Gene in a Retinoic Acid-Nonresponsive Embryonal Carcinoma Cell. M. A. CHRISTINE ...... closely related members of the steroid receptor gene family. (19). A natural ... The biochemical similarities between v-erbA and RARa' suggest that the.
MOLECULAR AND CELLULAR BIOLOGY, Dec. 1990, p. 6445-6453
Vol. 10, No. 12
0270-7306/90/126445-09$02.00/0 Copyright © 1990, American Society for Microbiology
A Dominant Negative Mutation of the Alpha Retinoic Acid Receptor Gene in a Retinoic Acid-Nonresponsive Embryonal Carcinoma Cell M. A. CHRISTINE PRATT, JARMILA KRALOVA, AND MICHAEL W. McBURNEY* Department of Medicine, University of Ottawa, 451 Smyth Road, Ottawa, Ontario, Canada KIH 8M5 Received 27 March 1990/Accepted 11 September 1990
Pluripotential embryonal carcinoma cells such as those of the P19 line differentiate when exposed to retinoic acid (RA). The RAC65 cell line is a mutant clone of P19 cells selected to be RA nonresponsive. RAC65 cells carry a rearrangement affecting one of the genes encoding a nuclear retinoic acid receptor (RARa). The mutant gene encodes a protein, RARa', that has lost its 70 C-terminal amino acids, thus truncating the RA-binding domain. The RARa' was found to be a dominant repressor of transcription from an RA-responsive target gene; however, expression of RARa' was insufficient to confer RA nonresponsiveness, suggesting that RAC65 cells carry an additional mutation(s) affecting RA-induced genes.
Retinoic acid (RA) has a wide variety of effects in biological systems (9, 44). In cell culture systems, RA can profoundly affect the state of differentiation (11), while in whole organisms RA is a potent teratogen (40) that adversely affects development of the amphibian central nervous system (16) and regenerating limbs (45). RA has also been shown to be a naturally occurring "morphogen" which plays an important role in chick wing morphogenesis (17, 59, 62). An enzyme capable of synthesizing RA has been identified
MATERIALS AND METHODS Cell culture. P19 (50) and RAC65 (35) cells were cultured as previously described (54). Transfection of plasmid DNA into these cells was performed by using the calcium phosphate technique (10). Chloramphenicol acetyltransferase (CAT) activities were performed as described elsewhere
(25).
(53).
How RA exerts its effects in various biological systems remains unclear. There is evidence that RA acts on the cell surface (67) and other evidence that a cytoplasmic binding protein might be the important site of RA action (57). Recently, three nuclear receptors for RA have been identified as members of the steroid receptor family (4, 8, 23, 37, 52, 68). These three nuclear receptors (RARa, RARPi, and RAR-y) are proteins with the same general structure: they have a 60-nucleotide region containing two zinc fingers comprising the DNA-binding domain as well as a region near the C terminus which binds the ligand, RA. The DNAbinding domains of all three RARs are highly homologous, and each binds the synthetic sequence element recognized by the thyroid hormone receptors called the thyroid-responsive element (TRE) (3). In the presence of RA, binding of any of the RARs to the TRE enhances transcription from adjacent promoters (63). In situ hybridization studies suggest that RA-responsive tissues express transcripts encoding nuclear RARs (15, 22), but there is no direct evidence that the RA-induced effects are mediated by these RARs. Murine embryonal carcinoma (EC) cells are among the cultured cell types that respond to RA (61). Cells of the P19 line of EC differentiate after exposure to RA and develop into neurons, glia, and smooth muscle cells (34, 49, 54a). The RAC65 subclone of P19 cells is a mutant line that fails to differentiate in RA (35). We set out to identify the site of the mutation(s) carried by RAC65 cells with the idea that this information might allow us to determine the site of action of RA responsible for the induction of differentiation of P19 cells. We report here that RAC65 cells carry a mutation affecting the RARa gene.
