regulation of specific rat liver messenger ribonucleic ... · * a preliminary report of these...

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 260, No. 10, Issue of May 25, pp. 590&5912,1985 Printed in U.S.A.

Regulation of Specific Rat Liver Messenger Ribonucleic Acids by Triiodothyronine*

(Received for publication, April 26, 1984)

Mark A. Magnuson, Beatrice Dozin, and Vera M. Nikodem From the Clinical Endocrinology Branch, National Institute of Arthritis, Diabetes, and Digestiue and Kidney Diseases, National Institutes of Health, Bethesda, Maryland 20205

A plasmid cDNA library was constructed using poly(A*) RNA isolated from the livers of rats treated with 3,5,3‘-triiodothyronine (T3) and fed a high carbohydrate diet. This library was screened by differential colony hybridization with [32P]cDNA probes made from hypothyroid and hyperthyroid rat liver poly(A+) RNA to obtain clones representing T3-inducible mRNAs. Using plasmid cDNAs to 4 different T3-inducible mRNAs, we have studied by hybridization assay the responses of these mRNAs to different thyroidal steady states and to a high carbohydrate diet. The fold of induction (hypothyroid to hyperthyroid) varied from about 4.0 (mRNA 5-8D) to 13.2 (mRNA 4-12B). The linearity of response with regard to nuclear receptor occupancy was estimated by assessing the relative mRNA levels in a euthyroid state. Three of the mRNAs demonstrated nonlinear responses with the largest portion of the induction occurring in the euthyroid to hyperthyroid transition. An induction by the high carbohydrate diet was clearly seen for only one mRNA (5-8D) suggesting that these two pathways of induction are independent. In a study of the response kinetics of each mRNA to a nuclear receptor saturating dose of T3 in hypothyroid animals, an increase was seen within 4 h (the earliest time point examined) for one of the mRNAs. The other 3 mRNAs did not increase significantly until 8 h after the Ts dose. Northern analysis showed a single mRNA corresponding to each of these 4 clones with sizes ranging from about 1375 to 7600 bases. Two mRNAs (5-9E and 4-12B) were shown by hybrid-selected translation to code for proteins of molecular mass of about 27 and 46 kDa, respectively. The availability of several different cDNA probes to T3 responsive liver mRNAs should facilitate future studies on the mechanism of action of this hormone.

The thyroid hormones can act at a pretranslational level selectively to affect gene expression (1). This action is thought to be mediated by the binding of T31 to a chromatin-associated receptor found in responsive tissue (2-4). The nuclear receptor-T3 complex presumably causes specific alterations in the synthetic rates of nuclear RNA. Subsequent to this, there are ~~

* A preliminary report of these studies was presented at the Meet- ing of the 7th International Congress of Endocrinology, Quebec City, Quebec, Canada, July, 1984 (Abstr. 1538). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The abbreviations used are: TI, 3,5,3’-triiodothyronine; SSC, 0.15 M sodium chloride, 0.015 M sodium citrate; SDS, sodium dodecyl sulfate.

changes in specific cytoplasmic mRNA levels, specific protein synthesis, and alterations in cell function. The thyroid hormones increase nuclear RNA polymerase activities (5, 6), which may lead to small increases in the cellular concentra- tions of total and poly(A+) RNA (7, 8). While the overall change in mRNA content is small, the relative levels of specific mRNAs can be dramatically altered.

Studies by two-dimensional gel electrophoresis of mRNA- encoded translational products demonstrated a broad spec- trum of response to T3 in liver, kidney, and heart (9-11). T3 is known to interact with other factors affecting gene expression. In liver, the mRNA levels for several lipogenic enzymes including malic enzyme (12), 6-phosphogluconate dehydrogenase (13), ATP citrate lyase (14), and fatty acid synthetase (15) are regulated by dietary carbohydrate content. Addition- ally, other hormones can also interact with T3 to modulate gene expression. For example, the maximal expression of aZu- globulin mRNA in liver depends upon androgens, glucocorti- coids, insulin, and growth hormone in addition to T3 (16-19). How and where these factors interact on a molecular level to effect gene expression is not known.

We have reported the molecular cloning of a rat malic enzyme cDNA and demonstrated the pretranslational regulation of malic enzyme mRNA by T3 in a hybridization assay (20). The availability of cDNA probes for other T3 regulated mRNAs might help to identify common characteristics in their response to T3 and perhaps help identify common fea- tures in the interaction of Ts with other regulatory factors. Furthermore, although the lipogenic enzymes serve as a good model system for the regulation of specific gene expression by T3, other mRNAs not in this class may reveal different patterns of regulation. Recently the molecular cloning and study of an mRNA from rat liver, termed “Spot 14,” has been reported (21). Although the function of this protein has not been determined, its mRNA appears to be rapidly regulated by changes in the animal’s thyroidal state. Here we report the isolation of 4 differentially hybridizing cDNA clones from a rat liver cDNA library. These cDNAs are used as probes to identify the size of their corresponding cytoplasmic mRNA, to study the regulation of their expression in different thyroidal and dietary steady states, and to determine the kinetics of their response to T3 nuclear receptor saturation in hypothyroid animals. Finally, two of these cDNAs were used to hybrid select mRNAs which, in an in vitro translation system, generated protein products allowing determination of their molecular mass and PI.

