Abstract
Microinjection of oligodeoxynucleotides (ODNs) complementary to cellular mRNAs has been advanced as an experimental approach to degrade target mRNAs in vivo and thereby obtain information as to the function of their cognate proteins. It is shown here that ODNs can induce a variety of aberrations in cell metabolism and structure when injected into Xenopus oocytes. Examination of histological sections of ODN-injected oocytes revealed the frequent abnormal accumulation of heavily staining basophilic material in the area of the germinal vesicle (gv). Ultrastructural analysis detected further abnormalities including blebbing of the plasma membrane, anomalous cytoskeletal structures, hyperorganised annulate lamellae, hyperinvagination of the gv, and formation of irregular nucleoli within the gv. Analysis of newly synthesised proteins by [35S]methio-nine radiolabelling of oocytes demonstrated that ODN injection can trigger a general decrease in both label uptake and protein synthesis. Qualitative effects on protein synthesis could also be observed, particularly a decrease in synthesis of high molecular weight proteins. The severity of ODN-induced effects is dose-dependent and highly variable from ODN to ODN. The previously reported delay in progesterone-induced maturation observed in oocytes depleted of the maternal mRNA D7 by ODN-directed degradation (Smith R.C., Dworkin M.B. and Dworkin-Rastl E.(1988) Genes and Devpt. 2, 1296-1306) is most likely a result of nonspecific ODN effects in the oocyte. Oocytes injected with effective antisense D7 ODNs that do not display detectable side effects matured with normal kinetics.
Introduction
A major focus of modern cell and developmental biology is the characterization of protein function. With respect to this goal, some of the most promising new approaches have been provided by the development of so-called ‘pseudogenetic’ techniques (Izant and Weintraub, 1984; Melton, 1985; Haseloff and Gerlach, 1988). These approaches rely on the ability of antisense nucleic acids to hybridize to endogenous target sequences in living cells and thereupon obstruct translation (often by mediating cleavage of the mRNA) for the cognate protein. One variation of these techniques is to inject cells with short (10–30 nucleotides) pieces of DNA complementary to a particular target sequence (Markus-Sekura, 1988). Complexes between these antisense oligodeoxynucleotides (ODNs) and their target mRNA sequences are recognized by endogenous RNase H, which digests the mRNA, and thus prevents expression of the protein (Dash et al. 1987; Shuttleworth and Colman, 1988). Under conditions of low transcription and high protein turnover, ODN-directed destruction of specific mRNAs allows one to create cells that are depleted of a specific protein.
ODN-directed mRNA degradation is a relatively new technique; nevertheless, it has already been employed in a number of important studies (Zamecnik et al. 1986; Stephenson and Zamecnik, 1978; Sagata et al. 1988; Minshull et al. 1989; Kloc et al. 1989; O’Keefe et al. 1989). Many studies to date have utilized Xenopus oocytes, eggs and embryos to either document the effects of ODN-directed mRNA degradation or to study the role of particular proteins in these cells (Dash et al. 1987; Shuttleworth and Colman, 1988; Shuttleworth et al. 1988; Jessus et al. 1988; Sagata et al. 1988; Smith et al. 1988; Kloc et al. 1989; Woolf et al. 1990). Xenopus oocytes and eggs are advantageous for these studies because their large size facilitates injection of ODNs, which do not readily penetrate the cell plasma membrane, and because of the absence of transcription during early embryogenesis. ODN-induced destruction of mRNA provides a means to generate ‘mutant’ oocytes or eggs lacking a particular mRNA, which can then be fertilized in vitro (Holwill et al. 1987; Kloc et al. 1989), and analyzed with respect to phenotype. Any abnormalities occurring during subsequent development are presumably the result of the absence of the target protein. In this way, the function of that protein in embryogenesis can be inferred.
It is of fundamental importance in the ODN-directed mRNA degradation technique that any effects caused by a particular ODN be specific to the sequence of that ODN, and not artifacts of the technique itself. The studies mentioned above, including a study from one of our laboratories (Smith et al. 1988), incorporated a number of controls that supported the conclusion that the effects of ODN-directed mRNA degradation were indeed sequence-specific. In our own study, injection of ODNs complementary to the Xenopus maternal mRNA D7 was shown to lead to the specific destruction of the endogenous D7 mRNA. Progesterone-induced meiotic maturation of these oocytes was significantly delayed compared to control ODN-injected oocytes, suggesting that the D7 mRNA translation product was involved in oocyte maturation (Smith et al. 1988). However, in the course of an ultrastructural analysis of ODN-injected Xenopus oocytes and eggs, a number of abnormalities that occurred in oocytes injected with D7- and control-ODNs were discovered. This report describes these abnormalities and presents new data on the effect of ODN-directed D7 mRNA degradation on oocyte maturation. The results lead to a réévaluation of the previous D7 mRNA ablation studies and demonstrate that meiotic maturation is unaffected by the absence of D7 protein.