*
Corresponding author. 6445
Northern (RNA) and Southern blot analyses. Total cellular RNA was isolated by the lithium chloride-urea procedure (2) and separated in 1% agarose gels containing 1.1 M formaldehyde (47). The RNA was transferred to Hybond N (Amersham) and hybridized with multiprimed labeled cDNA fragments. Genomic DNA was isolated as described previously (7). Samples of 10 ,ug were digested to completion with BamHI and separated on 0.8% agarose gels. The DNA was transferred to Hybond N membranes before being probed with radiolabeled cDNAs. The cDNAs used as probes were derived from human RARa (23), mouse RARa (68), human RAR, (4), mouse RAR-y (37), and mouse a-tubulin (42). RAC65 cDNA library. A cDNA library was constructed from RAC65 poly(A)+ RNA. Double-stranded cDNA was prepared by using a cDNA synthesis kit (Bethesda Research Laboratories) and cloned into lambda gtll (Amersham). Plaque screening was performed by using the RARax cDNA probe (23). Positive plaques were isolated, and cDNA inserts were removed after digestion with EcoRI. DNA sequence analyses were performed by the dideoxy method with reagents from the Sequenase DNA sequencing kit (U.S. Biochemical Corp.). The RARa' cDNA was subcloned into the EcoRI site of pTZ18/19R (Pharmacia), and nested deletions were created with exonuclease III. Singlestranded templates were produced from plasmid-transformed JM103 bacteria by superinfection with bacteriophage M13K07. DNA sequence was obtained either from both strands or from a minimum of three independent ladders from the same strand. Expression plasmids. The pTRE-CAT gene consists of the mouse mammary tumor virus promoter in which the glucocorticoid response element was replaced by the TRE (63). Other eucaryotic expression vectors were based on the regulatory sequences of the mouse pgk-J gene (1). The pgk-neo gene carries the bacterial neo gene between the promoter and polyadenylation signals of pgk-l (7a). Expression vectors for RARa', human RARI, and antisense
6446
PRATT ET AL.
MOL. CELL. BIOL. LO
aY)
_.
ua
aD
U< AL
c,c,
428S
L-P
RA RACj +1|+ RARc+1|-+RAR(|3+
RA+
_
9
foldinduction
14.0
_
6
0.05
a
5.6
RSV-CAT1
| _
5.4
FIG. 1. Inability of the endogenous RARs in RAC65 cells to activate transcription from pTRE-CAT. P19 or RAC65 cells were transfected with 10 ,ug of pTRE-CAT reporter plasmid, and the indicated samples were treated with or without 10-6 M RA. The RARa or RAR, expression vector (5 ,ug) was cotransfected into RAC65 cells along with the reporter plasmid where indicated. Induction values are means of at least four different experiments. Results are representative of a typical experiment.
constructed from pgk-neo by replacing the neo gene with the cDNA. The human RARa expression vector was driven by the Rous sarcoma virus (RSV) promoter (23). RARa'
418S
were
RESULTS RAC65 cells fail to activate pTRE-CAT. The three RARs are able to bind to the TRE and activate transcription from adjacent promoters in the presence of RA. Plasmid pTRECAT (63) carries the CAT gene driven by a promoter containing the TRE. After transfection into P19 cells, plasmid pTRE-CAT directed high levels of expression of CAT in the presence of RA (Fig. 1), suggesting the presence in these P19 cells of at least one RAR species. There was virtually no RA-induced CAT activity in transfected RAC65 cells, suggesting that these cells lack functional RARs. Transfection of RAC65 cells with pTRE-CAT along with expression vectors encoding either RARa or RARP resulted in RAinduced CAT activities in these transient expression experiments. RAC65 cells were transfected with the RARa or RARP expression vectors along with the pgk-neo gene, and transformed clones were selected in G418 in the presence or absence of RA. Under these conditions of cotransfection, at least 50% of the G418-resistant colonies express the cotransfected gene (data not shown), but none of the more than 200 G418-resistant colonies differentiated in the presence of RA, indicating that replacement of an RAR was insufficient to make RAC65 cells RA responsive. To clarify the nature of the RAC65 cell mutation(s), we used polyethylene glycol to fuse RAC65 cells with a ouabainand thioguanine-resistant derivative of P19 cells (35). Cell hybrids were selected in medium containing hypoxanthineaminopterin-thymidine (43) plus ouabain in the presence or absence of RA. The hybrid cells (more than 50 clones) resembled the RAC65 parent in their inability to differentiate in RA. Thus, the mutant phenotype of RAC65 cells appears to be dominant, consistent with our inability to reverse the mutant cell phenotype by transfection with an expression vector encoding the wild-type RARs.
