Content uploaded by James Trosko
Author content
All content in this area was uploaded by James Trosko on May 13, 2021
Content may be subject to copyright.
[Frontiers in Bioscience, 3, d208-236, February 15, 1998]
208
CELL-CELL COMMUNICATION IN CARCINOGENESIS
James E. Trosko1, Randall J. Ruch2
1Department of Pediatrics and Human Development, Michigan State University, East Lansing, Michigan 48824, 2Department of
Pathology, Medical College of Ohio, Toledo, Ohio
Received 2/2/98 Accepted 2/6/98
TABLE OF CONTENTS
1. Abstract
2. Introduction
3. Evolution and cancer
4. Theories of carcinogenesis: Overview
5. Stem cell versus the de-differentiation theories
6. Initiation/promotion/progression theory of carcinogenesis
7. Nature and nurture theory of carcinogenesis
8. Mutation versus the epigenetic theories of carcinogenesis
9. Oncogene and tumor suppressor gene theories
10. Gap junctions - ancient and ubiquitous mediators of cellular homeostasis
11. Structure of gap junctions
12. The connexin multigene family
13. Regulation of connexin gene expression
14. Gap junction formation, control of channel permeability, and mechanisms of disrupted GJIC
15. Multiple functions of GJIC
16. Role of GJIC in regulating cellular proliferation and neoplasia
16.1. Neoplastic cells have fewer gap junctions
16.2. Growth stimuli inhibit GJIC
16.2.1. Carcinogens
16.2.2. Oncogenes
16.2.3. Growth Factors
16.3. Growth inhibitors stimulate GJIC
16.4. Cell cycle-related changes in GJIC
17. Involvement of GJIC in the growth inhibition of neoplastic cells by nontransformed cells
18. Specific disruption and enhancement of GJIC
18.1. Connexin antisense studies
18.2. Connexin gene knockout
18.3. Dominant-negative inhibition of connexin function
18.4. Connexin transfection studies
19. GJIC and other growth control mechanisms
20. Growth regulation mediated by a gap junction signal
21. Modulation of GJIC for cancer therapy
22. Acknowledgment
23. References
1. ABSTRACT
To explain the complex carcinogenic process by
which a single normal cell in human beings can be
converted to an invasive and metastatic cancer cell, a
number of experimental findings, epidemiological
observations and their associated hypothesis/theories have
been integrated in this review. All cancers have been
generally viewed as the result of a disruption of the
homeostatic regulation of a cell’s ability to respond
appropriately to extra-cellular signals of the body which
trigger intra-cellular signal transducting mechanisms which
modulate gap junctional intercellular communication
between the cells within a tissue. Normal homeostatic
control of these three forms of cell communication
determines whether the cell: (a) remains quiescent (Go);
(b) enters into the cell proliferation phase; (c) is induced to
differentiate; (d) is committed to apoptose; or (e) if it is
already differentiated, it can adaptively respond.
During the evolution from single cell organisms
to multicellular organisms, new cellular/biological
functions appeared, namely, the control of cell proliferation
(“contact inhibition”), the appearance of the process of
Gap junctions and cancer
209
differentiation from committed stem cells of the various
tissues and the need for programmed cell death or
apoptosis. Interestingly, cancer cells have been
characterized as cells: (a) having been derived from a stem-
like cell; (b) without their ability to control cell growth or
without the ability to contact inhibit; (c) which can not
terminally differentiate under normal conditions; and (d)
having altered ability to apoptosis under normal conditions.
During that evolutionary transition from the single cell
organism to the multicellular organism, many new genes
appeared to accompany these new cellular functions. One
of these new genes was the gene coding for a membrane
associated protein channel (the gap junction), which
between coupled cells, allowed the passive transfer on ions
and small molecular weight molecules. A family of over a
dozen of these highly evolutionarily–conserved genes (the
connexin genes) coded for the connexin proteins. A
hexameric unit of these connexins in one cell (a connexon)
couples with a corresponding connexon in a contiguous
cell to join the cytoplasms. This serves to synchronize
either the metabolic or electrotonic functions of cells within
a tissue. Most normal cells within solid tissues have
functional gap junctional intercellular communication
(GJIC) (exceptions are free-standing cells such as red blood
cells, neutrophils, and several, if not all, the stem cells).
On the other hand, the cancer cells of solid tissues appear
to have either dysfunctional homologous or heterologous
GJIC. Therefore, among the many differences between a
cancer cell and its normal parental cell, the carcinogenic
process involves the transition from a normal, GJIC-
competent cell to one that is defective in GJIC.
The review examines how GJIC can be either
transiently or stably modulated by endogenous or
exogenesis chemicals or by oncogenes and tumor
suppressor genes at the transcriptional, translational, or
posttranslational levels. It also uses the gap junction as the
biological structure to facilitate cellular/tissue homeostasis
to be the integrator for the “stem cell” theory, “disease of
differentiation theory”, “initiation/promotion/progression”
concepts, nature and nurture concept of carcinogenesis, the
mutation/ epigenetic theories of carcinogenesis, and the
oncogene/ tumor suppressor gene theories of
carcinogenesis. From this background, implications to
cancer prevention and cancer therapy are generated.
2. INTRODUCTION: CANCER AS A ‘DISEASE OF
HOMEOSTASIS’
In order to understand the disease of cancer, one
must recognize that, while there are multiple cancer types
found in most organs, there are some “universal features”
associated with all cancers. Cells that are cancerous appear
not to respond to “contact inhibition”(1,2), fail to
terminally differentiate (3-5), appear to be clonally-derived
from a stem like cell (6-12), and continue to genotypically
and phenotypically change as the tumor grows (13,14).
More recently, the biological processes of “signal
transduction” (15,16) and programmed cell death or
apoptosis (17-19) appear also to be altered in cancer cells
compared to their normal parental cells.
During the course of evolution from single-celled
organisms to multi-cellular organisms, new genes and
cellular functions had to accompany that transition. Single-
cell organisms survived changes in the environment by
adaptively responding to physical (temperature, radiations)
and chemical (nutrients, toxins, toxicants) agents by
intracellular signals which led to cell proliferation
modifications. In the multi-cellular organism, a delicate
orchestration of the regulation of cell proliferation for
growth and tissue repair/wound healing and of the
differentiation of cells had to occur after the fertilization of
the egg cell, during embryonic/fetal development, sexual
maturation and adulthood/aging of the individual organism.
That orchestration of specific cell/tissue/organ and organ
system functions is referred to as “homeostasis”.
In the multi-cellular organism, homeostasis is
mechanistically governed by three major communication
processes: extracellular-communication via hormones,
growth factors, neurotransmitters and cytokines which
trigger intracellular-communication via alterations in
second messages (e.g., Ca++, diacylgycerol, pH, ceramides,
NO, c-AMP, reactive oxygen species) and activated signal
transduction systems to modulate intercellular-
communication mediated by gap junction channels (20)
(figure 1). Cell adhesion and cell-matrix interactions are
considered a subclass of intercellular communication
molecules.
All of these communication processes in a multi-
cellular organism are intimately interconnected to maintain
its normal development and health. In effect, this
communication processes must control a cell’s ability (a) to
proliferate; (b) to differentiate; (c); to apoptose; and (d) if
differentiated, to respond, adaptively. Disruption of any
one of these three forms of communication could lead to
increased or decreased proliferation; to abnormal
differentiation; to increased or decreased apoptosis and to
abnormal adaptive responses of differentiated cells.
3. EVOLUTION AND CANCER
Single cell organisms survive changes in their
environment by having in their populations a few
individual cells with mutations in some gene that might
give the individual cell a selective advantage. In turn, this
individual would survive to leave offspring and carry on
the species. Limitations in nutrients, as extracellular-
communication signals, can control cell proliferation of
single cellular organisms.
When, during the course of evolution, the first
multi-cellular organism appeared, new genes and biological
functions had to parallel the appearance of the control of
cell proliferation within the multi-cellular individual, as
well as the induction of differentiation of cells at critical
Gap junctions and cancer
210
Figure 1. Scheme of the postulated link between extracellular communication and gap junctional intercellular communication
via various intracellular signal transducing mechanisms (second message) mechanisms. Diagram illustrates how exogenous non-
genotoxic agents can either interfere with, or mimic, endogenous extracellular signals (Reprinted from J.E. Trosko and T. Inoue,
Stem Cells 15 (Suppl. 2), 59-67, 1997, AlphaMed Press; used with permission).
times, the control of programmed cell death and the
adaptive responses of the differentiated cells. One family
of highly evolutionary-conserved genes, the genes that code
for the gap junction proteins or connexins, appeared at the
time multi-cellularity appeared (21).
Philosophically, the appearance of cancer, a
disease of multi-cellular organisms, seems as though the
process of carcinogenesis is a “throw-back” in the
evolutionary process. The cancer cell, unlike the normal
multi-cellular counter-part, no longer has any growth
control (except by nutrient depletion) and can not
terminally differentiate. In effect, a cancer cell resembles a
bacterial cell that survives by uncontrolled cell
proliferation and can not differentiate. Normal cells of a
multi-cellular organism have connexin genes while single-
cellular organisms do not. Cancer cells, which do not
contact inhibit, do not have growth control, do not
terminally differentiate and usually have abnormal
apoptosis responses, do not appear to have functional gap
junctional intercellular communication (22-27). Is it just
coincidence that the control of cell growth, terminal
differentiation and the appearance of apoptosis appeared
when the gap genes appeared during the evolution of a
multi-cellular organism or is it causal?
4. THEORIES OF CARCINOGENESIS: OVERVIEW
In order to examine the major thesis of this
exercise, namely that reversible disruption of gap
junctional intercellular communication plays a role during
the tumor promotion phase of carcinogenesis and that
stable down-regulation of GJIC leads to the conversion of a
premalignant cell to an invasive and metastatic cancer cell,
a brief review of the major theories of carcinogenesis will
be undertaken. One of the first concepts that must be
understood when trying to unravel the complicated
carcinogenic process is the “hierarchical” nature of multi-
cellular organisms (28). The idea that the “whole is greater
than the sum of its parts” comes from this idea. From
Gap junctions and cancer
211
atoms and molecules which are organized into organelles
within cells, from cells which are organized within tissues
to form organs, and the integration of organs into organ
systems, the emergence of features not found in any of the
individual components of any one level can be found. The
negative and positive feedback of molecular, biochemical,
cellular and physiological information via the three forms
of communication processes (extra-, intra-, and inter-)
within and between the hierarchical levels is the basis for
the cybernetic concept (29). When these cybernetic
feedback systems help maintain the hierarchical nature of
multi-cellular organisms, then it can be said that
homeostasis is achieved and the organism is in a state of
health.
In the case of a human being, starting from a
single fertilized egg (the “toti-potent” cell), 100 trillion
cells, consisting of pluri-potent stem cells, progenitor cells
and terminally differentiated cells, are organized and
orchestrated via the communication processes to produce
an adult by cell proliferation, differentiation, apoptosis and
adaptive responses of the differentiated cells. The
pluripotent and progenitor cells are capable of further
growth, differentiation, wound healing and adaptive
responses before the break down of the hierarchy which
leads to the death of the higher level individuals (recall that
the death of the individual does not necessarily coincide
with the dead of cells). Early concepts, derived over the
decades by Claude Bernard, W.B. Cannon, P. Weiss, J.L.
Kavanaugh, O.H. Iverson, E.E. Osgood and V.R. Potter
(see reviews, 30,31), postulated the existence of positive
and negative regulatory factors that existed between stem
progenitor cells and their differentiated daughters to control
growth and differentiation.
The mechanistic basis for the cybernetic feedback
system consists of positive factors (growth factors,
hormones, cytokines, and neurotransmitters) [extra-cellular
signals] that are secreted by one cell type and that trigger
receptors and transmembrane signal transducting elements
in distal cells [intra-cellular signals]. These signals are, in
turn, either transmitted to, or blocked from, the neighboring
cells when the gap junction channels are up- or down-
regulated. After receiving these signals, the targeted cells
alter their physiology and produce negative extracellular
signals that feedback to the original positive-signalling
cells. If this basic view of homeostasis in a multi-cellular
organism is accepted, then, by logic alone, the breakdown
of any one of these three steps (extra-, intra- or inter-
cellular communication) should lead to the dysregulation of
cell proliferation, differentiation, apoptosis and adaptive
responses of differentiated cells. While this seems to
describe what has happened in cancer cells, it remains to be
experimentally verified that this is, indeed what happens
during carcinogenesis.
Any scientific hypothesis or theory must, by
definition, explain observations in order to produce testable
predictions which can falsify the hypothesis or theory.
Some of the major observations that would have to be
explained by any theory of carcinogenesis include: (a)
normal cells are contact inhibitable, while cancer cells are
not (1); (b) normal cells derived from stem and progenitor
cells are capable of terminal differentiation; cancer cells
under normal situations are not [teratomas represent a
special case] (32); most, if not all, tumors appear to be
derived from a single cell (6-12); and (d) during the long
carcinogenic process, the tumor cell acquires multiple
genotypic and phenotypic changes(13).
Over the decades, while many theories have
elements that explain some of the observations, none of
them provide the framework for a complete explanation.
From Boveri’s idea that cancers are formed because of
chromosomal abnormalities (33) to current ideas that
altered activation and de-activation of tumor suppressor
genes are the “cause” of carcinogenesis (34). The
following two quotations set the stage for understanding
the “reductionalistic “ versus a “holistic” view of the
problem:
“The understanding of the cellular basis
of cancer means being able to describe
the biochemical of the regulated
pathways between the cell surface and
the nucleus that control cell growth.”
(35).
“The cancer problem is not merely a
cell problem, it is a problem of cell
interaction, not only within tissues, but
with distant cells in other tissues.” (36)
Several the major theories have stimulated
research: (a) the idea that cancer is a “disease of
differentiation” (3-5); the “stem cell” theory of cancer (6-
12) has been pitted against the “de-differentiation” theory
of cancer (37); the idea that combines these former two
theories is found in “oncogeny as partially blocked
ontogeny” (5); the “initiation/promotion/progression”
concept of carcinogenesis was conceived as an operational
description to explain distinct steps during the multi-step
process (38); the “nature versus nurture” theory (39) has
been argued to explain whether genetics or the environment
was the determinant in the cause of cancer; classic
disagreements have appeared as to whether mutagenic
versus epigenetic mechanisms are responsible for
carcinogenesis (40); more recently, the “oncogene and
tumor suppressor gene” theory has been a driving force in
cancer research. The hypothesis that “ cancer was the
result of dysfunctional gap junctional intercellular
communication”, which was postulated by Loewenstein
(22), has been modified to integrate some of the
aforementioned ideas (40,41).
The working hypothesis of this review will that
the “dysfunctional gap junctional intercellular
communication” theory can integrate all of the other
theories because each of them can related to GJIC. The
hypothesis to be developed here is that, starting with a
pluripotent stem cell or early progenitor cell, a stable
alteration (a mutation or, in some cases, a epigenetic
Gap junctions and cancer
212
repression at the transcriptional level) of a gene (a proto-
oncogene or tumor suppressor gene) that controls terminal
differentiation of this cell but does not alter the control of
the proliferative ability of the cell (i.e., contact inhibition or
some form of GJIC would still remain in cells of solid
tissues and some form of growth inhibition would exist for
soft-tissue cells). This would explain the “ initiation “
phase of carcinogenesis . As long as this initiated stem-like
cell is communicating with other normal cells
(heterologous GJIC) or other initiated cells (homologous
GJIC), there will be no cell proliferation. When GJIC in
these initiated cells is inhibited in a sustained fashion by
endogenous chemicals (e.g., growth factors, hormones,
cytokines) or by exogenous non-genotoxic chemicals, then
clonal expansion of these initiated cells can occur. This
would constitute the tumor promotion phase. The process
of the down regulation of GJIC by both classes of
chemicals is reversible. Therefore, the process of tumor
promotion is either interruptible or even reversible (42).
Because of their inability to terminally differentiate (43)
these initiated cells will slowly accumulate as a focus of
benign mono-clonally derived focus of non-terminally
differentiated cells. The observation that tumor-promoting
chemicals seem to block apoptosis of cells, the
accumulation of initiated cells would also contribute to the
clonal expansion of these non-terminally differentiated
cells (19,44). Cell death or surgery could act as an
“indirect tumor promoter” by releasing a surviving initiated
cell from mitotic suppression (40).
When during this clonal expansion process of the
initiated, partially differentiated, but gap junctionally-
coupled cells accrues other stable events in the genome
(either mutational or epigenetic transcriptional repression
of genes) that brings about a genomic inhibition of GJIC
(e.g., activation of an oncogene; de-activation of a tumor
suppressor gene; mutation or transcriptional repression of a
gap junction gene or a cell adhesion molecule), then the
progression phase of carcinogenesis could occur.
