Proc. Natl. Acad. Sci. USAVol. 90, pp. 9601-9605, October 1993Developmental Biology

Loss of retinoic acid receptor y function in F9 cells by genedisruption results in aberrant Hoxa-1 expression anddifferentiation upon retinoic acid treatment

(embryonal carcinoma cells/gene targeting/homeobox gene expression)

JOHN F. BOYLAN*, DAVID LOHNESt, RESHMA TANEJAt, PIERRE CHAMBONt, AND LORRAINE J. GUDAS*t*Department of Pharmacology, Cornell University Medical College, New York, NY 10021; and tLaboratoire de Genetique Moleculaire des Eucaryotes duCentre National de la Recherche Scientifique, Unite 184 de Biologie Moleculaire et de Genie Genetique de l'Institut National de la Sante et de laRecherche Mddicale, Institut de Chimie Biologique, Faculte de Medecine, 11 rue Humann, 67085 Strasbourg Cedex, France

Contributed by Pierre Chambon, July 26, 1993

ABSTRACT Retinoic acid (RA) signal transduction is be-lieved to be mediated through several high-affinity nuclearreceptors [RA receptors (RARs) and retinoid X receptors],which are members of the steroid/thyroid/vitamin D super-family and function as transcription factors. Why multipleRARs exist and what gene targets are regulated by each of thethree receptors remain compelling questions in developmentalbiology. Through targeted disruption of both RARy alleles, wehave identified several differentiation-specific genes that areregulated either directly or indirectly by RARY in F9 embry-onal carcinoma cells. These include genes encoding Hoxa-1(Hox-1.6) and the extracellular matrix proteins laminin B1 andcollagen type IV (acl), all of which are RA inducible inwild-type F9 embryonal carcinoma cells but are not signifi-cantly induced in the RARy-/- lines. In contrast, transcriptsencoding Hoxb-1 (Hox-2.9) and cellular RA binding protein H(CRABPII) are activated by RA for a longer period of time inthe RARy-/- lines compared to the wild-type F9 line. Not allRA-responsive genes are aberrantly expressed; Rex-i, RARI3,and SPARC transcripts are regulated in the RARy-/- lines asthey are in F9 wild-type cells. Our results support the idea thateach RAR may regulate different subsets of RA-responsivegenes, which may explain, in part, the complex regulation ofdevelopmental processes by retinoids.

All-trans-retinoic acid (RA), one of the most potent naturalretinoids, is both an important signaling molecule in embry-onic development and cell differentiation and a useful drugfor the treatment of several types of cancers (1-3). Twoclasses of molecules are known to modulate the actions ofRA. The first class is composed of the cellular RA bindingproteins I (CRABPI) and II (CRABPII), which are small,cytoplasmic, high-affinity RA binding molecules (4-7). Thesecond class consists of the highly conserved nuclear recep-tors-the RA receptors (RARs) a, /3, and fy and the retinoidX receptors (RXRs) a, /3, and y-that function as ligand-inducible transcription factors through the formation of het-erodimers bound to specific RA response elements (RAREs)(refs. 1 and 8-10 and references therein). Both all-trans-RAand 9-cis-RA are effective ligands for the RARs, while theRXRs bind only 9-cis-RA with high affinity (refs. 9-11 andreferences therein). The existence of multiple RARs sug-gested that each may perform a unique function (8). This ideawas supported by the observation that in the developingmouse embryo each RAR exhibits distinct spatiotemporalexpression patterns (12, 13).Thus, one of the important questions in the field of devel-

opmental biology concerns the roles that each of the three

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

RARs and RXRs may play in transducing the RA signal.Since murine F9 embryonal carcinoma cells, when treatedwith RA, differentiate into endoderm cells resembling thoseof the mouse blastocyst, these cells represent an attractivemodel system in which to study the early events of mamma-lian development and retinoid signaling. Elimination of theRARy protein via targeted disruption ofboth RARy alleles inF9 cells affords the opportunity to study the function of theprotein. In this report, we demonstrate that F9 stem cellscarrying such a disruption fail to exhibit a normal differen-tiation morphology upon treatment with RA. At the molec-ular level, this abnormal differentiation is reflected by thealtered expression of the transcripts of several differentia-tion-specific genes, including the Hox genes Hoxa-J (Hox-1.6) and Hoxb-l (Hox-2.9), the CRABPII gene, and the genesencoding the extraceilular matrix proteins laminin Bi andcollagen type IV (al). However, Rex-i and SPARC mRNAlevels remain correctly regulated by RA in the RARy-/-lines, showing that not all differentiation-specific genes areaffected. This work identifies RA target genes that appear tobe specifically regulated by one of the RARs.

