effects of dexamethasone, heat shock, and serum responses on the inhibition of hsc70 synthesis by...

JOURNAL OF CELLULAR F'HYSIOLOGY 164:344355 (1995)

Effects of Dexamethasone, Heat Shock, and Serum Responses on the Inhibition of Hsc7O Synthesis by Antisense RNA in NIH 3T3 Cells

TING LI AND LAWRENCE E. HIGHTOWER* Department of Molecular and Cell Biology, The University of Connecticut,

Storrs. Connecticut 06269-3044

A dexamethasone (Dex)-inducible antisense RNA expression vector was constructed that contains the S'-untranslated region and one third of the coding sequence for the bovine hsc70 protein. This vector was used to transfect NIH 3T3 cells from which clonal cell lines expressing hsc70 antisense RNA were developed. Quantitative Northern blot analysis with strand-specific probes was used to demonstrate the Dex-inducible accumulation of hsc70 antisense RNA in proliferating cell cultures and the inhibition of hsc70 RNA levels. Surprisingly, antisense RNA was either much less effective in reducing the amounts of hsc70 RNA in Dex-treated cultures than in untreated controls or cells compensated by producing more hsc70 RNA in response to increasing amounts of antisense RNA. Hsc70 protein synthesis did not decrease in either Dex-treated or untreated cultures: it actually increased, again suggesting the activation of a compensatory response. In Dex-treated cultures subjected to heat shock, hsc70 antisense RNA blocked the induction of hsp70, indicating that newly synthesized RNA was targeted effectively before it became translationally active. To test this hypothesis further, Dex-treated cultures were made quiescent by serum deprivation and then restimulated with serum, which causes a burst of RNA and protein synthesis. Consistent with this hypothesis, increased synthesis of hsc70 was blocked in serum-stimulated cultures expressing antisense RNA. o 1905 Wiley-Liss. Inc.

The 70 kilodalton (kDa) heat shock cognate protein (hsc70) is a constitutively expressed member of a family of molecular chaperones which includes a stress-inducible nucleocytoplasmic member hsp7O and grp78 (BiP) which is located in the rough endoplasmic reticulum (ER). Hsc70 is thought to bind to nascent polypeptide chains to prevent misfolding and aggregation (reviewed in Hightower et al., 1994). As a component of protein-folding pathways, it functions in concert with accessory factors like the DnaJ homolog hsp40 and the chaperonin hsp6O in a large assembly in the cytoplasm of eukaryotic cells that appears to be a protein folding machine (Frydman et al., 1994). A second function of hsc70 is to chaperone proteins to organelles such as nuclei, mitochondria, and ER (reviewed in Brodsky and Schekman, 1994; Langer and Neupert, 1994). Most of the evidence for these roles comes from in vitro studies; there are only a few examples of data obtained from in vivo experiments (reviewed in Craig et al., 1994) and fewer still using mammalian cells (reviewed in Georgo- poulos and Welch, 1993).

We have used an antisense RNA approach to create pseudomutants of mammalian cells containing reduced amounts of hsc70. The long-term goal is to identify phenotypic alterations that will add to our understand- ing of the cellular functions of hsc70. Herein, we de- scribe the construction of an inducible antisense RNA expression vector and its effects on hsc7O RNA levels and hsc7O protein synthesis in cells in several different 1 1995 WILEY-LISS, INC

physiological states. Among existing antisense RNA and DNA strategies, nuclear expression of antisense RNA from an engineered antisense gene was chosen for this study because of the long half-life and abundance of mammalian hsc70. Antisense gene expression has been used in a number of experimental systems to in- hibit exogenous and endogenous gene expression (van der Krol et al., 1988; Helene and Toulme, 1990; Takayama and Inouye, 1990; Murray and Crockett, 1992). In mammalian cells, antisense RNA has been most successful in inhibiting expression of exogenous viral or bacterial genes or endogenous genes encoding low abundance regulatory proteins with relatively short half-lives, such as those encoded by protoonco- genes. The use of antisense RNA to target endogenous message as abundant as hsc70, which is approximately 25% of actin levels in cultured mammalian cells (Ting Li, unpublished observations), has been reported in two studies. In a transient assay, microinjection of an antisense actin gene or antisense RNA synthesized in vitro into cultured cells resulted in fewer actin microfil-

Received September 15,1994; accepted January 10, 1995 *To whom reprint requestsicorrespondence should be addressed at Department of Molecular and Cell Biology, The University of Connecticut, Storrs, CT 06269-3044.

Hsc70 ANTISENSE RNA 345

ament bundles (Izant and Weintraub, 1985). In another study, stable transformants constitutively expressing antisense RNA to y-actin and p-tubulin were isolated and shown to accumulate antisense RNA at levels similar to those of the highly abundant endogenous mRNA. However, no effect on actin mRNA level and protein synthesis nor on tubulin mRNA level was observed in these clones (Gunning et al., 1987).

Since the yeast homologs of hsc70 are known to be essential proteins, we decided to develop an inducible antisense vector to produce conditional pseudomutants. When we began construction of this vector, three relatively well-characterized inducible expression systems were available based on dexamethasone (Dex), metal- lothionein, and heat shock promoters. Since the latter two promoters require induction events that also induce the synthesis of stress proteins, a complication we wanted to avoid, the Dex-inducible promoter MMTV long terminal repeat (LTR) was chosen to express antisense RNA against hsc70. NIH 3T3 mouse embryo fibroblasts were chosen for this study because they have been used successfully to express antisense RNA from glucocorticoid-inducible promoters (Holt et al., 1986; Nishikura and Murray, 1987). We have reported recently that quiescent NIH 3T3 cells expressing hsc70 antisense RNA die faster than control cultures during serum deprivation, suggesting that hsc70 is essential in mammalian cells as well (Hightower and Li, 1994).

The mechanism of antisense RNA regulation of gene expression is not well understood, and this potentially powerful approach to revealing protein function is still in its infancy. It is generally believed that antisense RNA hybridizes to complementary endogenous RNA causing interference with nuclear processing such as splicing and export, and with mRNA translation in the cytoplasm. It is also thought that RNA:RNA duplexes formed in the nucleus, and possibly the cytoplasm as well, are rapidly degraded by double-strand-specific ribonuclease (dsRNase) or inactivated by duplex unwinding-modifying activities present in the nucleus, both of which contribute to the effectiveness of inhibition of target RNA by antisense RNA. In many cases, a large excess of antisense RNA over sense RNA is needed to achieve efficient inhibition. The effectiveness of antisense RNA also depends on the accessibility of both the antisense and endogenous RNA for efficient hybridization which is influenced by their secondary structure, protein associations, and other factors present in vivo which cannot be determined readily. The fact that antisense RNA can potentially block a t different steps during gene expression makes rational design of strategies for a given target rather difficult. Herein, we found unexpectedly that antisense RNA was either less effective in reducing hsc70 mRNA levels in Dex-treated cultures or cells mounted a compensatory response. An awareness of these responses should be helpful to other investigators designing inducible antisense vectors and antisense strategies.

