| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
Biological Sciences, State University of New York, Buffalo, New York 14260 [G. S. G., M. M., B. J. N.]; Department of Anatomy and Cell Biology, Medical Sciences Building, University of Western Ontario, London, Ontario, ON N6A 5C1 Canada [J. F. B., C. C. G. N.]; Department of Biochemical Cell Research, Tokyo Metropolitan Institute of Medical Science, Tokyo 113, Japan [Y. T., Y. S.]; Protein Chemistry Core Facility, Columbia University College of Physicians and Surgeons, New York, New York 10032 [M. A. G.]; Oncology Division, Hoffman-La Roche Inc., Nutley, New Jersey 07110 [R. N.]; Cell Signaling Technology, 1558 Cummings Center, Beverly, Massachusetts 01915 [Y. T.]; Unit of Multistage Carcinogenesis, International Agency for Research on Cancer, 69372 Lyon cedex 08, France [G. S. G., Y. O., H. Y.]; and Experimental Pathology and Chemotherapy Division, National Cancer Center Research Institute, Tokyo 104, Japan [G. S. G., H. T.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A role for gap junctions in transformed cell growth control was first proposed over 30 years ago by Loewenstein and Kanno (12) . Since that time, it has become increasingly clear that Cxs can indeed suppress transformed cell growth. Many transforming and mitogenic agents tend to decrease gap junctional communication. Conversely, some anticancer and antimitogenic agents can increase gap junctional communication (13) . In addition, Cx32 knockout mice display increased levels of spontaneous and chemically induced hepatoma compared with control mice (14) . Finally, normalization of transformed cells by transfection with specific Cxs provides direct examples of gap junctions acting as tumor suppressors (15) .
For example, Cx43 has been identified as a tumor suppressor gene that can reverse the transformed phenotype of mammary carcinoma (16) and glioma (17) cells. Others have shown that diffusable factors may be involved in Cx43 suppression of C6 glioma cell growth. However, beyond that, mechanisms responsible for this growth inhibition have not been elucidated. Even a role for gap junctional communication in Cx-mediated cell growth control has been disputed. Evidence for this assertion comes from reports that describe a lack of correlation between Cx-mediated dye transfer and growth control. For instance, Cx32-transfected C6 glioma cells transfer fluorescent dyes better than Cx43 transfectants (11 , 18) but show no autonomous growth suppression (19) . Other studies with Cx43-transfected human glioma cells (20) , as well as Cx43-transfected C6 glioma cells (21) subsequently transfected with Cx43 sequences engineered to act as dominant-negative constructs, also contradict correlations between gap junctional dye transfer and growth regulation.
However, as stated above, we and others have shown that gap junctions are selective in their ability to transfer specific dyes (9 , 22) , ions (6 , 22 , 23) , and most recently, natural metabolites (10 , 11) . Therefore, transfer of a particular dye may not accurately represent the transfer of molecules that may underlie the ability of a specific Cx to mediate cell growth control. This could explain paradoxical effects of Cx32 and Cx43, including heteromeric channels that would result from inhibition by dominant-negative constructs with respect to dye transfer and growth control.
Thus, we know that Cx43 suppresses glioma cell growth, possibly by diffusable factors. However, we do not know the nature of these factors or even whether gap junctional communication is involved in their modulation. In this report, we used the C6 glioma cell system to help define relationships between Cxs, gap junctional communication, and cell growth control. This included analysis of not only dye transfer but the transfer of endogenous metabolites between these cells. Our results suggest that gap junctional transfer of endogenous metabolites, as opposed to synthetic dyes, may indeed underlie the ability of a given Cx to regulate cell growth. Although Cx32 mediated more dye transfer between these cells, Cx43 preferentially allowed the transfer of endogenous metabolites. We followed the fate of these transjunctional metabolites to find that a portion of them were incorporated into larger cellular components. However, compared with Cx32, Cx43 mediated the transfer of molecules that were more readily incorporated into material with molecular weights Mr >50,000. These signals did not induce cell death, senescence, or suppress mitogenic signaling because no significant changes in cell viability, telomerase, MAPK activity, or even log phase growth rates were detected. Instead, Cx43 induced contact growth inhibition of these cells. This was potentiated by alterations of extracellular factors in response to Cx43 expression. MFG-E8 was the predominant extracellular factor affected by Cx43 in these cells.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Gap junctional communication was evaluated by the preloading method
with calcein (calcein acetoxymethyl ester; Molecular Probes, Eugene,
OR) and DiI (DiI C18; Molecular Probes) as described previously
(18)
. Calcein (623 daltons) is able to pass through gap
junctions, whereas DiI, a stable, lipophilic dye, does not pass to
neighboring cells. Fluorescently labeled cells were mixed and plated
with nonlabeled cells at a 1:500 ratio such that they settled into
confluent monolayers. Calcein and DiI fluorescence was visualized
through green and red attenuating filters. Gap junctional communication
was visualized by microscopic examination and quantitated as the number
of nonlabeled cells receiving calcein from an individual labeled cell.