MATERIALS AND METHODSCell Culture and Generation of Disrupted Lines. The RARy

disruption vector -)6.1 is described in detail elsewhere (14).Wild-type F9 cells (F9-Wt) were cultured under standardconditions (6); 1 x 107 cells per 500 ,ul suspended in electro-poration buffer plus 20 pg of linearized y6.1 plasmid wereelectroporated using a Bio-Rad gene pulser set at 200 V, 960,uF (15). Cells were plated at a density of 1 x 106 cells per150-cm2 tissue culture plate for 36 h followed by selection[G418 (300 ,g/ml active drug) plus 250 ,uM ganciclovir] for18-21 days. Individual colonies were isolated and propa-gated, and an aliquot was used for genomic Southern blotanalysis. The second allele was targeted by growing thesingle-copy disruption line F9-Wt-y-119 in high levels ofG418(2 mg/ml) for 21 days (15). Individual colonies were har-vested and analyzed as described above.

Differentiation of F9 Stem Cells. F9-Wt, F9-Wt-y-119, F9-Wt-y-119-14, F9-Wt-y-119-16, and F9-Wt-y-119-17 stem cellswere cultured and treated with 1 ,uM all-trans-RA with orwithout dibutyryl cAMP and theophylline as described (4, 6).No differences in doubling times were noted among the fivecell lines in the presence or absence ofRA (data not shown).For Northern blot analysis, the same stem cell lines weretreated with 1 ,uM all-trans-RA, and RNA was prepared at theindicated times. Southern and Northern blot analyses wereperformed as described elsewhere (6).

Abbreviations: RA, retinoic acid; RAR, RA receptor; RXR, retinoidX receptor; RARE, RA response element; wt, wild type.*To whom reprint requests should be addressed.

9601

Dow

nloa

ded

by g

uest

on

Oct

ober

21,

202

1

9602 Developmental Biology: Boylan et al.

Gel Mobility-Shift Assay. The expression vectors for themouse RAR coding sequences RARal (16, 17), RAR,B2 (16),and RARyl (16, 18) in pSG5 (19) have been described. Cellswere transfected with 10 ug of the murine RAR expressionvector plus 10 ,ug of carrier plasmid by the calcium phosphatetechnique (20). Cells were grown in the presence of 1 ,uMall-trans-RA for 24 h. Whole-cell extracts from cultures ofF9-Wt and F9-Wt-y-119-14 cells were prepared as reported(21). Mobility-shift assays were performed as described byGamer and Revzin (22) using the Hoxa-J/RAREf3 double-stranded oligodeoxynucleotide corresponding to the RAREof the Hoxa-l and RAR/32 genes (23, 24) as described byNicholson et al. (25). Protein extracts (4 ,ug) were preincu-bated with 50,000 cpm of 32P-labeled oligonucleotide probefollowed by addition of 1 ,u of ascites fluid monoclonalantibodies directed against the F region ofRARal, -(32, or -yl(21, 26, 27). The protein complexes were resolved on a 5%polyacrylamide gel.

RESULTSThe genomic sequence encoding the B domain of the RARyprotein was targeted with a replacement construct containing

Proc. Natl. Acad. Sci. USA 90 (1993)

6 kb of genomic sequence interrupted by the neomycin-resistance gene (Fig. la; ref. 14). Homologous recombinationwas detected by genomic Southern blotting with a genomicprobe flanking the targeting construct. This probe gives riseto a 4.5-kb BamHI genomic fragment specific for the target-ing event, compared to the wild-type (wt) 6-kb BamHIfragment (Fig. 1 a and b). Fig. lb shows that one copy of theRARy gene has been disrupted in the F9-Wt-y-119 cell line.The second allele was disrupted by selecting F9-Wt-y-119cells in high levels of G418 (15), resulting in 4 RARy-/- linesof 36 G418-resistant lines tested. Two of these cell lines,F9-Wt-y-119-14 and F9-Wt-y-119-16, were chosen for furtherstudy (Fig. lb).To confirm the disruption ofboth alleles ofthe RARygene,

