Role of Gene Overlap in the Regulation of mRNA Translation for Mitochondrial Cytochrome P-450c27/25 in the Rat*

(Received for publication, February 21, 1996, and in revised form, May 17, 1996)

Rass M. Shayiq

From the Drug Liver Unit, Department of Pharmacology, Temple University School of Medicine, Philadelphia, Pennsylvania 19140

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
Acknowledgments
REFERENCES


ABSTRACT

Previously published results have revealed sequence complementarity between the 5'-terminal regions of mRNAs for hepatic mitochondrial cytochrome P-450c27/ 25 (c27/25) and serine protease inhibitors (SPI) and predicted a role for this sequence overlap in both the regulation of c27/25 mRNA transcription and translation. The possibility that c27/25 mRNA forms an RNA duplex with complementary sequences of SPI mRNAs in vivo was demonstrated in the rat liver and COS-1 cells cotransfected with c27/25 and SPI2.1 plasmids. Quantitative evaluation of RNA duplex in COS-1 cells revealed that most of the c27/25 mRNA exists in duplex form when SPI2.1 mRNA was present at 5-10-fold that of c27/25 mRNA, a ratio comparable to that observed between these two RNAs in the liver. In cotransfected COS-1 cells with the same ratio of mRNAs, highly significant inhibition of the c27/25 mRNA translation (66-75%) was observed, while its transcription remained unaffected. The partial inhibition of c27/25 mRNA translation, even when most of it exists in duplex form, suggests that RNA duplex is undergoing some type of cytoplasmic processing to disengage c27/25 mRNA and make it available for translation. These results imply that abundant endogenous SPI RNAs are able to regulate the c27/25 gene expression.


INTRODUCTION

Two physiologically important reactions, 25-hydroxylation of vitamin D3 and 27-hydroxylation of cholesterol are catalyzed by a single rat hepatic mitochondrial enzyme, cytochrome P-450c27/25,1 a product of the CYP27 gene (1-4). The 27-hydroxylation of cholesterol is a rate-limiting step in the catabolic conversion of cholesterol to bile acids and 27-hydroxycholesterol is a regulatory oxysterol which indirectly modulates cholesterol synthesis in the liver and other tissues (5-7). The 25-hydroxylation of vitamin D3 is a first step in the conversion of inactive vitamin D3 to the 1alpha ,25-dihydroxy derivative, which is an active hormonal form (8-10). Vitamin D3 hormone, a known regulator of mineral metabolism and calcium homeostasis has more recently been implicated in a number of other biological activities not yet fully understood but clearly related to immunological responses, cell proliferation and differentiation (9-12).

Previously published results report two distinct mRNAs in both the rat (1, 13) and human (14) that code for the synthesis of a similarly sized 54-kDa c27/25 precursor protein (13). The larger 2.3-kb mRNA is identical to 1.9-kb mRNA except for a 400-nucleotide 5'-untranslated region (5'-UTR) extension. The terminal 291 nucleotides of this extension exhibit 100% complementarity with the 5'-translated region of SPI2.1 mRNA and about 99% complementarity with the 5'-translated regions of SPI2.2 and SPI2.3 mRNAs (13, 15, 16). The two mRNAs are most likely the products of two independent promoter regions as the gene that transcribes the smaller mRNA has been recently characterized (17). The sequence overlap between c27/25 and SPI2.1 mRNAs is unique and probably represents the first observation of 5' end overlap between two functional mRNAs. In all other previously reported cases of gene overlap, the overlapping region is located at or near the 3'-untranslated regions (18-22) or overlapping exists between a functional and a nonfunctional RNA (23-25). Furthermore, the higher abundance of SPI mRNAs in comparison to c27/25 mRNA in the liver suggests a role for this gene overlap in the regulation of CYP27 expression.

Antisense RNA has been shown to function as a regulator of DNA replication and mRNA translation in a number of experimental systems (26). The mechanisms by which this antisense RNA alters expression of the target gene has been variously attributed to effects on transcription, nuclear processing, nuclear transport, translation, and mRNA stability (26, 27). However, very little information exists to suggest a role for naturally occurring antisense RNA in eukaryotic gene regulation (22, 28-30). Formation of RNA duplex between cell RNAs and antisense transcripts or between two naturally occurring complementary mRNAs has been verified in several organisms and mammalian tissue culture cells (29, 31-35). A novel double strand-specific RNA unwindase/modificase activity, which can permanently separate the two complementary RNA strands, has also been detected in several organisms and mammalian tissue culture cells (36-39).

In this report, formation of an RNA duplex between the complementary sequences of c27/25 and SPI mRNAs has been investigated in the liver and COS cells carrying their genes. Quantitative evaluation of RNA duplexes formed in the transfected COS cells that expressed two mRNAs in a ratio comparable to that observed in liver revealed existence of most of the c27/25 mRNA in duplex form. Furthermore, the evidence that duplex formation plays an essential role in the control of c27/25 mRNA translation by blocking its 5' end is presented. The results suggest a regulatory role for the naturally occurring antisense RNA in the eukaryotic gene expression.


EXPERIMENTAL PROCEDURES

Materials

The chemicals and biochemicals used were of highest purity grades purchased from Sigma or Fisher. Radiolabeled materials such as [32P]dCTP, [32P]UTP (6000 Ci/mmol), and [35S]methionine (600 Ci/mmol) were purchased from Amersham Corp. Nytran and nitrocellulose membranes for various blot transfers were obtained from Schleicher & Schuell. Alkaline phosphate-conjugated secondary antibody, avian myeloblastosis reverse transcriptase, ribonuclease inhibitor (RNasin), rabbit reticulocyte lysate, and some restriction enzymes were purchased from Promega Biotech Corp. Amplitaq PCR fragment sequencing kit and PCR kits were supplied by Cetus Corp. In vitro transcription kit and other restriction enzymes were purchased from Stretagene. Reagents for polyacrylamide gel electrophoresis were purchased from Bio-Rad. The sources for pCMV4 vector and SPI2.1 have been described before (13).

Plasmid Construction

The 2.3-kb c27/25 cDNA was cloned into HindIII and XbaI sites of mammalian expression vector pCMV4 in the proper orientation (pCMVc27/25) as described before (13). The 1.9-kb SPI2.1 cDNA was excised from pBR322 by using PstI restriction enzyme, blunt ended using Klenow fragment, and cloned into SmaI site of pCMV4 in the proper orientation (pCMVSPI) and reverse orientation (pCMVSPIrev). CAT cDNA was cloned in front of the overlap region (OLR) in pCMV4 (OLRCAT), to replace the c27/25 translated region as a reporter gene. The OLRCAT in pCMV4 was prepared by isolating the 1.5-kb CAT fragment from pCAT basic vector (Promega biotech) by cutting with HincII and then ligating it in front of the OLR in pCMV4. pCMV4 vector with the OLR was prepared by excising the SmaI-SmaI 2.0-kb fragment from pCMVc27/25.

