Presence of cap structures in the messenger RNA of mouse eggs

A considerable body of experimental data has recently been accumulated which indicates that stable messenger RNA (mRNA) molecules are important templates for protein synthesis during early post-fertilization and cleavage stages of mouse development. Since there is little evidence for active mRNA synthesis immediately following fertilization, much of the stable mRNA may be maternally derived and stored in the oocyte (see Johnson, 1979, for review). A number of qualitative stage-specific changes in the kinds of proteins synthesized by these early embryos also occur (Van Blerkom, 1977). Some of these new proteins first detectable at the late 1-cell to early 2-cell stage are present in translates of mRNA extracted from unfertilized eggs although they are not normally synthesized by the egg itself (Braude, Pelham, Flach & Lobatto 1979). Furthermore, the appearance of these polypeptides is insensitive to transcriptional inhibition with α-amanitin. These results lead to the conclusion that some form of post-transcriptional control is operative for the expression of at least some of the egg mRNA molecules during early development. The question then arises as to what form such selective mRNA activation might take.

Most eukaryotic mRNAs contain at their 5′-ends a cap structure of the form rn⁷G⁵′ppp⁵′X^mp^y… which is added post-transcriptionally to the primary transcript (Shatkin, 1976; Revel & Groner, 1978). This modification protects the mRNA at its 5′-terminus against enzymatic attack by phosphatases and other nucleases. It also appears to be functionally important in mRNA translation since analogues such as m⁷GMP or m⁷GTP competitively inhibit translation of capped message (Canaani, Revel & Groner, 1976; Hickey, Weber & Baglioni, 1976a; Weber, Hickey & Baglioni, 1976b). Furthermore, chemical (Muthukrishnan et al. 1975; Both, Banerjee & Shatkin, 1975) and enzymic (Zan-Kowalczewska, Furuchi, Both & Shatkin, 1977) removal of cap structures reduces or eliminates translatability of mRNA both in vitro and in vivo (Lockard & Lane, 1978). Thus, modifications of the 5′-ends of mRNAs provide one way in which translational control might be exercised. Interestingly, in this context, newly fertilized mouse eggs incubated in media containing [³H]guanosine appear to incorporate a small amount of this precursor into cap structures (Young, 1977).

We have, therefore, assayed for the presence of cap structures in mRNA molecules of mouse eggs through (1) reduction in in vitro translatability in the presence of the cap analogue m⁷GTP, and (2) enzymatic cap removal and endlabelling with γ-³²P-ATP and polynucleotide kinase. The results suggest that the majority of mRNA in the unfertilized egg is capped. No change in the degree of capping of mRNA molecules was detected between the unfertilized egg and fertilized egg stage.

Six-to nine-week-old mice (outbred CFLP, Anglia Laboratory animals) were superovulated with an intra-peritoneal (i.p.) injection of 5 or. 10 i.u. Pregnant Mares Serum Gonadotrophin (Folligon, Intervet, Cambridge) followed 44 to 48 h later by 5 or 10 i.u. of Human Chorionic Gonadotrophin (HCG, Chorulon, Intervet) given i.p. Fertilized eggs were obtained from females which had vaginal plugs after being caged singly with CFLP males overnight. Both mated and unmated animals were sacrificed 13 –15 h post-HCG and the oviducts dissected out into 0·9% saline. Cumulus masses were liberated into a phosphate-buffered medium (PB1 containing 0·4% (w/v) bovine serum albumin, Whittingham & Wales, 1969). This was replaced for 5 –10 min by a medium containing 1 mg/ml hyaluronidase (Koch Light, Colnbrook, Bucks.) and 10 mg/ ml polyvinyl pyrrolidone to remove cumulus cells. Subsequently the eggs were transferred through two to three washes of PB1+BSA by use of a fine mouth pipette. Morphologically abnormal ova (or, in the case of fertilized eggs, eggs not displaying two pronuclei and second polar body extrusion) were discarded, and the remaining cells counted.