RAR,yRNA levels were measured by Northern blot analysis(Fig. lc). The F9-Wt-y-119-14 and F9-Wt-y-119-16 cell linesdo not contain RARy-specific transcripts as compared to thecontrol lines (Fig. lc, lanes 3 and 4). Thus, both copies of theRARy gene have been successfully disrupted by homologousrecombination initiated by the y6.1 targeting construct.The ability ofRARy-/- F9 cells to respond to RA was first

determined. F9-Wt stem cells and the RARy-/- (F9-Wt-y-

a

PimRARy

El E2

A B CDNA bindi

P2

2 E3 E5 E8

-1-ing

D Eh1igand binding

mRARy locus

E K B Ea E*. i

E5

probe

RARy Targeting Vector y6

RARy Recombinant

E K B K B E

E

K B Ea E K B K B EI |IT-I Ie -

E5

BamHl Digest [v-~Wt5kb--

ll

I;. ;-- ;-- l.

I

11~ L 11 LL LL

C

24 h 48 h 96 h1 2 3 41r1 2 3 4"'l 2 3 4

Rv

actin

FIG. 1. Targeted disruption of the RARygene by homologous recombination. (a) The RARydisruption vector y6.l is shown (14). The RARygenomic sequence spanning the region encoding the B domain of the RARy protein is interrupted by the neomycin-expression gene driven bythe GTI-II (simian virus 40 enhancer) plus the Rous sarcoma virus TATA box (14). The herpes simplex virus thymidine kinase (tk) gene is locatedat the 3' boundary of the targeting sequences. Using the probe indicated, a successful recombination event gives rise to a 4.5-kb BamHI genomicfragment, in contrast to the wt 6-kb genomic fragment. This targeting vector is designed to disrupt all isoforms of the RARygene. For full details,see ref. 14. B, BamHI; E, EcoRI; Ea, Eag I; K, Kpn I. (b) Southern blot analysis demonstrating disruption of the RARy gene. Positions of thetargeted genomic fragment and of the wt genomic fragment are indicated. Cell line F9-Wt-y-119-17 is also included as a control since this linewas derived from the high G418 selection but failed to undergo gene conversion. (c) Northern blot analysis showing the absence ofthe endogenousRARy mRNA. Cells were treated with 1 ,uM RA and RNA was prepared at the indicated times. Lanes: 1, F9-Wt; 2, F9-Wt-Y.119; 3,F9-Wt-y-119-14; 4, F9-Wt-y-119-16. The 3- to 4-fold reduction of RARy RNA in the control lines F9-Wt and F9-Wt-y119 is consistent withpreviously published reports (16, 28). There is no detectable RAR-y RNA in the RARy-/- cell lines F9-Wt-y-119-14 and F9-Wt-y-119-16.

F-I

BK

E8

E

b

6 kb

4.5 kb

Dow

nloa

ded

by g

uest

on

Oct

ober

21,

202

1

Proc. Natl. Acad. Sci. USA 90 (1993) 9603

119-14) line possess a similar morphology in the absence ofall-trans-RA (Fig. 2). However, in the presence ofRA or RAplus dibutyryl cAMP and theophylline, the F9-Wt (Fig. 2) andRARy+/- (data not shown) cells change their morphology,developing pronounced cell borders and extending cellularprocesses associated with differentiation (Fig. 2 a-d). Incontrast, the RARy-/- line exhibits fewer of the differenti-ation-specific morphological characteristics noted in theF9-Wt cells (Fig. 2 a'-d'). Interestingly, the RAR'y-/- linebegins to display some of the differentiation characteristics atlater times after RA treatment (data not shown). Thus, RARyprovides a critical function during the differentiation of F9stem cells, the absence of which significantly alters thenormal RA responsiveness with respect to the morphologicalchanges associated with differentiation.To investigate the function of RARy during RA-induced

differentiation, the two RARy-/- cell lines were treated with1 ,uM all-trans-RA and the differentiation was monitored byNorthern blot analysis (Fig. 3). In the lines that do notexpress RARy RNA, there is minimal induction of theHoxa-l gene at 24 h, and only low levels of Hoxa-1 RNA aredetected at 48 h after RA treatment. Moreover, no inductionof laminin Bi or collagen type IV (al) RNAs is observed inthe RARy-/- lines. In contrast, in the RARy-/- lines the