The overlap region of the c27/25 cDNA was amplified by using a sense primer PS-1 (5-ACAGTATCGGCTGCTCTTTG-3') and an antisense primer PAS-1 (5'-ATTGAGCCTCTACAAGAAGC-3') and the 280-bp amplified product was cloned into pCR3-Uni TA cloning vector (280pCR3) and used for the preparation of sense and antisense OLR cRNA probes as shown in Fig. 3A. Similarly, a 342-bp HindIII and SmaI fragment from the 5' end of the 2.3-kb c27/25 cDNA in Bluescript (330 bp from 5'-UTR and 12 bp from Bluescript vector) was cloned into pGEM-3Z (342pGEM) and used for the preparation of cRNA probe as shown in Fig. 2A.


Fig. 3. Quantitative analysis of RNA duplexes by double RNase A protection assay. A, the scheme of sense and antisense cRNA probe preparation for second RNase A protection assay. An open box represents the 280 bp of the OLR amplified with PAS-1 and PS-1 primers and cloned in pCR3-Uni vector. Sense and antisense cRNA probes were prepared from plasmid as described under "Experimental Procedures." B, the final pellets of C-RNAs and R-RNAs isolated from equal number of 1:4 COS cells were suspended in 25 µl of water and 5 µl from each sample were hybridized to 1 × 106 cpm of either sense or antisense probes for 16 h at 55 °C as described before (13). After a second RNase A digestion, protected species were melted and fractionated through a 5% acrylamide, M urea sequencing gel. Protected fragments from C-RNA (lane 1), tRNA (lane 2), R-RNA (lane 5), and heat denatured R-RNA (lane 6) with sense cRNA probe. Similarly, protected fragments from C-RNA (lane 3), tRNA (lane 4), R-RNA (lane 7), and heat denatured R-RNA (lane 8) with antisense cRNA probe. C, COS cells (1 × 107) cotransfected with c27/25:SPI2.1 plasmids in a 1:4 ratio were exposed to an access (5 × 106 cpm) of sense or antisense cRNA probes at the time of cell lysis and RNase A-resistant RNAs were isolated as described under "Experimental Procedures." 10 µg of resistant material were directly fractionated through a 5% acrylamide, 8 M urea sequencing gel. Lane 1, 1:4 COS cells exposed to antisense probe. Lane 2, 1:4 COS cells exposed to sense probe. Lane 3, normal COS cells exposed to sense probe. The expected position for a 280-bp protected fragment is indicated.
[View Larger Version of this Image (44K GIF file)]



Fig. 2. Detection of RNA duplex by double RNase A protection assay. A, the scheme of cRNA probe preparation for second RNase A protection assay. An open box represents the OLR (291 bp), shaded box represent downstream of the OLR (39 bp), and black box represents the vector derived extra nucleotides on the 5' leader of COS cell expressed c27/25 mRNA. H, HindIII; E, EcoRI; S, SmaI. The single bold line represents the length of the cRNA probe and the double lines represent protected fragments seen in lanes 1-4 corresponding to the length of the probe. B, 30 µg of C-RNAs and 10 µg of R-RNAs isolated from female rat liver and 1:4 COS cells were hybridized to 1 × 106 cpm of cRNA probe for 16 h at 55 °C and subjected to RNase A digestion. The protected species were melted and fractionated through a 5% acrylamide, 8 M urea sequencing gel. Lane 1, C-RNA from COS cells. Lane 2, R-RNA from COS cells. Lane 3, C-RNA from liver. Lane 4, R-RNA from liver. Lane 5, 342 nt antisense cRNA probe. M, marker (T and C nts of a sequencing ladder). The expected length of different protected fragments is indicated.
[View Larger Version of this Image (24K GIF file)]


DNA Transformation in COS-1 Cells

COS-1 cells were cultured in Dulbecco's modified Eagle's medium with high glucose supplemented with 10% fetal calf serum. Twenty-four hours after plating on 10-cm dishes, cells were subjected to transfection by the DEAE-dextran method as described previously (1, 13). The above described plasmid DNAs pCMVc27/25, pCMVSPI, pCMVSPIrev, and OLRCAT were isolated by CsCl density banding, mixed in different ratios to form 20 µg of DNA mixtures, ethanol precipitated, and used for cotransfection into COS cells. Each 10-cm COS cell dish was transfected with 20 µg of DNA mixture containing either pCMVc27/25:pCMVSPI or pCMVc27/25:pCMVSPIrev or OLRCAT:pCMVSPI or OLRCAT:pCMVSPIrev plasmids prepared in 1:1, 1:2, and 1:4 ratios. DNA mixtures were prepared by mixing 4 µg of pCMVc27/25 or OLRCAT plasmids with 4 or 8 or 16 µg of pCMVSPI or pCMVSPIrev plasmids and in lower ratio mixtures lesser amounts of SPI plasmids were compensated with pCMV4 vector DNA. The COS cells were harvested 44 h after transfection for RNA isolation and 65 h after transfection for CAT assay and isolation of crude mitochondrial fractions. Expression of c27/25 protein was monitored by Western blot analysis using monoclonal antibody to c27/25 as probe as described (40, 46). COS cell lysates were prepared and used for CAT assay essentially as described before (41).

RNA Preparation and Detection of RNA Duplex

RNase A-resistant RNA was isolated from the liver and COS cells cotransfected with c27/25 and SPI2.1 DNAs under nondenaturing conditions and used for the detection of RNA duplex by the double RNase protection technique and the reverse transcription-polymerase chain reaction (RT-PCR). Cytoplasmic RNA was crudely isolated without denaturing any potential double-stranded RNA from female rat (120-150 g Sprague-Dawley)) liver and COS-1 cells carrying c27/25 and SPI2.1 plasmids essentially as described before (29). Liver tissue and COS cells 44 h after transfection were homogenized with a Dounce homogenizer in 5 volumes of 10 mM Tris-HCl (pH 7.4) buffer containing 3 mM CaCl2, 2 mM MgCl2, and 0.5% Nonidet P-40 and the nuclei were pelleted at 500 × g for 10 min at 4 °C. The supernatant containing the cytoplasmic RNA was made 200 mM Tris-HCl (pH 7.4), 25 mM EDTA, 1% SDS; 0.4 mg of protease K/ml was added, and the incubation was carried out for 30 min at 42 °C. The supernatant was then phenol-chloroform extracted, and RNA and DNA were precipitated with equal volumes of isopropyl alcohol. The pellet was dissolved in 20 mM Tris (pH 7.4) containing 10 mM NaCl, and 6 mM MgCl2 and 100 µg/ml RNase A and 40 units/ml RQ1 DNase were added, and the mixture was incubated at 37 °C for 1 h. The protease K treatment was repeated for 15 min and the reaction mixture was phenol-chloroform extracted, the RNase A-resistant double-stranded RNA (R-RNA) was precipitated with equal volumes of isopropyl alcohol. The RNA pellet was suspended in water and concentration was determined by optical density at 260 nm. C-RNA was obtained by the same procedure, except that RNase A was replaced with RNasin-RNase inhibitor (1,000 units/ml).

R-RNAs and C-RNAs were denatured at 94 °C for 2 min in the presence of antisense primer PAS-1 or PAS-2 (5'-AAGCGAGATCCTTTCATAAC-3'), chilled on ice, and reverse transcribed using avian myeloblastosis reverse transcriptase as described (13). The DNA strand was then amplified by PCR with the help of a sense primer PS-1 from the 5' end of the OLR. The PCR amplification was carried out using a two-step program (94 °C, 30 s, 55 °C, 1 min) for 35 cycles with a final extension for 7 min at 72 °C. Amplified products of RT-PCR were visualized by ethidium bromide staining on 1.6% agarose gel.