Materials and methods
Oligonucleotide synthesis
ODNs were synthesized with an Applied Biosystems DNA synthesizer and purified by reverse-phase high performance liquid chromatography (HPLC) or, for EBI 1507, using an Applied Biosystems cartridge (OPC). Several independent preparations of ODNs EBI 896 and D7/2 were used during the course of this study with identical results. ODN concentration was calculated on the basis of 1 OD260 = 37μgmrl.
Collection and injection of oocytes
Stage VI oocytes (Dumont, 1972) were isolated manually from ovaries of Xenopus laevis and maintained in lx modified amphibian Ringer’s solution (MR) (Vincent and Gerhart, 1987). Oocytes were injected with 20–60 nl ODN solution in water. Maturation was induced by incubation of oocytes in 1–2 μg ml-1 progesterone (Sigma) in MR and assayed by germinal vesicle breakdown, visible as white spot formation at the animal pole of the oocyte.
Histology
Oocytes were fixed in 100% methanol at 4°C overnight, paraffin-embedded (Paraplast, Monoject) and sectioned at 14 μm by standard procedures. Deparaffinated sections were stained with Giemsa stain in phosphate-buffered saline (PBS) at pH 7.2, and photographed under bright-field illumination.
Electron microscopy
Oocytes were fixed for electron microscopy, as previously described (Bement and Capeo, 1989), in s-collidine buffer (34mM s-collidine, 70mM KC1, 15mM NaCl, ImM CaCl2, pH 7.4) containing 2.5% glutaraldehyde. Samples were postfixed in 0.1 M sodium cacodylate buffer (pH7.9) containing 1 % osmium tetroxide. After rinsing in cacodylate buffer alone, samples were dehydrated in a graded ethanol series and the ethanol was then replaced with acetone. Samples were embedded in Spurr embedding medium (Polysciences) and ultrathin sections were cut on glass knives with a Porter-Blum ultramicrotome. Sections were viewed and photographed with a Philips EM 201 transmission electron microscope.
Analysis of protein synthesis
Proteins synthesized in oocytes were labeled by addition of [35S]methionine (>1000 Ci mmol-1; Amersham) to the incubation medium (0.29 μCi/tl-1) and incubated for 6 to 15 h. Individual oocytes were homogenized in 10 μl MR at 4°C, spun in a microfuge for 30 s and the supernatant used directly for gel analysis. Samples were analyzed by electrophoresis on standard 10 % SDS-polyacrylamide Laemmli gels (Laemmli, 1970). The gels were then treated for fluorography, dried and exposed to film at —80°C (Chamberlain, 1979). Total radiolabel taken up by an oocyte was estimated by counting aliquots of oocyte supernatants in Aquasol (Dupont). Incorporation of [35S]methionine into protein was determined by measuring the TCA-precipitable counts of aliquots of the oocyte supernatants by standard methods.
Immunoblotting
Groups of three stage VI oocytes or matured oocytes were homogenized in 25 μl of 50mM Tris-HCl pH7.5, 0.5 M urea, 2% Nonidet P-40, ImM phenylmethylsulfonylfluoride (PMSF), 5% 2-mercaptoethanol at 4°C, spun for 30s in a microfuge and the resulting supernatants analyzed by Western blotting with purified anti-D7 antibodies as described previously (Smith et al. 1988).
Results
The ODNs used for the injections into Xenopus oocytes described in this report are listed in Table 1. They range from 19-22 nucleotides in length. Two of the ODNs (EBI 896 and EBI 1507) are not known to have any complementarity to Xenopus laevis DNA. ODNs D7/1, D7/2, D7β, EBI 2712, EBI 2724 and EBI 2735 are complementary to the Xenopus maternal mRNA D7 (Smith et al. 1988). ODN EBI 1116 is complementary to the Xenopus maternal mRNA 207H1 (Dworkin et al. 1985; E.D-R. and M.D., unpublished observations). Since both D7 and 207H1 mRNAs are translationally repressed in the stage VI oocyte (Dworkin et al. 1985; Smith et al. 1988), destruction of these mRNAs in the oocyte would, therefore, be expected to be without consequence to the physiology of the oocyte. An analysis of the effects of injection of a number of these ODNs upon the integrity of endogenous oocyte mRNAs and the kinetics of progesterone-induced maturation has been described previously (Smith et al. 1988).