RAR-Xt
RAR-Y
tubulin S.
1
2
3
4
FIG. 2. Northern blot analysis of P19 and RAC65 mRNA encoding RARa and RARy. Total cellular RNA (20 jig) from P19 or RAC65 cells was electrophoresed and transferred to nylon filters. One set of filters was probed with human RARa cDNA, and the other was probed with the mouse RAR-y cDNA. The blots were autoradiographed at -70°C for 3 days with an enhancing screen, after which they were stripped and rehybridized to an a-tubulin cDNA probe. Tubulin probed blots were exposed for 20 h.
RARa mRNA and genes are altered in RAC65 cells. To determine which of the three RARs are expressed in P19 and RAC65 cells, RNA blots from these two cells were hybridized with cDNA probes for the three RARs. Neither cell expressed RARPi (see Fig. 7), while both contained the RAR-y transcript (Fig. 2). P19 cells expressed the RARa as two abundant mRNA species of 3.3 and 4.0 kb along with a minor species of 2.1 kb. RAC65 cells contained both of the larger RARa transcripts at about half the level found in P19 cells (Fig. 2). However, the 2.1-kb transcript was much more abundant in RAC65 cells than in P19 cells. An additional, smaller band was apparent in some blots of RAC65 cell RNA. We believe that this band is artifactual and forms as a result of distortion of the band of abundant 2.1-kb mRNA by the 18S rRNA. Southern blots of genomic DNA from P19 and RAC65 cells were probed with the cDNAs of the three RARs. The restriction fragments hybridizing to RARP and RAR-y probes were identical for the two cell lines (data not shown). However, the blots probed with RARa cDNA detected restriction fragments in RAC65 cells that were absent from P19 cell DNA (Fig. 3). Two BamHI fragments of P19 cell DNA hybridized to the full-length RARa cDNA. Both bands were also present in RAC65 cell DNA; however, the upper band was present at approximately half the intensity, and an additional band of a still larger fragment was also present. Probing the blot with sequences specific to the 3' end of the cDNA indicated that these upper bands derived from the 3' end of the gene. The results of the Southern and Northern analyses suggested that RAC65 cells carried one normal and one rear-
RETINOIC ACID RECEPTOR DOMINANT NEGATIVE MUTATION
VOL. 10, 1990
B
A
O
40
0
cn
O
11 u
0
a
23*
4.2>
¢
6* *'10 _
2.0o 1.95 1.5m.
FIG. 3. Southern blot analysis of P19 and RAC65 genomic DNAs. Samples (10 Fg) of DNA digested with BamHI were separated on a 0.8% agarose gel, transferred to nylon, and hybridized to
full-length mouse RARa cDNA (A). The same blot was stripped and rehybridized to the ApaI-SmaI fragment of the mouse RARa cDNA (B), which consists of the 3' untranslated region of the RARa mRNA. This latter probe should hybridize exclusively to exon 7, the last exon of the RARa gene (Giguere, personal communication). DNA molecular size markers (in kilobase pairs) are shown on the left.
