[CANCER RESEARCH 43, 3034-3040,
July 1983]
Perspectives in Cancer Research
Role of Retinoids in Differentiation and Carcinogenesis
Michael B. Sporn and Anita B. Roberts
Laboratory of Chemoprevention, Division of Cancer Cause and Prevention, National Cancer Institute, Bethesda, Maryland 20205
It has been known for more than 50 years that retinoids, the
family of molecules comprising both the natural and synthetic
analogues of retinol, are potent agents for control of both cellular
differentiation and cellular proliferation (70). In their original clas
sic paper describing the cellular effects of vitamin A deficiency in
the rat, Wolbach and Howe clearly noted that there were distinct
effects on both differentiation and proliferation of epithelial cells.
During vitamin A deficiency, it was found that proper differentia
tion of stem cells into mature epithelial cells failed to occur and
that abnormal cellular differentiation, characterized in particular
by excessive accumulation of keratin, was a frequent event.
Furthermore, it was noted that there was excessive cellular
proliferation in many of the deficient epithelia. Although the
conclusion that an adequate level of retinoid was necessary for
control of normal cellular differentiation and proliferation was
clearly stated in the original paper by Wolbach and Howe, a
satisfactory explanation of the molecular mechanisms underlying
these effects on both differentiation and proliferation still eludes
us more than 50 years later.
It was inevitable that the basic role of retinoids in control of
cell differentiation and proliferation would eventually find practical
application in the cancer field, and there have been great ad
vances in this area, particularly for prevention of cancer. Many
studies have shown that retinoids can suppress the process of
carcinogenesis in vivo in experimental animals (for reviews, see
Refs. 7, 33, 51, 54, 56, and 57), and these results are now the
basis of current attempts to use retinoids for cancer prevention
in humans. Furthermore, there is now an extensive literature on
the ability of retinoids to suppress the development of the
malignant phenotype in vitro (for reviews, see Refs. 6, 8, 30, and
31 ), and these studies corroborate the use of retinoids for cancer
prevention. Finally, most recently, it has been shown that reti
noids can exert effects on certain fully transformed, invasive,
neoplastic cells, leading in certain instances to a suppression of
proliferation (30) and in other instances to terminal differentiation
of these cells, resulting in a more benign, nonneoplastic pheno
type (10,11,60,62).
Even though there are many types of tumor
cells for which this is not the case (33, 52) (indeed, at present
there are only a limited number of instances in which such
profound effects of retinoids on differentiation and proliferation
of invasive tumor cells have been shown), this finding neverthe
less has highly significant implications for the problem of cancer
treatment. It emphasizes that in many respects cancer is fun
damentally a disease of abnormal cell differentiation (36,44), and
it raises the possibility that even invasive disease may eventually
be controlled by agents which control cell differentiation rather
than kill cells. Since carcinogenesis is essentially a disorder of
cell differentiation, the overall scientific problem of the role of
retinoids in either differentiation or carcinogenesis is essentially
Received October 14,1982; accepted April 11,1983.
3034
the same problem and will be considered as a single problem in
this brief review.
MajorProblemsRelatingto RetinoidsandCancer
In the broadest sense, there are 2 major domains relating to
retinoids and the cancer problem: (a) the practical development
and use of retinoids for either cancer prevention or treatment;
and (b) the elucidation of the cellular and molecular mechanisms
underlying the first domain. The first domain has attracted a
great deal of attention in the past 5 years and requires the
coordinated efforts of synthetic organic chemists, cell biologists,
investigators in experimental carcinogenesis and chemotherapy,
pharmacologists, toxicologists, and clinical investigators in order
to synthesize new retinoids, test them both in vitro and in vivo
for useful biological activity, establish their pharmacokinetic and
lexicological properties, and then bring the best new retinoids to
clinical trial for either prevention or treatment of specific types of
cancer. This is a problem of immense scope, complexity, and
expense, which is currently being pursued with vigorous interest
throughout the world. We have reviewed some of the key issues
in this first domain in previous articles (54, 56, 57) and will not
discuss them further at this point. Instead, we will focus the rest
of this short review on the problem of cellular and molecular
mechanisms. Studies in this area are not only of great theoretical
interest but should also facilitate the practical development and
use of retinoids for prevention and treatment of cancer. Eluci
dation of the mechanism of action of retinoids may also lead to
new applications for their use. It is reasonable to suggest that
retinoids may find applications in the prevention or treatment of
diseases other than cancer, the pathogenesis of which involves
abnormalities of cell differentiation and/or proliferation (55). In
terms of the scientific challenge, it is again worth emphasizing
that the problem of cellular and molecular mechanism is still
unsolved more than 50 years after the initial description of the
overall biological activity of the retinoids. The problem of the
mechanism of action of retinoids may be studied at 3 levels,
namely, in the whole animal, at the cellular level, and finally at
the molecular level, which we shall now consider in turn, using
effects on both differentiation and carcinogenesis as markers.
Mechanism of Action of Retinoids in Differentiation and Car
cinogenesis Studied in the Whole Animal
Although the earliest studies on retinoid deficiency in the whole
animal emphasized its effects on epithelial cell differentiation and
proliferation, the possibility that retinoid deficiency caused ab
normalities in nonepithelial cells derived from mesenchymal ele
ments was also noted by several careful investigators. Indeed,
in the 1920s, it was reported that there was a reduction in
hematopoietic cells in the bone marrow of vitamin A-deficient
CANCER
RESEARCH
VOL. 43
Downloaded from cancerres.aacrjournals.org on July 12, 2017. © 1983 American Association for Cancer Research.
