Cellular factor affecting the stability of β-globin mRNA

Gene. 62 (1988) 65-74

65

Elsevier GEN 02256

Cellular factor affecting the stability of /?-globin mRNA (Recombinant half-life)

DNA; cell extract; ribonuclease inhibitor; gene expression;

synthetic transcripts;

mRNA

Catherine A. Stolle and Edward J. Benz Jr. Department of Internal Medicine, Hematology Section, Yale University School of Medicine, New Haven, CT 06510 (U.S.A.) Tel. (203) 785-4144 Received

22 May 1987

Accepted

8 October

Received

by publisher

1987 26 October

1987

SUMMARY

Messenger RNAs in eukaryotic cells exhibit a broad range of stabilities in vivo. Globin mRNA has a half life in excess of 50 h, but the half life of the c-myc oncogene mRNA is less than 20 min. Regulation of gene expression may be accomplished by a variety of mechanisms, including altering mRNA stability. We have examined the nuclear and cytoplasmic fractions of cells for factors affecting the metabolism of mRNA. Here we report that a HeLa whole-cell extract contains a factor that protects /_?-globinmRNA from attack by RNases in a mouse erythroleukemia cell cytoplasmic extract. The factor is non-dialysable, inactivated by proteinase K and heat treatment, and resistant to RNase and DNase digestion. The HeLa cell factor resembles placental RNase inhibitor in that the mRNA-protecting activity is effective against RNase A and that treatment of the extract with N-ethylmaleimide completely destroys the protective activity. However, puritied placental RNase inhibitor was unable to inhibit the RNase activity in the MELC cytoplasmic extract. These results suggest that the HeLa cell extract contains an RNase inhibitor (or inhibitors) with an activity or specificity that is distinct from that of placental RNase inhibitor.

Regulation of gene expression in eukaryotes can be achieved by several mechanisms. Although transcriptional regulation is a major determinant of gene expression (Brown, 1981; Darnell, 1982), the stability of the mRNA ultimately determines how long

it can function. The half lives of eukaryotic mRNAs vary greatly. While globin mRNA has a half life in excess of 50 h (Volloch and Housman, 198 l), the half life of the c-myc oncogene mRNA is less than 20 min (Dani et al., 1984). The rate of decay of some mRNAs can be altered in response to physiological signals such as hormone induction (Guyette et al.,

Correspondenceto: Dr. E.J. Benz Jr., Department

kilobases

INTRODUCTION

Abbreviations: Medicine,

Hematology

Section, Yale University

tine, 333 Cedar Street, LCI 812, New Haven,

of Internal

School of Medi-

CT 06510 (U.S.A.)

Tel. (203)785-4153.

0378-l 119/88/$03.50

NEM,

1988 Elsevier Science Publishers B.V. (Biomedical

Division)

base

N-ethylmaleimide;

phenylmethylsulfonyl phate.

0

bp,

pair(s);

or 1000 bp; MELC, PHA,

fluoride;

DTT,

mouse

dithiothreitol;