Clearly, an individual can inherit mutated or
possibly altered imprinted genes that directly affect the
initiation or promotion phases of carcinogenesis (e.g.,
xeroderma pigmentosum-DNA repair deficient and
hypermutable syndrome; Bloom’s syndrome) or that
affects an oncogene or tumor suppressor gene (i.e., Li-
Fraumeni syndrome). On the other hand, agents in the
environment (e.g., ultraviolet light and asbestos) can
influence one or more stages of carcinogenesis. These
illustrate that the “nature and nurture” theory of
carcinogenesis is the appropriate way to conceptualize
the role of the interaction of genes and environmental
factors (39).
5. STEM CELL VERSUS THE DE-
DIFFERENTIATION THEORIES.
One of the major debates in carcinogenesis
concerns the question whether all cells of the
multicellular organism are potential targets for
carcinogenesis or whether only few special cells can
given rise to cancer. One of the universal
characteristics of a cancer cell is that it appears to be
“immortalized” and partially, but not terminally-
differentiated. Normal cells appear to be “mortal” and to
have the capability to become terminally differentiated. It
is therefore important to define some terms. Stem cells
ought to be characterized: a “toti-potent cell is one that can
give rise to all cell types within the organism. A
pluripotent stem cell is derived from the toti-potent stem
cell and has been restricted (committed) to give rise by
what appears to be a finite number of cell divisions to only
a specific lineage of cell types within the organ in which it
gives rise. The daughter cells (progenitor cells) of these
pluripotent stem cells which are limited to give rise to one
cell type would give rise to the terminally-differentiated
cells of that lineage. By definition these terminally-
differentiated cells can never proliferate.
One of the prevailing paradigms in the cancer
field is that the first major step in carcinogenesis is for
a “mortal”, normal cell to be “immortalized” and then,
subsequently neoplastically transformed. While it
would seem obvious that some terminally differentiated
cells could not support the De-Differentiation theory
(e.g., red blood cells, polyploid hepatocytes, lense cells,
neurons, keratinocytes), proponents of this hypothesis
argue that some differentiated cells (progenitor cells)
could be plastic enough to revert back to a early
progenitor or pluripotent cell.
While cells of a tumor appear to be derived
from a single cell (“mono-clonal” theory), this does not
argue in favor of the stem cell over the de-
differentiation theory. While the evidence does not
rigorous support one theory over the other, several lines
of evidence seems mere consistent with the stem cell
theory. First, if one defines a stem cell as a cell that
has the capacity to divide asymmetrically (i.e., one
daughter is committed to terminally differentiate and
the other daughter must remain “stem-like”, then the
stem-like cell is, by definition “immortal”, while the
other is committed to become “mortal”. If that is the
case, then the first step in carcinogenesis is not to
“immortalize” a normal, “mortal” cell, but to prevent
the “ mortalization” of a normal, immortal stem cell
(see “initiation/ promotion/progression” theory below).
The second line of evidence in favor of the
stem cell theory is the observation made by Nakano et
al, (45) that there are only a few target cells in a
population of Syrian baby hamster cells which are
susceptible to neoplastic transformation. These cells
were operationally characterized as being “contact-
insensitive”.
The third line of evidence comes from
experiments which isolated and characterized
presumptive pluripotent stem cells from human kidney
and human breast tissue (32,46). These studies indicate
that pluripotent stem cell have no expressed connexin genes
or functional GJIC. Cancer cells are also characterized
Gap junctions and cancer
213
Figure 2. The initiation/promotion/progression model of
carcinogenesis. β1 = rate of terminal differentiation and
death of stem cell; β2 = rate of death, but not of terminal
differentiation of the initiated cell (à); α1 = rate of cell
division of stem cells; α2 = rate of cell division of initiated
cells; µ1 = rate of the molecular event leading to initiation
(i.e., possibly mutation); µ2 = rate at which second event
occurs within an initiated cell. (From Trsoko et al., In:
Modern Cell Biology; Vol. 7, Gap Junctions, E.L.
Hertzberg and R.G. Johnson, eds., pp. 435-448, 1998; with
permission from Alan R. Liss, Inc., New York).
with the same phenotype (41). The major difference is that
the normal human stem cell can be induced to differentiate
very easily, whereas, while some cancer cells can be
induced to differentiate, it is much harder to do. The fact
that some cancer cells can be induced to terminally
differentiate is, itself, evidence they behave as pluripotent
stem cells (47).
The recent hypothesis that telomerase activity is
necessary for immortality was derived from observations
that in normal cells the telomerase activity decreases as the
cell senescences while in immortal and neoplastic cells the
telomerase activity is “restored”(48). This hypothesis does
not seem to hold with recent demonstration that human
breast epithelial pluripotent stem cells have high telomerase
activity which remains high when they are prevented from
terminally differentiating by SV40 transformation. The
activity remains high even when these cells are
neoplastically transformed (49). Only when the normal
pluripotent stem cell is induced to terminally differentiate
does the telomerase activity decrease.
6. INITIATION/PROMOTION/PROGRESSION
THEORY OF CARCINOGENESIS
The multi-stage nature of carcinogenesis,
originally conceptualized during mouse skin chemical
carcinogenesis studies (42), consisted of the operational
concepts of “initiation”, “promotion”, and “progression”
(figure 2). No mechanism for each step can be directly
implied from these experiments. However, because the
initiation phase appears to be irreversible, and relatively
easily induced by DNA damaging agents, mutagenesis has
been the implied underlying mechanism responsible for this
step (50). It can not be rigorous ruled out, however, that
stable “epigenetic” events, which might transcriptionally
alter some proto-oncogene or tumor suppressor gene, could
also be an initiating agent.
Biologically, the initiating event, if it takes place
in a pluripotent stem cell, must prevent the “mortalization”
or terminal differentiation of a normal, immortal”
pluripotent stem cell. Some evidence from mouse skin
“initiation”/ promotion studies does seem to indicate that
initiated mouse skin cells do not terminally differentiate
under conditions were normal skin cells do and that these
non-differentiated cells, when place back into the mouse,
give rise to papillomas, indicating they still retain their
proliferative potential. This initiation event seems to tie the
stem cell theory to the multi-step theory of carcinogenesis.
The promotion phase of carcinogenesis,
operationally, is an interruptible process (and reversible up
to a certain point) (42). This implies that the initiated cell
can be stimulated to proliferate but not terminally
differentiate. The promotion process can be implied to be
an epigenetic process. Mitogenesis, rather than
mutagenesis, best describe the promotion process (40).
Also, since tumor promoters appear to block apoptosis,
these initiated cells can accumulate as dysfunctional, non-
differentiated cells within a tissue (e.g., enzyme-altered foci
in the liver; nodules in the mammary gland; polyps in the
colon or papillomas in the skin).
One of the first hypothesis concerning the
mechanism of tumor promotion was derived from
observations that the skin tumor promoters, phorbol esters,
could block gap junctional intercellular communication, at
non-cytotoxic levels, in a reversible fashion (51,52). More
recently, it was also postulated that tumor promoters could
inhibit apoptosis at the same time they blocked GJIC (19).
This infers that there might be a direct connection between
GJIC and the death signal being transmitted through gap
junctions.
Less is known about the mechanism(s) of
carcinogenic progression. It does appear to be the step
conferring autonomous growth of the initiated cell,
rendering it independent of exogenous tumor promoters. It
does seem to be an irreversible process, implying either a
mutagenic event or a stable epigenetic event.
7. NATURE AND NURTURE THEORY OF
CARCINOGENESIS
The idea that genetics might play a role in
carcinogenesis was not widely accepted until relatively
recently, in spite of the fact it was well known for decades
that there were hereditary syndromes that predisposed
individuals to cancer and that cancer cells could have
chromosomal aberrations. “Is cancer caused by the
individual’s genes or by the environment?” was the
question often heard in both scientific and lay circles.
When the paradigm, “ Carcinogen as mutagen” appeared
(52), an interesting new insight was formed that linked
“nature” and “nurture” together (40). If agents which were
Gap junctions and cancer
214
mutagens and they were found in the wide environment
(nurture), they must interact with DNA of the germ and/or
the somatic tissue (nature) to induce mutations in various
genes that affect the cancer process directly (e.g.,
oncogenes or tumor suppressor genes) or indirectly (e.g.,
DNA repair or initiator-prone genes; growth factor or
promoter prone genes). It was only when molecular
determinations of specific mutations were found in
oncogenes or tumor suppressor genes found in the tumor
cells that the question was resolved. The “nature” versus
“nurture” debate seemed irrelevant in favor of the “nature
and nurture” A classic example that would the xeroderma
pigmentosum syndrome which inherits a defective DNA
repair gene. The cells of these individuals are incapable of
removing sunlight induced pyrimidine dimers formed in
the DNA of skin cells, unlike the non-xeroderma
pigmentosum individual. As a result, these unrepaired
DNA lesion can act as substrates for both cell death or
mutations. If these mutations occurred in a skin stem
cell and in an oncogene or tumor suppressor gene in
these cells, initiation would take place. Other cells,
unable to repair these lesions could die. As a result, if
an initiated cells is stimulated to proliferate as a result
of other cells dying (i.e., compensatory hyperplasia),
then promotion of these surviving initiated cells could
occur (41,53).
This theory of carcinogenesis readily integrates
the stem cell and initiation/promotion/progression theories
of carcinogenesis.
8. MUTATION VERSUS THE EPIGENETIC
THEORIES OF CARCINOGENESIS
While the mutation theory of carcinogenesis
has had a long history [i.e., Boveri (33)], the idea that
non-mutagenic events might play a role during either
the whole of, or some phase of, carcinogenesis. As
mentioned above, the “Carcinogen as mutagen”
paradigm, together with the recent direct measurements
of mutations in oncogenes and tumor suppressor genes,
has almost become a “state religion”. By that we mean
it has paralyzed some investigators to think that
mutagenesis, alone, can explain all of carcinogenesis.
First, the definition of these two terms must be
understood. Mutagenesis is that process that brings
about a qualitative or quantitative alteration of the
genetic information. An epigenetic process is that
which alters the expression of the genetic
information at the transcriptional, translational or
posttranslational levels. In principle and in reality,
there can be chromosomal mutations (i.e., a
translocation or a non-disjunction of a chromosome)
that can induce an epigenetic event (i.e., The extra
chromosome 21 in Downs Syndrome can alter gene
expression without a point or gene mutation).
While the role of mutations in the carcinogenic
process appears well documented, the role of epigenetic
events is not as well documented or as widely accepted.
Yet, it is very well documented that in the
initiation/promotion/progression model of carcinogenesis,
tumor promoting chemicals, such as phorbol esters, DDT,
phenobarbital, saccharin, polybrominated biphenols,
peroxisome proliferators, etc., are non-mutagenic (55).
Promotion is the mitogenic process that brings about the
clonal expansion of initiated cells. Cell proliferation is an
epigenetic process by which a mitogenic stimulus triggers a
signal transduction pathway which can block gap junction
function, posttranslationally, and turn on cell cycle genes,
transcriptionally (56). The fact that tumor promoting
chemicals, such as phorbol esters, can activate signal
transducing protein kinases, alter the phosphorylation of
connexin proteins, at non-cytotoxic doses and effect 100%
of the exposed cells rules against it acting as a “
mutagen”(57,58). Also, the blockage of GJIC by the tumor
promoters is reversible. These are not characteristics of a
mutagen which acts in a random fashion on genes and is,
for practical purposes, an irreversible event.
If there must be multiple “hits” to give a normal
cell those phenotypic features, such as “immortality” or
inability to terminally differentiate, to lose contact
inhibition, to invade tissues and to metastasize, and if these
“hits” are in large part the result of mutations, then with the
average mutation rates of genes being relatively low, the
probability of accruing all these mutations in the one cell
that was originally initiated would be the product of the
individual low probabilities of each independent event.
One would never get a cancer if that were all that was
needed. However, if in the first cell, a mutation occurred in
a gene that prevented terminal differentiation but not
proliferation, then one needs only to stimulate this initiated
cell to proliferate a few times to produce a few million
cells. At that time, the target size of cells with one hit
increases the probability of a spontaneous mutation to
occur in one of these cells. This cell could then proliferate
a few times to create a target size of cell with the original
“hit” and a second “ hit”, increasing the likelihood that a
third “hit” could take please.
It seems clear that both mutagenic and epigenetic
mechanisms need to take place to bring about the
complicated, multi-step carcinogenic process. This
mutation and epigenetic theory of carcinogenesis can be
integrated into the stem cell, initiation/promotion/
progression and nature and nurture theories of
carcinogenesis.
9. ONCOGENE AND TUMOR SUPPRESSOR GENE
THEORIES
In recent years the concept of oncogenes was
derived from the fact that specific DNA sequences in tumor
cells, when injected into a non-tumorigenic recipient cell,
could transform it into a neoplastic cell. Over the years,
using this protocol, a number of “oncogenes” have been
identified and they have been characterized as falling into
four functional classes of genes. The counterpart of these
“activated” or mutated oncogenes in normal, non-
tumorigenic cells are called “ protooncogenes” and they
Gap junctions and cancer
215
code for (a) growth factors; (b) growth factor receptors; (c)
cellular signal transducing proteins; or (d) nuclear
transcription factors. These genes and their normal coded
proteins appear to be involved in the regulation of cell
proliferation, differentiation or apoptosis. If they are
somehow overexpressed or mutated to an unregulated form,
these cells contribute to the phenotype of cancer cells,
namely they do not contact inhibit or proliferate in an
uncontrolled fashion nor do they terminally differentiate or
apoptosis normally.
For these reasons, an oncogene-activated cell
resembles a stem cell which is initiated by an mutagenic
event (i.e., it can not terminally differentiate but can
proliferate) or which, in a limited way, it resembles a
normal progenitor cell treated with a tumor promoting
chemical such as phorbol ester (i.e., it transiently blocks
contact inhibition, enhances proliferation, modifies
differentiation) (59). It has been shown that phorbol ester-
treated myc- transformed cells [which do not form tumors
by itself] behave very much like the v-Ha-ras oncogene
(60-62). However, although by itself, the v-Ha-ras
oncogene product can trigger a mitogenic signal
transduction mechanism in cells, in the non-tumorigenic
cells used in the oncogene protocol system, it is not
sufficient to bring about full neoplastic transformation.
Together with an activated myc oncogene or phorbol ester,
the ras oncogene can induce the stable neoplastic
phenotype with myc or a transient phenotypic neoplastic
phenotype with phorbol ester. Both v-Ha-ras and phorbol
esters can also inhibit GJIC, stabily or transiently,
respectively.
Together with the observations that several
other oncogenes [e.g., src, neu, mos, raf] (41), can also
down-regulate GJIC concurrent with the neoplastic
transforming properties of their coded signal
transducting potential, the oncogene theory, not only is
linked to the stem cell theory, but the
initiation/promotion/progression theory.
Tumor suppressor genes, conceptually, must
do the opposite of the biological effects on oncogenes;
namely, they must suppress cell growth and assist in the
differentiation and apoptosis of cells. Operationally,
these DNA sequences are identified by transfecting
neoplastic cells and showing that they stop unregulated
cell growth, induce differentiation or apoptosis.
Recently, several studies have identified the existence
of tumor suppressor gene on the normal chromosome
11 (63-72). Interestingly, in these neoplastic cells
which do not have function GJIC, the non-neoplastic
cells containing a genetically-engineered normal
chromosome 11 have normal growth control, loss of
their neoplastic potential and enhanced GJIC (73).
These latter observation with a tumor suppressor
gene conceptual support the link between growth control
regulation, and its oncogene counterpart of the inhibition of
growth control, with GJIC and both the stem cell and
initiation, promotion, progression theories. More will be
said about the similarities of tumor suppressor genes, anti-
tumor promoters and GJIC.
10. GAP JUNCTIONS-ANCIENT AND UBIQUITOUS
MEDIATORS OF CELLULAR HOMEOSTASIS
As discussed above, the major premise of this
review is that dysfunctional GJIC plays a crucial role in the
tumor promotion phase of carcinogenesis and that all of the
major theories of carcinogenesis can be integrated if
examined from this perspective.
Cancer is a disease of “abnormal homeostasis”
mediated by defects in intra-, extra-, and intercellular forms
of communication that disrupt the delicate balances
between cellular proliferation, differentiation, apoptosis,
and adaptation. One of the most ubiquitous and ancient
forms of intercellular communication that is disrupted in
neoplastic cells is that mediated by gap junctions. In the
remainder of this review, we will discuss the structural and
functional aspects of gap junctions, the proteins that form
gap junction channels (connexins), the evidence that gap
junctions are involved in cellular growth regulation and
how this might work, and lastly how the enhancement of
GJIC and connexin expression might be beneficial in
cancer therapy.
11. STRUCTURE OF GAP JUNCTIONS
Gap junctions are one type of junctional complex
formed between adjacent cells and consist of aggregated
channels that directly link the interiors of neighboring cells.
Gap junctions have been detected in such primitive
invertebrates as jellyfish and hydra and similar structures
known as plasmodesmata are found in plants (74-76). In
the adult mammal, gap junctions are found in most cell
types with the known exceptions being skeletal muscle
fibers, certain neurons, and circulating blood cells (74,75),
although some blood cells may express gap junction
proteins and gap junction-like structures (77).