EXPERIMENTAL PROCEDURES

Materiak-Avian myeloblastosis virus reverse transcriptase was obtained from J. W. Beard, Life Sciences, Inc. Oligo(dT)-cellulose, type 2, and 0ligo(dT)~2-~~ were from Collaborative Research. The rabbit reticulocyte translation kit, [35S]methionine, [32P]deoxycyti-

5906

mRNA Regulation by Triiodothyronine 5907

dine triphosphate, and ["Clmethyl protein molecular weight standards were from New England Nuclear. All restriction enzymes and DNA molecular weight markers were from Bethesda Research Lab- oratories or New England Biolabs. The nick translation kit was from Bethesda Research Laboratories. BA-85 nitrocellulose filters were from Schleicher & Schuell. All other reagents were of analytical grade.

Treatment of Animals-Female Sprague-Dawley rats (Taconic Farms) weighing 120-150 g were surgically thyroidectomized by the supplier and placed on a low iodine diet (Teklad Diets). Hypothyroid- ism was confirmed after 4-6 weeks by the failure to gain weight and by elevated thyroid-stimulating hormone levels in serum obtained at killing. The animals were given doses of T3 by daily intraperitoneal injection as described for each experiment. Sham injected animals received only the alkaline saline vehicle. Sacrifice was between 800 A.M. and 12:OO P.M. In some cases the hypothyroid animals were fed a high carbohydrate test diet (ICN Biochemical Diets, Inc.).

Isolation of RNA-Rat liver RNA for use in the dot blot hybridization assays was prepared using urea/lithium (22) followed by the treatment of the RNA pellet with 0.1 mg/ml of proteinase K in 10 mM Tris, pH 7.6,5 mM EDTA, and 1% sodium dodecyl sulfate for 30 min at 42 "C. The solution was diluted 5-fold with cold HzO, adjusted to 0.3 M sodium acetate, pH 4.5, and ethanol precipitated to recover the RNA. Rat liver RNA used in the preparation of the cDNA library was prepared using guanidinium thiocyanate (23). Poly(A+) RNA was selected by binding to oligo(dT)-cellulose, type 2, as previously described (24). RNA was stored in sterile water a t -20 "C. For all RNA preparations, glassware was baked at 200 "C and solutions were autoclaved or filter sterilized.

Preparation of the cDNA Library-Poly(A+) RNA was isolated from the livers of normal rats treated for 7 days with T3 (15 pg/100 g, body weight) and fed a high carbohydrate diet. This RNA was used as the template to make cDNA with reverse transcriptase on a plasmid primer by the method of Okayama and Berg (25). Cyclized plasmids were used to transform Escherichia coli (HB101) as described by Hanahan (26). The transformed bacteria were allowed to grow in LB broth with ampicillin (50 pg/ml) at 37 "C for 5 h. The bacteria were concentrated by centrifugation then resuspended in fresh media containing 20% (v/v) glycerol. The liquid suspension of transformed bacteria was quick-frozen in a dry ice/ethanol bath and stored at -70 "C until used. A library of about 35,000 transformants was obtained. Judging from the number of colonies giving positive hybridization signals during the screening experiments, approximately 50% of the plasmids contained cDNA inserts. All work was done in compliance with the National Institutes of Health guidelines for work with recombinant DNA.

Screening by Differential Colony Hybridization-Aliquots of the cDNA library were grown on nitrocellulose filters, replicated twice, and grown until the colonies were visible again. The plasmid DNAs were amplified by overnight incubation of the filters on plates with chloramphenicol (200 pglml). The colonies were lysed (27), baked 2 h a t 80 "C in uacuo and prehybridized at least 4 h in a solution containing 50% formamide, 0.1% each of bovine serum albumin, Ficoll, polyvinylpyrrolidone, and sodium dodecyl sulfate, 0.9 M sodium chloride; 50 mM sodium phosphate, pH 7.4; 5 mM EDTA; 100 pg/ml of denatured salmon testes DNA; and 50 pg/ml of polyadenosine. [32P]cDNA probes were made from poly(A+) RNA (28) obtained from the livers of hypothyroid rats and normal rats treated for 7 days with T3 (15 pg/100 g, body weight). Specific activities of 2-8 x IO7 cpm/ pg of poly(A+) RNA were obtained. Hybridization of the duplicate filters was done in the prehybridization solution with the addition of 2.5-5 X 10' cpm/ml of the [32P]cDNA probes for 2-3 days at 42 "C. The filters were washed 3 times in 2 X SSC and 0.1% SDS at room temperature, twice in 1 X SSC and 0.1% SDS at 68 "C for 1 h each time, and finally in 0.2 X SSC and 0.1% SDS at 68 "C for 1 h. Images were obtained by autoradiography with an intensifying screen at -70 "C. Colonies which appeared to be differentially hybridizing were picked from the master plate and placed in microfilter wells containing LB broth, 15% glycerol, and 50 pg/ml of ampicillin. About 250 colonies were picked during this initial screening. These colonies were inoculated onto a new master filter, grown, replicated, amplified, and hybridized as above. Only those colonies which were differentially hybridizing in this second in situ hybridization were rescreened by differential dot hybridization.

