| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Pathology, National Children’s Medical Research Center, and
Department of Pediatrics, Juntendo University, School of Medicine, Tokyo, Japan
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The GEM is a local variation of the plasma membrane rich in glycosphyngolipid (GSL), cholesterol, sphingomyelin, and GPI-anchored proteins that organizes the functional microdomains (7, 8, 9). Biochemical and immunohistological observations have revealed the enrichment of a wide variety of signal transducing molecules, including Src family PTKs, trimeric G proteins, Ras (7, 8, 9), and linker for activation of T cells (10), in the GEM domains, suggesting their roles as platforms of signal transduction. Presently, it is assumed that cross-linking of GPI-anchored proteins within the external part of the GEM leads to coalescence of the microdomains and can cause simultaneous redistribution (approximation) of the Src family PTKs at their cytoplasmic side (1, 2). Consequently, Src family PTKs are activated via clustering-mediated autophosphorylation followed by phosphorylation of their substrates and thus triggering of signaling cascades (1, 2).
CD24, also referred to as heat-stable Ag (HSA) or nectadorin, is a member of the GPI-anchored protein family (11, 12, 13) that is expressed on several cell types, including B cells (11, 12, 13, 14, 15, 16, 17, 18). CD24 exhibits a differentiation-dependent expression pattern in B cells. During B cell ontogeny, CD24 is already expressed in very early stage B cell progenitors (19) and remains expressed on mature resting B cells. Resting B cells, such as the majority of B cells in circulating peripheral blood and mantle zone B cells in lymphoid follicles, express high levels of CD24. However, CD24 begins to disappear when B cells are activated and induced to further maturation (15, 17, 20, 21). In fact, expression of CD24 is markedly diminished in germinal center (GC) B cells in the secondary lymphoid follicles (20, 21, 22, 23), the site where activated B cells undergo terminal differentiation into Ab-secreting B cells (24, 25, 26, 27).
Both CD24+ and CD24- B cell subsets are present in GCs, with the latter being thought to be more differentiated (15, 17, 20, 21). Because the vast majority of GC B cells lack CD24 (20, 21, 22, 23), it was postulated that loss of CD24 expression is an essential step in the commitment of activated B cells to differentiate into Ab-secreting cells (13). Alternatively, expression of CD24 may be disadvantageous for the proliferation and/or further differentiation of activated B cells. GC B cells are entering a dramatic process that determines whether they will survive or die, changing their phenotypes quickly in a short period of time (20, 21, 24, 25, 26, 27). During this process, positive and negative signals for cell survival are believed to be cooperatively regulated by the Ag receptor and a number of accessory molecules (24, 25, 26, 27). Therefore, CD24 expression and disappearance may represent unknown aspects of this dramatic process.
Several lines of evidence suggest the functional involvement of CD24 in B cell development (14, 28, 29, 30). For example, Chappel et al. examined mouse B cells and demonstrated that the cross-linking of CD24 induces apoptosis in B cell precursors, including pro-B and pre-B cells (28, 29). They also observed that the cross-linking of CD24 does not induce apoptosis in mature resting B cells, but instead inhibits their ability to proliferate in response to anti-CD40 mAb and IL-4 (29). These findings suggest that CD24 is a potent negative regulator of B cell development and activation, whereas CD24 cross-linking has differential effects at several stages of B cell development (29).
Considering all of the above evidence together, CD24 probably mediates a negative signal for GC B cell survival via a GEM-mediated signaling mechanism, although the details of the mechanism are largely unknown. However, it has been well documented that GC B cells express a number of apoptosis-related molecules (31). When freshly isolated and placed in primary culture, multiple death-signaling cascades are activated, and hence GC B cells undergo rapid and spontaneous apoptosis (31, 32). Thus, primary cultures of GC B cells are unsuitable for characterizing the effect of a negative signal for cell survival operating via a specific molecule, such as CD24.
Burkitt’s lymphoma (BL) is a tumor that retains the properties of GC B cells (33, 34, 35, 36). Because BL cells undergo apoptosis upon cross-linking of the B cell receptor for Ag (BCR), they are often used to investigate the molecular and cellular signals for the cell death of B cells at the clonal level (35, 36, 37, 38). We previously reported that BL cells freshly obtained from patients consist of CD24+ and CD24- subtypes, and that CD24+ BL cells appear to be related to the CD24+ B cells in the GC (23). In the present study we chose BL cell lines expressing CD24 to investigate the functional role of CD24 in B cell development and its relation to GEM-mediated signaling. The results showed that Ab-mediated cross-linking of CD24 induces apoptosis in BL cells. Further evidence suggested that CD24 mediates a negative signal for B cell survival through a GEM-dependent mechanism that is closely related to the BCR-mediated apoptotic signal. The potent biological role of CD24-mediated apoptosis in B cell development is discussed.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CD24+ BL-derived cell lines P32/SH and Namalwa were obtained from the Japanese Cancer Research Resources Bank (Tokyo, Japan) and Institute for Fermentation (Osaka, Japan), respectively. The CD24- BL line Daudi from the Japanese Cancer Research Resources Bank was also used. Cells were cultured in RPMI 1640 supplemented with 10% FCS at 37°C in a humidified 5% CO2 atmosphere.