MATERIALS AND METHODS Construction of recombinant vectors

Since constitutive expression of hsc7O antisense RNA could lower cellular hsc70 concentrations to a lethal

level, an inducible mammalian expression vector capa- ble of expressing rat hsc7O RNA was constructed. Plas- mid pRC62 containing full length rat cognate hsc70 cDNA (OMalley et al., 1985) was digested with the restriction enzyme Pvu 11, generating five fragments with the expected sizes. A 689 bp fragment containing the 22 bp leader sequence and the first 33% of the coding sequence from the rat hsc70 cDNA was isolated and cloned into a mammalian expression vector pMSG (Pharmacia Biotech, Piscataway, NJ). The vector contains a 1.4 kb LTR from the mouse mammary tumor virus. A Dex-inducible promoter (GRE) within the LTR regulates the expression of any gene inserted into the downstream multiple cloning site. The 689 bp Pvu I1 fragment was inserted into the Sma I site in the oppo- site orientation with respect to the promoter. The eukaryotic transcription unit is completed by the required SV40 early splicing region and SV40 polyadenylation sequences in the vector. A dominant selectable marker, the Escherichia coli gpt gene, is expressed from the SV40 early promoter to ensure selection of stably transformed cells. The resulting plasmid pASC689 shown in Figure 1 was characterized by extensive computer-as- sisted restriction mapping. Both pASC689 and pMSG were linearized with EcoRl at the SV40-pBR322 junc- tion prior to transfection to facilitate stable integration of the intact transcription unit.

To synthesize strand-specific RNA probes for detect- ing antisense and sense hsc7O RNA expression, vector pBPC2034-7 was generated by insertion of the full length hsc70 cDNA into plasmid pBluescript KS+ (Stratagene, La Jolla, CAI. This was accomplished by isolating a 419 bp Pvu 11-Hind I11 fragment and a 1,614 bp Hind 111-Dra I fragment of the hsc70 cDNA from plasmid pRC62 and inserting both into the Sma I site of pBluescript KS+ in a single ligation reaction. The resulting plasmid pBPC2034-7 contains the 2 kb hsc70 cDNA in sense orientation to the T7 promoter and in antisense orientation to the T3 promoter. All recombinant DNA techniques were performed conventionally (Sambrook et al., 1989).

Cell culture, transfection, and selection NIH 3T3 cells were cultured in Dulbecco’s modified

Eagle’s medium (DME) supplemented with 4.5 g/l glu- cose and 10% calf serum (designated standard growth medium) a t 37°C in a humidified incubator with a 5% C02 atmosphere. Cells were transfected by the calcium phosphate precipitation method (Gorman, 1985). Sta- ble transformants were selected using MAX (Mycophe- nolic acid, Aminopterin, Xanthine) selection (Mulligan and Berg, 1981), and colonies were isolated and sub- cloned using the cloning ring technique. The resulting clones were progagated into mass culture and maintained as stably transformed cell lines under MAX selection. Cells were grown in standard growth medium without selection before analyses. Dex (Sigma Chemi- cals, St. Louis, MO) was added to culture medium to induce antisense RNA expression. To obtain quiescent, serum-starved cells, confluent monolayer cultures were washed with DME and incubated in medium containing 0.5% calf serum. For restimulation, serum-deprived cells were refed with standard growth medium.

LI AND HIGHTOWER 346

Dexmethssone

Inducible Promo

I” n 0

SV40 Early Splice Region

Fig. 1. Antisense hsc7O expression vector pASC689. The construction of this antisense expression vector is described in Materials and Methods. The diagram was drawn to scale based on the complete nucleotide sequence assembled into the MacGene computer program.

Northern blot analysis Total cellular RNA was isolated using the guanidin-

ium thiocyanate extractionlcesium chloride gradient centrifugation method. Equal amounts of RNA from different samples were size fractionated in 1% formaldehyde gels, transferred to nylon membranes (Genescreen, Dupont NEN, Boston, MA), hybridized with 32P-labeled strand-specific RNA probes, and visualized by autoradiography. All procedures were accomplished using standard protocols (Sambrook et al., 1989).

The 32P-labeled strand-specific RNA probes were synthesized from plasmid pBPC2034-7, linearized in the polylinker region immediately downstream of the insert, using the BIOBASICA I1 in vitro transcription system (International Biotechnologies, Inc., New Ha- ven, CT). Probes for antisense RNA were synthesized from EcoRI-linearized pBPC2034-7 using T7 RNA polymerase, whereas probes for endogenous hsc70 RNA were transcribed from Bam HI-cut pBPC2034-7 using T3 RNA polymerase. The “P-labeled DNA probes for glyceraldahyde phosphate dehydrogenase (GPDH) were synthesized from a 1.5 kb Pst I fragment of pGPDH, containing the full length GPDH cDNA, using the random primed oligolabeling method. Total RNA loading and transfer were monitored by methylene blue staining of the membrane prior to hybridization (Her- rin and Schmidt, 1988; Sambrook et al., 1989).

Metabolic labeling and cellular protein extraction

To measure rates of protein synthesis, cells were pulse labeled for 30 min in Selectamine medium (Gibco- BRL Life Sciences, Grand Island, NY) containing 2% dialyzed calf serum, 0.15 Fgirnl methionine, and 100

FCi/ml of Trans-”S-methionine (ICN Biomedicals, Costa Mesa, CA). Cells were solubilized either in gel buffer (62.5 mM Tris-HC1, pH 6.8,2.3% SDS, 5% 2-mercaptoethanol, 10% glycerol) for sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), or in O’Farrell buffer (62.5 mM Tris-HC1, pH 6.8,9.5 M urea, 2% Nonidet P-40,1.6% pH 3.5-10.0 and 0.4% pH four to six ampholytes, 5% 2-mercaptoethanol) for isoelectric focusing (IEF) in the first dimension followed by SDS- PAGE as a second dimension. Equal amounts of radio- activity in trichloroacetic acid-precipitable material were used in subsequent analyses. To determine total amounts of cellular protein, cells were lysed with 2% Nonidet P-40 (in 20 mM Tris pH 7.61, protein contents of extracts were determined by the Bradford dye-binding assay, and equal amounts of protein were subjected to polyacrylamide gel analyses after addition of equal volumes of 2x concentrates of the :sample buffers described above.

