Jan 9, 1989 - aminotransferase, phosphoenolpyruvate carboxykinase, and argininosuccinate synthetase is mediatedby a specific genetic locus (TSEI) that ...
Vol. 9, No. 7
MOLECULAR AND CELLULAR BIOLOGY, JUlY 1989, p. 2837-2846 0270-7306/89/072837-10$02.00/0 Copyright C) 1989, American Society for Microbiology
Hormonal Regulation of TSEI-Repressed Genes: Evidence for Multiple Genetic Controls in Extinction MATHEW J. THAYER AND R. E. K. FOURNIER* Department of Molecul(ar Medicine, Fred Hiutchinson Cancer Research Center, 1124 Columbia Street, Seattle, Washington 98104 Received 9 January 1989/Accepted 26 March 1989
Somatic cell hybrids formed by fusing hepatoma cells with fibroblasts generally fail to express liver functions, a phenomenon termed extinction. Previous studies demonstrated that extinction of the genes encoding tyrosine aminotransferase, phosphoenolpyruvate carboxykinase, and argininosuccinate synthetase is mediated by a specific genetic locus (TSEI) that maps to mouse chromosome 11 and human chromosome 17. In this report, we show that full repression of these genes requires a genetic factor in addition to TSE1. This conclusion is based on the observation that residual gene activity was apparent in monochromosomal hybrids retaining human TSEI but not in complex hybrids retaining many fibroblast chromosomes. Furthermore, TSE1repressed genes were hormone inducible, whereas fully extinguished genes were not. Analysis of hybrid segregants indicated that genetic loci required for the complete repression phenotype were distinct from TSE1.
Tissue-specific gene expression in mammalian cells is primarily regulated at the level of transcription (8). A particular gene may account for a large fraction of total transcription in one cell type yet be virtually silent in other cell lineages. Furthermore, expression of many tissue-specific genes is controlled by humoral agents, and the hormonal responses of a given gene may differ in different tissues. Although the regulation of tissue-specific and inducible gene expression clearly involves interactions between transacting factors and specific nucleotide sequences in target genes, the developmental events that establish these differentiated phenotypes have yet to be defined. Genetic tests can be used to identify trans-acting factors involved in the regulation of tissue-specific gene activity. Davidson et al. (7) first observed that somatic hybrids formed by fusing different cell types fail to express tissuespecific products, a phenomenon they termed extinction. This phenomenon proved to be both general, occurring in virtually all intertypic hybrid crosses, and bidirectional, affecting the differentiated products of both parental cells (32, 37, 40). Subsequently, Weiss and co-workers discovered that hybrid segregants that had lost parental chromosomes could reexpress previously extinguished functions (39, 40). These observations suggested that intertypic hybrids could provide a system with the potential to define genetic factors that affect tissue-specific gene activity in trans. With the development of methods for constructing karyotypically simple hybrids with predetermined genotypes (10, 17), that potential has been largely realized. Recent studies have identified two distinct genetic loci that are involved in extinction of liver gene activity in hepatoma hybrid cells (13, 15, 23, 24). The first such locus, tissuespecific extinguisher 1 (TSEI), is a discrete genetic entity that resides on mouse chromosome 11 and human chromosome 17. TSEI extinguishes tyrosine aminotransferase (TAT) (13), phosphoenolpyruvate carboxykinase (PEPCK) (15), and argininosuccinate synthetase (AS) (this report) expression in trans, but it has no effect on expression of most other liver genes. A second extinguisher locus has been mapped to mouse chromosome 1; rat hepatoma cells retain*
ing that single fibroblast chromosome are extinguished for both serum albumin and alcohol dehydrogenase gene activity (A. C. Chin and R. E. K. Fournier, submitted for publication). This locus (TSE2) is apparently identical to one previously assigned to an L-cell marker chromosome termed M1 (24). Deletion hybrids retaining fragments of human chromosome 17 have been used to show that the TAT, PEPCK, and AS extinction phenotypes behave as a single trait (15). That is, TSEI maps to a specific site on 17q distal to D17S4, and this region seems to affect expression of all three genes in a coordinate manner. This is interesting because TAT, PEPCK, and AS share other forms of gene control. Developmentally, all three genes are activated in the liver within a few hours of birth (27, 28, 33), and this process requires the function of a mouse chromosome 7-linked locus designated csdr-l (16, 31). Furthermore, humoral agents modulate expression of these genes in similar ways: both glucocorticoids and cyclic AMP (cAMP) induce PEPCK (29), TAT (30), and AS (22) expression in liver. Other studies suggest that cAMP stimulates the initial synthesis of PEPCK shortly after birth (28), and both glucocorticoids and cAMP have been implicated in the postnatal induction of TAT (11, 12, 33). In this report, we show that basal expression of the TAT, PEPCK, and AS genes was reduced but not abolished in the presence of fibroblast TSEJ. Furthermore, these TSEIrepressed genes were partially responsive to hormonal agents that normally induce their transcription. In marked contrast, neither residual activity nor hormone inducibilty could be demonstrated in genotypically complete hepatoma x fibroblast hybrids. These data indicate that multiple controls are involved in extinction and that complete transcriptional repression of the TAT, PEPCK, and AS genes requires a genetic factor(s) in addition to TSEI. MATERIALS AND METHODS Cell lines and culture conditions. FAO-i is an HPRT-, ouabain-resistant derivative of the highly differentiated rat hepatoma line H4IIEC3 (25); its properties have been reported (13). The microcell hybrid clone FH(17)I was constructed by transferring human chromosome 17 from diploid
Corresponding author. 2837
2838
THAYER AND FOURNIER
MOL. CELL. BIOL.