A double RNase A protection assay was performed essentially as described by Krystal et al. (29). Plasmid constructs 342pGEM and 280pCR3 were used to prepare sense and antisense radioactively labeled ([32P]UTP) cRNA probes by in vitro transcription using SP6 and T7 RNA polymerase as recommended by the manufacturer (Promega Biotech) after linearization of the plasmid in the polylinker as shown in Figs. 2A and 3A. Annealing of the different cRNA probes (1 × 106 cpm) to the RNA sample was performed at 55 °C overnight as described (13). The hybridized product was then subjected to a second RNase treatment (100 µg of RNase A/ml) in 200 mM NaCl, 100 mM LiCl, 10 mM Tris-HCl (pH 7.4), 1 mM EDTA for 30 min at 30 °C. The double RNase-protected RNA species were phenol-chloroform extracted, ethanol precipitated, and electrophoresed on a 5% acrylamide gel, 8 M urea.

Northern Blot Analysis

Total RNA was isolated from transfected COS-1 cells by the acid guanidine thiocynate-phenol chloroform technique of Chomezynski and Sacchi (42). Denatured total RNA was resolved on 1.2% formaldehyde-agarose gels and transferred to Nylon membranes as described (43). A 330-nt EcoRI-SmaI double-stranded fragment from the 5' end of c27/25 cDNA was randomly labeled with [32P]dCTP and used as a probe. This probe was capable of distinctly hybridizing to both c27/25 (2.3 kb) and SPI2.1 (1.8 kb) mRNAs on the same blot. Hybridizations in the presence of 50% formamide, at 42 °C for 18-24 h and post-hybridization washing of the blots were carried out as described (13, 43).

In Vitro Translation and Immunoprecipitation

COS-1 cells were transfected with pCMVc27/25 or pCMVSPI or pCMVSPIrev plasmids under identical conditions and used as an abundant and individual source for their RNAs. Poly(A)+ mRNA was isolated from these COS cell groups using Invitrogen Fast Tract mRNA isolation Kit and the manufacturers protocol. The levels of c27/25, SPI, and SPIrev RNAs in three RNA preparations were quantitated by dot blot analysis (43) using the same probe as described for Northern blot analysis. The c27/25 COS poly(A)+ RNA was mixed with SPI COS poly(A)+ RNA to give 1:2, 1:4, 1:6, and 1:8 ratios of c27/25:SPI2.1 RNAs in the RNA mixture. Similar ratios were prepared between c27/25 and SPIrev COS poly(A)+ RNAs to act as controls. The lesser amount of SPI2.1 COS RNAs in lower ratios was compensated with normal COS cell poly(A)+ RNA. Each RNA mixture (10 µg) was denatured at 65 °C for 5 min and then cooled to 37 °C in 20-30 min before subjecting them to in vitro translation in a rabbit reticulocyte lysate system. In vitro translation was carried out in a 24-µl reaction volume containing 10 µg of RNA mixture and 1 µCi of [35S]methionine/µl as described (44, 45). In vitro translation products containing about 1 × 107 acid precipitable counts/min were solubilized by boiling in a buffer containing 1% SDS and immunoprecipitated with polyclonal antibody (20 µg of ammonium sulfate-fractionated IgG) against c27/25 (46) essentially as described (44, 45). The immunoprecipitated proteins were released by boiling in Laemmli's sample buffer, resolved by electrophoresis on a 12% polyacrylamide gel containing SDS (47), and gels were subjected to fluorography.


RESULTS

P450c27/25 and SPI mRNAs Form Duplex in Vivo

RNA duplex formed between c27/25 and SPI mRNAs in vivo was detected by RT-PCR of RNase A-resistant RNA and the double RNase A protection procedure. COS-1 cells which do not normally express both c27/25 and SPI mRNAs (1, 13) provided an ideal system for studying the RNA duplex formation. Cytoplasmic RNA was crudely isolated without denaturing any potential double-stranded RNA from female rat liver and COS cells carrying c27/25 and SPI2.1 plasmids. Crudely isolated RNA was subjected to a high concentration RNase A digestion and RNase A-resistant RNA (R-RNA) was precipitated with isopropyl alcohol. C-RNA was prepared in the same fashion except that placental RNase inhibitor (RNasin) was added instead of RNase A during the initial incubation. DNA contamination was eliminated from both R-RNA and C-RNA samples by a high concentration DNase treatment.

If RNA duplex corresponding to specific sequences is formed in vivo, several predictions can be made as to the results of the procedure. First, if the initial RNase A digestion is complete, only the primers from within the OLR should yield amplified product with RT-PCR and no matter what the ratio of c27/25:SPI mRNAs is in the liver, equal amounts of the complementary sense and antisense sequences should be detected by double RNase A protection assay. Second, heat denaturation of cytoplasmic RNA before RNase A treatment should also destroy the RNA duplex and not yield any amplified product with RT-PCR or protected fragments with double RNase A assay. Third, no amplified product should be detected if reverse transcription reaction is omitted from RT-PCR. Finally, the complementary sense and antisense RNAs should be of the same size, providing that all RNase A is eliminated before denaturation of the resistant material. The results illustrated in Figs. 1, 2, 3 exactly conform with the predictions as follows.


Fig. 1. Detection of RNA duplexes by RT-PCR of RNase A-resistant RNA. Samples of R-RNAs and C-RNAs isolated from the liver and COS cells carrying c27/25 and SPI2.1 plasmids in 1:4 ratio were subjected to RT as described under "Experimental Procedures." Half of the RT reaction mixtures (10 µl) were amplified using PCR. Amplified products were visualized by ethidium bromide staining after electrophoresis on 1.6% agarose gel. Samples in lanes 1-3 and 8-11 were reverse transcribed with PAS-1 primer and amplified with PS-1 and PAS-1 primers. Samples in lanes 4-7 were reverse transcribed with PAS-2 primer and amplified with PS-1 and PAS-2 primers. Lanes 1, 2, 3, 7, and 11, rat liver R-RNAs and lanes 5 and 9, COS cell R-RNAs. Lanes 6 and 10, rat liver C-RNAs and lanes 4 and 8, COS cell C-RNAs. The marker is PUC18 digested with HaeIII. Fast moving bands represent primer dimers as they were also observed in the absence of the template. The construct at the top shows the position of different primers used.
[View Larger Version of this Image (37K GIF file)]


As shown in Fig. 1, RT-PCR amplification of R-RNA of liver (lane 11) and COS cells (lane 9) yielded an expected 280-bp amplified product. The same size of amplified product was identified using C-RNAs as template (lanes 10 and 8, respectively). The correct sequence of the amplified products was confirmed by direct sequencing of PCR amplified products (results not shown). When PAS-1 primer was replaced with outside antisense primer PAS-2, R-RNA from both liver (lane 7) and COS cells (lane 5) failed to yield amplified product, while both C-RNAs as expected gave 301-bp amplified product (lanes 6 and 4, respectively). RNase A treatment at 68 °C after denaturation of liver R-RNA at 94 °C abolished all protected species and failed to yield the amplified product (lane 3). RT-PCR of R-RNA of normal COS cells (lane 2) and direct PCR amplification without reverse transcription of R-RNA of cotransfected COS cells (lane 1) also failed to yield amplified product.