Electrophoretic analyses of proteins One-dimensional SDS-PAGE using 7.5% polyacryla-

mide gels was performed according l,o Laemmli (19701, and two-dimensional IEF-SDS-PAGE using 7.5% polyacrylamide gels was performed according to O’Farrell (1975) with modifications of the running conditions (Hightower, 1980; Duncan and Hershey, 1984). Gels were either fluorographed (Laskey and Mills, 1975) to visualize 35S-pulse-labeled patterns or silver stained (Sigma silver stain kit) to detect total protein. Signals were scanned and quantified by laser densitometry (LKB Ultroscan XL [Piscataway, NJ] for one-dimensional gels and Molecular Dynamics [Sunnyvale, CAI Computing Densitometer Model 300A for two-dimensional gels). Experiments were repeated at least three times and representative films were quantified.

347 Hsc7O ANTISENSE RNA

- - - - - - - - + ' - + - + - + - + - + - + - + - D e x

M i 3 ? A E C 7 Q a i n 1 1 1 2 1 3 1 4 1 5 1 6

e H e o 9 0 +,Grp78 4 - H s c 7 0

+ A c t i n

+ 30 kd

Fig. 2. Hsc70 and total proteins from antisense and control cell cultures separated on polyacrylamide gels. Aliquots oftotal cell protein (2 pg) from antisense or control transformants were separated by SDS- PAGE and visualized by silver staining. Subconfluent cells were cultured in the presence (+) or absence (-1 of l pM Dex for 10 days, with a

RESULTS Screening of clones expressing antisense

hsc70 RNA Introduction of inducible antisense expression vector

pASC689 into NIH 3T3 fibroblasts and selection by MAX resulted in stable transformants from which clonal cell lines were developed. Control lines were developed using the parental vector pMSG lacking the antisense sequence. Each transformed line was cultured in the presence or absence of l-2pM Dex for a period of 10 days with a medium change every 3 days. The amount of hsc70 in each culture was determined by silver staining of total cell protein extracts separated by one-dimensional SDS-PAGE (Fig. 2). This assay was used as a method of primary screening for promising clones since the ultimate objective was to isolate cell lines with reduced hsc70 protein after induction of antisense expression. Small reductions in hsc7O levels were found in many antisense clones, such as 3A-la, after treatment with Dex compared to untreated cultures of the same clone. No reduction in hsc7O was observed in control clones under the same treatment, and there were no observable differences in hsc70 levels between uninduced antisense clones and control clones.

medium change every 3 days. Lanes 1-8, and 9-16 were from two separate gels run and developed under identical conditions. Purified bovine hsc70 (100 pg) was used as a molecular mass marker (lane MI. Positions of relevant proteins are indicated.

proteins were visualized by silver staining and the dried gels were quantified by laser densitometry. This analysis showed that the accumulation of hsc7O after Dex treatment was unaffected in control clone 2-la and was reduced in the antisense clone 3A-la by 15% and clone 3B-2a by 10%. The amount of grp78 in Dex- treated cultures was reduced in the control clone by 35%, in antisense clone 3A-la by 63%, and in clone 3B-2a by 38%. In addition, analyses using two-dimensional gels (not shown) revealed that a member of the hsp90 family known to be hormonally regulated was induced by Dex in all clones along with a 30 kDa protein. The induction of hsp90 provided a useful indicator of a successful glucocorticoid response in these cells.

The growth characteristics of these antisense cell lines were similar to those of controls. Although Dex treatment slightly increased cell doubling time for both antisense and control cells, no significant difference in growth rate was observed between induced antisense clones and Dex-treated controls. No consistent morpho- logical changes associated only with antisense clones were found. Based on the results of the initial screening, the antisense clones 3A-la and 3B-2a and the control clone 2-la were selected for further analyses.

The amount of grp78 also appeared to be lower in the same clones that showed decreases in hsc70 levels after Dex treatment.

These qualitative observations were later confirmed

Expression Of antisense hsc70 RNA a reduction in the levels of hsc7O RNA in

growing cultures €or several clones selected for detailed study using two- dimensional IEF-SDS-PAGE to separate proteins from equal amounts of cell extracts (data not shown). The

To examine the expression of antisense hsc7O RNA and endogenous hsc7O mRNA, total RNA of selected clones was isolated from cells cultured in the presence

348 L1 AND HIGHTOWER

A

B

. . . - + - + - + 1 2 3 4 5 6

Fig. 3. Northern blot analysis of antisense hsc7O RNA expression. Equal amounts (20 kg) of total RNA isolated from antisense clones pASC689 3B-2a, pASC689 3A-la, and control clone pMSG 2-la treated with (+I or without i-) Dex (2 mM, 7 days) were size fractionated in a formaldehyde gel. RNA was transferred to a nylon membrane and hybridized first with '"P-labeled antisense-specific RNA probes (A) and then with "'P-labeled GPDH DNA probes (B). Autora- diographs of the two hybridizations are shown containing a 1.7 kb band ofhsc7O antisense RNA in A and a 1.45 kb band of GPDH mRNA in B.

or absence of Dex (2 pM) for 7 days and analyzed by Northern hybridization using strand-specific probes. The results from hybridizations with an antisense-specific RNA probe and a GPDH DNA probe used to control for loading differences are shown in Figure 3. As expected, antisense hsc70 RNA was detected in both antisense clones but not the control clone as a 1.71 kb transcript representing accurate transcription from the MMTV LTR of the pASC689 transgene. The signals detected probably represent only a fraction of the antisense RNA expressed in vivo since antisense RNAs may be unstable and antisense-sense RNA duplexes are rapidly degraded. Antisense hsc70 RNA was detected in uninduced antisense clones (lanes 1 and 31, demonstrating a basal transcription activity from the MMTV LTR. This basal expression was stimulated by Dex, indicating that the hormone-regulated promoter was functioning (lanes 2 and 4). The relative amounts of antisense RNA obtained from the antisense clones were quantified by laser densitometry, normalizing to the amounts of either GPDH mRNA or 28s rRNA with similar results (see Fig. 5A). Clone 3B-2a showed a very low but detectable basal level of antisense RNA accumulation which was induced four to sevenfold in different experiments. Clone 3A-la had a basal accumula-

c H s c 7 0 m R N A

B +28S

C 1 8 S

Fig. 4. Northern blot analysis of endogenous hsc70 RNA accumulation. Equal amounts (10 ( ~ g ) of total RNA isolated from antisense clones pASC689 3B-2a, pASC689 3A-la, and control clone pMSG2-la treated with ( + ) or without i-) Dex (2 mM, 5 days) were size fractionated in a formaldehyde gel, transferred to a nylon membrane, and hybridized with '"P-labeled sense-specific RNA probes. A Autoradio- graph of the resulting blot is shown. B: As a control for RNA loading and transfer, the membrane was stained with inethylene blue prior to hybridization to visualize 18s and 285 rRNA and photographed.

tion which was about fivefold higher than that of clone 3B-2a, and which was induced three to fourfold after Dex treatment. The induced level of antisense RNA in clone 3A-la was about six times higher than that of clone 3B-2a.