erythroleukemia

phytohemagglutinin; rNTP,

ribonucleotide

kb, cells; PMSF,

triphos-

66

1979; Wiskocil et al., 1980) or cell cycle (Old and Woodland, 1984; Luscher et al., 1985; Morris et al., 1986). Factors that affect the rate of decay of mRNAs can modulate the phenotypic impact resulting from expression of a particular gene. Covalent modifications of mRNAs [5’caps and poly(A) tails] appear to be necessary for the stability of most mRNAs (Shatkin, 1976; Baralle, 1983). In particular, 5 ’ caps may protect mRNAs from degradation by a 5’ to 3’ exonuclease (Furuichi et al., 1977; Shimotohno et al., 1977; Gedamu and Dixon, 1978). However, recent studies indicate that the presence of certain 5’ or 3’ untranslated sequences, specific mRNA degradative activities, and cellular ribonuclease inhibitor(s) may be involved in regulating mRNA turnover. Specific sequences within the 5’ (Morris et al., 1986) or 3’ (Georgiev and Birnstiel, 1985; Luscher et al., 1985; Shaw and Kamen, 1986) untranslated regions of certain mRNAs appear to protect these mRNAs from exoribonuclease attack or to target them for rapid degradation. The specific RNases responsible for mRNA turnover have not been identified; however, there is some evidence to suggest that eukaryotic mRNAs may be degraded in a 3 ’ to 5 ’ manner by processive exonuclease(s) analogous to the prokaryotic enzymes RNase II and polynucleotide phosphorylase (Donovan and Kushner, 1986; Ross and Kobs, 1986). Several studies suggest that the rate-limiting step in mRNA degradation is endonucleolytic cleavage near the 3’ end that leaves the mRNA susceptible to 3’ exonucleolytic attack (Bergman and Brawerman, 1977; Albrecht et al., 1984; Ross and Kobs, 1986). Finally, many mammalian tissues have been found to contain an inhibitor(s) of neutral RNase activity (for a review, see Blackburn, 1982). Equilibrium between RNases and their inhibitor(s) may thus play a role in the regulated turnover of mRNAs. We have encountered evidence for the presence of an RNase inhibitory activity in HeLa cells. We have recently reported that capped normal j?-globin and PO-39 thalassemic mRNAs are stable in a HeLa whole-cell in vitro transcription extract, but are rapidly degraded in a MELC cytoplasmic extract (Stolle et al., 1987). The stability of capped transcripts in these extracts is commonly ascribed to protection by the 5’ cap structure from a 5’ to 3’ exoribonuclease. Alternatively, the HeLa whole-cell

extract may contain an mRNA-stabilizing factor lacking in the MELC extract. In this paper, we report that the HeLa cell extract contains a protein factor similar, but not identical, to placental ribonuclease inhibitor that protects B-globin mRNA from degradation.

MATERIALS

AND METHODS

(a) Materials Pheny~ethylsu~onyl fluoride (PMSF) and NEM were purchased from Sigma. Proteinase K was purchased from Boehringer-Mannheim; placental ribonuclease inhibitor (RNasin) was from Promega Biotech. Other materials were obtained from suppliers noted below. (b) Cell-free extracts HeLa cells and MELC, strain 745, were cultured as previously described (Stolle et al., 1987). HeLa whole-cell extract was either prepared according to the method of Manley et al. (1980) or purchased from Bethesda Research Laboratories (Eukaryotic In Vitro Tr~sc~ption System) and was used according to the manufacturer’s instructions. A HeLa cell nuclear extract was prepared according to the method of Dignam et al. (1983). In preliminary experiments, we showed that these extracts had the appropriate mRNA splicing activities (Stolle et al., 1987). Cytoplasmic extracts of MELC were prepared either by the method of Dignam et al. (1983) or by nitrogen cavitation (Craine and Komberg, 198 1). (e) Synthesis and capping of j3-globin mRNA Labeled transcripts of normal fl-globin mRNA were prepared according to Green et al. (1983). Briefly, an SP6 plasmid containing a full-length normal pglobin cDNA insert (pSPk&, a gift of Dr. R. Spritz) was linearized with Hind111 and transcribed at 37 “C for 1 h with SP6 RNA polymerase (New England Nuclear) in 40 mM Tris buffer (pH 7.5) containing 6 mM MgCl,, 2 mM spermidine, 10 mM NaCl, 10 mM DTT, RNasin (1 unit/&, 500 PM of each rNTP, and 100 $i of [32P]GTP (400 mCi/mmol; Amersham). The DNA template was