c _

FIG. 2. RAR-y-/ cell lines fail to exhibit a complete differenti-ation morphology in culture. As stem cells, the F9-Wt and theRARy-/- cell line F9-Wt-y'.119-14 share a similar compact morphol-ogy. However, after treatment with either 1 AM RA or RA plusdibutyryl cAMP (250 i&M) and theophylline (500 ,uM) (RACT) for 96h, the F9-Wt cell line develops a differentiation phenotype charac-terized by the development of cellular processes, distinct cell bor-ders, and irregular cell shape. In contrast, in the RARy-/- line thisdifferentiation phenotype is much less pronounced. Thus, the ab-sence ofRARyattenuates the changes in morphology associated withRA-induced differentiation. (a) F9-Wt. (b) F9-Wt + RA, 9 h. (c)F9-Wt + RACT, 96 h. (d) F9-Wt + CT, 96 h. (a') F9-Wt-y.119-14.(b') F9-Wt-y-119-14 + RA, 96 h. (c') F9-Wt-y-119-14 + RACT, 96 h.(d') F9-Wt-y-119-14 + CT, 96 h.

Hoxb-1 and CRABPII RNAs are induced to the same level asin wt cells, but at 48 h after RA treatment these RNAs do notdecrease as rapidly as those in the wt cells.Not all differentiation-related genes are affected in the

RARy-/- lines. For instance, the RA responsive genes Rex-1and SPARC are regulated in the RAR-y-/- lines as they arein wt cells (Fig. 3). Interestingly, expression of RARa andRAR/3 (Fig. 3), as well as that of RXRa and RXR,6 (data notshown), remains unaltered in the RARy-/- lines. Theseresults suggest that RARy plays a specific role in the RAregulation of the "early" RA-inducible gene Hoxa-1, as wellas in that of the "late" RA-inducible laminin Bi and collagentype IV (al) genes.A gel mobility-shift assay was used with the addition of

monoclonal antibodies specific for RARa, -/3, and -y toinvestigate which RARs are present in the wt and RARy-/-F9 cells before and after RA treatment. With extracts ofuntreated F9-Wt cells, specific complexes are observed usingRARa and RARy but not RAR/3 antibodies, indicating thepresence of both RARa and RARy but not RAR,8 in thesecells (Fig. 4, lanes 1-3 and 7-9). Interestingly, the intensity ofthe complex formed with RARy is stronger than that of theRARa complex, suggesting that F9-Wt cells are richer inRARy than in RARa. In contrast, and as expected, no RARycomplex is detected in RARy-/- cells under conditionswhere the intensity of the RARa complex is similar to that ofwt cells (compare lanes 1 and 7). Control transfection exper-iments with vectors expressing either RARa, -,8, or -yindicate clearly that the absence of RARy in the "mutant"cells is due to the disruption of the RARy alleles (lanes 12-15and 22-24). After RA treatment, RAR/3 complexes are sim-ilarly "induced" in wt and RARy-/- cells (compare lanes4-6 and 10-12), in agreement with the Northern blot dataabove for RAR/3. As expected, the RA treatment does notaffect the formation of RARy complexes, which remainabsent in the RARy-/- cells (lane 12).Taken together the results described above indicate that

RARyis absent in the mutant cells, whereas RARP is inducedas in wt cells. Since the RAREs of the Hoxa-J and RAR,/genes appear to be identical (23, 24), this latter result indi-cates (i) that RAR,3 induction is autoregulatory, or (ii) thatdue to their different promoter context (29) the RAREs of theRAR/3 and Hoxa-l are bound and activated by differentreceptors (e.g., RARy in the case ofHoxa-l and RARa in thecase of RAR/3), and/or (iii) that there exists a greaterfunctional redundancy among RARs in the case of the RAR,8promoter than in the case of Hoxa-) (e.g., the RAR,8 pro-moter could be stimulated by both RARa and RARy,whereas the Hoxa-l promoter would be preferentially stim-ulated by RARy).

