| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Laboratory of Molecular Signaling, The Babraham Institute, Cambridge CB2 4AT, United Kingdom
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
gamete biology, gene regulation, meiosis, oocyte development
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
At least two conserved elements in the 3'-UTR, namely the hexanucleotide polyadenylation signal (AAUAAA) and one or more cytoplasmic polyadenylation elements (CPEs; U4–6 A1–2 U consensus sequence), are essential for translational control of stored transcripts in Xenopus and mouse oocytes [4, 5]. MOS is expressed almost exclusively in vertebrate germ cells [6] and is deadenylated and stored in a repressed form in immature oocytes of all vertebrates [2]. In Xenopus oocytes, translation of mos mRNA following polyadenylation is an early event during maturation [7]. In contrast, the polyadenylation of MOS in mouse oocytes appears after oocytes enter metaphase I [5]. Our report focuses on the means by which individual masked mRNAs in a high mammal are selected and translated at precisely defined stages during maturation.
Some masked mRNAs in pig oocytes exhibit a range of unique features in their 3'-UTRs. These characteristics are clearly seen in pig MOS mRNA [8]. First, the MOS 3'-UTRs in pig oocytes are heterogeneous at the 3'-end arising from alternative polyadenylation sites. Not only does the 3'-UTR in MOS mRNA in pig oocytes exist in three discrete lengths of 180, 480, and 530 nucleotides (nt), but none of the variants have CPE consensus sequences that exactly match those found in either the Xenopus or mouse MOS 3'-UTR. Third, a novel A-rich sequence precedes the sole nuclear polyadenylation signal sequence in the 3'-UTR. That this distinctive A-rich motif may be associated more widely with other masked mRNAs in pig oocytes is suggested by parallel studies in which comparisons have been made between the 3'-UTRs in CDC25 mRNA derived from somatic versus those from oocyte-enriched libraries [9]. Those studies showed that CDC25 transcripts from pig oocytes contain 3'-UTRs that are longer than their somatic counterparts and contain a similar A-rich motif (A5UA2(CA2)4A). These unique features of the 3'-UTRs in pig oocytes have led us to question whether translational control in this species is similar to, or more elaborate than, that in amphibians and mice.
We have used pig MOS mRNA to test the hypothesis that specialized features of the 3'-UTR in some stored transcripts in higher mammals may provide added complexity to the control of the translation process. Strategies involving mutational analysis and oocyte microinjection were used in the present study for four interrelated purposes. The first object was to test the possibility that the length of the 3'-UTR in porcine MOS transcripts influences both the timing and the degree of MOS mRNA mobilization. The second objective was to identify the extent to which the function of the hexanucleotide signal and U-rich elements differ in the three different-length MOS transcripts in pig oocytes. The third objective was to identify the role of the A-rich sequences in translational control. The final objective was to test the hypothesis that the effects on translation of the overall length of the 3'-UTR, together with that of the A-rich and U-rich elements, all act by influencing the polyadenylated status of the MOS mRNA.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The polymerase chain reaction (PCR) technique has been used to generate a variety of deletions and mutations in MOS 3'-UTRs by using the corresponding cDNA as template DNA [8]. pBluescript SK (Stratagen, UK) vector was used as a building vector to assemble MOS 3'-UTR variants, and the BamHI and XhoI sites were used as the cloning site. All the fusion genes were made by inserting a MOS 3'-UTR variant with or without mutations into StuI and XhoI sites of pGEM-luc vectors (Promega, UK) and were confirmed by sequencing analysis.
To construct the luciferase and wild-type MOS 3'-UTR fusion genes, three different MOS 3'-UTRs were generated by PCR using a pair of primers (forward, 5'-ggaattcgtgtggccggataagca-3'; reverse, 5'-taatacgactcactataggg-3'). The reverse primer is located in the cDNA cloning vector 3' polylinker site [8].
