Oxidized messenger RNA induces translation errors

Abstract

To investigate the effect of RNA oxidation on normal cellular functions, we studied the translation of nonoxidized and oxidized luciferase mRNA in both rabbit reticulocyte lysate and human HEK293 cells. When HEK293 cells transfected with nonoxidized mRNA encoding the firefly luciferase protein were cultured in the presence of paraquat, there was a paraquat concentration-dependent increase in the formation of luciferase short polypeptides (SPs) concomitant with an increase in 8-oxoguanosine. Short polypeptides were also formed when the mRNA was oxidized in vitro by the Fe–ascorbate–H₂O₂ metal-catalyzed oxidation system before its transfection into cells. Translation of the in vitro oxidized mRNA in rabbit reticulocyte lysate also led to formation of SPs. The SPs formed by either procedure contained the N-terminal and the C-terminal portions of the tagged luciferase. In addition, the oxidized mRNA was able to associate with ribosomes to form polysomes similar to those formed with nonoxidized mRNA preparations, indicating that the oxidized mRNAs are mostly intact; however, their translation fidelity was significantly reduced. Nevertheless, our results indicate that the SPs were derived from both premature termination of the translation process of the oxidized mRNA and the proteolytic degradation of the modified full-length luciferase resulting from translation errors induced by oxidized mRNA. In light of these findings, the physiological consequences of mRNA oxidation are discussed.

The generation of reactive oxygen species (ROS) is an unavoidable cellular event in normal respiring cells. ROS are known to play an important role in normal physiological functions and they can also oxidize cellular macromolecules, such as nucleic acids, proteins, and lipids, and lead to the loss of their physiological functions (1, 2). Growing evidence indicates that increased oxidative damage appears to be associated with a wide range of age-related neurodegenerative conditions. Although oxidation of DNA may ultimately have deleterious consequences on cellular homeostasis, multiple DNA repair systems are available to minimize such damage (3). Because of the fact that RNAs are less protected against ROS under oxidative stress conditions, intracellular RNA was found to be more susceptible to oxidation than DNA (4, 5). Similarly, irradiation of human skin fibroblasts with 765-kJ/m² UV A induced more extensive RNA oxidation than DNA oxidation (6). The oxidation of rRNA, tRNA, and mRNA may have serious effects on cellular homeostasis, because oxidation of these RNAs could impair the overall integrity of translational processes. Moreover, it was recently shown that oxidation of RNA is implicated in several neurodegenerative diseases (7–9). Nunomura et al. (8) reported that cytoplasmic RNA oxidation occurs to a greater extent in neurons that are implicated in the development of Alzheimer's disease (AD). Shan et al. (10) found that significant amounts of poly(A)⁺ mRNA were oxidized in AD brains, and the oxidation of RNA appeared to be highly selective, not random. However, it remains to be resolved mechanistically how oxidized RNA, particularly the relatively short-lived mRNA, affects cellular metabolism.

In this report, we investigated the translation of nonoxidized and oxidized luciferase mRNA in both rabbit reticulocyte lysate and human HEK293 cells. The results reveal that there is a direct correlation between the extent of mRNA oxidation and the frequency of translation errors, as indicated by the accumulation of 8-oxoguanosine derivatives and short peptides (SPs). These SPs are derived both from the rapid proteolytic degradation of the translation error-containing proteins and from the premature termination of the translational process. The physiological consequences of mRNA oxidation will be discussed in light of these findings.

Results

Translation Products Under Oxidative Stress Conditions.

To investigate the effects of intracellular ROS production on protein expression, firefly luciferase (F-luciferase) mRNA that had been fused with the FLAG tag at the 5′ terminus and the Myc tag at the 3′ terminus was transfected into HEK293 cells. The transfected cells were cultured for 6 h in the presence of two concentrations of paraquat, which, in the presence of oxygen, is known to produce superoxide radical anion. As shown in Fig. 1, Western blotting using FLAG- or c-Myc-specific antibodies demonstrated that SPs, which contained either the N terminus or the C terminus of the translated F-luciferase, were generated when transfected cells were treated with paraquat. The quantity of SPs formed appears to be dependent on the concentration of paraquat used. In addition, the accumulation of SPs depended upon the presence of the proteasome inhibitor MG132.

