Eur. J. Biochem. 90, 123- 130 (1978)

Partial Purification and Characterization of Rat-Liver Messenger RNA Coding for Phosphoendpyruvate Carboxykinase (GTP) Patrick B. IYNEDJIAN and Richard W. HANSON Fels Research Institute and Department of Biochemistry, Temple University School of Medicine, Philadelphia (Received March 20, 1978)

The mRNA coding for the gluconeogenic enzyme phosphoenolpyruvate carboxykinase (GTP) (EC 4.1.1.32) was partially purified from the liver of cyclic-AMP-treated rats by a procedure involving multiple oligo(dT)-cellulose chromatographies and sucrose gradient fractionations. The purification was monitored by translational assay using a wheat germ extract. Relative to RNA bound once to oligo(dT)-cellulose, the final material was enriched 20-fold in template activity for phosphoenolpyruvate carboxykinase synthesis. With this RNA preparation, cell-free enzyme synthesis amounted to 5 % of total mRNA-directed protein synthesis. The apparent sedimentation coefficient of phosphoenolpyruvate carboxykinase mRNA in sucrose gradients was between 20 and 22 S, corresponding to an average molecular weight of 0.93 x lo6. By formamide/polyacrylamide gel electrophoresis the molecular weight of the enzyme mRNA was estimated at between 0.91 x lo6 and 1.12 x lo6. From these estimates, it was concluded that considerable non-coding sequence(s) are present in the inRNA. Approximately 20 ”/, of the enzyme mRNA in rat liver failed to bind to oligo(dT)-cellulose, presumably because of the absence of a poly(A) segment. The translation of phosphoenolpyruvate carboxykinase mRNA by the wheat germ extract was inhibited in the presence of 7-methylguanosine 5‘-phosphate. The enzyme mRNA appears therefore to have a ‘cap’ at the 5’ end.

Recent work in our IabQratory has been concerned with the molecular mechanisms by which hormones and related factors control the rate of synthesis of the cytosolic form of the gluconeogenic enzyme Penolpyruvate carboxykinase. The mRNA coding for the enzyme directs the synthesis of a protein immunologically identical to P-enolpyruvate carboxykinase in heterologous cell-free translation systems [I, 21. Under appropriate conditions, such systems can serve for the assay of P-erzolpyruvate carboxykinase mRNA contained in crude RNA preparations extracted from rat tissues. Thus, we reported previously that the level of functional mRNA coding for P-enolpyruvate carboxykinase is increased in the liver during induction of the enzyme by cyclic AMP [2] and in the kidney after glucocorticoid administration or in metabolic acidosis [3]. Although suggestive, these observations do not prove changes in the abundance of P-enolpyruvate carboxykinase mRNA sequences, because increases in template activity for enzyme synthesis Enzyme. P-enolpyruvate carboxykinase or phospIio~,no/pyi-uvate carhoxykinase (GTP) (EC 4.1.1.32).

detected by translational assay might conceivably result from the activation of pre-existing nonfunctional mRNA. Changes in the ‘actual amount of Penolpyruvate carboxykinase mRNA according to the hormonal status of the animal could be measured by techniques involving molecular hybridization to synthetic DNA complementary to the enzyme messenger [4]. This, however, requires the prior purification of the messenger. Two types of approach have been used for the purification of specific eukaryotic mRNAs. One uses techniques such as density gradient centrifugation or gel electrophoresis where RNA species are fractionated according to their size. The other approach relies on the immunoprecipitation of polyribosomes involved in the synthesis of the protein of interest, followed by extraction of the isolated mRNA. Both methods have been successfully applied to the purification of mRNAs present in several thousand copies per cell in highly specialized tissues essentially committed to the synthesis of one protein [5-91. The isolation of less abundant mRNAs such as that coding for P-enolpyruvate carboxykinase, which are repre-

124

sented at most by only a few hundred copies in the liver cell even under induced conditions [lo] is obviously more problematic. In the present work, we have assessed the extent of purification of the enzyme mRNA which can be achieved by conventional sizing procedures. We furthermore report data on the poly(A) content and on the structure of the 5' end of rat liver P-enolpyruvate carboxykinase mRNA.