Total RNA was extracted essentially as described by Braude & Pelham (1979). Eggs were transferred in the minimum possible volume to a 1 ml conical centrifuge tube to which 2 μ1 of calf liver tRNA (2·5 mg/ml, Boehringer-Mannheim) was added as carrier to aid subsequent precipitation. Immediately 20 μl of phenol saturated with extraction buffer and 20 μl of extraction buffer (0·2 M-NaCl, 0·001 M EDTA, 0·02 M Tris, pH 7·4) were added. After thorough mixing, the suspension was drawn up a microhaematocrit tube by capillary action, one end of the tube was sealed with putty, and the two phases separated by centrifugation for 1 minute at 5000 rev./min in a haematocrit centrifuge. The tube was scored above the interface, and the aqueous phase transferred to a clean centrifuge tube followed by 2 vols. (50 –60 μ1) of cold absolute ethanol. The RNA precipitate was recovered by centrifugation at 28000g for 30 min in a Sorvall RC-5 centrifuge after being left overnight at –20°C. The extraction procedure was repeated for 2 μl of calf liver tRNA alone to provide a control.

The translation system used was the cell-free reticulocyte lysate described by Pelham & Jackson (1976) as adapted by Braude & Pelham (1979). For some experiments the same lysate was used without prior treatment with staphylococcal nuclease thus leaving endogenous translational activity.

Egg RNA was prepared for translation by dissolving it in [³⁵S]methionine (1000 –1300 Ci/mmole, 6-85 mCi/ml, Amersham); 5 μl of methionine were used for every 20 μl of lysate required in the incubation. The mixture was freeze-dried and the requisite volume of lysate added. The resulting translation mix was distributed in 20 μl aliquots into experimental tubes. Either GTP or the cap analogue 7-methylguanosine-5′-triphosphate (m⁷GTP, P-L Biochemicals Inc., Milwaukee, Wise.) were added to different incubation tubes in the amounts indicated in the Results section. Since nucleoside triphosphates can chelate Mg²⁺, wherever it is stated that m⁷GTP (or GTP) was added, an equimolar amount of MgCl₂ was also added to restore the magnesium concentration reduced by chelation. All tubes had a final volume of 22 μ1.

After incubation for 1 h at 37°C, 2 μl of RNAse A (2 mg/ml) in a solution of 20 mM methionine, 0·1 M EDTA was added. Tubes were incubated for a further 10 to 15 minutes after which 2 μl samples were taken using a Hamilton micro-syringe to assess quantitatively incorporation of [³⁵S]methionine into protein. The remainder of the sample was stored at –70°C until analysed by polyacrylamide gel electrophoresis. The 2 μl samples were diluted in 250 μl water. An equal volume of 1 N-NaOH, 0·5 M-H₂O₂ containing 1 mg/ml methionine was added and the samples incubated until decolorized. Samples were precipitated with 0·5 ml of 25% trichloroacetic acid (TCA), filtered on glassfibre filters (Whatman GF/C), washed with 8% TCA and dried. Filters were counted in a toluene-based scintillation cocktail (Cocktail N, Fisons, Loughborough) at an efficiency of approximately 70%.

Samples of translation products were analyzed qualitatively in one dimension on SDS polyacrylamide slab gels exactly as described by Braude and Pelham (1979) except that the separating gel was composed of a uniform 10% acrylamide rather than a gradient of 7 –15%. Radioactive proteins were visualized by fluorography (Laskey & Mills, 1975) using preflashed Fuji Rx X-ray film and exposure for 3 –14 days.

Because the terminal nucleotide of the cap structure has an inverted 5 –5′ linkage relative to the normal 3′ –5′ phosphodiester bonds in the rest of the polynucleotide chain, capped RNA molecules are not capable of being end-labelled with polynucleotide kinase and ATP. Non-capped RNA species with 5′-phosphates can be labelled after removal of the terminal phosphate with alkaline phosphatase. However the enzyme tobacco acid phosphatase (TAP) can cleave the cap structures to yield 5′-phosphate termini which subsequently can also be end-labelled. Therefore, by end-labelling RNA molecules with and without treatment with TAP, a method is provided to discriminate between capped and non-capped RNA molecules. This approach was applied to analysis of RNA extracted from unfertilized and fertilized eggs in the presence of tRNA carrier.