Rescreening by Differential Dot Hybridization-One hundred ap- parently differentially hybridizing colonies were grown in 25-ml minipreparations and their plasmid DNA was isolated (29). This

DNA was linearized with EcoRI endonuclease, denatured, and applied onto nitrocellulose (30) with the aid of a filtration manifold (Schleicher & Schuell). The dots were differentially hybridized with [32P]cDNA probes made from either hypothyroid or T3-induced poly(A+) RNA as above except that 1 X 10' cpm of [32P]cDNA was added per ml of hybridization solution. The filters were washed as described above and individual dots were counted by @-scintillation using a toluene-based mixture. The background hybridization to a dot with pBR322 plasmid DNA was subtracted and a ratio of induction was calculated. Twenty-five colonies which gave the highest ratio of induction were selected for further study. These colonies were grown in large preparations, amplified with chloramphenicol (200 pg/ ml), and lysed by the alkaline-SDS method (291, and their plasmid DNA was purified by banding in a CsCl density gradient (30). The plasmid DNA obtained was used in all further experiments. Two hundred fifty ng of linearized, denatured plasmid DNA was bound to nitrocellulose filter dots with the filtration manifold. Duplicate dots were hybridized with [32P]cDNA probes made from total RNA ex- tracted from hypothyroid rat livers at 0, 4, and 24 h after receiving T3 (200 pg/100 g, body weight). The dots were counted by p-scintillation, and background hybridization to pBR322 plasmid DNA was subtracted.

Analysis of Plusmid Insert Size and Cross-hybridization-Plasmid cDNA insert sizes were determined by digestion with Psti and PuuII under conditions described (30). The resulting fragments were run on 1-1.5% agarose gels, stained with 0.5 pg/ml of ethidium bromide and photographed under UV light. The size of the digestion products was determined by comparison with the migration of the fragments from a HaeIII digest of @X174 DNA. Cross-hybridization of cDNA clones was established by excising the cDNA inserts, fractionation of the products on 1.0% agarose gels, and blot transfer by the method of Southern (31). These filters were individually hybridized with the different nick translated recombinant plasmid [3zP]DNAs (32). After washing and autoradiography, cross-hybridization of the cDNA insert to the probe was assessed.

RNA Blot Analysis-For Northern blot analysis, 2-5 pg of poly(A+) RNA from T3-carbohydrate-induced rat livers was denatured with formaldehyde, separated on 1.2% agarose, 2.2 M formaldehyde gels (33), blot transferred to a nitrocellulose filter (34), and treated as described for the colony hybridization. For RNA dot blot hybridizations, 1-4 pg of poly(A+) RNA was denatured with formaldehyde and applied to nitrocellulose filters with the aid of the filtration manifold. Both the Northern and dot blots were prehybridized as above for at least 6 h and then hybridized with plasmids containing cDNA that were nick translated (32) to specific activities of 2-4 X 10' cpm/pg. Approximately 10' cpm/ml of plasmid [32P]DNA was added to each hybridization. The filters were allowed to hybridize at least 2 days and then washed as above. The filters were exposed to film for sufficient periods of time to obtain an image. The dots were cut from the filter and the bound radioactivity was quantitated by p-scintillation. Background from an equally sized piece of filter with no bound RNA was subtracted.

Identification of Protein Products by Hybrid-selected Translation- Recombinant plasmid DNA (10 pg) was linearized with EcoRI, denatured, and applied to 5-mm circles of nitrocellulose as described (35). The filters were baked 2 h at 80 "C in uacuo and washed with water at 100 "C for 1 min to remove unbound DNA. The filters were rinsed in a solution containing 50% formamide, 20 mM 1,4-pipera- zinediethanesulfonic acid, pH 6.4,0.4 M sodium chloride, 0.2% sodium dodecyl sulfate, 100 pg/ml of calf thymus tRNA, and 100 pg/ml of polyadenosine. Hybridization was in the same solution with 200 pg of T3-carbohydrate-stimulated poly(A+) RNA added. This solution was preheated to 70 "C for 5 min before adding the filters. Hybridi- zation was carried out at 45 "C for 16 h. The filters were rinsed at 60 "C 10 times with 0.5 X SSC and 0.5% sodium dodecyl sulfate and then 3 times in 2 mM EDTA. The RNA was eluted and translated as previously described (36). The molecular weights of the 35S-protein products were determined on 12% SDS-polyacrylamide gels by electrophoresis. The gels were fixed in 10% trichloroacetic acid then impregnated with dimethyl sulfoxide/2,5-diphenyloxazole (37), dried, and exposed at -70 "C. The isoelectric points of the hybrid-selected translation products were determined by the method of O'Farrell(38), except that the following percentage of ampholines (BioLyte, Biorad Co.) were used pH range 4-6, 0.8%; pH 6-8, 0.8%; and pH 3-10, 0.4%.

5908 mRNA Regulation by Triiodothyronine

RESULTS

Selection of Differentially Hybridizing Clones and Charac- terization of Their Plasmids-Differential colony hybridization (two cylces) and dot blot hybridization was used to select clones for study which appeared induced after T3 treatment. The dot blot hybridization was necessary to help eliminate falsely differentially hybridizing colonies obtained from the initial screening by colony hybridization. The differential hybridizations were done with single stranded [32P]cDNA probes made from hypothyroid and hyperthyroid rat liver poly(A+) RNA. Autoradiograms of two replica filters which were differentially hybridized are seen in Fig. lA. Arrows point to several colonies representative of those selected by this procedure. Autoradiograms from the differential dot blot hybridization are seen in Fig. 1B. Arrows point to several of the most strongly differentially hybridizing clones chosen for further study.