MATERIALS AND METHODS Materiuls Most reagents were bought from sources identified in previous papers [2,3]. Formamide was from Matheson, Coleman and Bell (Norwood, Ohio). Acrylamide and N,N '-methylene-bisacrylamide for RNA electrophoresis were supplied by Bio Rad Laboratories (Richmond, Cal.) and 'Stains All' was bought from Eastman Kodak (Rochester, N.Y.). Ribonuclease-free sucrose was purchased from Schwartz Mann (Orangeburg, N.Y .), and diethylpyrocarbonate from Sigma Chemical (St Louis, Mo.). Guanosine 5'-phosphate, 7-methylguanosine and 7methylguanosine 5'-phosphate were obtained from P. L. Biochemicals (Milwaukee, Wis.). General Mills provided the wheat germ for cell-free protein synthesis gratis. All solutions used for RNA extraction and purification were autoclaved except the sucrose solutions which were treated with diethylpyrocarbonate according to Stephenson et al. [l 11. Non-autoclavable equipment entering in contact with RNA was thoroughly rinsed prior to use with 0.4% (v/v) diethylpyrocarbonate. R N A Extraction

In order to raise the level of hepatic P-enoipyruvate carboxykinase mRNA, fasted, glucose-refed rats were injected with dibutyryl cyclic AMP and theophylline 75 min before killing [2]. Total tissue RNA was then extracted from the frozen livers by a conventional phenol/chloroform procedure. Subsequent steps involved precipitation of the large-molecular-weight RNAs by lithium chloride and selection of poly(A)containing RNA by oligo(dT)-cellulose chromatography, as described in detail previously [2]. The RNA eluted from the column in the low ionic strength buffer was ethanol-precipitated and dissolved in water. The solution was then heated at 70 "C for 5 min, quickly cooled to room temperature, made 10 mM in Tris-HC1, pH 7.5 and 0.5 M in NaCl and passed through a second oligo(dT)-cellulose column at room temperature. Poly(A)-containing RNA was eluted in 10 mM Tris-HC1 pH 7.5, precipitated by ethanol, dissolved in water and stored at - 80 "C.

Isolation of Phosphoenolpyruvate Carboxykinase mRNA

Sucrose Gradient Centrifugation Before being layered on the gradients, the RNA samples in water were brought to 80 (v/v) dimethylsulfoxide and heated at 65 "C for 3 min. The samples were then quickly cooled in an ice slurry, diluted by adding 3 vol. of cold gradient buffer (10 mM TrisHCI pH 7.2, 50 mM NaCl and 1 mM EDTA) and applied on 8 - 25 (w/w) sucrose gradients made up in gradient buffer. Centrifugation was performed at 2 "C at the centrifugal forces and for the periods given in the figure legends. After the run, the gradients were analyzed for light absorbance at 254 nm and fractionated using an Isco model 640 density gradient fractionator and an Isco model UA-4 absorbance monitor. RNA fractions were either assayed directly for template activity or were precipitated with ethanol for later use. Reference gradients containing rat liver RNA devoid of poly(A), isolated by two successive passages through oligo(dT)-cellulose, were included in all runs in order to determine the migration of 18-S and 28-S rRNAs. Polyacrylamide Gel Electrophoresis in Formamide The method of Duesberg and Vogt [12] was followed to prepare 3.5% (w/v) polyacrylamide gels in 97 (v/v) formamide. The ethanol-precipitated RNA samples were dissolved in a formamide/glycerol mixture (2/1), the formamide being buffered with 1 mM monosodium and disodium phosphates. Traces of bromphenol blue were added to serve as tracking dye. Electrophoresis was run for 6 h, at 80 V for the first 2.5 h and then at 130 V. The gels were stained with 'Stains All' according to Dahlberg et al. [13]. Rat liver 18-S and 28-S rRNAs purified by sucrose gradient centrifugation were used as markers, taking respectively 0.68 x lo6 and 1.74 x 106 as molecular weight values [14]. Translation of mRNA A non-preincubated wheat germ extract was used as source of ribosomes and protein-synthesis factors. Translation of rat liver mRNA was performed at 24 "C for 90 min in the presence of [3H]leucine using the conditions defined earlier [2,3]. Total protein synthesis was estimated by measuring hot trichloroacetic acid insoluble radioactivity. In order to obtain a high-speed supernatant containing polypeptide chains released by the ribosomes, the translation mixture was centrifuged in the Beckman 50 Ti rotor at 165000 x g for 1 h at 2 "C. The supernatant was used for the quantification of P-enolpyruvate carboxykinase synthesis by immunoprecipitation. To the supernatant were added non-radioactive P-enolpyruvate carboxykinase purified from rat liver cytosol as carrier, an

