Vol. 179, No. 3, 1991 September 30, 1991
AND BIOPHYSICAL RESEARCH COMMUNKATIONS Pages 1554-1561
BIOCHEMICAL
6’-SUBSTITUTED NAPHTHALENE-2-CARBOXYLIC ACID ANALOGS, A NEW CLASS OF RETINOIC ACID RECEPTOR SUBTYPE-SPECIFIC LIGANDS Gerhart Graupner, Gerard Malle*, Jean Maignan*, Gerard Lang*, Michel Prunieras* and Magnus Pfahl+ Cancer Center, La Jolla Cancer Research Foundation, 10901 North Torrey Pines Road, La Jolla, CA 92037 * Laboratoire
Received
July
The biological
22,
1991
response
by three nuclear as receptor Hybrid
de Recherche Fondamentale de L’Greal, 1 Avenue Eugene Schueller, 93600 Aulnay-sous-Bois, France
retinoic
subtype
receptors
receptors
ER-RARa, wild
of Q’-substituted
retinoic
type receptors
while
no activation
analogs
showed
a limited
retinoic
acid receptor
exhibited
strong
was observed.
subtype selectivity
can be defined
applications.
activation
different
transcriptional
Conversely,
assay.
measuring
acid, the three hybrid
the same induction
RARy. Three
We conclude
and for therapeutical
of retinoids
used to avoid
in response to retinoic
selectivity.
a
element
acid, elicited
of a receptor
the potential
were
RARp,and
naphthalene-2-carboxylic
is mediated
response
ER-RART
RARa,
TO explore
(retinoids)
a transcriptional
acid receptors. and
derivatives
we employed
an estrogen
ER-RARP
receptor
differentiation
RARa, p and 7.
activators,
that recognize
of endogenous
acid (RA) and synthetic
acid receptors,
selective
activities
corresponding
to retinoic
profile
retinoids,
as the analogs
activation
of y
two retinobenzoic
acid
that retinoids
with
unique
and tested for
their
impact on cellular
0 1991 Academic
profiles
of
Press, Inc.
Retinoic acid (RA) has been recognized as a pivotal regulatory compound in cell differentiation,proliferation and malignant transformation (1). These effects are mediated by nuclear receptors (RARa,RAR/3,RARy) which are members of the steroid/thyroid receptor family (2,3,4,5,6) and show tissue-specific expression patterns during embryogenesis and in adult life. The RARs am encoded by three different genes from which several isoforms can be derived ; these + To whom correspondence should be addressed . 0006-291X/91 Copyright All rights
$1.50
0 1991 by Academic Press, Inc. of reproduction in any form reserved.
1554
Vol.
179,
No.
BIOCHEMICAL
3, 1991
AND
isoforms appear to be regulated by tissue-specific
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
expression as well (7,8).-RA analogs (retinoids)
(9,lO) have long been employed to gain insight into molecular
parameters governing
seemingly unrelated properties as chemoprevention of tumorigenesis, and malformation. showing
differentiation
in the absence of cytoplasmic
such
Studies
retinoic acid binding protein (CRABP)
(11,12) and a lack of correlation between differentiation and retinoid CRABP binding (13) led to replacement of the initial concept of cytoplasmic proteins as the mediators of RA activity by the now established mechanism of action through nuclear receptors. Binding of synthetic retinoids to nuclear receptors has been correlated to biological activity (14), while other studies have shown transcriptional
activation by retinoids (15) and the ability to modulate RAR mRNA expression in
melanoma cells (16). Still, no significant differential transcriptional activation of RAR subtypes by synthetic retinoids has been demonstrated so far. Here we describe a class of 13-&r -RA analogs, derivatives
of the 6’-substituted
naphthalene-2-carboxylic acid, and present evidence that several of these analogs can preferentially induce hybrid receptors bearing the ligand binding domain of RAR subtypes. In comparison, two retinobenzoic acid derivatives (Am80 and Am580), representatives of a distinct class of all-rrunsRA analogs (13), show a different profile of RAR subtype selectivity.
Our data indicate that
retinoids can be synthesized that may govern specific biological functions by virtue of their RAR subtype specificity.
MATERIALS
AND METHODS
Retinoids : synthesis: The chloride of 6-methoxycarbonyl naphthalene carboxylic acid was obtained by the monosaponification reaction of 2,6-methyl naphthalene dicarboxylate and the formation of the chloride by the action of thionyl chloride. From the initial reactants, the 5,5,8,8-tetramethyl-5,6,7,8-tetrahydro naphthalene was as described in (17); the 1,1,2,3,3-pentamethylindane was prepared in accordance with the methods described in French Pat. # 1.392804. The 6-[(5,5,8,8-tetramethyl-5,6,7,8-tetrahydro-2-naph~yl) carbonyl] 2-methyl naphthalene carboxylate and the 6-[(1,1,2,3,3-pentamethyl-5-indanyl) carbonyl] 2-methyl naphthalene carboxylate were synthesized by condensation of bicyclic aromatic compound 5,5,8,8 tetramethyl-5,6,7,8 tetrahydronaphthalene and 1,1,2,3,3-pentamethylindane on 6-methoxycarbonyl 2naphthalene carboxylic acid chloride. (Prepared under conventional Friedel-Crafts reaction conditions). These keto-esters, by saponification, give the corresponding ketoacids 1 and 6, respectively. The keto-acid 1 is transformed into the amide 2 by ethylamine in the presence of N,N’carbonyldiimidaxole (CDI). The reduction of the keto-acid 1 by sodium botohydride in tetmhydrofumne provides the secondary alcohol 3. The reduction ,witb zinc, of the keto-acid 1 in acetic acid in the presence of HCl provides the acid 4. More details of the synthesis of these compounds have been described in U.S. Pat. 4, 826,969. The retinobenzoic acid compounds 8 and 9 (10) were synthesized and provided by GIRD Galderma (Sophia Antipolis, Valbonne, France). Stock solutions of 5x104 M in ethanol were prepared and kept under light protection at -2tYX All-trans retinoic acid was purchased from Sigma. Vectors : The expression vectors ER-RARa(pECE), ER-RARP(pECE), ER-RARy(pECE) and the reporter plasmid EREWAT were constructed as described (18). Cells : CV-1 cells were grown in Dulbecco’s modified Eagle’s medium, high glucose (Irvine Scientific , Irvine,CA) and 10% FCS (Tissue Culture Biologicals, Irvine,CA) treated with charcoal (12 mg/ml) to remove retinoids present in me serum. Transient transfections and CAT assays : Transient transfections of CV-1 cells were performed as described (19.29) with 25 ng expression vector, 100 ng reporter plasmid (EREtkCAT), 300 ng P-galactosidase plasmid @CH 110, Pharmacia), and 575 ng Bluescript (Stratagene). All retinoid dilutions were prepared in DME/lO% FCS(charcoal-treated) immediately before use. CAT assayswere controlled for conditions of linearity. CAT activities were normalized for transfection efficiency by the corresponding 8-galactosidase activity, the baseline value (activity in absence of tetinoid) was subtracted, and resulting activation values were expressed in percent of the highest level of activation by RA (10e6 M).