Removal of 5′-end cap structures, dephosphorylation, and subsequent end-group labelling with γ-³²P-ATP and polynucleotide kinase was performed in the same reaction mix using slight modifications of the procedure of Efstratiadis et al. (1977). Egg RNA or tRNA control preparations were divided into two equal aliquots and incubated in a 10 μ1 reaction mixture of 25 mM sodium acetate, pH 6·0, 10 mM mercaptoethanol, 0·2 mM EDTA in the presence ( + TAP) or absence (-TAP) of 1 unit of tobacco acid phosphatase at 37°C for 60 min. Subsequent procedures were identical for both the decapped (+TAP) and sham ( – TAP) aliquots. To each tube was added 2 μl of 0·5 M-Tris-HCI, pH 8·3, and 0·2 unit of bacterial alkaline phosphatase (Bethesda Research Laboratories, Inc., Rockville, Md.). Dephosphorylation was allowed to continue at 37°C for 30 min after which 6 μ1 of a solution containing 33 mM MgCl₂ and 33 mM dithiothreitol (DTT) was added. The alkaline phosphatase was inhibited by addition of 1 μl of 50 mM potassium phosphate, pH 9·5 and the entire mixture transferred to a clean tube in which 250 μ1Ci (approximately 80 pmoles) of γ-³²P-ATP (3000 Ci/m-mole, Amersham) had been taken to dryness by lyophilization. The reaction mix was brought to a final volume of 25 μ1 with H₂O and phosphorylation initiated by the addition of 1 μl (4 units) of poly-nucleotide kinase. Following further incubation at 37°C for 30 min, the reaction was terminated by addition of 5 μ1 of calf liver tRNA (2·5 mg/ml), 5 μl 2·4 M ammonium acetate and 100 μl ethanol. The mixture was allowed to precipitate overnight at −20°C and recovered by centrifugation.

The end-group-labelled RNA pellets were dissolved in 0·25 ml of 0·5 M ammonium acetate and applied to small columns (0·25 ml bed volume) of oligo-dT cellulose (Boehringer-Mannheim) made in cut-off Pasteur pipettes and equilibrated in the same buffer. The columns were washed with five 0·5 ml aliquots of 0·5 M ammonium acetate followed by four 0·5 ml aliquots of 0·1 M ammonium acetate such that no further radioactive material was eluted. Poly (A)-containing RNA (poly(A) + RNA) bound to the affinity column was then eluted with 1 ml of H₂O. For analysis on Polyacrylamide gels, RNA was recovered by lyophilization. For quantitative analysis, the RNA was precipitated with an equal volume of 10% TCA after addition of 100 μg carrier yeast RNA. The precipitates were trapped on GF/C filters, washed with 5%, TCA, dried and counted.

Tobacco mosaic virus (TMV), Cowpea mosaic virus (CPMV), and Tobacco rattle virus (TRV) mRNAs were kindly provided by Dr H. R. B. Pelham, Department of Biochemistry, University of Cambridge. Tobacco acid phosphatase was a gift from Dr A. Efstatiadis, Biological Laboratories, Harvard University, and polynucleotide kinase was a gift from Drs G. Chaconas and H. van de Sande, Division of Medical Biochemistry, University of Calgary.

Reticulocyte lysate with endogenous message activity was used to characterize a system to assess degree of capping of mRNA molecules by inhibition of translation with the cap analogue m⁷GTP. When increasing concentrations of m⁷GTP were added to this lysate, translation of mRNA into protein was reduced to 20% of control values with concentrations as low as 0-5 mM m⁷GTP (Fig. 1). This reduction in translation was not simply due to addition of guanosine nucleoside triphosphate since equivalent concentrations of GTP did not affect translatabiiity and at concentrations greater than 0·5 mM, even stimulated protein synthesis. Polyacrylamide gel separations of the translates were fluorographed to provide a qualitative visualization of the synthesized polypeptides (Fig. 2). Globin polypeptides which contain greater than 90% of the incorporated [³⁵S]methionine in these translations have been run off the bottom of the gel to allow clearer visualization of the remaining translation products. Clearly a reduction in the synthesis of all types of polypeptides coded for by the endogenous message occurs in the presence of m⁷GTP when compared with the equivalent effect of GTP.

Further characterization of the system was accomplished in message-dependent lysate by testing the effects of m⁷GTP on the translation of TMV mRNA which is known to be capped (Kieth & Fraenkel-Conrat, 1975) and on the translation of CPMV mRNA which is totally non-capped (Klootwijk, Klein, Zabel & Van Kammen 1977). Translation of the capped TMV mRNA was reduced to about 30% of control values using 0·5 or 1·0 mM m⁷GTP while the uncapped CPMV mRNA was virtually unaffected (Table 2).