fibroblasts into TK- FTO-2B rat hepatoma recipients (14) and selecting the TK+ phenotype in hypoxanthine-amino pterin-thymidine. FHB(17)1 was a backselectant population generated from FH(17)1 by selection in medium containing 30 ,ug of bromodeoxyuridine per ml; these cells no longer retain human chromosome 17. The FF-series whole-cell hybrids were prepared by fusing FAO-1 rat hepatoma cells with diploid fibroblasts from C57BL/6J mouse embryos. The properties of these clones have been reported (13). All cells were cultured in a 1:1 mixture of Ham F12Dulbecco modified Eagle medium supplemented with 10% fetal bovine serum (GIBCO Laboratories). FH(17)1 was maintained in medium containing hypoxanthine-aminopterin-thymidine. Antibiotics were not used, and the cells were free of mycoplasma, as judged by staining with the fluorochrome Hoechst 33258 (2). Plasmids. Plasmids were labeled to specific activities of 0.5 x 109 to 1.0 x 109 cpm/,ug by random priming. The rat PEPCK cDNA clone pPCK10 was provided by R. Hanson (42). The rat TAT cDNA clone pcTAT-3 was obtained from G. Schutz (34). The human cx-tubulin cDNA pKa-1 was provided by N. Cowan (5). The rat argininosuccinate synthetase cDNA clone pASR was from D. Mathieu-Mahul (19). RNA dot blots. (i) Time course analyses. Monolayers of cells in late exponential growth were exposed to 10-6 M dexamethasone or 5 x 10-' M dibutyryl cAMP (dBtcAMP) plus 1 mM theophylline for 2, 4, 8, 16, 24, or 36 h. Cytoplasmic RNAs (3 ,ug) extracted from the cells were applied in duplicate to Zetabind membranes and hybridized to cDNA probes as described previously (3). After autoradiography, each dot was cut out, and radioactivity was quantitated by liquid scintillation counting. Each time point represents the amount of specific hybridization normalized to the a-tubulin control. (ii) Dose-response curves. Monolayers of cells in late exponential growth were exposed to 5 x 10-6 X 10-5, 5 x 10-4, or 5 x 10-3 M dBtcAMP for 4 h, and cytoplasmic RNA was extracted and analyzed by dot blotting as described above. Each point was normalized to the value for the x-tubulin control, and values are expressed relative to maximal induction in each cell line. (iii) Cycloheximide experiments. Monolayers of cells in late exponential growth were exposed to 5 x 10-4 M dBtcAMP plus 1 mM theophylline for 0, 2, or 4 h in the presence or absence of 10 F.M cycloheximide. This concentration of cycloheximide inhibited protein synthesis by >90% at both 2 and 4 h of incubation, as judged by incorporation of [3H]leucine into trichloroacetic acid-precipitable material. Cytoplasmic RNA was extracted and analyzed by the dot blot procedure described above. Each measurement represents hybridization corrected for the cx-tubulin control. Northern (RNA) blot analyses. Cytoplasmic RNA (5 jig) was extracted from each cell line (9) and size fractionated on 1.2% agarose-formaldehyde gels. The RNAs were transferred to Zetabind nylon membranes (AMF Cuno) by capillary transfer, baked for 2 to 4 h at 80°C, and UV cross-linked by standard techniques (4). The blots were prehybridized for 5 min to several hours in hybridization buffer (50% formamide, 1% bovine serum albumin (fraction V), 1 mM EDTA, 0.5 M sodium phosphate [pH 7.2], 7% sodium dodecyl sulfate [SDS]). Hybridizations were for 17 to 24 h at 42°C in fresh hybridization buffer containing 1 x 108 to 8 x 108 cpm of randomly primed [32P]-labeled cDNA probe (specific activity, 0.5 x 109 to 1.0 x 109 cpm/,ug). The filters were washed in 2x SSC (SSC is 0.15 M NaCl plus 0.015 M sodium citrate)-0.1% SDS for 15 min at room temperature, in 0.lx
SSC-0.1% SDS for 15 min at room temperature, and in two changes of 0.lx SSC-0.1% SDS at 65°C for 30 min each. Autoradiography was for 1 to 5 days, using Kodak XAR or XRP film (Eastman Kodak Co.) and a single intensifying screen. The blots were stripped for reuse by boiling for 2 min in double-distilled water. Southern blot analyses. High-molecular-weight cellular DNA was extracted from each cell line as described previously (18). DNA (5 ,ug) was digested to completion with EcoRI (New England Bio-Labs, Inc.) and run through 0.7% agarose gels in 0.04 M Tris acetate-2 mM EDTA. The DNA was transferred to Zetabind or Magna Nylon 66 (Fisher Scientific Co.) by standard techniques (4, 36). Prehybridization and hybridization conditions were as described above for RNA blot analyses. TAT immunofluorescence. Cells were seeded onto 12mm-diameter glass cover slips and cultured in complete medium in the presence or absence of 10-6 M dexamethasone, 5 x 10-4 M dBtcAMP-1 mM theophylline, or both for 24 h. The cells were then fixed and stained essentially as described by Mevel-Ninio and Weiss (21), as follows. The cover slips were rinsed with phosphate-buffered saline and fixed in 3% formaldehyde for 1 min. The formaldehyde solution was diluted 1:1 with methanol and aspirated, and pure methanol was added. After 20 min at 4°C, the cover slips were rinsed in phosphate-buffered saline and incubated with rabbit anti-rat TAT antiserum for 30 min at 37°C in a humidified chamber. After two rinses with phosphatebuffered saline, bound immunoglobulin was stained by incubation with fluorescein-conjugated secondary antibody (sheep anti-rabbit immunoglobulin G; Organon Teknika). After three rinses in phosphate-buffered saline, the cover slips were mounted with buffered glycerol and examined under phase-contrast and epifluoresence illumination, using a Zeiss Axiophot photomicroscope.