To delete the A-rich sequence (CA13CA5CA5CA6) in the MOS 3'-UTRs, two PCR amplifications were carried out separately using two pairs of primers (pair 1: forward, 5'-ggaattcgtgtggccggataagca-3'; reverse, 5'-ggggtacctacatcttctcttgtt-3'; pair 2: forward, 5'-gggggtaccatagaggtaggatgga-3'; reverse, 5'-taatacgactcactataggg-3'). The PCR products were cut by KpnI-BamHI and KpnI-XhoI separately, and then the two fragments were ligated together and cloned into BamHI and XhoI sites of pBluescript vector.
To make hexanuleotide polyadenylation signal mutations (AAUAAA to AACUAA) in the MOS 3'-UTRs, two PCR amplifications were carried out separately using two pairs of primers (pair 1: forward, 5'-ggaattcgtgtggccggataagca-3'; reverse, 5'-gggtacctctatgttttttgtttttgtttttgtttttttttttttgc-3'; pair 2: forward, 5'-ggggtacctaggatggatttttagaaactaaagttacc-3'; reverse, 5'-taatacgactcactataggg-3'). Before these two PCR fragments were ligated together, KpnI was used to cut two fragments, followed by Klenow treatment to blunt the ends of the PCR fragment. Then, another PCR reaction was performed by using the above ligation products as template DNA and the following oligonucleotides as primers (forward, 5'-ggaattcgtgtggccggataagca-3'; reverse, 5'-taatacgactcactataggg-3').
To duplicate the A-rich sequence (CA13CA5CA5CA6) in MOS 3'-UTRs, two PCR amplifications were carried out separately using two pairs of primers (pair 1: forward, 5'-ggaattcgtgtggccggataagca-3'; reverse, 5'-gggtacctctatgttttttgtttttgtttttgtttttttttttttgc-3'; pair 2: forward, 5'-ggggtacctaggatggatttttagaaataaaagttacc-3'; reverse, 5'-taatacgactcactataggg-3'). After completion of the PCR reaction, two PCR products were cut by BamHI-KpnI and XhoI-KpnI, respectively. Then, the two fragments were ligated together and cloned into XhoI and BamHI sites of pGEM-7zf vector. After the recombinants were cut by KpnI and then blunt-ended by Klenow, the newly synthesized A-rich region (CA13CA5CA5CA6C) was cloned into the inactivated KpnI site.
To relocate the A-rich sequence (CA13CA5CA5CA6) to the HindIII site (Fig. 1) in the two long-form MOS 3'-UTRs (480 and 530 nt), a synthesized CA13CA5CA5CA6 DNA fragment was cloned into the HindIII site of the MOS 3'-UTR variant devoid of the A-rich region.
|
Two PCR reactions were used to mutate the U-rich sequences from ATTTTTA to ATTTGCA in MOS 3'-UTRs. To generate mutation in the second U-rich sequence 5' near to the hexanucleotide signal, a PCR reaction was carried out using a pair of primers (forward, 5'-gttcatttgctaacaagagaagatg-3'; reverse, 5'-taatacgactcactataggg-3' [T7 primer]). Then, another PCR reaction was performed using 5'-ggaattcgtgtggccggataagcaatttgtctttgttgctgttcatttgcaacaagagaag-3' as forward primer and the PCR products from the above PCR reaction as reverse primer to mutate the first U-rich sequence upstream of the A-rich sequence in the MOS 3'-UTR.
The control construct was the same as the experimental constructs except in the 3'-UTR. The 3'-UTR in control construct consisted of a short linker sequence following by a 30-adenine poly(A) tail. The linker sequence was as follows: taaaatgtaactgtattcagcgatgacgaaattcttagctattgtaatcctccgaggcct.
RNA Preparation and Translation Assay
All plasmid DNAs were prepared by using Wizard Maxipreps kits (Promega) and linearized by XhoI digestion. Chimeric RNA was synthesized in vitro by using Sp6 RNA polymerase (Roche, UK) as described by Krieg and Melton [10] and prepared for injection as described previously [9].