The results obtained from double RNase protection assay are presented in Fig. 2. As illustrated in Fig. 2A, a 342-nt antisense cRNA probe, consisting of 291 nucleotides of the OLR, 39 nucleotides from the 5'-UTR downstream of the OLR, and 12 nucleotides from the Bluescript vector was prepared from the HindIII-SmaI region of a 2.3-kb c27/25 cDNA cloned in pGEM-3Z vector. The probe was used to detect duplex RNA species from R-RNAs prepared from liver and cotransfected COS cells. As expected, a 330-nt protected fragment is observed with liver C-RNA (lane 3) and a 342-nt protected fragment with COS C-RNA (lane 1). The larger protection with COS C-RNA is due to the presence of 12 extra nucleotides on the 5' leader of COS cell expressed c27/25 mRNA derived from the vector. On the other hand, R-RNAs from both liver and COS cells show an identical 291-nt protected fragment corresponding to the whole length of the OLR (lanes 4 and 2, respectively). However, liver R-RNA also shows several smaller species in addition to the 291-nt product. The two prominent smaller 162 and 132 protected fragments could be a result of RNA duplex formed between c27/25 and SPI2.2 mRNAs, because there exists a 2-nt mismatch between SPI2.2 and c27/25 mRNAs at a position that would definitely result in 132- and 160-nucleotide fragments after RNase A digestion (15, 16). These results clearly suggest that RNA duplex is formed between SPI and c27/25 mRNAs in vivo and most of the c27/25 mRNA may exist in hybrid form.

Quantitative Evaluation of Duplexes Formed between c27/25 and SPI mRNAs in Vivo

The levels of SPI RNAs that can form duplexes with c27/25 mRNA in the liver are about 10-15-fold. At a similar ratio, the possibility that all c27/25 mRNA being involved in duplex formation was investigated in COS cells carrying c27/25 and SPI2.1 plasmids in a 1:4 ratio (1:4 COS cells) by double RNase protection assay and also RT-PCR. 1:4 COS cells express c27/25 and SPI2.1 mRNAs in a ratio comparable to that observed in rat liver (Figs. 3 and 4), and thus provided a good model for quantitative evaluations of RNA duplex formation between SPI and c27/25 mRNAs in the liver. Both C-RNA and R-RNA were isolated from the same number of 1:4 COS cells under nondenaturing conditions as described under "Experimental Procedures" and the final isopropyl alcohol-precipitated RNA pellet was suspended in an equal volume of water. The total yield of R-RNA was about 8-10% that of C-RNA. If all c27/25 RNA molecules exit in duplex form, then the number of c27/25 RNA molecules in the C-RNA preparation should be equal to the number of protected c27/25 RNA fragments in the R-RNA preparation. In other words, equal volumes of the two RNA preparations, if subjected to a second RNase protection assay with an antisense probe that hybridizes to c27/25 mRNA, should give protected fragments of comparable intensity as estimated by densitometric evaluation of an appropriately exposed autoradiogram. If only a fraction of the c27/25 mRNAs exists in duplex form, then the level of the protected fragment from R-RNA should be comparably low. However, the level of protected fragment that hybridizes to antisense probe should be similar to that hybridizing to sense probe under all experimental conditions. As is illustrated in Fig. 3B, the sense and antisense cRNA probes of equal specific activities appear to detect a large excess of native SPI mRNA compared to c27/25 mRNA from C-RNA (lanes 1 and 3, respectively) and equal amounts of resistant material from R-RNA (lanes 5 and 7, respectively), as was expected if duplex molecules exist. However, the intensity of protected fragments in lanes 5 and 7 is slightly higher, but comparable to the c27/25 mRNA level in lane 3. Although the levels of protected fragments from R-RNAs (lanes 5 and 7) should have been the same or less than the levels of c27/25 mRNA detected by the antisense probe from C-RNA (lane 3), their higher intensity can be attributed to better access of shorter protected fragments to hybridize to cRNA probes compared to the intact c27/25 mRNA. The results therefore suggest that most of the c27/25 mRNA exists in duplex form. Furthermore, as predicted, heating the cytoplasmic RNA to 94 °C for 2 min before RNase A treatment at 68 °C completely abolished protected fragments with both probes (lanes 6 and 8) indicating that the protected fragments represent double-stranded RNA and not RNA-DNA hybrids.


Fig. 4. Quantitative analysis of RNA duplexes by RT-PCR. A, total RNA was isolated from COS cells carrying c27/25 and SPI plasmids in different ratios and used for Northern blot analysis as described ("Experimental Procedures"). A 330-nt EcoRI-SmaI double-stranded fragment from the 5' end of c27/25 cDNA was randomly labeled with [32P]dCTP and used as probe. This probe was capable of distinctly hybridizing to both c27/25 (2.3 kb) and SPI2.1 (1.8 kb) mRNAs on the same blot. Lane 1, COS cells cotransfected with c27/25:SPIrev in 1:4 ratio. Lanes 2-4, COS cells cotransfected with c27/25:SPI2.1 plasmids in 1:1, 1:2, and 1:4 ratios, respectively. Lane 5, normal COS-1 cell RNA. B, 5 µg of C-RNAs from COS cell groups described in A, lanes 1-4, were subjected to RT-PCR. The primer PAS-1 was annealed to RNAs, prior to RT-PCR, after denaturation at 65 °C (panel I) or after melting at 94 °C for 2 min (panel II). Similarly, R-RNA from the same COS cell groups were subjected to RT-PCR after denaturation at 65 °C (panel III) and after melting at 94 °C for 2 min (panel IV). Corresponding lanes in all the four panels represent the same samples as mentioned at the top. The expected position for a 280-bp amplified PCR product is indicated by arrowheads.
[View Larger Version of this Image (34K GIF file)]


The possibility that the duplexes were formed after cell lysis and did not exist in vivo was investigated by adding an excess of labeled cRNA sense or antisense probe to the 1:4 COS cells during cell lysis. The R-RNAs isolated from these COS cells were directly analyzed on the sequencing gel. As shown in Fig. 3C, lane 1, the antisense probe is unable to protect any c27/25 mRNA fragment in the 1:4 COS cell lysate, while COS cell extracts exposed to the 280-nt sense probe show a minor protected fragment of the same size (lane 2). However, the same sense probe is unable to protect any fragment in control COS cells (lane 3) indicating that the minor protected fragment in lane 2 is a result of hybrid formation between sense probe and SPI2.1 mRNA. However, inefficient hybridization of sense probe to abundant free SPI2.1 mRNA and absence of protected fragments from COS cells exposed to antisense probe clearly demonstrate that the RNA duplexes between c27/25 and SPI mRNAs exist prior to cell lysis.