Sizing of the excised plasmid cDNA inserts was accomplished by fractionation on agarose gels and comparison with DNA standards (Table I). The sizes of the corresponding mRNA for the cDNA clones were determined by Northern blot analysis of poly(A+) RNA from T3-treated rats. Clones which hybridized to the same size mRNA and which showed similar response patterns after T3 treatment of hypothyroid rats were cross-hybridized to establish their common identity. This was accomplished by Southern blot analysis of the excised plasmid inserts (data not shown). The multiple isolation of cross-hybridizing clones occurred several times. Clones corresponding to 5-9E and 4-12B were each obtained three times. This is probably related both to the relatively high abundance of the parent mRNA in the cDNA library

Hypothyroid =P-cDNA Hyperthyroid SpaDNA

FIG. 1. Differential colony and dot hybridization of cDNA clones. A, differential colony hybridization. Transformed colonies, obtained from the cloning of cDNA synthesized from liver poly(A+) RNA from T3-treated and carbohydrate-fed rats were grown on master nitrocellulose filters and replicated twice. The filters were probed with [32P]cDNA made from liver poly(A+) RNA of hypothy: roid (left) and TI-treated, hyperthyroid rats (right). Autoradiography of the filters revealed colonies which were differentially hybridizing. Arrows point to several colonies representative of those chosen. B, differential dot hybridization. Plasmid DNA selected by differential colony hybridization was linearized, denatured, and bound to nitrocellulose filters. The filters were probed with [32P]cDNA made from liver poly(A') RNA of hypothyroid (left) and T3-treated hyperthyroid rats (right). Autoradiography shows differentially hybridizing colonies. Arrows point to some of the plasmids selected for further study.

TABLE I Summary o j p h m i d s

Plasmid Insert size mRNA size Estimated No. of times

isolated

bp" bases 5-9E 780 1375 3 4-12B 1320 2250 3 4-12D 870 4500 1 5-8D 1190 7600 1

bp, base pairs. Size total of fragments obtained when plasmid was digested with PuuII ad PstI endonucleases. The actual cDNA size is about 95-bp less. The difference represents plasmid-derived sequences, poly(A/T) and poly(C/G) sequences, contained between these two restriction sites.

28S- 23s -

18S- 16S-

1 2 3 4 FIG. 2. Northern blot analysis of mRNA size. Poly(A+) RNA

from T3-carbohydrate-induced rat liver was resolved on 1.2% agarose, 2.2 M formaldehyde gels, transferred to nitrocellulose, and hybridized with 3ZP-labeled plasmid DNAs. After washing the filters, images were obtained by autoradiography. Lane 1, 5-9E; lane 2, 4-12B; lane 3, 4-12D; and lane 4, 5-8D. Eukaryotic and prokaryotic ribosomal RNA size standards are 1.7 (16 S ) , 2.0 (18 S ) , 3.1 (23 S) , and 5.1 (28 S ) kilobases (45).

screened and to a strong differential hybridization signal for these clones.

A composite of the Northern blots for each different clone is shown in Fig. 2. All clones hybridized to a single sized mRNA. The sizes were estimated by comparison with ribosomal RNA standards. As can be seen in Table I, the estimated sizes ranged from about 1375 nucleotides for the mRNA of clone 5-9E to about 7600 nucleotides for the mRNA of clone 5-8D. The latter size is based upon an extrapolation past the 28 S ribosomal RNA standard. The clone 5-8D was originally isolated from a library made with poly(A+) RNA enriched for rat malic enzyme mRNA by immunopurification (20). The clone was subsequently shown, however, not to be coding for malic enzyme mRNA but corresponded to an mRNA which was responsive to T3 and a high carbohydrate diet. The cDNA insert sizes ranged from 780 to 1320 base pairs (Table I). The excised inserts represent cDNA except for 95 base pairs due to the cloning procedure and vector.

Relative mRNA Levels in Different Thyroid States and in Response to a High Carbohydrate Diet-Dot hybridization assay under conditions of cDNA probe excess was used to determine the relative levels of each mRNA under the different treatments. Under these conditions the signal is propor- tional to the quantity of hybridizable RNA. A euthyroid T3 replacement dose of 300 ng/100 g, body weight, daily was used in these experiments (9,lO). In our experiments this dose did not normalize the thyroid-stimulating hormone levels obtained at killing (data not shown). This is consistent with the