The results of Northern blot hybridization with the sense-specific RNA probe are shown in Figure 4A along with the results of methylene blue staining of the Northern blot as an internal control for RNA loading and transfer (Fig. 4B). The endogenous hsc7O RNA was detected as a 2.45 kb mRNA, as was predicted from its cDNA sequence. A background signal of nonspecific hybridization to the 28s ribosomal RNA was also detected in Figure 4A and i t corresponded to the level of rRNA shown by methylene blue staining. The signal intensities of antisense (Fig. 3) andl endogenous (Fig. 4A) hsc70 RNA described above were not directly comparable since these were from two different hybridizations. However, when all experimental conditions were taken into account, the levels of endogenous hsc70 mRNA were estimated to be much greater than those of the antisense RNA.

349 Hsc70 ANTISENSE RNA

Quantification of the relative levels of antisense RNA and hsc70 mRNA is provided in Figure 5A and 5B, respectively. Under basal antisense expression (-Dex, filled columns), the steady state levels of hsc70 mRNA in clones 3B-2a and 3A-la were reduced to 70% and 36% of the control level, respectively. Under induced antisense expression (+ Dex, hatched columns), however, there was no further reduction in hsc70 mRNA but rather an approximately 15% increase above the levels achieved by basal antisense expression. There was a slight reduction in hsc70 RNA levels in the control clone due to Dex treatment.

The rate of hsc70 protein synthesis in these clones was estimated by radioisotopic incorporation during a brief labeling period. Subconfluent cultures grown in the presence or absence of Dex were metabolically pulse labeled with 35S-methionine, and cell lysates containing an equal amount of radiolabeled proteins were analyzed by two-dimensional IEF-SDS-PAGE (not shown). The resulting fluorograms were quantified by laser densitometry using actin for normalization and the results are shown in Figure 5C. Comparison between -Dex and +Dex treatments within the same clone showed that the rate of hsc7O protein synthesis showed little or no change in the control clone and antisense clone 3B-2a, and decreased in antisense clone 3A-la only about 17%. These small decreases in hsc70 synthesis in the antisense cells treated with Dex are consistent with the small decreases in hsc70 amounts used to initially identity these clones are described in Figure 2.

Correlation plots of hsc7O RNA levels and relative rates of protein synthesis against antisense hsc70 RNA levels for the control and the two antisense clones com- bined provided additional insight into the effects of Dex in growing cultures (Fig. 6). Antisense RNA was either much less effective in reducing cellular levels of hsc7O RNA in Dex-treated cultures or cells could at least partially compensate by increasing hsc70 mRNA levels (Fig. 6A). And as shown in Figure 6B, when rates of hsc70 synthesis among clones were compared after normalization to actin, the rates of hsc70 synthesis were not reduced and in fact they were elevated by a maximum of about 25% in both Dex-treated and untreated cultures expressing antisense RNA. This observation suggests that there is a compensatory increase in the rate of hsc7O synthesis in response to the reduction in hsc70 RNA.

Hsc7O antisense RNA blocked newly synthesized hsp7O mRNA in heat shocked cells treated

with Dex The effect of antisense hsc7O RNA on hsp70 family

protein synthesis in heat shocked cells was investi- gated. Cells grown in the presence or absence of Dex for 5 days were heat shocked at the sublethal temperature of 42°C for 60 minutes and pulse labeled with "S-methionine for another 30 minutes a t 42°C. Cell lysates containing equal amounts of radiolabeled proteins were separated on two-dimensional gels (Fig. 7). Fluoro- graphs of these gels were quantified by laser densitometry as shown in Figure 8 for hsc7O synthesis (Fig. 8A) and hsp7O synthesis (Fig. 8B). In vector-transformed control clone 2-la, the most dramatic change after heat shock was the induction of hsp70, which was not detect-

1.25

1

0.75

0.5

0.25

0 - Control Antisense Antisense

2-1 a 3B-2a 3A-1 a

Control Antisense Antisense 2-1 a 3B-2a 3 A - l a

0 .1254 u) CI .- c 0.1 3

$ 0.075

0.05

9 0.025

n

c Q z

I

Control Antisense Antisense 2-1 a 38-2a 3A-1 a

Fig. 5. Comparison of levels of antisense and endogenous hsc7O RNA along with rates of protein synthesis. A Quantification by laser densitometry of the autoradiograms of antisense RNA accumulation shown in Figure 3. B Quantification by laser densitometry of the autoradiograms of hsc70 RNA accumulation shown in Figure 4. The absorbance value of hsc70 mRNA was normalized to that of the 28s ribosomal RNA shown in Figure 4A. The absorbance values between A and B are not directly comparable. C: Quantification of rates of hsc70 protein synthesis. Proteins were solubilized from control and antisense cultures after pulse labeling with 35S-methionine and they were separated on two-dimensional polyacrylamide gels (IEF-SDS-PAGE). Flu- orograms were prepared and the hsc70 spots were quantified by laser scanning densitometry. The absorbance values of hsc70 were normalized to that of actin (actin value set to 1.0). Values from the -Dex (filled columns) and +Dex (hatched columns) treatments of the same clone were analyzed on the same gel and are directly comparable. Values between clones can also be compared under the assumption that actin synthesis between clones does not vary. Relative values in absorbance units were taken as an estimate ofrelative rates ofprotein synthesis. The black-filled columns are data from untreated cultures and the diagonally hatched columns are data from Dex-treated cultures.

able under normal growth conditions. After Dex treatments, the level of hsp70 induction was reduced by about 36%, whereas the rate of hsc70 synthesis was not

LI AND HI( 350 3HTOWER

2 A

1 i' 0.5

3 s cn cn Q)

c

.- 5 zi 0 r- 0 u) I

0'

Antisense RNA (AU)

0 0.25 0.5 0.75 1 1.25

0 0.25 0.5 0.75 1 1.25

Antisense RNA (AU)

Fig. 6. Correlation plots of Hsc70 RNA levels (A) and estimated relative rates of protein synthesis (B) vs. amounts of antisense RNA in the two antisense clones and the control clone (zero amounts). Data points were taken from the experiments shown in Figure 5. Dex- treated cultures are designated by open circles and untreated cultures by open squares.

significantly altered. The fact that Dex treatment could somewhat reduce the induction of hsp70 by heat shock suggested that the hormone may have induced some thermoprotection in these cultures. Under these conditions, thermal damage and thus the requirement for hsps would be reduced.