The mouse mAbs used were anti-CD24 (ALB-9) and anti-CD48
(J4.57) from Coulter/Immunotech (Westbrook, MA), anti-CD48 (5-4.8)
from Santa Cruz Biotechnology (Santa Cruz, CA), anti-CD55 (IA10)
and anti-CD59 (p282H19) from PharMingen (San Diego, CA),
anti-phosphotyrosine (4G10) from Upstate Biotechnology (Lake
Placid, NY), anti-Lyn (clone 42) and anti-ERK1 (MK12) from
Transduction Laboratories (Lexington, KY), anti-GM3 (M2590) from
Wako Pure Chemical Industries (Osaka, Japan), and anti-
-actin
from Seikagaku (Tokyo, Japan). Anti-CD3 (OKT-3), anti-CD45 (9.4),
and anti-µ (DA4.4) from American Type Culture Collection
(Manassas, VA) and anti-CD24 (L30) (22) were also
used. The rabbit polyclonal Abs used were anti-Lyn from Serotec
(Oxford, U.K.); anti-CD55, anti-Csk, and anti-Blk from
Santa Cruz Biotechnology; anti-phospho-specific mitogen-activated
protein (MAP) kinase, anti-phospho-specific stress-activated
protein (SAP) kinase, anti-SAP kinase,
anti-phospho-specific p38 MAP kinase, and anti-p38 MAP kinase
from New England Biolabs (Beverly, MA);
anti-Gi
from Calbiochem-Novabiochem (San
Diego, CA); and anti-µ from Jackson ImmunoResearch (West Grove,
PA). The goat anti-cholera toxin B subunit (CT-B) polyclonal Ab
from Calbiochem was also used. Secondary Abs, including
fluorescein-conjugated and enzyme-conjugated Abs, and
fluorescein-conjugated avidin were purchased from either Jackson
ImmunoResearch or Molecular Probes (Eugene, OR). Biotinylation of mAb
was performed as described previously (23). Conjugation of
Abs to polystyrene beads was performed as described previously
(39). To cross-link CD24, cells were cultured in the
presence of Abs as indicated. In these cases, Abs were dialyzed in PBS
before addition to culture to remove additives. The recombinant CT-B
was obtained from Calbiochem. All chemical reagents were obtained from
Sigma-Aldrich (St. Louis, MO) unless otherwise indicated.
Immunofluorescence study
The cells were stained with FITC-labeled mAbs and analyzed by flow cytometry (EPICS-XL, Coulter) as described previously (23). In some experiments cells were stained with a combination of Alexa Fluor 488-labeled anti-CD24 mAb and Alexa Fluor 543-labeled either anti-Lyn, anti-Blk, or anti-Csk polyclonal Ab, then examined by confocal microscopy (GB-200, Olympus, Tokyo, Japan). Briefly, after a 30-min incubation with anti-CD24 mAb (L30) on ice, cells were washed intensively with ice-cold PBS and incubated with Alexa Fluor 488-conjugated goat anti-mouse polyclonal Ab for 30 min on ice. After intensive washing with ice-cold PBS, cells were either left on ice as an unpatched control or incubated for 30 min at 37°C to induce patching as described previously (10, 40). After intensive washing and fixation with 3% paraformaldehyde in PBS, cells were further stained with a combination of rabbit anti-Lyn, anti-Blk, or anti-Csk polyclonal Ab followed by Alexa Fluor 543-conjugated goat anti-rabbit polyclonal Ab and examined using the appropriate filter set. Because Alexa Fluor 488-conjugated goat anti mouse Ab absorbed any cross-reactivity, no cross-reaction against rabbit Ab was detected (data not shown). Similarly, Alexa Fluor 543-conjugated goat anti-rabbit Ab showed no cross-reactivity against mouse Ab (data not shown). No cross-talk was detected between Alexa Fluor 488 and Alexa Fluor 543 channels (data not shown).
Immunoblotting, immunoprecipitation, and immunocomplex kinase assay
Immunoblotting was performed as described previously (41). Briefly, cell lysates were prepared by solubilizing the cells in lysis buffer (containing 20 mM Na2PO4 (pH 7.4), 150 mM NaCl, 1% Triton X-100, 10 trypsin inhibitory units/ml aprotinin, 5 mM PMSF, 100 mM NaF, and 2 mM Na3VO4). After centrifugation, supernatants were obtained, and the protein concentration of each cell lysate was determined with a Bio-Rad protein assay kit (Bio-Rad, Hercules, CA). Fifty micrograms of each cell lysate was electrophoretically separated on an SDS-polyacrylamide gel and transferred to a nitrocellulose membrane using a semidry Transblot system (Bio-Rad). After blocking, the membranes were incubated with the appropriate combination of primary and secondary Abs as indicated, washed intensively, then examined using the enhanced chemiluminescence reagent system (ECL, Amersham, Aylesbury, U.K.). The results obtained from a 1-min exposure of the ECL-treated membrane to film are presented.