Extractions of RNA were performed as described previously [8]. For translation assays, luciferase protein assay kits (Promega) were used according to the manufacturer's instructions. For these assays, the injected and cultured oocytes were collected in groups of 20 and lysed in 20 µl of lysis buffer by freezing and thawing twice with accompanying pipetting.
Poly(A) Addition Analysis
The length of the poly(A) tail of MOS mRNA was measured using the poly(A) test developed by Salles and Strickland [11] with some modification. Total RNA extracted from a group of 30 oocytes was annealed with 25 ng of oligo(dT17) anchor (5' gactcgagtcgacatcga(T)17) in reverse-transcription reaction buffer at 42°C for 15 min. Then, ethanol was added to precipitate out RNA at –80°C overnight. The following day, the total RNA was collected by centrifugation and resuspended in 10 µl of water. After reverse transcriptions, the PCR amplifications were carried out using cDNA (equal to the total cDNA of one oocyte) as DNA template and the following primers (luciferase gene specific primer [forward], 5'-gccaagaagggcggaaagtcc-3'; an adapter oligonucleotide [reverse], 5'-gactcgagtcgacatcga-3') as PCR primers. The PCR conditions were 95°C for 5 min and then 20 cycles of 94°C for 15 sec, 55°C for 20 sec, and 72°C for 1 min followed by 7 min extension at 72°C. After amplification, 5 µl of the 50-µl reaction volume were electrophoresed on an agarose gel; thereafter, a Southern blot was performed using the 180-base pair fragment of mos 3'-UTR as probe.
Oocyte Preparation and Microinjection
Fully grown, G2-phase oocytes were aspirated from nonatretic follicles (>3 mm diameter) of pig ovaries obtained from a local abattoir and then cultured as described previously [12]. In the first experiment, aspirated oocytes were injected with chimeric mRNA constructs directly after aspiration from the ovaries (G2-phase injections). Injected oocytes in experiment 1 were then cultured for 18 h, 24 h (metaphase I), 48 h (metaphase II), or 72 h before being analyzed for luciferase expression. In some experiments, the culture of oocytes randomly allocated to the 72-h groups was briefly interrupted at 48-h to enable the cells to be electroactivated using methods described previously [13]. In experiment 2, cumulus-oocyte complexes were cultured for either 1 h (G2 phase), 24 h (metaphase 1), or 48 h (metaphase II) before injection with one of a variety of luciferase-MOS chimeric mRNA. Thereafter, injected oocytes were returned to culture for a further 24-h posttreatment period before analysis for luciferase expression.
Immediately before microinjection, oocytes were transferred to Hepes-buffered M199 medium to maintain a stable pH environment during manipulation. The oocyte preparations and injections followed our standard method [9]. Each experiment consisted of a set of different treatment groups (20 oocytes/group) comprising a control group (injected with luciferase mRNA), a group injected with luciferase hybrid mRNA and the wild-type 3'-UTR, and groups injected with luciferase hybrid mRNA and a mutant 3'-UTR. Each experiment was carried out on a single batch of oocytes injected, whenever possible, with the same micropipette to reduce intrareplicate variability. Each experiment was performed on five or more separate occasions. After injection, oocytes were washed four times in dissection medium before being replaced in culture. Freshly prepared pyruvate (40 µg/ml) was added to the oocytes in the postinjection period to compensate for any reduction in the gap junction-mediated entry of energy substrates because of the removal of follicle cells before microinjection.