Inhibition of dye transfer by 18
-carbenoxolone, a relatively
specific and noncytotoxic gap junction blocker (24)
, was
used to verify the efficacy of this procedure.
Gap junctional transfer of endogenous metabolites between C6 cells transfected with Cx32, Cx43, or empty vector was assayed by the capture protocol (11 , 25 , 26) as described in method 2 of Goldberg et al. (25) , except that instead of donors being plated in the absence of receivers for quantitation, transfer was quantitated based on donors and receivers from the same plates. Briefly, donor cells were metabolically labeled overnight with D-[U-14C]glucose and fluorescently labeled with DiI before being washed with PBS, trypsinized, and plated with receiver cells at a 1:6.25 ratio such that they settled into confluent monolayers. Cells were cocultured for 2–3 h, allowing them time to adhere to the plates and each other. Donors and receivers were separated from each other by fluorescence-activated cell sorting, collected, and lysed. Lysates were filtered through membranes with a Mr 50,000 nominal molecular weight cutoff. Filtrates were further fractionated over a membrane with a Mr 3,000 molecular weight cutoff. All retentates and filtrates were assayed for radioactivity by scintillation counting. Radioactivity was calculated as the cpm per cell for receivers and donors from the same plate. Ratios were then determined to ascertain the degree of transfer. The portion of radioactive material within a given size range that was derived from incorporated transjunctional metabolites that transferred from donors to receivers was calculated as the ratio of the radioactive material within the given size range in donors over that of the same size range in receivers from the same plate. The percentage of radioactive material within a given size range that was derived from transjunctional metabolites that transferred within 20 min of communication was calculated by dividing the portion of radioactive material within a given size range that was derived from incorporated transjunctional metabolites that transferred from donors to receivers within 20 min, by the portion of radioactive material within the same size range that was derived from incorporated transjunctional metabolites that transferred from donors to receivers within about 2 h (steady state), and multiplying by 100.
For analysis of growth, 20,000 cells were plated in 1 ml of standard medium/well on 12-well cluster plates and fed every 4 days. Cells were counted by removing medium, suspending in 0.25% trypsin and 1 mM EDTA in PBS, adding to Isotone (Coulter), and passing through a model ZM Coulter counter.
Telomerase Assays.
Telomerase extraction was performed on confluent cells according to a
3-[(3-cholamidopropyl)dimethylamino]-1-propansulfonate detergent
lysis protocol (27)
. Cells were washed twice with either
PBS (C6 cells) or TRAP washing buffer [10 mM HEPES-KOH (pH
7.5), 1.5 mM MgCl2, 1 mM
KCl, and 1 mM DTT; HeLa cells], homogenized on ice for
30 s in 50 µl of TRAP lysis buffer [1 mM Tris-HCl
(pH 7.5), 1 mM MgCl2, 1
mM EGTA, 5 mM -mercaptoethanol, 0.1
mM phenylmethylsulfonyl fluoride, and 10% glycerol] in
1.5 ml of microcentrifuge tubes with matching rotating pestles, left on
ice for 30 min, homogenized again for 30 s, and centrifuged at
4°C for 30 min at 15,000 x g. The
supernatants were kept at -30°C until use.
Telomerase activity was assessed by the TRAP as described previously
(27)
. The reaction mixtures (50 µl) contained 20
mM Tris-HCl (pH 8.3), 63 mM KCl, 1.5
mM MgCl2, 1 mM EGTA,
0.005% Tween 20, 5 µg of BSA, 50 µM deoxynucleotide
triphosphates, 1 µg of T4gene32 protein (Boehringer-Mannheim), 0.1
µg of TS forward primer (5'-AATCCGTCGAGCAGAGTT), 2 units of Taq
polymerase (Boehringer-Mannheim), and 4 µCi of
[-32P]dCTP (3000 Ci/mmol, 10 µCi/µl).