Analysis by light microscopy of the histological changes induced in stage VI oocytes by injection of ODNs
Histological sections of ODN-injected stage VI oocytes were prepared, stained with Giemsa stain and observed under the light microscope. Fig. 1 illustrates the effects of injection of ODNs D7/2, EBI 896 and EBI 1507 (40–130 ng/oocyte) in stage VI oocytes after overnight incubation. The effects observed ranged from the appearance of a disorganized basophilic area at a lateral position along the germinal vesicle (gv)-envelope (Fig. 1A) to a seeming invasion of strongly staining basophilic material into the inside of the gv (Fig. 1B,C), and finally to a stage where the area of the gv was completely taken up by this strongly staining basophilic material (Fig. 1D). This material was occasionally seen dispersed over the whole animal hemisphere (Fig. 1E). Control uninjected oocytes, in contrast, developed at most an evenly structured, basophilic area basal to the gv (Fig. 1F). Injection of high doses of ODNs sometimes caused some of the injected oocytes to develop a mottled surface pigmentation, depending on the batch of oocytes. Histological analysis of such oocytes revealed severe cytological abnormalities similar to those shown in Fig. 1D,E. Oocytes that exhibited this surface mottling were not included in the histological analyses presented except where indicated. In general, the oocytes with gross cytological abnormalities due to ODN injection described here had a perfectly normal external appearance and were outwardly indistinguishable from uninjected controls.
Histological analysis of ODN-injected oocytes. Sections were prepared as described in Materials and methods. Oocytes were incubated in MR overnight after ODN injection. (A) EB1 896, 130ng/oocyte; (B) EBI 1507, lOOng/oocyte; (C) D7/2, 40ng/oocyte; (D) D7/2, 130ng/oocyte; (E) D7/2, 130ng/oocyte; (F) uninjected. The right-hand side panels of A-F show a 4-fold higher magnification of part of the germinal vesicle. The diameter of a stage VI oocyte is about 1.2 mm.
Histological analysis of ODN-injected oocytes. Sections were prepared as described in Materials and methods. Oocytes were incubated in MR overnight after ODN injection. (A) EB1 896, 130ng/oocyte; (B) EBI 1507, lOOng/oocyte; (C) D7/2, 40ng/oocyte; (D) D7/2, 130ng/oocyte; (E) D7/2, 130ng/oocyte; (F) uninjected. The right-hand side panels of A-F show a 4-fold higher magnification of part of the germinal vesicle. The diameter of a stage VI oocyte is about 1.2 mm.
Table 2 summarises the proportion of oocytes that developed severe histological abnormalities (equivalent to those seen in Fig. 1 C-E) after injection with ODNs D7/2, EBI 896 and EBI 1507 at concentrations ranging from 40–170ng/oocyte. It is clear that despite the injection of constant doses of ODN per oocyte only a fraction of the oocytes developed severe abnormalities. The proportion of oocytes with abnormalities increased with increasing dose of injected ODN. It is striking that different ODNs differed dramatically in the severity of the effects they caused. ODNs EBI 916 and 1507 injected at high concentrations (>100 ng/oocyte) induced cytological effects in a significant proportion of oocytes, whereas no oocytes injected with EBI 896 at an equivalent concentration (if they had a normal outward appearance) displayed severe abnormalities. EBI 896, however, was not without effect in such oocytes and did induce less dramatic changes as shown in Fig. 1A when injected at high concentrations.
Analysis of the ultrastructural changes induced upon ODN injection in stage VI oocytes
An analysis of the ultrastructural changes that occur in ODN-injected oocytes confirmed and extended the results obtained with light microscopy. Electron microscopic analysis was conducted on oocytes injected with D7/2 or EBI 896 (80ng/oocyte) alone and, subsequently, on oocytes injected with D7/2 or EBI 896 and then treated with progesterone to induce meiotic maturation. ODN-injected oocytes exhibited numerous aberrations not observed in H2O-injected control oocytes in both the presence and absence of progesterone, although the aberrations in general occurred more rapidly in the presence of hormone. Within 3-8 h after injection, the plasma membrane took on a lobed appearance, with membrane-enclosed pieces of cytoplasm being apparent beneath the vitelline envelope (Fig. 2A). This apparent shedding of cytoplasm into the extracellular space was not observed in control oocytes (Fig. 2B). Next, several reorganisations became apparent in the cell interior. In the cytoplasm, highly elaborate arrays of cisternae were evident (Fig. 2C) and annulate lamellae became hyperorganised (Fig. 2D). The oocyte gv became hyperinvaginated such that long courses of cytoplasm penetrated the gv (Fig. 2E) and abnormal cytoskeletal structures were apparent in the vicinity of the invaginations (Fig. 2F). The nucleoli took on a fragmented appearance (Fig. 2G), which was not observed in control oocytes (Fig. 2H). In some ODN-injected oocytes, but not those injected with EBI 896, an extremely dense material sometimes accumulated in the area of the gv (Fig. 21,J). This material probably corresponds to the strongly staining basophilic material observed by light microscopy (Fig. 1).