Statistical Analysis
Subsets of oocytes from each treatment group were harvested at 24, 48, and 72 h after the initiation of culture to determine the level to which injected mRNA had been translated at each of these time points. Systematic differences between treatments in each replicate experiment were detected by using analysis of variance, and the statistical package GENSTAT was used to carry out additional analyses. When the overall analysis detected systematic group differences, a Student t-test was used for individual comparisons.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
When a luciferase RNA, linked directly by a short linker to a 30-adenine poly(A) tail (Figs. 1 and 2A) was injected into the cytoplasm of pig oocytes at the reinitiation of meiosis (0 h), uniformly low levels of luciferase activity (<2.5 optical density [OD] units) were detected throughout the 72-h experimental period (Fig. 2B). We next tested whether the addition of a pig MOS 3'-UTR to luciferase mRNA would alter the pattern of translation away from the control baseline levels and toward that observed for MOS translation in vivo. In addition, we also determined whether the three MOS 3'-UTR variants (180, 480, or 530 nt) differentially affected either the level of translation or the precise timing at which MOS translation is initiated or terminated. The results, which are summarized in Figure 2B, show that at prometaphase (18-h postinjection), only the 180-nt variant of the chimeric RNA is translated at a slightly, but not statistically significantly, higher level than that of the controls. At metaphase I (24-h postinjection), translation of luciferase-MOS transcripts had increased significantly in the short- and medium-form transcripts but not in the long-form chimeric transcripts (P < 0.05 for control vs. 180- or 480-nt chimeric variants). At metaphase II (48-h postinjection), luciferase levels in oocytes injected with luciferase-MOS chimeric mRNA were, in all cases, highly significantly different from that of oocytes injected with luciferase control mRNA (P < 0.001 for control vs. 180- or 480-nt chimeric variants, P < 0.01 for control vs. 530-nt chimeric variant). It was further evident that an inverse relationship exists between the length of the MOS 3'-UTR and the level of translation of the associated reporter mRNA (P < 0.05 for 180 vs. 480 or 530 nt). Luciferase activity in nonactivated oocytes at 72 h (late-metaphase II groups) was slightly higher than that at 48 h for each group of oocytes injected with chimeric luciferase-MOS mRNA, whereas in the control group, luciferase activity remained unchanged at basal levels. Activation of eggs at 48 h reduced, but did not eliminate, luciferase activity in the ensuing 24 h in all oocytes injected with chimeric mRNA.
|
To verify further the results outlined in Figure 2B, a second study was carried out using the same constructs but a different injection protocol. In these experiments, oocytes were injected at different stages during meiosis, namely at the resumption of maturation (0 h), at metaphase I (24 h), or at metaphase II (48 h), and then harvested 24 h thereafter. The overall pattern of translation is shown in Figure 2C. Compared with the control group, statistical analyses indicated that the presence of a MOS 3'-UTR significantly increased luciferase expression in oocytes injected at the resumption of meiosis and examined 24 h thereafter (P < 0.05 except for the mos 530-nt 3'-UTR group). Luciferase expression in oocytes injected at metaphase I or metaphase II with luciferase-MOS chimeric mRNA was, in all cases, significantly different from that of controls (P < 0.01). Luciferase activity from oocytes injected with luciferase MOS 180-nt 3'-UTR chimeric transcripts was, at all stages, higher than that from the 480- and 530-nt MOS 3'-UTR chimeric transcripts (P < 0.05), but no difference was observed in luciferase activity between the 480- and 530-nt chimeric transcript groups. Whereas the results between the two studies were not significantly different, a morphological examination showed that oocytes injected at different stages during maturation suffer from a number of disadvantages not experienced by oocytes injected at the resumption of meiosis (0 h). These disadvantages included a greater degree of nonspecific damage, particular to oocytes injected after germinal vesicle breakdown, and an attenuation of translational responses at late metaphase II. For these reasons, subsequent experiments aimed at identifying regulatory elements in each of the pig MOS 3'-UTR variants were all carried out on oocytes injected with mRNA constructs at the resumption of meiosis (0 h) and harvested for analysis at metaphase I (24 h), metaphase II (48 h), or late metaphase II (72 h).
