact
via their exodomains. The transmembrane CD58 has a short cytoplasmic tail of 12 amino acids. It is conceivable, therefore, that
the transmembrane CD58 also interacts with cytoplasmic signaling
elements via a linker protein. Since the two CD58 isoforms have
an identical extracellular sequence, they may interact with an identical linker protein and hence demonstrate similar association with
kinases. Therefore, kinase association of a particular GPI-anchored
protein may depend on its overall structure rather than on its possessing a GPI anchor. This suggestion is supported by the demonstration that a GPI-anchored form of CD4, constructed with the
GPI anchor of CD58, lost the signaling activity of transmembrane
CD4 (39), implying that membrane attachment by the CD58 GPI
anchor is not sufficient to direct association with kinase activity.
We have previously demonstrated that the turnover characteristics and the adhesive potential of CD58 were not directed by the
type of its of membrane anchor (29, 40). As it is demonstrated here
that association with signaling elements is independent of the type
of membrane anchor, the properties conferred upon a protein by a
GPI anchor remain elusive. Kinase complexes with GPI-anchored
CD58, in contrast to complexes with transmembrane CD58, may
be confined to the specialized membrane microdomains, in which
GPI-anchored proteins cluster in large complexes. Division of
membrane localization may be important for functional differences
between the two protein isoforms. GPI-anchored and transmembrane molecules internalize through distinct pathways (41). The
GPI-anchored proteins can be internalized via non-clathrin-coated
invaginations within the glycolipid-enriched domains (13). It has
been suggested that internalization is involved in signaling through
GPI-anchored molecules (42). Alternatively, it has been suggested
that complexed GPI-anchored proteins may be involved in activation-induced membrane vesiculation (43, 44).
References
1. Ferguson, M. A. J., and A. F. Williams. 1988. Cell surface anchoring of proteins
via glycosyl-phosphatidylinositol structures. Annu. Rev. Biochem. 57:285.
2. Low, M. G. 1989. The glycosyl-phosphatidylinositol anchor of membrane proteins. Biochem. Biophys. Acta 988:427.
3. Cross, G. A. M. 1990. Glycolipid anchoring of plasma membrane proteins. Annu.
Rev. Cell. Biol. 6:1.
4. Robinson, P. J. 1991. Phosphatidylinositol membrane anchors and T-cell activation. Immunol. Today 12:35.
5. . Brown, D. 1993. The tyrosine kinase connection: how GPI-anchored proteins
activate T cells. Curr. Opin. Immunol. 5:349.
6. Stefanova, I., V. Horejsi, I. J. Ansotegui, W. Knapp, and H. Stockinger. 1991.
GPI-anchored cell-surface molecules complexed to protein tyrosine kinases. Science 254:1016.
7. Thomas, P. M., and L. E. Samelson. 1992. The glycosylphosphatidylinositolanchored Thy-1 molecule interacts with the p60fyn protein tyrosine kinase in T
cells. J. Biol. Chem. 267:12317.
8. Rothberg, K. G., Y. S. Ying, B. A. Kamen, and R. G. W. Anderson. 1990.
Cholesterol controls the clustering of the glycophospholipid-anchored membrane
receptor for 5-methyltetrahydrofolate. J. Cell Biol. 111:2931.
9. Cinek, T., and V. Horejsi. 1992. The nature of large noncovalent complexes
containing glycosylphosphatidylinositol-anchored membrane glycoproteins and
protein tyrosine kinases. J. Immunol. 149:2262.
10. Anderson, R. G. W. 1993. Plasmalemmal caveolae and GPI-anchored membrane
proteins. Curr. Opin. Cell Biol. 5:647.
11. Fra, A. M., E. Williamson, K. Simons, and R. G. Parton. 1994. Detergent-insoluble glycolipid microdomains in lymphocytes in the absence of caveolae. J. Biol.
Chem. 269:30745.
12. Gorodinsky, A., and D. A. Harris. 1995. Glycolipid-anchored proteins in neuroblastoma cells form detergent-resistant complexes without caveolin. J. Cell Biol.
129:619.
13. Deckert, M., M. Ticchioni, and A. Bernard. 1996. Endocytosis of GPI-anchored
proteins in human lymphocytes: role for glycolipid-based domains, actin cytoskeleton, and protein kinases. J. Cell Biol. 133:791.
