THE JOURNAL OF ~ l O l . O C I C A I . CHEMISTRY

Prrnled in U. S. A Vol. 257, No. 11, Issue of June 10, pp. 6224-62:3(1, 1982

Purification and Characterization of the m’G(B’)pppN-grophosphatase from Human Placenta*

(Received for publication, December 21, 1981)

Donald L. Nus& Robert E. Altschuler, and Andrew J. Peterson From the Dioision of Laboratories and Research, New York State Department of Health, Albany, New York 12201

A purification scheme has been developed for the m7G(5’)pppN-pyrophosphatase from human placenta. The 1400-fold purified placental enzyme exhibited physical and enzymatic properties similar to those pre- viously reported for a crude preparation of the human m7G(5’)pppN-pyrophosphatase obtained from HeLa cells. Polyacrylamide gel analysis of enzyme fractions at different stages of purification revealed a M, = 40,000 polypeptide that increased in relative concentration as the specific activity of the enzyme fractions increased. Copurification of this polypeptide with m7G(5’)pppN- pyrophosphatase activity suggests the possibility that the 81,000-dalton native enzyme is a dimer composed of subunits of identical molecular weight. The highly purified placental enzyme, like the crude HeLa enzyme, failed to hydrolyze the cap moiety of intact mRNA even under conditions known to reduce mRNA secondary structure. Moreover, when a series of capped oligonu- cleotides that differed progressively in chain length by a factor of one nucleotide was tested as substrate, the rate of enzyme-catalyzed cap hydrolysis decreased as the chain length increased. The purified placental en- zyme failed to release m7pG from oligonucleotides con- taining the cap and 3 or more additional nucleotides. These results are discussed in terms of the probable biological function of the m‘G(5’)pppN-pyrophospha- tase.

An enzyme activity which specifically hydrolyzes the struc- ture derived from the 5’-terminal end of eukaryotic mRNA, the cap, was fist detected in HeLa cell extracts (1). A similar activity was subsequently reported in extracts of chick lens (2). The HeLa activity, partially purified and characterized with respect to physical properties and substrate specificity, was reported to catalyze the following reaction: m7G(5’)pppN + m7pG1 + ppN, where N = 2”O-methylated or unmethylated ribonucleosides or oligonucleotides less than 8 bases in length (3). G(5’)pppN and ring-opened m’G(5’)-pppN, which do not possess the positive charge at the N’ position of guanosine, were not hydrolyzed by enzyme preparations that readily hydrolyzed m7G(5‘)pppN. The 2-amino group of the 7-meth- ylguanosine moiety of m7G(5’)-pppN was also shown to be

* This work was supported by Grant PCM 78 12127 from the National Science Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ T o whom correspondence and reprint requests should be ad- dressed.

The abbreviations used are: ppG, guanosine 5’-diphosphate; m’pG, 7-methylguanosine 5’-monophosphate; G”’, 2”O-methylguano- sine; G(5‘)pppG, guanosine 5’-triphosphoryl-5”triphosphoryl-5’-gua- nosine; BAP, bacterial alkaline phosphatase; SDS, sodium dodecyl sulfate.

involved in substrate-enzyme interaction. Enzyme-catalyzed hydrolysis of m’G(5’)pppG proceeded in the absence of diva- lent cations at a pH optimum of 7.5 and a temperature optimum of 45 “C, with a Michaelis constant (K,) of 1.7 p ~ . The molecular weight of the native enzyme activity was determined to be 81,OOO.

As part of our continuing efforts to determine the biological function(s) of this unique enzyme, we have developed a scheme by which the enzyme can be purified more than 1400- fold and freed of contaminating ribonuclease activity. This report contains details of the purification procedure, physical and enzymatic properties of the purified enzyme, and results of studies aimed at understanding the inability of the enzyme to hydrolyze the cap moiety of intact mRNA.

EXPERIMENTAL PROCEDURES

Preparation of Substrate-[methyl-’H]miG(5’)pppG was prepared by methylation of G(5’)pppG with the reovirus-associated methyl- transferase using S-[methyZ-JH]adenosylmethionine (5-15 Ci/mmol) as the methyl donor (4). [methyl-’HIReovirus mRNA was prepared as previously described (I), while modificat.ions outlined by Kozak (5) were used to prepare IMP-substituted [methyZ-3H]reovirus mRNA. [methyl-”lH]Reovirus mRNA was converted to a double-stranded form by hybridization with an excess of purified reovirus genome RNA. Conditions used for genome RNA purification and mRNA- genome RNA hybridization were as recently described for wound tumor virus RNAs ( 6 ) . [a-32P]UMP-labeled wound tumor virus mRNA was prepared by in uitro transcription (6) directed by a variant population of wound tumor virus (-S2(70)) (7), which lacks the second largest genome segment. ”H-methyl-labeled vaccinia virus RNA was synthesized in vitro as described previously (8) in the presence of 20 pCi/ml of S-[methyZ-3H]adenosylmethionine at a con- centration of 1.6 p ~ . At this concentration of S-adenosylmethionine, methylation of vaccinia mRNA occurs primarily at the N-7 position of the terminal guanosine, i.e. the mRNAs contain primarily caps of the type m’G(5’)pppG and m7G(5’)pppA.2 ‘H-methyl-labeled, capped oligonucleotides of increasing chain length were generated by diges- tion with TI-ribonuclease. Oligonucleotides of specific chain lengths were isolated by chromatography on DEAE-cellulose in 7 M urea and subsequently desalted prior to use (1).