The possibility that a higher abundance of SPI mRNA is necessary for most of the c27/25 mRNA to exist in duplex form was analyzed in different groups of COS cells cotransfected with c27/25:SPI2.1 plasmids in 1:1 (1:1 COS), 1:2 (1:2 COS), and 1:4 (1:4 COS) ratios under identical conditions. COS cells cotransfected with c27/25:SPIrev in a 1:4 ratio (1:0 COS) were used as control. The c27/25 and SPI2.1 mRNA levels in these COS cell groups were analyzed by Northern blot analysis using a random labeled OLR probe capable of distinctly hybridizing to both c27/25 and SPI mRNAs on the same blot. As is evident from Fig. 4A, the levels of the 2.3-kb RNA species which represents c27/25 mRNA are almost the same in all COS cell groups. However, as expected, the 1.8-kb SPI mRNA levels increase with the increase in the ratio of transfected c27/25:SPI plasmids from 1:1 to 1:4 (lanes 2-4). As observed during double RNase A protection assay (Fig. 3B), Northern blot analysis also indicates that the COS cells carrying the c27/25 and SPI2.1 plasmids in 1:2 and 1:4 ratios express 5-10-fold higher levels of SPI2.1 mRNA than c27/25 mRNA.

The extent to which c27/25 mRNA exists in duplex form in these four groups of COS cells was analyzed by RT-PCR of C-RNAs and R-RNAs isolated from these COS cell groups under identical conditions. The annealing of primers to the RNAs before reverse transcription was carried out at two different temperatures. Normally, primers are annealed to RNA by heating them together at 65 °C for 5-10 min then cooling the mixture down to room temperature. But 65 °C for 5 min will not be enough to melt the 291 bp, generally more stable, RNA hybrid (48) formed between SPI2.1 and c27/25 mRNAs as the melting temperature for this hybrid lies above 90 °C. Thus, only those RNA molecules which do not exist in duplex form will be available to anneal with a primer from within the overlap region at 65 °C. Therefore, annealing of PAS-1 primer to equal amounts of C-RNAs from 1:1, 1:2, and 1:4 COS cell groups at 65 °C and subsequent RT-PCR should reveal the extent of RNA duplex formation by comparing the levels of their amplified products with the amplified product of similarly annealed C-RNA from 1:0 COS cells. Fig. 4B, panel I, shows PCR amplified products of C-RNAs from four COS cell groups with annealing carried out at 65 °C. There is almost 60% less amplified product seen in 1:1 COS and almost 80-90% less seen in 1:2 and 1:4 COS cells compared to 1:0 COS cells. When the same C-RNAs were melted at 94 °C for 2 min before annealing, the levels of amplified product are almost the same in all the four COS cell groups (panel II). This indicates that c27/25 mRNAs were present under both conditions but were able to anneal only after melting the duplex at 94 °C. R-RNAs from the same COS cell groups when annealed at 65 °C resulted in no amplification (panel III), but when heated to 94 °C before RT gave amplified products with cotransfected COS cell R-RNAs and, as expected, no product with R-RNA from 1:0 COS cells (panel IV). These results clearly demonstrate that most of the c27/25 mRNA exists in duplex form only when the levels of SPI mRNA are severalfold higher. Furthermore, very low levels of amplified products seen in 1:4 COS cells could be the result of RT-PCR of those c27/25 mRNAs that either escaped the duplex formation or most probably were released from the complex by unwindase/modificase activity.

RNA Duplex Formed in Vivo Blocks c27/25 mRNA Translation

Existence of most of the c27/25 mRNA in duplex form means that its 5' ends are blocked and unavailable for scanning ribosomal subunits to initiate translation from the downstream AUG codon. Therefore, to investigate the effect of RNA duplex formation on the translation of c27/25 mRNA, COS cells were cotransfected with c27/25 and SPI2.1 plasmids in different ratios and used as an in vivo model. The expression of c27/25 protein in the transfected COS cells was monitored by Western blot analysis using monoclonal antibody to c27/25 as probe. As shown in Fig. 5A, cotransfection of c27/25 and SPI2.1 plasmids into COS cells results in highly significant inhibition of c27/25 mRNA translation. COS cells carrying two plasmids in a 1:1 ratio show about 50% less c27/25 antigen than corresponding control COS cells (1:1, c27/25:SPIrev). The percent inhibition increases with the increase in c27/25:SPI ratio from 1:1 to 1:4 with about 75% inhibition observed in COS cells carrying plasmids in 1:4 ratio. The replacement of the c27/25 translated region with CAT as the reporter gene and its cotransfection into COS cells with the SPI2.1 plasmid resulted in a similar pattern of inhibition (Fig. 5B). The maximum of 66% inhibition of CAT activity was observed in the 1:4 COS cell group. Since c27/25 mRNA levels in all four of the COS cell groups are not affected by cotransfection, the only reason for the observed inhibition of translation could be the formation of duplex, which, by blocking the entrance of scanning ribosomes represses the initiation of translation. Furthermore, the percent inhibition observed in 1:1 COS cells correlates very well with the percent of c27/25 mRNA that exists in duplex form. However, only 66% inhibition of CAT activity in 1:4 COS cells compared to about 90% c27/25 mRNA engaged in duplex form suggests that c27/25 mRNA is released from the duplex by some type of cytoplasmic processing.


Fig. 5. Effect of cotransfection on c27/25 mRNA translation in vivo. A, mitochondrial fractions were prepared from COS cells carrying c27/25:SPI2.1 or c27/25:SPIrev plasmids in different ratios and used for Western blot analysis as described ("Experimental Procedures"). The immunodetection of c27/25 protein bands on the nitrocellulose membranes was carried out as described previously (13) using monoclonal antibody to c27/25 as probe and appropriate secondary antibodies conjugated to alkaline phosphatase. The relative band intensities on the immunoblots (Histogram) were scanned and quantified using a Scanmater 3 reflecting densitometer. Ratios of c27/25:SPI are: lane 3, 1:1; lane 5, 1:2; and lane 7, 1:4; and the ratios of c27/25:SPIrev are: lane 2, 1:1; lane 4, 1:2; and lane 6, 1;4. Lane 1, 1 µg of pure c27/25 protein. The expected position for a 52-kDa c27/25 protein has been indicated. B, cytosolic fractions were prepared from COS cells cotransfected with OLRCAT:SPI or OLRCAT:SPIrev plasmids in different ratios and subjected to in vitro reconstitution of CAT activity. The CAT activity was quantitated by PhosphorImager and autoradiographed on a x-ray film. The histogram represents the % relative CAT activities observed with different COS cell groups (C, chloramphenicol; AC, acetylchloramphenicol).
[View Larger Version of this Image (49K GIF file)]