1555
Vol.
179,
No.
RESULTS
BIOCHEMICAL
3, 1991
AND
AND BIOPHYSICAL RESEARCH
COMMUNICATIONS
DISCUSSION
RAR subtype specificities
are encoded in the ligand
-binding
domain
Most RA analogsused( 1 to 5 in Fig. 1) were closely related derivatives of 6’-substituted naphthalene-2-carboxylic acid. The overall structure was defined by a mostly planar systemof rings limited in degreesof rotational freedom, thereby imposingsevereconformationalrestrictions in comparison to RA (the target bonds for 9-c&and ll-cis-rruns-isomerisation are lost ) and offering less structural determinants for biochemical degradation than does RA (20) which is metabolized extensively in viva (21 and referencesherein). We reasonedthat the rather subtle differences between the retinoids selectedhere were more likely to allow detection of differences between the binding pockets of RAR subtypesthan more extensively modified retinoids usedin previous investigations (13,16). The domain structure of nuclear receptors (2,22) allows construction of hybrid receptors (2,23) which combine the inducibility by one ligand with the DNA binding properties of another receptor. The advantageof a hybrid receptor-basedexpressionsystem lies in the independence from activation by endogenousRA receptors,found in many mammaliancell lines (24), which do not activate transcription from an estrogenresponsiveelement(ERR) (23).
6
8
‘m.
9
O ;I
1; Ii
tid”u-
Fie. 1 . Structure and chemical nomenclature of the retinoids employed. 1. 6-[(5,5,8,8teuamethyl-5.6,7,8-tetrahydro-2-naphthyl)carbonyl]2-naphthalene carboxylic acid. 2. N-ethyl-6-[(5,5,8,8tetramethyl-5,6,7,8-tetrahydro-2-naphthyl)carbonyl]2-naphthalene carboxamide. 3.6-[(5,5,8,8-tetramethyl-5,6,7,8teuahydm-2-naphthyl)hydroxymethyl]2+uphthalene carboxylic acid. 4. l-(5,5,8,8-teuamethyl-5,6,7,8-tetrahy&2naphthyI)l-(6-carboxy-2-naphthyl) methane. 5. 6-~(1,1,2,3.3-pentamethyl-5-indnyI~~bonyl]2-naph~~ene carboxylic acid. 6. all-trans - retinoic acid. 7. 13-cis -retinoic acid. 8. (Am80) 4-(5,6,7,8-tetrahydro-5,5,8,8tetmmethyl-2-naphthalenylcarbamoyl)benzoic acid. 9. (Am580) 4-(5,6,7,8-tetrahydro-5,5,8,8-tetramethyl-2naphthamido) benzoic acid
1556
Vol.
179,
No.
3, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
Y
9
10
8 - log WI
7
6
10
9
8
7
6
- logInAl (Ml
VW
Fie. Comparison of hybrid receptor and wild type receptor activation by RA. CV-I cells were transfected with (a) expression vectors for ER-RARa, ER-RARP, or ER-RARy and an ERE-CAT reporter. Cells were then grown in the presence of varying concentrations of RA as described in Materials and Methods. CAT activity was measured as described (28,42). Receptor activiation is expressed as percent of maximal induction (measured at 10W6 M RA) after constitutive receptor activity was subtracted (n=5). (b) wild type. receptor expression vectors were cotransfected with a retinoic acid responsive reporter (TRE2CAQ (29.42) otherwise cultures were treated as in (a).
Here we report (Fig. 2a,2b) that over a wide concentration range of RA the particular dose response of each individual RAR subtype (15,25,26) was conserved in hybrid constructs of the type ER-RAR
(18) which contain the estrogen receptor amino terminus and DNA binding domain
connected to the ligand binding domain of a retinoic acid receptor subtype. Both ER-RARP ER-RARy
responded at 10-m M RA where the ER-RARa
was still inactive. ER-RARa
and
and ER-
RARy showed the steepest increase of activity between 10-g and 10-7 M RA, ER-RARP
in the
range of 10m9to 10-7 M RA. In the absence of ligand (data not shown ), the constitutive activity of the ER-RARy
was
the highest (30), the activity of ER-RARo:
the lowest of the three receptors.