Retinoids in Differentiation and Carcinogenesis
animals (21, 70). In the 1930s and 1940s, many studies were
performed on the need for retinoids for proper bone formation
(34, 39), and there was detailed investigation of the control of
osteoblasts and osteoclasts (cells derived from mesenchyme) by
retinoids (34). However, in the ensuing years, there was a much
greater emphasis on studies on the role of retinoids in control of
epithelial cell differentiation and proliferation, and the dogma that
retinoids were selectively involved in the control of epithelial cells
and were of relatively minor importance with respect to cells of
mesenchymal origin became scientific folklore.
In contrast, in the area of experimental embryology and tera
tology, there accumulated an impressive body of information
which indicated that retinoids had significant, selective teratogenic action on cells of mesenchymal origin in rat, mouse,
hamster, and chick embryos; we have reviewed these data
elsewhere (58). Particularly striking were the effects of either
retinol deficiency or retinoic acid excess on the development of
the very early vascular system of either the chick or the rat
embryo. In the 1-day-old chick embryo, retinol deficiency causes
failure of mesenchymal cells to proliferate and differentiate to
form the early vascular system (Refs. 66, 67; Fig. 1); treatment
of rat or chick embryos with excess retinol or retinoic acid has
similar effects (40).1 Furthermore, in the retinoid-deficient chick
embryo, normal development of the vascular system can be
restored by injection of appropriate amounts of various retinoids,
including esters of retinoic acid (66, 67). These results indicate
a very stringent requirement for retinoids, with either deficiency
or excess leading to abnormal development of tissues derived
from primitive mesenchyme. The effects of retinoids on the
developing vascular system of the early embryo appear to be
quite selective, since many other cell types do not appear to be
affected to anywhere near the same extent (58, 66, 67). In the
early embryo, a common stem cell type has long been believed
to be a precursor to both blood cells themselves and those cells
which will form the walls of the earliest blood vessels (50). The
preceding observations, made with only the simplest of morpho
logical techniques, suggest that retinoids play a role in controlling
the proliferation and differentiation of these mesenchymal pre
cursor cells or their early progeny; more recent work, using
sophisticated cell culture and recombinant DMA techniques, has
added important further information, as will be discussed later.
Mechanistic studies in the whole animal on the role of retinoids
in prevention of carcinogenesis have dealt largely with the pre
vention of epithelial carcinogenesis. There are especially con
vincing data on the efficacy of many different retinoids in preven
tion of skin, breast, and bladder cancer in experimental animals
(5, 7, 33, 37, 38, 57, 59, 65). Overall, these studies suggest that
retinoids exert a hormone-like control of either cell proliferation
or cell differentiation. However, in the whole animal, it is ex
tremely difficult to separate these 2 parameters; they are inti
mately linked with each other. Some investigators have stressed
the role of retinoids as antiproliferative agents (38), while others
have emphasized that the role of retinoids in control of differen
tiation, rather than proliferation, may be more important (4).
Whole-animal studies do not easily lend themselves to separate
analysis of these 2 parameters. Indeed, the problem of the
separation of the effects of retinoids on cell proliferation, as
contrasted to effects on cell differentiation, may turn out to be
more semantic than real, once a complete genetic analysis, using
1M. B. Sporn and D. L. Newton, unpublished results.
recombinant DMA methods, is available. The key problem is not
to deal with the semantics of whether retinoids preferentially
affect cell proliferation or cell differentiation but to identify the
specific genes, the function of which is ultimately controlled by
retinoids, either directly or indirectly. This problem cannot be
solved in the whole animal; it requires isolated cellular systems
and modern methods of molecular analysis. We will now discuss
the application of studies in these areas to understanding the
role of retinoids in differentiation and carcinogenesis.
Cellular Mechanism of Action of Retinoids in Differentiation
and Carcinogenesis
Significant advances in understanding the mechanism of action
of retinoids did not occur until in vitro systems were used as
experimental tools. The development of methods for organ cul
ture of tissues, such as the skin and the prostate, in which
retinoids are particularly active, was a major advance. In the
1950s, the classic studies of Fell and Mellanby (20) showed that
the differentiated phenotype of chick epidermis in organ culture
could be changed from keratinized to mucus producing by
treatment with retinol or retinyl acetate. In cultures treated with
retinoids, the keratinizing cells of the epidermis disappeared and
were replaced by mucus-producing cells, and in some instances
even by ciliated cells, which are not found in normal skin (20).
These organ culture experiments were essentially the reverse of
those performed by Wolbach and Howe in the 1920s in the
whole animal. In the Wolbach-Howe studies, retinoid deficiency
caused disappearance of normal mucociliary epithelium, with
replacement by keratinizing cells (keratinizing squamous meta
plasia); in the Fell-Mellanby experiments, retinoid excess caused
disappearance of normal keratinizing epithelium, with replace
ment by mucus and ciliated cells (mucus metaplasia).
The next significant advance in this area also used organ
culture methodology. Lasnitzki (26), working in the same labo
ratory as Fell and Mellanby, was able to show that the premalignant phenotype of mouse prostate glands that had been treated
with the carcinogen, 3-methylcholanthrene, could be altered by
retinoids. The atypical epithelial cells that were induced by the
carcinogen disappeared upon retinoid treatment of the organ
cultures, and they were replaced by cells with more normal
morphology (26). The effects of the retinoids were to suppress
abnormal cellular differentiation that had been induced by the
carcinogen in the epithelium of the prostate gland and to restore
a more normal pattern of epithelial differentiation. In other organ
culture studies, Lasnitzki (27) also made the important observa
tion that there were significant morphological similarities between
vitamin A-deficient prostatic epithelium and prostatic epithelium
in cultures that had been treated with methylcholanthrene; these
studies provide further evidence for the concept that the mech
anisms of action of retinoids in both differentiation and carcino
genesis are closely linked.