7 14

TABLE I1 Response of selected mRNA to changes in thyroid state and diet

Hypothyroid

Euthyroid

Hyperthyroid - 0 High Carbohydrate Diet

FIG. 3. Effect of thyroid state and carbohydrate feeding on hybridizable mRNA levels. Hypothyroid rats were given different treatments for 10 days and killed, and their livers were processed to obtain poly(A+) RNA. 1-2 pg of this poly(A+) RNA was bound to nitrocellulose filters, hybridized with each different nick translated 32P-plasmid DNA and washed, and the bound [32P]cDNA was determined by @-scintillation counting. The results are expressed as cpm/ pg of poly(A+) RNA applied (less background activity of an equally sized filter with no bound RNA). The hypothyroid rats (hatched bars) were untreated. A euthyroid state (solid bars) was obtained by treating hypothyroid rats with 300 ng of T3/100 g, body weight, for 10 days. A hyperthyroid state (shaded bars) was obtained by treating hypothyroid rats with 15 pg of T3/100 g, body weight, per day for 10 days. The carbohydrate-fed rats (open bars) received a high carbohydrate test diet. The results are expressed as the mean f S.D. of a t least 3 RNA preparations representing at least 6 individual animals.

observation by Seelig et al. (9) that the activities of thyroid hormone-stimulated hepatic enzymes were less than normal in rats receiving this dose. Nevertheless, this dose achieves a thyroidal state between hypothyroid and hyperthyroid, which is nearly euthyroid. To induce a hyperthyroid state, the hypothyroid animals were treated for 10 days with 15 pg of T3/ 100 g, body weight. This dose has been sufficient to induce malic enzyme mRNA approximately 10-fold as assessed by hybridization assay.’

The relative levels of each hybridizable mRNA in hypothyroid, euthyroid, and hyperthyroid states are shown in Fig. 3. The calculated ratios between euthyroid and hypothyroid, hyperthyroid and euthyroid, and hyperthyroid and hypothyroid are shown in Table 11. Over the full transition range (hypothyroid to hyperthyroid) the ratios of induction varied from about 4.0 for mRNA 5-8D to about 13.2 for mRNA 4- 12B. The ratios of induction over half the transition range, both hypothyroid to euthyroid and euthyroid to hyperthyroid, are also indicated in Fig. 2. Except for mRNA 4-12B, a higher ratio of induction is seen in the euthyroid to hyperthyroid transition than for the hypothyroid to euthyroid transition.

B. Dozin, M. A. Magnuson, and V. M. Nikodem, unpublished observations.

5-9E 11.3 2.5 4.5 0.15 2.4 4-12B 13.2 4.7 2.8 0.30 1.0 4-12D 5.2 1.2 4.2 0.06 0.4 5-8D 4.0 1.4 2.8 0.15 4.1

a Ratio of hybridizable mRNA level in hyperthyroid state (15 pg/ 100 g, body weight, for 10 days) and hypothyroid state.

* Ratio of hybridizable mRNA level in euthyroid state (300 ng/100 g, body weight, for 10 days) and hypothyroid state.

e Ratio of hybridizable mRNA level in hyperthyroid state (15 pg/ 100 g, body weight, for 10 days) and euthyroid state (300 ng/100 g, body weight, for 10 days).

Ratio of hybridizable mRNA levels for euthyroid state - hypothyroid state and hyperthyroid state - hypothyroid state.

‘Ratio of hybridizable mRNA level from high carbohydrate diet (HCD)-fed hypothyroid rats (10 days) and regular diet-fed hypothyroid rats.

The euthyroid hybridizable mRNA level expressed as a ratio of the total range of induction is calculated and expressed, euthyroid - hypothyroid + hyperthyroid - hypothyroid. These values range from a low of 0.06 for mRNA 4-12D to a high of 0.30 for mRNA 4-12B.

The response of each mRNA to the feeding of a high carbohydrate diet for 10 days was also examined and is also illustrated in Fig. 3. mRNA 5-8D was the most responsive to this treatment and was induced about 4.1-fold over hypothyroid rats fed a regular diet. The high carbohydrate diet- induced mRNA level observed was equal to that obtained by treatment with T3 (15 pg/100 g, body weight, for 10 days). The other mRNAs were generally less responsive to a high carbohydrate diet, although a 2.4-fold increase is seen for mRNA 5-9E. Little or no change is seen for mRNAs 4-12B and 4-12D.

Kinetic Response to Receptor-saturating Doses of T3 in Hy- pothyroid Animals-The time course of response for each mRNA to a replacement dose of T3 is shown in Fig. 4. Hypothyroid animals were given either T3 (200 pg/lOO g, body weight, every 24 h) or alkaline saline vehicle by intraperitoneal injection. This T3 dose has been shown to achieve >95% saturation of the T3 nuclear receptors for a period greater than 24 h (39). By using this dose of T3 one avoids the interpretational problems of dosing at more frequent intervals to maintain nuclear receptor saturation. All mRNAs selected showed dramatic changes by 24 h after the T, dose. This is entirely consistent with the increase in the levels of these mRNAs seen in the hypothyroid to hyperthyroid transition of Fig. 3. Variations in the pattern of induction are seen for the individual mRNAs. mRNA 4-12D appears to respond at 4 h in the T3-treated rats over the levels seen for the vehicle- injected rats, whereas for the others there are no clear changes until 8 h after the T3 dose. The significance of the increase at 4 h and decrease at 8 h for mRNAs 4-12B, 4-12D, and 5-8D is not established. Additional data points would be necessary to confirm the small changes seen. The relative rates of increase in hybridizable mRNA thereafter varies for each mRNA. mRNA 5-8D rises rapidly after 8 h to a peak at 24 h and then declines sharply. mRNA 5-9E rises sharply between 12 and 24 h then slows as it nears its plateau levels. mRNA 4-12B appears to steadily increase up to 48 h and probably beyond since the level at that time is only about half that seen in the 10-day T3-treated rats (data not shown).