14. Shaw, S., G. E. G. Luce, R. Quinones, R. E. Gress, T. A. Springer, and
M. E. Sanders. 1986. Two antigen-independent adhesion pathways used by human cytotoxic T cell clones. Nature 323:262.
15. Selvaraj, P., M. L. Plunkett, M. L. Dustin, M. E. Sanders, S. Shaw, and
T. A. Springer. 1987. The lymphocyte glycoprotein CD2 (LFA-2/T11/E rosette
receptor) binds to the cell surface ligand LFA-3. Nature 326:400.
16. Dustin, M. L., M. E. Sanders, S. Shaw, and T. A. Springer. 1987. Purified lymphocyte function-associated antigen-3 (LFA-3) binds to CD2 and mediates T
lymphocyte adhesion. J. Exp. Med. 165:677.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
FIGURE 7. coprecipitation of kinase activity with GPI-anchored CD58.
The following JY cells were used: wild-type cell line (lane 1), wild-type
clone (lane 2), GPI-deficient clone expressing only transmembrane CD58
(lane 3), CD58-negative clone incorporated with GPI-anchored CD58
(lane 4), and CD58-negative clone (lane 5). In vitro kinase assay was
performed on CD58 immunoprecipitates as described in the legend to Figure 4.
4365
4366
32. McHugh, R. S., S. N. Ahmed, Y. C. Wang, K. W. Sell, and P. Selvaraj. 1995.
Construction, purificiation, and functional incorporation on tumor cells of glycolipid-anchored human B7-1 (CD80). Proc. Natl. Acad. Sci. USA 92:8059.
33. Van den Berg, C. W., T. Cinek, M. B. Hallett, V. Horejsi, and B. P. Morgan.
1995. Exogenous glycosyl phosphatidylinositol-anchored CD59 associates with
kinases in membrane clusters on U937 cells and becomes Ca21-signaling competent. J. Cell Biol. 131:669.
34. Altmonte, M., C. Pucillo, G. Damante, and M. Maio. 1993. Cross-linking of HLA
class II antigens modulates the release of tumor necrosis factor a by the EBV-B
lymphoblastoid cell line JY. J. Immunol. 151:5115.
35. Robinson, P. J., M. Millrain, J. Antoniou, E. Simpson, and A. L. Mellor. 1989.
A glycophospholipid anchor is required for Qa-2-mediated T cell activation. Nature 342:85.
36. Su, B., G. L. Waneck, R. A. Flavell, and A. L. M. Bothwell. 1991. The glycosyl
phosphatidylinositol anchor is critical for Ly-6A/E-mediated T cell activation.
J. Cell Biol. 112:377.
37. Shenoy-Scaria, A. M., J. Kwong, T. Fujita, W. Olszowy, A. S. Shaw, and
D. M. Lublin. 1992. Signal transduction through decay-accelerating factor: interaction of glycosyl-phosphatidylinositol anchor and protein tyrosine kinases
p56lck and p59fyn. J. Immunol. 149:3535.
38. Resta, R., S. W. Hooker, A. B. Laurent, J. K. Shuck, Y. Misumi, Y. Ikehara,
G. A. Koretzky, and L. F. Thompson. 1994. Glycosyl phosphatidylinositol membrane anchor is not required for T cell activation through CD73. J. Immunol.
153:1046.
39. Sleckman, B. P., Y. Rozenstein, V. E. Igras, J. L. Greenstein, and S. J. Burakoff.
1991. Glycolipid-anchored form of CD4 increases intercellular adhesion but is
unable to enhance T cell activation. J. Immunol. 147:428.
40. Hollander, N. 1992. Membrane dynamics of the phosphatidylinositol-anchored
form and the transmembrane form of the cell adhesion protein LFA-3. J. Biol.
Chem. 267:5663.
41. Bamezai, A., V. S. Goldmacher, and K. L. Rock. 1992. Internalizaton of glycosyl-phosphatidylinositol (GPI)-anchored lymphocyte proteins. II. GPI-anchored
and transmembrane molecules internalize through distinct pathways. Eur. J. Immunol. 22:15.
42. Bamezai, A., V. Goldmacher, H. Reiser, and K. L. Rock. 1989. Internalization of
phosphatidylinositol-anchored lymphocyte proteins. I. Documentation and potential significance for T cell stimulation. J. Immunol. 143:3107.