Assay for m7G(5,1PPPN-pyrophosphatase Activity-At each puri- fication step, m’G(5’)pppN-hydrolyzing activity was located in indi- vidual fractions by use of the rapid bacterial alkaline phosphatase- coupled assay (3) which contained [methyL3H]m’G(5’)pppG (2500 cpm/pmol) a t a concentration of 200 nM. Analysis of reaction products was performed by chromatography on polyethyleneimine-cellulose sheets developed with H 2 0 (3). For determination of the specific activity of pooled active fractions obtained from each step in the purification scheme, [methyZ-3H]m’G(5’)pppG hydrolysis was fol- lowed in the absence of bacterial alkaline phosphatase by paper electrophoretic analysis of the reaction products in pyridine acetate buffer (pH 3.5) (3). In this case, [methyl-’H]m‘G(V)pppG (25 cpm/ pmol) was present a t a concentration of 20 NM. One unit of activity was defined as that amount of enzyme hydrolyzing 1 pmol of m’G(5’)pppG to m’pG and ppC in 10 min at 37 “C. A combination of paper electrophoresis in pyridine acetate buffer, pH 3.5, and descend- ing paper chromatography in isobutyric acid, 0.5 N NHdOH (10:6 v/v)

E. Paoletti, personal communication, and our analysis.

6224

by guest on April 5, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Purification of the Human miG(5')pppN-pyrophosphatase 6225

was used to analyze the reaction products after attempts to hydrolyze the cap moiety of intact mKNA with the pyrophosphatase.

Protein Determination-Throughout the purification procedure, protein concentration was determined by using the Bio-Kad protein assay kit with bovine serum albumin as a standard.

Poiyacrylamide Gel I;Iectrophoresis-Enzyme fractions at differ- ent stages of purification were analyzed by electrophoresis on 10% SDS-polyacrylamide gels according to Anderson et al. (9) by the method of Laemmli (10). Polypeptides were detected by the silver- staining technique of Switzer et al. (1 1 ) as simplified by Oakley et al. (12). '"1'-labeled wound tumor virus RNA treated with m'G(5')pppN- pyrophosphatase fractions was analyzed on 2.54% polyacrylamide gels containing 5 M urea using a Tris-borate-SDS buffer system, pH 8.3 (6). Following electrophoresis, gels were subjected to autoradiog- raphy on Kodak BB54 medical x-ray film at -TO "C with the aid of an intensifying screen.

Source of MateriaIs-S-[methyl-:lH]Adenosylmethionine (5-15 Ci/ mmol) and uridine 6'-[a-.'"P]triphosphate (400-600 Ci/mmol) were purchased from Amersham Corp. IIEAE-cellulose (DE-52) was ob- tained from Whatman. Ltd. Hydroxylapatite (Bio-Gel HT), G(5')pppG. and the protein assay kit were purchased from Bio-Rad. Hexylagarose was purchased from Miles-Yeda, Ltd., while polyeth- yleneimine-cellulose thin layer sheets were obtained from d . T. Raker Co. Bacterial alkaline phosphatase was purchased from Worthington. RNasin was a product of Riotec, Inc.; marker proteins for polyacryl- amide gel electrophoresis were purchased from Pharmacia. Vaccinia virus was a gift from Dr. Enzo I'aoletti. Division of Laboratories and Research, New York State Department of Health, Albany, NY.

RESULTS:'

Summary of Purification Scheme-Details for each step in the purification of the m'G(5')pppN-pyrophosphatase are pre- sented in the Miniprint. A summary of the purification param- eters for each step in a typical preparation is presented in Table I. A 1400-fold purification was obtained through the sucrose gradient centrifugation step with a recovery of ap- proximately 2.5%. Yields exceeding 100% were sometimes obtained following ammonium sulfate fractionation and chro- matography on DEAE-cellulose. Since greater than 100% yields were not consistently obtained, it is difficult to conclude that an inhibitor of m'G(5')pppN-pyrophosphatase activity is being removed at these steps. The losses taken after chroma- tography on hydroxylapatite and hexylagarose, as well as the sucrose gradient step, in this particular enzyme preparation, reflect sacrifices in yield for the sake of purity; only the most active fractions were pooled, assayed for specific activity, and subjected to further purification. Polyacrylamide gel electro- phoresis under nondenaturing conditions was used to further purify the pyrophosphatase. As will be discussed, this step yields a highly purified enzyme fraction free of detectable ribonuclease activity. However, the enzyme at this state of purity is stable a t -80 "C for only several weeks. In addition, unless a preparative gel procedure is employed, this step is impractical for routine preparation in terms of the quantity of enzyme that can be produced in a reasonable time frame.