These results are in agreement with observations on a different type of ER-RAR
hybrid receptor
(5) where the lowest concentration of RA sufficient for activation of the RARa ligand binding domain was tenfold higher than for the RARP ligand binding domain. We can conclude that the ligand binding domain determines the transcriptional activation by these RA hybrid receptors and that these hybrid receptors by retinoids. Modifications receptor
define a suitable system for comparative studies on receptor activation
of the retinoid
structure
allow
selective
activation
subtypes Five derivatives of the 6’ substituted naphthalene-2-carboxylic
of retinoic
acid
acid were analyzed for their
receptor activation capacity. Surprisingly, several compounds exhibited strong differential receptor activation (Fig. 3a-b). Compound 1 was a very potent inducer of ER-RARP
and ER-RARy
over
the entire range of concentrations investigated, while it activated ER-RARo only at concentrations around 10-7 M and above, to maximally 60% of the corresponding RA value. Compound 5 showed similar characteristics, but lower overall activation. Compound 2 activated ER-RARB only to 20% of the corresponding RA value, but showed about 3-4 times better activity with ERRARy (Fig. 3b,c). Most interestingly, it failed completely to activate ER-RARa (Fig. 3a). Compound 4 shared intermediate activation levels with ER-RARB and slightly higher activation levels on ER-RARy with compound 5, but failed to significantly activate ER-RARa at 10-7 M and 1557
BIOCHEMICAL
Vol. 179, No. 3, 1991
10
9 -log
6 [RETlNOID]
7 (M)
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
6 - log [RETINOID]
- log [REllNOID]
(M)
(M)
Fie. Activation of ER-RARa (a), ER-RARP (b), and ER-RARy (c) by five different 6’-substituted naphthalene-2-carboxylic acid derivatives, normalized to the highest induction by all-trans -RA (10-6 M).Transfections and CAT assays were performed as described in Fig. 2. Retinoids 2,3 and 4 showed no activity at certain concentrations with ER-RARa. Therefore no values are given in the graph (n=5).
below. Compound 3 failed to activate ER-RARa as well; ER-RARP was induced only at concentrationshigher than lo-8 M, never exceeding 30-40s of the correspondingRA value. ERRARy wasinduced up to the maximal level. Our data identify a chemically homogeneousclass of retinoids, derivatives of the 6’substituted naphthalene-2-carboxylic acid, that show selective activation of the ligand-binding domain of retinoic acid receptor subtypes. Consistently, activation of ER-RARa is either absent over five ordersof magnitudein retinoid concentrationor can be achievedonly at concentrationsof 10-7 M or higher. In a separatestudy, we comparedreceptor selectivity of a larger number of retinoids and alsoobservedthat structurally different derivatives of the naphthalenecarboxylic acid classare p/r selective (30). The RARP and RAR-y selectivity of thesecompoundscould be related to the higher degree of sequencesimilarity between the ligand binding domains of RARP and RARy comparedto the ligand binding domain of RARa (2-6). At 10-7M, compound3, and to a lesserdegree2, appearto distinguish betweenall three receptor subtype ligand binding domains, showing no activation of ER-RARa but four fold higher activation of ER-RARy than of ERRARP. The consistent linkage of ER-RARP inducibility with ER-RARyindicates that the derivatisation haschangedthe receptor selectivity in comparisonto the maternalcompound13-cisRA (7 in Fig. 1) fundamentally, since high ED50 values for RARcl and RARy, but fivefold lower 1558
Vol.
179,
No.
BIOCHEMICAL
3, 1991
AND
BIOPHYSICAL
E
60-
gj
‘lo-
g
m-
t?d
RESEARCH
COMMUNICATIONS
1
*
8 0. A CT 6 - log [RETINOID]
s = E s :lJ
10
9
(M)
0
7
- log [RETINOID]
0 60-
6
(hi)
R* 9 8
40-
6 - log [RETINOID]
(M)
&g& Activation of ER-RARu (a), ER-RARP (b), and ER-RARy B.Transfections, CAT assays and data normalization as in Fig. 2.
ED50 for RARP,
(C)
by the retinobenzoic
have been reported for 13-cis -RA (15). Consequently,
acid analogs g and
we reasoned that
molecular determinants of a given retinoid for receptor subtype selectivity may be linked to specific modifications of the polyolefinic side chain in RA by aryl substituents. To further study the impact of aryl substituent bulkiness and cishrans
configuration of the
carboxyl function, we selected two biologically well characterized retinobenzoic acid analogs Am80 and Am580 (13) ( 8 and 9 in Fig. 1) where the naphthalene carboxylic acid ring in cis configuration
has been replaced by a carbamoyl benzoic acid ring (AmgO), or an isocarbamoyl
benzoic acid ring (Am580), in mans configuration.
Transcriptional activation assays (Fig. 4) also
revealed a certain receptor subtype selectivity with these retinoids, but with completely different characteristics. At 10-8 M, 8 and 9 were selective inducers of ER-RARa. Activation by 8 was only about 10% of the maximum, whereas the isocarbamoyl derivative 9, which is structurally more closely related to 1, induced 25% of the maximal value. 8 maintained a high selectivity for ER-RARa over ER-RARy at 10-7 M (one order of magnitude difference in transcriptional activation), but a lower selectivity versus ER-RARP (two- to fivefold higher activation of the a This concentration-dependent ER-RARa subtype selectivity was not over the p receptor). unexpected since data on binding of 9 to bacterially expressed RARa ligand binding domain had been suggestive for some RARa preference over RARP (27). 1559
Vol.
179, No. 3, 1991
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
Table 1. ED50 values of transcriptional activation by seven retinoids in comparison to RA. Values were determined graphically from induction curves ; accuracy is estimated to be within a factor of 2 --*:Due to the complete lack of transcriptional activation, as seen in Fig. 3a, no half-maximal induction value could be derived. ** For mrne compounds maximal activity was measured at 10-5 M.
ER-RARa
ER-RARP
ER-RARv
RA
3 x IO-8
7 x 10-9
2 x 10-9
1
7 x IO-7
6 x 10-9
6 x 10-9
2
ND*
3 x 10-e”
3 x lo-h**
3
ND*
3 x lo-h**
3x10-7
4
ND*
3 x 10-7
7 x 10-s
5
1 x 10-6”
7 x 10-9
2 x 10-s
10-s
5 x 10-7
2 x 10-6~’
5 x 10-s
3 x 10-7”
Am 80
7x
Am 580
5 x 10-s
A comprehensive listing of ED50 values for each retinoid on each receptor subtype investigated is presentedin Table 1. In particular, selectivity of ER-RARP (ER-RAR$ over ERRARol occurs over several orders of magnitude in retinoid concentration and is consistently correlated with differences in ED50. The gene activation data presentedhere most likely reflect direct interaction of retinoids with the receptors. However, it cannot be excluded that the overall activity of certain retinoids can be affected in certain cell types containing other retinoid binding protein (for instance CRABP) or modifying enzymes. But the presenceof suchproteins should not affect the receptor selectivity of a compound. Our discovery of receptor specific retinoids lendsitself to further application in studiesof the biological roles of RAR subtypesand, possibly,the designof therapeutically active, receptorselectiveretinoids. Most likely, compounds1 and 5 could elicit in vivo-activation of the inducible RARP while restraining the transcriptional responsefrom the ubiquitous and constitutive RARa. Selective receptor activation shouldalsocontribute to unraveling the hierarchy of receptor subtype activation . Retinoids are usedas efficient therapeutic agentsagainstvarious skin diseases(21). 13-cis -RA has been reported to reduce the incidence of secondprimary tumors in patients with squamouscell carcinoma of head and neck, if given in high doses(28). However, therapy by retinoids is hamperedby severeundesirablesideeffects. It can now be determinedwhetherthe 6’substitutednaphthalene-2-carboxylic acid analogsof 13-cis -RA , by their strongreceptor subtype selectivity, will show a different spectrumof biological responses,with fewer sideeffects. In that case,somemembersof this retinoid classmight turn out to be valuable candidatesfor improved cancertherapy.