In antisense cells, there was a greater reduction in the heat-induced level of hsp70 synthesis under Dex treatment than in vector-transformed control cultures. The heat-induced rate of hsp70 synthesis under Dex was reduced by 61% in antisense clone 3A-la and by 45% in clone 3B-2a compared to the heat shocked cultures without Dex treatment (Fig. 8B). The rate of hsc70 synthesis was also slightly decreased by Dex treatment by about 24% in clone 3A-la and 13% in clone 3B-2a (Fig. 8A). The antisense clone that produced the highest amount of hsc70 antisense RNA (clone 3A-la) thus had the highest percent inhibition of both hsc70 and hsp70 synthesis. These data indicated that antisense hsc70 RNA, in addition to inhibiting

hsc70 expression, also targeted the highly homologous hsp70 RNA and partially blocked its induction during heat shock.

A comparison of the relative rates of hsc70 synthesis both with and without Dex and hsp7O synthesis without Dex among clones in Figure 8 indicated that the rates were higher in the two antisense clones compared to the control clone, again suggesting the presence of a compensatory response to the expression of hsc7O antisense RNA. Relative to the control clone, the rate of hsc70 synthesis was elevated by 50-60% and hsp7O by 20-30% in the antisense cultures without Dex treatment.

Antisense hsc7O RNA effectively blocked hsc7O synthesis directed by newly synthesized hsc7O

transcripts in serum-stimulated cultures Inhibition of hsp70 inductions by antisense hsc70

RNA in heated shocked cultures suggested that antisense RNA may be more efficient in targeting newly synthesized RNA transcripts than in blocking mRNA already engaged in protein synthesis. To test this hypothesis for new hsc70 transcripts without the complication of cross-hybridization of hsc70 antisense RNA to hsp70 transcripts, cells were made quiescent by serum deprivation of confluent cultures and then stimulated with serum to reenter the cell cycle. Our strategy in using serum deprivation was to reduce the synthesis of hsc70 as part of the down-regulation of protein synthesis under these conditions, and to build up a larger pool of antisense RNA. Serum stirnulation of quiescent cultures would cause a transient, rapid increase in hsc70 synthesis as part of an overall expansion of protein synthetic capacity (Takenaka and Hightower, 1993). Under these conditions, newly transcribed hsc7O RNA would be effectively targeted by antisense hsc70 RNA, if our hypothesis was correct.

Upon serum stimulation, the rate of hsc70 synthesis increased rapidly during the first 5 hours and subse- quently declined back to basal level by 48 hours (data not shown). As shown in the fluorograms of two-dimensional gels in Figure 9, this initial 5-hour increase was blocked effectively in cells induced for antisense expression by Dex. Whereas Dex treatment had little effect on the control clone (Fig. 9A,B), it substantially reduced the rate of hsc70 synthesis in the antisense clones (compare Fig. 9C with Fig. 9D and Fig. 9E with Fig. 9F). These fluorograms were quantified by laser densitometry and Figure 10 shows the relative rates of hsc7O synthesis, each measured by a 30-minute period of "S-methionine incorporation 5-5.5 hours after serum addition. At 48 hours of serum deprivation (Fig. lOA), there were no significant differences between antisense and control cells in their expression profiles. After 72 hours of serum deprivation (Fig. lOB), the serum-stimulated rates of hsc7O synthesis in uninduced (- Dex) antisense clones were significantly higher (3346%) than those in the untreated control clone. This increase was blocked in antisense clones treated with Dex to up-regulate antisense expression. By 96 hours of serum deprivation (Fig. lOC), inhibition of hsc70 up-regulation by serum was even greater. The serum-stimulated rate of hsc70 synthesis was reduced by 52% and 33% in Dex-treated antisense clone 3A-la

Hsc7O ANTISENSE RNA

- D e x

351

+ D e x

Fig. 7. Two-dimensional gel patterns of newly synthesized proteins from heat shocked cultures. Cells grown in the presence or absence of Dex (2 p M for 5 days) were incubated a t the heat shock temperature of 42°C for 60 minutes followed by metabolic pulse labeling with "S-methionine also at 42°C for 30 minutes, all in the presence or absence of Dex. Cell lysates were then analyzed by IEF-SDS-PAGE and the resulting fluorograms are shown for proteins from -Dex (A,C,E) and +Dex (B,D,F) cultures.

and 3B-2a, respectively, but it was reduced by only 13% in control clone 2-la. Therefore, the longer the cells were serum deprived, the more complete the inhibition to the induction of hsc70 synthesis was upon readdition of serum. Similar changes were observed for grp78, pre- sumably brought about by cross-hybridization of hsc7O antisense RNA to grp78 transcripts. The data also suggested that uninduced, basal antisense RNA expression during serum deprivation further reduced the level of hsc70, which was compensated by a greater increase in

the rate of hsc70 synthesis in the antisense clones compared to control cultures during subsequent serum stimulation (compare the -Dex columns in Fig. 1OC).

DISCUSSION Antisense RNA approaches have been used sparingly

in the study of stress proteins. In one study, hsp26 antisense RNA expression from an hsp7O promoter resulted in over 80% reduction in synthesis of Drosophila hsp26 (McGarry and Lindquist, 1986). However, no


A 0 3

v) CI .- -

0.2 8

z s

c m 0.1

0 .o Control Antisense Antisense

2-1 a 3A-la 3B-2a

0.3 v)

c 3

E m

CI .- 8 0.2

9 g 0.1

0.0

Fig. 8. Comparison of hsc7O and hsp70 protein synthesis under heat shock. The fluorograms in Figure 7 were quantified by laser densitometry as described in the Figure 5C legend. The absorbance units reflect rates of incorporation of radioactive methionine into protein during the pulse label period and are taken as an estimate of relative rates of protein synthesis for hsc70 (A) and hsp70 (B). The black-filled columns are data from untreated cultures and the diagonally hatched columns are data from Dex-treated cultures.

changes occurred in the patterns of protein synthesis during recovery from heat shock, an indicator of acquired thermotolerance, suggesting that either hsp26 is not an important determinant in this process or other members of the family can compensate for its reduction. The effect of inhibition of Drosophilia-inducible hsp70 synthesis during heat shock was also studied (Lind- quist et al., 1988). Heat-inducible expression of hsp7O antisense RNA suppressed the heat shock induction of hsp70 synthesis to below 25% of that observed in the vector control. In compensation, hsp70 synthesis was extended for a longer period of time during recovery along with the rate and duration of synthesis of hsp68, hsp83, and small hsps. The recovery of normal protein synthesis was also delayed in antisense cells after a severe heat shock. These results suggested a role for hsp70 in the regulation of the heat shock response.