Fig. 1.

Detection of SPs in oxidatively stressed cells. HEK293 cells transfected with mRNA encoding both the FLAG- and c-Myc-tagged luciferase gene were cultured in the presence of paraquat for 6 h. (Upper) Immunoblots using anti-FLAG antibody (Left two gels) or anti-c-Myc antibody (Right two gels) were prepared from cells treated with or without MG132. (Lower) Immunoblots with anti-β-actin after stripping out the anti-FLAG and anti-c-Myc to verify equal loading.

To confirm that paraquat treatment induces oxidation of mRNA in cells, tritium-labeled luciferase mRNA was transfected into HEK293 cells, which were then treated with 20 μM paraquat for 6 h. Data summarized in Fig. 2A show that paraquat treatment had little effect, if any, on the degradation of luciferase mRNA. However, as shown in Fig. 2B, paraquat treatment led to the formation of mRNA 8-oxoguanosine derivatives as revealed by its binding to 8-oxoguanosine-specific antibody. The data in Fig. 2B further show that 8-oxoguanosine is associated with RNA and not with DNA, because antibody binding was unaffected by DNase and almost completely eliminated by RNase treatment. Moreover, evidence that antibody binding is specific for 8-oxoguanosine derivatives follows from the fact that no radioactive material was bound to antibody that was previously saturated with unlabeled 8-oxoguanosine.

Fig. 2.

Oxidation of mRNA transfected to cells under oxidative stress. (A) Tritium-labeled mRNA extracted from transfected HEK293 cells treated with (solid bars) or without (open bars) 20 μM paraquat for 6 h was subjected to 1% agarose gel electrophoresis as described in Materials and Methods. The lane on a gel was excised into nine blocks from the top to the front end. The radioactivity in each block was measured. (B) Extracted mRNA (≈6,000 cpm) from transfected cells was immunoprecipitated with anti-8-oxoguanosine antibody. Where indicated, the mRNA was pretreated with RNase or DNase, or the antibody was pretreated with 8-oxoguanosine. The generation of 8-oxoguanosine was quantitated and expressed as the ratio of the bound radioactivity to that of the unbound form. Results were from three independent RNA samples.

Translation of Oxidized mRNA in Vivo.

To examine the effects of metal-catalyzed oxidation of mRNA on its translation capacity in HEK293 cells, oxidized F-luciferase mRNA obtained by treating the mRNA with increasing concentrations (0–10 μM) of 1:1 mixtures of iron and ascorbate for 1 h was transfected into the cells. The luciferase activity was assayed after the transfected cells were incubated for 12 h. Results summarized in Fig. 3A show that oxidation of the luciferase mRNA by the iron–ascorbate system led to a dose-dependent decrease in the F-luciferase activity, expressed as a ratio to the activity of the nonoxidized R-luciferase. However, if 1 mM EDTA was added before the luciferase mRNA was treated with iron–ascorbate, it prevented its oxidation, as indicated by the resulting luciferase activity. In essence, the relative luciferase activity observed for the translation product of this nonoxidized mRNA is significantly higher than that obtained with oxidized mRNA. Fig. 3B shows that under anaerobic conditions, where oxidation of F-luciferase mRNA was prevented, the relative luciferase activity remained at high levels, independent of the presence or absence of iron and ascorbate. Moreover, under aerobic conditions, no oxidation of mRNA occurred in samples treated with ascorbate alone, and treatment with iron (ferrous chloride) alone was less effective than that obtained with both iron and ascorbate.

Fig. 3.

Translation of oxidized mRNA in HEK293 cells. (A) Relative luciferase activity [F-luciferase/Renilla luciferase (R-luciferase)] observed with samples containing F-luciferase mRNA incubated for 1 h in the presence of the indicated amounts of a 1:1 mixture of ferrous iron and ascorbate, in the absence or presence of 1 mM EDTA before transfection as indicated. After transfection, cells were incubated for 12 h before measurement of luciferase activity. (B) Data obtained with F-luciferase mRNA incubated with 10 μM ascorbate (asc), 10 μM ferrous chloride (Fe), 10 μM ferrous chloride and ascorbate (Fe-asc), or no addition of ferrous iron and ascorbate (ctrl), under aerobic or anaerobic conditions before transfection.