Identification of mRNA Protein Products by Hybrid-selected Translation-Attempts to identify the protein products for

5910 mRNA Regulation by Triiodothyronine I 5-9E

I 5-8D

l t b 40 20 0 4 812 24 48

I 4-12B

0 4 8 12 24 48

I 4-12D

601 /

I I 0 4 8 12 24 48

~~

TIME (HOURS)

5-8D

* e*..* 0 * .a.

0 4V 4 8 12 24 48 TIME AFTER T3 DOSE

(HOURS)

FIG. 4. Kinetic response of hybridizable mRNA levels to a receptor-saturating dose of Ts. Hypothyroid rats, three/group except 24 and 48 h which contained two, were given either T3 (200 pg/100 g, body weight) or alkaline saline at time 0. At the times indicated, the rats were killed, and equal portions of their livers were combined and then processed together to obtain poly(A+) RNA. 1-4 pg of this poly(A+) RNA was bound to duplicate dots on nitrocellulose filters, hybridized with each different nick translated [32P]cDNA, and washed, and the bound [32P]cDNA was determined by @-scintillation counting. The results are expressed as cpm/pg of poly(A+) RNA applied (less background activity of an equally sized filter with no bound RNA). The solid line is the response to the Ts dose. The dashed line is the response to the vehicle a t 4 h only. A second dose of T3 was given at 24 h to those rats killed at 48 h. The inset (bottom) shows an autoradiogram of the duplicate dots for clone 5-8D after hybridization and washing. 4V designates the poly(A+) RNA from the rats killed 4 h after the vehicle injection. Each dot represents the signal obtained from 1 pg of poly(A+) RNA.

the mRNAs corresponding to several of the cDNA clones were made. In two cases recombinant plasmid DNA, bound to filters, selected an mRNA which directed the synthesis of 35S- labeled proteins in an in vitro system, as can be seen in Fig. 5. A filter with pBR322 DNA (lane 2) yielded only a few products of relatively low intensity. The filter with clone 4- 12B plasmid DNA (lane 3) selected an mRNA which directed the synthesis of a protein corresponding to molecular mass of about 46 kDa. The filter with clone 5-9E plasmid DNA (lane 4) similarly yielded a protein of about 27 kDa. As can be seen, the bands generated by the selected mRNA were of much higher intensity than bands thought to be derived from non- specific filter binding of mRNA. The isoelectric points of these products were separately determined by isoelectric fo- cusing to be about 4.4 and 4.5 for clone 4-12B and 5-9E, respectively (data not shown). Judging from the mass of the translated proteins (27 and 46 kDa), the coding region for mRNAs 5-9E and 4-12B was estimated to be 53 and 56%, respectively. Thus, these mRNAs may have substantial 3’ and/or 5’ untranslated regions.

5 30+

1 2 3 4 FIG. 5. Hybrid-selected translation of clones 5-9E and 4-

12B. Recombinant plasmid DNAs were separately bound to nitrocellulose filters and hybridized to 200 pg of poly(A+) RNA from T3- carbohydrate-treated rats. The filters were washed and the bound mRNA was eluted as described under “Experimental Procedures.” The %-labeled translation products were analyzed by electrophoresis on 12% SDS-polyacrylamide gels. Total translation products visual- ized by fluorography are shown. Lane 1, no added mRNA; lane 2, translation products from mRNA eluted from filter containing pBR322 DNA; lane 3, translation products from mRNA eluted from the filter containing 4-12B plasmid DNA; and lane 4 , translation products from mRNA eluted from the filter containing 5-9E plasmid DNA. “C-methylated molecular mass marker proteins (in kDa) were bovine serum albumin (69), ovalbumin (46), carbonic anhydrase (30), and @-lactoglobulin (18).

DISCUSSION

The mechanism by which T3 can act to regulate the expression of specific genes is unknown. It is possible that each of these genes possesses a similar DNA sequence in the 5‘ regulatory region which is recognized by the thyroid hormone receptor complex and which regulates the transcriptional activity of the gene. A number of different T3 responsive genomic clones would be useful in investigating the gene regions which possibly interact with the Tg-receptor complex. Although genomic clones and sequence data are not yet avail- able, studies utilizing different cDNAs as probes to study changes in mRNA levels can give some preliminary insights into T3 action. To begin to accomplish these goals, cDNA clones to several Tg-regulated genes from rat liver have been obtained by differential hybridization of a cDNA plasmid library. These have been used to study the response of their corresponding mRNAs to various thyroidal states, the kinetic response to an acute dose of T3 in hypothyroid animals, and the effect of changes in dietary carbohydrate.