For immunoprecipitation, 500 µg of the cell lysates were incubated with 1 µg of the indicated Ab and 30 µl of 50% protein G-agarose (Roche, Mannheim, Germany) for 1 h. After intensive washing, the immunoprecipitates were separated by electrophoresis and analyzed as described above. Immunocomplex kinase assay was also performed using enolase, as exogenous substrate, as described previously (41).
Detection of apoptosis
To detect apoptosis, cells were assayed for DNA ladder formation by gel electrophoresis. After treatment of cells as indicated, DNA was extracted from the cells, separated by 1% agarose gel electrophoresis, and examined under UV light as described previously (37). To quantitate the incidence of apoptotic cells, cells were stained with FITC-labeled annexin V using the MEBCYTO Apoptosis Kit (Medical & Biological Laboratories, Nagoya, Japan) and then analyzed by flow cytometry according to the manufacturer’s protocol.
Sedimentation of the GEM fraction and lipid analysis
The GEM fractions were prepared by linear sucrose gradient centrifugation as described previously (40, 42). Briefly, 1 x 108 cells were lysed and homogenized in Tris-buffered saline (25 mM Tris (pH 7.5) and 150 mM NaCl) containing 1% Triton X-100, 1 mM PMSF, 1 mM Na3Vo4, and aprotinin (10 trypsin inhibitory units/ml). Cell extracts were adjusted to 40% sucrose in Tris-buffered saline. A linear gradient (5–30% sucrose in Tris-buffered saline) was formed above the lysate and centrifuged at 39,000 rpm for 18 h at 4°C in a Beckman SW40Ti rotor (Palo Alto, CA). In some experiments cells were biotinylated before homogenization as described previously (42). The lipids extracted from the gradient fractions as described previously (40, 42) were separated on a high performance TLC plate (Merck, Darmstadt, Germany), followed by blotting to polyvinylidene difluoride membrane (Millipore, Bedford, MA) and immunostained with anti-GM3 mAb as described previously (40, 42).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We first examined whether BL cell lines express CD24. Among eight
BL lines tested, including P32/SH, Namalwa, Ramos, Daudi, CA-46,
Jiyoyie, EB-3, and SCC-3, only the first two lines expressed CD24. As
shown in Fig. 1
, both P32/SH and Namalwa
lines expressed high levels of CD24 by flow cytometry. In addition,
other GPI-anchored proteins, including CD48, CD55, and CD59 (Fig. 1),
but not CD90 (data not shown), were expressed on both lines. Flow
cytometric analysis also revealed that both lines showed a similar
pattern of expression of B cell differentiation Ags, including surface
IgM (sIgM; Fig. 1), CD19, CD20, CD72, CD80, CD81, and HLA-DR (data not
shown). A GC B cell-specific marker, CD10 (20, 21, 23, 36), and a protein tyrosine phosphatase, CD45, were also
expressed on both lines (Fig. 1). We used these two
CD24+ BL lines in the following experiments.
|
Next we tested whether cross-linking of CD24 mediated by specific
Abs induces any physiological responses in BL cells. When P32/SH cells
treated with a combination of anti-CD24 mAb, L30, and rabbit
secondary polyclonal anti-mouse Ig Ab for 24 h were examined
morphologically by light microscopy, a portion of cells exhibited
cleaved nuclei, a finding typical of apoptosis (Fig. 2A). No such figures were
observed in untreated cells (data not shown). The following results
clearly show that cross-linking of CD24 indeed induced apoptosis in BL
cells. First, DNA prepared from P32/SH cells on which CD24 is
cross-linked showed oligonucleosomal ladder fragmentation on agarose
gel electrophoresis (Fig. 2B). Second, cleavage of the
nuclear DNA was also confirmed by detection of subploid cells with
propidium iodide staining (data not shown). Third, there was a
significant increase in cells binding to annexin V after CD24
cross-linking by immobilized anti-CD24 mAb on polystyrene beads
(Fig. 2C).
|
A, when
P32/SH cells were treated with the anti-CD24 mAb L30 alone for
24 h, <10% of the cells underwent apoptosis as assessed by
annexin V binding. However, in the presence of the secondary rabbit
polyclonal anti-mouse Ig Ab L30 induced the appearance of a
significant number of annexin V-bound cells (Fig. 3A),
suggesting that strong cross-linkage of CD24 is required to induce
apoptosis in these cells. The fact that either a combination of
biotinylated L30 and streptavidin (Fig. 4B) or L30 immobilized
on polystyrene beads (Fig. 2
C) also induced a comparable
level of apoptosis further supports the above idea. We also examined
the effect of ALB-9, another anti-CD24 mAb, and observed identical
results (Fig. 3A). These phenomena were specific to CD24,
because anti-CD45, which binds to BL cells, did not induce
apoptosis (Fig. 3A). Isotype-matched control mouse Ig (Fig. 2C) or anti-CD3 did not induce apoptosis (Fig. 3A). We similarly examined Namalwa cells and observed mostly
identical results (Fig. 3B). By contrast, when
CD24- Daudi BL cells, in which signaling for
apoptosis can occur by BCR cross-linking, were similarly examined,
treatment with a combination of L30 and rabbit anti-mouse Ig Ab
failed to induce apoptosis (Fig. 3C). Thus, these data
suggest that sufficient CD24 cross-linking specifically induces
apoptosis in CD24+ BL cells, and that this is not
due to the nonspecific binding of either mouse Ig or secondary rabbit
polyclonal anti-mouse Ig Ab.