Each gap junction channel is comprised of two
hemichannels or connexons and each connexon is formed
by the aggregation of six protein subunits known as
connexins (figure 3) (78). Connexins are folded in the
plasma membrane in the approximate shape of an “M”.
The amino and carboxyl termini project into the cytoplasm
while the remainder of the molecule traverses the plasma
membrane four times. These membrane-spanning regions
lie in parallel. The third one contains a high proportion of
hydrophilic amino acids and is thought to line the interior
of the channel. The four membrane-spanning domains and
the extracellular loops are highly conserved between the
many different connexins that have been cloned (described
below). More varying are the cytoplasmic regions. As will
Gap junctions and cancer
216
Figure 3. Diagrammatic representations of gap junctions
and connexins. (A) Gap junction between two cells. Gap
junction channels are too small to enable the cell-to-cell
passage of macromolecules such as proteins, but will allow
free cell-to-cell diffusion of micromolecules (<2,000 Da)
such as calcium ion, water, and cyclic nucleotides. (B)
Structure of a connexin. All known connexins are
transmembrane proteins with four transmmbrane domains
and cytoplasmic amino and carboxyl termini. The third
transmembrane domain contains many hydrophilic amino
acids and is thought to line the gap junction channel. Some
connexins are phosphorylated on the carboxyl tail. (C)
“End-on” view of a gap junction hemichannel (connexon).
Two connexons align end-to-end to form a complete
channel between neighboring cells. Each connexon is
formed from six connexins; the third transmembrane
domain of each connexin lines the channel.
Table 1. Mammalian connexins and organs or tissues in
which they are found.
CONNEXIN TISSUE
Cx26 Parenchymal liver (hepatocytes),
pancreas, endometrium
Cx30 Brain, skin
Cx30.3 Skin
Cx31 Skin, placenta
Cx31.1 Skin
Cx32 Parenchymal liver (hepatocytes),
kidney, pancreas
Cx33 Testes
Cx37 Endothelium, lung, ovary
Cx40 Endothelium, smooth muscle,
myocardium, lung
Cx43 Most epithelia, heart, uterus, connective
tissue, brain
Cx45 Kidney, skin
Cx46 Lens
Cx50 Lens
be discussed, these different regions may be involved in the
cellular regulation of gap junction formation and channel
permeability. Connexin folding as well as connexin-
connexin and connexon-connexon interactions are
mediated through disulfide bonds, hydrophobic protein
interactions, and other more poorly understood forces.
Gap junction channels have a diameter of
approximately 1.5-2 nm depending upon the type of
junction-forming protein and are large enough to permit the
direct diffusion of small (<2,000 Da) molecules and ions
between cells (74). Many substances such as ions, water,
sugars, nucleotides, amino acids, fatty acids, small
peptides, drugs, and carcinogens are small enough to move
between cells through gap junction channels. However,
proteins, complex lipids, polysaccharides, RNA, and other
large molecules are not. Channel passage does not require
ATP and appears to result from passive diffusion. This
flux of materials between cells via gap junction channels is
known as gap junctional intercellular communication
(GJIC).
One of the most significant physiological
implications for GJIC is that gap junction "coupled" cells
within a tissue are not individual, discrete entities, but are
highly integrated with their neighbors. This property
facilitates homeostasis and also permits the rapid, direct
transfer of second messengers between cells to coordinate
cellular responses within the tissue. On the other hand, the
channel permeability size limit enables cells to maintain
enzymatic and other macromolecular functions through the
retention of specific enzymes, receptors, and other large
molecules. Thus, GJIC can be viewed as an ancient,
common, and important mechanism of cellular homeostasis
and integrator of cellular functions within complex tissues.
12. THE CONNEXIN MULTIGENE FAMILY
The gap junction channel-forming connexins
comprise a multi-gene family with at least thirteen
mammalian connexins discovered thus far (78). Several
homologous DNAs have been identified in other vertebrate
species. Several connexins and tissues in which they are
highly expressed are listed in table 1. The number
associated with each connexin indicates its molecular mass.
Connexins are expressed in a cell-, tissue-, and
development-specific manner. For instance, connexin43 is
the predominant connexin expressed in cardiac muscle and
was first cloned from this tissue (79), although other
connexins (connexin40, connexin45, and connexin46) have
also been detected in cardiac tissue. In adult liver, the
predominant connexins are connexin32 and connexin26
(80-82) which are expressed by adult parenchymal liver
cells (hepatocytes). However, nonparenchymal liver
epithelial cells, hepatic fat storing (Ito) cells, and hepatic
connective tissue cells express connexin43 (83-85).
Connexin 37 and connexin40 are highly expressed in lung
endothelial cells (86,87), whereas connexin43 is the
predominant connexin expressed by human and murine
lung epithelial cells (88). Lee et al. (89), however,
reported that connexin32 was the most abundant connexin
Gap junctions and cancer
217
in freshly isolated rat type II alveolar epithelial cells and
that connexin43 expression arose after culture. Connexin
expression is also related to cellular differentiation. For
example, the types of connexins expressed by epidermal
cells changes as the cells differentiate and move away
from the basal layer (90,91).
Despite the presence of conserved sequences
within connexins, the diversity of these proteins is not
due to alternative splicing of RNAs. Instead, there
appears to be one gene per connexin. Many connexin
genes have been mapped and are located on several
chromosomes (92) suggesting their distribution is random
throughout the genome. Why there are so many
connexins is not clear, but may reflect differences in their
function(s) and/or the regulation of their expression,
formation, and permeability.
In those cells where multiple connexins are
expressed, gap junction hemichannels may be comprised
of more than one connexin (heteromeric connexons) or
may be homomeric (93-96). Heteromeric channels may
have different permeability and regulatory properties than
homomeric ones and may provide numerous additional
options for regulating the type of signals that pass
through its gap junction channels. In addition,
heterotypic channels consisting of two different
connexons have been described (97,98). Interestingly,
only certain heterotypic pairings form functional
channels. For example, hemichannels of connexin43 and
connexin26 and of connexin32 and connexin26 form
permeable channels, but those of connexin43 and
connexin32 do not (99).
13. REGULATION OF CONNEXIN GENE
EXPRESSION
The structures of nearly all connexin genes
identified thus far are similar and consist of two exons
separated by a long intron (78). The first exons are quite
short (about 100 base pairs), contain no protein coding
information, and are separated from the second exons by
long introns (6-8 kilobasepairs). All of the coding
information for the proteins resides within the second
exons. The promoters that have been identified for these
genes are located upstream of exon 1.
The rat connexin32 gene is unusual, however.
Initial work indicated that it consisted of a small first
exon separated from a second exon by an approximately
6 kb intron and that the promoter was located upstream of
exon 1 (100); this is the typical connexin gene structure
just described. Recently, however, two additional
promoters have been identified which are active in neural
tissue, but not in the liver. These promoters are located
more proximal to exon 2 and give rise to mRNAs
containing the noncoding, alternate exons 1A or 1B and
exon 2 (101,102). Thus, in contrast to the other connexin
genes, three promoters control the expression of
connexin32 and function in a tissue-specific manner.
The regulation of connexin gene expression is
poorly understood. Besides the connexin32 gene, many
other connexin gene promoters have been sequenced and
several putative regulatory sites have been identified
(103-111). Few of these sites have been examined for
function, however.
In the rat and mouse connexin32 promoters
located upstream of exon 1 and active in liver tissue,
there does not appear to be a canonical TATA-box
sequence (100,103). Instead, several GC-rich sequences
near the transcription start sites appear to bind the
transcription factor Sp1 and positively regulate
connexin32 transcription (107,108). In addition, there is
a novel 25 bp element within this region that also appears
to be necessary for connexin32 transcription (108).
Hepatic nuclear factor-1 and nuclear factor-1 consensus
sequences are also present in this basal transcriptional
region. DNAse I hypersensitive sites reside further
upstream in this promoter and may silence connexin32
expression in nonexpressing cells (108). The alternate
connexin32 promoters active in neural tissues contain
canonical TATA boxes (101,102), but their functional
activity has not been reported.
The promoters of human and mouse connexin26
have been sequenced and several interesting elements
have been noted (103,106). Like the connexin32 hepatic
promoter, there does not appear to be a canonical TATA
box, but a TTAAAA motif is present in the human
promoter approximately 20 bp upstream of the
transcription start sites (106). Additionally, several GC-
rich sequences, a Yin-Yang (YY)-1-like element, and
consensus mammary gland factor binding sites are
present further upstream (103,106). The functional
activity of these elements has not been reported, however.
The promoters of the human, rat, and mouse
connexin43 genes have been cloned (104,105,109). All
three promoters contain a TATA box located near the
transcriptional start sites and several activator protein
(AP)-1 sites further upstream. The AP-1 site(s) in the
human connexin43 promoter function positively since
phorbol ester treatment of human uterine myometrial
cells increased connexin43 expression within 6-8 h and
mutation of the most proximal AP-1 site reduced the
response (110). The increase in connexin43 expression
was preceded by transient increases of c-Fos and c-Jun
which also supports the functional role of the these sites;
in response to phorbol esters, c-Fos and c-Jun
heterodimerize then bind to AP-1 sites. The
enhancement of connexin43 expression by phorbol ester
may seem paradoxical since 12-O-tetradecanoylphorbol-
13-acetate (TPA), the classical phorbol ester tumor
promoter, blocks GJIC in most connexin43-expressing
cells. However, this block occurs rapidly (within
minutes) and is due to activation of calcium-
phospholipid-dependent protein kinase (protein kinase C;
PKC) phosphorylation of connexin43 rather than to
decreased connexin43 gene transcription (112,113).
Gap junctions and cancer
218
Figure 4. Formation of gap junctions containing
connexin43. The connexin43 gene is expressed and the
messenger is translated into protein. Six connexin43
subunits oligomerize in the Golgi apparatus into a
connexon (hemichannel) and which is transported to the
cell surface. There the connexons align with connexons
from the neighboring cell to form complete channels.
These aggregate into large gap junction plaques that may
contain hundreds to thousands of channels. During the
pairing of connexons and aggregation into plaques,
connexin43 is phosphorylated at least twice. This results in
three species of connexin43 that can be detected by
Western blot, namely connexin43-NP which is not
phoshorylated and connexin43-P1 and connexin43-P2
which are phosphorylated.
The rodent and human connexin43 promoters
also contain several half-palindromic estrogen-responsive
elements (EREs) that enhance connexin43 transcription in
the presence of estrogen (111). In addition, negative and
positive elements were identified within approximately the
first 100 nucleotides of the mouse promoter (109).
Several physiological, pharmacological, and
dietary factors alter the expression of connexins. These
factors often affect connexin expression in a cell-specific
manner. For example, the expression of connexin32 and
connexin26 in hepatocytes, but not in liver epithelial cells,
is increased by glucocorticoids (114-116). The
mechanism(s) have not been identified. As noted above,
estrogen increases connexin43 transcription presumably
through putative estrogen-responsive elements (111).
Cyclic AMP, forskolin, and glucagon increase connexin32
and connexin43 transcription in certain cells (117-121) and
this may occur through activation of cAMP-responsive
enhancer elements located in the connexin32 and
connexin43 promoters (90,103).
Connexin messenger RNA and protein stability
also appear to be important in the expression of connexins.
Sequences located in the 3’-untranslated region of the
connexin43 mRNA appear to stabilize this messenger
(104). Besides increasing connexin gene transcription,
cyclic AMP may also increase the permeability of
preexisting gap junction channels, enhance the formation of
gap junctions from preexisting connexin, and decrease
connexin degradation (117,122,123). These effects may
involve connexin phosphorylation by cAMP-dependent
protein kinase (protein kinase A; PKA) (117).
In summary, several transcriptional and
posttranscriptional mechanisms are involved in the
regulation of connexin expression. A better understanding
of these mechanisms will lead to the development of
therapies designed to alter connexin expression for the
treatment of diseases such as cancer (see below).
14. GAP JUNCTION FORMATION, CONTROL OF
CHANNEL PERMEABILITY, AND MECHANISMS
OF DISRUPTED GJIC
The mechanisms of gap junction formation
and regulation of channel permeability are poorly
understood. The most complete knowledge regarding
gap junction formation is for connexin43 (Cx43) and
is based largely upon he work of Musil and
Goodenough (57,124) (figure 4). Their data suggest
that six Cx43 subunits oligomerize into connexons in
the Golgi apparatus and are then transported to the
plasma membrane. At this point, the connexons are
closed to prevent leakage of cellular contents and
entry of extracellular materials. At the plasma
membrane, the connexons are attracted to those on the
adjacent cell by poorly understood forces and two
connexons join in an end-to-end manner to form a
complete channel. Subsequently, the channels
aggregate into large gap junction plaques and the
channels open to connect the two cells, although the
order of these last two steps is controversial.
Coincident with the formation of open gap junctional
channels and aggregation of particles into junctional
plaques, connexin43 is phosphorylated in at least two
positions. The kinase(s) that perform these
phosphorylations are not known, but a likely
candidate is PKA. This phosphorylation may increase
particle aggregation, channel permeability, and/or
connexin43 stability as discussed above.
Gap junction channel formation also requires
appropriate cell-cell adhesion. The cadherins appear
to be especially important in this regard since gap
junction formation was induced in connexin-
expressing, cadherin-deficient, noncommunicating
cells after transfection with cadherin expression
vectors and cadherin antibodies blocked gap junction
formation (57,125). In addition, membrane protein
glycosylation can also impair gap junction formation.
Treatment of noncommunicating cells with an
inhibitor of glycosylation induced gap junction
formation (127,128).
Rates of connexin synthesis and degradation
and the disassembly and removal of gap junctions
from the cell surface are also important in the
regulation of GJIC. Biochemical studies have
demonstrated that connexin43 and connexin32 have
half-lives (hours) that are much shorter than most
plasma membrane proteins (days) (129-133).
Gap junctions and cancer
219
Gap junction degradation proceeds by several
pathways. Gap junctions containing connexin43 are
internalized and degraded in lysosomes or are ubiquinated
and degraded in proteosomes in response to various
physiological or toxicological cues. For example, the loss
of gap junctions from rabbit granulosa cells during
maturation of the ovarian follicle occurs by gap junctional
internalization and lysosomal degradation (134). We have
reported that certain pesticides induce connexin43 gap
junction internalization and lysosomal degradation in rat
liver epithelial cells (135). In addition, connexin43 has
been reported to be ubiquinated and degraded via the
proteosome pathway and to be degraded in lysosomes
(136,137). The relative contribution of each pathway may
be cell specific. Degradation of connexin32-containing gap
junctions can be mediated by the calcium-activated
proteases, µ-calpain and m-calpain (138,139) and
connexin50 is processed in the ocular lens by calpain
(140). Thus, there appear to be several pathways of
connexin degradation and the contribution of each are cell-
and connexin-specific.
After aggregation into plaques, gap junction
particles are dispersed in the plasma membrane in response
to certain physiological and toxicological cues. For
instance, partial hepatectomy causes dispersal of hepatocyte
gap junction particles (141) as well as altering connexin
expression (discussed above). We have shown that a
compound from licorice root, 18β-glycyrrhetinic acid,
causes the disaggregation of connexin43-containing gap
junction particles and their dispersal in the plasma
membrane (142). Interestingly, this was associated with
the dephosphorylation of connexin43 and supports the role
of phosphorylation in gap junction particle aggregation
(57,124). Another group has observed gap junction
particle dispersal induced by the alpha isomer of this
compound, but did not see changes in connexin43
phosphorylation (143).
Once formed, gap junction channels open and
close and this “gating” is controlled by several mechanisms
including connexin phosphorylation. Several protein
kinases phosphorylate connexins and this may alter
connexin tertiary structure to either open or close the
channels (144). Many protein kinases have been identified
that phosphorylate various connexins or that may indirectly
stimulate connexin phosphorylation by other kinases.
Activation of PKA induces connexin43 and connexin32
phosphorylation on serine residues located in the
cytoplasmic carboxyl tail and this is associated with
increased gap junction number and channel permeability
(145). The phorbol ester-activated protein kinase, PKC,
also leads to increased phosphorylation of connexin43 on
serine residues, but this is associated with decreased
channel permeability (146). In hepatocytes, PKC has also
been reported to induce connexin32 phosphorylation and
an enhancement of channel permeability (147), but in
connexin32-transduced rat liver epithelial cells, PKC
activation results in loss of connexin32-mediated GJIC
(113). The src oncogene product, pp60src, which has
tyrosine kinase activity, blocks GJIC when expressed in
many types of cells and this is due to the phosphorylation
of connexin43 on tyrosine residues in the carboxyl tail
(148). Some connexins such as connexin26 are not
phosphorylated so that gating must be regulated in other
ways.
Other mechanisms regulating channel gating
include intracellular levels of hydrogen and calcium ions,
transjunctional voltage, and free radicals (145). Decreased
pH or pCa induce channel closure in a cell- and connexin-
specific manner. Residues in the intracellular regions of
the protein have been reported to be important in the pH
gating effect (150-151). It is less clear how calcium
regulates gap junction channel permeability (152-154).