The fractions from different stages of purification were examined for polypeptide composition by SDS-polyacryl- amide gel electrophoresis (Fig. 1). A polypeptide migrating with apparent M , = 40,000 was the major polypeptide ob- served in the active fraction recovered from nondenaturing polyacrylamide gel. This polypeptide was fist distinguishable in the fraction recovered from hydroxylapatite chromatogra- phy and increased in relative staining intensity in parallel with

' I Portions of this paper (including part of Kesults and Figs. 1-5) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Kockville Pike, Bethesda, MD 20814. Kequest Document No. 81M-3114. cite the authors, and include a check or money order for $2.80 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.

1'AHI.E I Purification of m'G(j')pppN.p~roph~).sr)h~:tnse from human

placenta The starting material was loo0 g of placental tis..ue.

- ~- ~~ ~~

Fraction Spyifir Total ,,,, t9 ar,l\.lt\ 1~r~Over9 -~ - .

rnl ~~ ~

mp units/n,p "';::",x ';

30,OOO X g superna- 1.804 9,922 800 5.958 loo

20-40'3 ammonium 210 4.620 1.700 7.854 132 tant

sulfate fractiona- tion

DEAE-cellulose 210 540 20,ooO 10,080 169 Hydroxylapatite 15.6 85.5 3o.oo0 2.565 48 Hexylagarose 2.0 1.12 :383,0(KI 427 Sucrose gradient 1.2 0.05 857,000 48 2.5"

- I

peak fraction

" T h e recovery of enzyme activity in the peak fraction after the sucrose gradient sedimentation step is a corrected value reflecting the fact that only 307 of the hexylagarose fraction was further purified.

-~ " ~" ~

1 2 3 4 5 6

94.000 -

67,000-

43,000-

30,000-

- 67,000 - 60,000

- *

- 36,000 +

FIG. 1. SDS-polyacrylamide gel analysis of enzyme f ract ions at different stages of purification. Enzyme fractions at different stages of purification were analyzed on 105 SIIS-polyacrylamide gels as described under "Experimental I'rocedures." Lone I. 20-45,'; am- monium sulfate fraction; lnnr 2, IIEAE-cellulose-purified fraction: lane 3 , hydroxylapatite-purified fraction: Innr 4. hexylagarose-puri- fied fraction: Innr 5. peak fraction recovered after sucrose gradient sedimentation: lnnr 6. peak fraction recovered from native polyacryl- amide gel. Lanes 1-4 received approximately 1 pg of protein each. while lane 5 received 0.6 pg of protein. The estimated amount o f protein applied to lane 6 was 0.2-0.3 pg. I'olypeptides were detected by the silver-staining technique (11. 12). The migration positions of marker polypeptides are indicated on the right and lrft n!c:rgins. The major polypeptide (40.000 daltons) associated with nl'G(5')pppN- pyrophosphatase activity is indicated by an arrorc' at the right of lane 6 while a minor polypeptide (53,OOo daltons). also present in this fraction, is indicated by an asterisk.

enz-yme activity as the purification proceeded through the hexylagarose and sucrose gradient steps and represented greater than 95% of the protein present in the most purified fraction. Copurification of this polypeptide and m'G(5')pppN- hydrolyzing activity suggests the possibility that the native m'G(5')pppN-pyrophosphatase (81,000 daltons, see Table 11) is composed of two identical subunits of 40,000 daltons. How- ever, a polypeptide of 53,000 daltons was also observed as a minor component in fractions from the later purification steps (Fig. 1). Thus, confirmation that the M , = 40,000 polypeptide constitutes identical subunits of the m'G(5')pppN-p.wophos- phatase must await further experimentation.

by guest on April 5, 2018

http://ww

w.jbc.org/

Dow

nloaded from

6226 Purification of the Human m'G(57pppN-pyrophosphatase

Physical Properties and Substrate Specificity-Physical and enzymatic properties of the purified placental m'G(5')pppN-pyrophosphatase were determined for compar- ison with those previously established for the crude HeLa preparation (3). The methods used for determining these parameters were identical to those used for the HeLa enzyme analysis except that the effective substrate concentration was 1.8 PM (ie. above the I C n g ) rather than 80 nM. As expected, the purified placental enzyme and the crude HeLa preparation exhibited very similar chromatographic and sedimentation values as well as reaction optima (Table 11). Furthermore, similar values for the apparent molecular weight (-80,OOO) and K,, (1.5-1.7 PM) were calculated for both enzyme prepa- rations.

The HeLa m'G(5')pppN-pyrophosphatase activity ex- hibited a high degree of substrate specificity (3). G(5')pppG and ring-opened m'G(5')pppC"' were not hydrolyzed under conditions where m'G(5')pppC" was readily hydrolyzed. Gua- nosine cogeners methylated at the N' position were effective competitive inhibitors of enzyme-mediated cap hydrolysis while unmethylated guanosine cogeners were not. The 2- amino group of the 7-methylguanosine portion of m'G(5')pppN was also shown to participate in enzyme-sub- strate interactions: the trimethylated cap structure derived from U-2 RNA (methylated at the 2-amino position) was not a substrate for the pyrophosphatase and 7-methylinosine co- geners were not effective inhibitors of cap hydrolysis.