removed by digestion with RNase-free DNase (25 pg/ml; Promega Biotech) for 15 min at 37°C. The RNA was then extracted with phenol-chloroform, centrifuged through a Sephadex G-50 column to remove unincorporated label, ethanol-precipitated, and resuspended in sterile distilled water. Typical yields were 2-6 pg of RNA per pg of DNA template, with a specific activity of approx. 1.7 x 10’ dpm/,ug RNA. A 5’-terminal cap was added to the SP6 RNA transcripts with guanylytransferase (1 unit/pg RNA; Bethesda Research Laboratories) in a 30 ~1 reaction containing 50 mM Tris - HCl (pH 7.9), 1.25 mM MgCl,, 6 mM KCl, 3 mM DTT, RNasin (1 unit/p& 100 PM S-adenosyl methionine, and 300 PM GTP. After incubation at 37 ‘C for 60 min, the RNA was extracted with phenol-chloroform, precipitated with ethanol, and resuspended in sterile distilled water. The efficiency of the capping reaction was estimated by the intensity of a band corresponding to full-length /?-globin mRNA on autoradiograms of denaturing agarose gels of synthetic transcripts incubated for 60 min in a HeLa whole-cell extract (see Fig. 1). Under the conditions used for these measurements, noncapped mRNA is rapidly degraded, but capped mRNA is completely preserved. For the experiments described in this report, only preparations exhibiting approx. SO-loo% stability in HeLa extract after capping were used. Other methods of capping, such as inclusion of cap analogues in the transcription reaction, did not result in efficient synthesis of fulllength transcripts. In addition, we eschewed this approach because cap analogs not removed from the synthesis reaction may interfere with mRNA metabolism by binding factors which recognize cap structures (Filipowicz et al., 1976; Patzelt et al., 1983). The P-globin transcripts prepared as described above contained 20 non-globin bases in the 5’-flanking region and a small number of non-globin bases at the 3’ terminus, but otherwise were identical to naturally occurring j$globin mRNAs in that they contained a 5’ cap, a poly(A) tail, and were readily translated into /?-globin chains in a rabbit reticulocyte lysate in vitro translation extract. (d) Stability

in cell-free extracts

The stability of the labeled transcripts was assessed by electrophoresis on denaturing agarose

gels after incubation in cell-free extracts as described in the legend to Fig. 1.

RESULTS

AND

DISCUSSION

(a) Stability of /I-globin mRNA in cell-free extracts

Equal amounts of capped and non-capped /?-globin mRNAs were incubated in either a HeLa whole-cell extract (prepared as an in vitro transcription/splicing extract) or a MELC cytoplasmic extract for 2 h at 30°C. The stability of the transcripts was assessed as described in the legend to Fig. 1. In this assay, the amount of full-length mRNA remaining at the end of the incubation reflects the relative stabilities of the transcripts in vitro. Capped /?-globin mRNAs were stable in the HeLa cell extract in that most of the input RNA remained full-length for up to 2 h (Fig. 1). Non-capped transcripts, however, were readily degraded in this extract. After only 20 min, no full-length non-capped transcripts remained. In contrast, a cytoplasmic extract prepared from MELC rapidly degraded both capped and noncapped transcripts. This result was consistently observed with several different preparations of mRNA and extracts. The data shown in Fig. 1 agree with previous observations by other laboratories. Green et al. (1983) and Kramer et al. (1984), during studies of globin mRNA splicing, observed that capped transcripts were more stable in HeLa cell extracts and frog oocyte nuclei than uncapped transcripts. These investigators credited the stabilizing effect of the 5’ cap structure to protection of the mRNA from a 5’ to 3’ exoribonuclease activity. An alternative hypothesis is that the HeLa cell extract might contain an mRNA-protecting factor that stabilizes capped transcripts and is lacking in the MELC extract. If this were the case, the HeLa cell extract might be expected to protect mRNAs from degradation by the MELC extract. To test this hypothesis, capped mRNA was preincubated for 15 min in the HeLa whole-cell extract and then challenged by the addition of an equal volume of MELC extract. Pre-incubation of the capped mRNA with HeLa whole-cell extract was sufficient to protect the mRNA from subsequent