,
RESULTS Properties of the cell lines. FH(17)1 is a hepatoma microcell hybrid that selectively retains human chromosome 17 derived from diploid fibroblasts. Its isolation and properties were similar to those of the HF(17)-series clones described previously (13). This particular hybrid was used as a prototype in the studies reported here because the population had a stable karyotype and contained few segregant cells. FH(17)1 cultures at early passage were challenged with medium containing bromodeoxyuridine, and TK- clones that had segregated human chromosome 17 arose at a frequency of 11%. Approximately 100 such clones were pooled to form a mass population of backselected cells designated FHB(17)1. The presence or absence of human chromosome 17 in FH(17)1 and FHB(17)1, respectively, was confirmed by cytogenetic analysis and by Southern marker analysis, using cloned probes from four loci on human chromosome 17: MYHI, D17SJ, D17S4, and ERBAI (data not shown). FH(17)1 populations retained a single copy of an apparently intact human chromosome 17 in >90% of the cells; FHB(17)1 cells had segregated that single human chromosome. The FF-series clones FF3-3 and FF5-1 are intertypic
hybrids formed by fusing highly differentiated rat hepatoma cells (FAO-1) with diploid mouse embryo fibroblasts; these hybrids have been characterized extensively (3, 13). Cytogenetic analysis has shown that these hybrids are karyotypically complete: they retain virtually all parental chromosomes. Liver-specific gene activity is extinguished in
HORMONAL EFFECTS AND EXTINCTION
VOL. 9, 1989
Tf rT
P EPCiL'(: Ix
i
I
Z-
A ,..
j
2839
II..
-
I
Z--
_g RNA...
_,,
.,_
7-
.-
.Z_
_
10. ;i).
-.
d. 410wo
1.25
0.62
0.31 0.16
_
_ -
_O
_~
_
4
:_ _
_~
_
0.08
0.04'
0.. 0212__L
.-
,
FIG. 1. Quantitation of basal PEPCK, TAT. and AS mRNA expression in a TSE1+ microcell hybrid and its backselectant. Serial twofold dilutions of cytoplasmic RNA from FH(17)1 and FHB(17)1 were applied to nylon membranes and hybridized with labeled cDNA probes from the PEPCK, TAT, AS, and oA-tubulin (TU) genes.
these hybrids, as judged by expression of 20 different liver mRNAs. Finally, TAT and serum albumin mRNA levels in these clones are repressed 500- to 1,000-fold relative to levels in parental hepatoma cells. Basal expression of PEPCK, TAT, and AS mRNAs in the presence of fibroblast TSEI. As indicated above, extinction in whole-cell hybrids is a quantitatively large effect, with tissue-specific mRNA levels being reduced 500- to 1,000fold. In contrast, extinguished mRNAs are generally repressed only 10- to 20-fold in microcell hybrid clones. These differences are largely due to the fact that microcell hybrids contain many more segregant cells than are generally present in whole-cell hybrid populations. These segregants have lost the donor chromosome under selection, are reexpressing previously extinguished functions, and are dying in the selective medium. The segregants typically constitute 5 to 15% of cells in microcell hybrid populations at steady state, and this is sufficient to account for the residual levels of gene activity detected in these clones (13). The presence of segregants in most microcell hybrids has precluded an accurate assessment of the true magnitude of TAT, PEPCK, and AS gene repression by fibroblast TSEI. To circumvent this problem, we quantitated basal expression of TAT, PEPCK, and AS mRNAs in a microcell hybrid that was largely free of reexpressing segregant cells. Serial twofold dilutions of FH(17)1 and FHB(17)1 cytoplasmic RNAs were applied to nylon membranes, and the filters hybridized with labeled PEPCK, TAT, AS, and cx-tubulin cDNA probes (Fig. 1). Specific hybridization of the PEPCK, TAT, and AS probes to FHB(17)1 RNA was detected over a wide concentration range. In contrast, hybridization of these probes with FH(17)1 RNA was detected only at significantly higher RNA concentrations. Hybridization of the a-tubulin probe was similar at equivalent concentrations of FH(17)1 and FHB(17)1 RNA. We estimate that PEPCK mRNA levels in FH(17)1 were decreased 64- to 128-fold relative to those of its FHB(17)1 backselectant, whereas TAT and AS mRNA levels were reduced 8- to 16-fold. Thus, it appeared that TSEJ had a greater effect on PEPCK mRNA expression than on accumulation of TAT or AS mRNA. This result indicated
that segregants alone could not account for the residual gene activity in FH(17)1, as segregants would contribute equally to the accumulation of all three mRNAs. Furthermore, the -100-fold reduction of PEPCK mRNA expression suggested that fewer than 1% of cells in this population were segregants, a value in agreement with the frequency of TK- cells determined by selective challenge with bromodeoxyuridine. Thus, residual TAT, AS, and possibly PEPCK gene expression was detected in the presence of fibroblast TSEJ. We next assayed whether gene activity could be induced by the positive factors that control expression of these genes in liver. Hormonal induction of TSEJ-repressed genes. The effects of TSEI on TAT, PEPCK, and AS mRNA accumulation are largely mediated at the level of gene transcription, as judged by nuclear run off assays (unpublished data). Furthermore, both glucocorticoids and cyclic nucleotides induce expression of these genes by increasing the rate of transcription initiation (29, 30). Therefore, we determined whether either of these positive regulators of PEPCK and TAT gene transcription could induce gene activity in the presence of fibroblast TSEJ. The kinetics of PEPCK, TAT, and AS mRNA accumulation in cells induced with dexamethasone or cAMP were determined by an RNA dot blot assay. Cytoplasmic RNA was extracted from the primary clone [FH(17)1] and its backselectant [FHB(17)1] after various intervals in the presence of either dexamethasone or dBtcAMP plus theophylline. The RNA (3 ,ug) was applied to nylon membranes and hybridized with radioactively labeled cDNA probes. Duplicate dots from each time point were quantitated by liquid scintillation counting. Each value represents the amount of specific hybridization to the TAT, PEPCK, or AS cDNA probe normalized for hybridization to the ot-tubulin control, whose expression remained constant throughout the time course of the experiment in both cell lines. Results of the dexamethasone inductions are shown in Fig. 2. In the absence of TSEJ [i.e., in FHB(17)1], induction of all three mRNAs was detected within 2 to 4 h of hormone addition, and steady-state levels increased 10- to 20-fold
2840
THAYER AND FOURNIER
0.4
6 5 4 3
PEPCK A .A A
Po, 2
Z $sz
MOL. CELL. BIOL.