SPI mRNA Blocks c27/25 mRNA Translation in Vitro

The effect of SPI mRNA abundance on the translation of c27/25 mRNA was further investigated in rabbit reticulocyte lysate in the presence of 35S-labeled methionine. The COS cells carrying c27/25 or SPI2.1 or SPIrev plasmids provided an individual and abundant source of their mRNAs. After normalizing their mRNA levels (see "Experimental Procedures"), c27/25 COS cell poly(A)+ RNA was mixed with either SPI COS cell RNA or SPIrev COS cell RNA in 1:2, 1:4, 1:6, and 1:8 ratios, respectively. The RNA mixtures were subjected to in vitro translation in the rabbit reticulocyte system and translation products were electrophoresed either directly or after immunoprecipitation with c27/25 antiserum on a polyacrylamide gel. Fluorography results are presented in Fig. 6. Lanes 1 and 2 show total translation products of c27/25:SPI and c27/25:SPIrev RNA mixtures, respectively. As expected, immunoprecipitation of total translation products yielded a 54-kDa component representing the c27/25 precursor form. The intensity of immunoprecipitated bands remained the same when RNA mixtures of c27/25:SPIrev in 1:2 to 1:6 ratios were translated in vitro and immunoprecipitated (lanes 3-5, respectively). However, the effect of SPI2.1 mRNA on the c27/25 mRNA translation resembles the inhibition of translation observed in vivo, but the degree of inhibition is much higher. The inhibition of translation increases with the increase in the ratio from 1:2 to 1:6 (lanes 6-8) but levels somewhat thereafter (1:8 ratio; lanes 9). At a ratio of 1:6, about 90% of c27/25 mRNA translation is suppressed compared to about 60% inhibition in 1:2 COS cell group with comparable ratios of the two mRNAs. Similar degree of inhibition was observed when in vitro transcribed c27/25 and SPI transcripts were subjected to translation in the rabbit reticulocyte system in vitro (results not included), thus directly demonstrating the specificity of the two mRNAs to anneal and interrupt translation. The higher rate of inhibition observed in vitro may be due to an inability of the rabbit reticulocyte lysate to process the RNA duplex as is presumed to be happening in the COS cells. These in vitro cotranslation results further suggest that the inhibition of c27/25 mRNA translation is the result of RNA duplex formation. In summary, the complementary sequences of c27/25 and SPI mRNAs can form hybrids both in vivo and in vitro and result in down-regulation of c27/25 mRNA translation.


Fig. 6. Effect of SPI mRNA on the translation of c27/25 mRNA in vitro. Poly(A)+ mRNA isolated from COS cells carrying pCMVc27/25 plasmid was mixed with poly(A)+ RNA isolated from COS cells carrying either pCMVSPI or pCMVSPIrev plasmids as described under "Experimental Procedures." 10 µg of each RNA mixture were translated in a rabbit reticulocyte system using 1 µCi/µl [35S]methionine. The lysates were subjected to centrifugation at 15,000 rpm for 15 min in an Eppendorf centrifuge and clear supernatants (107 cpm) were used for immunoprecipitation with polyclonal antibody to c27/25 as described before (13). Lanes 1 and 2, 2 µl of in vitro translation product of c27/25:SPI and c27/25:SPIrev RNA mixtures, respectively. Lanes 3-5, immunoprecipitates of c27/25:SPIrev RNA mixtures in 1:2, 1:4, and 1:6 ratios, respectively. Lanes 6-9, immunoprecipitates of c27/25:SPI RNA mixtures in 1:2, 1:4, 1:6, and 1:8 ratios, respectively. The expected position for a 54-kDa c27/25 precursor protein has been indicated.
[View Larger Version of this Image (45K GIF file)]



DISCUSSION

The results presented here show that the complementary sequences between c27/25 and SPI mRNAs form RNA duplexes in vivo and result in the suppression of c27/25 mRNA translation. Several approaches were used to demonstrate the existence of RNase A-resistant transcripts corresponding to sense and antisense sequences derived from the 5'-overlap region between c27/25 and SPI mRNAs in rat liver and also in COS cells carrying their genes. The results presented clearly indicate that the RNase resistance does not arise from self-protection by the individual transcripts. If that were the case, one would expect the large excess of SPI transcripts to reflect its higher mRNA abundance in the liver; instead, equivalent amounts and equal sizes of c27/25 and SPI2.1 RNAs were detected. Furthermore, the results clearly demonstrate that the duplex in liver is not formed only between c27/25 and SPI2.1 mRNAs which show 100% sequence complementarity in the overlap region but also with SPI2.2 which shows a two-nucleotide mismatch with c27/25, hence resulting in two protected fragments after double RNase protection assay (Fig. 2). The RNA duplexes identified are unlikely to be the result of DNA contamination or RNA-DNA hybrids since cytoplasmic supernatants were routinely treated with DNase I. Furthermore, instead of RT-PCR, direct amplification of R-RNA by PCR did not yield any amplified product, clearly indicating that the amplification of R-RNA by RT-PCR is not a result of DNA contamination. All these observations strongly suggest that the most likely explanation for the RNase A resistance is formation of an RNA duplex between the complementary sequences of c27/25 and SPI mRNAs. Denaturation of cytoplasmic RNA before RNase A digestion at 68 °C, which results in the complete loss of amplification, supports this conclusion. Finally, only the primers from within the OLR yielding amplified product with RT-PCR from R-RNA and their sequence analysis demonstrates that the duplex consists of complementary sequences that form the overlap region between these two mRNAs.

Detection of duplex formation between cell RNAs and antisense transcripts has been reported in several cell systems (29, 31-35). Formation of RNA duplexes between cellular RNAs is further addressed by the presence of unwindase/modificase enzyme activity in several mammalian cell lines which not only unwinds long RNA duplexes but also inhibits their reformation through the conversion of adenosine to inosine residues (36-39). In mouse L cells transfected to express an antisense herpes simplex virus thymidine kinase RNA, most herpes simplex virus thymidine kinase RNA was found in duplex form and the thymidine kinase activity was reduced 80-90% under these conditions (31). Yokoyama and Imamoto (32) reported a 70% reduction in the constitutive production of Myc protein as a result of duplex formation between Myc RNA and antisense transcripts, when intracellular antisense RNA transcript was present at 10-20-fold that of Myc RNA in human promyelocytic leukemia cell line HL-60. The existence of an RNA duplex formed between N-Myc mRNA and its endogenous antisense transcript was first detected by Krystal et al. (29). They found most of the cytoplasmic antisense RNA involved in duplex formation with about 5% of the sense N-Myc mRNA and suggested that duplex formation could modulate RNA processing by preserving a population of N-Myc mRNA. Most of these (29, 31-35) and other (49) studies have concluded that an excess of antisense RNA is needed for the native sense RNA to be completely involved in duplex formation and obtain phenotypic effect. Furthermore, it has also been suggested that in cells, the efficiency of interaction between sense and antisense RNAs may vary depending on the length and/or the primary structure of the RNAs involved (49). The possibility that all c27/25 mRNA could be involved in hybrid formation in the presence of abundant SPI2.1 mRNA was investigated in COS cells carrying c27/25 and SPI2.1 plasmids. Both double RNase protection assay (Fig. 3) and RT-PCR of RNase A-resistant RNA (Fig. 4) revealed that almost all c27/25 mRNA exists in duplex form in the presence of abundant SPI2.1 mRNA. Furthermore, the addition of an excess of labeled sense or antisense cRNA probes to the cells during lysis shows that RNA duplexes did not form during RNA preparation but existed in COS cells before cell lysis was carried out.