The calcium-binding protein, calmodulin, has been
reported to be associated with connexins and this may be
important in calcium-gating of gap junction channels (155-
159). Free radicals are highly reactive molecules and ions
generated during normal cellular metabolism or following
exposure to certain toxic agents. Excessive intracellular
levels of free radicals decrease gap junction channel
permeability (160,161), but the mechanisms are complex.
The radicals may directly attack connexins or other plasma
membrane components (e.g., lipids), or may induce
changes in cellular calcium levels or redox status. It is also
unclear whether radicals are important physiologically in
the regulation of gap junction channel permeability and
turnover. From a pathological standpoint, however, free
radicals may be important in the dysfunctional GJIC in
cardiac myocytes that occurs during ischemic injury
(161,163) and in hepatocytes during certain types of toxic
injury (164).
From the above discussion, it is apparent that the
ability of cells to establish and maintain GJIC is dependent
not only on connexin gene expression, but also on many
other factors including cell-cell adhesion, gap junction
assembly/disassembly, connexin stability, and channel
gating. Not surprisingly, the disruption of GJIC by various
agents often involves these regulatory mechanisms.
15. MULTIPLE FUNCTIONS OF GJIC
Many physiological roles besides growth control
have been proposed for GJIC and several are briefly
reviewed:
1. Homeostasis. GJIC permits the rapid equilibration of
nutrients, ions, and fluids between cells. This might
be the most ancient, widespread, and important
function for these channels (74).
2. Electrical coupling. Gap junctions serve as
electrical synapses in electrically excitable cells
such as cardiac myocytes, smooth muscle cells, and
neurons (74,75). In these tissues, electrical
coupling permits more rapid cell-to-cell
transmission of action potentials than chemical
synapses. In myocytes, this enables their
synchronous contraction.
3. Tissue response to hormones. GJIC may enhance the
responsiveness of tissues to external stimuli (165,166).
Gap junctions and cancer
220
Second messengers such as cyclic nucleotides,
calcium, and inositol phosphates are small enough to
pass from hormonally activated cells to quiescent cells
through junctional channels and activate the latter.
Such an effect may increase the tissue response to an
agonist.
4. Regulation of embryonic development. Gap junctions
may serve as intercellular pathways for chemical
and/or electrical developmental signals in embryos
and for defining the boundaries of developmental
compartments (167). GJIC occurs in specific patterns
in embryonic cells and the impairment of GJIC has
been related to developmental anomalies and the
teratogenic effects of many chemicals (168).
16. ROLE OF GJIC IN REGULATING CELLULAR
PROLIFERATION AND NEOPLASIA
More important to this review is that GJIC is also
involved in the regulation of cellular growth and expression
of the neoplastic phenotype. For some time, gap junctions
have been proposed to serve as passageways for the cell-to-
cell exchange of low molecular weight growth regulatory
molecules (169). GJIC is frequently reduced in neoplastic
and carcinogen-treated cells. It was hypothesized that this
contributed to dysregulated cellular growth by isolating
cells from their neighbors (170,171). As will be described
below, there is now compelling evidence that this is true.
16.1 Neoplastic Cells Have Fewer Gap Junctions
Gap junction size and number, connexin
expression, and cell-cell coupling (GJIC) have been studied
in many neoplastic cells using ultrastructural, biochemical,
and immunological means and by introduction of
fluorescent or radioactive tracers and determination of
tracer passage into adjacent cells. The vast majority of
neoplastic cells have fewer and smaller gap junctions,
express less connexins, and have reduced GJIC compared
to their nonneoplastic counterparts (74,88,172,173).
There are exceptions, however, in that some
neoplastic cells have normal or greater gap junction
expression and cell-cell coupling (88,170,173-178). This
does not indicate, however, that such cells communicate
normally. Loewenstein (170) has pointed out that cells can
have defective GJIC at several levels. First, they may lack
functional (permeable) gap junctions. Secondly, they may
have functional gap junctions amongst themselves
(homologous GJIC), but are incapable of gap junction
formation with nontransformed cells (heterologous GJIC).
Thirdly, they may form gap junctions, but are insensitive to
the gap junction signals that control growth and phenotype.
Finally, they may have other defects that are sufficient to
neoplastically transform the cell despite functional GJIC.
In support of the second possibility, our group (88,173)
and others (175-178) have identified several neoplastic cell
lines that have extensive gap junction formation, connexin
expression, and homologous GJIC, but little heterologous
GJIC with their nontransformed counterparts. If this
deficiency occurs in vivo, the tumor cells would be isolated
from the junctional regulatory influences of their
surrounding normal cells. This inability of tumor cells to
communicate with normal cells is not due to the expression
of different connexins, but possibly to differences in the
cell surfaces (e.g., glycosylation or cell-cell adhesion
molecules) that prevent adequate cell-cell contact needed
for gap junction formation.
16.2 Growth Stimuli Inhibit GJIC
16.2.1. Carcinogens
A variety of chemical carcinogens have been
identified that enhance or “promote” neoplastic
transformation through mechanisms that do not involve
direct damage of DNA. Many of these “nongenotoxic
tumor promoters” selectively induce the proliferation
and/or inhibit the programmed death (apoptosis) of
preneoplastic cells (179). This leads to clonal expansion of
the preneoplastic cell population and increased risk of
additional genetic changes occurring in these cells that lead
to full neoplastic transformation. Importantly, GJIC may
play a role in this proliferative response. Most of the tumor
promoters that have been examined (over 100) inhibit GJIC
in cultured cells and cells within target tissues (180-182).
The ability of tumor promoters to inhibit GJIC is one of
their most common properties. These compounds are
chemically diverse and include pesticides such as DDT,
dieldrin, and lindane; pharmaceuticals such as
phenobarbital and diazepam; dietary additives such as
saccharin and butylated hydroxytoluene; polyhalogenated
hydrocarbons such as dioxin; and peroxisome proliferators
such as clofibrate. Not surprisingly, these agents inhibit
GJIC through several diverse mechanisms (discussed
below).
Preneoplastic cells are more sensitive than
normal cells to the inhibitory effects of tumor promoters on
GJIC (183-186). This differential response agrees with the
selective proliferation and clonal expansion of
preneoplastic cells versus normal cells as discussed above.
In contrast to nongenotoxic tumor promoters,
most studies indicate that mutagenic (“genotoxic”)
carcinogens do not inhibit GJIC or induce cell proliferation
(182,187) although another group has reported that several
classical, mutagenic carcinogens can reduce GJIC (188).
Genotoxic carcinogens instead probably induce neoplastic
transformation by mutationally activating proto-oncogenes
and inactivating tumor suppressor genes (189). Thus,
chemical carcinogens appear to have different effects on
DNA, GJIC, and cell proliferation that correlate with their
genotoxic/nongenotoxic carcinogenic mechanisms.
16.2.2. Oncogenes
Oncogenes are genes derived from normal
cellular genes (proto-oncogenes) that have been
mutationally activated and/or are overexpressed and that
function in the transformation of a normal cell into a
neoplastic one. The protein products of these genes
function in signal transduction, gene regulation, growth
control, and many other facets of tissue and cellular
homeostasis, and not surprisingly, many block GJIC.
Gap junctions and cancer
221
Oncogene products that block GJIC include Ras, Neu,
and Src (173,189). Oncogenes also cooperate in their
ability to reduce GJIC and transform cells. Rat liver
epithelial cells that expressed only raf or myc
oncogenes did not have reduced GJIC and were not
transformed, but GJIC was strongly inhibited and the
cells were highly tumorigenic when both oncogenes
were expressed (176). Similarly, overexpression of the
c-Ha-ras oncogene in rat liver epithelial cells partially
reduced GJIC, but v-myc overexpression did not and
coexpression of the two resulted in nearly complete
inhibition of GJIC and highly malignant cells (62).
16.2.3. Growth Factors
Many growth factors such as epidermal
growth factor, platelet derived growth factor, basic
fibroblast growth factor, hepatic growth factor, and
transforming growth factor-alpha inhibit GJIC when
applied to cultured cells (25). This effect may occur
rapidly (minutes to hours) or may be delayed (days).
The mechanism(s) of inhibition of GJIC by epidermal
growth factor may be related to the stimulation of
connexin phosphorylation by mitogen-activated protein
kinase (MAPK) and closure of gap junctional channels
(190). Platelet-derived growth factor induced the rapid
inhibition of GJIC coincident with the phosphorylation
of connexin43 (191). Basic fibroblast growth factor,
however, inhibited GJIC after long exposures (>8 h)
coincident with decreased connexin43 expression
(192). Some growth factors such as transforming
growth factor beta enhance GJIC in some types of cells,
but decrease it in others (193-196).
Several oncogenes code for growth factors,
growth factor receptors, or mitogenic signal transducing
elements and several tumor promoters act as growth
factors since they induce cell proliferation. As noted
above, oncogenes and tumor promoters have also been
associated with dysfunctional GJIC. Thus, it is evident
that growth factors, oncogenes, and tumor promoters
share the common properties of increasing cell
proliferation and inhibiting GJIC.
16.3. Growth Inhibitors Stimulate GJIC
In contrast to the effects of growth factors,
oncogenes, and tumor promoters on GJIC, many growth
inhibitors and anticancer agents increase GJIC and
connexin expression in target cells (25). Retinoids,
carotenoids, green tea extract, certain flavonoids,
dexamethasone, and cyclic AMP analogues and
agonists inhibit neoplastic transformation and/or tumor
cell growth and can block neoplastic transformation in
some tissues. These agents also increase connexin
expression and gap junction formation in target tissues
(115,197-199) or block the inhibitory effects of tumor
promoters on GJIC (200-202).
Certain tumor suppressor gene products also
increase GJIC in neoplastic cells. As discussed above,
the human chromosome 11 carries one or more tumor
suppressor genes (63-72). Introduction of this
chromosome into neoplastic cells restored normal
growth control, reduced tumorigenicity, and enhanced
GJIC mediated by connexin43 (73) despite the fact that
the connexin43 gene is located on human chromosome
6 (92). This result also suggests that tumor suppressor
gene products inhibit neoplastic transformation by
enhancing GJIC in addition to their known actions on
cell cycle genes, signal trandsuction pathways, and gene
expression (203).
16.4. Cell Cycle-Related Changes in GJIC
GJIC may also have a role in the progression
of dividing cells through the cell cycle. In several
model systems, cell cycle-related changes in GJIC have
been noted (25). For example, this has been
demonstrated clearly in regenerating rat liver. The
adult rat liver parenchymal cells (hepatocytes) are
highly coupled by gap junctions and only a small
proportion (<1%) are undergoing cellular replication.
The majority of the cells are in stationary (G0) phase,
but will enter the cell cycle and begin dividing in a
synchronous manner when two-thirds of the liver is
removed surgically (two-thirds partial hepatectomy).
When followed throughout the cell cycle, gap junctions
and connexin expression decrease dramatically in S-
phase then reappear and persist throughout the rest of
the cycle (204,205).
Cell culture studies have also demonstrated
changes in GJIC at specific stages in the cell cycle (206-209).
Cultured cells normally replicate in an asynchronous manner.
They can be synchronized, however, by first blocking cell
cycle progression at a particular point using chemical agents
or nutrient or growth factor depletion then releasing the cells
from the block. When such cells were analyzed for GJIC,
reductions were noted in late G1 and in mitosis. In one
study, the reduction of GJIC in G1 was correlated with a
change in the phosphorylation state of connexin43 which
could be prevented by inhibitors of protein kinase C (209).
As discussed above, this kinase is one of many that
phosphorylate connexins and alter gap junction permeability.
Thus, both in vivo and in vitro studies have
documented cell cycle-related changes in GJIC. The
specific cell cycle stage(s) at which GJIC are reduced
are not the same in all studies, however. Connexin
expression and phosphorylation may be involved in this
reduced GJIC, but this is not fully established for all
systems nor is it understood how changes in GJIC
contribute to cell cycle regulation.
17. INVOLVEMENT OF GJIC IN THE GROWTH
INHIBITION OF NEOPLASTIC CELLS BY
NONTRANSFORMED CELLS
Cell culture model systems have illustrated
that the growth of neoplastic cells can be inhibited by
coculture with nonneoplastic cells. This effect has been
attributed to noncontact-dependent and contact-
dependent phenomena. For instance, normal cells may
secrete growth inhibitors (e.g., TGF-β) into the culture
Gap junctions and cancer
222
medium that inhibit neoplastic cell growth (210).
Contact with extracellular matrix or plasma membrane
components such as integrins may also trigger growth-
inhibiting processes in neoplastic cells (211-214).
The inhibition of neoplastic cell growth by
normal cells also appears to involve GJIC, however
(215-217). We have studied this using cultured rat
liver epithelial cells (WB-F344 cells) as a model system
(217). In normal WB-F344 cells, high levels of
connexin43 are expressed, numerous gap junctions are
formed, and the percentage of communicating cells is
high (95-100%). Neoplastic transformants of these
cells which were generated by ras and neu oncogene
transfection formed few gap junctions and had low
incidences of communication (20-25%). Coculture of
the two types of cells resulted in the growth
suppression of the neoplastic cells. Inhibition occurred
only when the two types of cells were permitted to
make contact with each other and not when they were
physically separated in the culture dish. In addition,
medium from normal cells was incapable of reducing
the growth of neoplastic cells. These data suggested
that cell-cell contact, not secretion of extracellular
factors was required for growth inhibition. To
investigate whether GJIC between normal and
neoplastic cells was required, GJIC-incompetent
mutants of the normal cells were utilized. These cells
did not inhibit the growth of neoplastic cells despite the
presence of cell-cell contact. When GJIC was restored
in the mutant cells by their transfection with a
functional connexin43 gene, however, neoplastic cell
growth was again inhibited. Thus, the inhibition of
neoplastic WB cell growth by normal cells required
GJIC.
18. SPECIFIC DISRUPTION AND
ENHANCEMENT OF GJIC
The above discussion has reviewed indirect
evidence that GJIC controls cell growth and the
expression of the neoplastic phenotype. The problem
with those studies is that many of the growth inhibitors,
growth enhancers, carcinogens, oncogene products, etc.
have many effects on cells besides altering GJIC so that
the evidence is correlative. Cause and effect cannot be
shown unequivocally. To demonstrate a role for GJIC
in growth control and neoplasia directly, methods to
specifically alter GJIC must be utilized.
Recently, many techniques have been
developed to specifically inhibit or enhance the
expression of a target gene in cultured cells and
animals. Several approaches have been used to
specifically disrupt the function or expression of
connexins and GJIC in nonneoplastic cells. These
include antisense blockage of connexin gene
expression, targeted disruption of connexin genes
(connexin “gene knockout”), and transfection of
defective connexin genes whose protein products block
the function of normal connexin proteins (“dominant-
negative” connexin expression). Additionally, the
specific enhancement of GJIC in neoplastic cells has
been achieved by transfection of functional connexin
genes. As discussed below, these types of approaches
have provided strong, direct evidence that GJIC is
involved in growth regulation and neoplastic
transformation.
18.1. Connexin Antisense Studies
Antisense approaches include treating cells or
animals with short (usually 15-25 nucleotides), single-
stranded DNA or RNA molecules that are
complementary to targeted heteronuclear or messenger
RNA, or transfecting cells with vectors that
continuously generate complementary RNA. Antisense
molecules are thought to inhibit gene expression by
binding to heteronuclear RNA or messenger RNA and
inducing their degradation or by inhibiting translation.
Using antisense approaches, two groups have
inhibited connexin expression in nonneoplastic cells
and shown that this affected their growth. In one study,
connexin antisense-transfected, nontransformed cells
lost their ability to inhibit the growth of cocultured
neoplastic cells (218). In another study, cells treated
with connexin antisense oligonucleotides grew to a
much higher saturation density (i.e., the maximal
number of cells per dish) (219).
18.2. Connexin Gene Knockout
“Gene knockout” involves the disruption of a
target gene by the insertion of a noncoding sequence
through homologous recombination. A connexin43
knockout mouse has been developed (220). This
mutation, which was lethal at birth, resulted in
offspring that had enlarged, abnormally developed
hearts. No neoplasms were evident in the embryos,
possibly because of the young age at death or because
of compensatory communication by the expression of
other connexins. However, cell lines developed from
the embryos exhibited abnormal patterns of growth
(221,222). A connexin32 knockout mouse has also
been developed (223). Interestingly, these mice
exhibited elevated rates of endogenous hepatocyte
proliferation and were more susceptible to
spontaneous and carcinogen-induced hepatic tumor
formation (223).