In spite of the advances that were made in the above experi
ments, organ culture methods have definite liabilities for analysis
of mechanism; they use mixed-cell populations, from which it is
very difficult, if not impossible, to obtain replicate samples of
homogeneous cells. The introduction of cell culture methodology
to studies of retinoid mechanism was therefore of great impor
tance and now is allowing molecular investigation of the role of
retinoids in differentiation and carcinogenesis. In contrast to the
organ culture studies, in which the emphasis was on the role of
JULY 1983
Downloaded from cancerres.aacrjournals.org on July 12, 2017. © 1983 American Association for Cancer Research.
3035
M. B. Sporn and A. B. Roberts
retinólas in control of epithelial differentiation, œil culture studies
have emphasized the role of retinoids in cells of mesenchymal
origin, if only because such cells are grown more easily in culture.
One may suppose that, as better systems are developed for
epithelial cell culture, there will be increasing investigation of
retinoids in many different types of epithelium using these meth
ods; this has already happened with epidermal cell culture (72).
Continuous cell lines of mesenchymal origin have been widely
used to study the effects of retinoids on both differentiation and
carcinogenesis; the cell lines which have been used are both
neoplastic and nonneoplastic. The experiments which opened
up this area of investigation were those of Merriman and Bertram
(35) and Harisiadis ef al. (24), which showed that retinoids can
act directly on nonneoplastic cells to suppress the process of
malignant transformation induced by either chemicals or radia
tion. In the case of the experiments done with suppression of
chemical carcinogenesis, it was clearly shown that retinoids were
effective in suppressing transformation even when they were
applied to cells a full week after original exposure of the cells to
carcinogen. Whatever the genetic damage caused by the carcin
ogen, it had already occurred. The role of the retinoids in these
experiments was thus clearly shown to be a suppressor of the
expression of the malignant phenotype in cells that had been
previously initiated by a carcinogen (35). Furthermore, in these
experiments, continuous presence of the retinoids was required
to suppress the malignant phenotype; removal of the retinoids
from the culture allowed expression of the transformed state.
Retinoids can also change the differentiation of invasive neo
plastic cells growing in either monolayer or suspension culture.
The most striking example of this phenomenon is the induction
of terminal differentiation in murine F9 teratocarcinoma cells (60,
62) or human promyelocytic leukemia cells (Refs. 10 and 11 ; Fig.
2); in these cases, the differentiated phenotype is drastically
changed from neoplastic to nonneoplastic, and proliferation of
the induced cells is permanently suppressed. In the F9 system,
retinoids induce terminal differentiation of teratocarcinoma stem
cells to cells which resemble parietal endoderm; a variety of new
proteins is induced in the differentiated cells (62, 63). In the
human promyelocytic leukemia system, retinoids can induce
terminal differentiation of malignant leukemia cells, leading to
formation of morphologically mature granulocytes, which have
functional markers of the mature neutrophil (10,11); these results
have been obtained with the established HL60 cell line (11), as
well as with primary cultures of other promyelocytic leukemia
cells (10). These studies with neoplastic leukemia cells in turn
have had a major influence on studies on possible effects of
retinoids on normal myeloid differentiation. Since some of the
leukemias may be viewed as diseases in which there is a block
or arrest in normal myeloid differentiation and maturation (14,
23) and since retinoids can apparently overcome this block in
certain leukemia cells, it has been suggested that retinoids may
also be involved in normal hematopoiesis (11,18).
The mechanism of all of these effects of retinoids, whether
they be to alter the differentiated phenotype in nonneoplastic or
preneoplastic epithelial cells in organ culture, to suppress the
appearance of the neoplastic phenotype in nonneoplastic mes
enchymal cells in monolayer culture, or to induce the terminally
differentiated phenotype in fully neoplastic cells in monolayer
culture, is not known. It is tempting to believe that there is a
common mechanism (or limited number of mechanisms), which
underlies all of these phenomena. In cell culture studies, as we
3036
noted before in the studies on whole animals, various investiga
tors again have chosen to emphasize the role of retinoids in
control of either cell proliferation (30) or cell differentiation (72).
It would appear that retinoids control both processes and that
any dispute over which is more important is relatively fruitless at
present. Rather, it would seem more productive to focus on the
specific molecular processes involved, to which we shall now
turn.
Molecular Mechanism of Action of Retinoids in Differentiation
and Carcinogenesis
Over the years, numerous hypotheses on the molecular mech
anism of action of retinoids in control of differentiation have been
proposed, but none has stood up to the experimental data. In
particular, any hypothesis relating to molecular mechanism of
action must take into account the evidence, now overwhelming,
that retinole acid will support growth in the whole animal as
effectively as retinol (74), that retinoic acid is more active than
retinol or retinal in numerous in vitro test systems (11, 30, 57,
62), and that there is no evidence that the mammalian organism
can convert retinoic acid to retinol (19). In many test systems,
retinoic acid is at least 100 to 1000 times more active than is
retinol (11, 57, 62), and biological activity can be measured at
levels as low as 10~11M (57). Thus, the hypothesis, proposed in
the 1960s, that retinol directly modifies membrane structure (16)
to exert its biological effects is now of only historical interest.