|
|
Next we examined whether the induction of apoptosis is specific
for CD24 or is a common feature of GPI-anchored proteins expressed on
these BL cells. As shown in Fig. 5, when
P32/SH cells were treated with a combination of anti-CD48 mAb and
secondary rabbit anti-mouse Ig, a number of annexin V-binding cells
comparable to that obtained with anti-CD24 mAb was induced. In
contrast, the mAbs against either CD55 or CD59 could only induce a low
number of annexin V-binding cells, even in the presence of secondary
rabbit anti-mouse Ig Ab (Fig. 5). These data suggest that induction
of apoptosis upon cross-linking is a common feature of GPI-anchored
proteins expressed on these BL cells, while the magnitude of cell death
is variable in each case.
|
It was reported that sufficiently strong cross-linking of BCR also
induces apoptosis in BL cells (37, 38). Therefore, we
tested the synergism between cross-linking of BCR and that of CD24 on
the induction of apoptosis. Although the DA4.4 mAb against human µ
heavy chain was able to induce marked apoptosis in P32/SH cells at full
dose (5 µg/ml), it did not induce apoptosis at a suboptimal dose (0.1
µg/ml) even in the presence of secondary rabbit anti-mouse Ig Ab
(Fig. 4A, compare columns 3, 7,
9, and 10). Similarly, a suboptimal dose (0.5
µg/ml) of anti-CD24 mAb, L30, also failed to induce significant
apoptosis in P32/SH cells (Fig. 4A, columns 2 and
6). However, when suboptimal doses of both DA4.4 and L30
were mixed, a significant level of apoptosis was induced in P32/SH
cells with the help of secondary rabbit anti-mouse Ig Ab (Fig. 4A, columns 6–8). The data indicate a synergism
between cross-linking of CD24 and that of BCR on the effect of
apoptosis induction in BL cells.
To further examine whether cross-linking between CD24 and BCR is
necessary for the synergistic effect on apoptosis induction, we used
the biotin-avidin system. As presented in Fig. 4B, a
combination of a full dose of biotinylated L30 (20 µg/ml) and avidin
(1 µg/ml) is sufficient to induce apoptosis in P32/SH cells
(columns 7 and 8). Although a suboptimal dose of
biotinylated L30 (5 µg/ml) induced a low level of apoptosis in the
presence of avidin (Fig. 4B, columns 3 and
4), simultaneous addition of a suboptimal dose (0.1 µg/ml)
of DA4.4 remarkably enhanced the induction of apoptosis (Fig. 4B, columns 5 and 6). Based on these
results, we concluded that independent cross-linking of each molecule
is sufficient to induce synergistic induction of apoptosis and that
cross-linking between CD24 and BCR is unnecessary.
Distribution of CD24 and other molecules in linear sucrose gradient centrifugation
GPI-anchored proteins are thought to be localized in the GEM
domain. To investigate the participation of GEM in CD24-mediated
apoptosis, we examined the distribution of CD24 in linear sucrose
gradient centrifugation. When cell lysates prepared from biotinylated
P32/SH cells were fractionated, most of the biotinylated proteins were
liberated with 1% Triton X-100 from the plasma membrane and thus
recovered in fractions 11–13 as high density detergent-soluble
fractions (DS) (Fig. 6A). Some
were not solubilized from plasma membrane and thus were recovered in
buoyant, low density fractions 5–9, which include the GEM (Fig. 6A). By immunoprecipitation, most CD24 molecules were
recovered in buoyant, low density fractions 6 and 7 (Fig. 6B). Consistent with previous reports, immunoprecipitated
CD24 molecules appeared as a broad band on SDS-PAGE gels (5, 16, 43), because of their extensive glycosylation (13, 18). By contrast, when CD24- Daudi BL
cells were similarly examined, anti-CD24 mAb L30 failed to
precipitate any biotinylated proteins from either the GEM or DS (data
not shown), suggesting the specificity of this mAb. In contrast, CD45
molecules, which have been reported to be found predominantly in the DS
(10, 44), were immunoprecipitated only from the DS in both
P32/SH cells (Fig. 6B) and Daudi cells (data not shown).
These findings indicate that the failure to immunoprecipitate CD24 from
the DS is not due to disruption of immunoprecipitation in the presence
of high sucrose concentrations and that CD24 molecules are indeed
distributed only in the GEM fractions.
|
C), and Syk
(data not shown), were only recovered in DS. In contrast, approximately
half of all Lyn protein, an Src family PTK, was recovered in the GEM
(Fig. 6C). Furthermore, immunoprecipitation revealed that a
portion of Lyn was detected in anti-CD24 immunoprecipitates from
the GEM fractions (Fig. 7). Because no
Lyn was detected in control mouse Ig immunoprecipitates from the same
fractions (data not shown), the results indicate that CD24 and Lyn are
specifically associated in the GEM.
|
B, when cell lysates prepared from P32/SH
cells treated with the anti-CD24 mAb L30 immobilized on polystyrene
beads were fractionated, the distribution of CD24 shifted to a higher
density site, and it was recovered mainly in fractions 7 and 8. It is
noteworthy that the distribution of Lyn in sucrose gradient
centrifugation also shifted similarly (Fig. 6C). In
contrast, cross-linking of CD24 did not affect the other proteins,
including Blk, Csk (Fig. 6C), Syk, or CD45 (data not shown),
all of which remained in the DS. No such shift was induced by
anti-CD3 or anti-CD45 mAb immobilized on polystyrene beads
(data not shown). From these results, mobility shift of CD24 and Lyn by
CD24 cross-linking was regarded to be a phenomenon specific to CD24 and
was not due to nonspecific binding of Ab.