,. .
11'11(1'f
1,,:
10
)
20
l
30
44
A l
A
1
TAT
A .A
.00
I
A
z V
A
* -l
A
...
K
.A'
(
z 10 6 r z 5 . AS 4. 3 * 2 )
20
A
S4
A
44 0 |~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ K1
30
A
A
U
K-I
U
-4A
a K1
1z
Iy
_"W
)
10
20
30
w
40
_m ___ Llk
TIME (hours) FIG. 2. Dexamethasone induction of PEPCK, TAT, and AS mRNAs in a TSE1+ microcell hybrid (L) and its backselectant (A). Cells were incubated in medium containing 10-6 M dexamethasone for various intervals, and RNA was prepared and analyzed as described in Materials and Methods. RNA concentrations were normalized to a-tubulin mRNA levels, which remained constant throughout the experiment.
within 24 h. This induction phenotype is typical of rat hepatoma lines of the H4IIEC3 family. However, a different response was observed in hepatoma cells retaining fibroblast TSEJ. Whereas both TAT and AS mRNA levels increased in FH(17)1 cells treated with dexamethasone, induced levels were only 15 to 30% those of the FHB(17)1 backselectant; that is, dexamethasone-induced gene activity was decreased in cells containing TSEJ. Induction of PEPCK mRNA under these conditions was barely detected. From these results, it appeared that TAT and AS expression was partially dexamethasone inducible in the presence of fibroblast TSEJ, whereas PEPCK gene activity was refractory to glucocorticoid induction under these conditions. Furthermore, the magnitude of the TAT and AS responses in FH(17)1 was such that these responses could not have been due solely to segregant cells in the population. These induction phenotypes were very different from those of karyotypically complete whole-cell hybrids, which were refractory to glucocorticoid induction (13; see below). Induction of these genes by cAMP was even more dramatic. A rapid increase in steady-state levels of all three mRNAs occurred in both the primary hybrid and its backselectant when cells were incubated in medium containing
U
U
I
mA
FIG. 3. cAMP induction of PEPCK, TAT, and AS mRNAs in a TSE1+ microcell hybrid (U) and its backselectant (A). Cells were incubated in medium containing 5 x 10-4 M dBtcAMP-1 mM theophylline for various intervals and analyzed as described in Materials and Methods. RNA concentrations were normalized to o-tubulin mRNA levels. Insets show the early kinetics of induction in the TSE1+ microcell hybrid [FH(17)1] after incubation in medium with dBtcAMP plus theophylline for 0, 0.5, 1.0. 1.5, 2, 3, 4, 5, 6, 8, and 10 h.
dBtcAMP plus theophylline (Fig. 3). Although basal levels of PEPCK mRNA differed by about 100-fold in FH(17)1 versus FHB(17)1, cAMP treatment abolished this difference: within 2 h of cAMP addition, PEPCK mRNA levels were comparable in the two cell lines. Similar results were obtained for the TAT and AS genes; the differences in basal expression in the primary hybrid and its backselectant ('10-fold) were eliminated by incubating the cells in the presence of cyclic nucleotide. Thus, cAMP induced expression of the PEPCK, TAT, and AS genes to the same high levels in the presence or absence of fibroblast TSE1. dBtcAMP dose response. To determine whether the ability
HORMONAL EFFECTS AND EXTINCTION
VOL. 9, 1989
2841
0 80
A
6040-
20-
-6
-5
-4
-3
-6
-5
-4
-3
-6
-5
-4
-3
log [cAMP] (M) FIG. 4. cAMP dose response in a TSE1+ microcell hybrid (-) and its backselectant (A). Cells were incubated in medium containing the indicated concentrations of dBtcAMP for 4 h, and RNA was prepared and analyzed as described in Materials and Methods. RNA levels are expressed relative to maximum induction of PEPCK (A), TAT (B), and AS (C) achieved in each cell line.