The substrate specificity of the purified placental enzyme was examined by determining the relative effect of methylated and unmethylated guanosine cogeners on enzyme-catalyzed m'G(5')pppG hydrolysis (Table 111). As previously shown with the HeLa enzyme, only the methylated guanosine cogeners were effective inhibitors.

Test for Contaminating RNAase Actiuity-The HeLa m'G(5')pppN-pyrophosphatase failed to hydrolyze the cap

TABLE I1 Comparison of physical and enzymatic properties of the HeLa and

placental m'G(5')pppN-pyrophosphatases Values for the HeLa enzyme are taken directly from Ref. 3. Meth-

ods used to obtain the values for the placental enzyme were identical to those described (3) except that the concentration of "H-methyl-cap substrate was 1.8 p ~ , i.e. above the Kn,. The sucrose gradient fraction of the placental enzyme was used to determine all parameters except Stokes radii which were determined with the hydroxylapatite fraction.

~~

Properties He12 enzyme Placental enzyme

Stokes radii 3.9 3.9-4.0 Sedimentation coefficient (S) 4.9 4.9 Estimated molecular weight 81,000 80.000-81,000 pH optimum 7.5 7.5-8.0 Temperature optimum ("C) 50 50-60 Michaelis constant (IC,#) ( p ~ ) 1.7 1.5 - ~. .

TABLE 111 Substrate specificity of the placental m'G(5')pppN-pyrophosphatase

Reaction conditions were as described for determining the sub- stfate specificity of the HeLa enayme (3) except that the [methyl-:'H] m'G(5')pppC concentration was 1.8 p~ (933 cpm/pmol) rather than 80 nM. Data are presented as the concentration of inhibitor giving 50% inhibition of cap hydrolysis ( I d and as the ratio of inhibitor concentration/substrate concentration a t LA,. The sucrose gradient fraction of the placental enzyme was used in this series of experiments.

Inhibitor 1 i l l Conc inhibitor/conc

substrate at IW

PC. PPG > I 0 0 mM >55,555 PPPG 60 mM 33,333 m:PG 2 PM 1.1 m'PPPG <2 pM <1.0

moiety of intact mRNA under a variety of in uitro conditions (1, 3). Since contaminating RNAase activity interferes with experiments aimed at examining conditions under which the enzyme might recognize and remove m'pG from intact mRNA, a major aim of this purification scheme was to provide enzyme preparations free of RNAase activity. Consequently, enzyme fractions were checked for RNAase activity using a very sensitive test. Wound tumor virus mRNA uniformly labeled with [a-:"P]UMP was incubated with 500 units of enzyme activity from each purification step and subsequently analyzed by polyacrylamide gel electrophoresis (Fig. 2) to score for endo- and exonucleolytic activity. Wound tumor

FIG. 2. Test for RNAase contamination in enzyme fractions at different stages of purity. Enzyme fractions (500 units) were incubated for 20 min at 37 "C with 50,000 cpm of [a""P]UMP-labeled wound tumor virus [-Sz(70)] mRNA (2 X IO" cpm/pg) in a final volume of 25 p1 containing 20 mM Tris, pH 7.5. At the end of the reaction, samples were diluted with an equal volume of 7 M urea, 90 mM Tris, pH 8.3, 90 mM boric acid, 4 mM EDTA, 0.58 SDS, and 20'; sucrose and boiled for 1 min. The samples were then subjected to polyacrylamide gel electrophoresis and subsequent autoradiography as described under "Experimental Procedures." Lane I , unincubated RNA; lane 2, RNA incubated in the absence of enzyme; lane 3, RNA incubated in the presence of the 20-45% ammonium sulfate fraction; lane 4, RNA incubated in the presence of the DEAF,-cellulose-puri- fied fraction; lane 5, RNA incubated in the presence of the hydrox- ylapatite-purified fraction; lane 6, RNA incubated in the presence of the hexylagarose-purified fraction; lane 7, RNA incubated in the presence of enzyme fraction recovered after sucrose gradient sedi- mentation; lane 8, RNA incubated in the presence of the enzyme fraction recovered from native polyacrylamide gel electrophoresis. Wound tumor transcripts possess molecular weights ranging between 1.5 X 10" and 1.6 X 1 0 , indicated by the upper and lower arrows at the right margin, respectively.

by guest on April 5, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Purification of the Human m7G(5')pppN-pyrophosphatase 6227

TABLE IV Failure of human m'G(5')pppN-pyrophosphatase to hydrolyze the

cap moiety of intact mRNA [:IH]m'pC released

from Substrate and/or conditions

Exogenous RNA" 0 pmol 9 pmol

45 pmol 90 pmol

+ RNasin - RNasin

- Dimethyl sulfoxide + Dimethyl sulfoxide

[nethyL3H]IMP-substituted RNAb

Dimethyl sulfoxide-treated [methyl-3H]RNA'

CH3HgOH-treated [methyl-'H]RNA" - CH'HgOH + CH3HgOH

methyl-'HlReovirus double-stranded RNA'

pmol

33.3 36.2 34.4 42.0

52.6 53.7

46.3 35.8

49.7 46.4

Effect of RNA on cap hydrolysis. Reaction mixtures (25 pl) contained 40 mM Tris, pH 7.5, 58 pmol of [methyl-'H]m7GpppG, 140 units of sucrose gradient-purified pyrophosphatase, and varying amounts of in vitro synthesized vaccinia virus or reovirus RNA (0-90 pmol). Incubation was for 10 min at 37 "C. The reaction was termi- nated by boiling for 2 min and reaction products were analyzed by paper chromatography in isobutyric acid, 0.5 N NH40H (106 v/v).