To assess the biochemical basis of the difference between GEM from
untreated and CD24 cross-linked cells, we examined the distribution of
the other presumably GEM-associated molecules in sucrose gradient
centrifugation before and after treatment with anti-CD24 mAb. GSL
is a major constituent of the GEM domain. In P32/SH cells, GM3 was
found to be the most abundantly expressed GSL as assessed by lipid
analysis (data not shown). As shown in Fig. 6D, TLC blotting
analysis showed that the majority of GM3 was recovered in the low
density GEM fractions. A dually acylated heterodimeric G protein
subunit, Gi, is known to reside in GEM
(45, 46, 47), and indeed, immunoblotting analysis revealed
that Gi was recovered only in low density
fractions 5–7 after sucrose gradient centrifugation (Fig. 6D). Other GPI-anchored proteins, CD48 and CD55, were also
recovered only in GEM (Fig. 6D). However, when the cells
were CD24 cross-linked the distribution of all these molecules shifted
to a higher density site, and after sucrose gradient centrifugation
they were recovered in fractions 6–8 (Fig. 6D), the same as
CD24 (Fig. 6B). The results indicate that all GEM-associated
molecules show the same distribution shift after sucrose gradient
centrifugation following CD24 cross-linking.
We further examined whether the association between CD24 and Lyn in the
GEM fractions was affected by CD24 cross-linking. As shown in Fig. 7C, the total amount of Lyn protein precipitated with
anti-Lyn mAb did not change significantly after CD24 cross-linking.
However, the recovery of Lyn in anti-CD24 immunoprecipitates was
markedly increased after CD24 cross-linking (Fig. 7A).
Moreover, CD24 cross-linking resulted in the enhanced kinase activity
of Lyn in GEM fractions as revealed by increased transphosphorylation
of enolase, an exogenous substrate, by in vitro kinase assay (Fig. 7D).
Colocalization of CD24 and Lyn, but not Blk, in cell membrane after cross-linking of CD24
Because biochemical data have indicated that CD24 is closely
correlated with Lyn, but not with Blk or Csk, in the GEM domain, we
next examined the correlation between CD24 and those PTKs in the cell
membrane by using immunostaining and confocal microscopy. In cells left
on ice throughout the staining process, CD24 and those PTKs were evenly
distributed (Fig. 8, B,
D, and F). However, when the cells were incubated
at 37°C, CD24 became concentrated to distinct patches within the
membrane (Fig. 8, A, C, and E).
Significantly, Lyn was found to clearly colocalize with the
CD24-patches as presented in a overlay picture in Fig. 8A.
In contrast, Blk and Csk did not colocalize with the CD24 patches (Fig. 8, C and E). These results are consistent with
the biochemical data and indicate that Lyn, but not Blk and Csk, is
preferentially colocalized with CD24 patches after Ab-mediated CD24
cross-linking.
|
Upon cross-linking the BCR has been shown to move into GEM in
mouse and human B cells (44, 47, 48, 49). Therefore, we
examined whether the same was true in the P32/SH cells used in the
present study. As shown in Fig. 9, when
cell lysates prepared from P32/SH cells were fractionated, µ heavy
chain was recovered only in DS after sucrose gradient centrifugation.
However, after BCR cross-linking by treating cells with the
anti-µ heavy chain mAb immobilized on polystyrene beads, a small
portion of µ heavy chain was translocated in low density fractions 5
and 6 (Fig. 9). Gi, a marker protein for GEM,
was also recovered in fractions 5–7 regardless of BCR cross-linking
(Figs. 6D and 9). The results suggest that BCR cross-linking
induces recruitment of a fraction of BCR itself in the GEM fraction,
but that it does not affect the distribution of GEM on sucrose gradient
centrifugation.
|
, CD24 cross-linking
also resulted in translocation of a small portion of the µ heavy
chain in the low density fractions. However, in contrast to BCR
cross-linking the translocated µ heavy chain was recovered in
fractions 6 and 7, where the other GEM-associated molecules were
located after CD24 cross-linking (Fig. 6, B and
D). Treatment with CT-B induces apoptosis in BL cells
As demonstrated in this study, cross-linking of GPI-anchored
proteins that reside in GEM induces apoptosis in BL cells. We also
investigated whether the induction of apoptosis is specific to
GPI-anchored proteins or is a common feature of GEM-resident molecules.