P. B. Iynedjian and R. W. Hanson

125

excess of specific antibody against the enzyme and Triton X-405 to a final concentration of 0.6%. After an incubation of 30 min at 37 "C and 15 h at 4 ' C , the irnmunoprecipitate was collected by centrifugation through a discontinuous two-layer density gradient made of 0.3 ml of 1 M and 0.15 ml of 0.5 M sucrose solutions in 10 mM sodium phosphate buffer, pH 7, containing 10 mM L-leucine and 0.6 Triton X-405. The precipitate was further washed three times in the above solution and then processed for analysis by sodium dodecylsulfate/polyacrylamide gel electrophoresis. The antigen-antibody complexes were dissociated by incubation for 15 min at 100 "C in 7.5 mM sodium phosphate, pH 7, 2 % (w/v) sodium dodecylsulfate, 5 % (v/v) 2-mercaptoethanol and 20 % (w/v) glycerol. Electrophoresis was on cylindrical 11 %, polyacrylamide gels containing 0.1 % sodium dodecylsulfate. After electrophoresis the gels were fractionated in I-mm slices with a gel extrusion device and the radioactivity in each fraction measured by liquid scintillation as described [3]. A single prominent radioactive polypeptide co-migrating with the carrier P-enolpyruvate carboxykinase was detected in the gels, as shown in our previous publications [2,3,15]. The immunological identity of the cell-free translation product with P-enolpyruvate carboxykinase has been demonstrated [2]. RESULTS Partial Purijication Rat Liver P-enolpyruvate Carboxykinase rn R N A

of

RNA prepared by oligo(dT)-cellulose chromatography of total tissue RNA from the liver of cyclicAMP-treated rats has considerable messenger activity for the synthesis of P-enolpyruvate carboxykinase in

a cell-free translation system derived from wheat germ [2]. Such RNA served therefore as starting material for our purification procedure. In response to this RNA, the synthesis of P-enolpyruvate carboxykinase by the wheat germ system represented between 0.2 % and 0.3 % of the total mRNA-directed amino acid incorporation (Table I), in excellent agreement with previous results [2]. When the RNA was passed a second time through oligo(dT)-cellulose, only 40 % of the applied material was retained. The rest was presumably mostly rRNA, which is known to heavily contaminate RNA preparations bound once to oligo(dT)-cellulose [I 61. As expected, the elimination of this extraneous RNA resulted in a roughly 2-fold increase in the specific messenger activity for overall protein synthesis (Table 1). The specific template activity for P-enolpyruvate carboxykinase synthesis, however, was not enhanced to the same extent. This might reflect the loss of enzyme message either in the flow-through fraction of the column chromatography or by inactivation. Alternatively, the decrease in enzyme mRNA activity might be due to the loss of an RNA species stimulating the translation of the template by the wheat germ ribosomes. Incomplete recoveries of specific mRNAs during oligo(dT)-cellulose chromatography have been reported by others P61. The poly(A)-containing RNA obtained after two rounds of chromatography was centrifuged through sucrose gradients after denaturation in the presence of 80 % dimethylsulfoxide. As may be seen in Fig. 1 (A), the RNA spread over most of the gradient, with two distinct peaks, presumably 18-S and 28-S rRNAs, superimposed on the heterodisperse background typical of mammalian mRNA. The RNA with apparent sedimentation coefficient of 20- 26.5 S was precipitated with ethanol, and a small sample was assayed

Table 1. Partial purification of P-enolpyruvate carboxykinase mRNA Total liver RNA was extracted from the liver of cyclic-AMP-treated rats and further processed as described in Materials and Methods and in the text. Overall mRNA activity was determined by trichloroacetic acid precipitation of the high-speed supernatant obtained from the wheat germ translational assay. [3H]Leucine incorporation was corrected for incorporation occurring in the absence of added mRNA. P-enolpyruvate carboxykinase mRNA activity was measured by immunoprecipitation of the enzyme contained in the high-speed supernatant obtained after translation by the wheat germ extract. Recovery was calculated by multiplying P-enolpyruvate carboxykinase mRNA activity by the yield of RNA Step

RNA yield

Overall mRNA activity (A)

P-enolpyruvate carboxykinase mRNA ___._

activity (B)

Pg

dis. min-' (ng RNA)-'

purification

-fold

purity (B/A x 100)