The response patterns of Tg-regulated proteins as a function of nuclear receptor occupancy are variable. The simplest relation between the concentration of the Tg-receptor complex in the nucleus (so-called receptor occupancy) and the gene product level would be linear. This appears to be the case for the relationship of growth hormone content of pituitary cells in culture and to receptor occupancy of T3 (40). Hepatic malic enzyme and a-glycophosphate dehydrogenase show nonlinear responses defined as showing a greater increase in the 50% to >95% receptor occupancy transition (euthyroid to hyperthyroid) than in the 5-15% to 50% receptor occupancy transition


(hypothyroid to euthyroid) (41). This nonlinear response has been termed amplification. Recently the response of Spot 14 mRNA (21) and malic enzyme mRNA2 have been shown to reflect a nonlinear response. The response of 3 of the T3 responsive mRNAs reported here are similar to those seen for malic enzyme mRNA and Spot 14 mRNA. mRNAs 5-9E, 4- 12D, and 5-8D respond most in the euthyroid to hyperthyroid transition with less than 15% of their induction occurring in the hypothyroid to euthyroid transition. mRNA 4-12B has a more nearly linear response with respect to predicted nuclear receptor occupancy since about 30% of its induction occurs in the hypothyroid to euthyroid transition. A strictly linear response would show about 50% of its induction in both the hypothyroid to euthyroid and euthyroid to hyperthyroid transition. Since the euthyroid state in these animals is actually slightly hypothyroid, the true euthyroid level for this mRNA may be higher, making the responses even more linear. A model of T3-mediated regulation of gene expression needs to deal with the response patterns seen, both linear and nonlinear. How different mRNAs can respond with different patterns to the same level of T3 nuclear occupancy is not under- stood. Furthermore, although we deal only with mRNAs which are induced by T3, one must also account for attenua- tion in the levels of specific mRNAs which has also been observed for T3 (9-11). Additional studies are necessary to address these questions central to the action of T3.

All of these mRNAs show substantial inductions within 24 h after an acute dose of TI. The earliest onset of induction is seen for mRNA 4-12D (clearly induced a t 4 h) with the other 3 mRNAs showing no significant increases until 8 h after the T3 dose. These responses are generally in the same time range seen for malic enzyme mRNA (42) but distinctly longer than the response seen for Spot 14 mRNA (21) which is induced significantly at 30 min under similar conditions. It is unknown if this means that other factors are responsible for the induction of these mRNAs or if these responses are simply a function of transcription rate and/or mRNA half-life. All of these kinetic responses to T3 must be interpreted with appre- ciation that the observed rapidity of increase in mRNA levels is due not only to a change in the rate of synthesis but also to the half-life of the mRNA. If these mRNA had a long half- life, even an immediate major change in synthetic rate would appear as a gradual increase in mRNA level.

In liver those mRNAs regulated by T3 are often also regulated by diet (10). In this study we examined the response of these mRNAs to feeding a high carbohydrate diet. The dietary response seen in hypothyroid animals is often attenuated compared to that of normal animals (10). Despite this, we found one of the mRNAs, 5-8D, to be highly responsive to dietary factors. The other mRNAs which were regulated by T3 appeared minimally or unaffected by dietary changes. Thus, it appears for these mRNAs that T3 and diet do not necessarily act by a common pathway to regulate the cytoplasmic mRNA level.

When an mRNA level is altered in response to treatment with T3, there are several possible mechanisms by which this could occur. If the change is due to an altered rate of transcription, this could be governed directly by the level of the T3-nuclear receptor complex. Alternately, the T3-receptor complex may act by regulating the level of another mRNA coding for a protein which regulates the transcriptional activity of several genes or which affects the stability of certain mRNAs. These possibilities are speculatory since there are many sites of proven mRNA regulation (43), and where T3 acts in this series is unknown. Spot 14 mRNA, which responds the most rapidly of any T3-responsive mRNA thus far studied,

is not induced in the presence of cyclohexamide (44). This finding suggests the possibility of a short-lived protein inter- mediate(s) as yet unidentified, which may be important in the regulation of mRNA levels by T3. Similar factors may be important for the induction of the mRNAs in this report.

Several of the clones we have identified may be useful for future studies of the mechanism of action of thyroid hormones and the effects of diet on mRNA regulation. mRNAs 5-9E and 4-12B are the best candidates for the study of T3 action since their protein product size has been determined and they are both induced to moderate levels with T3. It will be important to determine if there is a common regulatory site for these mRNAs, and what the site of this regulation is. Future studies with these probes can focus on some of these problems and allow extension of our investigations to the structural genes encoding them.

Acknowledgments-We would like to thank Joan Todd for her excellent assistance in the preparation of this manuscript and Dr. J. E. Rall for his critical reading of it.

REFERENCES 1. Towle, H. C. (1983) in Molecular Basis of Thyroid Hormone Action

(Oppenheimer J. H., and Samuels, H. H., eds) pp. 179-212, Academic Press, New York

2. Oppenheimer, J. H. (1979) Science (Wash. DE.) 203 , 971-979 3. Samuels, H. H., Perlman, A. J., Raaka, B. M., and Stanley, F.

(1982) Recent Prog. Horm. Res. 38,557-559 4. Baxter, J. D., Eberhardt, N. L., Apriletti, J. W., Johnson, L. K.,

Ivarie, R. D., Schachter, B. S., Morris, J. A., Seeburg, P. H., Goodman, H. M., Latham, K. R., Polansky, J. R., and Martial, J. A. (1979) Recent Prog. Horm. Res. 35,97-153

5. Tata, J. R., and Widnell, C. C. (1966) Biochem. J. 98 , 604-620 6. Widnell, C. C., and Tata, J. R. (1966) Biochem. J. 98 , 621-629 7. Towle, H. C., Dillmann, W. H., and Oppenheimer, J. H. (1979)

J. Biol. Chem. 254,2250-2257 8. Dillmann, W. H., Mendecki, J., Koerner, D., Schwartz, H. L.,

and Oppenheimer, J. H. (1978) Endocrinology 102, 568-575 9. Seelig, S., Liaw, C., Towle, H. C., and Oppenheimer, J. H. (1981)

Proc. Natl. Acad. Sci. U. S. A. 78,4733-4737 10. Liaw, C., Seelig, S., Mariash, C. N., Oppenheimer, J. H., and

Towle, H. C. (1983) Biochemistry 22,213-221 11. Dillmann, W. H., Barrieux, A., Neeley, W. E., and Contreras, P.