DISCUSSIONWe show here that the inactivation ofRARy in F9 stem cellsprevents several aspects of the RA-associated differentiationpathway observed in F9-Wt cells. RARy gene disruptionspecifically results in a drastically reduced induction ofHoxa-1 RNA and in the lack of induction of the laminin Biand collagen type IV (al) RNAs. In contrast, Hoxb-l andCRABPII show normal RA induction in the RARy-/- linesbut are not correctly down regulated. RAR/3, SPARC, andRex-i regulation are unaffected by the lack of RARy. Thiswork represents identification of some of the direct and/orindirect gene targets of RARy. In the developing mouse,RARa, RAR/3, and RARy are expressed in distinct embry-onic regions, suggesting that each RAR performs a specificdevelopmental function (12, 13). Our observations supportthe idea that each RAR regulates the expression of specificgenes. We have also shown that RARy is involved in themorphological changes associated with RA-induced F9 stem

Developmental Biology: Boylan et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

21,

202

1

9604 Developmental Biology: Boylan et al.

F9-Wt V-119+/- Y-119-14-/- Y-119-16-/- y-119-17+/-

E.e = E . E o = E E c c

W CN t r- ' m e%4 - r 0 '

ma: AA. ....sR. .......

tR #... *.

|Jw~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..........

Hoxa-I

Hoxb-

CRABP-II

RARp

! RARaX ,wr :" ~~~~~~~~~~~~~~... ... " #...Z... ...:.:..