.. ____ recovery

% ~-

Oligo(dT) 1 Oligo(dT) I1 Gradient I Gradient 111

3850 1750 300 11

Poly(Aj-free RNA (18-24 S)

-

246 447 431 281

U.68 I 0.952 4.595 12.265

25

0.027

I .U 1.4 6.7 18.0

0.28 0.21 1.07 4.36

100 63 52 5

-

0.11

-

126

Isolation of Phosphoenolpyruvate Carboxykinase m R N A

1

'04

8

12 16 m Eluate (ml)

1

28

24 t

I

I

Eluate (ml)

Fig. 1 . Sucrose gradient ,fiactiona/ion o / P-enolpyvuvcrfe carboxykinase mRNA. (A) R N A chromatographed twice on oligo(dT)-cellulose was centrifuged after denaturation by dimethylsulfoxide through & 2 5 % sucrose gradients for 21 h at 2 "C and 22000 rev./min in a Beckman SW27 rotor. Sedimentation in the gradient was from left (top of tube) to right (bottom of tube). The bracket indicates the zone of the gradient collected for further purification. (B) Third density gradient fvactionation. RNA was centrifuged through an 8 - 25 "/,sucrose gradient Fractions a s indicated for 21 h at 2 "C and 26000 rev./min in the BeckmanSW40 rotor. The light absorbance at 254 nm was recorded (-). were assayed in the wheat germ system for overall template activity (open bars) and P-enolpyruvate carboxykinase mRNA activity (black bars). Overall template activity is incorporation of [3H]leucine into trichloroacetic-acid-insolublematerial of total reaction mixture (before highspeed centrifugation). Enzyme mRNA activity is incorporation of [3H]leucineinto immunoprecipitable material contained in the high-speed supernatant of the reaction mixture. Arrows indicate the position of 18-S and 28-S rRNA

for template activity. At this stage, P-enolpyruvate carboxykinase messenger activity was enhanced about 7-fold above the initial level (Table 1). The bulk of the 20-26.5s RNA was then further processed through two additional cycles of sucrose gradient centrifugation, each time after treatment with dimethylsulfoxide. Fig. 1 (B) depicts the distribution of the RNA in the last gradient. Eleven RNA fractions ranging in sedimentation coefficient from 16.5-S to 27-S were collected and assayed for their ability to direct total protein and P-enolpyruvate carboxykinase synthesis. Overall messenger activity per weight of RN A declined in RNA fractions of increasing size (Fig. 1B, open bars), an observation consistent with the limited efficiency of the wheat germ system in the translation of long messengers. As we were primarily interested in the relative enrichment or the various fractions in P-enolpyruvate carboxykinase message, the specific activity of the RNA (leucine incorporation into P-enolpyruvate carboxykinase per weight of RNA) rather than the total messenger activity per fraction was plotted in Fig. 1. The apparent skewness of the distribution of P-enolpyruvate carboxykinase mRNA would be largely suppressed in a plot showing the amount of enzyme mRNA in the fractions. With

the two fractions most enriched in enzyme messenger, the incorporation of radioactive leucine into P-enolpyruvate carboxykinase amounted to 12000 dis. min (pg RNA)-', a value 18-fold higher than with the original material. Enzyme synthesis, as percentage of overall protein synthesis directed by exogenous mRNA, was more than 4'x as opposed to 0.28 % before purification, again indicating a more than 15-fold purification of the specific enzyme mRNA. Although the successive density gradient centrifugations resulted in considerable purification of Penolpyruvate carboxykinase mRNA, contamination by other niRNAs was still present, as evidenced by the fact that the relative rate of cell-free P-enolpyruvate carboxykinase synthesis at the direction of the purest fraction did not exceed 5 % of total protein synthesis. Sodium dodecylsulfate/polyacrylamide gel electrophoresis of RNAse-treated total cellfree products revealed a heterogeneous mixture of peptides, the majority of which migrated faster than P-enolpyruvate carboxykinase, as shown in Fig. 2 (A); the position of carrier P-enolpyruvate carboxykinase is indicated by the arrow. Part of the small-molecularweight peptides originated from the translation of messengers endogenous to the wheat germ extract,