There are only two examples of antisense RNA used to study mammalian stress proteins, both of which target the ER protein grp78 in Chinese hamster ovary (CHO) cells. In one study, cells were selected that constitutively expressed high levels of antisense RNA against grp78, resulting in significant reductions in levels of grp78 mRNA and protein (Dorner e t al., 1988). Reduction of grp78 in cells producing tissue plasmino- gen activator (tPA) led to increased secretion of tPA. In a second study (Li et al., 1992), CHO cells expressing high levels of antisense RNA against grp78 from a strong constitutive promoter (RSV LTR) were selected through coamplification with a DHFR gene. Despite the presence of high levels of antisense RNA, these cells

have a normal basal level of grp78 protein synthesis but produce a 10-fold higher level of endogenous grp78 mRNA, suggesting the presence of a compensatory response. However, the stress induction of grp78 synthesis by the calcium ionophore A23187 was blocked and the cells became sensitive to A23187 treatment.

The antisense sequence used in the present study was derived from the rat hsc70 cDNA and covers 22 bp of the leader sequence and the N-terminal one third of coding sequence. The 689 n t antisense sequence shares overall 95.4% nucleotide sequence identity to the mouse hsc70 mRNA, within which there are five regions of 50-79 n t in length with uninterrupted complete homology. Our results showed that this antisense sequence is effective in targeting hsc70 RNA in mouse cells. It has been shown previously in animal cells that antisense RNA of 45 n t and 52 n t is effective (Izant and Weintraub, 1984, 1985; Melton, 1985), and that all regions of homology need not be contiguous (Kim and Wood, 1985; Daugherty et al., 1989). Unexpectedly, reduction in cellular levels of hsc7O RNA did not result in decreased synthesis of hsc70 protein in either Dex- treated or untreated antisense cells in proliferating cultures. One interpretation is that initial reduction of hsc7O RNA levels due to expression of antisense RNA triggered a compensatory increase in hsc70 protein synthesis. Alternatively, hsc7O mRNA may be present in excess of that needed to sustain a high rate of hsc70 protein synthesis in growing cells.

Two hypotheses can be offered to esplain the data in Figure 6 showing that there is much more hsc70 RNA remaining in Dex-treated clone 3A-la than would be predicted using the correlation derived for untreated cultures which indicates a reduction in hsc70 mRNA to levels essentially undetectable by Northern blotting. Either antisense RNA is less effective in targeting hsc7O mRNA in Dex-treated cells, due to a change in target accessibility for example, or the cells increase their levels of hsc70 mRNA to compensate for losses due to inactivation by the antisense RNA. As noted above, CHO cells expressing high levels of antisense RNA against grp78 maintained normal basal levels of grp78 protein synthesis by producing a 10-fold higher level of grp78 mRNA as a compensatory response, so there is a precedent for the latter possibility (Li et al., 1992). Our results together with these provide evidence that the expression of the two constitutive members of the mammalian hsp70 family, the cytosolic hsc70 and the ER grp78, are both tightly autoregulated and resist pseudogenetic manipulation.

The reduction in grp78 synthesis observed in cells when hsc70 synthesis was inhibited is interesting. A trivial explanation is that the ant tsense hsc70 RNA also interacts with the grp78 transcript which is homologous to hsc70 particularly a t the N-terminal domains of the proteins. However, sequence analysis revealed only a 59.5% overall nucleotide sequence identity between rat hsc70 and rat grp78 in the region of antisense targeting, with the maximum length of continuous complete homology to only 17 nt. This degree of homology is probably below the level required for antisense- sense interaction. For example, antisense RNA to hsp26 did not affect the expression of two other mem-

Hsc7O ANTISENSE RNA 353

- D e x + D e x

7

a m

Fig. 9. Two-dimensional gel analysis of newly synthesized hsc7O and grp78 from quiescent cultures stimulated with serum. Confluent cultures were made quiescent by serum deprivation for 96 hours and then were stimulated with serum for 5 hours either in the absence (A,C,E) or presence (B,D,F) of Dex. Fluorograms of proteins labeled with 35S-methionine 5-5.5 hours after serum addition. Labeling conditions and analysis by IEF-SDS-PAGE were as described for Figure 7.

bers of the family, hsp23 and hsp28, which has 72% homology to the hsp26 in the 210 bp antisense target region (Lindquist et al., 1988). Thus the inhibition of grp78 synthesis may be due to other mechanisms. Per- haps hsc70 and grp78 synthesis are coordinately regulated. One functional link that may require a coordi- nate regulation would be in translocation of polypeptides across the ER membrane, where both proteins are required (Brodsky and Schekman, 1994).

Basal hsc70 antisense RNA expression in cells under heat shock resulted in increased hsp70 and hsc70 synthesis relative to control, indicating again a compensatory response. In contrast, the induction of hsp70 during heat shock was partially blocked under Dex- induced antisense expression, suggesting a cross-inhibition of newly induced hsp70 RNA by the antisense hsc7O RNA. We have directly compared the region of the rat hsc7O nt sequence which is complementary to our


Control 2-la Antisense 3A-la Antisense 3B-2a

2 In

Control 2-1 a Antisense 3A-la Antisense 3B-2a

Control 2-la Antlsense 3A-la Antisense 3B-2a

Fig. 10. The fluorograms shown in Figure 9 containing newly synthesized proteins from cultures deprived of serum for 96 hours (C) along with those from 48 hours (A) and 72 hours (B) of serum deprivation were quantified by laser densitometry as described in the legend of Figure 5C. The absorbance units reflect relative rates of protein synthesis. The data for hsc70 are represented by black-filled columns for -Dex cultures and diagonally hatched columns for +Dex cultures. The data for grp78 are represented by grey-filled columns for -Dex cultures and white columns for +Dex cultures.

antisense RNA sequence to that of mouse hsp7O (Gen- Bank Accession No. M35021). There is 45% nt sequence identity in the untranslated region and 76% identity in the translated region with the longest stretch of unbro- ken matches being 21 nt.

The extent of inhibition of hsc70 synthesis by Dex- induced antisense RNA was slightly greater in heat shocked cells than in unheated controls, despite the added cross-interaction of the antisense RNA to the hsp70 RNA. These data from heat shocked cells suggest that hsc70 antisense RNA is more effective in targeting newly synthesized transcripts than in blocking mRNA engaged in protein synthesis.