To test this, we selected CT-B. CT-B has been well characterized as a
specific ligand for GSL GM1 that resides in the GEM domain (10, 39). Cross-linking of GEM with CT-B has been reported to induce
intracellular signaling events in T cells via a GEM-mediated mechanism
(10, 39). Flow cytometric analysis revealed binding of
CT-B to P32/SH cells (Fig. 10A), indicating that this
cell line expresses GM1. As shown in Fig. 10B, when P32/SH
cells were treated with CT-B alone for 24 h, the percent increase
in apoptotic cells, assessed by annexin V binding, was not significant.
However, in the presence of the goat anti-CT-B Ab, CT-B induced a
significant increase in number of annexin V-bound cells (Fig. 10B), indicating that sufficient cross-linking of GM1 also
can induce apoptosis in BL cells. We also attempted to determine
whether GM3 cross-linking with specific Ab also induces apoptosis of BL
cells, but flow cytometric analysis revealed that the anti-GM3 mAb
we used for TLC blotting in this study does not recognize the
extracellular portion of GM3 and thus is not useful for cross-linking
(data not shown).
|
In the cases of other cell types, it has been reported that
cross-linking of CD24 as well as other GPI-anchored proteins results in
intracellular signaling events such as an activation of protein kinases
accompanied by an increased level of tyrosine phosphorylation of
intracellular proteins (1, 2, 4, 6, 12). Thus, we
attempted to observe similar changes in BL cells expressing CD24.
Inconsistent with the previous reports, no significant change in
tyrosine phosphorylation level of total cellular proteins by CD24
cross-linking detected in immunoblotting analysis using
anti-phosphorylated tyrosine Ab (data not shown). However, when
individual molecules were examined by immunoblotting using
phospho-specific Abs, clear elevations of phosphorylations of ERK1, SAP
kinase, and p38 MAP kinase were detected (Fig. 11), suggesting that CD24 cross-linking
results in the activation of these kinases.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
However, inconsistent with our observation, it has been reported that no impairment of immune function could be detected in CD24-deficient mice in a variety of immunization and infection models, indicating that the peripheral immune system is not grossly affected by the absence of CD24 (30). Although the exact reason remains unclear, it has been speculated that compensatory changes and/or functionally overlapping systems might be responsible for the mild changes in phenotype of CD24-deficient mice, as is the case in regard to numerous other genes examined by targeted mutations in mice (30). Because we have observed that cross-linking of other GPI-anchored proteins, including CD48, CD55, and CD59, also leads apoptosis in BL cells, it is possible that other molecules compensate for the function of CD24 in CD24-deficient mice.
In addition to the GPI-anchored proteins, GSL GM1 also induces apoptosis in BL cells upon cross-linking with its specific ligand, CT-B. We and others recently reported that cross-linking of Gb3, another GSL expressed on a portion of GC B cells and BL cells, also induces apoptosis upon cross-linking (40, 51, 52). Considering the fact that all these molecules reside in the GEM domain, a close correlation has been suggested between apoptosis induction and cross-linking of GEM-associated molecules. However, Chappel et al. demonstrated that the effect of CD24 cross-linking differs according to the stage of B cell development (29). They observed that CD24 cross-linking induces apoptosis in B cell precursors in the mouse system, but not in mature resting B cells. In view of their report, it is possible that cross-linking of GEM may have different effects at different stages of B cell development.
We also demonstrated in the present study that GEM-mediated signaling is involved in the CD24-mediated apoptotic process. Increased coisolation of Lyn in CD24 immunoprecipitates from a biochemically separated GEM fraction after CD24 cross-linking suggests a close correlation between these molecules in the GEM domain. A microscopic cocapping of CD24 and Lyn in the cell membrane after CD24 cross-linking further supports this idea. Together with the evidence of elevated Lyn kinase activity in the GEM fraction mediated by CD24 cross-linking, it is likely that CD24 cross-linking mediates intracellular signals that lead to apoptosis of BL cells through activation of Lyn kinase in a GEM-dependent manner. It is noteworthy that Blk, another member of the Src family of PTK, was not recovered in the GEM fraction after sucrose gradient centrifugation. Microscopic observation also revealed that Blk did not colocalize with CD24 after cross-linking of CD24. Consistently, Sammar et al. reported that CD24 is associated with Lck, Hck, and Lyn, but not Fyn, in mouse monocytic cells (6). The data indicate the distinct detergent solubility as well as the distinct role in GEM-mediated signaling of each Src family member despite their structural and functional similarity.
Interestingly, we observed that the cross-linking of CD24 leads to altered distribution of the GEM fraction in sucrose gradient centrifugation, and the recovery of GEM-associated molecules shifted to a higher density site, suggesting an increase in detergent solubility of the molecular complex of the GEM domain. The precise explanation for this phenomenon is not known at this moment. However, because the buoyancy and the detergent insolubility of the GEM fraction on sucrose density gradients are mainly due to the properties of the lipids that organize GEM domains (53), we speculate that the cross-linking of CD24 induces aggregation of the GEM domains involving a large number of proteins, which leads to higher detergent solubility. Recently, Harder et al. reported that cross-linking of GPI-anchored CD59 protein induces tyrosine phosphorylation-dependent accumulation of filamentous actin into GEM patches in Jurkat T cells, suggesting the interaction of the actin cytoskeleton with GEM domains in the course of GEM-mediated cell signaling (54). Additional experiments are clearly necessary, but their observations may support our hypothesis.