P. B. Iynedjian and R. W. Hanson

since they were synthesized also in the absence of added inRNA (data not shown). For comparison, Fig. 2 B illustrates the electrophoretic analysis of an immunoprecipitate obtained from the translation mixture programmed with the partially purified Penolpyruvate carboxykinase mRNA. Clearly, the immunoprecipitation procedure results in the isolation of a single major polypeptide, the mobility of which corresponds to that of authentic P-enolpyruvate carboxykinase. As discussed in an earlier paper [2], the presence of a small amount of radioactive peptides shorter than the enzyme in the immunoprecipitate might be due to premature release of unfinished Penolpyruvate carboxykinase molecules. Alternatively, it could reflect a low level of contamination of the precipitate by cell-free products unrelated to the enzyme. In order to evaluate more directly the degree of purity of the RNA and to assess the potential for further purification, the RNA most enriched in Penolpyruvate carboxykinase template was displayed on polyacrylamide gels in formamide. The partially purified mRNA migrated as a band centered between 18-S and 28-S rRNAs (Fig.3). The majority of the RNA had a molecular weight ranging from 0.87 to 1.4 x lo6. Relative to adjacent gradient fractions, the RNA with peak template activity for P-enolpyruvate carboxykinase synthesis was found to be enriched in RNA species banding between M , values of 0.91 x lo6 and 1 . 1 2 lo6. ~ This range can therefore serve as an estimate of the molecular weight of P-enolpyruvate carboxykinase mRNA as deduced from polyacrylamide gel electrophoresis under partially denaturing conditions. Because of the limited amount of gradientpurified RNA, we did not attempt preparative gel electrophoresis for the further purification of the mRNA.

127

Migration (mrn) Fig. 2. Sodium dodc~.vlsulfate/polya~ry~ami~e grl elecirophoresis of ,cheat germ trans1ationproduct.s synthesized in the presence ofpartially purified P-enolpyruvute carboxykinase mRNA. (A) Total protein synthesis products obtained from the high-speed supernatant of the translation mixture. The sample was incubated for 10 min at 2 4 ° C with 0.0040/; (w/v) ribonuclease A and 1 0 m M EDTA before addition of trichloroacetic acid. The precipitate was washed with ethanol and ether before dissolution in the sodium dodecylstilhtejmercaptocthanol buffer described under Materials and Methods. The arrow indicates the migration of authentic P-enolpyruvate carboxykinase. (B) Immunoprecipitated cell-free product from the same translation mixture as in A. Details on imtnunoprecipitation and sample preparation for electrophoresis are given in Materials and Methods

Poly(A) Content of Rat Liver P-enolpyuuvute Cauboxykinase mRNA The retention of P-enolpyruvate carboxykinase mRNA by oligo(dT)-cellulose strongly suggests the presence of a poly(A) segment in at least part of the messenger contained in the liver cell. It was of interest, however, to determine whether there are substantial amounts of enzyme message lacking poly(A). Rat liver RNA eluting in high-ionic-strength buffer from two successive oligo(dT)-cellulose columns was further fractionated on sucrose gradients to select material sedimenting between 18 and 24-S. After a third passage through oligo(dT)-cellulose in a buffer containing 0.5 M sodium chloride, the RNA was incubated in the wheat germ translation system. As shown in Table 1, a weak stimulation of total amino acid incorporation above the endogenous level was observed. Moreover, low levels of P-molpyruvate carboxykinase could be

Fig. 3. Foirnamid~~~polyacrylami~e gc4 electrophoresis of purriafly purifkd P-enolpyruvaie carho.xykinase mRNA. Experimental conditions were as described in Materials and Methods. The gels were 0.5 cm by 7 cm. Gel 1, 2.0 pg 28-S rRNA from rat liver; gel 2, 2.3 pg 18-S r R N A ; gel 3, 1 pg 28-S 1 pg IS-S rRNAs; gel 4, 3.7 pg of RNA from the 4th assayed fraction in Fig. 1

+

128

Isolation of Phosphoenolpyruvate Carboxykinase mRNA

B

P

I

12

-.-

. K

E

9

G 0

;

0,

c

6 0 ,

.-C I 3 ?