18.3. Dominant-Negative Inhibition of Connexin
Function Dominant-negative disruption of normal gene
product function involves introduction of a mutated gene
whose protein product interferes with the function of the
normal gene product. Because of the oligomerization of
connexins into hemichannels and the precise architecture
of the gap junction channel, this approach may be
especially fruitful to block GJIC. Recently, several
mutated connexins that posses dominant-negative activity
have been identified (224-227). One of these, a
dominant-negative mutant of connexin26, inhibited
connexin26-mediated GJIC, enhanced cell growth, and
Gap junctions and cancer
223
Figure 5. Illustration of the concept that growth regulatory
signals may pass through gap junction channels. (A)
Negative signals would enter from neighboring cells and
suppress the growth of the receiving cell. (B) Positive,
growth stimulatory signals would diffuse into neighboring
cells resulting in a substimulatory level in the signal-
generating cell. If a cell lacks functional gap junctions, its
growth would not be suppressed by neighboring negative
signals or would be stimulated by its accumulation of
positive signals. This could lead to dysregulated growth.
increased the tumorigenicity of cells transfected with the
mutant gene (227).
18.4. Connexin Transfection Studies
Connexin gene expression has been enhanced in
several poorly expressing malignant cell lines by
transfection of connexin cDNAs (228-236). In these
“recommunicating” neoplastic cells, growth rates in vitro
and/or tumor formation were highly reduced; these effects
often correlated with the extent of reestablished GJIC. In
some cases, the connexin-transfected cells also expressed
altered levels of cell cycle regulatory proteins (233) or
more differentiated functions (235). In another study
(231), the connexin-transfected cells secreted a soluble,
peptide growth inhibitor. This latter finding suggests that
GJIC is integrated with other mechanisms of growth
control.
19. GJIC AND OTHER GROWTH CONTROL
MECHANISMS
Thus, four different approaches have been used
successfully to inhibit or enhance GJIC in a specific way.
These studies provide strong evidence that GJIC is
involved in growth regulation and neoplasia, but do not
exclude the importance of other mechanisms. With most
important biological control processes, multiple, redundant
mechanisms are present to prevent cellular dysfunction
should a single mechanism become defective. Similarly,
we believe GJIC is only one mechanism of growth control
that functions coordinately with others (e.g., growth factor
responses and signal transduction pathways, cell cycle
controls, responses of cells to the extracellular matrix and
cell-cell adhesion molecules, etc.) as previous studies
suggest (231,233). The next decade should provide much
insight into the interplay between these various regulatory
processes.
20. GROWTH REGULATION MEDIATED BY A
GAP JUNCTION SIGNAL
Two basic schemes by which growth may be
regulated by a gap junction signal molecule are shown in
figure 5 and are adapted from Loewenstein (170). Both
inhibitory and stimulatory low-molecular weight signals
might be produced in cells and diffuse to adjacent cells
through gap junction channels. The negative signals would
inhibit cell division and maintain differentiation whereas
positive signals would stimulate growth and prevent
differentiation. The loss of gap junctions or reductions in
channel permeability would isolate cells from the inhibitory
signals of neighboring cells and/or enable the accumulation
of positive signals in signal-generating cells. These effects
would result in the loss of growth inhibition or the
stimulation of proliferation, respectively. If the reduction
of GJIC were sustained, a noncommunicating cell could
expand by clonal growth into a tumor.
While this model is clearly an oversimplification
of growth and tumor formation, conceptually it provides
the tissue with the ability to fine-tune cell number and
function in a closed system. By regulating the amount and
type of positive and negative signals generated, and the
number and permeability of gap junction channels, a
steady-state level of signals and cell number could be
maintained. Changes in cell number following cell loss
(due to toxicity, wounding, etc.) or cell gain (due to cell
proliferation, impaired cell death, etc.) would change the
size of the communicating network. This would result in
an altered steady-state level of signals and would trigger
cell growth or death until the system (tissue) reached its
previous homeostatic, steady-state level.
This gap junction communicating cell network is
a closed system that incorporates all the cells in the tissue
and is, therefore, different than growth control mediated by
extracellular factors and other types of cell-cell or cell-
extracellular matrix interactions. Theoretically, growth
control mediated by GJIC would enable more strict control
Gap junctions and cancer
224
of cell number than would extracellular growth control
mediated by growth factors, hormones, and chalones or
cell-cell and cell-extracellular matrix regulation through
integrins and other membrane components. The
concentrations of the former could not be controlled
precisely amongst all the cells and the latter would be local
and would not integrate all cells in the tissue.
What is the gap junction growth regulatory signal
and how does it work? This is obviously a difficult
question to answer experimentally because there are so
many molecules and ions capable of passing through gap
junction channels. Multiple stimulatory and inhibitory
signals might be involved. The answer, however,
represents the “Holy Grail” of the gap junction/growth
control field. Several criteria for a gap junction growth
regulatory molecule can be proposed to simplify the
problem:
1. The signal must be capable of passing between cells
through gap junction channels.
2. The signal should affect cell growth in some manner.
3. The levels of the signal should oscillate within the cell
that generates the signal and these oscillations should
be important in growth control.
4. The amplitude of the oscillations should be dampened
by GJIC.
5. The oscillatory signal should regulate cell growth
through a known mechanism since it is unlikely that
the passage of a signal through gap junctions per se
would affect growth.
As first suggested by Loewenstein (170), one
possible growth regulatory, gap junction signal molecule is
cAMP. This molecule fulfills the above criteria for a gap
junction growth regulatory signal:
1. Cyclic AMP, which has a molecular weight of 329,
can pass through gap junction channels (165,166).
2. Treatment of cells with cell-permeable cAMP
analogues or agonists elevates cAMP levels and
inhibits the growth of many types of cells (237).
Often, this is due to a block of cell cycle progression
in G1 (237).
3. The level of cAMP normally fluctuates throughout
the cell cycle. In many types of cells, cAMP content
is highest in G1 then decreases as cells enter S-phase
and this change is essential for the cells to exit G1
and undergo DNA synthesis (237).
4. In high density, nonproliferating cells, however,
cAMP content is more uniform throughout the cell
population (238) which suggests that GJIC can
dampen cAMP oscillations.
5. Cyclic AMP impedes cell cycle progression in G1 by
at least two mechanisms. First, it decreases the
expression of cyclin D1 (239). This cyclin
associates with cyclin-dependent kinase (cdk) 4 or 6,
depending upon the cell type, and is active in G1.
The active complex phosphorylates the
retinoblastoma protein which then releases E2F
transcription factors and these activate the
expression of genes needed in S-phase (240). In this
way, cyclin D1 is essential for cell cycle progression
through G1 into S-phase. Secondly, cAMP increases
the expression of the cyclin-dependent kinase
inhibitor, p27/kip-1 (241), which blocks the activity of
cyclin D1/cdk 4/6 and other cylin/cdk complexes
(240). Neoplastic cells generally have reduced
contents of cAMP and p27/kip-1 and higher
cyclin/cdk activity (237,240).
Thus, cAMP fulfills the five criteria listed
above for a gap junction, growth regulatory signal
molecule. Cyclic AMP levels also fluctuate at other
points in the cell cycle besides G1 in a cell-specific
manner and other second messengers similarly oscillate
throughout the cell cycle (237). This indicates that the
growth regulatory role of cAMP and other potential gap
junction signals is undoubtedly very complex. Here we
will only further consider the role of cAMP.
Figure 6 illustrates how an oscillating, growth
inhibitor such as cAMP might inhibit cell growth via
GJIC; this model is based upon previous ones (170,216).
First, to inhibit growth, the signal must be sustained
above a threshold; periodic oscillations above the
threshold would not suppress growth. In cells at low
density where cell-cell contact and GJIC would be
minimal, the cells would generate these oscillatory,
inhibitory signals in an asynchronous manner (panel A).
The graph depicts such fluctuation within a single cell
and illustrates that the signal would not be sustained
above a theoretical growth inhibitory threshold. Growth
would result in this cell. Similarly, if cell-cell contacts
are extensive but GJIC is low or does not exist (as in
neoplastic cells or cells exposed to growth factors or
tumor promoters), the signal would still oscillate
asynchronously in individual cells and growth would
occur (panel B). In contrast, if GJIC is present, the signal
oscillations would be dampened by cell-cell diffusion and
a steady-state signal level would arise in the cell
population (panel C). If this steady-state level is above
the inhibitory threshold, growth would be suppressed.
Below the threshold, growth would ensue. This situation
is analogous to nonneoplastic cells that have extensive
GJIC and stop growth when reaching high density
(contact inhibition of growth).
Several facets of this model need additional
consideration. First, it permits growth in cells that are
coupled well by gap junctions, i.e., nonneoplastic cells.
Growth in such cells could be achieved by raising the
growth inhibitory threshold, by decreasing the steady-
state level of the inhibitor, or by reducing GJIC
(uncoupling the cells). In the case of cAMP, an increase
in the threshold would occur if cyclin D1/cdk 4/6 activity
was activated by another factor. More cAMP might be
required to inhibit the complex. A decrease in the
steady-state level of cAMP could be achieved by
reducing adenylate cyclase activity or increasing
phosphodiesterase activity. Finally, gap junction
uncoupling of a cell in G1 (described above)
Gap junctions and cancer
225
Figure 6. Illustration of how an oscillating, growth
inhibitory signal such as cAMP could suppress cell growth
only when GJIC was present. (A) in the absence of
extensive cell-cell contact (and GJIC), signal levels would
oscillate asynchronously in individual cells. The
concentration of the signal in an individual cell over time is
plotted on the graph. The signal would rise and fall above
a growth inhibitory threshold, but because of the lack of
sustained signal above the threshold, cell replication would
continue. This situation is analogous to low density,
proliferating cell cultures. (B) Similar to the case with
dispersed cells, the presence of extensive cell-cell
contact, but no GJIC would also result in cells with
oscillating signal levels and continuous growth. This
situation is similar to neoplastic cells that continue to
proliferate at high cell density, i.e., they are not
contact-inhibited. (C) The signal level in cells that
are extensively coupled by gap junctions, however,
would reach a steady-state because any oscillation in
an individual cell would be dampened (buffered) by
the neighboring cells. This steady-state signal level,
if above the growth inhibitory threshold, would
suppress growth.
would enable the cAMP level to fall because it could
not be buffered by neighboring cells. Second, the
model also accounts for the dysregulated growth of
neoplastic cells. In general, these cells have deficient
GJIC, lower cAMP content, and higher cyclin/cdk
activities (237,240). This model is consistent with the
"Neighborhood Coherence Principle" model of tissue
homeostasis mediated by gap junctions (242).
21. MODULATION OF GJIC FOR CANCER
THERAPY
Several factors can influence the response of
a tumor to chemotherapy (243). One is the
vascularization of the tumor. Many tumors have a
disorganized, inadequate vascular supply. Many of the
cells within such tumors are deficient in oxygen and
nutrients. Unfortunately, hypoxic cells may be
resistant to certain chemotherapeutic drugs such as
adriamycin that generate oxygen free radicals and
other reactive oxygen species as a mechanism of cell
killing. Poor vascularization may also result in
limited penetration of drugs away from tumor blood
vessels. This is especially true for water-soluble
agents such as methotrexate, adriamycin, and
vinblastine. Nutrient-deprived cells in poorly
vascularized tumors may also have poor uptake of
drugs through active transport and have altered drug
metabolism.
One possible mechanism to increase drug
penetration and dispersal in tumors would be to
increase GJIC. Oxygen and nutrients would
theoretically pass through gap junctions from cells
adjacent to blood vessels to those further away. This
increase in GJIC might be achieved by increasing tumor
cell connexin expression pharmacologically (e.g., with
forskolin, steroids, and retinoids) or by introducing
active connexin genes (gene therapy approach). As
demonstrated by the numerous in vitro studies
described above, the enhancement of tumor cell GJIC
might have the additional benefit of reducing tumor cell
growth, besides enhancing chemothrerapy.
GJIC may also improve a more novel type of
cancer therapy that involves introducing a lethal gene
such as the Herpes Simplex virus thymidine kinase
(HSV-TK) gene into tumor cells (244,245).
Expression of this gene renders the tumor cells
susceptible to the thymidine analogue, ganciclovir
(GCV) which is phosphorylated readily by HSV-TK,
but not by endogenous TK. The phosphorylated
metabolite is incorporated into the DNA of
proliferating cells such as cancer cells resulting in cell
death. Cells that express the gene are susceptible to
GCV, but nonexpressing and nonproliferating cells
are very resistant. Introduction of the gene into tumor
cells is usually accomplished by infection with a virus
that has been genetically engineered to express the
gene. Interestingly, however, only a small percentage
of the tumor cells take up and express the HSV-TK
gene yet a much higher percentage of the cells are
killed following GCV treatment. This is known as the
“bystander effect” (244,245).
One mechanism for the bystander effect is that
phosphorylated GCV can pass between tumor cells through
the few gap junction channels that often exist (246-248). It
follows then that the enhancement of GJIC in such cells
might improve the bystander effect. This has recently been
demonstrated using connexin-transfected neoplastic cells
(249,250). These results suggest that the enhancement of
GJIC might improve the bystander effect in certain types of
cancer gene therapies as well as in more conventional
chemotherapies.
Gap junctions and cancer
226
22. ACKNOWLEDGEMENTS
The authors gratefully acknowledge the excellent
secretarial assistance of Ms. Kathy Deanda and Mrs.
Robbyn Davenport. The authors were supported by grants
from the NIH (CA57612, CA21104) and the U.S. Air Force
Office of Scientific Research (F49620-97-1-0022) and the
NIEHS (P42 ES04911-09).
23. REFERENCES
1. Borek C & L Sachs: The difference in contact
inhibition of cell replication between normal cells and cells
transformed by different carcinogens. Proc. Natl. Acad.
Sci. USA 56, 1705-1711 (1966)
2. Abercrombie M: Contact inhibition and malignancy.
Nature 281, 259-262 (1979)
3. Markert C: Neoplasia: a disease of cell
differentiation. Cancer Res 28, 1908-1914 (1968)
4. Pierce GP: Neoplasms, differentiation and mutations.
Am J Pathol 77, 103-1118 (1974)
5. Potter VR: Phenotypic diversity in experimental
hepatomas: the concept of partially blocked ontogeny. Br J
Cancer 38, 1-23 (1978)
6. Fialkow PJ: Clonal origin of human tumans. Ann Rev
Med 30, 135-176 (1979)
7. Greaves MF: Differentiation-linked leukemogenesis
in lymphocytes. Science 234, 697-704 (1986)
8. Nowell PC: The clonal evolution of tumor cell
populations. Science 194, 23-28 (1976)
9. Till JE: Stem cells in differentiation and neoplasia. J
Cell Physiol 1, 3-11 (1982)
10. Kondo S: Carcinogenesis in relation to the stem cell
mutation hypothesis. Differentiation 24, 1-8 (1983)
11. Buick RN & MN Pollak: Perspectives on clonogenic
tumor cells, stem cells and oncogenes. Cancer Res 44,
4909-4918 (1984)
12. Trosko JE, CC Chang & BV Madhukar: Cell-cell
communication: Relationship of stem cells to the
carcinogenic process. In: Mouse Liver Carcinogenesis:
Mechanisms and Species Comparisons. Eds: DE
Stevenson, JA Popp, JM Ward, RM McClain, TJ Slaga, HC
Pipot, Alan R. Liss, NY, 259-276 (1990)
13. Kerbel RS, P Frost, R Liteplo, DA Carlow & BE
Elliot: Possible epigenetic mechanisms of tumor
progression: Induction high frequency heritable but
phenotypically unstable changes in the tumorigenic and
metastatic properties of tumor cell populations by 5-
azacytidine treatment. J Cell Physiol 3, 87-97 (1984)
14. Cairns J: Mutation selection and the natural history
of cancer. Nature 255, 197-200 (1975)
15. Hill CS & R Treisman: Transcriptional regulation by
extracellular signals: Mechanisms and specificity. Cell 80,
199-211 (1995)
16. O’Brian CA, NE Ward & CG Ioannides: Altered
signal transduction in carcinogenesis. In: Advances in
Molecular and Cell Biology, Vol.7, 61-88 (1993)
17. Kerr JFR, AH Willie & AR Currie: Apoptosis: a
basic biological phenomenon with wide-ranging
implications in tissue kinetics. Br J Cancer 68, 239-257
(1972)
18. Thompson CB: Apoptosis in the pathogenesis and
treatment of disease. Science 267, 1456-1462 (1995)
19. Trosko JE & JA Goodman: Intercellular
communication may facilitate apoptosis: implications for
tumor promotion. Molec Carcinogen 11, 8-11 (1994)
20. Trosko JE, BV Madhukar & Chang CC: Endogenous
and exogenous modulation of gap junctional intercellular
communication: Toxicological and pharmacological
implications. Life Sci 53, 1-19 (1993)
21. Revel JP: The oldest multicellular animal and its
junction. In: Gap Junctions. Eds: Hertzberg EL, Johnston
RG, Alan R. Liss, NY, 135-149 (1988)
22. Loewenstein WR: Permeability of membrane junctions.
Ann NY Acad Sci 137, 441-472 (1966)
23. Kanno Y, Modulation of cell communication and
carcinogenesis. Jpn J Physiol 35, 693-707 (1985)
24. Yamasaki H, M Mesnil, Y Omori, N Mironov & V
Krutovskikh: Intercellular communication and carcinogenesis.