SPs Are Formed by Translation of Oxidized mRNA in Vitro.

The effects of oxidized mRNA on protein production were also investigated by using an in vitro translation system, where proteolytic degradation can be minimized by the addition of multiple protease inhibitors. The mRNA encoding epitopes-tagged F-luciferase was oxidized with indicated concentrations of iron–ascorbate and hydrogen peroxide for 1 h. The translation of these oxidized mRNAs was carried out in rabbit reticulocyte lysate containing multiple protease inhibitors. The translation products of oxidized luciferase mRNA were monitored by using the anti-FLAG and anti-c-Myc antibodies that recognize the luciferase N terminus and the C terminus, respectively. The intensity of the bands revealed by anti-FLAG or anti-c-Myc was normalized to correct for the differentials in antibody sensitivity with the assumption that the full-length luciferase contained both N- and C-terminal tags. Quantitation of the full-length luciferase generated as a function of time revealed that mRNA oxidation impaired its translation efficiency as monitored by Western blotting (Fig. 4A). A similar time course was also observed by monitoring the luciferase activity (data not shown). These data reveal that oxidation of mRNA causes a decrease in its translation fidelity. Fig. 4B shows the ratios of the sum of SPs formed to total protein and peptide generated as monitored with anti-FLAG antibody (N-terminal) or anti-c-Myc antibody (C-terminal). Comparing the ratios detected with anti-FLAG and anti-c-Myc antibodies, it is clear that oxidized mRNA leads to an increase in generation of SPs, with N-terminus-containing SPs proportionally higher than those SPs containing the C terminus. These results indicate that there are increasing quantities of prematurely terminated peptides formed when the template used is oxidized mRNA, and the quantities of SPs generated are correlated with the extent of mRNA oxidation. The ratio observed with anti-c-Myc antibody also increases, but with a much lower amplitude relative to that observed for the N terminus. The C-terminal SPs may derive, in part, from the degradation of the full-length protein, a process expected to be more favorable for the abnormal proteins.

Fig. 4.

In vitro translation of oxidized mRNA. mRNA oxidized by iron–ascorbate and hydrogen peroxide was incubated in rabbit reticulocyte lysate supplemented with protease inhibitors as described in Materials and Methods. After incubation for the indicated times, the reaction mixture was subjected to Western blot analysis using anti-FLAG antibody or anti-c-Myc antibody. The Western blot intensity obtained with anti-FLAG and anti-c-Myc antibodies was normalized with the assumption that the full-length proteins contained both FLAG and c-Myc tags. (A) Time course for the relative amount of full-length protein formed with nonoxidized mRNA (♦), mRNA oxidized for 1 h with 5 μM 1:1 iron–ascorbate and 1 μM H₂O₂ (■), or mRNA oxidized for 1 h with only 5 μM H₂O₂ (▴). (B) Ratio of the SPs to total protein/peptide generated, as monitored by Western blotting (WB) with anti-FLAG (N-terminal) (Left) or anti-c-Myc (C-terminal) (Right) antibodies. Solid black bar, gray bar, and open bar represent data for nonoxidized mRNA, mRNA oxidized with 1 μM H₂O₂, or mRNA oxidized with 5 μM H₂O₂, respectively, as described above.

SPs Are Formed by Translation of Oxidized mRNA in Vivo.