More recently, it has been suggested that a primary biological
role of the retinoids is to participate in sugar transfer reactions
by means of the intermediate retinyl phosphate mannose, which
is a metabolite of retinol (1, 15, 71). This hypothesis cannot be
rationalized with the experimental data on retinoic acid, sum
marized above. Neither is there any convincing evidence at
present for a metabolite of retinoic acid which is involved in sugar
transfer reactions. The recent synthesis of a new series of
retinoids (29), which may be viewed as retinoidal benzoic acid
derivatives (Chart 1) and which are even more potent than
retinoic acid in many test systems both in vivo and in vitro,
provides even further experimental evidence against any essen
tial role for retinyl phosphate mannose in control of differentiation
or carcinogenesis. The new analogue shown in Chart 1 (or its
derivatives) will support growth in the whole animal fed a vitamin
A-deficient diet (29), is at least 1000 times more active than
retinol in suppressing skin carcinogenesis in the mouse (29), is
more than 100 times as active as retinol in the hamster trachéal
organ culture system (57), and is more than 1000 times as active
as retinol or retinyl acetate in the F9 teratocarcinoma or HL60
promyelocytic leukemia test systems (61). With data such as
these at hand, it is unreasonable to believe that retinyl phosphate
mannose plays any universally critical role in the control of
differentiation or carcinogenesis. Although one cannot exclude
the possibility that there may be some situations in which retinyl
phosphate mannose may play some role, currently available
information would relegate this metabolite to a minor role in
control of differentiation and carcinogenesis.
If one wishes to develop a molecular hypothesis of retinoid
mechanism that is compatible with the broadest range of exper
imental data, then the simplest one that can be proposed at
present is to suggest that retinoids modify gene expression.
This, of course, is not a new idea. If one takes this as a general
proposition, then 2 important questions follow: (a) which genes
CANCER
RESEARCH
VOL. 43
Downloaded from cancerres.aacrjournals.org on July 12, 2017. © 1983 American Association for Cancer Research.
Retinoids in Differentiation and Carcinogenesis
COOH
Chart 1. Structure of a new retinoidal benzole acid derivative, which is 1000
times more active than is retinol or retinyl acetate in several test systems, both in
vitro and in vivo.
are controlled by retinoids; and (b) how are these genes con
trolled by retinoids? (is the mechanism one of direct or indirect
control?). We will provide only an outline of what is known
regarding these 2 questions.
With respect to which genes are known to be controlled by
retinoids, one is impressed by the number of recent reports
which indicate that retinoids control the expression of many
proteins which either are direct constituents of the cytoskeleton
and extracellular matrix or participate in the formation of cyto
skeleton and matrix. These proteins include keratins (22), colla
gen (62, 63), collagenase (12), transglutaminase (53, 73), and
laminin (63). Determination of the specific types of cytoskeletal
or matrix proteins which are produced in cells is now being used
as a specific marker for cell differentiation, and it would appear
that retinoids are intimately involved in this process. Other pro
teins the expression of which is known to be controlled by
retinoids include plasminogen activator (62, 63), alkaline phosphatase (48, 63), and the receptor for epidermal growth factor
(25, 47). Furthermore, the important observation has recently
been made in the HL60 system that retinoic acid controls the
expression of the myc oncogene. Using a specific molecular
probe for the myc gene, it has been shown that physiological
levels of all-frans-retinoic acid suppress myc gene expression in
HL60 cells (Ref. 68; Fig. 3). Although the specific molecular
function for the myc gene product has not yet been elucidated,
it is presumed that in some way the excessive expression of this
gene and its product is correlated with the excessive proliferation
(and perhaps with the arrested differentiation) of the HL60 cell.
However, the possibility must be considered that an oncogene
other than myc may also contribute to the neoplastic behavior
of the HL60 cell2 and that retinoic acid may have significant
interactions with this gene as well. Furthermore, as of the
present, the kinetics of the interaction of retinoic acid with myc
in HL60 has not been determined, and it is not yet clear whether
retinoic acid controls myc expression directly or indirectly.
Thus, at the gene level, it appears that retinoids affect the
expression of genes or gene products involved with both differ
entiation and proliferation. The remaining, and most difficult,
question is, "How do they do it?" The overall problem of the
control of gene expression is beyond the scope of this article;
one may conceive of both direct and indirect mechanisms that
involve either regulation of gene transcription itself, regulation of
processing of primary gene transcripts, or regulation of transla
tion of processed message. Little is known about retinoids in
any of these areas. Recent experiments have demonstrated
effects of retinoids on genomic expression in retinoid-deficient
rat tissues (43); however, the complexity of the experimental
system precludes analysis of the molecular mechanisms in
volved. By analogy with the steroids, it has been suggested that
the effects of retinoids in controlling gene expression are me2 R. A. Weinberg, personal communication.
ro
HL60
HL60 + DMSO
HL60 + Ret. Acid
HL60
~
Fig. 3. Hybridization of myc probe to RNA from HL60 cells Induced to differen
tiate. Degree of differentiation was judged by the percentage of cells able to reduce
nitroblue tetrazolium, as well as morphological criteria (see Fig. 2). Left to right:
Lane 1, RNA from uninduced HL60 (less than 2% of cells nitroblue tetrazolium
positive); Lane 2, RNA from HL60 induced to differentiate with dimethyl sulfoxide
(DMSO) (87% of cells nitroblue tetrazolium positive); Lane 3. RNA from HL60
induced to differentiate with retinoic (ReÃ-.)acid (40% of cells nitroblue tetrazolium
positive); Lane 4, a second isolate of RNA from uninduced HL60 (less than 2% of
cells nitroblue tetrazolium positive). On a molar basis, retinoic acid is approximately
1 million times more active than is dimethyl sulfoxide in inducing differentiation in
HL60 (11). Reprinted with permission from Ref. 68.