Mutat Res 333, 181-188 (1995)
25. Ruch RJ: The role of gap junctional intercellular
communication in neoplasia. Ann Clin Lab Sci 24, 216-231
(1994)
26. Trosko JE, CC Chang, BV Madhukar & E Dupont:
Intercellular communication: A paradigm for the interpretation
of the initiation/promotion/ progression model of
carcinogenesis. In: Chemical Induction of Cancer: Modulation
and Combination Effects. Ed: Arcos JC, Birkhauser, Boston,
205-225 (1996)
27. Holder JW, E Elmore & JC Barrett: Gap junction
function and cancer. Cancer Res 53, 3475-3485 (1993)
Gap junctions and cancer
227
28. Brody H: A systems view of man: Implications for
medicine, science and ethics. Perspect Biol Med 17, 71-92
(1973)
29. Potter VR: Probabilistic aspects of the human
cybernetic machine. Perspect Biol Med 17, 164-183
(1974)
30. Iversen OH: Cybernetic aspects of the cancer
problem. In: Progress in Cybernetics. Eds: Weiner N,
Schade JP, Elsevier, Amsterdam, 76-110 (1965)
31. Potter VR: The present status of the blocked
ontogeny hypothesis of neoplasia: the thalassemia
connection. Oncodevelop Biol Med 2, 243-266 (1981)
32. Chang CC, JE Trosko, MH El-Fouly, Gibson-
D’Ambrosio & SM D’Ambrosio: Contact insensitivity of a
subpopulation of normal human fetal kidney epithelial cells
and of human carcinoma cell lines. Cancer Res 47, 1634-
1645 (1987)
33. Boveri T: Zur Frage Entshehung maligner tumoren.
Gustav Fischer, Jena (1914)
34. Weinberg RA: Tumor suppressor genes. Science
254, 1138-1146 (1991)
35. Hunter T: Cell growth control mechanisms. Nature
322, 14-15 (1986)
36. Potter VR: Biochemistry of cancer. In: Cancer
Medicine. Eds: Holland J, Frei E, Lea and Febiger
Publishers, Philadelphia, 178-192 (1973)
37. Sell S: Cellular origin of cancer: de-differentiation or
stem cell maturation arrest. Environ Health Perspect 101,
15-26 (1993)
38. Pitot HC, T Goldsworthy & S Moran: The natural
history of carcinogenesis: implications of experimental
carcinogenesis in the genesis of human cancer. J Supramol
Struct Cellul Biochem 17, 133-146 (1981)
39. Trosko JE and CC Chang, (1979) Genes, pollutants
and human diseases. Quart Rev Biophys 11, 603-627.
40. Trosko JE, CC Chang & A Medcalf: Mechanisms of
tumor potential role of intercellular communication. Cancer
Invest 1, 511-526 (1983)
41. Trosko JE, CC Chang, BV Madhukar & E Dupont:
Oncogenes, tumor suppressor genes and intercellular
communication in the ‘oncogeny as partially blocked
ontogeny’ hypothesis. In: New Frontiers in Cancer
Causation. Ed: Iversen OH, Taylor and Francis Publishers,
Wash. DC, 181-197 (1993)
42. Boutwell RK: The function and mechanism of promoters
of carcinogenesis. In: CRC Critical Reviews in Toxicology.
Chemical Rubber Co., Cleveland, pp. 419-443 (1974)
43. Yuspa SH and DL Morgan: Mouse skin cells resistant to
terminal differentiation associated with initiation of
carcinogenesis. Nature 293, 72-74 (1981)
44. Bursch W, B Lauer, I Timmermann-Trosiener, G
Barthel, J Schuppler & R Schulte-Hermann: Controlled death
and apoptosis of normal and preneoplastic cells in rat liver
following withdrawal of tumor promoters. Carcinogenesis 5,
453-458 (1984)
45. Nakano S, H Ueo, SA Bruce & POP Ts’o: A contact-
insensitive subpopulation in Syrian hamster cell cultures with a
greater susceptibility to chemically-induced neoplastic
transformation. Proc Natl Acad. Sci USA 82, 5005-5009
(1985)
46. Kao C-Y, K Nomata, CS Oakley, CW Welsch & CC
Chang: Two types of normal human breast epithelial cells
derived from reduction mammoplasty: phenotypic
characterization and response to SV40 transfection.
Carcinogenesis 16, 531-538 (1995)
47. Block A: Induced cell differentiation in cancer therapy.
Cancer Treat Rep 68, 199-206 (1984)
48. Shay J & WE Wright: Telomerase activity in human
cancer Curr Opin Oncol 8, 66-71 (1996)
49. Kang KS, JE Trosko & CC Chang: Telomerase activity
in a normal human breast epithelial cell type with stem cell
characteristics and its neoplastically transformed cells. Proc
Amer Assoc Cancer Res 38, 504 (1997)
50. Weinstein, IB, S Gattoni-Celli, P Kirschmeirer, M
Lambert, W Hsiao, J Backer & A Jeffrey: Multistage
carcinogenesis involves multiple genes and multiple
mechanisms. J Cellul Physiol 3, 127-137 (1984)
51. Yotti LP, CC Chang & JE Trosko: Elimination of
metabolic cooperation in Chinese hamster cells by a tumor
promoter. Science 206, 1089-1091 (1979)
52. Murray AW and DJ Fitzgerald: Tumor promoters inhibit
metabolic cooperation in cocultures of epidermal and 3T3 cells.
Biochem Biophys Res Commun 91, 395-401 (1979)
53. Ames BN, WE Durston, E Yamasaki & FD Lee:
Carcinogenesis as mutagens: a simple test system combining
liver homogenates for activation and bacteria for detection.
Proc Natl Acad Sci USA 70, 2281-2285 (1973)
54. Argyris TS: Regeneration and the mechanism of
epidermal tumor promotion. CRC Crit Rev Toxicol 14, 211-
258 (1985)
Gap junctions and cancer
228
55. Trosko JE and CC Chang: Nongenotoxic mechanisms in
carcinogenesis: Role of inhibited intercellular communication.
In: Banbury Report 31: Carcinogen Risk Assessment: New
Directions in the Qualitative and Quantitative Aspects. Cold
Spring Harbor Laboratory, Cold Spring Harbor, NY, 139-170
(1988)
56. Lau AF, WE Kurata, MY Kanemitsu, LW Loo, BJ
Warn-Cramer, W Ekhart & Lampe P: Regulation of connexin
43 function by activated tyrosine protein kinases. J. Bioenerg.
Biomembr., 28, 359-367 (1996)
57. Musil LS & DA Goodenough: Biochemical analysis of
connexin43 intracellular transport, phosphorylation and
assembly into gap junctional plaques. J Cell Biol 115, 1357-
1374 (1991)
58. Godwin AJ, LM Green, MP Walsh, JR McDonald, DA
Walsh & WF Fletcher: In situ regulation of cell-cell
communication by the cAMP-dependent protein kinase and
protein kinase C. Molec Cellul Biochem 127/128, 293-307
(1993)
59. Brissette JL, NM Kumar, NB Gilula & GP Dotto: The
tumor promoter 12-o-tetradecanoylphorbol-13-acetate and the
ras oncogene modulat expression and phosphorylation of gap
junction proteins. Molec Vellul Biol 11, 5364-5371 (1991)
60. Land H, LF Parada & RA Weinberg: Tumorigenic
conversion of primary embryo fibroblasts requires at least two
cooperating oncogenes. Nature 304, 596-602 (1983)
61. Hsiao WLW, S Gattoni-Celli & I Weinstein: Oncogene-
induced transformation of C3H 10T1/2 cells is enhanced by
tumor promoters. Science 226, 552-554 (1984)
62. Hayashi T, K Nomata, CC Chang, RJ Ruch & JE
Trosko: Cooperative effects of V-Myc and C-HA-Ras
oncogenes on gap junctional intercellular communication
and tumorigenicity in rat liver epithelial cells. Cancer Lett
(in press)
63. Standbridge EJ & J Wilkinson: Dissociation of
anchorage independence from tumorigenicity in human cell
hybrids. Int J Cancer 26, 1-8 (1980)
64. Standbridge EJ, RR Flandermeyer, DW Daniels &
WA Nelson-Rees: Specific chromosome loss associated
with the expression of tumorigenicity in human cell
hybrids. Somat Cell Mol Genet 7, 699-712 (1981)
65. Misra BC & ES Srivatsan: Localization of HeLa cell
tumor suppressor gene to the long arm of chromosome 11.
Am J Hum Genet 45, 565-577 (1989)
66. Loh WE, HJ Scrable, E Livanos, MJ Arboleda, WK
Cavenee, M Oshimura & BE Weissman: Human
chromosome 11 contains two different growth suppressor
genes for embryonal rhabdomyosarcoma. Proc Natl Acad
Sci USA 89, 1755-1759 (1992)
67. Bystrom C, C Larsson, C Blomberg, K Sandelin, U
Flakmer, B Skogseid, K Oberg, S Werner & M
Nordenskjold: Localization of the MEN1 gene to a small
region within chromosome 11q13 by deletion mapping in
tumors. Proc Natl Acad Sci USA 87, 1968-1972 (1990)
68. Radford DM, SW Ashley, SA Wells & DS Gerhard:
Loss of heterozygosity of markers on chromosome 11 in
tumors from patients with multiple endocrine neoplasia
syndrome type 1. Cancer Res 50, 6529-6533 (1990)
69. Larsson C, B Skogseid, K Oberg, Y Nakamura & M
Nordenskjold: Multiple endocrine neoplasia type 1 gene
maps to chromosome 11 and is lost in insulinoma. Nature
332, 85-87 (1988)
70. Gatti RA, I Berkel, E Boder, G Braedt, P Charmley,
P Concannon, F Ersoy, T Foroud, NG Jaspers & K Lange:
Localization of an ataxia telangiectasia gene to to
chromosome 11q22-23. Nature 336, 577-580 (1988)
71. Hampton GM, LA Penny, RN Baegen, A Larson, C
Brewer, S Lia, RM Busby-Earle, AW Williams, CM Steel
and CC Bird: Loss of heterozygosity in cervical carcinoma:
Subchromosomal localization of a putative tumor
suppressor gene at 11q22-q24. Proc Natl Acad Sci USA 91,
6953-6957 (1994)
72. Coleman WB, KD McCullough, GL Esch, CJ
Civalier, E Livanos, BE Weissman, JW Grisham & GJ
Smith: Suppression of the tumorigenic phenotype of a rat
liver epithelial cell line by the p11.2-p12 region of human
chrosome 11. Molec Carcinogen 13, 220-232 (1995)
73. de Feijter-Rupp HL, T Hayashi, GH Kalimi, P
Edwards, JL Redpath, CC Chang, EJ Standbridge & JE
Trosko: Restored gap junctional communication in non-
tumorigenic HeLa-Normal human fibroblast hybrids.
Carcinogenesis (in press)
74. Loewenstein WR: Junctional intercellular
communication: the cell-cell membrane channel. Physiol
Rev 61, 829-913 (1981)
75. Bennett MVL, LC Barrio, TA Bargiello, DC Spray,
E Hertzberg & JC Saez: Gap Junctions: new tools, new
answers, new questions. Neuron 6, 305-320 (1991)
76. Robards AW and WJ Lucas: Plasmodesmata. Ann
Rev Plant Physiol Plant Mol Biol 41, 369-419 (1990)
77. Beyer EC and TH Steinberg: Evidence that the gap
junction protein connexin-43 is the ATP-induced pore of
mouse macrophages. J Biol Chem 266, 7971-7974 (1991)
Gap junctions and cancer
229
78. White TW, R Bruzzone & DL Paul: The connexin
family of intercellular channel forming proteins. Kidney Int
48, 1148-1157 (1995)
79. Beyer EC, DL Paul & DA Goodenough:
Connexin43: a protein from rat heart homologous to a gap
junction protein from liver. J Cell Biol 105, 2621-2629
(1987)
80. Paul DL: Molecular cloning of cDNA for rat liver
gap junction protein. J Cell Biol 103, 123-134 (1986)
81. Kumar NM and NB Gilula: Cloning and
characterization of human and rat liver cDNAS coding for
a gap junction protein. J Cell Biol 103, 767-776 (1986)
82. Zhang J-T & BJ Nicholson: Sequence and tissue
distribution of a second protein of hepatic gap junctions,
Cx26, as deduced from its cDNA. J Cell Biol 109, 3391-
3401 (1989)
83. Stutenkemper R, S Geisse, HJ Schwarz, J Look, O
Traub, BJ Nicholson & K Willecke: The hepatocyte-
specific phenotype of murine liver cells correlates with
high expression of connexin32 and connexin26 but very
low expression of connexin43. Exp Cell Res 201, 43-54
(1992)
84. Spray DC, M Chanson, AL Moreno, R Dermietzel &
P Meda: Distinctive gap junction channel types connect
WB cells, a clonal cell line derived from rat liver. Am J
Physiol 260, C513-C527 (1991)
85. Rojkind M, P Novikoff, P Greenwel, J Rubin, L
Rojas-Valencia, A Carvalho, R Stockert, D Spray, E
Hertzberg & A Wolkoff: Characterization and
functional studies on rat liver fat-storing cell line and
freshly isolated hepatocyte coculture system. Am J Path
146(6), 1508-1520 (1995)
86. Willecke K, R Heynkes, E Dahl, R Stutenkemper,
H Hennemann, S Jungbluth, T Suchyna & BJ
Nicholson: Mouse connexin37: cloning and functional
expression of a gap junction gene highly expressed in
lung. J Cell Biol 114, 1049-1057 (1991)
87. Hennemann H, T Suchyna, H Lichtenberg-Frat, S
Jungbluth, E Dahl, J Schwarz, BJ Nicholson & K
Willecke: Molecular cloning and functional expression
of mouse connexin40, a second gap junction gene
preferentially expressed in lung. J Cell Biol 117, 1299-
1310 (1992)
88. Cesen-Cummings K, MJ Fernstrom, AM
Malkinson & RJ Ruch: Frequent reduction of gap
junctional intercellular communication and connexin43
expression in human and mouse lung carcinoma cells.
Carcinogenesis (in press)
89. Lee YC, CE Yellowley, ZY Li, HJ Donahue & DE
Rannels: Expression of functional gap junctions in cultured
pulmonary alveolar epithelial cells. Am J Physiol Lung Cell
Mol Physiol 272, L1105-L1114 (1997)
90. Risek B, FG Klier & NB Gilula: Multiple gap
junction genes are utilized during rat skin and hair
development. Development 116, 639-651 (1992)
91. Butterweck A, C Elfgang, K Willecke & O Traub:
Differential expression of the gap junction proteins
connexin45, -43, -40, -31 & -26 in mouse skin. Eur J Cell
Biol 65, 152-163 (1994)
92. Willecke K, S Jungbluth, E Dahl, H Hennemann, R
Heynkes & K-H Grzeschik: Six genes of the human
connexin gene family coding for gap junctional proteins are
assigned to four different human chromosomes. Eur J Cell
Biol 53, 275-280 (1990)
93. Nicholson B, R Dermietzel, D Teplow, O Traub, K
Willecke & J-P Revel: Two homologous protein
components of hepatic gap junctions. Nature 329, 732-734
(1987)
94. Sosinsky G: Mixing of connexins in gap junction
membrane channels. Proc Natl Acad Sci USA 92, 9210-
9214 (1995)
95. Stauffer KA: The gap junction proteins α1-connexin
(connexin-32) and α2-connexin (connexin-26) can form
heteromeric hemichannels. J Biol Chem 270, 6768-6772
(1995)
96. Jiang JX and DA Goodenough: Heteromeric
connexons in lens gap junction channels. Proc Natl Acad
Sci USA 93, 1287-1291 (1996)
97. Rubin JB, VK Verselis, MVL Bennett & TA
Bargiello: Molecular analysis of voltage dependence of
heterotypic gap junctions formed by connexins 26 and 32.
Biophys J 62, 183-195 (1992)
98. Bukauskas FF, C Elfgang, K Willecke & R
Weingart: Heterotypic gap junction channels (connexin26 -
connexin32) violate the paradigm of unitary conductance.
Pflugers Arch 429, 870-872 (1995)
99. Bruzzone R, TW White & DL Paul: Connections
with connexins: the molecular basis of direct intercellular
signaling. Eur J Biochem 238, 1-27 (1996)
100. Miller T, G Dahl & R Werner: Structure of a gap
junction gene: rat connexin-32. Biosci Rep 8, 455-464
(1988)
101. Neuhaus IM, G Dahl & R Werner: Use of alternate
promoters for tissue-specific expression of the gene coding
for connexin32. Gene 158, 257-262 (1995)
Gap junctions and cancer
230
102. Sohl G, C Gillen, F Bosse, M Gleichmann, HW
Muller & K Willecke: A second alternative transcript of the
gap junction gene connexin32 is expressed in murine
Schwann cells and modulated in injured sciatic nerve. Eur
J Cell Biol 69, 267-275 (1996)
103. Hennemann H, G Kozjek, E Dahl, B Nicholson & K
Willecke: Molecular cloning of mouse connexins26 and -
32: Similar genomic organization but distinct promoter
sequences of two gap junction genes. Eur J Cell Biol 58,
81-89 (1992)
104. Sullivan R, C Ruangvoravat, D Joo, J Morgan, BL
Wang, XK Wang & CW Lo: Structure, sequence and
expression of the mouse Cx43 gene encoding connexin 43.