Production of SPs was systematically observed when cells were transfected with oxidized F-luciferase mRNA. Western blot analysis using FLAG antibody, which recognizes the N-terminal region of the luciferase, showed that in vivo translation of oxidized mRNA leads to the generation of a number of low molecular weight peptides, but these peptides are observed almost exclusively only in the presence of the protease inhibitor MG132 (Fig. 5A and B). To rule out the possibility that the SPs observed may be derived from an MG132-insensitive protease, experiments were carried out with MG132 plus leupeptin, a lysosomal serine/cysteine protease inhibitor; pepstatin A, a lysosomal aspartate protease inhibitor; or benzyloxycarbonyl-Val-Ala-Asp(OMe) fluoromethyl ketone (zVAD-fmk), a general caspase inhibitor. Results of these experiments showed that these additional protease inhibitors exerted no effect on the formation of SPs (data not shown). However, in the absence of MG132, there was no accumulation of SPs (Fig. 5B). This raises the possibility that translation of oxidized mRNA leads directly to generation of SPs. However, the fact that MG132 was needed to detect the SPs implies that the SPs are readily degraded by the proteasome. It should be pointed out that under the oxidizing conditions used to obtain the data shown in Fig. 5, the F-luciferase mRNA appeared to be intact (Fig. 5C).

Fig. 5.

Generation of SPs by translation of oxidized mRNA. (A and B) (Upper) HEK293 cells were transfected with in vitro oxidized mRNA, mock mRNA (mRNA unrelated to luciferase), or truncated mRNA and incubated with (A) or without (B) MG132. The translation products were subjected to Western blot analysis using the anti-FLAG antibody. (Lower) Anti-β-actin immunoblots to verify equal loading for each lane. (C) The mRNA samples were examined by agarose gel electrophoresis followed by ethidium bromide staining. (D) The expression levels of the full-length proteins, quantified by using a fluorescence-based assay (black bars) together with luciferase activity (gray bars) in cell lysates, were compared in cells treated either with (Left) or without (Right) MG132.

The possibility that SP formation reflects the translation of mRNA fragments generated during oxidation of the mRNA is discounted by results of studies showing that SPs were not formed in fractions containing fragmented mRNA molecules after limiting RNase digestion (data not shown) or truncated mRNA with its 3′ terminus removed. In addition, the relative quantities of the full-length luciferase shown in Fig. 5 A and B are summarized in Fig. 5D, together with its catalytic activity. Note that in the presence of the proteasome inhibitor MG132 luciferase activity appears to decrease with increasing hydrogen peroxide to ≈30%, whereas the full-length protein declined only to ≈65% (Fig. 5D Right). However, in the absence of MG132, the data in Fig. 5D Left relative to those shown in Fig. 5D Right reveal that the inactive full-length luciferase produced by oxidized mRNA appears to be more susceptible to proteasome-mediated degradation.

Oxidized mRNA Is Associated with Polysomes.

To investigate whether the observed effect of mRNA oxidation or its translation could be derived from the alteration of its ability to form a complex with a polysome, the binding of oxidized mRNA to polysomes was investigated by using a sucrose density gradient method. Fig. 6A shows that the radioactively labeled oxidized mRNA accumulated in both the polysome and monosome fractions. The distribution was similar to that of nonoxidized mRNA. Quantitatively, Fig. 6B shows that there is little change in polysome formation by mRNA oxidized by 2 μM or 10 μM H₂O₂, despite the fact that the activity of the translated luciferase was greatly reduced because of mRNA oxidation (Fig. 6C). This observed reduction in luciferase activity can be attributed to the translation errors caused by the oxidized mRNA and leading to the production of SPs and inactive full-length luciferase. Furthermore, as shown in Table 1, the association of oxidized mRNA to polysomes exhibited a response similar to that of nonoxidized mRNA to puromycin. In both cases, puromycin caused a reduction of mRNA binding to polysomes, whereas cycloheximide showed no effect on the stability of the mRNA–polysome complex. These data, together with those shown in Fig. 2, reveal that oxidized mRNA is mostly intact and mRNA oxidation does not alter its ability to bind polysomes.

Fig. 6.

Binding of oxidized mRNA to polysomes. (A) In vitro oxidized ³H-labeled RNA was transfected into HEK293 cells. The cell lysate was subjected to sucrose density centrifugation and fractionated as described in Materials and Methods. The radioactivity was measured for each fraction. (B) Percent recovery of the oxidized mRNA in the polysome fraction. (C) Relative luciferase activity in the cell lysate.

Table 1.