Sample extracts (5 µl) were added, and the tubes were kept at 23°C
for 30 min to allow TS primer extension before addition of 0.1 µg of
CX reverse primer (5'-[CCCTTA]3CCTAA). A
30-cycle PCR (94°C for 30 s, 50°C for 30 s, and 72°C
for 90 s) was then performed in a Perkin-Elmer Thermo Cycler.
Twenty-five µl of PCR reaction products were analyzed on 8%
nondenaturing acrylamide gels. Gels were dried at 80°C, exposed to
phosphor screens or X-ray film, and visualized on a Molecular Dynamics
PhosphorImager or by autoradiography.
Evaluation of MAPK Levels.
Total ERK1 and ERK2 protein levels were evaluated by Western blot
analysis of cell lysates with mouse monoclonal antiserum directed
against amino acids 219–358 of rat ERK2 (Transduction Laboratories,
Lexington, KY), which is pan-specific for ERK1 and ERK2. Levels of ERK1
and ERK2 activity were evaluated with affinity-purified antibody
directed against tyrosine-phosphorylated amino acids 196–209 of human
ERK1 [DHTGFLTEY(P)VATRWC; New England Biolabs, Beverly, MA], which is
specific for active phosphoforms of ERK1 and ERK2 because
phosphorylation of Y204 in the TXEYX consensus site contained in this
sequence is known to activate MAPK (28, 29, 30)
. Confluent
cells were washed once with PBS, lysed on the plates in SDS-PAGE sample
buffer at a concentration of 5,000 cells per µl, sonicated, boiled
for 5 min, and centrifuged at 15,000 x g for
5 min. The supernatants were stored at -80°C until resolution by
SDS-PAGE on 12% gels and transfer to Immobilon-P membranes
(Millipore). Antiserum was used according to the protocols suggested by
the suppliers and detected by chemiluminescence (ECL; Amersham).
Evaluation of Effects of Conditioned Media on Cell Growth.
Serum-free medium (DMEM) was conditioned for 24 h by 1 day
postconfluent cells. This medium was then concentrated to 
100-fold
by centrifugal filtration through membranes (Amicon) with different
molecular weight cutoffs to obtain the fractions indicated in the text
and related figures. Nontransfected C6 cells (20,000 cell/well) were
plated in standard medium on multiwell tissue culture plates. After the
cells settled (2–3 h), the medium was replaced with media fortified
with conditioned media fractions, as indicated in the text and related
figures. Fortified medium was created by adding concentrated
conditioned media such that equal concentrations were achieved
(i.e., concentrated conditioned medium was diluted
100-fold into standard medium). Cells were counted after 90 h
of growth in the conditioned medium. Control cells were grown in
standard medium. Cells were counted by removing medium and suspending
in 0.25% trypsin and 1 mM EDTA in PBS, which was
subsequently added to Isotone (Coulter) and examined with a Coulter
counter.
Analysis of Conditioned Media by FPLC, SDS-PAGE, and MALDI-TOF
Mass Spectrometry.
Serum-free medium supplemented with 100 µCi/ml
[35S]Met and [35S]Cys
was conditioned and concentrated to obtain proteins between
Mr 50,000–100,000 as described above.
The DMEM was exchanged for 20 mM Tris (pH 7.5) by
concentration over Centricon 50 membranes before the material was
resolved by FPLC through an anion exchange matrix (DEAE Sepharose;
Pharmacia) with a 0–1 M gradient of NaCl.
Fractions were resolved by SDS-PAGE on 12% gels, which were Coomassie
stained, dried, exposed to X-ray film (XAR-5; Kodak), and visualized by
autoradiography.
Bands were then excised from the gel, transferred to an acid-washed
tube, rehydrated with water, crushed, washed three times for 20 min
with 50 mM Tris (pH 8.0)/50% acetonitrile, dried, and
incubated overnight at 32°C with 0.67 ng/µl trypsin in 25
mM Tris (pH 8.5) to completely digest the protein. Peptides
were then extracted with 50% acetonitrile/0.1% trifluoroacetate,
dried, suspended in 10 mg/ml 4-hydroxy--cyanocinnamic acid in 50%
acetonitrile/0.1% trifluoroacetate containing angiotensin as an
internal standard, and applied to a MALDI sample plate, which was dried
and washed with water to remove excess buffer salts. MALDI mass
spectrometric analysis was performed on a PerSeptive Voyager DE-RP mass
spectrometer in the linear mode (31)
. Profiles were
queried against the entire nr database by ProFound
(32)
.4
Northern Blotting.