IMP-substituted [methyl-"H]reovirus mRNA as substrate for the pyrophosphatase. The 25-pl reaction mixtures contained 40 mM Tris, pH 7.5, 58 pmol of [methyl-'H]m%pppG or 2 pmol of [methyL'H] IMP-substituted reovirus RNA, and 140 units of the sucrose gradient- purified pyrophosphatase. Since IMP-substituted RNA is more sus- ceptible to RNAase activity than unsubstituted RNA (5), some re- actions received 20 units of RNasin and 1 mM dithiothreitol to ensure continued stability of the RNAase inhibitor. Incubation was per- formed at 37 "C for 30 min. The reaction was terminated by boiling the mixture containing m'GpppG and diluting the mixture containing RNA with 125 pl of 50 mM Tris, pH 7.5,O.l M mercaptoethanol, 0.5% SDS, and 0.1 M NaCI. The mixtures containing RNA were further processed by passage through a column (0.9 X 30 cm) of G-50 Sephadex equilibrated with the dilution buffer. In each case, the majority (greater than 95%) of the 3H-methyl-radioactivity eluted with the excluded column volume. RNA in this fraction was precipi- tated with ethanol, resuspended in 100 pl of 5 mM NaOAc, pH 6.0, treated sequentially with nuclease P, and bacterial alkaline phospha- tase (l), and analyzed by paper chromatography to determine the relative quantities of m% and G" in the cap moiety of the RNAs, i.e. to detect any removal of m7pG from intact mRNA. In subsequent experiments, the cap moiety derived from IMP-substituted reovirus RNA by nuclease P, digestion was shown to be a substrate for the m7G(5')pppN-pyrophosphatase. ' Dimethyl sulfoxide-treated [methyl-'Hlreovirus RNA as substrate

for the m7G(5')pppN-pyrophosphatase. [methyl-'H]Reovirus RNA (24 pmol) or [methyZ-3H]m%pppG (170 pmol) was dried in uucuo, resuspended in 6 p1 of 90% dimethyl sulfoxide, 25 mM Tris, pH 7.5, and incubated for 15 min at 50 "C. The denatured substrates were diluted into reaction mixtures containing 40 mM Tris, pH 7.5, and 140 units of the sucrose gradient-purified pyrophosphatase at 50 "C, and incubated at 50 "C for 15 min. The final reaction volume was 25 pl, the fiial concentration of substrate was 58 pmol of m'GpppG or 10 pmol of RNA/reaction, respectively, and the final dimethyl sulfoxide concentration was 7.2%. The reactions were terminated by boiling reaction mixture containing rn7GpppG or by diluting reaction mixture containing RNA as described for the IMP-substituted RNA, and the reaction products were analyzed. ' CHaHgOH-treated [methyl-'H]reovirus RNA. [methyl-'H]Reo-

virus RNA (24 pmol) or [methyG3H]m7GpppG (170 pmol) was dried and resuspended in 6 pl of 5 mM CH3HgOH in 40 mM borate buffer, pH 7.6, and incubated 20 min at 50 "C. The denatured substrates were diluted into reaction mixture containing 40 mM borate buffer, pH 7.6, and 140 units of sucrose gradient-purified pyrophosphatase at 50 "C. Incubation was performed at 50 "C for 15 min. The final reaction volume was 25 pl, and the final concentrations of the sub-

virus [a-32P]mRNA was completely degraded by fractions obtained through the hexylagarose column step (lunes 3-6, Fig. 2). Sucrose gradient sedimentation of the hexylagarose fraction removed the majority of RNAase contamination (lane 7, Fig. 2) while the enzyme fraction recovered after polyacrylamide gel electrophoresis under nondenaturing con- ditions was free of any RNAase contamination (lane 8, Fig. 2). Thus, this purification scheme is useful for preparing enzyme fractions suitable for studies concerning the inability of the m7G(5')pppN-pyrophosphatase to recognize the cap moiety of intact mRNA.

Failure of the m7G(5')pppN-pyrophosphatase to Hydrolyze the Cup Moiety of Intact mRNA-Inability of the m7G(5')pppN-pyrophosphatase to hydrolyze the cap moiety of intact mRNA could result from (i) the direct inhibition of pyrophosphatase activity by the polynucleotide chain of the mRNA molecule owing to its polyanionic character, (ii) mRNA secondary structure restraints that prevent access of the enzyme to the cap moiety, or (iii) steric hindrance by nucleotides immediately adjacent to the cap. The fist possi- bility was investigated by testing the effect of exogenous RNA on pyrophosphatase activity. The failure of even high concen- trations of exogenous RNA to inhibit cap hydrolysis (Table IV) makes it unlikely that the polyanionic character of the polynucleotide chain of a mRNA molecule contributes to the inability of the pyrophosphatase to hydrolyze the cap moiety of that molecule.