0

"

0.2

0.4

0.8

Nucleotide concn

0.2

0.4

0.8

0

(rnM)

Fig. 4. Effeci oj 7-mriiiylpunosine 5'-phosphate on trunslation of P-enolpyruvute carhoxykinase mRNA. Total tissue RNA from cyclicAMP-treated rats was chromatographed once through oligo(dT)cellulose and incubated in the wheat germ translation assay either without addition or in the presence of various concentrations of 7-methylguanosine 5'-phosphate ( L O ) or of guanosine 5'phosphate (0).(A) Total leucine incorporation into hot-trichloroacetic-acid-precipitable material. (B) Leucine incorporation into material precipitated from the high-speed supernatant of the translation mixture by antibody against P-enolpyruvate carboxykinase. Two assays were performed at each concentration; the range between the two determinations is shown by the brackets

detected among the cell-free products. From the respective yields of RNA retained on oligo(dT)-cellulose and of 18- 24-S poly(A)-lacking RNA, we calculated that no more than 20% of rat liver P-enolpyruvate carboxykinase mRNA was devoid of poly(A), or contained a poly(A) sequence too short to hybridize to oligo(dT)-cellulose. 5 ' - Terminal Structure of Rat Liver P-enolpyruvate Carboxykinase mRNA

The majority of poly(A)-containing RNA molecules in mammalian cells have a characteristic oligonucleotide sequence at the 5' end containing 7-methylguanosine linked by a 5'-5' triphosphate bond to a second methylated nucleoside [17,18]. This structure, called a 'cap', appears to play an essential role in the translation of the messengers by the protein synthetic machinery [19]. For instance, it has been shown that 7-methylguanosine 5'-phosphate [20] or synthetic polynucleotides containing a cap structure [21] compete with capped mRNAs for ribosome binding and thereby inhibit the translation of those messengers, while they do not affect the translation of uncapped mRNAs. In the present work we used 7-methylguanosine 5'-phosphate as a probe to determine whether P-enolpyruvate carboxykinase mRNA contained a 5'-terminal cap. Fig. 4 (B) demonstrates that the cellfree synthesis of P-enolpyruvate carboxykinase by

the wheat germ extract at the direction of rat liver poly(A)-containing RNA was drastically reduced in the presence of the methylated nucleotide, with 50 inhibition occurring between 0.05 and 0.1 mM and virtually total suppression at 0.4 mM. Overall protein synthesis was similarly inhibited (Fig. 4A), in agreement with the observation of others [22]. In the experiment depicted in Fig. 4, a slight depression of protein synthesis was observed with 0.8 mM of guanosine 5'-phosphate used as a control; neither guanosine 5'-phosphate nor 7-methylguanosine had any inhibitory effect in a subsequent experiment (not shown). From these data, we tentatively conclude that the great majority of rat liver mRNAs, including P-enolpyruvate carboxykinase mRNA, have a cap structure at the 5' end. DISCUSSION The structural features of rat liver P-enofpyruvate carboxykinase mRNA emerging from this study are typical of most eukaryotic mRNAs. One such feature is the presence of a poly(A) sequence at the 3' end of the molecule [23]. Our data indicate that approximately 80 % of the mRNA coding for P-enolpyruvate carboxykinase in the liver cell contains a poly(A) segment, presumably at the 3' end, while the rest either is devoid of it or has only very short oligo(A) tracts unable to hybridize to oligo(dT). Other wellcharacterized mRNAs are partitioned in the same way between poly(A)-containing and poly(A)-lacking RNA fractions [24,25]. The presence of a cap at the 5' end of several mRNAs has been inferred from the inhibition by 7methylguanosine 5'-phosphate of the translation of these messengers in cell-free systems [26,27]. Using the same argument, we propose that rat liver P-enolpyruvate carboxykinase mRNA is capped. The validity of this argument is of course predicated on the assumption that 7-methylguanosine 5'-phosphate inhibits messenger translation specifically because of its structural analogy to the 5' cap [28]. Evidence that this is so comes from the demonstration that uncapped mRNAs such as satellite tobacco necrosis virus RNA [20,29] or encephalomyocarditis virus RNA [26] are translated well in the presence of the nucleotide. Data presented here allow a relatively accurate estimate of the size of rat liver P-enolpyruvate carboxykinase mRNA. After denaturation in dimethylsulfoxide the messenger sedimented on sucrose gradients standardized with 18-S and 28-S rRNAs with an apparent sedimentation coefficient of 20 - 22 S, corresponding to a molecular weight of 0.84- 1.02 x lo6 [30]. By formamide/polyacrylamide gel electrophoresis, the molecular weight range for P-enolpyruvate carboxykinase mRNA was estimated to be 0.91 - 1.12 x lo6. The slight discrepancy between the