ranged RARa gene, that the rearrangement affects the 3' end gene, and that this rearranged gene encodes the abundant 2.1-kb transcript. The RARa' transcript encodes a truncated receptor. A cDNA library was constructed from RAC65 cell RNA, and a clone carrying the novel transcript was isolated and sequenced (Fig. 4). This cDNA, called RARa', is a 1.7-kb fragment consisting of 404 nucleotides of 5' untranslated region followed by a coding region identical to that reported for the mouse RARa (68). This coding region, however, extends only to codon 390. At codon 391 the expected glycine is replaced by alanine, and codon 392 is a translation termination signal. Following this region is another 187 nucleotides of unrelated sequence containing a polyadenylation signal and a poly(A) tail. A search of GenBank reof the
vealed that this downstream sequence is derived from the
long terminal repeat region of the
mouse ETn
element. This
ETn element is a retroviruslike sequence that is repeated more than 1,000 times in the mouse genome and is expressed at very high levels in EC cells (60). The presence of the novel RARa transcript in RAC65 cells
6447
was confirmed by polymerase chain reaction analysis. Oligonucleotide primers for the polymerase chain reaction were synthesized from sequences on either side of the predicted RARa-ETn fusion. These primers amplified a fragment of the expected size and sequence from reverse-transcribed RAC65 cell RNA, indicating the authenticity of the RARa cDNA. RARa' is a dominant suppressor of RA-induced transcription. The protein encoded by the RARa' transcript is predicted to be 70 amino acids shorter than the wild-type receptor. These 70 amino acids are lost from the C-terminal region that contains the RA-binding domain (Fig. 4B). The nature of the RARa' sequence suggests that it might encode a truncated protein capable of binding the TRE sequence but incapable of binding RA. The entire RARa' cDNA was subcloned into a eucaryotic expression vector (Boer et al., in press), and this construct was cotransfected into P19 cells along with the pTRE-CAT reporter gene. In the absence of RARa' there was strong RA-induced CAT expression from pTRE-CAT, but this induction was essentially obliterated by cotransfection with the vector encoding RARa' (Fig. 5). No inhibition resulted from cotransfection of pTRE-CAT with plasmid encoding normal receptor or no receptor. Interestingly, the vector encoding the antisense to RARa' also inhibited RA-induced CAT induction. This antisense effect in a transient expression experiment suggests that the normal RARoa protein is relatively unstable, a conclusion consistent with its estimated half-life of 120 min (32). The expression of RARa' strongly inhibited RA-induced expression from pTRE-CAT but had no effect on the expression from unrelated promoters such as those from the murine pgk-J gene and RSV. Cotransfection of the RARa' vector along with an expression vector encoding the normal RARa indicated that substantial suppression of RA-induced transcription was achieved