Gene 130, 191-199 (1993)
105. Chen Z-Q, D Lefebvre, X-H Bai, A Reaume, J
Rossant & SJ Lye: Identification of two regulatory
elements within the promoter region of the mouse connexin
43 gene. J Biol Chem 270, 3863-3868 (1995)
106. Kiang DT, N Jin, ZJ Tu & HH Lin: Upstream
genomic sequence of the human connexin26 gene. Gene
199, 165-171 (1997)
107. Bai S, DC Spray & RD Burk: Identification of
proximal and distal regulatory elements of the rat
connexin32 gene. Biochim. Biophys. Acta Gene Struct
Expression 1216, 197-204 (1993)
108. Bai S, A Schoenfeld, A Pietrangelo & RD Burk: Basal
promoter of the rat connexin 32 gene: Identification and
characterization of an essential element and its DNA- binding
protein. Mol Cell Biol 15, 1439-1445 (1995)
109. De Leon JR, PM Buttrick & GI Fishman: Functional
analysis of the connexin43 gene promoter in vivo and in vitro. J
Mol Cell Cardiol 26, 379-389 (1994)
110. Geimonen E, W Jiang, M Ali, GI Fishman, Garfield RE
& Andersen J: Activation of protein kinase C in human uterine
smooth muscle induces connexin-43 gene transcription through
an AP-1 site in the promoter sequence. J Biol Chem 271,
23667-23674 (1996)
111. Yu W, G Dahl & R Werner: The connexin43 gene is
responsive to oestrogen. Proc R Soc Lond [Biol.] 255, 125-132
(1994)
112. Matesic DF, HL Rupp, WJ Bonney, RJ Ruch & JE
Trosko: Changes in gap-junction permeability, phosphorylation
& number mediated by phorbol ester and non-phorbol-ester
tumor promoters in rat liver epithelial cells. Mol Carcinog 10,
226-236 (1994)
113. Ren P, PP Mehta & RJ Ruch: Inhibition of gap
junctional intercellular communication by tumor promoters
in connexin43 and connexin32-expressing liver cells: Cell
specificity and role of protein kinase C. Carcinogenesis (in
press)
114. Kwiatkowski AP, TK Baker & JE Klaunig:
Comparison of glucocorticoid-mediated changes in the
expression and function of rat hepatocyte gap junctional
proteins. Carcinogenesis 15, 1753-1758 (1994)
115. Ren P, AW de Feijter, DL Paul & RJ Ruch:
Enhancement of liver cell gap junction protein expression
by glucocorticoids. Carcinogenesis 15, 1807-1814 (1994)
116. Ren P & RJ Ruch: Inhibition of gap junctional
intercellular communication by barbiturates in long-term
primary cultured rat hepatocytes is correlated with liver
tumour promoting activity. Carcinogenesis 17, 2119-2124
(1996)
117. Saez JC, DC Spray, AC Nairn, E Hertzberg, P
Greengard & MVL Bennett: cAMP increases junctional
conductance and stimulates phosphorylation of the 27-kDa
principal gap junction polypeptide. Proc Natl Acad Sci
USA 83, 2473-2477 (1986)
118. Traub O, J Look, D Paul & K Willecke: Cyclic
adenosine monophosphate stimulates biosynthesis and
phosphorylation of the 26 KDA junction protein in
cultured mouse hepatocytes. Eur J Cell Biol 43, 48-54
(1987)
119. Mehta PP, M Yamamoto & B Rose: Transcription of
the gene for the gap junctional protein connexin43 and
expression of functional cell-to-cell channels are regulated
by cAMP. Mol Biol Cell 3, 839-850 (1992)
120. Burghardt RC, R Barhoumi, TC Sewall & JA
Bowen: Cyclic AMP induces rapid increases in gap
junction permeability and changes in the cellular
distribution of Connexin43. J Membr Biol 148, 243-253
(1995)
121. Matesic DF, T Hayashi, JE Trosko & JA Germak:
Upregulation of gap junctional intercellular communication
in immortalized gonadotropin-releasing hormone neurons
by stimulation of the cyclic AMP pathway.
Neuroendocrinology 64, 286-297 (1996)
122. Saez JC, WA Gregory, T Watanabe, R Dermietzel, EL
Hertzberg, LM Reid, MVL Bennett & DC Spray: cAMP delays
disappearance of gap junctions between pairs of rat hepatocytes
in primary culture. Am J Physiol 257, C1-C11 (1989)
123. Atkinson M, P Lampe, H Lin, R Kollander, X-R Li & D
Kiang: Cyclic AMP modifies the cellular distribution of
connexin43 and induces a persistant increase in the junctional
permeability of mouse mammary tumor cells. J Cell Sci 108,
3079-3090 (1995)
Gap junctions and cancer
231
124. Musil LS & DA Goodenough: Multisubunit assembly of
an integral plasma membrane channel protein, gap junction
connexin43, occurs after exit from the ER. Cell 74, 1065-1077
(1993)
125. Jongen WMF, DJ Fitzgerald, M Asamoto, C Piccoli, TJ
Slaga, D Gros, M Takeichi & H Yamasaki: Regulation of
connexin43-mediated gap junctional intercellular
communication by Ca+2 in mouse epidermal cells is controlled
by E-cadherin. J Cell Biol 114, 545-555 (1991)
126. Frenzel EM & RG Johnson: Gap junction formation
between cultured embryonic lens cells is inhibited by antibody
to N-cadherin. Dev Biol 179, 1-16 (1996)
127. Wang Y & PP Mehta: Facilitation of gap-junctional
communication and gap-junction formation in mammalian cells
by inhibition of glycosylation. Eur J Cell Biol 67, 285-296
(1995)
128. Wang YJ, PP Mehta & B Rose: Inhibition of
glycosylation induces formation of open connexin-43 cell-to-
cell channels and phosphorylation and Triton X-100
insolubility of connexin-43. J Biol Chem 270, 26581-26585
(1995)
129. Fallon RF & DA Goodenough: Five-hour half-life of
mouse liver gap-junction protein. J Cell Biol 90, 521-526
(1981)
130. Yancey SB, BJ Nicholson & J-P Revel: The dynamic
state of liver gap junctions. J Supramol Struct Cell
Biochem 16, 221-232 (1981)
131. Traub O, J Look, R Dermietzel, F Brummer, D
Hulser & K Willecke: Comparative characterization of the
21-kDa and 26-kDa gap junction proteins in murine liver
and cultured hepatocytes. J Cell Biol 108, 1039-1051
(1989)
132. Laird DW, KL Puranam & J-P Revel: Turnover and
phosphorylation dynamics of connexin43 gap junction
protein in cultured cardiac myocytes. Biochem J 273, 67-72
(1991)
133. Laird DW, M Castillo & L Kasprzak: Gap junction
turnover, intracellular trafficking & phosphorylation of
connexin43 in brefeldin A-treated rat mammary tumor
cells. J Cell Biol 131, 1193-1203 (1995)
134. Larsen W & N Hai: Origin and fate of cytoplasmic
gap junctional vesicles in rabbit granulosa cells. Tissue Cell
10(3), 585-598 (1978)
135. Guan XJ & RJ Ruch: Gap junction endocytosis and
lysosomal degradation of connexin43-P2 in WB-F344 rat
liver epithelial cells treated with DDT and lindane.
Carcinogenesis 17, 1791-1798 (1996)
136. Laing JG & EC Beyer: The gap junction protein
connexin43 is degraded via the ubiquitin proteasome
pathway. J Biol Chem 270, 26399-26403 (1995)
137. Laing JG, PN Tadros, EM Westphale & EC Beyer:
Degradation of connexin43 gap junctions involves both the
proteosome and the lysosome. Exp Cell Res 236, 482-492
(1997)
138. Elvira M, JA Diez, KKW Wang & A Villalobo:
Phosphorylation of connexin-32 by protein kinase C
prevents its proteolysis by æ-calpain and m-calpain. J Biol
Chem 268, 14294-14300 (1993)
139. Elvira M, KKW Wang & A Villalobo:
Phosphorylated and non-phosphorylated connexin-32
molecules in gap junction plaques are protected against
calpain proteolysis after phosphorylation by protein kinase
C. Biochem Soc Trans 22, 793-796 (1994)
140. Lin JS, S Fitzgerald, YM Dong, C Knight, P
Donaldson & J Kistler: Processing of the gap junction
protein connexin50 in the ocular lens is accomplished by
calpain. Eur J Cell Biol 73, 141-149 (1997)
141. Yancey S, D Easter & J-P Revel: Cytological
changes in gap junctions during liver regeneration. J
Ultrastr Reas 67, 229-242 (1979)
142. Guan XJ, S Wilson, KK Schlender & RJ Ruch: Gap-
junction disassembly and connexin 43 dephosphorylation
induced by 18(-glycyrrhetinic acid. Mol Carcinog 16, 157-
164 (1996)
143. Goldberg GS, AP Moreno, JF Bechberger, SS Hearn,
RR Shivers, DJ MacPhee, YC Zhang & CCG Naus:
Evidence that disruption of connexon particle arrangements
in gap junction plaques is associated with inhibition of gap
junctional communication by a glycyrrhetinic acid
derivative. Exp Cell Res 222, 48-53 (1996)
144. Brink PR: Gap junction channel gating and
permselectivity: Their roles in co-ordinated tissue function.
Clin Exp Pharmacol Physiol 23, 1041-1046 (1996)
145. Saez JC, VM Berthoud, AP Moreno & DC Spray:
Gap junctions: Multiplicity of controls in differentiated and
undifferentiated cells and possible functional implications.
Adv Second Messenger Phosphoprotein Res 27, 163-198
(1993)
146. Berthoud VM, MB Rook, O Truab, EL Hertzberg &
JC Saez: On the mechanisms of cell uncoupling induced by
a tumor promoter phorbol ester in clone 9 cells, a rat liver
epithelial cell line. Eur J Cell Biol 62, 384-396 (1993)
147. Saez JC, AC Nairn, AJ Czernik, DC Spray, EL
Hertzberg, P Greengard & MVL Bennett: Phosphorylation of
Gap junctions and cancer
232
connexin 32, a hepatocyte gap-junction protein by
cAMP-dependent protein kinase, protein kinase C and
Ca+2/calmodulin-dependent protein kinase II. Eur J
Biochem 192, 263-273 (1990)
148. Crow DS, WE Kurata & AF Lau: Phosphorylation
of connexin43 in cells containing mutant src
oncogenes. Oncogene 7, 999-1003 (1992)
149. Ek JF, M Delmar, R Perzova & SM Taffet: Role
of histidine 95 on pH gating of the cardiac gap junction
protein connexin43. Circ Res 74, 1058-1064 (1994)
150. Ek-Vitorin JF, G Calero, GE Morley, W Coombs,
SM Taffet & M Delmar: pH regulation of connexin43:
Molecular analysis of the gating particle. Biophys J 71,
1273-1284 (1996)
151. Wang XG, LQ Li, LL Peracchia & C Peracchia:
Chimeric evidence for a role of the connexin
cytoplasmic loop in gap junctionchannel gating.
Pflugers Arch 431, 844-852 (1996)
152. Spray DC, AP Moreno, B Eghbali, M Chanson &
GI Fishman: Gating of gap junction channels as
revealed in cells stably transfected with wild type and
mutant connexin cDNAs. Biophys J 62, 48-50 (1992)
153. Lazrak A & P Camillo: Gap junction gating
sensitivity to physiological internal calcium regardless
of pH in Novikoff hepatoma cells. Biophys J 65, 2002-
2012 (1993)
154. Bruzzone R, TW White & DL Paul: Expression of
chimeric connexins reveals new properties of the
formation and gating behavior of gap junction channels.
J Cell Sci 107, 955-967 (1994)
155. Hertzberg EL and NB Gilula: Liver gap junctions
and lens fiber junctions: comparative analysis and
calmodulin interaction. Cold Spring Harbor Symp
Quant Biol 46, 639-645 (1982)
156. Arellano RO, F Ramon, A Rivera & GA
Zampighi: Calmodulin acts as an intermediary for the
effects of calcium on gap junctions from crayfish lateral
axons. J Memb Biol 101, 119-131 (1988)
157. Toyama J, H Sugiura, K Kamiya, I Kodama, M
Terasawa & H Hidaka: Ca2+-calmodulin mediated
modulation of the electrical coupling of ventricular
myocytes isolated from guinea pig heart. J Mol Cell
Cardiol 26, 1007-1015 (1994)
158. Peracchia C, XG Wang, LQ Li & LL Peracchia:
Inhibition of calmodulin expression prevents low-pH-
induced gap junction uncoupling in Xenopus oocytes.
Pflugers Arch 431, 379-387 (1996)
159. Torok K, K Stauffer & WH Evans: Connexin 32 of gap
junctions contains two cytoplasmic calmodulin-binding
domains. Biochem J 326, 479-483 (1997)
160. Saez JC, MVL Bennett & DC Spray: Carbon
tetrachloride at hepatotoxic levels blocks reversibly gap
junctions between rat hepatocytes. Science 236, 967-969
(1987)
161. Ruch RJ and JE Klaunig: Inhibition of mouse hepatocyte
intercellular communication by paraquat-generated oxygen free
radicals. Toxicol Appl Pharmacol 94, 427-436 (1988)
162. Peters NS, CR Green, PA Poole-Wilson & NJ Severs:
Reduced content of connexin43 gap junctions in ventricular
myocardium from hypertrophied and ischemic human hearts.
Circulation 88, 864-875 (1993)
163. Owens LM, TA Fralix, E Murphy, WE Cascio & LS
Gettes: Correlation of ischemia-induced extracellular and
intracellular ion changes to cell-to-cell electrical uncoupling in
isolated blood-perfused rabbit hearts. Circulation 94, 10-13
(1996)
164. Miyashita T, A Takeda, M Iwai & T Shimazu: Single
administration of hepatoxic chemicals transiently decreases the
gap junction protein levels of connexin 32 in rat liver. Eur J
Biochem 196, 37-42 (1991)
165. Lawrence TS, WH Beers & NB Gilula: Transmission of
hormonal stimulation by cell-cell communication. Nature 272,
501-506 (1978)
166. Murray SA & WH Fletcher: Hormone-induced
intercellular signal transfer dissociates cyclic amp-
dependent protein kinase. J Cell Biol 98, 1710-1719
(1984)
167. Kalimi GH and CW Lo: Communication
compartments in the gastrulating mouse embryo. J Cell
Biol 107, 241-255 (1988)
168. Loch-Caruso R and JE Trosko: Inhibited intercellular
communication as a mechanistic link between teratogenesis
and carcinogenesis. CRC Crit Rev Toxicol 16, 157-183
(1986)
169. Loewenstein WR & Y Kanno: Intercellular
communication and the control of growth. Lack of
communication between cancer cells. Nature 209, 1248-
1249 (1966)
170. Loewenstein WR: Junctional intercellular
communication and the control of growth. Biochim
Biophys Acta 560, 1-65 (1979)
171. Loewenstein WR: Cell-cell communication and the
control of growth. Am Rev Respir Dis 142, S48-S53 (1990)
Gap junctions and cancer
233
172. Weinstein RS, FB Merck & J Alroy: The structure
and function of intercellular junctions in cancer. Adv
Cancer Res 23, 23-89 (1976)
173. Yamasaki H: Gap junctional intercellular
communication and carcinogenesis. Carcinogenesis 11,
1051-1058 (1990)
174. Garber SA, MJ Fernstrom, GD Stoner & RJ Ruch:
Altered gap junctional intercellular communication in
neoplastic rat esophageal epithelial cells. Carcinogenesis
18, 1149-1153 (1997)
175. Mehta PP & WR Loewenstein: Differential
regulation of communication by retinoic acid in
homologous and heterologous junctions between normal
and transformed cells. J Cell Biol 113, 371-379 (1991)
176. Kalimi GH, LL Hampton, JE Trosko, SS
Thorgeirsson & AC Huggett: Homologous and
heterologous gap-junctioal intercellular communication in
v-raf-, v-myc- & v-raf/v-myc-transduced rat liver epithelial
cell lines. Mol Carcinogenesis 5, 301-310 (1992)
177. Mikalsen S-O: Heterologous gap junctional
intercellular communication in normal and morphologically
transformed colonies of Syrian hamster embryo cells.