OWLELIGMENTS We thank Uwe Reichert (GIRD Galderma) for the generous gifts of the Am80 and Am580 compounds. This work was supported by NIH grants DK 35083 and CA 50676 (to M.P.).
1560
Vol.
179,
No.
3, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
REFERENCES
1.1 2.1 3.) 4.) 5.) 6.) 7.)
8.) 9.) 10.) 11.) 12.) 13.) 14.) 15.) 16.) 17.) 18.) 19.) 20.) 21.) 22.) 23.) 24.) 25.) 26.) 27.) 28.)
29.) 30.)
Sherman, M.I.(Ed) (1986) Retinoids and Cell Differentiation, CRC Press, Boca Raton Benbrook, D., Lemhardt, E., and Pfahl, M. (1988) Nature m, 669 - 672 Giguere, V., Ong, E.S., Seguy. P., and Evans, R. (1987) Nature m, 624 - 629 Petkovich, M., Brand N.J., Krust, A., and Chambon, P. (1987) Nature m, 444 - 450 Brand, N., Petkovich, M., Krust, A., Chambon, P., de The, H., Marchio, A., Tiollais, P., and Dejean, A. (1988) Nature m, 850 - 853 Krust, A., Kastner, P., Petkovich, M., Zelent, A., and Chambon, P. (1989) Proc. Natl. Acad. Sci. USA. &, 5310 - 5314 Zelent, A., Mendelsohn, C., Kastner, P., Krust, A., Gamier, J-M., Ruffenach, F., Leroy, P., and Chambon, P. (1991) EMBO J. fi,71- 81 Leroy, P., Krust, A., Relent, A., Mendelsohn, C., Gamier, J-M., Kasmer, P., Dierich, A.,and Chambon, P. (1991) EMBO J. 111.59 - 69 Loeliger, P., Bollag, W., and Mayer, H.(1980) Eur. J. Med. Chem. - Chim. Ther. fi, 9 - 15 Kagechika, H., Kawachi, E., Hashimoto, Y., Himl, T., and Shudo, K. (1988) J. Med. Chem. a,2182 2192 Breitman, T., Selonick, S., and Collins, S. (1980) Proc. Natl. Acad. Sci. USA. =2936-2940 Dotter, D., and Koeffler, H.P. (1982) J. Clin. Invest. &, 277 - 283 Jetten, A.M., Anderson, K., Deas, M.A., Kagechika, H., Lotan, R., Rearick, J.L., and Shudo, K. (1987) Cancer Res. 47.3523 - 3527 Darmon, M., Rocher, M.K., Carey, M.T., Martin, B., Rabillaud, T., DelescluseC, and Shroot, B. (1988) Skin Pharmacology 1,161 - 175 Astrom, A., Pettersson, U., Krust, A., Chambon, P., and Vorhees, J.J. (1990) B&hem. Biophys. Res. Comm. m, 339 - 345 Clifford, J.L., Petkovich, M., Chambon, P., and Lotan, R. (1990) Mol. Endocrinol. u, 1546 - 1555 Brunson, H.Q., and Kroger, J.W.(1940) J. Am. Chem. Sot. &&36 - 44 Pfahl, M., Tzukerman, M., Zhang, X.-k., Lehmann, J., Hermann, T, Wills, K., and Graupner, G. (1990) Methods Enzymol. m, 256 - 270 Graupner, G., Zhang, X-k., Tzukerman, M., Wills, K., Hermann, T., and Pfahl, M. (1991) Mol. Endocrinol. 5,365 - 372 Vane, F.M., and Bugge C.J.L. (1981) Drug Metab. Dispos. 9,515 - 520 Chytil, F. (1986) J. Am. Acad. Dermatol. 15,741 - 747 Evans, R.M. (1988) Science m, 889 - 895 Graupner, G., Wills, K.N., Tzukerman, M., Zhang, X-k., and Pfahl, M. (1989) Nature &@, 653 - 656 Hoffmann, B., Lehmann, J.M., Zhang, X-k., Hermann, T., Husmann, M., and Pfahl, M. (1990) Mol. Endocrinol. 4, 1734 - 1743 Giguere, V., Shago, M., Zimgibl, R.,Tate, P.,Rossant, J., and Varmuza, S. (1990) Mol. Cell. Biol. m, 2335 - 2340 Lehmann, J.M., Hoffmann, B., and Pfahl, M. (1991) Nucl. Acid Res. 2,573 - 578 Crettaz, M., Baron, A., Siegenthaler, G., and Hunziker, W. (1990) Biochem. J. 272,391 - 397 Hong, W.K., Lippman, S.M., Itri, L.M., Karp, D.D.,Lee, J.S., Byers, R.M., Schantz, S.P., Kramer, A.M., Lotan, R., Peters, L.J., Dimery, I.W., Brown, B.W., and Goepfert, H. (1990) New Engl. J. Med. m, 795 - 801 Husmann, M, Lehmann, J., Hoffman, B., Hermann, T., Tzukerman, M., and Pfahl, M. (1991) Mol. Cell. Biol. 11, 4097-4103. Lehmann, J., Dawson, MI., Hobbs, P.D., Husmann, M., and Pfahl, M. (1991) Cancer Res. 5-l 48044809.