dialed by specific intracellular binding proteins (13). However,
retinoids have significant effects in control of both differentiation
and carcinogenesis in 2 important cell systems, namely HL60
and 10T1/2fibroblasts, in which no retinoid-binding protein (anal
ogous to steroid-binding proteins) can be detected (17, 28).3
The alternative to a steroid-like mechanism for retinoids is to
suggest that they control gene expression via interactions with
protein kinases, both cyclic AMP4 dependent and cyclic AMP
independent. Retinoids have been shown to increase cyclic AMP1dependent protein kinase activity in B16 melanoma cells (32),
which are highly sensitive to their antiproliferative effects (30),
as well as in F9 teratocarcinoma cells which are induced to
differentiate by retinoic acid (45). Furthermore, A/6,O2'-dibutyryl
cyclic adenosine 3':5'-monophosphate
markedly potentiates the
differentiating effects of retinoic acid in both F9 teratocarcinoma
cells (63) and HL60 leukemia cells (42). Very recent work has
also suggested that a secondary effect of retinoic acid may be
the induction in the F9 system of a calcium- and phospholipiddependent, cyclic AMP-independent protein kinase activity (2,
41, 64) and that some of the interactions between retinoids and
phorbol esters may be mediated through this system (2). These
latest studies on a calcium-dependent protein kinase system
provide an important link between retinoids and calcium, which
is now assuming an increasing importance in control of cell
proliferation and differentiation (46, 69). Thus, although studies
on the interactions between retinoids and the various protein
kinase systems of the cell have only recently begun, they have
already yielded significant new data which will need to be inte
grated into an overall hypothesis of mechanism of action.
Ultimately, it would appear that the problem of the molecular
mechanism of action of retinoids in control of differentiation and
carcinogenesis is converging on one of the central problems in
all of biology, namely, the control of gene expression. There may
3 T. R. Breitman, personal communication.
4 The abbreviation used is: cyclic AMP, cyclic adenosine 3':5'-monophosphate.
JULY 1983
Downloaded from cancerres.aacrjournals.org on July 12, 2017. © 1983 American Association for Cancer Research.
3037
M. B. Sporn and A. B. Roberts
be new mechanisms, yet to be discovered, that may be critically
involved in this process. For example, yet another question that
awaits further experimentation is the functional relationship be
tween retinoids and polypeptide growth factors that control cell
proliferation and differentiation (58). The role of peptide growth
factors in controlling these processes is of fundamental impor
tance (for a review, see Ref. 9); for example, there is evidence
that a new family of transforming growth factors (type a and
type ß),which we have recently described (3, 49), may play a
key role in the processes of carcinogenesis and differentiation.
Clearly, any future hypothesis dealing with mechanism of action
of retinoids will need to integrate the role of retinoids, peptide
growth factors, and specific genes controlling differentiation and
proliferation. In particular, it must define the precise relationship
between retinoids, transforming growth factors, and cellular on-
15. De Luca, L. M., Bhat, P. V., Sasak, W., and Adamo, S. Biosynthesis of
phosphoryl and glycosyl phosphoryl derivatives of vitamin A in biological
membranes. Fed. Proc., 38: 2535-2539,1979.
16. Dingle, J. T., and Lucy, J. A. Vitamin A, carotenoids, and cell function. Biol.
Rev., 40:422-461,1965.
17. Douer, D., and Koeffler, H. P. Retinoic acid—inhibition of the donai growth of
human myeloid leukemia cells. J. Clin. Invest., 69: 277-283,1982.
18. Douer, D., and Koeffler, H. P. Retinoic acid enhances growth of human early
erythroid progenitor cells in vitro. J. Clin. Invest., 69:1039-1041,1982.
19. Dowling, J. E., and Wald, G. The biological function of vitamin A acid. Proc.
Nati. Acad. Sei. U. S. A., 46: 587-608,1960.
20. Fell, H. B., and Mellanby. E. Metaplasia produced in cultures of chick ectoderm
by high vitamin A. J. Phystol., JÕ9:470-488,1953.
21. Findlay, G. M., and McKenzie, R. D.The bone marrow in deficiency diseases.
J. Pathd. Bacteriol., 25: 402-403,1922.
22. Fuchs, E., and Green, H. Regulation of terminal differentiation of cultured
human keratinocytes by vitamin A. Cell, 25: 617-625,1981.
23. Gallo, R. C. On the origin of human acute myeloblastic leukemia: virus-"hot
spot" hypothesis. In: R. Neth, R. C. Gallo, S. Spiegelman, and F. Stohlman
cogenes. The breadth, potency, and specificity of retinoids in the
control of cell function all suggest that retinoids will be valuable
tools for the experimental scientist to unravel molecular mecha
nisms, in addition to their practical usefulness in controlling
differentiation and carcinogenesis.
24.
Acknowledgments
27.
We thank Ruth Morsillo for expert assistance
manuscript.
with the preparation
of the
25.
26.
28.
29.
References
30.
1. Adamo, S., De Luca, L. M., Silverman-Jones, C. S., and Yuspa, S. H. Mode of
action of ret mol. Involvement in glycosylation reactions of cultured mouse
epidermal cells. J. Btol. Chem., 254: 3279-3287,1979.
2. Anderson, W. B., Kraft, A. S., and Evain-Brion, D. Effect of retinole acid and
phorbol ester treatment of embryonal carcinoma cells on calcium, phospholipiddependent protein kinase activity. Cold Spring Harbor Conf. Cell Proliferation,
10: in press, 1983.