68

Fig. 1. Stability oft-Robin mRNA transcripts in cell-free extracts. Equal amounts ofcapped @cap) or non-capped (8) mRNA transcripts were incubated in either a HeLa whole-cell extract (HeLa) or a MELC cytoplasmic extract (MELC) for 2 h at 30°C. A typical assay consisted of 3 x IO’ cpm of mRNA (20 ng) in 30 ~1 of cell extract. Aliquots (5 ~1) removed at the time points indicated were d&ted in 300 ~1TE buffer ( 10mM Tris *HCl, pH 7.5; 1 mM EDTA), extracted with an equal volume of phenol-c~orofo~, ethanol precipitated, and resuspended in a small volume of sterile distilled water. The samples were mixed with 3 ~01s. of a denaturing buffer containing 50% deionized formamide, 17% formaldehyde, and 1 x MOPS buffer (40 mM morpholinopropanesulfonic acid, pH 7.0; 10 mM sodium acetate; 1 mM EDTA), heated to 65°C for 10 mm, and electrophoresed on a 1.5% agarose gel containing 6.7% formaldehyde in 1 x MOPS buffer. The labeled mRNA bands were transferred to nitrocellulose paper by a Northern-blotting procedure (Thomas, 1980), and the blots were exposed to x-ray film (Kodak XAR-5). Variability in loading of labeled mRNA can be observed at some time points and is due to differences in recovery of mRNA after the phenol-chloroform extraction and ethanol precipitation steps. Degradation of capped and non-capped mRNA at zero time in MELC extract was a consistent observation with several preparations of extract and mRNA and can be attributed to degradation that occurred within the time necessary for processing of the 0 time-point sample.

degradation by the MELC extract (Fig. 2). The stability of the mRNA in the HeLa-MELC mixture was not due to a buffer component in the HeLa whole-cell extract since the capped mRNA was completely degraded in a 1: 1 mixture of MELC extract and buffer of the same composition as that used to prepare the HeLa extract (HeLa cell dialysis buffer). Identical results were obtained when a HeLa cell nuclear extract was used in place of the whole-cell extract. The protecting activity was also not unique to HeLa cells, since a similar protecting activity was present in a rabbit reticulocyte lysate, but at a lower concentration (not shown). HeLa extract thus protected mRNAs from degradative activities in the MELC extract that were capable of hydrolyzing capped mRNAs. The stability of capped mRNA under these conditions was clearly not due to the

presence of cap alone. Therefore, we hypothesized that the HeLa extract must contain additional factors that stabilize mRNA. @) Characterization of the protecting com~nent

Since the HeLa cell extract appeared to contain a component that protected mRNAs from degradation, we endeavored to ascertain the nature of the component. To determine whether the protecting activity required the presence of a protein, the HeLa cell extract was treated with 0.3 mg/ml of proteinase K for 60 min at 42’ C, followed by inhibition of the proteinase IS with 10 mM PMSF. Alternatively, the HeLa cell extract was heated to 68” C for 60 min and then cooled before addition of capped mRNA. The

69

HeLa min

0 20 60

HeLa MELC 0

20 60

MELC 0 20 60

Buffer MELC 0

2060

Fig. 2. Protection of/I-globin mRNA by HeLa cell extract. Capped /I-globin mRNA was pre-incubated for 15 min at 30°C in HeLa cell extract (HeLa) or HeLa cell dialysis buffer (12 mM Hepes, pH 7.9; 60 mM KCI; 7.2 mM MgCl,; 0.2 mM EDTA; 1.2 mM DTT; 10.2% glycerol) (Buffer) followed by the addition of an equal volume of MELC extract (MELC). Aliquots taken during a l-h incubation at 30°C were treated as described in Fig. 1. Controls consisted of transcripts incubated in HeLa cell extract alone or MELC extract alone. The p-globin mRNA in HeLa or HeLa-MELC extract mixtures is largely full-length with some smaller material that migrates as a smear. The shorter labeled transcripts in these extracts were most likely due to radiochemical decay of labeled RNA transcripts that were used in several experiments over a period of time and not to RNA degradation during the incubation. Note that the amount of full-length mRNA is the same at 0 and 60 min and that the amount of shorter transcripts does not increase with time. In contrast, reaction mixtures containing MELC extract exhibit a decrease in full-length mRNA from 0 to 60 mm.