Electron microscopie analysis of ODN-injected oocytes. Oocytes were injected with 80 ng of ODN and then incubated for the time indicated in either medium alone, or medium containing progesterone (1 μgml-1). (A) Cortex of an oocyte injected with EBI 896 and incubated in progesterone-containing medium for 3h. Lobes of cytoplasm (arrows) are apparent in the extracellular space. Y denotes yolk platelets. (B) Cortex of a control oocyte injected with H2O and incubated in progesterone-containing medium for 3h. No lobes are present. Y denotes yolk platelets. (C) Interior cytoplasm of an oocyte injected with D7/2 and incubated in progesterone-containing medium for 6h. An array of parallel cisternae is present (C). Y denotes yolk platelets. (D) Annulate lamellae (AL) in the interior cytoplasm of an oocyte injected with D7/2 and incubated in medium alone for 3h. The annulate lamellae appear hyperorganized because the pores (arrows) of adjacent lamellae are in perfect register. Y denotes yolk platelets. (E) The germinal vesicle (GV) of an oocyte injected with D7/2 and then incubated in progesterone-containing medium for 8h. Long invaginations of cytoplasm (C) which contain electron-dense material (ED) penetrate the nucleoplasm. (F) High magnification view of the germinal vesicle of an oocyte injected with D7/2 and then incubated in progesterone-containing medium for 3h. An abnormal parallel array of filaments (F) is present next to the germinal vesicle envelope (GV). (G) Low-magnification view of the germinal vesicle of an oocyte injected with D7/2 and then incubated in progesterone for 14 h. A fragmented nucleolus (Nu) is present in an area of nucleoplasm between two cytoplasmic invaginations (I). (H) Low-magnification view of a control oocyte injected with H2O and incubated in progesterone-containing medium for 3h. A normal, round nucleolus (Nu) is present in nucleoplasm (N). Note the absence of the extensive invaginations seen in experimental oocytes. M denotes mitochondria. (I) Low magnification view of the germinal vesicle area of an oocyte injected with D7/2 and then incubated in progesterone-containing medium for 5h. A large patch of electron-dense material (ED) is present. Y denotes yolk platelets. (J) Low magnification view of the germinal vesicle area of an oocyte injected with D7/2 and incubated in progesterone-containing medium for 3h. Mitochondria (M) are interspersed between clumps of electron-dense (ED) material. Y denotes yolk platelets. (K) Cortex of an oocyte injected with EBI 1116 and incubated in progesterone-containing medium for Ih. Blebs of cytoplasm (B) and yolk platelets (Y) are present outside the oocyte, trapped beneath the coelomic envelope (CE).
Electron microscopie analysis of ODN-injected oocytes. Oocytes were injected with 80 ng of ODN and then incubated for the time indicated in either medium alone, or medium containing progesterone (1 μgml-1). (A) Cortex of an oocyte injected with EBI 896 and incubated in progesterone-containing medium for 3h. Lobes of cytoplasm (arrows) are apparent in the extracellular space. Y denotes yolk platelets. (B) Cortex of a control oocyte injected with H2O and incubated in progesterone-containing medium for 3h. No lobes are present. Y denotes yolk platelets. (C) Interior cytoplasm of an oocyte injected with D7/2 and incubated in progesterone-containing medium for 6h. An array of parallel cisternae is present (C). Y denotes yolk platelets. (D) Annulate lamellae (AL) in the interior cytoplasm of an oocyte injected with D7/2 and incubated in medium alone for 3h. The annulate lamellae appear hyperorganized because the pores (arrows) of adjacent lamellae are in perfect register. Y denotes yolk platelets. (E) The germinal vesicle (GV) of an oocyte injected with D7/2 and then incubated in progesterone-containing medium for 8h. Long invaginations of cytoplasm (C) which contain electron-dense material (ED) penetrate the nucleoplasm. (F) High magnification view of the germinal vesicle of an oocyte injected with D7/2 and then incubated in progesterone-containing medium for 3h. An abnormal parallel array of filaments (F) is present next to the germinal vesicle envelope (GV). (G) Low-magnification view of the germinal vesicle of an oocyte injected with D7/2 and then incubated in progesterone for 14 h. A fragmented nucleolus (Nu) is present in an area of nucleoplasm between two cytoplasmic invaginations (I). (H) Low-magnification view of a control oocyte injected with H2O and incubated in progesterone-containing medium for 3h. A normal, round nucleolus (Nu) is present in nucleoplasm (N). Note the absence of the extensive invaginations seen in experimental oocytes. M denotes mitochondria. (I) Low magnification view of the germinal vesicle area of an oocyte injected with D7/2 and then incubated in progesterone-containing medium for 5h. A large patch of electron-dense material (ED) is present. Y denotes yolk platelets. (J) Low magnification view of the germinal vesicle area of an oocyte injected with D7/2 and incubated in progesterone-containing medium for 3h. Mitochondria (M) are interspersed between clumps of electron-dense (ED) material. Y denotes yolk platelets. (K) Cortex of an oocyte injected with EBI 1116 and incubated in progesterone-containing medium for Ih. Blebs of cytoplasm (B) and yolk platelets (Y) are present outside the oocyte, trapped beneath the coelomic envelope (CE).