1561
AND BIOPHYSICAL RESEARCH COMMUNKATIONS Pages 1554-1561
BIOCHEMICAL
6’-SUBSTITUTED NAPHTHALENE-2-CARBOXYLIC ACID ANALOGS, A NEW CLASS OF RETINOIC ACID RECEPTOR SUBTYPE-SPECIFIC LIGANDS Gerhart Graupner, Gerard Malle*, Jean Maignan*, Gerard Lang*, Michel Prunieras* and Magnus Pfahl+ Cancer Center, La Jolla Cancer Research Foundation, 10901 North Torrey Pines Road, La Jolla, CA 92037 * Laboratoire
Received
July
The biological
22,
1991
response
by three nuclear as receptor Hybrid
de Recherche Fondamentale de L’Greal, 1 Avenue Eugene Schueller, 93600 Aulnay-sous-Bois, France
retinoic
subtype
receptors
receptors
ER-RARa, wild
of Q’-substituted
retinoic
type receptors
while
no activation
analogs
showed
a limited
retinoic
acid receptor
exhibited
strong
was observed.
subtype selectivity
can be defined
applications.
activation
different
transcriptional
Conversely,
assay.
measuring
acid, the three hybrid
the same induction
RARy. Three
We conclude
and for therapeutical
of retinoids
used to avoid
in response to retinoic
selectivity.
a
element
acid, elicited
of a receptor
the potential
were
RARp,and
naphthalene-2-carboxylic
is mediated
response
ER-RART
RARa,
TO explore
(retinoids)
a transcriptional
acid receptors. and
derivatives
we employed
an estrogen
ER-RARP
receptor
differentiation
RARa, p and 7.
activators,
that recognize
of endogenous
acid (RA) and synthetic
acid receptors,
selective
activities
corresponding
to retinoic
profile
retinoids,
as the analogs
activation
of y
two retinobenzoic
acid
that retinoids
with
unique
and tested for
their
impact on cellular
0 1991 Academic
profiles
of
Press, Inc.
Retinoic acid (RA) has been recognized as a pivotal regulatory compound in cell differentiation,proliferation and malignant transformation (1). These effects are mediated by nuclear receptors (RARa,RAR/3,RARy) which are members of the steroid/thyroid receptor family (2,3,4,5,6) and show tissue-specific expression patterns during embryogenesis and in adult life. The RARs am encoded by three different genes from which several isoforms can be derived ; these + To whom correspondence should be addressed . 0006-291X/91 Copyright All rights
$1.50
0 1991 by Academic Press, Inc. of reproduction in any form reserved.
1554
Vol.
179,
No.
BIOCHEMICAL
3, 1991
AND
isoforms appear to be regulated by tissue-specific
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
expression as well (7,8).-RA analogs (retinoids)
(9,lO) have long been employed to gain insight into molecular
parameters governing
seemingly unrelated properties as chemoprevention of tumorigenesis, and malformation. showing
differentiation
in the absence of cytoplasmic
such
Studies
retinoic acid binding protein (CRABP)
(11,12) and a lack of correlation between differentiation and retinoid CRABP binding (13) led to replacement of the initial concept of cytoplasmic proteins as the mediators of RA activity by the now established mechanism of action through nuclear receptors. Binding of synthetic retinoids to nuclear receptors has been correlated to biological activity (14), while other studies have shown transcriptional
activation by retinoids (15) and the ability to modulate RAR mRNA expression in
melanoma cells (16). Still, no significant differential transcriptional activation of RAR subtypes by synthetic retinoids has been demonstrated so far. Here we describe a class of 13-&r -RA analogs, derivatives
of the 6’-substituted
naphthalene-2-carboxylic acid, and present evidence that several of these analogs can preferentially induce hybrid receptors bearing the ligand binding domain of RAR subtypes. In comparison, two retinobenzoic acid derivatives (Am80 and Am580), representatives of a distinct class of all-rrunsRA analogs (13), show a different profile of RAR subtype selectivity.
Our data indicate that
retinoids can be synthesized that may govern specific biological functions by virtue of their RAR subtype specificity.
MATERIALS
AND METHODS
Retinoids : synthesis: The chloride of 6-methoxycarbonyl naphthalene carboxylic acid was obtained by the monosaponification reaction of 2,6-methyl naphthalene dicarboxylate and the formation of the chloride by the action of thionyl chloride. From the initial reactants, the 5,5,8,8-tetramethyl-5,6,7,8-tetrahydro naphthalene was as described in (17); the 1,1,2,3,3-pentamethylindane was prepared in accordance with the methods described in French Pat. # 1.392804. The 6-[(5,5,8,8-tetramethyl-5,6,7,8-tetrahydro-2-naph~yl) carbonyl] 2-methyl naphthalene carboxylate and the 6-[(1,1,2,3,3-pentamethyl-5-indanyl) carbonyl] 2-methyl naphthalene carboxylate were synthesized by condensation of bicyclic aromatic compound 5,5,8,8 tetramethyl-5,6,7,8 tetrahydronaphthalene and 1,1,2,3,3-pentamethylindane on 6-methoxycarbonyl 2naphthalene carboxylic acid chloride. (Prepared under conventional Friedel-Crafts reaction conditions). These keto-esters, by saponification, give the corresponding ketoacids 1 and 6, respectively. The keto-acid 1 is transformed into the amide 2 by ethylamine in the presence of N,N’carbonyldiimidaxole (CDI). The reduction of the keto-acid 1 by sodium botohydride in tetmhydrofumne provides the secondary alcohol 3. The reduction ,witb zinc, of the keto-acid 1 in acetic acid in the presence of HCl provides the acid 4. More details of the synthesis of these compounds have been described in U.S. Pat. 4, 826,969. The retinobenzoic acid compounds 8 and 9 (10) were synthesized and provided by GIRD Galderma (Sophia Antipolis, Valbonne, France). Stock solutions of 5x104 M in ethanol were prepared and kept under light protection at -2tYX All-trans retinoic acid was purchased from Sigma. Vectors : The expression vectors ER-RARa(pECE), ER-RARP(pECE), ER-RARy(pECE) and the reporter plasmid EREWAT were constructed as described (18). Cells : CV-1 cells were grown in Dulbecco’s modified Eagle’s medium, high glucose (Irvine Scientific , Irvine,CA) and 10% FCS (Tissue Culture Biologicals, Irvine,CA) treated with charcoal (12 mg/ml) to remove retinoids present in me serum. Transient transfections and CAT assays : Transient transfections of CV-1 cells were performed as described (19.29) with 25 ng expression vector, 100 ng reporter plasmid (EREtkCAT), 300 ng P-galactosidase plasmid @CH 110, Pharmacia), and 575 ng Bluescript (Stratagene). All retinoid dilutions were prepared in DME/lO% FCS(charcoal-treated) immediately before use. CAT assayswere controlled for conditions of linearity. CAT activities were normalized for transfection efficiency by the corresponding 8-galactosidase activity, the baseline value (activity in absence of tetinoid) was subtracted, and resulting activation values were expressed in percent of the highest level of activation by RA (10e6 M).