The possible role of mRNA secondary structure restrictions in preventing access to the cap moiety was explored by testing pyrophosphatase activity on capped mRNA under conditions known to remove or reduce mRNA structure. mRNA second- ary structure was reduced by treatment with 90% dimethyl sulfoxide (14), with 5 m~ CH3HgOH or by substituting IMP for GMP in the polynucleotide chain (5), and the mRNAs were then treated with the purified pyrophosphatase. Al- though [methyl-'H]m7G(5')pppG was readily hydrolyzed un- der all conditions tested, the cap moiety of intact mRNA remained unhydrolyzed (Table IV). The m7G(5')pppN-pyro- phosphatase also failed to hydrolyze the cap of mRNA in a more rigid double-stranded form (Table IV) or in an initiation complex with wheat germ ribosome (16) (data not shown).

To test the effect of additional nucleotides immediately adjacent to the cap in the 3' direction on the ability of the pyrophosphatase to hydrolyze the cap moiety, a series of 5'- terminal capped oligonucleotides were generated by T1-ribo- nuclease digestion of 'H-methyl-labeled vaccinia virus mRNA. Advantage was taken of the fact that vaccinia virus particle synthesizes primarily monomethylated mRNA at low S-aden- osylmethionine concentrations.2 Since the penultimate bases in the cap of vaccinia mRNA are G and A (13), TI-ribonuclease digestion of the monomethylated mRNA generated the fol- lowing series of 'H-methyl-labeled 5'-terminal fragments with the same specific activity: m7GpppGp, m7GpppApGp,

strates were 58 pmo1/25 pl for m7GpppG and 10 pmo1/25 pl for RNA while CH:jHgOH was at 0.4 mM. The reactions were terminated and reaction products analyzed as described for the IMP-substituted RNAs.

[methyl-'HH]Reovirus double-stranded RNA. In vitro synthesized [methyl-'H]reovirus RNA was put into a double-stranded form by hybridization with reovirus genome RNA using conditions described recently for wound tumor virus (6) and subsequently tested as sub- strate for the m'G(5')pppN-pyrophosphatase. The 25-pl reaction mix- ture contained 40 mM Tris, pH 7.5,80 mM NaCl to stabilize the RNA in a double-stranded configuration, 4 pmol of 'H-methyl-labeled reovirus double-stranded RNA, and 140 units of the sucrose gradient- purified pyrophosphatase. Incubation was performed at 37 "C for 15 min. Reactions were terminated and reaction mixtures processed as described for IMP-substituted RNA.

by guest on April 5, 2018

http://ww

w.jbc.org/

Dow

nloaded from

6228 Purification of the Human m‘G(5’)pppN-pyrophosphatase

CHAl N LENGTH FIG. 3. Relationship between oligonucleotide chain length

and the rate of m’G(5‘)pppN-pyrophosphatase-catalyzed hydrolysis of the cap moiety. A series of ‘”H-methyl-labeled capped oligonucleotides of the following types were generated by ribonuclease TI digestion of monomethylated vaccinia virus mRNA (see “Experimental Procedures”): m‘GpppGp, m’GpppApGp, m7GpppApXpGp, etc. These capped oligonucleotides of Identical specific activities were tested as substrate for the purified m’G(5’)pppN-pyrophosphatase. The reaction mixture of 25 1.11 con- sisted of 40 mM Tris, pH 7.5, ”H-methyl-labeled oligonucleotides at a concentration of 100 nM, 70 units of the sucrose gradient-purified m’G(5’)pppN-pyrophosphatase fraction, and 20 units of the RNAase inhibitor RNasin. Following incubation for 5 min at 37 “C, the reac- tion mixture was boiled for 1 min to stop the reaction and the products were analyzed by paper chromatography using isobutyric acid, 0.5 N NH,OH (10% v/v) as the solvent. The rate of hydrolysis was deter- mined by scoring for the amount of ,’H-methyl-labeled m‘pG liberated from the oligonucleotide during the 5-min reaction (1). The chain length, as presented on the abscissa, indicates the number of nucleo- tides adjacent to the cap in the 3’ direction, i.e. m‘GpppGp = a chain length of 0, while m’GpppApXpGp = a chain length of 2.

m7GpppApXpGp, etc. These 5’-terminal fragments were iso- lated by DEAE-cellulose chromatography in 7 M urea ( l ) , desalted, and tested under identical conditions as substrate for the m7G(5’)pppN-pyrophosphatase. Cap hydrolysis, as measured by m’pG release from the oligonucleotides, de- creased as the oligonucleotide chain length increased (Fig. 3). The purified pyrophosphatase failed to release m7pG from oligonucleotides containing the cap and 3 or more nucleotides even upon prolonged incubation (60 min). Thus, m’G(5’)pppN-pyrophosphatase-catalyzed hydrolysis between the p and y phosphate of m’G(5’)pppN (where N is any base, methylated or not) (3) involves an enzyme-substrate interac- tion that is prevented by the addition of 3 or more nucleotides to the N moiety, i.e. enzyme attack appears to be from the penultimate base side.

r l h

DISCUSSION

There exists considerable evidence which suggests that the 5’-terminal structure, or cap, of eukaryotic mRNA plays a significant role in polypeptide chain initiation (17-19) and may significantly influence the stability of mRNA by confer- ring resistance to the action of cellular 5’ + 3’ exoribonuclease activities (20-22). Cap formation is also intimately associated with initiation of cellular (23) and viral (24-26) mRNA syn- thesis. In light of the importance of the cap in mRNA synthe- sis, translation, and stability, and our current level of under-

standing of these events, enzymes or factors which specifically recognize the cap structure must be scrutinized as potential effectors of these processes.