Oocytes injected with ODNs EBI 1116 and 1507 exhibited the same ultrastructural abnormalities observed in D7/2 or EBI 896-injected oocytes (Fig. 2K and data not shown). Although the relative frequency of oocytes displaying the observed abnormalities with the different ODNs was not assessed in this ultrastructual study, it was noticed that higher doses of ODNs (120 ng/oocyte) induced both a greater frequency and a more rapid onset of the various abnormalities than lower doses of ODNs (80ng/oocyte).
Effects of ODN injection upon oocyte protein synthesis
The form and progression of ultrastructural reorganization observed in response to ODN injection bears a striking similarity to pseudomaturation, a phenomenon described by Steinert et al. (1974). Since pseudomatu ration is reported to be associated with a general decrease in Xenopus oocyte protein synthesis (Baltus et al. 1973), it was appropriate to monitor protein synthesis in ODN-injected oocytes. The patterns of proteins synthesized in ODN-injected oocytes, labelled by incubation in [35S]methionine-containing medium, and analyzed by gel electrophoresis are shown in Fig. 3. It is clear that certain ODNs injected at 110–130 ng/oocyte can cause a dramatic decrease in the amount of incorporation of label into proteins, but the effects vary greatly from ODN to ODN (Fig. 3A,B). ODN EBI 896 elicited only minor observable effects; both quantitatively and qualitatively protein synthesis profiles derived from EBI 896-injected oocytes were similar to those from uninjected control oocytes, except for a slight, but reproducible, decrease in incorporation of label into high molecular weight proteins. This decrease in the synthesis of large proteins was characteristic of protein synthesis profiles of oocytes injected with a number of individual ODNs. ODNs D7/1, D7/2 and EBI 1507 showed the most dramatic decrease in overall signal (Fig. 3A,B). Longer exposure of the gels revealed that these ODNs, in addition, caused differences in the synthesis pattern of some individual oocyte proteins. The effects of ODN injection on oocyte protein synthesis profiles were highly reproducible from oocyte to oocyte within any particular batch. The exact degree of inhibition of incorporation, however, varied between different batches of oocytes. Nevertheless, the relative inhibition of the different ODNs with respect to each other remained constant within the following hierarchy: D7/2, D7/1, EBI 1507>EBl 1116, D7β>EBI 896. The protein synthesis profiles of oocytes injected with a selection of the ODNs at lower ODN concentrations (40ng/oocyte) are shown in Fig. 3C. The ODN-mediated effects are much reduced compared to the effects of higher ODN doses, but there still remain quantitative and qualitative effects, again with individual ODNs differing with respect to the extent of the effects they cause.
Protein synthesis profile of ODN-injected oocytes labeled by incubation in [35S]methionine. (A) 110 ng ODN/oocyte. Oocytes were incubated for 3h postinjection prior to labeling for 6h. (B) 130ng ODN/oocyte. Oocytes were incubated for 3h postinjection prior to labeling for 13 h. (C) 40ng ODN/oocyte. Injected oocytes were incubated for 25 h postinjection prior to labeling for 6h. Experiments A–C used different batches of oocytes. Each lane represents the protein synthesis profile of individual oocytes (0.3 oocyte equivalent/lane). In A and B, longer exposures of certain panels (as indicated) are shown at the right-hand side.
Protein synthesis profile of ODN-injected oocytes labeled by incubation in [35S]methionine. (A) 110 ng ODN/oocyte. Oocytes were incubated for 3h postinjection prior to labeling for 6h. (B) 130ng ODN/oocyte. Oocytes were incubated for 3h postinjection prior to labeling for 13 h. (C) 40ng ODN/oocyte. Injected oocytes were incubated for 25 h postinjection prior to labeling for 6h. Experiments A–C used different batches of oocytes. Each lane represents the protein synthesis profile of individual oocytes (0.3 oocyte equivalent/lane). In A and B, longer exposures of certain panels (as indicated) are shown at the right-hand side.