The translation of mRNA extracted from unfertilized eggs was also reduced to about 20% of control values in the presence of 0·5 to 1·0 mM m⁷GTP (Fig. 3). Qualitative analysis of the products on polyacrylamide gels indicated that the synthesis of all polypeptides coded for by the egg mRNA population was affected by the m⁷GTP (Fig. 4). The one radioactive band remaining even in the tRNA background sample is the consequence of addition of [³⁵S]methionine to an endogenous reticulocyte polypeptide by a ribosome-independent reaction (Pelham & Jackson, 1976), therefore, explaining its insensitivity to m⁷GTP. The quantitative aspects of experiments on other unfertilized egg mRNA preparations are summarized in Table 1 along with data on the non-inhibitory effect of GTP.

End-labelling procedures with and without treatment with TAP were utilized to discriminate between capped and non-capped mRNA molecules as described in Materials and Methods. To assess that the enzymes and reaction conditions employed did not lead to mRNA degradation, end-labelling was performed with TAP removal of cap structures of tobacco rattle virus (TRV) mRNA and the electrophoretic mobility of this labelled material as detected by autoradiography compared to that of untreated TRV mRNA detected by staining procedures (channels A and B, Fig. 5). With the gel conditions utilized, two high molecular weight mRNA components are resolved in the stained pattern. These two mRNA species remain intact with identical mobility to the starting material after the end-labelling procedures (Fig. 5).

The same procedures were then applied to analysis of RNA extracted from unfertilized and fertilized eggs in the presence of tRNA carrier. Under the conditions utilized, about 20 × 10⁶ c.p.m. of γ-³²P-ATP were transferred to 5′-ends of RNA molecules although clearly the predominant transfer was into the tRNA carrier which was in large excess to that of egg RNA (Table 2). When this material was subjected to affinity chromatography on oligo-dT cellulose to purify the poly(A) + RNA (putative message) component of the egg RNA, less than 0·1% was bound and this varied with the number of eggs extracted. An example of an experiment conducted on RNA extracted from 616 unfertilized eggs is summarized in Table 2. End-labelled tRNA alone (whether TAP treated or not) binds non-specifically to the oligo-dT cellulose at the level of 0·012 and 0·011%, respectively. Over several experiments this background value varied from 0·009% to 0·014%. Egg RNA extracted in the presence of the tRNA carrier bound to the level of 0·088% above the background value when end-labelled after ‘cappase’ treatment with TAP. In contrast, in the absence of ‘cappase’ treatment, only 0·021% of the ³²P-label purified with poly(A) + RNA. Assuming the EGG+TAP sample represents end-labelled poly(A) + RNA molecules which were both capped and uncapped while EGG-TAP represents endlabelling of only uncapped molecules, the ratio between the two values leads to the calculation that 76·1% of the poly(A) + RNA molecules (putative mRNA) of this unfertilized egg RNA preparation are capped (Table 2). The amount of radioactivity bound to the oligo-dT cellulose was converted to percentages to correct for minor differences in total radioactivity applied to the column from one sample to the next. The results of a number of such analyses performed on sets of 400 –700 unfertilized or fertilized eggs are summarized in Table 3. In each case, the proportion of poly(A)-containing RNA which was end-labelled after decapping procedures was always much higher than in the absence of decapping procedures.

In order to demonstrate that the quantitative differences in ³²P-incorporation observed in the presence or absence of TAP treatment did not simply reflect variation in the amount of tRNA carrier bound to oligo-dT cellulose columns, poly(A)-containing, end-labelled RNAs from an experiment similar to that summarized in Table 2 were resolved on polyacrylamide gels (Fig. 5). A small amount of labelled control carrier tRNA does bind to the oligo-dT cellulose column and is resolved as a few bands in the lower part of the gel (channel E, Fig. 5). In the EGG + TAP sample, a number of labelled poly(A) + RNA bands of higher molecular weight than the carrier tRNA are resolved and clearly are of egg origin (channel C, Fig. 5). As expected, few labelled poly(A) + RNA bands are resolved in the EGG —TAP sample under identical conditions of electrophoresis and autoradiographic exposure (channel D, Fig. 5).