of the cells to respond to suboptimal concentrations of cAMP was somehow altered by the presence of fibroblast TSEJ (resulting in lower basal expression), dose-response experiments were performed. The FH(17)1 primary hybrid and its FHB(17)1 backselectant were incubated in the presence of various concentrations of dBtcAMP for 4 h, and cytoplasmic RNA was extracted. Levels of PEPCK, TAT, and AS mRNAs were quantitated as described above. Inductions of all three genes were virtually identical in the two cell lines at dBtcAMP concentrations between 0.05 and 5 mM (Fig. 4). We conclude that the ability of the cells to respond to cAMP was not altered by fibroblast TSEI. To determine whether changes in endogenous cAMP levels might be responsible for the decrease in basal PEPCK, TAT, and AS mRNA expression in the presence of fibroblast TSEJ, cAMP concentrations within the cells were measured. Whole-cell extracts from induced (0.5 mM dBtcAMP, 4 h) and uninduced cells were prepared, and the cAMP concentrations in the extracts were determined by radioimmunoassay (RIANEN; Dupont, NEN Research Products). This assay detected both cAMP and dBtcAMP with equal efficiency. The concentration of cAMP in uninduced FH(17)1 cells was 17.2 pM/mg of cell protein, and this increased to 121 pM/mg of protein after incubation of the cells in medium containing 0.5 mM dBtcAMP for 4 h. Corresponding values for uninduced FHB(17)1 cells were 16.9 and 188 pM/mg of protein. These results indicated that intracellular cAMP concentrations were not affected by the presence of fibroblast TSEJ; the two cell lines contained similar cAMP concentrations both before and after induction. Induction of extinguished genes does not require protein synthesis. Glucocorticoid induction of most genes is a primary response; i.e., protein synthesis is not required (41). Similarly, cyclic nucleotide induction of PEPCK mRNA synthesis does not require protein synthesis (29). To determine whether active versus TSEI-repressed genes could be discriminated on the basis of protein synthesis requirements for induction, cycloheximide experiments were performed. Cells were induced with dBtcAMP in the presence or absence of 10 ,uM cycloheximide for 2 or 4 h. This concen-
tration of cycloheximide inhibited protein synthesis by >90% at both times, as judged by incorporation of [3H]leucine into acid-precipitable material. PEPCK and TAT mRNAs were induced in both the primary hybrid and its backselectant after 2 and 4 h in medium containing dBtcAMP, and induction was not affected by cycloheximide (Fig. 5). AS mRNA was induced 3- and 10-fold in FH(17)1 cells after 2 and 4 h in medium containing dBtcAMP, whereas little induction in the backselectant was detected at these early times. Again, the presence of cycloheximide in the medium had no effect on the induction of AS mRNA. Thus, protein synthesis was not required for induction of these genes in either the presence or the absence of fibroblast TSEJ. Therefore, transcriptional activation of extinguished genes by dBtcAMP must be a primary response
mediated through preexisting factors. Induction of TAT protein in cells containing fibroblast TSE1. The results summarized above indicate that TAT
expression was reduced -10-fold in the presence of fibroblast TSEJ but that gene activity could be induced with either glucocorticoids or cAMP. The magnitude of these responses was such that it seemed likely that most cells of the hybrid population were responding to the inducer. However, it remained possible that a subpopulation of highly inducible cells accounted for the inducible phenotype. To distinguish between these possibilities, we determined the fraction of cells expressing TAT under inducing conditions by immunofluorescence. FH(17)1 and its backselectant were plated onto cover slips and incubated with dexamethasone, dBtcAMP, or both for 24 h. The cells were fixed and stained, using a TAT-specific primary antibody and a fluorescein-labeled secondary antibody. No specific labeling of either the primary clone or its backselectant was detected with uninduced cells. Thus, TAT specific activities of 90% of the
2842
THAYER AND FOURNIER
MOL. CELL. BIOL.
C
B
A 0
C)0.9
0.6
C.) Z
0.3
0
2
4
2
4
2
4
TIME (hours) FIG. 5. Protein synthesis requirements for cAMP induction of PEPCK (A), TAT (B), and AS (C) mRNAs in a TSE1+ microcell hybrid and its backselectant. RNA from FH(17)1 cells incubated with 5 x 104 M dBtcAMP in the presence (i) or absence (I) of 10 ,uM cycloheximide for 0, 2, or 4 h was prepared and analyzed as described in Materials and Methods. RNA from FHB(17)1 cells treated with dBtcAMP plus cycloheximide (LI) or dBtcAMP alone (Fl) was analyzed similarly.
cells in each population were clearly expressing TAT (Fig. 6). In dexamethasone-treated cultures, TAT-specific immunofluorescence was evident in >80% of the cells of either FH(17)1 or FHB(17)1 populations. In this case, however, the intensity of staining was clearly less in the primary hybrid than in the backselectant. The primary hybrid did contain a small fraction of cells (=1%) with backselectant levels of fluorescence; we presume that these cells were segregants. In the presence of both inducers, >90% of cells in each population were highly fluorescent. As TAT was clearly induced in most FH(17)1 cells by either dexamethasone or cAMP, induction of gene activity was necessarily occurring in cells that retained fibroblast TSEJ. Extinction of inducible and basal expression in whole-cell hybrids. Previous studies indicated that both basal and dexamethasone-inducible TAT expression were extinguished in karyotypically complete hepatoma x fibroblast hybrids (13). Extinction in this context corresponded to a >500-fold reduction in steady-state levels of TAT mRNA (3). This behavior contrasts with that of microcell hybrids, as described above. To more fully document these apparent differences, we compared the glucocorticoid and cyclic nucleotide induction phenotypes of microcell versus wholecell hybrids. FF5-1 and FF3-3 are intertypic hepatoma x fibroblast hybrids with essentially complete karyotypes; each contains approximately 50 chromosomes derived from its rat hepatoma parent plus 20 to 30 mouse fibroblast chromosomes. Detailed karyotyping has shown that these hybrids retain each of the 20 different mouse chromosomes in >90% of the cells (13). Furthermore, mouse chromosome 11 was present in 23 of 24 FF5-1 metaphases and 35 of 36 FF3-3 metaphases examined. Most of the cells in each hybrid population (23 of 23 and 30 of 35 metaphases for FF5-1 and FF3-3, respectively) retained only a single copy of that particular mouse chromosome. Expression of PEPCK, TAT, and AS mRNAs in these hybrids under various inducing conditions was assayed by RNA blot hybridization. The cells were incubated in serumfree medium containing dexamethasone, dBtcAMP, or both
for 8 h, and RNA was extracted and analyzed as described in Materials and Methods. FAO-1 parental hepatoma cells expressed PEPCK, TAT, and AS mRNAs, and expression was inducible by both glucocorticoids and cAMP (Fig. 7). In contrast, PEPCK and TAT mRNAs were not detectable in mouse embryo fibroblasts under any conditions. AS mRNA was expressed at low levels in MEF cells (90% of FH(17)1 cells under similar inducing conditions, karyotypically complete hybrids must retain a genetic factor(s) in addition to TSEI that results in a more complete extinction phenotype.