The reduced autoradiographic signal in the protein synthesis profiles of ODN-injected oocytes illustrated in Fig. 3 results from a combination of two effects: an inhibition of uptake of [35S]methionine into the oocyte as well as an inhibition of protein synthesis itself. Injection of ODNs D7/2 and EBI 1507 decreased label uptake by oocytes from the surrounding medium to approximately 25% of that of uninjected oocytes, whereas injection of ODN EBI 896 had no effect on this parameter (Fig. 4A). Direct effects of ODN injection upon oocyte protein synthesis were obtained by comparing the ratio of TCA-precipitable counts to total counts in ODN-injected and control oocytes. Injection of ODNs D7/2 and EBI 1507 caused a decrease in total oocyte protein synthesis to 38 % and 64 % of control values, respectively. EBI 896 was without effect (Fig. 4B).
Effect of ODN injection on oocyte methionine uptake and protein synthesis. (A) Total uptake of pSjmethionine into ODN-injected oocytes expressed as % uptake of control (uninjected) oocytes. Each point represents the average of duplicate measurements taken from 3-4 oocytes. (B) Ratio of [35S]methionine incorporated into TCA-precipitable counts to the total [35S]methionine taken up by ODN-injected oocytes expressed as % of the value obtained from control (uninjectcd) oocytes. Each value represents the average of duplicate measurements taken from 2–4 oocytes. The same experimental oocyte supernatants as displayed in Fig. 3A were used for both analyses.
Effect of ODN injection on oocyte methionine uptake and protein synthesis. (A) Total uptake of pSjmethionine into ODN-injected oocytes expressed as % uptake of control (uninjected) oocytes. Each point represents the average of duplicate measurements taken from 3-4 oocytes. (B) Ratio of [35S]methionine incorporated into TCA-precipitable counts to the total [35S]methionine taken up by ODN-injected oocytes expressed as % of the value obtained from control (uninjectcd) oocytes. Each value represents the average of duplicate measurements taken from 2–4 oocytes. The same experimental oocyte supernatants as displayed in Fig. 3A were used for both analyses.
To ensure that the observed effects of ODN injection were not caused by impurities in the ODN preparation, three different methods of ODN purification were employed for the protein synthesis studies. Injection of ODNs purified by gel elution, HPLC, or oligonucleotide purification cartridge (OPC) gave identical effects on protein synthesis (data not shown). Further, analysis by end-labeling of ODNs used in these studies demonstrated that the vast proportion of each ODN preparation was full length. This, in addition to the fact that injection of deoxynucleosidemonophosphates (dAMP, dCMP, dGMP or TMP, each tested separately at 110 ng/oocyte) had no effect on protein synthesis (data not shown), indicated that the generation of ODN breakdown products is unlikely to be responsible for ODN-induced protein synthesis inhibition.
Effects of ODN injection on oocyte maturation
Since some experimental designs require in vitro fertilization and production of embryos from ODN-injected oocytes, the effects on oocyte maturation of the metabolic and cytological changes induced by ODN injection described above are of interest. The results of an analysis of the kinetics of progesterone-induced maturation of ODN-injected oocytes are summarized in Table 3. Whereas low doses (40 ng/oocyte) of ODNs EBI 896 and D7/2 did not delay oocyte maturation substantially, at 130 ng/oocyte a pronounced delay of maturation (as assayed by white spot formation at the animal pole) was observed for D7/2, but not for EBI 896. This is in agreement with previous data (Smith et al. 1988). EBI 1507 delayed maturation even at a low ODN dose. Light microscopic as well as ultrastructural analysis of ODN-injected maturing oocytes confirmed that these oocytes underwent a similar progression of cytological abnormalities as seen in ODN-injected oocytes not treated with hormone (see above and data not shown). Furthermore, gel analysis of the protein synthesis profiles of these progesterone-treated oocytes showed inhibitory effects similar to those of ODN-injected oocytes without hormone treatment. The degree of inhibition appeared to correlate with the extent of maturation delay exhibited by a population of oocytes (data not shown).
Effects of selected D7 ODNs upon oocyte maturation
The above results make it likely that the observed delay in maturation of oocytes depleted of D7 mRNA by ODN injection is a consequence of nonspecific ODN effects in the oocyte. This puts into question the previous interpretation of the maturation delay as suggesting a role for the D7 gene product in oocyte maturation (Smith et al. 1988). In order to evaluate further the role of D7 in the maturation process, antisense D7 ODNs that displayed no apparent nonspecific effects in the oocyte were identified. Three new D7 ODNs (EBI 2712, 2724 and 2735), when injected at 110 ng/oocyte, did not inhibit oocyte protein synthesis (Fig. 5A). The result of Western blotting with anti-D7 antibodies of extracts derived from these ODN-injected oocytes after hormone-induced maturation is shown in Fig. 5B. D7 protein was absent from stage VI oocytes but accumulated during oocyte maturation in control uninjected oocytes, in accordance with previous data (Smith et al. 1988). Oocytes matured after injection with any of the three new D7 ODNs did not accumulate detectable D7 protein. Amido black staining of the blot confirmed that equal amounts of protein were loaded per lane (data not shown). Despite the lack of accumulation of D7 protein, these oocytes matured with normal kinetics upon treatment with progesterone (Fig. 5C).