TWO activities that specifically recognize the cap have been reported. A cap-specific binding activity was detected in ex- tracts of Artemia salina (27) and rabbit reticulocytes (28). The latter activity, the 24,000-dalton cap-binding protein, was purified to apparent homogeneity (29). An enzyme which specifically recognizes and hydrolyzes the cap was first de- tected in extracts of HeLa cells (1). This m7G(5’)pppN-pyro- phosphatase was partially purified and characterized (3). A similar activity was reported in crude extracts of chick lens but was not characterized (2). Nucleotide pyrophosphatase activities which hydrolyze m7G(5’)pppN have also been de- scribed in tobacco (30) and potato (31). However, it must be stressed that these enzymes, unlike the human m7G(5’)pppN- pyrophosphatase, are not specific for the cap and will readily hydrolyze unmethylated G(5’)pppN as well as the ring-opened derivative of m7G(5’)pppN.

In this communication we described a scheme for the puri- fication, from placental tissue, of the human m7G(5’)pppN- pyrophosphatase to a high degree of purity, free of contami- nating RNAase activity. The purified placental enzyme is similar to the HeLa m7G(5’)pppN-pyrophosphatase with re- gard to physical properties and substrate specificity. Further- more, the purified placental enzyme, like the HeLa activity, failed to hydrolyze the cap moiety of intact mRNA even under conditions known to reduce secondary structure (Table IV). With respect to this last point, the detection of an mRNA- decapping enzyme in the ribosomal fraction of Saccharomyces cereuisiae has recently been reported (32). This activity was reported to release m7ppG, rather than m7pG, from intact yeast mRNA. However, the author stressed that the enzyme fraction did contain significant ribonuclease activity. Most interesting was the observation that the yeast enzyme was unable to hydrolyze m7G(5’)pppA(G) derived from the 5’- termini of the yeast mRNA that was used as substrate t.o detect the activity. It will, therefore, be most interesting to compare the yeast enzyme, once suitably purified, with the human m7G(5’)pppN-pyrophosphatase.

It was previously reported (1) that the HeLa m7G(5’)pppN- pyrophosphatase activity hydrolyzed the cap moiety of oli- gonucleotides containing as many as 8 nucleotides. The puri- fied placental enzyme, in contrast, failed to hydrolyze the cap moiety when extended in the 3’ direction by the addition of 3 nucleotides (Fig. 3). The probable reason for this discrepancy lies in the fact that in the previous study an unresolved preparation of reovirus 5”terminal ribonuclease TI fragments {4-10 nucleotides in length) was used as substrate and the HeLa preparation contained ribonuclease activity that would reduce the chain length of the oligonucleotides during the reaction. The progressive decrease in the rate of cap hydrolysis with increasing oligonucleotide chain length suggests that the pyrophosphatase achieves hydrolysis of the 5”5’ pyrophos- phate Linkage by attacking from the penultimate base side of the molecule, In this regard, the purified m7G(5’)pppN-pyro- phosphatase represents a potentially useful model system with which to determine precisely how cap-specific enzymes and factors interact with this unique and functiondly impor- tant structure.

Although the substrate specificity of the purified human m7G(5’)pppN-pyrophosphatase is now well documented, the biological function of the enzyme remains unclear. We have previously suggested that the enzyme functions a t some step in the chain of events that results in mRNA decay (I, 3). While the enzyme is unable to release m7pG from intact mRNA in uitro, it may perform this role in vivo, in association

by guest on April 5, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Purification of the Human m

with other factors rendering mRNA susceptible to 5’ -+ 3’ exoribonucleolytic attack. More likely, the m7G(5’)pppN-py- rophosphatase is involved in the metabolism, or finishing, of capped oligonucleotide intermediates of the mRNA degrada- tive pathway. Certainly, a mechanism must exist for prevent- ing the accumulation of caps or capped oligonucleotides as mRNA decays since these intermediates would act as inhibi- tors of polypeptide chain initiation (33). The m7G(5’)pppN- pyrophosphatase may represent the eukaryotic cells answer to the problem of how to metabolize caps resulting from mRNA decay without, at the same time, hydrolyzing the cap structure in intact functional mRNA. In any event, the ability, as described in this communication, to prepare highly purified, ribonuclease-free, preparations of the human m7G(5’)pppN- pyrophophatase represents an important step toward the ul- timate elucidation of its functional role in eukaryotic cells.