DISCUSSION
Tissue-specific gene activity is generally suppressed in intertypic hybrids formed by fusing different cell types of equal ploidy (reviewed in reference 6). The extinction phenotype displayed by such hybrids is remarkable in that most, if not all, tissue-specific transcription is suppressed (3). However, extinction is reversible, and the tissue-specific genes of one parent tend to be reexpressed as chromosomes of the other are lost (39, 40). These phenomena have formed a basis for attempts to identify genetic factors that regulate tissue-specific gene expression in trans. TSEJ was the first trans-acting locus of this type to be defined (13). Although clearly involved in PEPCK, TAT, and AS extinction in hepatoma microcell hybrids, TSEJ seems to require the activity of other genetic factors for complete repression of these genes. This conclusion is supported by two main observations. First, the magnitude of PEPCK, TAT, and AS mRNA repression was less in microcell hybrids (10- to 100-fold) than in whole-cell hybrids (>500- to 1,000-fold), and segregant cells alone could not account for the residual expression in microcell hybrid clones. Second, the PEPCK and TAT genes could be induced by dexamethasone and dBtcAMP in TSEJ-containing microcell hybrids but not in karyotypically complete whole-cell hybrids. Furthermore, whole-cell hybrids acquired an extinction phenotype similar to that of microcell hybrids as fibroblast chromosomes were lost. Several observations suggest that conversion of whole-cell hybrids from the noninducible to the inducible phenotype involved the segregation of genetic loci distinct from TSEJ. First, most cells in fully extinguished hybrid populations retained only a single copy of mouse chromosome 11, as did microcell hybrids and hybrid segregants. Thus, TSEJ gene dosage cannot account for the different extinction phenotypes of these clones. Second, hybrid segregants displayed an induction phenotype that was more similar to that of microcell hybrids (TSE1+) than parental hepatoma cells (TSEl-). For example, cAMP and dexamethasone induced similar levels of PEPCK mRNA in hepatoma cells, but only cAMP induced significant PEPCK mRNA accumulation in microcell hybrids. Hybrid segregants (e.g., FF5-1, passage 13) displayed the latter phenotype. Similarly, dexamethasone induced more AS mRNA than did cAMP in hepatoma cells, but the reverse was true in microcell hybrids and hybrid segregants. Third, although the induction phenotype of hybrid FF5-1 changed between passages 5 and 13, basal expression of AS mRNA was constant (and low) during this interval. This finding suggests that TSEJ was not being segregated from the cells. Finally, TAT immunofluorescence data clearly demonstrated that whole-cell hybrids contained both cAMP-inducible as well as noninducible cells; TSEIcontaining microcell hybrids contained only cAMP-inducible cells. Although cAMP reversed TSEI-mediated extinction in microcell hybrids, it is unlikely that TSEJ functions through the cyclic nucleotide induction pathway. Results presented here show that neither intracellular cAMP levels nor cAMPinducible gene transcription was affected by fibroblast TSEJ. Furthermore, TSEI affected both basal and glucocorticoidinducible TAT expression, and neither of these activities is cyclic nucleotide dependent (1). Finally, cAMP-responsive elements lie within 120 base pairs of the PEPCK cap site (26, 35), whereas sequences required for TSE1 regulation map further upstream (unpublished observations). Thus, TSEI
VOL. 9, 1989
and cAMP affect PEPCK, TAT, and AS gene transcription in mechanistically distinct ways. The extinction phenotype of TSEJ-containing microcell hybrids is interesting. The PEPCK, TAT, and AS genes of these cells are in a functional state that is different from that of either parent. These loci are neither silent, as they are in fibroblasts, nor fully active, as in hepatic cells. However, expression of all three genes can be restored to hepatoma levels in response to a strong transcriptional inducer such as cAMP. This suggests that TSEJ-repressed genes are competent for transcription but that a factor(s) required for full basal activity is not functioning in the presence of fibroblast TSEI. According to this model, the TSEI-sensitive factor(s) would no longer be limiting for transcription under inducing conditions. Although TSEJ is likely to alter the array of transcription factors that interact with the PEPCK, TAT, and AS gene promoters in hybrid cells, no evidence for a direct interaction between the TSEJ product(s) and these genes has been obtained. Furthermore, it is not clear whether TSEI functions by generating a negative factor (necessarily leaky) that represses transcription or by neutralizing the expression or function (or both) of a positive factor required for gene activity. In this regard, it is interesting that growth hormone extinction seems to operate via repression of the gene encoding GHF-1, a primary activator of growth hormone gene transcription (20). Whether this is the sole mechanism operating in growth hormone extinction has yet to be determined (38). In any case, these results suggest that a primary function of extinguisher loci may be to regulate expression of genes encoding tissue-specific transcription factors. Further insight into this mechanism of gene control will require the isolation of individual TSE loci and their products. Finally, the results presented here demonstrate that extinction is a polygenic phenomenon. Not only are different tissue-specific genes controlled by distinct extinguisher loci (13, 24), but repression may require the action of more than one trans-acting locus. We do not yet know whether the two-layered extinction phenotype displayed by the PEPCK, TAT, and AS genes will prove general, nor is the genetic basis of full repression apparent. Indeed, hepatoma microcell hybrids retaining most chromosomes of the murine complement have been screened for PEPCK, TAT, and AS extinction phenotypes, and only chromosome 11 (carrying TSEJ) has been implicated. This may indicate that full repression can only be achieved in the presence of TSE1 or that the second-level phenotype itself is polygenic. Genetic resolution of these questions will require the analysis of cell populations with carefully constructed hybrid genotypes. ACKNOWLEDGMENTS We thank our colleagues for many helpful discussions, especially Hal Weintraub and Jon Cooper for their comments on the manuscript. We are also grateful to N. Cowan, R. Hanson, D. MathieuMahul, and G. Schutz for providing reagents. This study was supported by Public Health Service grant GM26449 from the National Institute of General Medical Sciences. R.E.K.F. is the recipient of an American Cancer Society faculty research award. LITERATURE CITED 1. Boney, C., D. Fink, D. Schlichter, K. Carr, and W. D. Wicks. 1983. Direct evidence that protein kinase catalytic subunit mediates the effects of cAMP on tyrosine aminotransferase synthesis. J. Biol. Chem. 258:4911-4918. 2. Chen, T. R. 1977. In situ detection of mycoplasma contamination in cell cultures by fluorescent Hoechst 33258 stain. Exp.