(1983) J. Biol. Chem. 258 , 7738-7745 12. Siddiqui, U. A., Goldflam, T., and Goodridge, A. G. (1981) J. Biol.

Chem. 256,4544-4550 13. Miksicek, R. J., and Towle, H. C. (1983) J. Biol. Chem. 258,

9575-9579 14. Sul, H. S., Wise, L. S., Brown, M. L., and Rubin, C. S. (1984) J.

Biol. Chem. 259,1201-1205 15. Morris, S. M., Jr., Nilson, J. H., Jenik, R. A., Winberry, L. K.,

McDevitt, M. A., and Goodridge, A. G. (1982) J. Biol. Chem.

16. Roy, A. K., and Dowbenko, D. J. (1977) Biochemistry 16 , 3918- 257,3225-3229

3922 17. KUkZy D. T., Chan, K.-M., and Feigelson, P. (1978) J. Biol. Chem.

253,7886-7890 18. Kurtz, D. T., Sippel, A. E., and Feigelson, P. (1976) Biochemistry

19. Roy, A. K., Chatterjee, B., Prasad, M. S. K., and Unakar, N. J.

20. Magnuson, M. A., and Nikodem, V. M. (1983) J. Biol. Chem.

21. Jump, D. B., Narayan, P., Towle, H., and Oppenheimer, J. H.

22. Auffray, C., and Rougeon, F. (1980) Eur. J. Biochem. 107, 303-

23. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., and Rutter,

24. Krystosek, A., Cawthon, M. L., and Kabat, D. (1975) J. Biol.

25. Okayama, H., and Berg, P. (1982) Mol. Cell. Biol. 2, 161-170 26. Hanahan, D. (1983) J. Mol. Biol. 166 , 557-580

15,1031-1036

(1980) J. Biol. Chem. 255, 11614-11618

258,12712-12717

(1984) J. Biol. Chem. 259,2789-2797

314

W. J. (1979) Biochemistry 18, 5294-5299

Chem. 250,6077-6084

5912 mRNA Regulation by Triiodothyronine

27. Grunstein, M., and Hogness, D. S. (1975) Proc. Natl. Acad. Sci. M. S., and Goldstein, J. L. (1982) Proc. Natl. Acad. Sci. U. S.

28. St. John, T. P., and Davis, R. W. (1979) Cell 16, 443-452 37. Bonner, W. M., and Laskey, R. A. (1974) Eur. J. Biochem. 46, 29. Birnboim, H. C., and Doly, J. (1979) Nucleic Acids Res. 7, 1513- 83-88

1523 38. O’Farrell, P. H. (1975) J. Biol. Chem. 250,4007-4021 30. Maniatis, T., Fritsch, E. F., Sambrook, J. (eds) (1982) Molecular 39. Oppenheimer, J. H., Silva, E., Schwartz, H. L., and Surkes, M. I.

Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, (1977) J. Clin. Invest. 59, 517-527 Cold Spring Harbor, NY 40. Tsai, J. S., and Samuels, H. H. (1974) Biochem. Biophys. Res.

31. Southern, E. (1975) J. Mol. Biol. 98, 503-517 Commun. 59,420-428 32. Rigby, P. W. J., Dieckman, M., Rhodes, C., and Berg, P. (1977) 41. Coulombe, P., Schwartz, H. L., and Oppenheimer, J. H. (1978) J.

J. Mol. Biol. 113, 237-251 Clin. Invest. 62, 1020-1028 33. Lehrach, H., Diamond, D., Wozney, J. M., and Boedtker, H. 42. Towle, H. C., Mariash, C. N., and Oppenheimer, J. H. (1980)

(1977) Biochemistry 16, 4743-4751 Biochemistry 19,579-585 34. Thomas, P. S. (1980) Proc. Natl. Acad. Sci. U. S. A. 77, 5201- 43. Darnell, J. E. (1982) Nature 297,365-371

5205 44. Seelig, S., Jump, D. B., Towle, H. C., Liaw, C., Mariash, C. N., 35. Ploegh, H. L., Orr, H. T., and Strominger, J. L. (1980) Proc. Natl. Schwartz, H. L., and Oppenheimer, J. H. (1982) Endocrinology

Acad. Sci. U. S. A. 77,6081-6085 110,671-673 36. Chin, D. J., Luskey, K. L., Farst, J. R., MacDonald, R. J., Brown, 45. Loening, U. E. (1968) J. Mol. Biol. 38, 355-365

U. S. A. 72,3961-3965 A. 79,7704-7708

regulation of specific rat liver messenger ribonucleic ... · * a preliminary report of these...

Documents