43. Butikofer, P., F. A. Kuypers, C. M. Xu, D. T. Y. Chiu, and B. Lubin. 1989.
Enrichment of two glycosyl-phosphatidylinositol-anchored proteins, acetylcholinesterase and decay accelerating factor, in vesicles released from human red
blood cells. Blood 74:1481.
44. Whitlow, M., K. Iida, P. Marshall, R. Silber, and V. Nussenzweig. 1993. Cells
lacking glycan phosphatidylinositol-linked proteins have impaired ability to vesiculate. Blood 81:510.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
17. Le, P. T., L. W. Vollger, B. F. Haynes, and K. H. Singer. 1990. Ligand binding
to the LFA-3 cell adhesion molecule induces IL-1 production by human thymic
epithelial cells. J. Immunol. 144:4541.
18. Webb, D. S. A., Y. Shimizu, G. A. Van Seventer, S. Shaw, and T. L. Gerrard.
1990. LFA-3, CD44, and CD45: physiologic triggers of human monocyte TNF
and IL-1 release. Science 249:1295.
19. Diaz-Sanchez, D., S. Chegini, K. Zhang, and A. Saxon. 1994. CD58 (LFA-3)
stimulation provides a signal for human isotype switching and IgE production
distinct from CD40. J. Immunol. 153:10.
20. Dustin, M. L., P. Selvaraj, R. J. Mattaliano, and T. A. Springer. 1987. Anchoring
mechanisms for LFA-3 cell adhesion glycoprotein at membrane surface. Nature
329:846.
21. Seed, B. 1987. An LFA-3 cDNA encodes a phospholipid-linked membrane protein homologous to its receptor CD2. Nature 329:840.
22. Wallner, B. P., A. Z. Frey, R. Tizard, R. J. Mattaliano, C. Hession, M. E. Sanders,
M. L. Dustin, and T. A. Springer. 1987. Primary structure of lymphocyte function-associated antigen 3 (LFA-3): the ligand of the T lymphocyte CD2 glycoprotein. J. Exp. Med. 166:923.
23. Sanchez-Madrid, F., A. M. Krensky, C. F. Ware, E. Robbins, J. L. Strominger,
S. J. Burakoff, and T. A. Springer. 1982. Three distinct antigens associated with
human T lymphocyte-mediated cytolysis: LFA-1, LFA-2, and LFA-3. Proc. Natl.
Acad. Sci. USA 79:7489.
24. Barnstable, C. J., W. F. Bodmer, G. Brown, G. Galfre, C. Milstein,
A. F. Williams, and A. Ziegler. 1978. Production of monoclonal antibodies to
group A erythrocytes, HLA and other human cell surface antigens—new tools for
genetic analysis. Cell 14:9.
25. Leo, O., M. Foo, D. H. Sachs, L. E. Samelson, and J. A. Bluestone. 1987. Identification of a monoclonal antibody specific for a murine T3 polypeptide. Proc.
Natl. Acad. Sci. USA 84:1374.
26. Keisari, Y. 1992. A colorimetric microtiter assay for the quantitation of cytokine
activity on adherent cells in tissue culture. J. Immunol. Methods 146:155.
27. Cooper, J. A., B. M. Sefton, and T. Hunter. 1983. Detection and quantification of
phosphotyrosine in proteins. Methods Enzymol. 99:387.
28. Klausner, R. D., J. J. O’Shea, H. Loung, P. Ross, J. A. Bluestone, and
L. E. Samelson. 1987. T. cell receptor tyrosine phosphorylation: variable coupling for different activating ligands. J. Biol. Chem. 262:12654.
29. Hollander, N., P. Selvaraj, and T. A. Springer. 1988. Biosynthesis and function
of LFA-3 in human mutant cells deficient in phosphatidylinositol-anchored proteins. J. Immunol. 141:4283.
30. Miyata, T., J. Takeda, Y. Iida, N. Yamada, N. Inoue, M. Takahashi, K. Maeda,
T. Kitani, and T. Kinoshita. 1993. The cloning of PIG-A, a component in the
early steps of the GPI-anchor biosynthesis. Science 259:1318.
31. Medof, M. E., T. Kinoshita, and V. Nussenzweig. 1984. Inhibition of complement activation on the surface of cells after incorporation of decay-accelerating
factor (DAF) into their membranes. J. Exp. Med. 160:1558.
ASSOCIATION OF CD58 WITH PROTEIN KINASES