HORMONAL EFFECTS AND EXTINCTION
2845
Cell. Res. 104:255-262. 3. Chin, A. C., and R. E. K. Fournier. 1987. A genetic analysis of extinction: trans regulation of 16 liver-specific genes in hepatoma-fibroblast hybrid cells. Proc. Natl. Acad. Sci. USA 84: 1614-1618. 4. Church, G. M., and W. Gilbert. 1984. Genomic sequencing. Proc. Natl. Acad. Sci. USA 81:1991-1995. 5. Cowan, N. J., P. R. Dodner, E. U. Fuchs, and D. W. Cleveland. 1983. Expression of human alpha-tubulin genes: interspecies conservation of 3' untranslated regions. Mol. Cell. Biol. 3:
1738-1745. 6. Davidson, R. L. 1974. Gene expression in somatic cell hybrids. Annu. Rev. Genet. 8:195-218. 7. Davidson, R. L., B. Ephrussi, and K. Yamamoto. 1966. Regulation of pigment synthesis in mammalian cells as studied by somatic hybridization. Proc. Natl. Acad. Sci. USA 56:14371440. 8. Derman, E., K. Krauter, L. Walling, C. Weinberger, M. Ray, and J. E. Darnell. 1981. Transcriptional control in the production of liver-specific mRNAs. Cell 23:731-739. 9. Favaloro, J., R. Freisman, and R. Kamen. 1980. Transcription maps of polyoma virus-specific RNA: analysis by two-dimensional nuclease S1 gel mapping. Methods Enzymol. 65:718749. 10. Fournier, R. E. K., and J. A. Frelinger. 1982. Construction of microcell hybrid clones containing specific mouse chromosomes: application to autosomes 8 and 17. Mol. Cell. Biol. 2:539-552. 11. Ghisalberti, A. V., J. G. Steele, M. H. Cake, M. C. McGrath, and I. T. Oliver. 1980. Role of adrenaline and cyclic AMP in appearance of tyrosine aminotransferase in perinatal rat liver. Biochem. J. 190:685-690. 12. Holt, P. G., and I. T. Oliver. 1969. Studies on the mechanism of induction of tyrosine aminotransferase in neonatal rat liver. Biochemistry 8:1429-1437. 13. Killary, A. M., and R. E. K. Fournier. 1984. A genetic analysis of extinction: trans-dominant loci regulate expression of liverspecific traits in hepatoma hybrid cells. Cell 38:523-534. 14. Killary, A. M., T. G. Lugo, and R. E. K. Fournier. 1984. Isolation of thymidine kinase deficient rat hepatoma cells by selection with bromodeoxyuridine, Hoechst 33258, and visible light. Biochem. Genet. 22:201-213. 15. Lem, J., A. C. Chin, M. J. Thayer, R. J. Leach, and R. E. K. Fournier. 1988. Coordinate regulation of two genes encoding gluconeogenic enzymes by the trans-dominant locus Tse-J. Proc. Natl. Acad. Sci. USA 85:7302-7306. 16. Loose, D. S., P. A. Shaw, K. S. Krauter, C. Robinson, S. England, R. W. Hanson, and S. Gluecksohn-Waelsch. 1986. Trcans regulation of phosphoenolpyruvate carboxykinase (GTP) gene, identified by deletions in chromosome 7 of the mouse. Proc. Natl. Acad. Sci. USA 83:5184-5188. 17. Lugo, T. G., B. Handelin, A. M. Killary, D. E. Housman, and R. E. K. Fournier. 1987. Isolation of microcell hybrid clones containing retroviral vector insertions into specific human chromosomes. Mol. Cell. Biol. 7:2814-2820. 18. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 19. Mathieu-Mahul, D., D. Q. Xu, S. Saule, J. R. Lidereau, F. Galibert, R. Berger, M. Mauchauffe, and C. J. Larsen. 1985. An EcoRl restriction fragment length polymorphism (RFLP) in the human c-erbA locus. Hum. Genet. 71:41-44. 20. McCormick, A., D. Wu, J.-L. Castrillo, S. Dana, J. Strobl, E. B. Thompson, and M. Karin. 1988. Extinction of growth hormone expression in somatic cell hybrids involves repression of the specific trans-activator GHF-1. Cell 55:379-389. 21. Mevel-Ninio, M., and M. C. Weiss. 1981. Immunofluorescence analysis of the time-course of extinction, reexpression, and activation of albumin production in rat-hepatoma fibroblast heterokaryons and hybrids. J. Cell. Biol. 90:339-350. 22. Morris, S. M., C. L. Moncman, K. D. Rand, G. J. Dizikes, S. D. Cederbaum, and W. E. O'Brien. 1987. Regulation of mRNA levels for five urea cycle enzymes in rat liver by diet, cyclic
2846
23.