Further studies are required to elucidate whether the in vitro events
mediated by Ab-based cross-linking of GPI-anchored proteins, including
CD24-mediated apoptosis in BL cells, mimic any physiological phenomena.
Recently, however, data have accumulated on the physiological roles of
GPI-anchored proteins in various aspects of cell biology. One possible
role of GPI-anchored proteins is as machinery for immunoreceptor
signaling. T cell clones deficient in GPI anchor biosynthesis showed
impaired TCR signaling (55) and reduced CD3 
-chain
phosphorylation (56). In agreement with these
observations, it was reported that GEM domains are essential for TCR,
BCR, and Fc
receptor signaling (44, 47, 48, 49, 57, 58, 59, 60).
Thus, it was postulated that immediately after ligation the aggregated
immunoreceptors become associated with the GPI-anchored proteins and
thus become approximated to Src family PTKs that can phosphorylate the
immunoreceptor family tyrosine-based activation motif on the receptors,
leading to intracellular signaling (1, 2, 60). The
evidence showing that a portion of BCR is translocated into the GEM
domain following cross-linking, as presented in previous studies
(44, 47, 48, 49) and in this report, seems to support this
idea. Translocation of some BCR into the GEM domain following CD24
cross-linking, as demonstrated in this study, also suggests a close
correlation between the GEM domain and the BCR-mediated signaling
system.
Alternatively, GPI-anchored proteins themselves can act as a receptor for natural ligands. For example, CD14 expressed on monocytes is a receptor for the bacterial cell wall component LPS and mediates cytokines secretion. CD16b expressed on neutrophils is a low affinity receptor for IgG. CD62P, also called P-selectin, has been identified as a CD24 ligand (61), and it was reported that CD24 mediates rolling of breast cancer cells on immobilized CD62P (62). These data suggest that CD24 mediates the apoptotic signal in B cells upon ligand stimulation. However, our preliminary study revealed that the soluble form of CD62P could not bind effectively to CD24+ BL cells (data not shown). Additional experiments, including evaluation of the effect of immobilized CD62P and the discovery of an alternate ligand, are now underway.
Although the downstream cascade of CD24-mediated apoptotic signaling remains to be elucidated, our data suggest the involvement of MAP kinases in this process. We showed that cross-linking of CD24 induces the activation of MAP kinase family members, including ERK1, SAP kinase, and p38 MAP kinase. Because Src family PTKs are located upstream of MAP kinase cascades in several receptor signaling systems (63, 64, 65, 66), these activation events are likely to be mediated by Lyn kinase that is activated in GEM domain after cross-linking of CD24. Recently, it was shown that the activation of SAP kinase and/or p38 MAP kinase mediates apoptotic signaling in a variety of cell types (38, 67, 68). For example, Sugawara et al. showed that the p38 MAP kinase signaling pathway is critically involved in inducing negative selection of developing T cells using mouse thymocytes (68). In the case of B cells, Graves et al. reported that delayed and sustained activation of both SAP kinase and p38 MAP kinase is an essential requirement for BCR-induced apoptosis using the human B lymphoma B104 cell line (38). Based on these results, it is possible that both SAP kinase and p38 MAP kinase are involved in the CD24-mediated apoptotic process in BL cells. Further study of how both kinases participate in this apoptotic process is now underway.
In conclusion, the findings described here indicate that CD24 induces apoptosis in BL cells via a GEM-mediated signaling system. Additional studies are clearly necessary, but our finding provides a model for the study of death of human B cells, particularly focusing on the roles of GEM and its constituents.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Junichiro Fujimoto, Department of Pathology, National Children’s Medical Research Center, 3-35-31 Taishido, Setagaya-ku, Tokyo 154-8509, Japan.
3 Abbreviations used in this paper: PTK, protein tyrosine kinase; GEM, glycolipid-enriched membrane; GSL, glycosphyngolipid; HSA, heat-stable Ag; BCR, B cell receptor for Ag; GC, germinal center; BL, Burkitt’s lymphoma; DS, detergent-soluble fractions; MAP, mitogen-activated protein; SAP, stress-activated protein; CT-B, cholera toxin B subunit.
Received for publication August 1, 2000. Accepted for publication February 16, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
subunit (Gi1) and c-Src tyrosine kinase to caveolae membranes: clarifying the role of N-myristoylation. Cell. Mol. Biol. 3:293.