Total cellular RNA was purified, resolved by gel electrophoresis (15
µg/lane), and blotted to nylon membranes as described previously
(33)
. Blots were hybridized with
[32P]dCTP-labeled (NEN Life Science Products)
2.0-kb, full-length mouse MFG-E8 cDNA (34)
at 42°C in
hybridization buffer (5x SSPE, 5x Denhardt’s solution, 50%
formamide, 0.5% SDS, and 0.1 mg/ml salmon sperm DNA), washed at high
stringency, and examined by autoradiography. Membranes were then
stained with methylene blue (35)
. Bands were quantitated
by image densitometry (ImageQuant; Molecular Dynamics).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A, Cx43 transfectants displayed significantly higher
(>2-fold) levels of communication than control transfectants. However,
Cx32 transfectants passed dye to several times as many cells as the
Cx43 transfectants. In all cases, dye transfer was completely inhibited
by ACO, a relatively specific and noncytotoxic gap junction blocker
(24)
. Communication that occurred within 20 min (after
removal for the blocker) displayed similar results as communication
seen without ACO.
|
A, all cells showed similar log phase growth rates. Instead,
growth suppression resulted from an induction of contact inhibition of
growth, reflected in a 2-fold decrease in cell saturation densities of
the Cx43 transfected cells compared with Cx32 or control transfectants
(Fig. 2B
).
|
. This presented an
apparent paradox with respect to gap junctional communication and
growth control. However, transfer of a particular dye may not
accurately represent the communication potential of a specific channel
for functionally relevant metabolites. To address this situation, we
used the capture approach to examine the transfer of endogenous
metabolites through these channels. We used glucose, which is the major
food source for these and many other cells, to label a wide range of
metabolites, thereby allowing the detection of endogenous
transjunctional molecules in a nonbiased manner (11
, 25
, 26)
.
As shown in Fig. 1
B, no significant difference in the
ability of metabolites derived from glucose to travel through gap
junctions composed of Cx43 or Cx32 was seen. Transfer of these
metabolites was blocked by ACO. However, examination of molecules that
transferred within 20 min of communication (after ACO removal) revealed
that Cx43 transfectants were much better coupled with respect to
metabolites than Cx32 cells. This was in stark contrast to
communication measured by fluorescent dye shown in Fig. 1A
.
These molecules are natural permeants that may underlie the effects of
gap junctions in cells, in this case, cell growth control. As a first
step in investigating potential function, we sought to determine the
fate of these transjunctional compounds. This was done by measuring the
radioactivity of size fractionated radioactive material derived from
transjunctional metabolites that transferred from donors to receivers
(as described in "Materials and Methods"). As shown in Fig. 1
B, this suggested that a significant portion of the
transjunctional metabolites in these cells were used anabolically by
the receivers.
However, Cx32 and Cx43 may mediate the transfer of compounds with
different fates with respect to incorporation into macromolecules. As
shown in Fig. 1
C, about 22–35% of the radioactive material
of Mr <50,000 in control and Cx32
transfectants coupled for up to 2 h was derived from metabolites
that transferred from donors to receivers within 20 min of
communication. In contrast, approximately twice this amount, 72–78%,
of the material within this size range was derived from transjunctional
metabolites that transferred within 20 min between Cx43 transfectants.
Although 34% of labeled material Mr >50,000 in Cx32 receivers was
derived from metabolites that transferred within 20 min, 50–57% of
this material was derived from transjunctional precursors that passed
within 20 min between control or Cx43 transfectants. This suggests that
similar molecules passed through the channels made from endogenous or
exogenous Cx43 in the control and Cx43 transfectants, respectively,
whereas Cx32 mediated the transfer of compounds that were not as
readily incorporated into macromolecules
Mr >50,000. This effect was
significant, with 57% of the total variance accounted for by the Cx
phenotype (P < 0.05) by ANOVA.
Effect of Cx43 on Telomerase Activity.
The Cx-induced contact growth inhibition described above could have
been attributable to either inhibition of cell division or induction of
cell death. One indication of changes in cell senescence that
accompanies neoplastic transformation is the activity of telomerase,
proposed to be responsible for the chromosome maintenance that is
necessary for immortalization of transformed cells (37)
.