1555
Vol.
179,
No.
RESULTS
BIOCHEMICAL
3, 1991
AND
AND BIOPHYSICAL RESEARCH
COMMUNICATIONS
DISCUSSION
RAR subtype specificities
are encoded in the ligand
-binding
domain
Most RA analogsused( 1 to 5 in Fig. 1) were closely related derivatives of 6’-substituted naphthalene-2-carboxylic acid. The overall structure was defined by a mostly planar systemof rings limited in degreesof rotational freedom, thereby imposingsevereconformationalrestrictions in comparison to RA (the target bonds for 9-c&and ll-cis-rruns-isomerisation are lost ) and offering less structural determinants for biochemical degradation than does RA (20) which is metabolized extensively in viva (21 and referencesherein). We reasonedthat the rather subtle differences between the retinoids selectedhere were more likely to allow detection of differences between the binding pockets of RAR subtypesthan more extensively modified retinoids usedin previous investigations (13,16). The domain structure of nuclear receptors (2,22) allows construction of hybrid receptors (2,23) which combine the inducibility by one ligand with the DNA binding properties of another receptor. The advantageof a hybrid receptor-basedexpressionsystem lies in the independence from activation by endogenousRA receptors,found in many mammaliancell lines (24), which do not activate transcription from an estrogenresponsiveelement(ERR) (23).
6
8
‘m.
9
O ;I
1; Ii
tid”u-
Fie. 1 . Structure and chemical nomenclature of the retinoids employed. 1. 6-[(5,5,8,8teuamethyl-5.6,7,8-tetrahydro-2-naphthyl)carbonyl]2-naphthalene carboxylic acid. 2. N-ethyl-6-[(5,5,8,8tetramethyl-5,6,7,8-tetrahydro-2-naphthyl)carbonyl]2-naphthalene carboxamide. 3.6-[(5,5,8,8-tetramethyl-5,6,7,8teuahydm-2-naphthyl)hydroxymethyl]2+uphthalene carboxylic acid. 4. l-(5,5,8,8-teuamethyl-5,6,7,8-tetrahy&2naphthyI)l-(6-carboxy-2-naphthyl) methane. 5. 6-~(1,1,2,3.3-pentamethyl-5-indnyI~~bonyl]2-naph~~ene carboxylic acid. 6. all-trans - retinoic acid. 7. 13-cis -retinoic acid. 8. (Am80) 4-(5,6,7,8-tetrahydro-5,5,8,8tetmmethyl-2-naphthalenylcarbamoyl)benzoic acid. 9. (Am580) 4-(5,6,7,8-tetrahydro-5,5,8,8-tetramethyl-2naphthamido) benzoic acid
1556
Vol.
179,
No.
3, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
Y
9
10
8 - log WI
7
6
10
9
8
7
6
- logInAl (Ml
VW
Fie. Comparison of hybrid receptor and wild type receptor activation by RA. CV-I cells were transfected with (a) expression vectors for ER-RARa, ER-RARP, or ER-RARy and an ERE-CAT reporter. Cells were then grown in the presence of varying concentrations of RA as described in Materials and Methods. CAT activity was measured as described (28,42). Receptor activiation is expressed as percent of maximal induction (measured at 10W6 M RA) after constitutive receptor activity was subtracted (n=5). (b) wild type. receptor expression vectors were cotransfected with a retinoic acid responsive reporter (TRE2CAQ (29.42) otherwise cultures were treated as in (a).
Here we report (Fig. 2a,2b) that over a wide concentration range of RA the particular dose response of each individual RAR subtype (15,25,26) was conserved in hybrid constructs of the type ER-RAR
(18) which contain the estrogen receptor amino terminus and DNA binding domain
connected to the ligand binding domain of a retinoic acid receptor subtype. Both ER-RARP ER-RARy
responded at 10-m M RA where the ER-RARa
was still inactive. ER-RARa
and
and ER-
RARy showed the steepest increase of activity between 10-g and 10-7 M RA, ER-RARP
in the
range of 10m9to 10-7 M RA. In the absence of ligand (data not shown ), the constitutive activity of the ER-RARy
was
the highest (30), the activity of ER-RARo:
the lowest of the three receptors.
These results are in agreement with observations on a different type of ER-RAR
hybrid receptor
(5) where the lowest concentration of RA sufficient for activation of the RARa ligand binding domain was tenfold higher than for the RARP ligand binding domain. We can conclude that the ligand binding domain determines the transcriptional activation by these RA hybrid receptors and that these hybrid receptors by retinoids. Modifications receptor
define a suitable system for comparative studies on receptor activation
of the retinoid
structure
allow
selective
activation
subtypes Five derivatives of the 6’ substituted naphthalene-2-carboxylic
of retinoic
acid
acid were analyzed for their
receptor activation capacity. Surprisingly, several compounds exhibited strong differential receptor activation (Fig. 3a-b). Compound 1 was a very potent inducer of ER-RARP
and ER-RARy
over
the entire range of concentrations investigated, while it activated ER-RARo only at concentrations around 10-7 M and above, to maximally 60% of the corresponding RA value. Compound 5 showed similar characteristics, but lower overall activation. Compound 2 activated ER-RARB only to 20% of the corresponding RA value, but showed about 3-4 times better activity with ERRARy (Fig. 3b,c). Most interestingly, it failed completely to activate ER-RARa (Fig. 3a). Compound 4 shared intermediate activation levels with ER-RARB and slightly higher activation levels on ER-RARy with compound 5, but failed to significantly activate ER-RARa at 10-7 M and 1557
BIOCHEMICAL
Vol. 179, No. 3, 1991
10
9 -log
6 [RETlNOID]
7 (M)
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
6 - log [RETINOID]
- log [REllNOID]
(M)
(M)
Fie. Activation of ER-RARa (a), ER-RARP (b), and ER-RARy (c) by five different 6’-substituted naphthalene-2-carboxylic acid derivatives, normalized to the highest induction by all-trans -RA (10-6 M).Transfections and CAT assays were performed as described in Fig. 2. Retinoids 2,3 and 4 showed no activity at certain concentrations with ER-RARa. Therefore no values are given in the graph (n=5).