P. B. Iynedjian and R. W. Hanson

two values was not unexpected, several previous studies having shown that when rRNAs are used as standards, mRNA sizes estimated by gel electrophoresis are consistently larger than values obtained by density gradient centrifugation [8,9,25,31]. The reason for such discrepancies appears to reside in the strong tendency of rRNA to assume a secondary structure even under denaturing conditions. This residual secondary structure would result in artefactually high mobility during gel electrophoresis and artefactually rapid sedimentation during gradient centrifugation, in turn resulting in the overestimation and underestimation respectively of the size of fully denatured species [32]. We therefore propose as our best estimate for the molecular weight of P-enolpyruvate carboxykinase a value of 0.99 x lo6, intermediary between the average density gradient and gel values. Accordingly, the mRNA should contain 2900 nucleotides. Since a protein of approximately 80 000 molecular weight like P-enolpyruvate carboxykinase [1,33] is totally encoded in a sequence of 1750 nucleotides, we calculate that the messenger contains 1150 untranslated nucleotides. Assuming that the poly(A) sequence has an average length of 150 nucleotides, we conclude that the mRNA molecule contains 1000 additional noncoding nucleotides. Even the smallest M , estimate derived from sucrose gradient fractionation, 0.84 x lo6, would accomodate 600 non-coding nucleotides in addition to poly(A). Thus, by the presence of considerable non-coding sequence(s), as well as both a 5’ cap and a 3’ poly(A) tail, rat liver P-enolpyruvate carboxykinase mRNA conforms to the typical model for eukaryotic mRNAs. The isolation procedure for P-enolpyruvate carboxykinase mRNA used here is a conventional scheme including multiple affinity chromatography and density gradient centrifugation steps. Earlier studies had shown a 20-fold enrichment in enzyme mRNA by a single passage of polysomal RNA through oligo(dT)cellulose 1341. In the present work, a better than 15fold purification with respect to other mRNAs was achieved. The overall purification of P-enolpyruvate carboxykinase mRNA from total cellular RNA was therefore in excess of 300-fold. At this stage, P-enolpyruvate carboxykinase synthesis in the wheat germ system amounted to more than 4 % of total mRNAdirected protein synthesis. This ratio is a minimal estimate of the purity of the messenger for two reasons. Firstly, enzyme synthesis in the translational assay may have been underestimated, because only fulllength immunoprecipitable cell-free products were scored as enzyme, while shorter peptides resulting from premature termination of translation were not taken into account. Secondly, overall mRNA-directed protein synthesis (calculated by subtracting endogenous amino acid incorporation in the absence of added mRNA from the total incorporation observed in the

129

presence of liver mRNA) may have been overestimated because the addition of exogenous mRNA to the nonpreincubated wheat germ extract might possibly stimulate the translation of endogenous messengers [35]. In any case, the data from the template activity assay indicated that the partially purified mRNA was still very heterogeneous. This conclusion was confirmed by analytical formamide/polyacrylamide gel electrophoresis, in which the RNA was resolved over a relatively broad zone. Indeed, our results suggest that gel electrophoresis might usefully be employed at the preparative scale for the isolation of P-enolpyruvate carboxykinase mRNA of purity substantially higher than obtained after density-gradient fractionation. We thank Dr F. John Ballard for providing the antibody to P-enolpyruvate carboxykinase used in this study. This research was supported by grant AM-I8034 from the National Institutes of Health and by the Samuel S. Fels Fund.