Hexanucleotide and U5A Signals Regulate Translation
The MOS 3'-UTR was subjected to a mutational analysis to test, first, whether the two U5A motifs, which are situated 3 and 75 nt upstream of the hexanucleotide signal (AAUAAA), function as CPEs in pig oocytes and, second, whether the degree of control exerted by these U-rich sequence differ in the three different-length MOS variants. The constructs used to test both these questions are shown in Figures 1 and 3A, whereas the consequences of mutating the MOS 3'-UTR are presented in Figure 3B. Deletion of the hexanucleotide signal terminated translation in each of the three variants (P < 0.001). By contrast, mutations to U-rich sequences in the short form (180-nt) chimeric mRNA significantly alters both the pattern and level of luciferase activity (P < 0.01) but without entirely suppressing luciferase-MOS mRNA translation (Fig. 3B, i). Thus, translation in short-form transcripts containing mutations to the U5A motifs increased significantly during the first 24 h to levels indistinguishable from those of their nonmutated counterparts. Thereafter, translational activity in the mutated short-form variant plateaued, whereas that in the nonmutated group underwent a further threefold rise (P < 0.001). We conclude that the hexanucleotide signal is indispensable for translation but that the importance of the U5A elements depends partly on the length of the MOS 3'-UTR.
|
Mutations in the A-Rich Motif of the MOS 3'-UTR Block Translation
The presence of a distinctive A-rich element in some masked mRNAs in pig oocytes raises the question of whether this motif acts as an additional regulator of translation in this species. To investigate this, we compared full-length luciferase-MOS chimeric mRNA with constructs in which the A-rich region had been deleted or mutated (Figs. 1 and 4A). Luciferase activity in oocytes injected with luciferase control RNA was invariably low (<3.0 OD units) and highly significantly different (P < 0.001) from that in oocytes injected with wild-type luciferase-MOS chimeric constructs (Fig. 4B). The effect of introducing mutations into the A-rich tract, shown in Figure 4B (i and iii), was to highly significantly reduce (P < 0.01) translation in both the 180- and 530-nt variants (not tested in the 480-nt variant). Similarly, the complete deletion of the A-rich tract significantly (P < 0.01) depressed luciferase translation in the 480- and 530-nt chimeric mRNA constructs (Fig. 4B, ii and iii). By contrast, deletion of the entire A-rich tract from the 180-nt variant (Fig. 4B, i) had only a small suppressive effect on luciferase translation during the initial 48 h after injection. However, by 72 h, translation of the 180-nt variant containing the A-rich deletion was significantly (P < 0.05) lower than that of the corresponding wild-type construct (Fig. 4B, i). We conclude that the sequence integrity of the A-rich tract in all three variants is essential for MOS translation but that the effect of its total deletion is least in the 180-nt variant and greatest in the longest-form (530-nt) variant.
|
Although the A-rich region in the MOS 3'-UTR is located 71 nt downstream of the stop codon and 24 nt upstream of the polyadenylation hexanucleotide signal in all three MOS variants, the polyadenylation site is, by contrast, 43, 343, and 393 nt downstream of the A-rich tract in the short-, intermediate-, and long-form MOS 3'-UTR variants, respectively. To determine whether the position of the A-rich tract is a critical determinant of translational control, two sets of experiments were carried out in which the A-rich region was either duplicated or else moved downstream of the hexanucleotide signal in both the long- and short-form chimeric constructs (Figs. 1 and 5A). The effect on translation of these positional modifications within the MOS 3'-UTR shows that the duplication of the A-rich tract severely depressed translation (P < 0.01) of chimeric luciferase transcripts (Fig. 5B, i and iii). Similarly, the insertion of a 17-nt fragment of random sequence between the A-rich region and the hexanucleotide signal (AAUAAA) equally effectively suppressed luciferase translation (data not shown). By relocating the A-rich sequence to a site approximately 200 nt downstream of the hexanucleotide signal in the intermediate- and long-form chimeric mRNAs (not feasible in the short-form variant), we next investigated whether repositioning rather than duplication of the A-rich region would influence translation. The results showed that the repositioning of the A-rich motif terminates translation (Fig. 5B, ii and iii) by demonstrating that the precise position of this regulatory element is of crucial importance in regulating translation of MOS mRNA in oocytes.