below. Compound 3 failed to activate ER-RARa as well; ER-RARP was induced only at concentrationshigher than lo-8 M, never exceeding 30-40s of the correspondingRA value. ERRARy wasinduced up to the maximal level. Our data identify a chemically homogeneousclass of retinoids, derivatives of the 6’substituted naphthalene-2-carboxylic acid, that show selective activation of the ligand-binding domain of retinoic acid receptor subtypes. Consistently, activation of ER-RARa is either absent over five ordersof magnitudein retinoid concentrationor can be achievedonly at concentrationsof 10-7 M or higher. In a separatestudy, we comparedreceptor selectivity of a larger number of retinoids and alsoobservedthat structurally different derivatives of the naphthalenecarboxylic acid classare p/r selective (30). The RARP and RAR-y selectivity of thesecompoundscould be related to the higher degree of sequencesimilarity between the ligand binding domains of RARP and RARy comparedto the ligand binding domain of RARa (2-6). At 10-7M, compound3, and to a lesserdegree2, appearto distinguish betweenall three receptor subtype ligand binding domains, showing no activation of ER-RARa but four fold higher activation of ER-RARy than of ERRARP. The consistent linkage of ER-RARP inducibility with ER-RARyindicates that the derivatisation haschangedthe receptor selectivity in comparisonto the maternalcompound13-cisRA (7 in Fig. 1) fundamentally, since high ED50 values for RARcl and RARy, but fivefold lower 1558
Vol.
179,
No.
BIOCHEMICAL
3, 1991
AND
BIOPHYSICAL
E
60-
gj
‘lo-
g
m-
t?d
RESEARCH
COMMUNICATIONS
1
*
8 0. A CT 6 - log [RETINOID]
s = E s :lJ
10
9
(M)
0
7
- log [RETINOID]
0 60-
6
(hi)
R* 9 8
40-
6 - log [RETINOID]
(M)
&g& Activation of ER-RARu (a), ER-RARP (b), and ER-RARy B.Transfections, CAT assays and data normalization as in Fig. 2.
ED50 for RARP,
(C)
by the retinobenzoic
have been reported for 13-cis -RA (15). Consequently,
acid analogs g and
we reasoned that
molecular determinants of a given retinoid for receptor subtype selectivity may be linked to specific modifications of the polyolefinic side chain in RA by aryl substituents. To further study the impact of aryl substituent bulkiness and cishrans
configuration of the
carboxyl function, we selected two biologically well characterized retinobenzoic acid analogs Am80 and Am580 (13) ( 8 and 9 in Fig. 1) where the naphthalene carboxylic acid ring in cis configuration
has been replaced by a carbamoyl benzoic acid ring (AmgO), or an isocarbamoyl
benzoic acid ring (Am580), in mans configuration.
Transcriptional activation assays (Fig. 4) also
revealed a certain receptor subtype selectivity with these retinoids, but with completely different characteristics. At 10-8 M, 8 and 9 were selective inducers of ER-RARa. Activation by 8 was only about 10% of the maximum, whereas the isocarbamoyl derivative 9, which is structurally more closely related to 1, induced 25% of the maximal value. 8 maintained a high selectivity for ER-RARa over ER-RARy at 10-7 M (one order of magnitude difference in transcriptional activation), but a lower selectivity versus ER-RARP (two- to fivefold higher activation of the a This concentration-dependent ER-RARa subtype selectivity was not over the p receptor). unexpected since data on binding of 9 to bacterially expressed RARa ligand binding domain had been suggestive for some RARa preference over RARP (27). 1559
Vol.
179, No. 3, 1991
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
Table 1. ED50 values of transcriptional activation by seven retinoids in comparison to RA. Values were determined graphically from induction curves ; accuracy is estimated to be within a factor of 2 --*:Due to the complete lack of transcriptional activation, as seen in Fig. 3a, no half-maximal induction value could be derived. ** For mrne compounds maximal activity was measured at 10-5 M.
ER-RARa
ER-RARP
ER-RARv
RA
3 x IO-8
7 x 10-9
2 x 10-9
1
7 x IO-7
6 x 10-9
6 x 10-9
2
ND*
3 x 10-e”
3 x lo-h**
3
ND*
3 x lo-h**
3x10-7
4
ND*
3 x 10-7
7 x 10-s
5
1 x 10-6”
7 x 10-9
2 x 10-s
10-s
5 x 10-7
2 x 10-6~’
5 x 10-s
3 x 10-7”
Am 80
7x
Am 580
5 x 10-s
A comprehensive listing of ED50 values for each retinoid on each receptor subtype investigated is presentedin Table 1. In particular, selectivity of ER-RARP (ER-RAR$ over ERRARol occurs over several orders of magnitude in retinoid concentration and is consistently correlated with differences in ED50. The gene activation data presentedhere most likely reflect direct interaction of retinoids with the receptors. However, it cannot be excluded that the overall activity of certain retinoids can be affected in certain cell types containing other retinoid binding protein (for instance CRABP) or modifying enzymes. But the presenceof suchproteins should not affect the receptor selectivity of a compound. Our discovery of receptor specific retinoids lendsitself to further application in studiesof the biological roles of RAR subtypesand, possibly,the designof therapeutically active, receptorselectiveretinoids. Most likely, compounds1 and 5 could elicit in vivo-activation of the inducible RARP while restraining the transcriptional responsefrom the ubiquitous and constitutive RARa. Selective receptor activation shouldalsocontribute to unraveling the hierarchy of receptor subtype activation . Retinoids are usedas efficient therapeutic agentsagainstvarious skin diseases(21). 13-cis -RA has been reported to reduce the incidence of secondprimary tumors in patients with squamouscell carcinoma of head and neck, if given in high doses(28). However, therapy by retinoids is hamperedby severeundesirablesideeffects. It can now be determinedwhetherthe 6’substitutednaphthalene-2-carboxylic acid analogsof 13-cis -RA , by their strongreceptor subtype selectivity, will show a different spectrumof biological responses,with fewer sideeffects. In that case,somemembersof this retinoid classmight turn out to be valuable candidatesfor improved cancertherapy.