a. "* :6laiWOB1Rex

~~~III Il ~~~~~~~~~~~~lamininBI

*-...A* w collagen type IV

. . .* *> *-i,,, i *; :' 't! j_ j SPARC

W & -41P A I actin. " " " ".~q

FIG. 3. The absence of RARyRNA results in a reduction of RA-induced Hoxa-1, laminin B1, andcollagen type IV (al) RNA ex-pression. The ability of F9 stemcells lacking the RARy gene torespond to RA-induced differenti-ation was tested by Northern blotanalysis. The RARy-/- cell linesF9-Wt-y-119-14 and F9-Wt-y-119-16 exhibit reduced levels ofthe Hoxa-1 RNA and do not ex-press detectable levels of lamininB1 or collagen type IV (al) RNAscompared to the F9-Wt andRARy+/- cell lines F9-Wt-y-119and F9-Wt-y-119-17. Hoxb-1 andCRABPII RNA levels are inducedto the same level as in the wt line;however, they do not decrease asrapidly as in wt cells. Expressionof RARa, RAR,B, Rex-1, andSPARC RNAs is not significantlyaffected in the RARy-/- lines.This experiment was performedtwice with similar results.

Transfected RAR

(x Ia yI a- , yaAb (x,P "x,B Y" (x ,B'(xa y (x [i7 ty -y__ _Yaliyc' B(t

D12 14 16 18 20 22 24 2611 13 15 17 19 21 23 25 27

it I__

RA - + - +

F9-Wt y-1-119-14-/-*+

F9-Wt y-119-14'/-

FIG. 4. Lack of binding of RARy to the Hoxa-1 RARE in aRARy-/- line. Binding of the RARs present in extracts of F9-Wt andRARy-/- cell lines (with or without RA treatment) was investigatedby a gel mobility-shift assay. A radiolabeled oligonucleotide repre-senting the Hoxa-1/RAR,8 RARE was incubated with nuclear ex-tracts prepared from F9-Wt cells (lanes 1-3, RA untreated; lanes 4-6,RA treated), F9-Wt-y-119-14 RARy-/- cells (lanes 7-9, RA un-treated; lanes 10-12, RA treated), F9-Wt cells (lanes 13-15, RAuntreated; lanes 6-18, RA treated), and F9-Wt-y-119-14 RARy-/-cells (lanes 19-24, RA untreated; lanes 25-27, RA treated) trans-fected with murine RARa (lanes 13, 16, 19, 22, and 25), RAR,8 (lanes14, 17, 20, 23, and 26), or RARy (lanes 15, 18, 21, 24, and 27)expression vectors. Arrow indicates the shifted complex formed inthe presence of mouse monoclonal antibodies Ab9a(F) (lanes 1, 4, 7,10, 13, 16, 22, and 25), Ab8,3(F)2 (lanes 2, 5, 8, 11, 14, 17, 23, and 26),and Aby(mF) (lanes 3, 6, 9, 12, 15, 18, 24, and 27) specifically directedagainst the F region of RARa, -P3, and -y, respectively.

cell differentiation. This abnormal morphological differenti-ation of RARy-/- cells may result in part from reducedHoxa-J expression, as we have previously demonstrated thatectopic expression of the Hoxa-J gene in the absence ofRAleads to a morphological change, which is, however, clearlydistinct from that induced by RA (30). In this respect, we notethat Hoxa-J expression does not induce the morphologicaldifferentiation of embryonal carcinoma P19 cells (31).

Disruption of the RARy gene in mice does not produce thealterations associated with Hoxa-1 disruption (14, 32, 33).There may be several explanations for this apparent discrep-ancy. It is possible that the control of Hoxa-J expression inthe context of the whole animal is different from that in atissue culture model or that the functional redundancy amongthe three RAR types is greater in the developing mouse,providing a means for compensating for the loss of onereceptor. Note that RARa but not RARy transcripts weredetected in the rhombencephalic region that is affected byHoxa-l disruption (34). The possibility of functional redun-dancy among the three RARs in vivo is supported by theresults of additional RAR disruptions in the mouse. Micelacking the RARal isoform possess no developmental orgrowth abnormalities (35, 36), suggesting that the loss ofRARal can be compensated for during development. Indeed,even in the F9 RARy-/- cells, there appears to be some levelof redundancy, since Hoxa-1 is detectable at low levelsduring the latter phase of RA treatment. The low levels ofHoxa-l expression at late times of RA treatment either mayresult from RARa action or may be the result of the increasein RAR/3 expression at later times. Further studies with F9embryonal carcinoma cells disrupted for the RARa and/orRAR,8 genes should be useful in resolving these questions offunctional redundancies among RARs.

We are grateful to Dr. Andree Krust for gel mobility-shift assaysand to Drs. Cecile Egly and Marie-Pierre Gaub for RAR antibodies.We thank Mr. Alex Langston, Dr. Lap Ho, Dr. Anna Means, and Dr.Ker Yu for helpful discussions. We are grateful to Mr. Alex Langstonfor assistance with the photographs of the cell lines. J.F.B. is therecipient of a postdoctoral fellowship from the National Institutes ofHealth (1F32 CA09251-01). D.L. was supported by fellowships fromthe Medical Research Council of Canada and from the Centre

AL1 1

Proc. Natl. Acad. Sci. USA 90 (1993)

Dow

nloa

ded

by g

uest

on

Oct

ober