Effects of D7 ODNs EBI 2712, 2724 and 2735 on oocyte protein synthesis, accumulation of D7 protein during oocyte maturation and kinetics of maturation. (A) Protein synthesis profile of D7 ODN-injected oocytes labeled by incubation with [î5S]methionine. Oocytes were each injected with 110 ng ODN and then incubated for 5h 15 min prior to labeling for 14 h 35 min. Each lane represents the protein synthesis profile of individual oocytes (0.3 oocyte equivalent/lane). (B) Western blot analysis with anti-D7 antibodies of oocytes injected with D7 ODNs and subsequently matured in vitro. Oocytes were injected with D7 ODNs (HOng/oocyte), incubated for 3h in MR and then induced to mature with progesterone. Oocytes were taken for analysis 19 h after hormone addition at which time all hormone-treated oocytes were mature. Egg=uninjected mature oocyte, st 6=uninjected stage VI oocyte. (C) Maturation kinetics of oocytes injected with D7 ODNs. Oocytes were injected (110 ng ODN/oocyte) and incubated in MR for 3h prior to induction of maturation with progesterone. 58 uninjected and 10–14 ODN-injected oocytes were assayed for each maturation determination.
Effects of D7 ODNs EBI 2712, 2724 and 2735 on oocyte protein synthesis, accumulation of D7 protein during oocyte maturation and kinetics of maturation. (A) Protein synthesis profile of D7 ODN-injected oocytes labeled by incubation with [î5S]methionine. Oocytes were each injected with 110 ng ODN and then incubated for 5h 15 min prior to labeling for 14 h 35 min. Each lane represents the protein synthesis profile of individual oocytes (0.3 oocyte equivalent/lane). (B) Western blot analysis with anti-D7 antibodies of oocytes injected with D7 ODNs and subsequently matured in vitro. Oocytes were injected with D7 ODNs (HOng/oocyte), incubated for 3h in MR and then induced to mature with progesterone. Oocytes were taken for analysis 19 h after hormone addition at which time all hormone-treated oocytes were mature. Egg=uninjected mature oocyte, st 6=uninjected stage VI oocyte. (C) Maturation kinetics of oocytes injected with D7 ODNs. Oocytes were injected (110 ng ODN/oocyte) and incubated in MR for 3h prior to induction of maturation with progesterone. 58 uninjected and 10–14 ODN-injected oocytes were assayed for each maturation determination.
Discussion
The absence of transcription until the mid-blastula stage of embryogenesis, the ease of injection into oocytes, eggs and embryos, and the fact that several proteins are known to be expressed during well-defined temporal windows during meiotic maturation and early embryogenesis, makes the Xenopus system attractive for ODN-directed destruction of mRNA. The technique has been used to study the roles of c-mos (Sagata et al. 1988), tubulin (Jessus et al. 1988) and D7 (Smith et al. 1988) in meiotic maturation and of xlgv7 in embryogenesis in Xenopus (Kloc et al. 1989). ODN-directed mRNA destruction has also been employed to study the role of c-mos in meiotic maturation of mouse oocytes (O’Keefe et al. 1989).
In the majority of these studies, the specificity of ODN action was assessed by examination of target mRNA and control mRNA integrity, by analysis of the quantity of target protein recognized by immunological techniques, or by both (Jessus et al. 1988; Sagata et al. 1988; Smith et al. 1988; O’Keefe et al. 1989). Experiments were also controlled by injection of sense ODNs or other ODNs with no apparent complementarity to cellular mRNAs (Sagata et al. 1988; Smith et al. 1988). In a small number of these studies, total protein synthesis was also examined. Jessus et al. (1988) noted that both control and experimental ODN injection into Xenopus oocytes resulted in a decrease in synthesis of high molecular weight proteins. Minshull et al. (1989), using Xenopus egg extracts, noted that both control and experimental ODNs reduced overall protein synthesis in Xenopus egg extracts by 14–37 %. Akagi et al. (1989), when injecting 30-mer ODNs into Xenopus oocytes, noted that doses of higher than 2.5 ng/oocyte had nonspecific effects on protein synthesis. O’Keefe et al. (1989) observed that protein synthesis in mouse oocytes injected with control and experimental ODNs was only 80% of that in uninjected oocytes.