OWLELIGMENTS We thank Uwe Reichert (GIRD Galderma) for the generous gifts of the Am80 and Am580 compounds. This work was supported by NIH grants DK 35083 and CA 50676 (to M.P.).
1560
Vol.
179,
No.
3, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
REFERENCES
1.1 2.1 3.) 4.) 5.) 6.) 7.)
8.) 9.) 10.) 11.) 12.) 13.) 14.) 15.) 16.) 17.) 18.) 19.) 20.) 21.) 22.) 23.) 24.) 25.) 26.) 27.) 28.)
29.) 30.)
Sherman, M.I.(Ed) (1986) Retinoids and Cell Differentiation, CRC Press, Boca Raton Benbrook, D., Lemhardt, E., and Pfahl, M. (1988) Nature m, 669 - 672 Giguere, V., Ong, E.S., Seguy. P., and Evans, R. (1987) Nature m, 624 - 629 Petkovich, M., Brand N.J., Krust, A., and Chambon, P. (1987) Nature m, 444 - 450 Brand, N., Petkovich, M., Krust, A., Chambon, P., de The, H., Marchio, A., Tiollais, P., and Dejean, A. (1988) Nature m, 850 - 853 Krust, A., Kastner, P., Petkovich, M., Zelent, A., and Chambon, P. (1989) Proc. Natl. Acad. Sci. USA. &, 5310 - 5314 Zelent, A., Mendelsohn, C., Kastner, P., Krust, A., Gamier, J-M., Ruffenach, F., Leroy, P., and Chambon, P. (1991) EMBO J. fi,71- 81 Leroy, P., Krust, A., Relent, A., Mendelsohn, C., Gamier, J-M., Kasmer, P., Dierich, A.,and Chambon, P. (1991) EMBO J. 111.59 - 69 Loeliger, P., Bollag, W., and Mayer, H.(1980) Eur. J. Med. Chem. - Chim. Ther. fi, 9 - 15 Kagechika, H., Kawachi, E., Hashimoto, Y., Himl, T., and Shudo, K. (1988) J. Med. Chem. a,2182 2192 Breitman, T., Selonick, S., and Collins, S. (1980) Proc. Natl. Acad. Sci. USA. =2936-2940 Dotter, D., and Koeffler, H.P. (1982) J. Clin. Invest. &, 277 - 283 Jetten, A.M., Anderson, K., Deas, M.A., Kagechika, H., Lotan, R., Rearick, J.L., and Shudo, K. (1987) Cancer Res. 47.3523 - 3527 Darmon, M., Rocher, M.K., Carey, M.T., Martin, B., Rabillaud, T., DelescluseC, and Shroot, B. (1988) Skin Pharmacology 1,161 - 175 Astrom, A., Pettersson, U., Krust, A., Chambon, P., and Vorhees, J.J. (1990) B&hem. Biophys. Res. Comm. m, 339 - 345 Clifford, J.L., Petkovich, M., Chambon, P., and Lotan, R. (1990) Mol. Endocrinol. u, 1546 - 1555 Brunson, H.Q., and Kroger, J.W.(1940) J. Am. Chem. Sot. &&36 - 44 Pfahl, M., Tzukerman, M., Zhang, X.-k., Lehmann, J., Hermann, T, Wills, K., and Graupner, G. (1990) Methods Enzymol. m, 256 - 270 Graupner, G., Zhang, X-k., Tzukerman, M., Wills, K., Hermann, T., and Pfahl, M. (1991) Mol. Endocrinol. 5,365 - 372 Vane, F.M., and Bugge C.J.L. (1981) Drug Metab. Dispos. 9,515 - 520 Chytil, F. (1986) J. Am. Acad. Dermatol. 15,741 - 747 Evans, R.M. (1988) Science m, 889 - 895 Graupner, G., Wills, K.N., Tzukerman, M., Zhang, X-k., and Pfahl, M. (1989) Nature &@, 653 - 656 Hoffmann, B., Lehmann, J.M., Zhang, X-k., Hermann, T., Husmann, M., and Pfahl, M. (1990) Mol. Endocrinol. 4, 1734 - 1743 Giguere, V., Shago, M., Zimgibl, R.,Tate, P.,Rossant, J., and Varmuza, S. (1990) Mol. Cell. Biol. m, 2335 - 2340 Lehmann, J.M., Hoffmann, B., and Pfahl, M. (1991) Nucl. Acid Res. 2,573 - 578 Crettaz, M., Baron, A., Siegenthaler, G., and Hunziker, W. (1990) Biochem. J. 272,391 - 397 Hong, W.K., Lippman, S.M., Itri, L.M., Karp, D.D.,Lee, J.S., Byers, R.M., Schantz, S.P., Kramer, A.M., Lotan, R., Peters, L.J., Dimery, I.W., Brown, B.W., and Goepfert, H. (1990) New Engl. J. Med. m, 795 - 801 Husmann, M, Lehmann, J., Hoffman, B., Hermann, T., Tzukerman, M., and Pfahl, M. (1991) Mol. Cell. Biol. 11, 4097-4103. Lehmann, J., Dawson, MI., Hobbs, P.D., Husmann, M., and Pfahl, M. (1991) Cancer Res. 5-l 48044809.
1561