|
Relationships Between 3'-UTR Mutations, Polyadenylation, and Translation of Luciferase-MOS Chimeric mRNA
A poly(A) analysis was carried out on total RNA extracted from oocytes at the G2, metaphase I, and metaphase II stages to determine the timing and extent of polyadenylation in pig MOS mRNA during maturation. The results showed that a global increase in length of the MOS poly(A) tail was detected as meiosis progressed from metaphase I to metaphase II (data not shown). Ascribing polyadenylation changes to individual MOS variants, however, was not feasible, because all three variants coexist in pig oocytes. To circumvent this problem, we determined the extent of polyadenylation in individual MOS variants by injecting the short- or long-form luciferase-MOS chimeric mRNA variants into pig oocytes and measuring the extent of polyadenylation 48 h later (metaphase II). The baseline set of results in these experiments is shown in Figure 6A, lane 1 (180-base pair variant) and in Figure 6B, lane 1 (530-nt variant). The control luciferase transcripts, which are devoid of an associated MOS 3'-UTR, are not shown, because they do not undergo measurable polyadenylation change during maturation. In marked contrast to the luciferase controls, wild-type chimeric luciferase-MOS constructs underwent extensive polyadenylation during maturation; the poly(A) tail increased by approximately 300 adenines in the luciferase-MOS 180-nt UTR variant and by 400 adenines in the luciferase-MOS 530-nt UTR variant (Fig. 6, A and B, lane 2 [wild type]). These findings show that the polyadenylation signal element in pig oocytes resides in the MOS 3'-UTR, that each of the MOS variants is heavily polyadenylated during maturation despite each variant arising from a different polyadenylation site, and that the 530-nt variant becomes more heavily polyadenylated during maturation than the 180-nt variant despite the polyadenylation site in the long-form variant being separated from the hexanucleotide signal by 404 nt.
|
Oocytes were injected with wild-type and mutated forms of luciferase-MOS chimeric mRNA to test whether mutations to the hexanucleotide signal, the two U-rich sequences, or the A-rich region uniformly suppressed translation by preventing polyadenylation during maturation. Mutations to the hexanucleotide signal (AAUAAA to AACUAA) in both the short-form (180-nt) and long-form (530-nt) MOS 3'-UTR variants totally abolished poly(A) tail elongation during oocyte maturation (Fig. 6, A and B, lane 3 [AAUAAA mutation]). By contrast, mutations to the U-rich sequences (UUUUUA to UUUGCA) or to the A-rich elements in both the short- and long-form luciferase-MOS chimeric transcripts reduced by approximately 40% but did not totally inhibit polyadenylation during oocytes maturation (Fig. 6, A and B, lanes 4 [U-tract mutant] and 5 [A-tract mutant]).