This activity was measured by the TRAP in Cx43 and control
transfectants. On the basis of the relative intensities of the ladder
of bands obtained (representing amplified telomeric sequences), shown
in Fig. 3
A, the expression of Cx43 was not accompanied by any
decreased telomerase activity. This suggested that the decreased cell
saturation density associated with Cx-mediated growth suppression was
not associated with enhanced cell senescence. There was also no
morphological evidence for increased levels of necrotic cell death or
detachment of cells from the culture dishes at higher densities.
Therefore, we chose to examine the effects of these Cxs on the
inhibition of mitogenic pathways.
|
. If anything, a slight increase in
active MAPK may appear to have accompanied Cx43 expression in C6 cells.
However, this was not consistent nor statistically significant upon
repeated analysis. Preliminary evidence also suggests that levels of
MAPK kinase 1/2, MAPK kinase kinase, signal transducers and activators
of transcription 3, and cyclic AMP-responsive element binding protein
were not significantly affected by Cx-mediated growth inhibition of
these cells (data not shown).
Examination of Secreted Factors Modulated by Cx43.
Although no overt effects on mitogenic signaling or processes related
to cell senescence were evident, earlier reports have linked growth
suppression of C6 cells by Cx43 to the production of a diffusable
growth-inhibitory factor (36)
. Because this might provide
insight into the intracellular processes that may be affected by gap
junctions, we sought to further characterize this activity. Using a
series of size exclusion filters (with
Mr 3,000, 10,000, 30,000, 50,000, and
100,000 nominal molecular weight cutoffs), it was found that the
growth-inhibitory activity consistently associated with fractions
containing material Mr >50,000 and
Mr <100,000 (Fig. 4)
. Parallel fractions of media from control transfectants showed no such
activity. The size range of the active fraction suggested that the
soluble factor was likely to be a complex polypeptide. Consistent with
this, the activity of this fraction was refractive to nuclease but
sensitive to heat treatment (95°C for 5 min; data not shown).
|
470
mM NaCl (Fig. 5A)
. However, a minor peak and a shoulder that eluted at about
550 and 700 mM, respectively, were increased in
Cx43-coupled cells. In contrast, a major peak eluting at 900
mM NaCl (arrow in Fig. 5A
)
was suppressed by Cx43.
|
) identified polypeptides that were
differentially induced (arrowheads) or
suppressed (arrows) by Cx43. These bands
were metabolically labeled and, therefore, represented direct products
of the cells, as opposed to exogenous components of the media. The
major difference was seen as a Mr
60,000 extracellular compound that was selectively suppressed by Cx43
(arrow in Fig. 5B
). Resolution
of this peptide from other components in the conditioned medium enabled
its identification by MALDI-TOF peptide mass mapping of
trypsin-generated fragments eluted from the gel (32)
. As
shown in Fig. 6
, with >50% coverage, this peptide mass mapping revealed the protein
to be rat MFG-E8 (38)
.
|
, MFG-E8 mRNA was suppressed to 50% of the levels seen in either
Cx32, empty vector, or nontransfected cells. As stated in the
introduction, the dominant-negative Cx43 mutants L160M and A253V both
ablate growth regulation by wild-type Cx43. Northern blot analysis in
Fig. 7
reveals that, along with ablating growth suppression, these
mutants also ablate MFG-E8 suppression by wild-type Cx43. This
observation is consistent with MFG-E8 repression playing a functional
role in Cx43 induction of contact growth inhibition. Suppression of
MFG-E8 by Cx43 carries some interesting implications, as discussed
below.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Because Cx-mediated cell growth suppression occurred at the level of contact growth inhibition, we assayed mechanisms at cell confluence that may have been involved. Telomerase activity was not decreased by Cx43, which was not surprising because, as stated above, cell death did not accompany growth suppression. In the absence of cell death or senescence, attention was turned to mitogenic signaling through the MAPK pathway. MAPK is both necessary and sufficient for cell transformation and is a key element of serum induced growth (28 , 39) . However, we found no significant alterations in MAPK levels to accompany Cx-mediated growth inhibition.
Gap Junctional Communication.