by RARa' in the presence of equal or higher concentrations of the normal RARa (Table 1). Thus, RARa' appears to be the product of a dominant negative mutation (30), and the presence of RARa' in RAC65 cells appears to be responsible for the failure of these cells to express the pTRE-CAT gene (Fig. 1). RARa' is insufficient to account for the RAC65 cell phenotype. To determine whether the presence of RARa' accounts for the RA-nonresponsive phenotype of RAC65 cells, we cotransfected P19 cells with the pgk-neo gene along with either plasmid vector alone, expression vector encoding RARa, or expression vector encoding RARa'. Colonies were selected in G418 in the presence or absence of RA. In eight different experiments, more than 1,000 G418-resistant clones were tested for the ability to form colonies of EC cells in the presence of RA. In the absence of RA, large macroscopic colonies EC cells formed after 7 days of culture. In the presence of RA, not a single macroscopic colony of EC cells was present at this time. Since the RAC65 cells formed large colonies of EC cells in the presence or absence of RA, it appears that expression of RARa' is insufficient to confer the RA-nonresponsive phenotype on P19 cells. Although no macroscopic colonies of EC cells formed in RA-treated cultures of G418-resistant colonies, it was possible to establish from these cultures lines of EC cells that grew continuously in RA and were partially nonresponsive to RA (i.e., not responsive to RA doses sufficient to induce differentiation of P19 cells, 10' M, but still responsive to higher doses to which RAC65 cells are not responsive). To establish these lines, populations of cells growing in the presence of G418 and 10-7 M RA were pooled and grown continuously in the presence of RA. After a number of
A
v40 v50 v70 v30 v60 vlo v20 agacaaattgcatccc :cgettcgagcaccacaggacatgcctccccccagccacctacctgggggcccttctgggagt vlO0
v90
vllO
v120
v140
v130
V150
ggcatcctctttgata^cctgaagaccagctctggaccttcctacaaaagtgtctcccaggaccgcttaaccaaaggacq v180
v170
v190
v200
v220
v210
v230
gattcttgggacatcccccaatccacctggccccctccggctagggagggggacccagaagactaaagttgactaacttg v250
v290
v280
v270
v260
v300
v310
ggacttgcagctccag;gacagaagggggtgggggtgggctgaccacccaaccccccctctgggccccccccggccccat v330
v350
v340
v360
v380
v370
v390
gccccgaggaagagggactgcgtgaaggccatcacaactacctgccagactgtttgcctgctcttctgactgtggctgct v410
v430
v420
v440
v450
v470
v460
tggcatggccagcaatagcagttcctgcccaacacctgggggcgggcacctcaatgggtacccagtacccccttacgcct S
A
M
N
S
S
F
F
P
S
C
P
v500
v490 P
L
M
G
P
T
G
G
v510 G
L
P
S
G
L
H
P
G
Y
G
N
v530
v520 L
A
S
T
P
P
V
P
Q
L
Q
H
F
A
Y
v550
v540 L
P
G
S
V
v570 v580 v590 v600 v610 v630 v620 tacagcacaccgtccccagccaccatcgagacccagagcagcagttccgaagagatagtacccagccctccctcaccacc Y S T P S P A T I E T Q S S S E E I V P S P P S P P v650
v670
v660
v690
v680
v700
v710
gcccctgccccgcatc:acaagccttgctttgtttgtcaagacaaatcatccggctaccactatggggtcagtgcctgtg L
P
P
R
I
K
Y
v730
P
C
F
C
V
Q
D
v750
v740
S
K
S
v760
G
H
Y
Y
G
V
S
v780
v770
A
C
E
v790
agggctgtaagggcttcttccgacqaagcatccagaagaacatggtgtatacgtgtcaccgggacaagaactgcatcatc G