Carcinogenesis 1485, 2085-2090 (1993)
178. Krutovskikh V, G Mazzoleni, N Mironov, Y Omori,
A-M Aguelon, M Mesnil, F Berger, C Partensky & H
Yamasaki: Altered homologous and heterologous gap-
junctional intercellular communication in primary human
liver tumors associated with aberrant protein localization
but not gene mutation of connexin 32. Int J Cancer 56, 87-
94 (1994)
179. Goldsworthy TL, R Fransson-Steen & RR Maronpot:
Importance of and approaches to quantification of
hepatocyte apoptosis. Toxicol Pathol, 24, 24-35 (1996)
180. Trosko JE: Mechanisms of tumor promotion:
possible role of inhibited intercellular communication. Eur
J Cancer Clin Oncol 23, 19-29 (1987)
181. Klaunig JE and RJ Ruch: Role of intercellular
communication in nongenotoxic carcinogenesis. Lab Invest
62, 135-146 (1990)
182. Budunova IV and GM Williams: Cell culture assays
for chemicals with tumor promoting or inhibiting activity
based on the modulation of intercellular communication.
Cell Biol Toxicol 10, 71-116 (1994)
183. Klaunig JE, RJ Ruch, CM Weghorst & JA Hampton:
Role of inhibition of intercellular communication in hepatic
tumor promotion. In Vitro Toxicol 3, 91-107 (1990)
184. Krutovskikh VA, M Mesnil, G Mazzoleni & H
Yamasaki: Inhibition of rat liver gap junction intercellular
communication by tumor-promoting agents in vivo:
Association with aberrant localization of connexin proteins.
Lab Invest 72, 571-577 (1995)
185. Klann RC, DJ Fitzgerald, C Piccoli, TJ Slaga & H
Yamasaki: Gap-junctional intercellular communication in
epidermal cell lines from selected stages of sencar mouse skin
carcinogenesis. Cancer Res 49, 699-705 (1989)
186. Budunova IV, S Carbajal & TJ Slaga: Effect of diverse
tumor promoters on the expression of gap-junctional proteins
connexin (Cx) 26, Cx31.1 & Cx43 in SENCAR mouse
epidermis. Mol Carcinog 15, 202-214 (1996)
187. Ruch RJ & JE Klaunig: The effects of tumor promoters,
genotoxic carcinogens & hepatocytotoxins on mouse
hepatocyte intercellular communication. Cell Biol Toxicol 2,
469-483 (1986)
188. Na MR, SK Koo, DY Kim, SD Park, SK Rhee, KW
Kang & CO Joe: In vitro inhibition of gap junctional
intercellular communication by chemical carcinogens. Toxicol
98, 199-206 (1995)
189. Trosko JE, CC Chang, BV Madhukar & JE Klaunig:
Chemical, oncogene and growth factor inhibition of gap
junctional intercellular communication: An integrative
hypothesis of carcinogenesis. Pathobiol 58, 265-278
(1990)
190. Kanemitsu MY and AF Lau: Epidermal growth factor
stimulates the disruption of gap junctional communication and
connexin43 phosphorylation independent of 12-0-
tetradecanoylphorbol 13-acetate-sensitive protein kinase C: The
possible involvement of mitogen-activated protein kinase. Mol
Biol Cell 4, 837-848 (1993)
191. Bukauskas FF, C Kempf & R Weingart: Cytoplasmic
bridges and gap junctions in an insect cell line (Aedes
albopictus) Exp Physiol 77, 903-911 (1992)
192. Shiokawa-Sawada M, H Mano, K Hanada, S Kakudo, T
Kameda, K Miyazawa, Y Nakamaru, T Yuasa, Y Mori, M
Kumegawa & Y Hakeda: Down-regulation of gap junctional
intercellular communication between osteoblastic MC3T3-E1
cells by basic fibroblast growth factor and a phorbol ester (12-
O-tetradecanoylphorbol-13-acetate) J Bone Miner Res 12,
1165-1173 (1997)
193. van Zoelen EJJ & LGJ Tertoolen: Transforming growth
factor-B enhances the extent of intercellular communication
between normal rat kidney cells. J Biol Chem 266, 12075-
12081 (1991)
194. Chandross KJ, M Chanson, DC Spray & JA Kessler:
Transforming growth factor-á1 and forskolin modulate gap
Gap junctions and cancer
234
junctional communication and cellular phenotype of cultured
Schwann cells. J Neurosci 15, 262-273 (1995)
195. Gibson DFC, MZ Hossain, GS Goldberg, P Acevedo &
JS Bertram: The mitogenic effects of transforming growth
factors beta 1 and beta 2 in C3H/10T1/2 cells occur in the
presence of enhanced gap junctional communication. Cell
Growth Diff 5, 687-696 (1994)
196. Rudkin GH, DT Yamaguchi, K Ishida, WJ Peterson, F
Bahadosingh, D Thye & TA Miller: Transforming growth
factor-á, osteogenin & bone morphogenetic protein-2 inhibit
intercellular communication and alter cell proliferation in
MC3T3-E1 cells. J Cell Physiol 168, 433-441 (1996)
197. Rogers M, JM Berestcky, MZ Hossain, H Guo, R Kadle,
BJ Nicholson & JS Bertram: Retinoid-enhanced gap junctional
communication is achieved by increased levels of connexin 43
mRNA and protein. Molec Carcinogenesis 3, 335-343 (1990)
198. Mehta PP, JS Bertram & WR Loewenstein: The actions
of retinoids on cellular growth correlate with their actions on
gap junctional communication. J Cell Biol 108, 1053-1065
(1989)
199. Zhang L-X, RV Cooney & JS Bertram: Carotenoids
up-regulate connexin43 gene expression independent of
their provitamin A or antioxidant properties. Cancer Res
52, 5707-5712 (1992)
200. Ruch RJ, SJ Cheng & JE Klaunig: Prevention of
cytotoxicity and inhibition of intercellular
communication by antioxidant catechins isolated from
Chinese green tea. Carcinogenesis 10, 1003-1006
(1989)
201. Sigler K & RJ Ruch: Enhancement of gap
junctional intercellular communication in tumor
promoter-treated cells by components of green tea. Cancer
Lett 69, 15-19 (1993)
202. Chaumontet C, V Bex, I Gaillard-Sanchez, C Seillan-
Heberden, M Suschetet & P Martel: Apigenin and
tangeretin enhance gap junctional intercellular
communication in rat liver epithelial cells. Carcinogenesis
15, 2325-2330 (1994)
203. Weinberg R: Tumor supressor genes. Neuron 11,
191-196 (1993)
204. Meyer DJ, SB Yancey, J-P Revel & A Peskoff:
Intercellular communication in normal and regenerating rat
liver: a quantitative analysis. J Cell Biol 91, 505-523
(1981)
205. Dermietzel R, SB Yancey, O Traub, K Willecke & J-
P Revel: Major loss of the 28-kD protein of gap junction in
proliferating hepatocytes. J Cell Biol 105, 1925-1934
(1987)
206. Gordon RE, BP Lane & M Marin: Regeneration of
rat tracheal epithelium: Changes in gap junctions during
specific phases of the cell cycle. Exp Lung Res 3, 47-56
(1982)
207. Shiba Y, Y Sasaki & Y Kanno: Inhibition of gap-
junctional intercellular communication and enhanced
binding of fibronectin-coated latex beads by stimulation of
DNA synthesis in quiescent 3T3-L1 cells. J Cell Physiol
145, 268-273 (1990)
208. Stein LS, J Boonstra & RC Burghardt: Reduced cell-
cell communication between mitotic and nonmitotic
coupled cells. Exp Cell Res 198, 1-7 (1992)
209. Koo SK, DY Kim, SD Park, KW Kang & CO Joe:
PKC phosphorylation disrupts gap junctional intercellular
communication at G0/S phase in clone 9 cells. Mol Cell
Biochem 167, 41-49 (1997)
210. Theodorescu D, M Caltabiano, R Grieg, D Rieman &
RS Kerbel: Reduction of TGF-beta activity abrogates
growth promoting tumor cell-cell interactions in vivo. J
Cell Physiol 148, 380-390 (1991)
211. Gille J & RA Swerlick: Integrins: Role in cell
adhesion and communication. Ann NY Acad Sci 797, 93-
106 (1996)
212. Martin W, G Zempel, D Hulser & K Willecke: Growth
inhibition of oncogene-transformed rat fibroblasts by
cocultured normal cells: relevance of metabolic cooperation
mediated by gap junctions. Cancer Res 51, 5348-5354 (1991)
213. Borchers AH, MB Powell, NE Fusenig & GT Bowden:
Paracrine factor and cell-cell contact-mediated induction of
protease and c-ets gene expression in malignant
keratinocyte/dermal fibroblast cocultures. Exp Cell Res 213,
143-147 (1994)
214. Saeki T, T Morioka, M Arakawa, F Shimizu & T Oite:
Modulation of mesangial cell proliferation by endothelial cells
in coculture. Am J Physiol 139, 949-957 (1991)
215. Stoker MGP: Transfer of growth inhibition between
normal cells: autoradiographic studies using marked cells. J
Cell Sci 2, 293-304 (1967)
216. Mehta PP, JS Bertram & WR Loewenstein: Growth
inhibition of transformed cells correlates with their junctional
communication with normal cells. Cell 44, 187-196 (1986)
217. Esinduy C, CC Chang, JE Trosko & RJ Ruch: In vitro
growth inhibition of neoplastically transformed cells by non-
Gap junctions and cancer
235
transformed cells: requirement for gap junctional intercellular
comunication. Carcinogenesis 16, 915-921 (1995)
218. Goldberg GS, KD Martyn & AF Lau: A connexin 43
antisense vector reduces the ability of normal cells to inhibit the
foci formation of transformed cells. Mol Carcinog 11, 106-114
(1994)
219. Ruch RJ, XJ Guan & K Sigler: Inhibition of gap
junctional intercellular communication and enhancement of
growth in BALB/c 3T3 cells treated with connexin43 antisense
oligonucleotides. Mol Carcinog 14, 269-274 (1995)
220. Reaume AG, PA De Sousa, S Kulkarni, BL Langille, D
Zhu, TC Davies, SC Juenja, GM Kidder & J Rossant: Cardiac
malformation in neonatal mice lacking connexin43. Science
267, 1831-1834 (1995)
221. Martyn KD, WE Kurata, BJ Warn-Cramer, JM Burt, E
TenBroek & AF Lau: Immortalized connexin43 knockout cell
lines display a subset of biological properties associated with
the transformed phenotype. Cell Growth Differ 8, 1015-1027
(1997)
222. Naus CCG, JF Bechberger, YC Zhang, L Venance, H
Yamasaki, SC Juneja, GM Kidder & C Giaume: Altered gap
junctional communication, intercellular signaling, and growth
in cultured astrocytes deficient in connexin43. J Neurosci Res
49, 528-540 (1997)
223. Temme A, A Buchmann, HD Gabriel, E Nelles, M
Schwarz & K Willecke: High incidence of spontaneous and
chemically induced liver tumors in mice deficient for
connexin32. Curr Biol 7, 713-716 (1997)
224. Sullivan R & CW Lo: Expression of a connexin
43/á-galactosidase fusion protein inhibits gap junctional
communication in NIH3T3 cells. J Cell Biol 130, 419-429
(1995)
225. Paul DL, K Yu, R Bruzzone, RL Gimlich & DA
Goodenough: Expression of a dominant negative inhibitor
of intercellular communication in the early Xenopus
embryo causes delamination and extrusion of cells.
Development 121, 371-381 (1995)
226. Omori Y, M Mesnil & H Yamasaki: Connexin 32
mutations from X-linked Charcot-Marie-Tooth disease
patients: Functional defects and dominant negative effects.
Mol Biol Cell 7, 907-916 (1996)
227. Duflot-Dancer A, M Mesnil & H Yamasaki:
Dominant-negative abrogation of connexin-mediated cell
growth control by mutant connexin genes. Oncogene 15,
2151-2158 (1997)
228. Mehta PP, A Hotz-Wagenblatt, B Rose, D Shalloway
& WR Loewenstein: Incorporation of the gene for a cell-
cell channel protein into transformed cells leads to
normalization of growth. J Memb Biol 124, 207-225 (1991)
229. Zhu D, S Caveney, GM Kidder & CCG Naus:
Transfection of C6 glioma cells with connexin 43 cDNA:
Analysis of expression, intercellular coupling & cell
proliferation. Proc Natl Acad Sci USA 88, 1883-1887 (1991)
230. Eghbali B, JA Kessler, LM Reid, C Roy & DC Spray:
Involvement of gap junctions in tumorigenesis: Transfection of
tumor cells with connexin 32 cDNA retards growth in vivo.
Proc Natl Acad Sci USA 88, 10701-10705 (1991)
231. Naus CCG, K Elisevich, D Zhu, DJ Belliveau & RF Del
Maestro: In vivo growth of C6 glioma cells transfected with
connexin43 cDNA. Cancer Res 52, 4208-4213 (1992)
232. Rose B, PP Mehta & WR Loewenstein: Gap-junction
protein gene suppresses tumorigenicity. Carcinogenesis 14,
1073-1075 (1993)
233. Masuda M, K Yamazaki, T Matsunaga, J Kanzaki &
Y Hosoda: Melanocytes in the dark cell area of human
vestibular organs. Acta Otolaryn 519, 152-157 (1995)
234. Mesnil M, V Krutovskikh, C Piccoli, C Elfgang, O
Traub, K Willecke & H Yamasaki: Negative growth control
of HeLa cells by connexin genes: Connexin species
specificity. Cancer Res 55, 629-639 (1995)
235. Hirschi KK, CE Xu, T Tsukamoto & R Sager: Gap
junction genes Cx26 and Cx43 individually suppress the
cancer phenotype of human mammary carcinoma cells and
restore differentiation potential. Cell Growth Differ 7, 861-
870 (1996)
236. Ruch RJ, K Cesen-Cummings & AM Malkinson:
Role of gap junctions in lung neoplasia. Exp Lung Res (in
press)
237. Pastan IH, GS Johnson & WB Anderson: Role of
cyclic nucleotides in growth control. Annu Rev Biochem
44, 491-522 (1975)
238. Wiemelt AP, MJ Engleka, AF Skorupa & FA
McMorris: Immunochemical visualization and quantitation
of cyclic AMP in single cells. J Biol Chem 272, 31489-
31495 (1997)
239. L’Allemain G, JN Lavoie, N Rivard, V Baldin & J
Pouyssegur: Cyclin D1 expression is a major target of the
cAMP-induced inhibition of cell cycle entry in fibroblasts.
Oncogene 14, 1981-1990 (1997)
240. Pines J: Cyclins from sea urchins to HeLa cells:
making the human cell cycle. Biochem Soc Trans 24, 15-33
(1996)
Gap junctions and cancer
236
241. Kato J, M Matsuoka, K Polyak, J Massague & CJ
Sherr: Cyclic AMP-induced G1 phase arrest mediated by
an inhibitor (p27kip1) of cyclin-dependent kinase 4
activation. Cell 79, 487-496 (1994)
242. Phipps M, J Darozewski & J Phipps: How the
neighborhood coherence principle (NCP) can give
riese to tissue homeostasis: A cellular automation
approach. J Theor Biol 185, 475-487, 1997
243. Durand RE: Distribution and activity of
antineoplastic drugs in a tumor model. J Natl Cancer Inst
81, 146-152 (1989)
244. Denning C & JD Pitts: Bystander effects of different
enzyme-prodrug systems for cancer gene therapy depend
on different pathways for intercellular transfer of toxic
metabolites, a factor that will govern clinical choice of
appropriate regimes. Hum Gene Ther 8, 1825-1835 (1997)
245. Paillard F: Bystander effects in enzyme/prodrug gene
therapy - Commentary. Hum Gene Ther 8, 1733-1735
(1997)
246. Bi WL, LM Parysek, R Warnick & PJ Stambrook: In
vitro evidence that metabolic cooperation is responsible for
the bystander effect observed with HSV tk retroviral gene
therapy. Hum Gene Ther 4, 725-731 (1993)
247. Pitts JD: Cancer gene therapy: A bystander effect
using the gap junctional pathway. Mol Carcinog 11, 127-
130 (1994)
248. Fick J, FG Barker, II, P Dazin, EM Westphale, EC
Beyer & MA Israel: The extent of heterocellular
communication mediated by gap junctions is predictive of
bystander tumor cytotoxicity in vitro. Proc Natl Acad Sci
USA 92, 11071-11075 (1995)
249. Mesnil M, C Piccoli & H Yamasaki: A tumor
suppressor gene, Cx26, also mediates the bystander effect
in HeLa cells. Cancer Res 57, 2929-2932 (1997)
250. Dilber MS, MR Abedi, B Christensson, B
Birkstrand, GM Kidder, CCG Naus, G Gahrton & CIE
Smith: Gap junctions promote the bystander effect of
herpes simplex virus thymidine kinase in vivo. Cancer Res
57, 1523-1528 (1997)
Key words: Connexins; Connexons; Gap junctions;
Tumor promotion; Carcinogenesis; Stem cells; Intercellular
communication; Oncogenes.
_______________________________________________
Send correspondence to: James E. Trosko, Ph.D., Michigan
State University, Department of Pediatrics and Human
Development, 240 NFSTC Bldg., East Lansing, Michigan
48824, Tel: (517) 353-6346, Fax: (517) 432-6340; E-mail:
trosko@pilot.msu.edu