The combined results from both the polyadenylation analysis (Fig. 6) and luciferase expression assays (Figs. 2– 5) show that the hexanucleotide signal is essential for both polyadenylation and translation of the luciferase-MOS transcripts. By contrast, mutations to either the A-rich or the U-rich sequences had a relatively limited but comparable effect on polyadenylation in both the short- and long-form MOS variants, with polyadenylation being reduced to approximately 60% of that observed in wild-type constructs. Nevertheless, these same mutations totally abolished translation in the long-form, but not in the short-form, MOS variants. These results suggest that the correlation between translation and polyadenylation in pig MOS constructs can be relatively poor.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
That chimeric mRNA constructs are translated in a pattern dictated by the 3'-UTR is well established [19–21]. Our results add to this extensive body of evidence by establishing that the pig MOS 3'-UTR confers a MOS-like translational pattern on an unrelated mRNA. In pig oocytes, the pattern of MOS mRNA translation is characterized by a relatively extended period of continued translational repression of 18–20 h after the initiation of oocyte maturation (unpublished results). Once derepressed, pig MOS mRNA becomes totally polyadenylated in the 6-h period between 18 and 24 h after the induction of maturation; no further polyadenylation occurs after entry into metaphase I (data not shown). However, MOS mRNA translation, as indicated by the levels of MOS protein (or luciferase activity after luciferase-MOS chimeric mRNA injection), increases steadily from metaphase I (24 h) to metaphase II (48 h). In the absence of fertilization or experimental egg activation, translational levels are sustained for at least a further 24 h after the formation of the second metaphase plate. Two questions arise from these observations; namely what suppresses translation for 18–20 h before nuclear membrane breakdown, and what mechanisms stimulate translation during metaphase? The mechanisms of translational repression of oocyte RNAs involve elements in the 3'-UTR that often overlap with those involved in translational activation [3, 21, 22] and also specific protein–RNA interactions [23]. We had anticipated that one or more of the mutations introduced into the pig 3'-UTR might prevent translation repression of the luciferase hybrid transcripts after injection at the G2 stage, especially because a variety of unmasked RNAs injected into G2-stage sheep oocytes are rapidly translated [24]. However, none of the mutations induced precocious translation of hybrid constructs. One interpretation of these findings is that the pig MOS 3'-UTR, like its Xenopus counterpart, is devoid of repressor elements [21]. Alternatively, it is possible that repressor elements were present but were not identified, either because our mutational analysis was too restricted or because of insensitivities in the experimental protocols used in our experiments.
The inability to identify repressors in the 3'-UTR of pig MOS RNA contrasted sharply with the finding that this region is rich in sequences that control RNA activation. The role for each of the activator elements differs, with some acting as translational switches whereas others appear to be controlling the level rather than the induction of translation. Indeed, these multiple levels of control distinguish the pig MOS 3'-UTR from that of either the mouse or frog. The polyadenylation signal and CPEs act as common switches in all three species, and mutations or deletions in either of these elements induce a severe or total inhibition of translation. The two U-rich tracts mutated in the present study, by base substitution of the U preceding the A, are located 3 and 76 nt upstream from the hexanucleotide signal and 17 nt each side of the CA13CA5CA5CA6 element. These U-rich sequences represent the total complement of these elements in the 180-nt pig MOS 3'-UTR, but a total of five additional U3–5 A elements exist downstream of the hexanucleotide in the longer form MOS 3'-UTRs. The effects of these elements on MOS mRNA regulation have not, as yet, been investigated and provide further potential control mechanisms for the 480- and 530-nt MOS mRNA variants.
Our results suggest that the presence of three different-length 3'-UTRs in MOS mRNA from pig oocytes endows this mRNA with additional translational characteristics that extend beyond those described in Xenopus and mouse oocytes. In the present study, we questioned whether the length of the MOS 3'-UTR determined either the time of translational initiation or the subsequent level of translation. The evidence on the timing of translation is equivocal. At 24 h, luciferase activity in oocytes injected with the short-form chimeras was higher (P < 0.05) than that found in oocytes from the other groups, but it is uncertain whether this reflected a slightly earlier onset or a higher rate of translation. Clear evidence from a number of experiments suggest, however, that the level of translation in metaphase-staged oocytes was highest in oocytes injected with luciferase-MOS 190-nt 3'-UTR transcripts and lowest when the chimeras contained the 530-nt 3'-UTR. That multiple polyadenylation sites are not unique to MOS mRNA but also occur in some other mRNAs, and that these can affect translation, already have been highlighted [8 and references therein].