REFERENCES 1 Tilghman, S. M., Ballard, F. J. & Hanson, R. W. (1976) in Gluconeogenesis, Its Regulation in Mammalian Species (Hanson, R. W., Mehlman, M. A,, eds) pp. 47-91, John Wiley and Sons, New York. 2 Iynedjian, P. B. & Hanson, R. W. (1977) J . B i d . Chem. 252, 655 - 662. 3 Iynedjian, P. B. & Hanson, R. W. (1977) J . Biol. Chem. 252, 8398 - 8403. 4 Parish, J. H. (1972) in Principles und Practice uf Experiments wifh Nucleir Acids, pp. 355 - 379, John Wiley and Sons, New York. 5 Honjo, T., Packman, S., Swan, D., Nau, M. & Leder, P. (1974) Proc. Nut1 Acad. Sci. U.S.A. 71, 3659-3663. 6 Kazazian, H. H., Jr, Snyder, P. G . & Cheng, T. (1974) Biochem. Biophys. Res. Commun. 59, 1053- 1061. 7. Morrison, M. R., Brinkley, S. A,, Gorski, J. & Lingrel, J. B. (1974) J . Bid. Chem. 249, 5290-5295. 8. Rosen, J. M., Woo, S. L. C., Holder, J. W., Means, A. R. & O’Malley, B. W. (1975) Biochemisrry, 14, 69-78. 9. Shapiro, D. J. & Schimke, R . T. (1975) J . B i d . Chem. 250, 1759 - 1764. 10. Galau, G . A,, Klein, W. H., Britten, R. J. & Davidson, E. H. (1977) Arch. Biochem. Biophys. 179, 584- 599. 11. Stephenson, M. L., Wirthlin, L. S., Scott, J. F. & Zamecnik, P. C. (1972) Proc. Nut1 Acad. Sci. U.S.A. 69, 1176-1180. 12. Duesberg, P. H. & Vogt, P. K. (1973) J . Virol. 12, 594-599. 13. Dahlberg, A. E., Dingman, C. W. & Peacock, A. C. (1969) J . Mol. Biol. 41, 139-147. 14. Wellauer, P. K., David, I. B., Kelley, D. E. & Ferry, R. P. (1 974) J . Mol. Biol. 89, 397 - 407. 15. Kioussis, D., Reshef, L., Cohen, H., Tilghman, S. M., lynedjian, P. B., Ballard, F. J. & Hanson, R. W. (1978) J . Biol. Chem. 253,4327-4332. 16. Gielen, J., Aviv, H. & Leder, P. (1974) Arch. Biochem. Biophys. 163, 146-154. 17. Wei, C. M., Gershowitz, A. &Moss, B. (1975) Ce114, 379-386. 18. Adams, J. M. & Cory, S. (1975) Nature (Lond.) 255, 28-33. 19. Busch, H., Hirsch, F., Gupta, K. K., Rao, M., Spohn, W. & Wu, B. C. (1976) Pvogr. Nucfeic Acid Res. Mol. Bzol. 19, 39-63.

130

P. B. Iynedjian and R. W. Hanson : Isolation of Phosphoendpyruvate Carhoxykinase mRNA

20. Hickey, E. D., Weber, L. A. & Baglioni, C. (1976) Proc. Nail Acad. Sci. U.S.A. 73, 19-23. 21. Both, G. W., Furuichi, Y., Muthukrishnan, S. & Shatkin, A. J. (1976) J . Mol. Biol. 104, 637-658. 22. Rao, M. S., Blackstone, M. & Busch, H. (1977) Biochemistry, 16,2756- 2762. 23. Brawerman, G. (1976) Prog. Nucleic Acid Res. Mol. Biol. 17, 117- 147. 24. Rhoads, R. E. (1975) J . Biol. Chem. 250, 8088-8097. 25. Taylor, J. M. & Tse, T. P. H. (1976) J . B i d . Chem. 251,7461 7467. 26. Weher, L. A,, Feman, E. R., Hickey, E. D., Williams, M. C. & Baglioni, C. (1976) J. B i d . Chem. 251, 5657-5662. 27. Sharma, 0. K., Hruby, D. E. & Beezley, D. N. (1976) Biochem. Biophys. Res. Commun. 72, 1392 - 1398.

28. Hickey, E. D., Weber, L. A. & Baglioni, C., Kim, C. H. & Sarma, H. (1977) J . Mol. Biol. 109, 173-183. 29. Roman, R., Brooker, J. D., Seal, S . N. & Marcus, A. (1976) Nature (Lond.) 260, 359-360. 30. McConkey, E. H . (1967) Methods Enzymol. 12, 620-634. 31. Shapiro, D. J. & Baker, H. J. (1977) J . B i d . Chem. 252, 52445250. 32. Woo, S. L. C., Rosen, J. M., Liarakos, C. D., Choi, Y. C., Busch, H., Means, A. R., O’Malley, B. W. & Robberson, D. L. (1975) J . Biol. Chem. 250, 7027-7039. 33. Ballard, F. J. & Hanson, R. W. (1969) J . Biol. Chem. 244, 5625 - 5630. 34. Tilghman, S. M., Fisher, L. M., Reshef, L., Ballard, F. J. & Hanson, R. W. (1976) Biochem. J . 156, 619-626. 35. Senger, D. R. & Gross, P. R. (1976) Devel. Biol. 53, 128-133.

P. B. lynedjian, Institut de Pharmacologie, Universitk de Lausanne, Rue du Bugnon 21, CH-1031 Lausanne, Switzerland R. W. Hanson, Department of Biochemistry, Case Western Reserve University School of Medicine, 2109 Adelbert Road, Cleveland, Ohio, U.S.A. 44106