C
G
K
F
F
R
v810
S
R
K
Q
I
N
V
T
Y
C
H
D
R
v850
v840
v830
v820
M
N
K
I I
C
v870
v860
aacaaggtgacccgga3ccgctgccagtactgccggctgcagaaatgtttcqacgtggqcatgtccaaggagtcggtgcg N
K
T
V
!.
R
R
v890
C
Q
C
Y
v900
L
R
K
Q
v910
C
D
F
v920
G
V
S
M
K
E
v940
v930
S
V
R
v950
aaacgatcgaaagaaaaagaagaaagaggcacccaagcccgagtgctcagagagctactcgctgacgcctgaggtgggcq R
D
N
K KK
K
E
A
v980
v970
P
K
P
E
v990
C
S
S
E
vlOOO
Y
S
v10
L
T
P
E
v1020
V
G
E
v1030
agctcattgagaagqttcgcaaagcgcaccaggagaccttcccggccctctgccagctgggcaagtacactacgaacaac L
E
I
K
R K A
V
V1OS0 S
S
E
Q
Q
H
v1060
R V
S
v1130
L
F P
T
E
I
D
L
W
Q L
G
v1090 F
K
D
vl l50
v1140
C
v1080
v1070
D
L
A
E
S
T
Y
v1lO L
S
T
v1170
v1160
K
T
N
N
vlllO C
K
I
I
v1180
T
K
v1190
tgtggagttcgccaaz;-agcttcccggcttcaccaccctcaccatcgccgaccagatcaccctcctcaaggctgcctgcc V
A K Q
F
E
L
G
P
F
v1220
v1210
T
L
T
I
T
A
v1240
v1230
Q
D
T
I
L
L
K
A A
v1260
v1250
C
L
v1270
tggatatcctgattctgcgaatctgcacgcggtacacgcctgagcaagacacactgaccttctcagatggactgaccctg I
D
I
L
L
R
I
v1290
C
T
R
T
Y
P
Q
E
D
T
v1320
v1310
v1300
T
L
F
v1330
S
G
D
L
L
T
v1350
v1340
aaccggactcagatgcacaacgctggctttggccccctcaccgacttggtctttgccttcgccaaccagctgctgcccct N
R
T
Q
N
H
N
A
G
F
G
L
T
D
L
V
F
A
F
v1410
v1400
v1390
v1380
v1370
P
A
Q
N
L
v1420
L
P
L
v1430
ggagatggacgatgctgagactggactgctcagtgccatctgcctcatctgtggagaccgacaggacctggagcagccag E H D
D A E
v1450 K
N
D
V
L
G
T
L
v1460 L
Q
v1530
E
S
C
I
A
L
L
E
A
L
v1550
v1540
K
G
C
I
D
V
v1560
V
Y
Q
R
v1490
v1480
v1470
P
L
D
E
L
v1500
R
K
R
R
P
S
D
P
R
P
H
v1590
v1580
v1570
Q
V1510
atgttccccaagatgctgatgaagatcacagaccttcggagcatcagcgccaaggTcctgaaacctgcttgctcggggtgg
M
F
P
K
M
L M K
I
T
v1620
v1610
D
L
S I S
R
v1630
A
K A
A
v1650
v1640
v1660
v1670
agcttcctgctcattcgttctgccacgcccactgctggaacctgcggagccacacccgtgcacctttctactggaccaga v1700
v1690
v1710
v1720
v1730
v1750
v1740
gattattcggcgqga3tcgggtcccctcccccttccttcataactgqtgtcacaacaataaaatttgagccttgatcaga aaaaaaa
B
461 154
88
-r-
I
A/B
200 U,,,,,, 11 ~/
C
419
,,, ,
,+
-
4
_
A
ECTn
An
nAn
E
391 FIG. 4. The RARa' cDNA. (A) Nucleotide and derived amino acid sequences of the RARa' cDNA. The junction between the RARa and ETn sequences is indicated after nucleotide 1575 with an inverted triangle. The ETn polyadenylation signal at nucleotides 1737 to 1742 is underlined. The closed triangle after nucleotide 1593 denotes the position of an additional A in the published ETn sequence, and the open triangles at nucleotides 1656 and 1726 indicate single-base changes relative to the published sequence (60). (B) Schematic diagram of RARa' cDNA. Letters indicate the functional receptor domains (19): A/B, a putative transcriptional activation region; C, the DNA-binding zinc finger region; E, the RA-binding domain. The boundaries of the domains are indicated by amino acid residue numbers. RARa encodes a 461-amino-acid protein, while RARa' terminates at amino acid 391 and is followed by an ETn-derived sequence. 6448
RETINOIC ACID RECEPTOR DOMINANT NEGATIVE MUTATION
VOL. 10, 1990
1
3
*
*
0
2
-28S
_4-
. 4'
RA
4
Cj PGEMf RAR a' RAR a'IRSV-CATI R1RAR 0~~~~~~~~~ +
-
_
+
_
+
+
-
+
.