treated extracts were then tested for protecting activity by addition of MELC extract as before. In contrast to mRNA incubated in an untreated HeLa cell extract, mRNA incubated in HeLa extract digested with proteinase K showed considerable degradation (Fig. 3). Very little full-length mRNA remained after only the brief exposure to MELC extract required for manipulation of the zero time point sample. Messenger RNA incubated in a HeLa cell extract treated identically except that proteinase K was omitted showed little degradation. The presence of PMSF alone in the reaction mixture thus had no effect on the degradative activity of the MELC extract. The proteinase K and PMSF preparations were found to be free of ribonuclease activity since mRNA incubated with these components, in

the absence of cell extract, remained full-length for more than 60 min (not shown). Heating the HeLa cell extract to 68°C was also effective in decreasing the protective activity of the extract. However, treatment of the HeLa cell extract with DNase I or RNase A (followed by inhibition of the enzyme with RNasin) did not affect the protecting activity of the HeLa cell extract (not shown). Therefore, the only treatments that were effective in reducing or eliminating the protective activity of the HeLa cell extract were those that digested or denatured proteins. (c) Biochemical characteristics

of HeLa cell factor

We next addressed the question of whether or not the protecting component interacted directly with the

70

min

Fig. 3. Treatment of HeLa cell extract with proteinase K or heat. The HeLa cell extract was treated with proteinase K (0.3 mg/ml) for 60 min at 42°C followed by inhibition of the enzyme with 10 mM PMSF. Alternatively, the HeLa cell extract was heated to 68°C for 60 min and then chilled. Capped #?-globinmRNA was preincubated for 15 min at 30°C in proteinase K digested extract (HeLa - Pro K - PMSF), HeLa cell dialysis buffer containing 10 mM PMSF (Buffer - PMSF), and heat-treated HeLa cetl extract (HeLa - 68°C/60’). After the addition of an equal volume of MELC extract (MELC), the mixtures were incubated at 30°C for 60 min. An untreated HeLa cell extract without addition of MELC extract (HeLa) served as control. Samples taken at the time points indicated were treated as described in Fig. 1. M, RNA size markers of 1.38 kb, 0.55 kb, and 0.22 kb.

capped mRNA. Although several independent lines of investigation were undertaken in an effort to demonstrate a direct protein:mRNA interaction, none of these approaches provided any evidence for such an interaction. For example, protection of the labeled mRNA by the HeLa cell extract did not depend on the order of addition of MELC extract or mRNA to the HeLa extract. Messenger RNAs exposed to the HeLa extract and then separated from the bulk of the extract by sucrose gradient centrifugation or oligo(dT) column chromatography were only slightly protected from degradation. Furthermore, dilution of the HeLa extract with increasing amounts of MELC extract resulted in gradual loss of protecting activity (not shown). Finally, addition of increasing amounts of unlabeled poly(A) + RNA did not compete with labeled mRNA

for the protecting activity in the HeLa cell extract (Fig. 4). These experiments suggest that the protecting activity is not an mRNA-b~d~g protein. Rather, the data raise the possibility that the protecting component may be a proteinaceous RNase inhibitor. A class of such proteins has been isolated from human placenta, rat reticulocytes and other rn~rn~~ and non-mammalian tissues. The ribonuclease inhibitor from the human placenta is an acidic protein of 50 kDa that forms a 1: 1 complex with bovine pancreatic RNase A (Blackb~ et al., 1977). Normally, more than 95% of the available RNase activity in the cell is in an inactive complex maintained by a six- to eight-fold molar excess of free inhibitor over enzyme. There is evidence that eq~~b~~ between the neutral ribonuclease and its inhibitor may have a role

71

Fig. 4. Unlabeled poly(A) + RNA fails to compete with labeled @-globin mRNA for HeLa cell protecting activity. Increasing amounts of unlabeled poly(A)+ RNA were added to HeLa-MELC mixtures containing labeled &globin mRNA, and the reactions were incubated at 30°C for 60 min. Aliquots were removed at 0 and 60 min and treated as described in Fig. 1. The MELC extract used in this experiment was one previously shown to degrade mRNA. H, HeLa cetl extract without addition of MELC extract; Poly(A) + RNA, rat liver RNA twice eluted from an oligo(dT) column (i.e., 100 x poly(A)+ RNA = 2 pg of o~g~dT~pu~~ed RNA).