REFERENCES 1. N u s , D. L., Furuichi, Y., Koch, G., and Shatkin, A. J. (1975) Cell

2. Lavers, G. C. (1977) Mol. Biol. Rep. 3,413-420 3. Nuss, D. L., and Furuichi, Y. (1977) J. Biol. Chem. 252, 2815-

4. Furuichi, Y., Muthukrishnan, S., Tomasz, J., and Shatkin, A. J.

5. Kozak, M. (1980) Cell 19, 79-90 6. N u s , D. L., and Peterson, A. J. (1981) Virology 114, 399-404 7. Reddy, D. V. R., and Black, L. M. (1977) Virology 80, 336-346 8. Nuss, D. L., and Paoletti, E. (1977) J. Virol. 23, 110-116 9. Anderson, C. W., Baum, P. R., and Gesteland, R. F. (1973) J.

6.21-27

2821

(1976) J. Biol. Chem. 251, 5043-5053

Virol. 12, 241-252 10. Laemmli, U. K. (1970) Nature (Lond) 227,680-685 11. Switzer, R. C., 111, Merril, C. R., and Shifrin, S. (1979) Anal.

12. Oakley, B. R., Krisch, D. R., and Morris, N. R. (1980) Anal. Biochem. 98, 231-237

Biochem. 105, 361-363

7G(S)pppN-pyrophosphutuse 6229

13. Wei, C. M., and Moss, B. (1975) Proc. Natl. Acad. Sci. U. S. A.

14. Strauss, J. H., Kelly, R. B., and Sinsheimer, R. L. (1968) Biopol-

15. Bailey, J. M., and Davidson, N. (1976) Anal. Biochem. 70,75-85 16. Kozak, M., and Shatkin, A. J. (1976) J. Biol. Chem. 251, 4259-

4266 1’1. Both, G . W., Banerjee, A. K., and Shatkin, A. J. (1975) Proc.

Natl. Acad. Sci. U. S. A . 72, 1189-1193 18. Filipowicz, W. (1978) FEBS Lett. 96, 1-11 19. Kozak, M. (1978) Celt 15, 1109-1123 20. Furuichi, Y., LaFiandra, A., and Shatkin, A. J. (1977) Nature 266,

21. Lockard, R. E., and Lane, C. (1978) Nucl. Acids Res. 5,3237-3247 22. Shimotohno, K., Kodama, Y., Hashimoto, J., and Miura, K.-I.

23. Darnell, J. E., Jr. (1979) Prog. Nucl. Acid. Res. Mol. Biol. 22,

24. Banerjee, A. K., Abraham, G., and Colonno, R. J. (1977) J. Gen.

25. Krug, R. M., Broni, B. A., and Boulay, M. (1979) Cell 18,329-334 26. Wertheirner, A. M., Chen, S.-Y., Borchardt, R. T., and Furuichi,

Y. (1980) J. Biol. Chem. 255, 5924-5930 27. Filipowicz, W., Furuichi, Y., Sierra, J. M., Muthukrishnan, S.,

Shatkin, A. J., and Ochoa, S. (1976) Proc. Natl. Acad. Sci. U.

28. Sonenberg, N., Morgan, M. A., Testa, D., Colonno, R. J., and Shatkin, A. J. (1979) Nucl. Acids Res. 7, 15-29

29. Sonenberg, N., Rupprecht, K. M., Hecht, S. M., and Shatkin, A. J. (1979). Proc. Natl. Acad. Sci. U. S. A. 76,4345-4349

30. Shinshi, H., Miwa, M., Kato, K., Noguchi, M., Matsushima, T., and Sugimura, T. (1976) Biochemistry 15,2185-2190

31. Zan-Kowalczewska, M., Bretner, M., Sierakowska, H., Szczesna, E., Fillipowicz, W., and Shatkin, A. J. (1977) Nucl. Acids Res. 4,3065-3081

32. Stevens, A. (1980) Bochem. Biophys. Res. Commun. 96, 1150- 1155

33. Hickey, E. D., Baglioni, C., and Weber, L. A. (1976) Nature 261, 71-72

72,318-322

ymers 6, 793-801

235-239

(1977) Proc. Natl. Acad. Sci. U. S. A. 74,2734-2738

327-353

Virol. 34, 1-8

S. A . 73, 1559-1563

by guest on April 5, 2018

http://ww

w.jbc.org/

Dow

nloaded from

6230 Purification of the Human m7G(5')pppN-pyrophosphatase

I 1

FRACTION NUMBER

and t h d d n g , when stoked a t -20'

P

80TT0M TOP

FRACTION NUMBER

p l y a c r y l a m l d e e l e c t r o p h a r e r l r ?

to l e f t

by guest on April 5, 2018

http://ww

w.jbc.org/

Dow

nloaded from

D L Nuss, R E Altschuler and A J Petersonhuman placenta.

Purification and characterization of the m7G(5')pppN-pyrophosphatase from

1982, 257:6224-6230.J. Biol. Chem. 

  http://www.jbc.org/content/257/11/6224Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/257/11/6224.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on April 5, 2018

http://ww

w.jbc.org/

Dow

nloaded from