Despite a large number of investigations, mechanisms underlying the
ability of Cxs to suppress transformed cell growth have not been
defined. Even the most basic questions regarding this have gone
unanswered. The only bona fide function of Cxs is to allow cells to
share small hydrophilic molecules (1, 2, 3)
. Rather than
decrease growth rates, Cx43-mediated growth suppression resulted from
increased contact growth inhibition of these cells. This observation is
consistent with the formation of gap junctions and intercellular
communication networks being required for cell growth inhibition by
Cxs.
However, as discussed in the "Introduction," a role for gap junctional communication in Cx-mediated cell growth inhibition has been disputed. This claim has arisen from reports that the ability of Cx26 and Cx43 to normalize HeLa and C6 cell growth does not always correlate with their ability to transfer fluorescent dye (20 , 21 , 40) . At first, the data reported here supported this contention, because Cx32 transfectants were better coupled than Cx43 transfectants when assayed by fluorescent dye transfer. However, use of the capture technique revealed that, in contrast to synthetic dyes, the Cx43 transfectants shared endogenous metabolites more efficiently than Cx32 transfectants. In support of this, we have shown recently that ATP and glutamate may travel 160 or 29 (respectively) times more efficiently through channels made of Cx43 than through channels made from Cx32 (11) . We propose that the data reported here are consistent with gap junctional communication, rather than other unknown Cx-mediated interactions, underlying the ability of a given Cx to mediate growth suppression of a particular cell type.
Tracking the fates of transjunctional metabolites revealed that a significant portion of Cx43 permeants were incorporated into macromolecules more readily than those that transferred via Cx32. This is also consistent with preferential transfer of ATP and glutamate through Cx43, which could be used for RNA and protein production by receiver cells. However, the significance of this finding in terms of signals that mediate growth suppression is not clear. In the absence of specific growth regulatory transjunctional metabolites, we sought to define their targets and mechanisms as described below.
Conditioned Media and Distinct Mechanisms.
As reported previously (36)
, growth-inhibitory activity
was found in conditioned medium from Cx43-transfected C6 cells. These
factors could present a handle to mechanisms by which gap junctions, or
more specifically Cx43, can suppress the growth of particular cells (in
this case, glioma). As a step in this direction, we resolved components
in the conditioned media of Cx43 and control transfectants to find
those that may play a part in this. Although actual growth-inhibitory
factors induced by Cx43 were not identified here, direct and
unbiased analysis revealed that suppression of MFG-E8 represented a
major effect of Cx43 in these cells. This corroborates earlier work
indicating that Cx43 modulates expression of extracellular proteins
such as insulin-like growth factor binding protein 4 (41)
.
However, our results are unique in that the proteins identified by this
direct approach represented the major component of the extracellular
solution that was altered by Cx43 (i.e., the peak containing
MFG-E8 in fraction 8 of Fig. 5
).
Cx43-transfected C6 cells that were secondarily transfected with dominant-negative Cx43 constructs were used to further define relationships between Cx43, MFG-E8, and glioma cell growth. These constructs restore tumorigenicity of C6 cells in the face of Cx43 expression. The mutant Cx43L160 M, in which lysine 160 is changed to Met, is analogous to the Cx32V139 M CMTX mutant. In addition to ablating growth regulation of C6 cells by Cx43 (21) , the Cx43L160 M construct also ablates Cx26 growth suppression of HeLa cells (40) . Another mutant, Cx43A253V, represents a human polymorphism and has been found in metastatic meningiomas (42) . Interestingly, although both constructs ablate growth regulation by wild-type Cx43, only Cx43L160 M blocks dye transfer between these cells. In contrast, the Cx43A253V mutant increases tumorigenicity of the cells without affecting their ability to share Lucifer Yellow (21) . However, both of these mutants elevated MFG-E8 expression in the Cx43 transfectants. This result is consistent with MFG-E8 suppression playing an important role in Cx43-mediated growth suppression of these cells.
In addition to providing a potential handle to mechanisms by which Cx43 inhibits glioma cell growth, this finding may also shed light on reasons behind the cell type dependence of specific Cxs to regulate cell growth. For example, if suppression of MFG-E8 is functionally related to Cx43 growth suppression, HeLa cells would not be susceptible because they do not produce significant levels of MFG-E8 (43) . This seems to be the case because Cx43 does not suppress HeLa cell growth (44) . Interestingly, as described below, mammary carcinoma cells do express high levels of MFG-E8 (45 , 46) and, similar to glioma cells, are normalized by Cx43 expression (16) . However, although Cx43 suppressed MFG-E8 expression in C6 cells, additional studies would be needed to address cell and tumor-specific effects. Also, although Cx32 did not affect MFG-E8 expression, other Cxs were not tested; thus, it is not clear whether MFG-E8 is a specific target of Cx43.