in the regulated turnover of mRNA. The ratio of inhibitor to neutral RNase activity is high in tissues characterized by high rates of RNA synthesis and accumulation (i.e., regenerating liver, PHA-stimulated lymphocytes, and hormonally induced tissues). Conversely, tissues with decreased protein synthesis and increased protein turnover demons~ate lower levels of ~hibitor and elevated RNase activity (Blackbum et al., 1977; reviewed in Blackbum, 1982). Placental ribonuclease inhibitor contains 30 half-cystine plus cystine residues and is irreversibly inactivated by sulfhydryl reagents. Inactivation causes dissociation of the RNase-inhibitor complex into active RNase and inactive inhibitor (Blackbum et al., 1977). To determine if the protecting activity from HeLa cells is similar to the human placental RNase inhibi-

tor, the HeLa cell extract was treated with 5 mM NEM to reduce sulfhydryl groups and then tested for protecting activity as before. NEM treatment completely eliminated the protecting activity of the HeLa cell extract in HeLa-MELC mixtures, but had no effect on the RNase activity of the MELC extract alone (Fig. 5). Also, since NEM inactivated the ribonuclease inhibitors in the HeLa cell extract, previously latent HeLa cell RNase was activated (presumably by release from the RNase inhibitor), and mRNA degradation was observed in HeLa cell extract alone. The NEM, which was prepared with sterile distilled H,O and filtered through a 0.22~pm Millipore f&err, was found to be Free of ribonuclease activity in that labeled mRNA remained full-length during a 60-min incubation in HeLa cell dialysis buffer containing 5 mM NEM (not shown).

12

Fig. 5. Treatment of HeLa cell extract with 5 mM NEM. Capped @lobin mRNA was preincubated for 5 min at 30°C in untreated (HeLa) or NEM-treated (HeLa-NEM) cell extract. Where indicated, an equal volume of MELC extract was added to the incubation mixtures. Alternatively, capped /J-gIobin mRNA was added directly to MELC extract treated with 5 mM NEM (MELC-NEM). Ahquots taken at the time points indicated were treated as described in Fig. 1.

The similarity of the mRNA-prot~t~g activity in HeLa cell extracts and placental RNase inhibitor was also examined by determining if HeLa cell extract could inhibit the activity of bovine pancreatic RNase A (Sigma). Labeled mRNA was incubated at 30 ‘C for up to 60 min in HeLa cell extract containing increasing concentrations of RNase A (Fig. 6). The HeLa cell extract ~~bit~ RNase A at a concentration of 1 pg/ml; full-length mRNA was detectable in this reaction mixture after 60 min of incubation. In contrast, mRNA incubated in HeLa cell dialysis buffer containing 1 pg/ml of RNase A was rapidly degraded. These results suggest that the HeLa-protecting activity is due to a protein similar to placental RNase inhibitor in that the activity is effective against RNase A and sensitive to NEM. To determine if placental ribonuclease inhibitor could substitute for HeLa cell extract in our mRNA protection assay, labeled mRNA was added to HeLa cell dialysis

buffer cont~ing increasing mounts of the ribonuclease inhibitor RNasin and then challenged by addition of either MELC extract or RNase A (1 pg/ml). RNasin was unable to protect the mRNA from degradation by concentrations of MELC extract or RNase A shown to be inhibited by HeLa cell extracts, even when RNasin was present at concen~ations of greater than 32000 units of ribonuclease inhibitor per ml, or approx. 5% of the protein content of the HeLa cell extract (not shown). The inability of placental RNase inhibitor to substitute for HeLa cell extract suggests that HeLa cells contain a ribonuclease inhibitor or inhibitors with an activity or specificity that is different from that of placental RNase inhibitor and that the MELC preparation contains nucleases insensitive to RNasin. The RNase in~bitor(s) in HeLa cells, therefore, is similar, but not identical, to placental RNase inhibitor.