Introduction of antisense transcripts into cells or whole organisms, either by microinjection or by transformation with appropriate gene constructs, has been widely used to interfere with endogenous gene expression (26, 27). This has not only provided insight into the function of specific genes but also had enormous impact on applied aspects, especially in plants (50). Despite the usefulness of artificial antisense constructs in experimental regulation of gene expression, there is only limited evidence for the participation of endogenous antisense transcripts in eukaryotic gene expression (28-30), and no evidence for their participation in translational regulation. However, it should be noted that in all cases of reported gene overlaps, except c27/25, the overlapping region is located at or near the 3'-untranslated regions (18-22) or overlapping exists between a functional and a poorly expressed nonfunctional mRNA (23-25, 29). In this regard the overlap at the 5'-UTR of c27/25 mRNA with the 5'-coding region of SPI2.1 mRNA is unique and probably represents the first observation of 5' overlap between two functional mRNAs. One of these RNAs, SPI2.1, is highly abundant. Higher abundance of SPI mRNA in comparison to c27/25 mRNA suggests that this gene overlap might have a role to play in both regulation of c27/25 mRNA transcription and translation.

The scanning hypothesis of eukaryotic protein synthesis postulates that the initial contact between 40 S ribosomal subunit, associated factors, and mRNA occurs at the capped 5' end of the polynucleotide chain (51, 52). The ribosomal subunit then scans the mRNA until the appropriate initiator codon is reached (52). In the case of c27/25, mRNA translation has to be initiated at the 7th AUG codon resulting in 54-kDa precursor protein (13). Therefore, the scanning hypothesis predicts that the efficiency of c27/25 mRNA translation could be impaired by RNA duplex formation. Deletion mapping of the antisense RNA has shown that the critical region for blocking mRNA translation with an antisense transcript is the 5'-region of the mRNA (53). Formation of an RNA duplex between c27/25 and SPI RNAs should block the translation of both classes of the mRNA species involved, but the effect should be more significant on the translation of c27/25 mRNA as most of it exists in duplex form. Translational control of c27/25 gene expression by RNA duplex formation was investigated in COS cells carrying c27/25 and SPI2.1 plasmids in different ratios and the appearance of c27/25 protein in the COS cell mitochondria was analyzed by Western blot analysis. As is evident from Fig. 5, c27/25 protein synthesis was severely reduced, but not abolished in cotransfected COS cells. The c27/25 mRNA translated region was also replaced with more easily quantitated CAT enzyme to confirm the implication of RNA duplex in the repression of c27/25 translation. However, in the same COS cells c27/25 mRNA transcription was not affected. The partial inhibition of c27/25 translation observed in COS cells could be easily explained if all c27/25 mRNAs were not complexed. But, in 1:4 COS cells where the ratio of c27/25 and SPI2.1 mRNAs is comparable to that observed in liver, and most of the c27/25 mRNAs exist in duplex form, the partial inhibition of c27/25 mRNA translation indicates that RNA duplex must be undergoing some type of cytoplasmic processing to release c27/25 mRNA and make it available for translation. The available cellular mechanisms suggest three possible ways for 5' end blocked c27/25 mRNA to undergo translation, which are summarized as follows. First, subject to the presence of unwindase/modificase activity in the rat liver, translation can occur following the separation of two strands of a double-stranded RNA. Second, translation occurs following the action of a double-strand specific RNase (23, 54) resulting into conversion of the 2.3-kb mRNA into an mRNA of approximately 2.0 kb size. The 2.0-kb c27/25 mRNA should have no difficulty in getting translated as it still possesses more than 100 nucleotides of the 5'-UTR. Finally, translation takes place by less efficient internal ribosome binding mechanism (55, 56) while the two RNAs are still in hybrid form. So far, only polio virus RNA and a cellular RNA for human immunoglobulin heavy chain-binding protein (GRP-78) have been shown to undergo translation through a mechanism of internal ribosome binding (55, 56). Although, no sequence homologies between the 5'-UTR of polio virus RNA and GRP-78 RNA in the internal ribosome-binding regions have been found, we have observed a stretch of 34 nucleotides, 98 nucleotides upstream of the GRP-78 initiation codon (57), showing 74% positional identity with the -38 to -70 nucleotide stretch of c27/25 mRNA (13). This stretch of 34 nucleotides is not involved in hybrid formation, is far from the OLR, and is available for ribosomes to bind and initiate translation. Several approaches are being adopted to identify the mechanism/mechanisms involved in the processing of c27/25:SPI RNA duplex and to establish the mode of c27/25 mRNA translation in the rat liver.

In summary, the existence of an RNA duplex formed between the complementary sequences of two functional cellular mRNAs has been demonstrated. The existence of most of the c27/25 mRNA in duplex form has been shown to severely inhibit the c27/25 protein synthesis in COS cells carrying c27/25 and SPI2.1 mRNAs in a ratio comparable to that observed between these mRNAs in the rat liver. In the same COS cells transcription of the c27/25 mRNA is not affected which, however, may not correctly represent the conditions under which the two mRNAs are transcribed in the liver. This is probably the first report where one functional cellular RNA regulates the translation of another cellular RNA by way of forming RNA duplexes in vivo.


FOOTNOTES

*   The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1    The abbreviations used are: c27/25, cytochrome P-450c27/25; SPI, serine protease inhibitor; OLR, overlap region; 5'-UTR, 5'-untranslated region; RT-PCR, reverse transcription-polymerase chain reaction; CAT, chloramphenicol acetyltransferase; kb, kilobase(s); bp, base pair(s); R-RNA, RNase-A resistant RNA; C-RNA, control RNA; nt, nucleotide(s).

Acknowledgments

I am thankful to Drs. James M. Leonardo and Dinesh Tewari for critical reviews of the manuscript. I thank Dr. Xiang Ma for his help in the preparation of this manuscript and Remus Beretta for helping with the illustrations. I also thank Dr. Narayan G. Avadhani in whose laboratory this work was initiated and the DNA clones and antibody were generated. I am particularly grateful to Dr. Martin Black, Director Drug Liver Unit, for providing me necessary research facilities, without his support this study could not be completed.