Implications of MFG-E8 suppression by Cx43.
MFG-E8 contains a signal sequence that is followed by an epidermal
growth factor-like domain that contains an RGD site involved in
integrin binding. In addition, the carboxyl part of the protein
contains C1 and C2 domains of blood clotting factor 5/8, which may be
involved in phospholipid binding (47)
. The human
MFG-E8 gene is located on chromosome 15Q25
(48)
, but homologues bearing the same motifs have been
cloned from bovine (49)
, pig (50)
, mouse
(34)
, rat (38)
, and horse
(51)
genomes. The protein promotes cell attachment via the
RGD sequence, whereas the epidermal growth factor moieties may augment
breast cancer growth (52)
. Accordingly, MFG-E8 is detected
mainly on breast tumor cells (45)
. For example, human
MFG-E8 (also called lactadherin or BA46) was found in patients with
metastatic breast tumors but not melanoma or healthy controls
(46)
. Indeed, MFG-E8 antisera has been patented for use in
the detection and treatment of tumors (53)
. Antisera
against MFG-E8 inhibited the growth of human tumors in nude mice
(54
, 54) , in one report actually curing six of seven mice
with human breast cell tumors (56)
. Although MFG-E8 may be
produced by transformed ovary and lung cells, HeLa cells do not produce
significant levels (43)
, which, as discussed above, is of
particular relevance to this study. In any case, although the RGD
sequence has been shown to bind integrin vß3 in some cells
(52)
, it may also be involved in a wide array of cell
adhesion and signaling events, including contact inhibition of growth,
that are unpredictable at this time (57)
.
Thus, both breast carcinoma and glioma cells display decreased levels of Cx43 and increased MFG-E8 levels. Furthermore, both cell types are normalized by Cx43, whereas HeLa cells, which do not express MFG-E8, are not. Moreover, normalization of glioma cells is accompanied by suppression of MFG-E8. When taken together, this report presents evidence for a connection between Cx43 expression, modulation of extracellular factors including MFG-E8, and gap junctional transfer of endogenous metabolites rather than fluorescent dye or other undefined interactions. Examination of biophysical properties and permeabilities of specific gap junction channels may be needed to address the relevance and function of Cx43 in the repression of cell growth and MFG-8.
|
FOOTNOTES |
|---|
1 This work was supported in part by a
Yamagiwa-Yoshida Memorial International Cancer Study Grant from the
International Union Against Cancer, a grant from the Wendy Will Case
Cancer Research Fund, a Grant-in-Aid and Research Fellowship from for
the Foundation for the Promotion of Cancer Research as part of the
Second Term Comprehensive 10-Year Strategy for Cancer Control from the
Japanese Ministry of Health and Welfare (to G. S. G), a Grant-in-Aid
from the Canadian Institutes of Health Research (to C. C. G. N.),
and grants from the Whittaker Foundation and the NIH (to B. J. N.). 
2 To whom requests for reprints should be
addressed, at Physiology and Biophysics, Basic Science Building, Health
Science Complex, State University of New York at Stony Brook, Stony
Brook, NY 11794-8661. Phone: (631) 431-6332; Fax: (631) 444-3432;
E-mail: camgsg{at}Buffalo.edu
3 The abbreviations used are: Cx, connexin; MAPK,
mitogen-activated protein kinase; MFG-E8, milk fat globule epidermal
growth factor 8; DiI,
1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyaine perchlorate;
TRAP, telomere repeat amplification protocol; ERK, extracellular
signal-regulated kinase; FPLC, fast protein liquid chromatography; ACO,
-carbenoxolone; MALDI-TOF, matrix assisted desorption
ionization-time of flight.
4 Internet address:
http://Prowl.rockefeller.edu/cgi-bin/ProFound.