13

Fig. 6. Inhibition concentrations removed

ofRNase

A by HeLa cell extract.

of bovine pancreatic

at the time points indicated

buffer (Buffer) containing

RNase

were treated

1 pg/ml RNase

Labeled &globin mRNA was added to HeLa cell extract (HeLa) containing

A (O-10 pg/ml),

and the reaction

as described

mixtures

in Fig. 1. Labeled

mRNA

were incubated was similarly

increasing

at 30°C for 60 min. Aliquots incubated

in HeLa cell dialysis

A.

(d) Conclusions

(1) HeLa whole-cell and nuclear extracts contain a component that protects mRNA from degradation by RNases in an MELC cytoplasmic extract. (2) The cellular component is non-dialysable, susceptible to proteinase K digestion and heat treatment, and is unaffected by digestion with RNase or DNase. (3) The cellular component does not appear to interact directly with the mRNA. (4) The mRNA-protecting component is similar to placental ribonuclease inhibitor in that it is effective against RNase A and it is inactivated by the sulfhydryl reagent NEM. (5) The HeLa cell ribonuclease inhibitor is not identical to placental ribonuclease inhibitor, since the placental protein could not substitute for HeLa cell extract in our assay.

(6) The HeLa cell ribonuclease inhibitor appears to be active against RNases capable of degrading capped transcripts.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge Drs. Aaron Shatkin, Nahum Sonenberg, and Peter Blackburn for helpful discussions, Drs. Nancy Berliner and Susan Baserga for critical reading of the manuscript, Sharlene Ivory for excellent technical assistance, and Nicole E. Sykes for preparation of the manuscript. This work was supported by grant HL 24385 from the National Institutes of Health and the Cooley’s Anemia Foundation. C.A.S. is supported by an NRSA Fellowship from the National Institutes of Health and E.J.B. is the recipient of a Research

74

Career Development Institutes of Health.

Award from the National

REFERENCES Albrecht, G., Krowczynska, A. and Brawerman, G.: Configuration of&$obin messenger RNA in rabbit reticulocytes: identification of sites exposed to endogenous and exogenous nucleases. J. Mol. Biol. 178 (1984) 881-896. Baralle, F.E.: The functional significance of leader and trailer sequences in eukaryotic mRNAs. Int. Rev. Cytol. 81 (1983) 71-106. Bergen, I.E. and Brawerman, G.: Control ofbre~do~ ofthe polyadenylate sequence in mammalian polyribosomes: role of poly(adenylic acid) - protein interactions. Biochemistry 16 (1977) 259-264. Blackburn, P. and Moore, S.: Pancreatic ribonuclease. In Boyer, P.D. (Ed.), The Enzymes, Vol. XV. Academic Press, New York, 1982, pp. 317-433. Blackbum, P., Wilson, G. and Moore, S.: Ribonuctease inhibitor from human placenta: purification and properties. J. Biol. Chem. 252 (1977) 5904-5910. Brown, D.D.: Gene expression in eukaryotes. Science 2llfl981) 667-674. Craine, B.L. and Komberg, T.: Activation of the major Drasophifia heat shock genes in vitro. Cell 25 (1981) 671-681. Dani, C., Blanchard, J.M.,Piechaczyk, M., Sabouty, SE., Marty, L. and Jeanteur, P.: Extreme instability of myc mRNA in normai and transformed human cells. Proc. Natl. Acad. Sci. USA 81 (1984) 7046-7050. Dame11 Jr., J.E.: Variety in the level of gene control in eukaryotic cells. Nature 297 (1982) 365-371. Dignam, J.D., Lebovitz, R.M. and Roeder, R.G.: Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated m~rna~an nuclei. Nucl. Acids Res. 11 (1983) 1475-1489. Donovan, W. and Kushner, S.R.: Polynucleotide phosphorylase and ribonuclease II are required for cell viability and mRNA turnover in Escherkhia coli K-12. Proc. Natl. Acad. Sci. USA 83 (1986) 120-124. Filipowicz, W., Fun&hi, Y., Sierra, J.M., Muthukrishn~, S., Shatkin, A.J. and Ochoa, S.: A protein binding the methylated 5’-terminal sequence, m’GpppN, of eukaryotic messenger RNA. Proc. Natl. Acad. Sci. USA 73 (1976) 1559-1563. Furuichi,Y., LaFiandra,A. and Shatkin, A.J.: 5’-Terminal structure and mkNA stability. Nature 266 (1977) 235-239. Gedamu, L. and Dixon, G.H.: Effect of enzymatic ‘decapping’ on protamine mRNA translation in wheat germ S-30. B&hem. Biophys. Res. Commun. 85 (1978) 114-124. and Bimstiel, M.L.: The conserved Georgiev, 0. CAAGAAAGA spacer sequence is an essential element for the formation of 3’ termini of the sea urchin H3 histone mRNA by RNA processing. EMBO J. 4 (1985) 481-489.