Two approaches were utilized to investigate the nature of 5′-terminal structures on mRNAs in mouse eggs. The rationale for the first approach was based on the evidence that cap structures on mRNA molecules appear to augment efficiency of initiation in translation and the finding that cap analogues like m⁷GTP can reduce the translatability of capped mRNAs in vitro (Shatkin, 1976; Revel & Groner, 1978; Kozak, 1978). However, the degree of discrimination between translation of capped and uncapped mRNAs and the specificity of action of cap analogues for the inhibition of translation of capped messages is dependent upon a number of parameters including the choice of in vitro system (Lodish & Rose, 1977), the temperature and concentration of monovalent ions during translation (Kemper & Stolarsky, 1977; Weber et al. 1978; Chu & Rhoads, 1978), and sources of initiation factors (Held, West & Gallagher 1977; Bergmann et al. 1979). The nature of the specificity of inhibition of translation by the cap analogue m⁷GTP in the reticulocyte lysate system utilized in this study was assessed by comparing its effect on the translation of mRNAs from two plant viruses, TMV which has capped mRNA (Keith & Fraenkel-Conrat, 1975) and CPMV which has uncapped mRNA (Klootwijk et al. 1977). Both contain mRNA components of similar molecular weight (1·3 –2·2 × 10⁶ daltons) and incorporate similar amounts of [³⁵S]methionine into protein per μ1g mRNA under standard conditions in the message-dependent reticulocyte system (Table 1). The capped TMV RNA is markedly inhibited by m⁷GTP while the uncapped CPMV mRNA is largely unaffected over the range of cap analogue concentrations utilized. In addition, the importance of the methyl group in the m⁷GTP cap analogue was reconfirmed since GTP itself had no inhibitory effect on translation (Fig. 1, Table 1).

The translation of mRNAs from mouse eggs, like the endogenous mRNAs of non-nuclease treated reticulocyte lysate, is also reduced considerably in the presence of the cap analogue m⁷GTP. All the polypeptide translation products resolved in one-dimensional polyacrylamide gels were reduced in quantity suggesting that virtually all the translatable mRNA species are capped. Included within this set would be the ‘masked mRNAs’ which code for the characteristic 35000 dalton molecular weight 2-cell polypeptides described by Braude et al. (1979). Since most of the other polypeptides resolved on two-dimensional gels of egg mRNA translates are also present in two-dimensional electropherograms of polypeptides synthesized by intact eggs (Braude et al. 1979), it appears that nearly every template active mRNA (within the limits of sensitivity of detection in these experiments) in unfertilized eggs is capped. The fact that cap analogues do not reduce translation completely to background levels is a common observation in these types of experiments and likely reflects the fact that initiation may occur by mechanisms not involving the cap structure but at reduced efficiency (Kozak, 1978). The results of these experiments do not allow us to conclude that every mRNA molecule in the unfertilized egg is capped. Some messenger RNAs can be expected to be in too low a concentration to stimulate detectable amounts of protein synthesis. Also, if the egg contains some uncapped mRNAs which do not translate (or translate with low efficiency) in vitro, they would not be detected as being insensitive to inhibition by m⁷GTP. For this reason, another set of experiments which do not rely on the translatability of mRNA were undertaken to assess degree of capping.

Poly(A)-containing RNA (putative message) was purified subsequent to end-labelling procedures performed before and after removal of 5′-terminal cap structures with tobacco and phosphatase. The total RNA content of 500 mouse eggs is about 275 ng (Olds, Stern & Biggers, 1973). On the basis of poly (A) content of the mouse egg (Levey, Stull & Brinster 1978) and the assumption that the average mRNA molecule in mouse eggs is about 1500 nucleotides long (molecular weight of approximately 5 × 10⁵ daltons) of which about one tenth (150 nucleotides) is poly(A) tract, it can be calculated that 2 –4% of the total RNA is poly(A) + RNA or putative message. In absolute terms, this is 5 –10 ng (or on a molar basis, 10 –20 fmoles) of poly(A) + RNA per 500 eggs. We extracted this RNA in the presence of 5 μg of carrier tRNA from calf liver, which is in 500 to 1000-fold excess over that of the poly(A) + RNA. It is therefore reasonable to expect that something less than 0·5% of the resulting total RNA preparation is represented by poly(A)-containing RNA.

The specific activity of the γ-³²P-ATP used in end-labelling was 6·5 × 10³ cpm/fmole. Theoretically, if every 5′-end of poly(A) + RNA was labelled, approximately 6·5 –13·0 × 10⁴ c.p.m. of ³²P-label should be transferred to poly (A) + RNA molecules. In the case of the experiment summarized in Table 2, RNA was prepared from 616 eggs, and half was end-labelled after removal of cap structures and half in the absence. In the former case, all poly(A)-containing RNA molecules should have the potential to be end-labelled with γ-³²P-ATP and poly-nucleotide kinase while in the latter, only non-capped species should label. The RNA sample which was decapped with TAP (and represented RNA extracted from 308 eggs) contained 18·14 × 10³ c.p.m. of incorporated label in poly(A) + RNA molecules (Table 2). On the basis of the calculations above, as much as 40 to 80 × 10³ c.p.m. was theoretically possible. The reduced level obtained is not unreasonable, however, since recovery of RNA during extraction must be assumed to be less than 100% and since enzymatic decapping and endlabelling procedures do not go to total completion in the reaction times utilized even when enzymes and ATP substrate are provided in excess (Efstratiadis et al. 1977).