fold-
11.25
induction
8.6
0.05
0.6
passages, colonies of EC cells emerged which appeared to be nonresponsive to RA. Because of the means by which they were isolated, it was not possible to determine the frequency with which these variants emerged; however, they were isolated from cultures transfected with all three constructs. EC cells derived from transfections involving the RARa' vector invariably expressed the RARa' mRNA, as determined by Northern blot hybridization (Fig. 6). EC cultures transfected with either vector alone or vector encoding the normal RARa expressed only the two normal RARa transcripts. One clone of these RA-nonresponding P19(RARa') cells carrying and expressing the RARa' vector was transfected with the pTRE-CAT reporter gene, and the level of RAinduced CAT expression was significantly reduced relative to that seen in P19 cells but still much higher than that seen in RAC65 cells (Table 2). However, the low frequency with which RA-nonresponsive cells were recovered from P19 cells transfected with the RARa' vector and the much higher frequency with which the vectors are known to be expressed in these cotransfected cells suggest that expression of the RARat' is insufficient to confer a nonresponsive phenotype to an otherwise normal P19 cell. The fact that high levels of RARa' expression were found in the RA-nonresponding cells suggests that RARa' contributes to, but is not entirely TABLE 1. Repression by RARa' of transcription of pTRE-CAT in the presence of RARaa CAT activity with given amt (ILg) of RARaIa
(4g)
5.0
2.5
1.0
0.5
0.1
0
0 5
0.03 0.5
0.05 2.2
2.3 8.6
3.6 10.5
3.7 ND
10.5 10.4
p.g
r -18S
I
FIG. 5. Demonstration that RARa' is a dominant repressor of pTRE-CAT expression. P19 cells were transfected with 7 p.g of pTRE-CAT along with 10 ,ug of plasmid-carrying expression vectors driving the RARa or RARa' cDNA in the sense (+) or antisense (-) orientation. The lanes labeled pGEM4 were transfected with pTRECAT along with 10 .Lg of plasmid vector alone; those labeled RSV-CAT were transfected with RSV-CAT alone. Induction values are the means of at least three independent experiments. Results of a typical experiment are shown.
RARat
6449
a P19 cells were cotransfected with 7.5 of pTRE-CAT and different amounts of plasmid carrying the RARa' expression vector in the presence or absence of 5 .ag of the RARa expression vector. Total amounts of DNA in each transfection were equalized with pGEM4. b Expressed as the ratio of activities obtained in the presence and absence of RA. ND, Not determined.
_p
*
*
tubulin
t
FIG. 6. Northern blot analysis of RNA from P19(RARa') cells. Total cellular RNA was derived from P19 (lane 1), RAC65 (lane 2), or P19(RARa') (lane 3) cells, the latter being a stable transformant of P19 cells selected for RA nonresponsiveness after transfection of the RARa' expression vector. A 20-pLg sample of RNA from each cell was subjected to Northern blot analysis. The blot was probed with the RARa cDNA.
responsible for, the RA-nonresponsive phenotype of RAC65 cells. RA-induced RARP gene expression. Although RARP transcripts were not detected in P19 cells, exposure of these cells to RA resulted in the appearance of RARP transcripts within 1 h of RA exposure (Fig. 7A). The induction of RARP expression was not inhibited by the protein synthesis inhibitor emetine (Fig. 7B), indicating that this gene is RA responsive in EC cells as it is in hepatoma cells (13). Unexpectedly, the RAR, transcript also appeared in RAtreated RAC65 cells with essentially the same kinetics as in P19 cells (Fig. 7C), and this induction was also not dependent on protein synthesis (Fig. 7D). Another early event in P19 cells following RA exposure is the rapid decline in the abundance of small RNA polymerase III transcripts from the murine B2 short interspersed repetitive sequence (5). This decline in small B2 transcripts also occurs in RAC65 cells (6a). Thus, despite the inability of RAC65 cells to differentiate in RA and despite the expression in these cells of the RARa dominant negative mutation, TABLE 2. Reduction of RA-induced expression of pTRE-CAT in P19(RARa') cells RA-induced CAT activityb
Cell linea
P19 P19(RARa')
RAC65
±g
Expt 1
Expt 2
Expt 3
7.2 3.4 0.5
8.7 3.2 ND
15.5 3.4 0.7
a Transfected with 10 of pTRE-CAT. b Expressed as the ratio of activities in transfected cells cultured in the presence and absence of RA. Results are from three independent experiments.
6450
PRATT ET AL.
MOL. CELL. BIOL.
A
B
Time (hr) in RA
1
0 9
2
3
6
I.*
lt
t
18
E
C
24 s*III
+RA
RA
E
A . . . i-