REFERENCES

  1. Su, P., Rennert, I., Shayiq, R. M., Yamamoto, R., Zheng, Y. M., Addya, S., Strauss, J. F., and Avadhani, N. G. (1990) DNA Cell Biol. 9, 657-665 [Medline] [Order article via Infotrieve]
  2. Usui, E., Noshiro, M., Ohyama, Y., and Okuda, K. (1990) FEBS Lett. 274, 175-177 [CrossRef][Medline] [Order article via Infotrieve]
  3. Sakaki, T., Akiyoshi-Shibata, M., Yabusaki, Y., and Ohkawa, H. (1992) J. Biol. Chem. 267, 16497-16502 [Abstract/Free Full Text]
  4. Wikvall, K. (1993) Handb. Exp. Pharmacol. 105, 705-718
  5. Jelinek, D., Andersson, S., Slaughter, C. A., and Russell, D. W. (1990) J. Biol. Chem. 265, 8190-8197 [Abstract/Free Full Text]
  6. Bjorkhem, J. (1985) in Sterols and Bile Acids (Danielsson, H., and Sjovall, J., eds), pp. 231-278, Elsevier, Amsterdam
  7. Goldstein, J., and Brown, M. (1990) Nature 343, 425-430 [CrossRef][Medline] [Order article via Infotrieve]
  8. Deluca, H. F., and Schnoes, H. K. (1976) Annu. Rev. Biochem. 45, 631-666 [CrossRef][Medline] [Order article via Infotrieve]
  9. Friedlander, E. J., and Norman, A. W. (1975) Arch. Biochem. Biophys. 170, 731-738 [Medline] [Order article via Infotrieve]
  10. Series 75, 333-379Deluca, H. F. (1980) in Harvey Lectures, Series 75, 333-379
  11. Henery, H. L. (1980) in Vitamin D: Molecular Biology and Clinical Nutrition (Norman, A., ed), pp. 127-148, Marcel Dekker, New York
  12. Cancela, L., Theofan, G., and Norman, A. (1988) in Hormones and Their Action, Part 1 (Cooke, B. A., King, R. J. B., and Van der Molen, H. J., eds), pp. 269-289, Elsevier, Amsterdam
  13. Shayiq, R. M., and Avadhani, N. G. (1992) J. Biol. Chem. 267, 2421-2428 [Abstract/Free Full Text]
  14. Cali, J. J., and Russell, D. W. (1991) J. Biol. Chem. 266, 7774-7778 [Abstract/Free Full Text]
  15. Rossi, V., Rouayrenc, J. F., Paquereau, L., Vilarem, M., and LeCam, A. (1992) Nucleic Acids Res. 20, 1061-1068 [Abstract]
  16. Yoon, J.-B., Berry, S. A., Seelig, S., and Towle, H. C. (1990) J. Biol. Chem. 265, 19947-19954 [Abstract/Free Full Text]
  17. Mullick, J., Addya, S., Sucharov, C., and Avadhani, N. G. (1995) Biochemistry 34, 13729-13742 [Medline] [Order article via Infotrieve]
  18. Spencer, C. A., Gietz, R. D., and Hodgetts, R. B. (1986) Nature 322, 279-281 [Medline] [Order article via Infotrieve]
  19. Adelman, J. P., Bond, T., Douglass, J., and Herbert, T. (1987) Science 235, 1514-1517 [Medline] [Order article via Infotrieve]
  20. Williams, T., and Freed, M. (1986) Nature 322, 275-279 [Medline] [Order article via Infotrieve]
  21. Lazar, M., Hodin, R., Darling, D., and Chin, W. (1989) Mol. Cell. Biol. 9, 1128-1136 [Medline] [Order article via Infotrieve]
  22. Morel, Y., Bristow, J., Gitelman, S. E., and Miller, W. L. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 6582-6586 [Abstract]
  23. Hildebrandt, M., and Nellon, W. (1992) Cell 69, 197-204 [Medline] [Order article via Infotrieve]
  24. Nepveu, A., and Marcu, K. B. (1986) EMBO J. 5, 2859-2865 [Abstract]
  25. Kindy, M. S., McCormack, J. E., Buckler, A. J., Levine, R. A., and Sonenshein, G. E. (1987) Mol. Cell. Biol. 7, 2857-2862 [Medline] [Order article via Infotrieve]
  26. Green, P. J., Pines, O., and Inouye, M. (1986) Annu. Rev. Biochem. 55, 569-597 [CrossRef][Medline] [Order article via Infotrieve]
  27. Van der Krol, A. R., Mol, J. N. M., and Stuitje, A. R. (1988) Biotechniques 6, 958-976 [Medline] [Order article via Infotrieve]
  28. Khochbin, S., and Lawrence, J. J. (1989) EMBO J. 8, 4107-4114 [Abstract]
  29. Krystal, G. W., Armstrong, B. C., and Battey, J. F. (1990) Mol. Cell. Biol. 10, 4180-4191 [Medline] [Order article via Infotrieve]
  30. Kimelman, D., and Kirschner, M. W. (1989) Cell 59, 687-696 [Medline] [Order article via Infotrieve]
  31. Kim, S. K., and Wold, B. J. (1985) Cell 42, 129-138 [Medline] [Order article via Infotrieve]
  32. Yokoyama, K., and Imamoto, F. (1987) Proc. Natl. Acad. Sci, U. S. A. 84, 7363-7367 [Abstract]
  33. Holt, J. T., Redner, R. L., and Nienhuis, A. W. (1988) Mol. Cell. Biol. 8, 963-973 [Medline] [Order article via Infotrieve]
  34. Bunch, T. A., and Goldstein, L. S. (1989) Nucleic Acids Res. 17, 9761-9782 [Abstract]
  35. Melton, D. A. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 144-148 [Abstract]
  36. Bass, B. L., and Weintraub, H. (1987) Cell 48, 607-613 [Medline] [Order article via Infotrieve]
  37. Rebagliati, M. R., and Melton, D. A. (1987) Cell 48, 599-605 [Medline] [Order article via Infotrieve]
  38. Wagner, R., and Nishikura, K. (1988) Mol. Cell. Biol. 8, 770-777 [Medline] [Order article via Infotrieve]
  39. Bass, B. L., and Weintraub, H. (1988) Cell 55, 1089-1098 [Medline] [Order article via Infotrieve]
  40. Shayiq, R. M., and Avadhani, N. G. (1990) Biochemistry 28, 7546-7554
  41. Gorman, C. M., Moffat, L. F., and Howard, B. H. (1982) Mol. Cell. Biol. 2, 1044-1051 [Medline] [Order article via Infotrieve]
  42. Chomezynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156-159 [CrossRef][Medline] [Order article via Infotrieve]
  43. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd Ed., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY
  44. Niranjan, B. G., Raza, H., Shayiq, R. M., Jefcoate, C. R., and Avadhani, N. G. (1988) J. Biol. Chem. 263, 575-580 [Abstract/Free Full Text]
  45. Shayiq, R. M., and Avadhani, N. G. (1990) Biochemistry 29, 866-873 [Medline] [Order article via Infotrieve]
  46. Addya, S., Zheng, Y., Shayiq, R. M., Fan, J., and Avadhani, N. G. (1991) Biochemistry 30, 8323-8330 [Medline] [Order article via Infotrieve]
  47. Laemmli, U. K. (1970) Nature 227, 680-685 [Medline] [Order article via Infotrieve]
  48. Bartles, J. R., and Hubbard, A. L. (1984) Anal. Biochem. 140, 284-292 [Medline] [Order article via Infotrieve]
  49. Wang, S., and Dolnick, B. C. (1993) Nucleic Acids Res. 21, 4383-4391 [Abstract]
  50. Nellen, W., Hildebrandt, M., Mahal, B., Mohrle, A., Kroger, P., Maniak, M., Oberhauser, R., and Sadiq, M. (1992) Biochem. Soc. Trans. 20, 750-754 [Medline] [Order article via Infotrieve]
  51. Kozak, M. (1992) Annu. Rev. Cell Biol. 8, 192-225
  52. Kozak, M. (1986) Cell 44, 283-292 [Medline] [Order article via Infotrieve]
  53. Melton, D. A. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 144-148 [Abstract]
  54. Blomberg, P., Wagner, E. G. H., and Nordstrom, K. (1990) EMBO J. 9, 2331-2340 [Abstract]
  55. Pelletier, J., and Sonenberg, N. (1988) Nature 334, 320-325 [CrossRef][Medline] [Order article via Infotrieve]
  56. Macejak, D., and Sarnow, P. (1991) Nature 353, 90-94 [Medline] [Order article via Infotrieve]
  57. Ting, J., and Lee, A. S. (1988) DNA 7, 275-286 [Medline] [Order article via Infotrieve]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.