Having established that the length of the MOS 3'-UTR affects the level of translation, we next tested the hypothesis that these differences in translation effectiveness result from differences in the action of the U-rich sequences or the hexanucleotide signal in the three MOS transcripts. To investigate this, we first determined whether the length of each of the three MOS 3'-UTRs was correlated with corresponding levels of polyadenylation during maturation. Our results show that the long-form transcripts are approximately 30% more extensively polyadenylated than the short-form variants. We next compared the effect on translation of inserting the same mutation in each of the different-length hybrid constructs. While mutating the U-rich sequences significantly depressed translation in all three chimeric MOS mRNAs, our results indicate differences in the degree to which translational activity is still possible even after suppression of the U-rich sequence function. Translational activity is abolished in the two long-form transcripts after mutation of the U-rich sequence, but it persists at a reduced level in the short-form MOS mRNA. By contrast, mutating the single consensus hexanucleotide signal reduces translation to comparable basal levels in all three MOS 3'-UTRs. Although the 480- and 530-nt long-form 3'-UTRs possibly might contain additional noncanonical polyadenylation signals preceding the second and third polyadenylation sites, these, if they exist, would be incapable of sustaining translation when the AAUAAA signal is mutated. Furthermore, it is interesting to note that as the distance between the intact hexanucleotide signal and the polyadenylation site increases from 18 to more than 400 nt, the overall level of translation declines, but the extent of polyadenylation increases.
The distinctive A-rich region, CA13CA5CA5CA6, has become a focus of our attention, because other masked mRNAs contain similar A-rich sequences in mammalian oocytes [9]. The combined results of our deletion, point mutation, and relocation assays demonstrated that the nucleotide sequence and location of this CA13CA5CA5CA6 element is essential for the translation, but not for the repression, of transcripts containing either of the two long-form porcine MOS 3'-UTRs. The role of the A-rich element in the 180-nt short-form MOS 3'-UTR is more complex: The total deletion of the A-rich element does not significantly alter either the timing or the level of translation in short-form transcripts during the first 48 h after injection, but it does depress it thereafter. In contrast, base substitutions, relocations, or duplications of this element in chimeric mRNA constructs containing the short-form MOS 3'-UTR suppresses translation in a highly significant manner. Whereas demonstrating the overall importance of the A-rich region in MOS mRNA and, by extension, in at least some other masked mRNAs, the findings do not illustrate the mechanisms by which this element functions. Our working hypothesis, supported by computer modeling, is that the A-rich motif has a key role in maintaining the required secondary structure of 3'-UTRs in some masked porcine mRNAs.
|
FOOTNOTES |
|---|
2 Correspondence: Yanfeng Dai, The Babraham Institute, Laboratory of Molecular Signaling, Babraham, Cambridge CB2 4AT, UK. FAX: 44 0 1223 496043; daiy{at}bbsrc.ac.uk
Received: 19 April 2005.
First decision: 8 May 2005.
Accepted: 28 June 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Uzbekova, Y. Arlot-Bonnemains, J. Dupont, R. Dalbies-Tran, P. Papillier, S. Pennetier, A. Thelie, C. Perreau, P. Mermillod, C. Prigent, et al. Spatio-Temporal Expression Patterns of Aurora Kinases A, B, and C and Cytoplasmic Polyadenylation-Element-Binding Protein in Bovine Oocytes During Meiotic Maturation Biol Reprod, February 1, 2008; 78(2): 218 - 233. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Miller and G.C. Ostermeier Towards a better understanding of RNA carriage by ejaculate spermatozoa Hum. Reprod. Update, November 1, 2006; 12(6): 757 - 767. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) shows the hexanucleotide polyadenylation signal; the underlined sequence (
) shows the A-rich sequence. Duplicate A-rich sequences (not shown) were inserted immediately following the native A-rich region. The bold italicized sequence AAGCUU (HindIII site) show the site to which the A-rich sequence was translocated