Our results demonstrate that ODNs can indeed cause nonspecific effects, including inhibition of protein synthesis. The results were obtained regardless of the method of ODN purification, and do not appear to be a result of contamination with (or generation of) ODN degradation products. Different ODNs varied enormously in the severity of the effects they elicited upon injection into oocytes. The active component within the ODN responsible for the observed effects is completely unknown. There is no obvious correlation with primary sequence or with base composition of the ODN. The quantities of ODNs employed in this study were within the range employed in other reports (Sagata et al. 1988; Shuttleworth and Colman, 1988; Shuttleworth et al. 1988; Kloc et al. 1989), and the results of Agaki et al. (1989) suggest that even much lower concentrations of ODNs can still inhibit protein synthesis.
ODNs, in addition to their effect on protein synthesis, were also found to cause dramatic changes to oocyte morphology. ODNs cause shedding of membrane-enclosed cytoplasmic bodies into the extracellular medium, cytoskeletal reorganizations, hyperinvagination and disruption of the nuclear envelope, hyperorganization of the annulate lamellae, and accumulation of densely staining basophilic material in the region of the gv. The ultrastructural effects of ODN injection do not resemble the normal ultrastructural course of progesterone-induced meiotic maturation (Bement and Capco, 1989, 1990) but they are similar to those previously described to occur in oocytes that have undergone pseudomaturation (Baltus et al. 1973; Steinert et al. 1974). They resemble in many ways the cytological features of cells undergoing apoptosis, or programmed cell death (Wyllie et al. 1980). This, in addition to the observation that occasionally mottled oocytes found in a control population of oocytes exhibited cytological abnormalities similar to those induced by ODN injection (E.D-R. and R.S., unpublished observations), leads to the possibility that ODNs are triggering cell death. It is unlikely that inhibition of protein synthesis is the sole explanation for the cytological effects observed upon ODN injection since treatment of oocytes even with high doses (100 μg ml-1) of cycloheximide does not lead to the development of these cytological abnormalities, as detectable using light microscopy (E.D-R. and R.S., unpublished observations).
Populations of ODN-injected oocytes displaying reduced protein synthesis and cytological abnormalities showed delayed maturation kinetics. Since it is known that inhibition of protein synthesis by cycloheximide treatment can inhibit progesterone-induced maturation at high doses and delay maturation at lower doses (Wasserman and Masui, 1975; R.S., unpublished observations), it is possible that the delay in maturation observed in ODN-injected oocytes is a result of the ODN-induced inhibition of protein synthesis. It is, however, also expected that the cytological abnormalities seen in the ODN-injected oocytes would interfere with the maturation process. This has been confirmed in a preliminary analysis of mature (according to white spot formation) oocytes that had been injected with 40 ng EBI 1507 prior to progesterone treatment. A high percentage of these oocytes were unable to undergo an activation response when challenged with the Ca2+ ionophore A23187, and these, oocytes exhibited gross cytological abnormalities, in particular a system of extensive interconnected yolk-free areas in the animal hemisphere (data not shown).
The delay in oocyte maturation caused by injection of D7 ODNs was previously interpreted as suggesting a role for D7 in oocyte maturation (Smith et al. 1988). However, it is apparent from the results reported here that the D7 ODNs most effective at delaying maturation are, at similar concentrations, also those producing the most severe abnormalities in the oocyte. The fortuitous use of particular control ODNs in earlier experiments that cause only minor abnormalities and, therefore, do not affect maturation kinetics, was unfortunate. A search for new D7 ODNs that do not produce obvious toxic effects (as assayed by effects on oocyte protein synthesis patterns) led to the identification of three effective D7 ODNs. Injection of these ODNs into oocytes prevented the accumulation of D7 protein during hormone-induced oocyte maturation. Furthermore, such oocytes were found to mature with normal kinetics demonstrating conclusively that the previously reported delay in oocyte maturation caused by injection of D7 ODNs was due to nonspecific ODN effects.
Clearly, careful selection of ODNs for experiments utilizing ODN-mediated RNA destruction is recommended. For example, only ODNs with negligible effects on patterns of protein synthesis should be used. To reduce the risk of non-specific effects the minimum effective ODN dose should be employed. Ideally any phenotype obtained with this technique should be supported independently, or if possible rescued by injection of the targeted mRNA. In particular, phenotypes that could be achieved by a general reduction in protein synthesis should be viewed with caution. It is encouraging that Kloc et al. (1990) have recently reported the successful production of normal embryos derived from ODN-injected oocytes.
Acknowledgements
We wish to acknowledge Rudi Hauptmann and Hermann Mahr for synthesis of oligodeoxyribonucleotides and Brenda Demeties for excellent technical assistance. This work was supported in part by NIH grants HD 23686 and HD 00598 to D.G.C.