In heat shocked control clone 2-la treated with Dex, the induction of hsp70 was attenuated suggesting that heat produced less protein damage under these conditions resulting in a smaller demand for new chaperones. It is known that Dex treatment induces a small amount of thermotolerance in cultured mammalian cells in the absence of hsp70 induction (Fisher et al., 1986); and in NIH 3T3 cells, Dex induces aB-crystallin, a member of the small hsp family, along with increased

thermotolerance (Aoyama et al., 1993). The existence of a state of acquired thermotolerance in our Dex-treated cells would explain the attenuated heat shock response that was observed in cultures without antisense RNA.

In addition to evidence from our heat shock experiments and Lee’s work on inhibition of calcium ionophore- stimulated grp78 synthesis, suggesting that newly synthesized RNA is effectively targeted, we demonstrated an effective block of serum-stimulated increases in hsc70 synthesis in Dex-treated quiescent cells expressing antisense RNA. After serum stimulation, large amounts of newly synthesized and processed hsc70 RNA are present in the vertebrate nucleus. Serum stimulation causes a rapid increase in the steady state level of hsc7O mRNA up to sixfold within 5-24 hours in quiescent cells (Sorger and Pelham, 1987; Takenaka and Hightower, 1992, 1993). These newly synthesized nuclear hsc7O transcripts must be rapidly processed and transported out to the cytoplasm, and thus may open up many accessible targets for antisense:sense RNA interactions, leading to more efficient inhibition of hsc7O synthesis under these conditions. It has been shown directly in at least one study that target avail- ability for interaction with antisense RNA is a critical determinant for effective antisense i:nhibition. In primary oocytes, the 5’ and middle regions of the dormant maternal tPA mRNA is not accessible for antisense RNA interaction until meiotic maturation occurs and tPA mRNA becomes polyadenylated and translated. In contrast, the 3‘ region is accessible for hybrid formation at all stages (Strickland et al., 1988). Studies have shown that RNA (or mRNA) molecules in the nucleus are localized in discrete “transcript domains” containing proteinaceous structures (Carter et al., 1993; Xing et al., 1993). RNA a t these sites or in transit to the nuclear membrane may be in configurations that per- mit more effective interactions with hsc70 antisense RNA. Another possible contributor to more efficient antisense RNA activity is the cell cycle regulated expression of double-stranded RNA unwindingimodifying activity in nuclei of mouse 3T3 fibroblasts (Wagner and Nishikura, 1988). This activity increases sharply upon serum stimulation of quiescent cells and exhibits an intermediate level of activity in continuously proliferating cells. It has been suggested that the modification of adenosine to inosine residues in thie RNA duplex by this “unwindase” gradually unwinds the duplex, resulting in the loss of ability of both strands to rehybridize and the loss of translatability of the modified mRNA (Bass and Weintraub, 1988; Wagner et al., 1989; Bass, 1992).

ACKNOWLEDGMENTS We thank Carol Norris for suggestions on the manu-

script. This work was supported by grants from the National Science Foundation and the National Insti- tutes of Health. We benefited from use of the Cell Cul- ture Facility of the University of Connecticut Biotech- nology Center.

LITERATURE CITED Aoyama, A., Frohli, E., Schafer, R., and Klemenz, R. (1993) nB-Crys-

tallin expression in mouse NIH 3T3 fibroblasts: Glucocorticoid re- sponsiveness and involvement in thermal protection. Mol. Cell. Biol., 13;1824-1835.

Hsc70 ANTISENSE RNA 355

Bass, B.L. (1992) The double-stranded RNA unwindinglmodifying activity. In: Antisense RNA and DNA. J.A.H. Murray, eds. Wiley- Liss, Inc., New York. pp. 159-174.

Bass, B.L., and Weintraub, H. (1988) An unwinding activity that covalently modifies its double-stranded RNA substrate. Cell, 55: 1089-1098.

Brodsky, J.L., and Schekman, R. (1994) Heat shock cognate proteins and polypeptide translocation across the endoplasmic reticulum membrane. In: The Biology of Heat Shock Proteins and Molecular Chaperones. R.I. Morimotor, A. Tissieres, and C. Georgopoulos, eds. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. pp. 85-110.

Carter, K.C., Bowman, D., Carrington, W., Fogarty, K., McNeil, J.A., Fay, F.S., and Lawrence, J.B. (1993) A three-dimensional view of precursor messenger RNA metabolism within the mammalian nucleus. Science, 259t1330-1335.

Craig, E.A., Baxter, B.K., Becker, J . , Halladay, J., and Ziegelhoffer, T. (1994) Cvtosolic hso70s of Saccharomvces cereuisiaet Roles in oro- tein synchesis, protein translocation, pioteolysis, and regulation In: The Biology of Heat Shock Proteins and Molecular Chaperones. R.I. Morimoto, A. Tissieres, and C. Georgopoulos, eds. Cold Spring Har- bor Laboratory, Cold Spring Harbor, New York, pp. 31-52.

Daughtery, B.L., Hotta, K., Kumar, C., Ahn, Y.H., Zhu, J.D., and Pestka, S. (1989) Antisense RNA: Effect of ribosome binding sites, target location, size, and concentration on the translation of specific mRNA molecules. Gene. Anal. Tech., 6:l-16.

Dorner, A.J., Krane, M.G., and Kaufman, R.J. (1988) Reduction of endogenous GRP78 levels improves secretion of a heterologous protein in CHO cells. Mol. Cell. Biol., 8:40634070.

Duncan, R., and Hershey, W.B. (1984) Evaluation of isoelectric focusing running conditions during two-dimensional isoelectric focusing/ sodium dodecyl sulfate-polyacrylamide gel electrophoresis: Varia- tion of gel patterns with changing conditions and optimized isoelectric focusing conditions. Anal. Biochem., 138:144-155.

Fisher, G.A., Anderson, R.L., and Hahn, G.M. (1986) Glucocorticoid- induced heat resistance in mammalian cells. J . Cell. Physiol., 128: 127-132.

Frydman, J., Nimmesgern, E., Ohtsuka, K., and Ulrich Hartl., F. (1994) Folding of nascent polypeptide chains in a high molecular mass assembly with molecular chaperones. Nature, 370.111-117.

Georgopoulos, C., and Welch, W.J. (1993) Role of the major heat shock proteins as molecular chaperones. Annu. Rev. Cell Biol., 9r601-634.

Gorman, C. (1985) High efficiency gene transfer into mammalian cells. In: DNA Cloning: A Practical Approach. D.M. Glover, eds. IRL Press, Oxford, Washington, DC, pp. 143-190.

Gunning, P., Leavitt, J., Muscat, G., Ng, S.-U., and Kedes, 1. (1987) A human 6-actin expression vector system directs high-level accumulation of antisense transcripts. Proc. Natl. Acad. Sci. USA, 84:4831- 4835.