3. Anzano, M. A., Roberts, A. B., Meyers, C. A., Komoriya, A., Lamb, L C.,
Smith, J. M., and Sporn, M. B. Synergistic interaction of two classes of
transforming growth factors from murine sarcoma cells. Cancer Res., 42:
4776-4778,1982.
4. Astrup, E. G., and Paulson, J. E. Effect of retinoic acid pretreatment on 12-Otetradecanoylphorbol-13-acetate-induced
cell population kinetics and polyamine biosynthesis in hairless mouse epidermis. Carcinogenesis (Lond.), 3:
313-320, 1982.
5. Becci, P. J., Thompson, H. J., Grubbs, C. J., Squire, R. A., Brown, C. C.,
Sporn, M. B., and Moon, R. C. Inhibitory effect of 13-c/s-retinoic acid on
urinary bladder carcinogenesis induced in C57BL/6 mice by N-butyl-W-(4hydroxybutylXiitrosamine. Cancer Res., 38: 4463-4466,1978.
6. Bertram, J. S., Mordan, L. J., Domanska-Janik, K., and Bernacki, R. J. Inhibition
of In vitro neoplastic transformation by retinoids. In: M. S. Amott, J. van Eys,
and Y. M. Wang (eds.), Molecular Interrelations of Nutrition and Cancer, pp.
315-335. New York: Raven Press, 1982.
7. Bollag, W. Retinoids and cancer. Cancer Chemother. Pharmacol., 3:207-215,
1979.
8. Borek, C. Vitamins and micronutrients modify carcinogenesis and tumor pro
motion in vitro. In: M. S. Amott, J. van Eys, and Y. M. Wang (eds.). Molecular
Interrelations of Nutrition and Cancer, pp. 337-350. New York: Raven Press,
1982.
9. Bradshaw, R. A., and Sporn, M. B. (eds.). Symposium: Polypeptide growth
factors and the regulation of cell growth and differentiation. Fed. Proc., 42:
2590-2634,1983.
10. Breitman, T. R., Collins, S. J., and Keene, B. R. Terminal differentiation of
human promyelocytic leukemia cells in primary culture in response to retinoic
acid. Blood, 57: 1000-1004,1981.
11. Breitman, T. R., Setonick, S. E., and Collins, S. J. Induction of differentiation
of the human promyelocytic leukemia cell line (HL-60) by retinoic acid. Proc.
Nati. Acad. Sei. U. S. A., 77: 2936-2940,1980.
12. Brinckertioff, C. E., and Harris, E. D., Jr. Modulation by retinoic acid and
corticosteroids of collagenase production by rabbit synovial fibroblasts treated
with phorbol myristate acetate or polyethylene glycol). Biochim. Biophys. Acta,
677:424-432,1981.
13. Chytil. F., and Ong, D. E. Cellular retinol- and retinoic acid-binding proteins in
vitamin A action. Fed. Proc., 38: 2510-2514,1979.
14. Clarkson, B. C. Acute myelocytic leukemia in adults. Cancer (Phila.), 30:15721582,1972.
31.
3038
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
(eds.), Modem Trends in Human Leukemia, pp. 227-236. Munich: J. F.
Lehmanns Verlag, 1974.
Harisiadis, L., Miller, R. C., Hall, E. J., and Borek, C. A vitamin A analogue
inhibits radiation-induced oncogenic transformation. Nature (Lond.), 274: 486487,1978.
Jetten, A. M. Retinoids specifically enhance the number of epidermal growth
factor receptors. Nature (Lond.), 284:626-629,1980.
Lasnitzki, I. The influence of A hypervitaminosis on the effect of 20-methylcholanthrene on mouse prostate glands grown in vitro. Br. J. Cancer, 9: 434441,1955.
Lasnitzki, I. Hypovitaminosis-A in the mouse prostate gland cultured in chem
ically defined medium. Exp. Cell Res., 28: 40-51,1962.
LJbby, P. R., and Bertram, J. S. Lack of intracellular retinoid-binding proteins
in a retinol-sensitive cell line. Carcinogenesis (Lond.), 3: 481-484,1982.
Loeliger, P., Bollag, W., and Mayer, H. Arotinoids. a new class of highly active
retinoids. Eur. J. Med. Chem., 75:9-15,1980.
Lotan, R. Effects of vitamin A and its analogs (retinoids) on normal and
neoplastic cells. Biochim. Biophys. Acta, 605:33-91,1980.
Lotan, R., Thein, R., and Lotan, D. Suppression of the transformed cell
phenotype expression by retinoids. In: f. L. Meyskens and K. Prasad (eds.),
The Modulation and Mediation of Cancer by Retinoids. Basel: S. Karger AG,
in press, 1983.
Ludwig, K. W., Lowey, B., and Niles, R. M. Retinoic acid increases cyclic AMPdependent protein kinase activity in murine melanoma cells. J. Biol. Chem.,
255: 5999-6002,1980.
Mayer, H., Bollag, W., Hanni, R., and Rüegg,R. Retinoids, a new class of
compounds with prophylactic and therapeutic activities in oncology and der
matology. Experientia (Basel). 34:1105-1119,1978.
Mellanby, E. Vitamin A and bone growth: the reversibility of vitamin A deficiency
changes. J. Phystol., r05: 382-399,1947.
Merriman, R. L., and Bertram, J. S. Reversible inhibition by retinoids of 3methylcholanthrene-induced
neoplastic transformation in C3H/10TV2 CL8
cells. Cancer Res., 39:1661-1666,1979.