21,

202

1

Proc. Natl. Acad. Sci. USA 90 (1993) 9605

National de la Recherche Scientifique (CNRS). R.T. was supportedby a CNRS fellowship. This work was supported by GrantsROIHD24319 and R01CA43796 from the National Institutes ofHealth (L.J.G.), by the CNRS, the Institut National de la Sante et dela Recherche Mddicale, the Centre Hospitalier Universitaire Re-gional, the Association pour la Recherche sur le Cancer, the LigueNationale Francaise contre le Cancer, the Fondation pour la Re-cherche Medicale Francaise, and the Human Science Frontier Pro-gram (D.L., R.T., and P.C.).

1. De Luca, L. M. (1991) FASEB J. 5, 2924-2933.2. Hong, W. K., Lippman, S., Itri, L. M., Karp, D. D., Lee,

J. S., Byers, R., Schantz, S. P., Kramer, A., Lotan, R., Peters,L. J., Dimery, I. W., Brown, B. W. & Goepfert, H. (1990) N.Engl. J. Med. 323, 795-801.

3. Warrell, R. P., Jr., Frankel, S. R., Miller, W. H., Jr., Schein-berg, D. A., Itri, L. M., Hittelman, W. N., Vyas, R., Andreeff,M., Tafuri, A., Jakubowski, A., Gabrilove, J., Gordon, M. S.& Dmitrovsky, E. (1991) N. Engl. J. Med. 324, 1385-1393.

4. Stoner, C. M. & Gudas, L. J. (1989) Cancer Res. 49, 1497-1504.

5. Giguere, V., Lyn, S., Yip, P., Siu, C.-H. & Amin, S. (1990)Proc. Natl. Acad. Sci. USA 87, 6233-6237.

6. Boylan, J. F. & Gudas, L. J. (1991) J. Cell Biol. 112, 965-979.7. Boylan, J. F. & Gudas, L. J. (1992) J. Biol. Chem. 267,

21486-21491.8. Chambon, P., Zelent, A., Petkovich, M., Mendelsohn, C.,

Leroy, P., Krust, A., Kastner, P. & Brand, N. (1991) inRetinoids: 10 Years On, ed. Saurat, J. H. (Karger, Basel), pp.10-27.

9. Mangelsdorf, D. J., Borgmeyer, U., Heyman, R. A., Zhou,J. Y., Ong, E. S., Oro, A. E., Kakizuka, A. & Evans, R. M.(1992) Genes Dev. 6, 329-344.

10. Leid, M., Kastner, P. & Chambon, P. (1992) Trends Biochem.Sci. 17, 427-433.

11. Allenby, G., Bocquel, M. T., Saunders, M., Kazmer, S.,Speck, J., Rosenberger, M., Lovey, A., Kastner, P., Grippo,J. F., Chambon, P. & Levin, A. A. (1993) Proc. Natl. Acad.Sci. USA 90, 30-34.

12. Dolle, P., Ruberte, E., Kastner, P., Petkovich, M., Stoner,C. M., Gudas, L. J. & Chambon, P. (1989) Nature (London)342, 702-705.

13. Ruberte, E., Dolle, P., Chambon, P. & Morriss-Kay, G. (1990)Development 111, 45-60.

14. Lohnes, D., Kastner, P., Dierich, A., Mark, M., LeMeur, M.& Chambon, P. (1993) Cell 73, 643-658.

15. Mortensen, R. M., Conner, D. A., Chao, S., Geisterfer-Low-

rance, A. A. T. & Seidman, J. G. (1992) Mol. Cell. Biol. 12,2391-2395.

16. Zelent, A., Krust, A., Petkovich, M., Kastner, P. & Chambon,P. (1989) Nature (London) 339, 714-717.

17. Leroy, P., Krust, A., Zelent, A., Mendelsohn, C., Gamier,J.-M., Kastner, P., Dierich, A. & Chambon, P. (1991) EMBOJ. 10, 59-69.

18. Kastner, P., Krust, A., Mendelsohn, J., Gamier, J. M., Zelent,A., Leroy, P., Staub, A. & Chambon, P. (1990) Proc. Natl.Acad. Sci. USA 87, 2700-2704.

19. Green, S., Issemann, I. & Sheer, E. (1988) Nucleic Acids Res.16, 369.

20. Brand, N., Petkovich, M., Krust, A., Chambon, P., de The, H.,Marchio, A., Tiollais, P. & Dejean, A. (1988) Nature (London)332, 850-853.

21. Rochette-Egly, C., Lutz, Y., Saunders, M., Scheuer, I., Gaub,M.-P. & Chambon, P. (1991) J. Cell Biol. 115, 535-545.

22. Gamer, M. M. & Revzin, A. (1981) Nucleic Acids Res. 9,3047-3060.

23. deThe, H., del Mar Vivanco-Ruiz, M., Tiollais, P., Stunnen-berg, H. & Dejean, A. (1990) Nature (London) 343, 177-180.

24. Langston, A. W. & Gudas, L. J. (1992) Mech. Dev. 38, 217-228.

25. Nicholson, R. C., Mader, S., Nagpal, S., Leid, M., Rochette-Egly, C. & Chambon, P. (1990) EMBO J. 9, 4443-4454.

26. Gaub, M. P., Rochette-Egly, C., Lutz, Y., Ali, S., Metthes, H.,Scheuer, I. & Chambon, P. (1992) Exp. Cell Res. 201, 335-346.

27. Rochette-Egly, C., Gaub, M.-P., Lutz, Y., Ali, S., Scheuer, I.& Chambon, P. (1992) Mol. Endocrinol. 6, 2197-2209.

28. Hu, L. & Gudas, L. J. (1990) Mol. Cell. Biol. 10, 391-396.29. Shen, S. S. (1991) J. DNA Sequence Mapp. 2, 111-119.30. Goliger, J. A. & Gudas, L. J. (1992) Gene Expression 2,

147-159.31. Pratt, M. A. C., Langston, A. W., Gudas, L. J. & McBumey,

M. W. (1993) Differentiation 53, 105-113.32. Lufkin, T., Dierich, A., LeMeur, M., Mark, M. & Chambon,

P. (1991) Cell 66, 1105-1119.33. Chisaka, O., Musci, T. S. & Capecchi, M. R. (1992) Nature

(London) 355, 516-520.34. Ruberte, E., Dolle, P., Chambon, P. & Morriss-Kay, G. (1991)

Development 111, 45-60.35. Li, E., Sucov, H. M., Lee, K., Evans, R. M. & Jaenisch, R.

(1993) Proc. Natl. Acad. Sci. USA 90, 1590-1594.36. Lufkin, T., Lohnes, D., Mark, M., Dierich, A., Gorry, P.,

Gaub, M. P., LeMeur, M. & Chambon, P. (1993) Proc. Natl.Acad. Sci. USA 90, 7225-7229.

Developmental Biology: Boylan et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

21,

202

1