Helene, C., and Toulme, J.-J. (1990) Specific regulation of gene expression by antisense, sense and antigene nucleic acids. Biochem. Bio- phys. Acta, 1049:99-125.

Herrin, D.L., and Schmidt, G.W. (1988) Rapid, reversible staining of Northern blots prior to hybridization. Biotechniques, 6~196-200.

Hightower, L.E. (1980) Cultured animal cells exposed to amino acid analogues or puromycin rapidly synthesize several polypeptides. J. Cell. Physiol., 102:407427.

Hightower, L.E., and Li, T. (1994) Structure and function of the mammalian hsp7O family. In: Heat Shock Proteins in the Nervous Sys- tem. R.J. Mayer and I.R. Brown, eds. Academic Press, London, pp. 1-30.

Hightower, L.E., Sadis, S.E., and Takenaka, I.M. (1994) Interactions of vertebrate hsc70 and hsp70 with unfolded proteins and peptides. In: The Biology of Heat Shock Proteins and Molecular Chaperones. R.I. Morimoto, A. Tissieres, and C. Georgopoulos, eds. Cold Spring Harbor Laboratory Cold Spring Harbor, New York, pp. 179-208.

Holt, J.T., Gopal, T.V., Moulton, A.D., and Nienhuis, A.W. (1986) Inducible production of c-fos antisense RNA inhibits 3T3 cell prolif- eration. Proc. Natl. Acad. Sci. USA, 83t4794-4798.

Izant, J.G., and Weintraub, H. (1984) Inhibition of thymidine kinase gene expression by anti-sense RNA: A molecular approach to ge- netic analysis. Cell, 36:1007-1015.

Izant, J.G., and Weintraub, H. (1985) Constitutive and conditional suppression of exogenous and endogenous genes by antisense RNA. Science. 229:345352.

Kim, S.K., and Wood, B.J. (1985) Stable reduction ofthymidine kinase activity in cells expressing high levels of antisense RNA. Cell, 42: 129-138.

Laemmli, U.K. (1970) Cleavage of the structural proteins during the assembly of the head of bacteriophage T4. Nature, 277r680-685.

Langer, T., and Neupert, W. (1994) Chaperoning mitochondria1 bio- genesis. In: The Biology of Heat Shock Proteins and Molecular Chaperones. R.I. Morimoto, A. Tissieres, and C. Georgopoulos, eds. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp. 53-84.

Laskey, R.A., and Mills, A.D. (1975) Quantitative film detection of 3H and I4C in polyacrylamide gels by fluorography. Eur. J. Biochem., 56:335--341,

Li, L.-J., Li, X., Ferrario, A,, Rucker, N., Liu, E.S., Wong, S., Gomer, C.J., and Lee, A.S. (1992) Establishment ofa Chinese hamster ovary cell line that expresses grp78 antisense transcripts and suppresses A23187 induction of both GRP78 and GRP94. J. Cell. Physiol., 153: 575-582.

Lindquist, S., McGarry, T.J., and Golic, K. (1988) Use of antisense RNA in studies of the heat-shock response. In: Antisense RNA and DNA. D.A. Melton, eds. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp. 71-77.

McGarry, T.J., and Lindquist, S. (1986) Inhibition of heat shock protein synthesis by heat-inducible antisense RNA. Proc. Natl. Acad. Sci. USA, 83t399-403.

Melton, D.A. (1985) Injected anti-sense RNAs specifically block messenger RNA translation in vivo. Proc. Natl. Acad. Sci. USA, 82t144- 148.

Mulligan, R.C., and Berg, P. (1981) Selection for animal cells that express the Escherichia coli gene coding for xanthine-guanine phos- phoribosyltransferase. Proc. Natl. Acad. Sci. USA, 78t2072-2076.

Murray, J.A.H., and Crockett, N. (1992) Antisense techniques: An overview. In: Antisense RNA and DNA. J.A.H. Murray, ed. Wiley- Liss, Inc., New York, pp. 1 4 9 .

Nishikura. K., and Murray, J.M. (1987) Antisense RNA of oroto-onco- gene c-fos blocks renewed growth of quiescent 3T3 cells: Mol. Cell. Biol., 7t639-649.

OFarrell, P.H. (1975) High resolution two-dimensional electrophoresis of proteins. J . Biol. Chem., 250:4007-4021.

OMalley, K., Mauron, A., Barchas, J.D., and Kedes, L. (1985) Consti- tutively expressed rat mRNA encoding a 70-kilodalton heat shock- like protein. Mol. Cell. Biol., 5t34763483.

Sambrook, J., Maniatis, T., and Fritsch, E.F. (1989) Molecular Clon- ing (a Laboratory Manual). Cold Spring Harbor, Laboratory Cold Spring Harbor, New York.

Sorger, P.K., and Pelham, H.R.B. (1987) Cloning and expression of a gene encoding hsc73, the major hsp-like protein in unstressed rat cells. EMBO J., 6t993-998.

Strickland, S., Hurate, J., Belin, D., Vassalli, A,, Rickles, R.J., and Vassalli, J.D. (1988) Antisense RNA directed against the 3’ noncod- ing region prevents dormant mRNA activation in mouse oocytes. Science, 241t680-684.

Takayama, K.M., and Inouye, M. (1990) Antisense RNA. Crit. Rev. Biochem. Mol. Biol., 25r155-184,

Takenaka, I.M., and Hightower, L.E. (1992) Transforming growth factor-pl rapidly induces hsp7O and hsp90 molecular chaperones in cultured chicken embryo cells. J . Cell. Physiol., 152:56%577.

Takenaka, I.M., and Hightower, L.E. (1993) Regulation of chicken hsp7O and hsp90 family gene expression by transforming growth factor-61. J . Cell. Physiol., 155:5&62.

van der Krol, A.R., Mol, J.N.M., and Stuitje, A.R. (1988) Modulation of eukaryotic gene expression by complimentary RNA or DNA sequences. Biotechniques, 6:958-976.

Wagner, R.W., and Nishikura, K. (1988) Cell cycle expression of RNA duplex unwindase activity in mammalian cells. Mol. Cell. Biol., 8:770-777.

Wagner, R.W., Smith, J.E., Cooperman, B.S., and Nishikura, K. (1989) A double-stranded RNA unwinding activity introduces structural alterations by means of adenosine to inosine conversions in mammalian cells and Xenopus eggs. Proc. Natl. Acad. Sci. USA, ~- 86:2647-2651.

Xing, Y.,Johnson, C.V., Dobner, P.R., andLawrence, J.B. (1993) High level organization of individual gene transcription and RNA splicing. Science, 259:1326-1330. -

effects of dexamethasone, heat shock, and serum responses on the inhibition of hsc70 synthesis by...

Documents