Mintz, B., and Fleischman, R. A. Teratocarcinomas and other neoplasms as
developmental defects in gene expression. Adv. Cancer Res., 34: 211-278,
1981.
Moon, R. C., Grubbs, C. J., Sporn, M. B., and Goodman, D. G. Retinyl acetate
inhibits mammary carcinogenesis induced by N-methyl-W-mtrosourea. Nature
(Lond.), 267: 620-621,1977.
Moon, R. C., Thompson, H. J., Becci, P. J., Grubbs, C. J., Gander, R. J.,
Newton, D. L., Smith, J. M., Phillips, S. L., Henderson, W. R., Mullen, L. T.,
Brown, C. C., and Spom, M. B. W-(4-Hydroxyphenyl)retinamide, a new retinoid
for prevention of breast cancer in the rat. Cancer Res., 39:1339-1346,1979.
Moore, L. A. Relationship between carotene, blindness due to constriction of
the optic nerve, papillary edema, and nyctalopia in calves. J. Nutr., 17: 443459,1939.
Morriss. G. M., and Steele, C. E. Comparison of the effects of retinol and
retinoic acid on postimplantation rat embryos in vitro. Teratology, 15: 109119,1977.
Nishizuka, Y., and Takai, Y. Calcium and phospholipid turnover in a new
receptor function for protein phosphorylation. Cold Spring Harbor Conf. Cell
Proliferation, 8:237-249,1981.
Olsson, I. L., Breitman, T. R., and Gälte,R. C. Priming of human myeloid
leukernic cell lines HL-60 and U-937 with retinoic acid for differentiation effects
of cyclic adenosine 3':5'-monophosphate-inducing
agents and a T-lymphocyte-derived differentiation factor. Cancer Res., 42: 3928-3933,1982.
Omori, M., and Chytil, F. Mechanism of vitamin A action: gene expression in
retinol deficient rats. J. Bid. Chem., 257: 14370-14374,1982.
Pierce, G. B., Shikes, R., and Fink, L. M. Cancer—A Problem of Developmental
Biology, Englewood Cliffs, N. J.: Prentice Hall, 1978.
Plet, A., Evain, D., and Anderson, W. B. Effect of retinoic acid treatment of F9
embryonal carcinoma cells on the activity and distribution of cyclic AMPdependent protein kinase. J. Biol. Chem., 257: 889-893,1982.
Rasmussen, H. Calcium and cAMP as Synarchic Messengers. New York: John
CANCER
RESEARCH
VOL. 43
Downloaded from cancerres.aacrjournals.org on July 12, 2017. © 1983 American Association for Cancer Research.
Retinoids in Differentiation and Carcinogenesis
Wiley & Sons, Inc., 1981.
47. Rees, A. R., Adamson, E. D., and Graham, C. F. Epidermal growth factor
receptors increase during the differentiation of embryonal carcinoma cells.
Nature (Lond.), 281: 309-311,1979.
48. Reese, D. H., Fiorentino, G. J„
Claflin, A. J., Malinin, T. I., and Poiitano, V. A.
Rapid induction of alkaline phosphatase activity by retinoic acid. Biochem.
Biophys. Res. Commun., 702:315-321,1981.
49. Roberts, A. B., Frolik, C. A., Anzano, M. A., and Sporn, M. B. Transforming
growth factors from neoplastic and non-neoplastic tissues. Fed. Proc., 42:
2621-2626,1983.
50. Romanoff, A. L. The Avian Embryo. New York: The Macmillan Co., 1960.
51. Saffiotti, U., Montesano, R., Sellakumar, A. R., and Borg, S. A. Experimental
cancer of the lung. Inhibition by vitamin A of the induction of tracheobronchial
squamous metaplasia and squamous cell tumors. Cancer (Phila.), 20: 857864,1967.
52. Schroder, E. W., Rapaport, E., and Black, P. H. Retinoids and cell proliferation.
Cancer Surv., in press, 1983.
53. Scott, K. F. F., Meyskens, F. L., and Russell, D. H. Retinoids increase
transglutaminase activity and inhibit omithine decarboxylase activity in Chinese
hamster ovary cells and in melanoma cells stimulated to differentiate. Proc.
Nati. Acad. Sei. U. S. A., 79: 4093-4097,1982.
54. Sporn, M. B., Dunlop, N. M., Newton, D. L., and Smith, J. M. Prevention of
chemical carcinogenesis by vitamin A and its synthetic analogs (retinoids). Fed.
Proc., 35: 1332-1338,1976.
55. Sporn, M. B., and Harris, E. D., Jr. Proliferative diseases. Am. J. Med., 70:
1231-1236,1981.
56. Sporn, M. B., and Newton, D. L. Chemoprevention of cancer with retinoids.
Fed. Proc., 38: 2528-2534,1979.
57. Sporn, M. B., and Newton, D. L. Retinoids and Chemoprevention of cancer.
In: M. S. Zedeck and M. Upkin (eds.), Inhibition of Tumor Induction and
Development, pp. 71-100. New York: Plenum Publishing Corp., 1981.
58. Sporn, M. B., Newton, D. L., Roberts, A. B., De Larco, J. E., and Todaro, G.
J. Retinoids and suppression of the effects of polypeptide transforming fac
tors—a new molecular approach to Chemoprevention of cancer. In: A. C.
Sartorelli, J. S. Lazo, and J. R. Benino (eds.), Molecular Actions and Targets
for Cancer Chemotherapeutic Agents, pp. 541-554. New York: Academic
Press, Inc., 1981.
59. Sporn, M. B., Squire, R. A., Brown, C. C., Smith, J. M., Wenk, M. L., and
Springer, S. 13-c/s-Retinoic acid: inhibition of bladder carcinogenesis in the
rat. Science (Wash. D. C.), 195: 487-489,1977.