Received 4/21/00. Accepted 9/18/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y. Shen, P. R. Khusial, X. Li, H. Ichikawa, A. P. Moreno, and G. S. Goldberg Src Utilizes Cas to Block Gap Junctional Communication Mediated by Connexin43 J. Biol. Chem., June 29, 2007; 282(26): 18914 - 18921. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Shen, Z. Jia, R. G. Nagele, H. Ichikawa, and G. S. Goldberg Src Uses Cas to Suppress Fhl1 in Order to Promote Nonanchored Growth and Migration of Tumor Cells Cancer Res., February 1, 2006; 66(3): 1543 - 1552. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Lahlou, M. Fanjul, L. Pradayrol, C. Susini, and S. Pyronnet Restoration of Functional Gap Junctions through Internal Ribosome Entry Site-Dependent Synthesis of Endogenous Connexins in Density-Inhibited Cancer Cells Mol. Cell. Biol., May 15, 2005; 25(10): 4034 - 4045. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. T. Fu, J. F. Bechberger, M. A. Ozog, B. Perbal, and C. C. Naus CCN3 (NOV) Interacts with Connexin43 in C6 Glioma Cells: POSSIBLE MECHANISM OF CONNEXIN-MEDIATED GROWTH SUPPRESSION J. Biol. Chem., August 27, 2004; 279(35): 36943 - 36950. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. B. Alexander, H. Ichikawa, J. F. Bechberger, V. Valiunas, M. Ohki, C. C. G. Naus, T. Kunimoto, H. Tsuda, W. T. Miller, and G. S. Goldberg Normal Cells Control the Growth of Neighboring Transformed Cells Independent of Gap Junctional Communication and Src Activity Cancer Res., February 15, 2004; 64(4): 1347 - 1358. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. W. Doble, X. Dang, P. Ping, R. R. Fandrich, B. E. Nickel, Y. Jin, P. A. Cattini, and E. Kardami Phosphorylation of serine 262 in the gap junction protein connexin-43 regulates DNA synthesis in cell-cell contact forming cardiomyocytes J. Cell Sci., January 22, 2004; 117(3): 507 - 514. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. S. Goldberg, D. B. Alexander, P. Pellicena, Z.-Y. Zhang, H. Tsuda, and W. T. Miller Src Phosphorylates Cas on Tyrosine 253 to Promote Migration of Transformed Cells J. Biol. Chem., November 21, 2003; 278(47): 46533 - 46540. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Iacobas, M. Urban-Maldonado, S. Iacobas, E. Scemes, and D. C. Spray Array analysis of gene expression in connexin-43 null astrocytes Physiol Genomics, November 11, 2003; 15(3): 177 - 190. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. S. Rehman, S. Yin, B. A. Mayhew, R.A. Word, and W. E. Rainey Human myometrial adaptation to pregnancy: cDNA microarray gene expression profiling of myometrium from non-pregnant and pregnant women Mol. Hum. Reprod., November 1, 2003; 9(11): 681 - 700. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Olbina and W. Eckhart Mutations in the Second Extracellular Region of Connexin 43 Prevent Localization to the Plasma Membrane, but Do Not Affect Its Ability to Suppress Cell Growth Mol. Cancer Res., July 1, 2003; 1(9): 690 - 700. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-W. Zhang, K. Nakayama, K.-I. Nakayama, and I. Morita A Novel Route for Connexin 43 to Inhibit Cell Proliferation: Negative Regulation of S-Phase Kinase-associated Protein (Skp 2) Cancer Res., April 1, 2003; 63(7): 1623 - 1630. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H.-C. Lin, J. Yang, S. Liu, T. Takano, X. Wang, Q. Gao, K. Willecke, and M. Nedergaard Connexin Mediates Gap Junction-Independent Resistance to Cellular Injury J. Neurosci., January 15, 2003; 23(2): 430 - 441. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Koval Sharing signals: connecting lung epithelial cells with gap junction channels Am J Physiol Lung Cell Mol Physiol, November 1, 2002; 283(5): L875 - L893. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. S. Goldberg, A. P. Moreno, and P. D. Lampe Gap Junctions between Cells Expressing Connexin 43 or 32 Show Inverse Permselectivity to Adenosine and ATP J. Biol. Chem., September 20, 2002; 277(39): 36725 - 36730. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Huang, Y. Lin, C. C. Wang, J. Gano, B. Lin, Q. Shi, A. Boynton, J. Burke, and R.-P. Huang Connexin 43 Suppresses Human Glioblastoma Cell Growth by Down-Regulation of Monocyte Chemotactic Protein 1, as Discovered Using Protein Array Technology Cancer Res., May 1, 2002; 62(10): 2806 - 2812. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
0.05).