Green, M.R., Maniatis, T. and Melton, D.A.: Human b-globiu pre-mRNA synthesized in vitro is accurately spliced in Xenopus oocyte nuclei. Cell 32 (1983) 681-694. Guyette, W.A., Matusik, R.J. and Rosen, J.M.: Prolactinmediated transcriptional and post-transcriptional control of casein gene expression. Ceil 17 (1979) 1013-1023. Krainer, A.R., Maniatis, T., Ruslcin, B. and Green, M.R.: Normal and mutant human /I-globin pre-mRNAs are faithfully and efficiently spliced in vitro. Cell 36 (1984) 993-1005. Luscher, B., Stauber, C., Schindler, R. and Schumperli, D.: Faithful cell-cycle regulation of a recombinant mouse histone H4 gene is controlled by sequences in the 3’-terminal part of the gene. Proc. Natl. Acad. Sci. USA 82 (1985) 4389-4393. Manley, J.L., Fire, A., Cane, A., Sharp, P.A. and Getter, M.L.: DNA-de~ndent tr~sc~ption of adenovirus genes in a soluble whole-cell extract. Proc. Natl. Acad. Sci. USA 77 (1980) 3855-3859. Morris, I., Malashi, F., Weber, L., Hickey, E., Greenspan, D., Bonner, J., Stein, J. and Stein, G.: Involvement of the 5’-leader sequence in coupling the stability of a human H3 histone mRNA with DNA replication. Proc. Natl. Acad. Sci. USA 83 (1986) 981-985. Old, R.W. and Woodland, H.R.: Histone genes: not so simple after all. Cell 38 (1984) 624-626. Patzeit, E., Blaas, D. and Kuechler, E.: CAP binding proteins associated with the nucleus. Nucl. Acids Res. 11 (1983) 5821-5835. Ross, J. and Kobs, G.: H4 histone messenger RNA decay in cell free extracts initiates at or near the 3’ terminus and procedes 3’ to 5’. J. Mol. Biol. 188 (1986) 579-593. Shatkin, A.J.: Capping of eucaryotic mRNAs. Cell 9 (1976) 645-653. Shaw, G. and Kamen, R.: A conserved AU sequence from the 3 ’ untranslated region of GM CSF mRNA mediates selective mRNA degradation. Cell 46 (1986) 659-667. Shimotohno, K., Kodama, Y., Hashimoto, J. and Miura, K.: Importance of 5’-terminal blocking structure to stabilize mRNA in eukaryotic protein synthesis, Proc. Natl. Acad. Sci. USA 74 (1977) 114-124. Stolle, CA., Payne, M.S. and Benz Jr., E.J.: Equal stabilities of normal p globin and non-translatable PO-39 thalassemia transcripts in cell-free extracts. Blood ‘70(1987) 293-300. Thomas, P.: Hybrid~ation of denatured RNA and small DNA fragments transferred to nitrocellulose. Proc. Natl. Acad. Sci. USA 77 (1980) 5201-5205. Volloch, V. and Housman, D.: Stability of globin mRNA in terminally differentiating murine erythroleukemia cells. Cell 23 (1981) 509-514. Wiskocil, R., Bensky, P., Dower, W., Goldberger, R.F., Gordon, J.I. and Deeley, R.G.: Coordinate regulation of two estrogendependent genes in avian liver. Proc. Natl. Acad. Sci. USA 77 (1980) 4474-447s. Communicated by S.R. Kushner.