60. Strickland, S. Mouse teratocardnoma cells: prospects for the study of embryogenesis and neoplasia. Cell, 24: 277-278,1981.
61. Strickland, S., Breitman, T. R., Frickel, F., Nürrenbach,A., Hädicke, E., and
Sporn, M. B. Structure-activity relationships of a new series of retinoidal
JULY 1983
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
benzoic acid derivatives as measured by induction of differentiation of murine
F9 teratocardnoma cells and human HL-60 promyelocytic leukemia cells.
Cancer Res., in press, 1983.
Strickland, S., and Mandavi, V. The induction of differentiation in teratocarcinoma stem cells by retinoic acid. Cell, 15: 393-403, 1978.
Strickland, S., Smith, K. K., and Marotti, K. R. Hormonal induction of differen
tiation in teratocardnoma stem cells: generation of parietal endoderm by
retinoic acid and dibutyryl cAMP. Cell, 21: 347-355,1980.
Takai, Y., Kishimoto, A., Kawahara, Y., Minakuchi, R., Sano, K„
Kikkawa, U.,
Mori, T., Yu, B., Kaibuchi, K., and Nishizuka, Y. Calcium and phosphatidylinositol turnover as signalling for trans-membrane control of protein phosphorylation. Adv. Cyclic Nucleotide Res., 14: 301-308,1981.
Thompson, H. J., Becci, P. J., Brown, C. C., and Moon, R. C. Effect of the
duration of retinyl acetate feeding on inhibition of 1-methyM-nitrosoureainduced mammary carcinogenesis in the rat. Cancer Res., 39: 3977-3980,
1979.
Thompson, J. N. The role of vitamin A in reproduction. In: H. F. DeLuca and
J. W. Suttie (eds.), The Fat-Soluble Vitamins, pp. 267-281. Madison, Wis.:
The University of Wisconsin Press, 1969.
Thompson, J. N., Howell, J. M., Pitt, G. A. J., and Mclaughlin, C. I. The
biological activity of retinoic add in the domestic fowl and the effects of vitamin
A deficiency on the chick embryo. Br. J. Nutr., 23: 471-490,1969.
Westin, E. H., Wong-Staal, F., Gelmann, E. P., Dalla Pavera, R., Papas, T. S.,
Lautenberger, J. A., Eva, A., Reddy, E. P., Tronick, S. R., Aaronson, S. A.,
and Gallo, R. C. Expression of cellular homologues of retroviral one genes in
human hematopoietic cells. Proc. Nati. Acad. Sei. U. S. A., 79: 2490-2494,
1982.
Whitfield, J. F., Boynton, A. L., MacManus, J. P., Rixon, R. H., Sikorska, M.,
Tsang, B., and Walker, P. R. The roles of calcium and cyclic AMP in cell
proliferation. Ann. N. Y. Acad. Sci., 339: 216-240, 1980.
Wolbach, S. B., and Howe, P. R. Tissue changes following deprivation of fat
soluble A vitamin. J. Exp. Med., 42: 753-777,1925.
Wolf, G., Kiorpes, T. C., Masushige, S., Schreiber, J. B., Smith, M. J., and
Anderson, R. S. Recent evidence for the participation of vitamin A in glycoprotein synthesis. Fed. Proc., 38: 2540-2543, 1979.
Yuspa, S. H. Retinoids and tumor promotion. In: D. A. Roe (ed.), Diet and
Cancer: From Basic Research to Policy Implications. New York: Alan R. LJss,
in press, 1983.
Yuspa, S. H., Ben, T., and Steinert, P. Retinoic acid induces transglutaminase
activity but inhibits comification of cultured epidermal cells. J. Biol. Chern.,
257:9906-9908,1982.
Zile, M., and DeLuca, H. F. Retinoic acid: some aspects of growth-promoting
activity in the albino rat. J. Nutr., 94: 302-308,1968.
3039
Downloaded from cancerres.aacrjournals.org on July 12, 2017. © 1983 American Association for Cancer Research.
M. B. Sporn and A. B. Roberts
Jb. 2
Fig. 1. Left, normal 3-day avian embryo. Note well-developed extraembryonic circulatory system, with large major blood vessels surrounding the embryo. Right,
abnormal 2-day avian embryo, resulting from retinoid deficiency. Although many structures, such as the somites, have formed relatively normally, there is a conspicuous
absenceof an extraembryonic circulatory system. Treatment of early embryos with retinoic acid methyl ester will restore normal development. Reprinted with permission
from Ref. 66.
Fig. 2. Retinoic acid induces morphological and functional maturation of leukemic cells obtained from a patient with promyetocytic leukemia.A, cells cultured without
retinoic acid consisting of promyelocytes with characteristic cytoplasmic granules, x 860. B, cells cultured with retinoic acid showing maturation to bandedand segmented
neutrophils. x 860. C, absence of nitroblue tetrazolium reduction by cells cultured without retinoic acid, x 315. 0, nitroblue tetrazolium reduction by cells incubated with
retinoic acid, x 315. Reprinted with permission from Ref. 10.
3040
CANCER
RESEARCH
VOL. 43
Downloaded from cancerres.aacrjournals.org on July 12, 2017. © 1983 American Association for Cancer Research.
Role of Retinoids in Differentiation and Carcinogenesis
Michael B. Spom and Anita B. Roberts
Cancer Res 1983;43:3034-3040.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/43/7/3034.citation
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on July 12, 2017. © 1983 American Association for Cancer Research.