24.
25. 26.
27.
28.
29.
30.
31.
32.
THAYER AND FOURNIER
AMP, and glucocorticoids. Arch. Biochem. Biophys. 256:343353. Peterson, T. C., A. M. Killary, and R. E. K. Fournier. 1985. Chromosomal assignment and trans regulation of the tyrosine aminotransferase structural gene in hepatoma hybrid cells. Mol. Cell. Biol. 5:2491-2494. Petit, C., J. Levilliers, M.-O. Ott, and M. C. Weiss. 1986. Tissue-specific expression of the rat albumin gene: genetic control of its extinction in microcell hybrids. Proc. Natl. Acad. Sci. USA 83:2561-2565. Pitot, H. C., C. Peraino, P. A. Morse, and V. R. Potter. 1964. Hepatoma in tissue culture compared with adapting liver in vivo. Nat]. Cancer Inst. Monogr. 13:229-242. Quinn, P. G., T. W. Wong, M. A. Magnuson, J. B. Shabb, and D. K. Granner. 1988. Identification of basal and cyclic AMP regulatory elements in the promoter of the phosphoenolpyruvate carboxykinase gene. Mol. Cell. Biol. 8:3467-3475. Raiha, N. C. R., and J. Suikoren. 1968. Factors influencing the development of urea synthesizing enzymes in rat liver. Biochem. J. 107:793-795. Ruiz, J. P. G., R. Ingram, and R. W. Hanson. 1978. Changes in hepatic messenger RNA for phosphoenolpyruvate carboxykinase (GTP) during development. Proc. Nati. Acad. Sci. USA 75:4189. Sasaki, K., P. P. Cripe, S. R. K. Koch, T. L. Andreone, D. D. Petersen, E. G. Beale, and D. Granner. 1984. Multihormonal regulation of phosphoenolpyruvate carboxykinase gene transcription. J. Biol. Chem. 259:15242-15251. Schmid, E., W. Schmid, M. Jantsen, D. Mayer, B. Jastroff, and G. Schutz. 1987. Transcription activation of the tyrosine aminotransferase gene by glucocorticoids and cAMP in primary hepatocytes. Eur. J. Biochem. 165:499-506. Schmid, W., G. Muller, G. Schutz, and S. Gluecksohn-Waelsch. 1985. Deletions near the albino locus on chromosome 7 of the mouse affect the level of tyrosine aminotransferase mRNA. Proc. Natl. Acad. Sci. USA 82:2866-2869. Schneider, J. A., and M. C. Weiss. 1971. Expression of differentiated functions in hepatoma cell hybrids. 1. Tyrosine aminotransferase in hepatoma-fibroblast hybrids. Proc. Natl. Acad.
MOL. CELL. BIOL.
Sci. USA 68:127-131. 33. Sereni, F., F. T. Kenney, and N. Dretchmer. 1959. Factors influencing the development of tyrosine-a-ketoglutarate transaminase activity in rat liver. J. Biol. Chem. 234:609-612. 34. Shinomiya, T., G. Scherer, W. Schmid, H. Zentgraf, and G. Schutz. 1984. Isolation and characterization of the rat tyrosine aminotransferase gene. Proc. Natl. Acad. Sci. USA 81:13461350. 35. Short, J. M., A. Wynshaw-Boris, H. P. Short, and R. W. Hanson. 1986. Characterization of the phosphoenolpyruvate carboxykinase (GTP) promoter-regulatory region. II. Identification of cAMP and glucocorticoid regulatory domains. J. Biol. Chem. 261:9721-9726. 36. Southern, E. M. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98:503-517. 37. Sparkes, R. S., and M. C. Weiss. 1973. Expression of differentiated functions in hepatoma cell hybrids: alanine aminotransferase. Proc. Natl. Acad. Sci. USA 70:377-381. 38. Tripputi, P., S. Guerin, and D. D. Moore. 1988. Two mechanisms for extinction of gene expression in hybrid cells. Science 241:1205-1207. 39. Weiss, M. C., and M. Chaplain. 1971. Expression of differential functions in hepatoma cell hybrids: reappearance of tyrosine aminotransferase inducibility after loss of chromosomes. Proc. Natl. Acad. Sci. USA 68:3026-3030. 40. Weiss, M. C., R. S. Sparkes, and R. Bertolotti. 1975. Expression of differentiated functions in hepatoma cell hybrids. IX. Extinction and reexpression of liver-specific enzymes in rat hepatomachinese hamster fibroblast hybrids. Somat. Cell Genet. 1:2740. 41. Yamamoto, K. R. 1985. Steroid receptor regulated transcription of specific genes and gene networks. Annu. Rev. Genet. 19: 209-252. 42. Yoo-Warren, H., J. E. Monahan, J. Short, H. Short, A. Bruzed, A. Wynshaw-Boris, H. M. Meisner, D. Samols, and R. W. Hanson. 1983. Isolation of the gene coding for cytosolic phosphoenolpyruvate carboxykinase (GTP) from the rat. Proc. Natl. Acad. Sci. USA 80:3656-3660.