The finding that much more ³²P is incorporated into poly(A)-containing RNA during end-labelling procedures on molecules subjected to enzymatic decapping than in those which were sham treated (Tables 2 and 3) supports the previous results which demonstrated that the majority of translatable mRNAs in mouse eggs is capped. Further, there is no observable difference between unfertilized and fertilized egg RNA. The validity of these conclusions depends upon the fact that the enzymatic decapping and end-labelling procedures do not result in fragmentation of RNA molecules to yield other 5′-termini which can be labelled. It is imperative that the enzymes utilized are free of contaminating nucleases. The TAP and polynucleotide kinase preparations utilized here have been previously used to end-label μ1-globin (Efstratiadis et al. 1977) and protamine (Gedamu & Dixon, 1978) mRNA without degradation. In this study, mRNA components from TRV have been enzymically decapped and end-labelled without observable degradation when compared to starting material on polyacrylamide gels (Fig. 5). Because the TRV RNA molecules resolved are very large, generation of smaller molecules because of nuclease cleavage should have been readily detected as a number of labelled bands in the lower part of the gel. In addition, affinity chromatography on oligo-dT cellulose does not appear to lead to degradation since poly(A)-containing mRNA purified from mouse eggs by this procedure still has translational activity (Johnson, Schultz & Braude, unpublished results). Qualitative analysis of poly(A) + RNA molecules on polyacrylamide gels (Fig. 5) also reveals a marked increase in the intensity and number of RNA species end-labelled in the presence of TAP compared to those end-labelled in the absence of TAP treatment. These are clearly distinguishable from the background of control carrier tRNA preparations. Finally, if extensive fragmentation of RNA molecules occurred during the reactions, many additional 5′-termini available for end-labelling would be generated, a result which is not consistent with the amount of ³²P-incorporation we observed. That a small amount of RNA cleavage has occurred during end-labelling procedures cannot be ruled out. The consequence of such events would be the generation of 5′-termini which would behave as uncapped molecules. Therefore, the observation that about 80% of the poly(A) + RNA molecules are capped must be regarded as a minimal estimate and conceivably all of the molecules could be capped.

In summary, virtually all the mRNA molecules extracted from unfertilized mouse eggs that lead to detectable polypeptide products during in vitro translation are sensitive to inhibition by m⁷GTP. In experiments in which poly(A) + RNA (putative message) was analysed after end-labelling procedures with and without treatment with TAP, a minimum of 80% of the molecules were observed to contain cap structures in each case. Taken together, the results suggest that the majority of mRNAs in mouse eggs contain cap structures. These findings are not unexpected as caps provide protection from 5′-exonuclease degradation and appear to augment translational efficiency in eukaryotic cells. The mRNA of unfertilized sea-urchin eggs has also been shown to be predominantly capped (Hickey et al. 1976b). The methodology utilized in this study is not sufficiently sensitive to rule out the existence of some uncapped mRNA species in mouse eggs or minor alterations in capping which might accompany fertilization events, and may well be compatible with the small amount of [³H]guanosine incorporated into cap structures in newly fertilized mouse eggs observed by Young (1977). However, no quantitative data were presented in the latter study making it difficult to assess the significance of this observation in respect to what percentage of mRNA molecules might be affected. Our results indicate that a considerable proportion of mRNA molecules in the mouse egg contains cap structures and argue against addition of cap structures to pre-existent mRNA molecules of maternal origin as a major mechanism of post-transcriptional regulation in early post-fertilization events.

This work was supported by grants from the Ford Foundation and the Medical Research Council to Dr M. H. Johnson, Department of Anatomy, University of Cambridge, and in part by an operating grant from the Medical Research Council of Canada to G.A.S. The authors are grateful to all individuals who provided enzymes and viral mRNAs as gifts. Valuable discussion and helpful criticism from Dr P. R. Braude and Dr H. R. B. Pelham is also noted.