RI to the Rac1-JNK pathway. J. Biol. Chem. 272:10751.This article has been cited by other articles:
|
|
 |
|
  E. Sagiv, A. Starr, U. Rozovski, R. Khosravi, P. Altevogt, T. Wang, and N. Arber Targeting CD24 for Treatment of Colorectal and Pancreatic Cancer by Monoclonal Antibodies or Small Interfering RNA Cancer Res., April 15, 2008; 68(8): 2803 - 2812. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Jin, Q. Tao, J. Peng, H. M. Soo, W. Wu, J. Ying, C. R. Fields, A. L. Delmas, X. Liu, J. Qiu, et al. DNA methyltransferase 3B (DNMT3B) mutations in ICF syndrome lead to altered epigenetic modifications and aberrant expression of genes regulating development, neurogenesis and immune function Hum. Mol. Genet., March 1, 2008; 17(5): 690 - 709. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Yoon, A. Terada, and H. Kita CD66b Regulates Adhesion and Activation of Human Eosinophils J. Immunol., December 15, 2007; 179(12): 8454 - 8462. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Roda-Navarro and H. T. Reyburn Intercellular protein transfer at the NK cell immune synapse: mechanisms and physiological significance FASEB J, June 1, 2007; 21(8): 1636 - 1646. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. M. Popovici, N. K. Betzler, M. S. Krause, M. Luo, J. Jauckus, A. Germeyer, S. Bloethner, A. Schlotterer, R. Kumar, T. Strowitzki, et al. Gene Expression Profiling of Human Endometrial-Trophoblast Interaction in a Coculture Model Endocrinology, December 1, 2006; 147(12): 5662 - 5675. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. C. Smith, G. Oxford, Z. Wu, M. D. Nitz, M. Conaway, H. F. Frierson, G. Hampton, and D. Theodorescu The Metastasis-Associated Gene CD24 Is Regulated by Ral GTPase and Is a Mediator of Cell Proliferation and Survival in Human Cancer Cancer Res., February 15, 2006; 66(4): 1917 - 1922. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Schabath, S. Runz, S. Joumaa, and P. Altevogt CD24 affects CXCR4 function in pre-B lymphocytes and breast carcinoma cells J. Cell Sci., January 15, 2006; 119(2): 314 - 325. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Gu, J. Yamashiro, E. Kono, and R. E. Reiter Anti-Prostate Stem Cell Antigen Monoclonal Antibody 1G8 Induces Cell Death In vitro and Inhibits Tumor Growth In vivo via a Fc-Independent Mechanism Cancer Res., October 15, 2005; 65(20): 9495 - 9500. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Weichert, C. Denkert, M. Burkhardt, T. Gansukh, J. Bellach, P. Altevogt, M. Dietel, and G. Kristiansen Cytoplasmic CD24 Expression in Colorectal Cancer Independently Correlates with Shortened Patient Survival Clin. Cancer Res., September 15, 2005; 11(18): 6574 - 6581. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. C. Jung, W. S. Park, H. J. Kim, E. Y. Choi, M.-C. Kook, H.-W. Lee, and Y. Bae TCR-Independent and Caspase-Independent Apoptosis of Murine Thymocytes by CD24 Cross-Linking J. Immunol., January 15, 2004; 172(2): 795 - 802. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Wu, S.-T. Pang, L. Sahlin, A. Blanck, G. Norstedt, and A. Flores-Morales Gene Expression Profiling of the Effects of Castration and Estrogen Treatment in the Rat Uterus Biol Reprod, October 1, 2003; 69(4): 1308 - 1317. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. G. Kohlgraf, A. J. Gawron, M. Higashi, J. L. Meza, M. D. Burdick, S. Kitajima, D. L. Kelly, T. C. Caffrey, and M. A. Hollingsworth Contribution of the MUC1 Tandem Repeat and Cytoplasmic Tail to Invasive and Metastatic Properties of a Pancreatic Cancer Cell Line Cancer Res., August 15, 2003; 63(16): 5011 - 5020. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Yuan, M. Liang, Z. Yang, E. Rute, N. Taylor, M. Olivier, and A. W. Cowley Jr. Gene expression reveals vulnerability to oxidative stress and interstitial fibrosis of renal outer medulla to nonhypertensive elevations of ANG II Am J Physiol Regulatory Integrative Comp Physiol, May 1, 2003; 284(5): R1219 - R1230. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Taguchi, N. Kiyokawa, K. Mimori, T. Suzuki, T. Sekino, H. Nakajima, M. Saito, Y. U. Katagiri, N. Matsuo, Y. Matsuo, et al. Pre-B Cell Antigen Receptor-Mediated Signal Inhibits CD24-Induced Apoptosis in Human Pre-B Cells J. Immunol., January 1, 2003; 170(1): 252 - 260. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Hase, Y. Kanno, H. Kojima, C. Morimoto, K. Okumura, and T. Kobata CD27 and CD40 Inhibit p53-independent Mitochondrial Pathways in Apoptosis of B Cells Induced by B Cell Receptor Ligation J. Biol. Chem., November 27, 2002; 277(49): 46950 - 46958. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Polyak and J. P. Deans Alanine-170 and proline-172 are critical determinants for extracellular CD20 epitopes; heterogeneity in the fine specificity of CD20 monoclonal antibodies is defined by additional requirements imposed by both amino acid sequence and quaternary structure Blood, May 1, 2002; 99(9): 3256 - 3262. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Ayllon, A. Fleischer, X. Cayla, A. Garcia, and A. Rebollo Segregation of Bad from Lipid Rafts Is Implicated in the Induction of Apoptosis J. Immunol., April 1, 2002; 168(7): 3387 - 3393. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Alam and M. Gorska Genomic Microarrays . Arraying Order in Biological Chaos? Am. J. Respir. Cell Mol. Biol., October 1, 2001; 25(4): 405 - 408. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