Distribution of oxidized mRNA to monosome or polysome fraction after treatment with protein synthesis inhibitor

mRNA	Inhibitor	Monosome fraction, %	Polysome fraction, %
Control	None	25.6	23.5
	Cycloheximde	24.2	25.1
	Puromycin	42.6	6.9
Oxidized	None	25.6	24.8
	Cycloheximide	23.5	23.7
	Puromycin	42.5	6.6

Discussion

Free radicals and ROS are known to oxidize cellular macromolecules, such as proteins, lipids, and nucleic acids, which could impair their physiological functions. Much attention has been focused on the oxidative effects of proteins, lipids, and DNAs (1, 2, 11), whereas the oxidation of RNAs likely plays a major role in exerting the biological effects of ROS because RNAs are more vulnerable to oxidation than DNAs, and oxidized mRNAs may impair the translation process. Furthermore, oxidation of RNA has been implicated in several neurodegenerative disorders (7–9, 12) and has been shown to occur in the early stages of diseases (13). In this study, we investigated the effect of mRNA oxidation on its translational integrity. To this end, F-luciferase mRNA was synthesized and oxidized in vitro by using the iron–ascorbate–hydrogen peroxide system. The translation of oxidized luciferase mRNA was investigated both in HEK293 cells and in the rabbit reticulocyte lysate translation system. Alternatively, the nonoxidized form of mRNA was transfected into HEK293 cells followed by incubation in the presence or absence of paraquat, which generates superoxide in the presence of oxygen. Results of these studies led to the following observations: (i) Both in vitro and in vivo oxidation of luciferase mRNA leads to the formation of SPs, suggesting the possibility that oxidation of mRNA causes premature termination of the translation process and/or the formation of modified luciferase molecules caused by translation errors that are more susceptible to proteolytic degradation (Figs. 1, 4B, and 5A). (ii) Accumulation of SPs is observed almost exclusively in the presence of the proteasome inhibitor, indicating that the SPs are highly susceptible to degradation catalyzed by proteasome. (iii) The formation of SPs depends on the concentration of paraquat used to oxidize the mRNA in HEK293 cells (Fig. 1) and the concentration of the iron–ascorbate–H₂O₂ system used for the in vitro oxidation of mRNA (Figs. 4B and 5A). (iv) SPs were not detected in cells transfected with nonoxidized mRNA that had been mildly digested with RNase (data not shown) or with truncated mRNA (Fig. 5A). These results indicate that the fragmented mRNA may not form stable polysomes. (v) In translation in vitro, SPs that contain the N terminus are disproportionately more abundant than those containing the C terminus, and this observation is consistent with the notion of premature termination (Fig. 4B). The possibility that the observed lower level of C-terminus-containing SPs is due to the cleavage site occurring at the c-Myc tag, such that the C-terminus-containing SP becomes too small to be retained on the SDS/polyacrylamide gel, is quite unlikely because the sizes of the N-terminus-containing SPs varies over a wide range (Fig. 1). (vi) Formation of SPs was not altered when the proteasome inhibitor MG132 was added with either lysosomal protease inhibitor leupeptin or pepstatin A or with the general caspase inhibitor zVAD-fmk. (vii) It seems likely that SPs generated in response to paraquat treatment after transfection of mRNA are due to mRNA oxidation, because mRNA 8-oxoguanosine derivatives were generated and the SPs were similarly formed relative to those formed after transfection of mRNA oxidized in vitro by the iron–ascorbate–hydrogen peroxide system (Figs. 2 and 5). However, the possibility that oxidation of other RNA species, such as tRNA and rRNA, contributes to SP production is not excluded. (viii) Paraquat treatment leads to in vivo mRNA oxidation without observable mRNA degradation, and yet oxidized mRNA exhibits a binding capacity to polysomes similar to that of its nonoxidized counterpart (Fig. 2, Fig. 6, and Table 1). (ix) The oxidized forms of luciferase mRNA impair the formation of the full-length luciferase and its fully active enzyme in a dose-dependent manner with respect to the concentration of the iron–ascorbate–hydrogen peroxide system used to oxidize the mRNA (Figs. 3, 4A, and 5D). (x) In the absence of proteasome inhibitors, the inactive full–length luciferase is preferentially degraded (Fig. 5D).

Together, our data revealed that the translation fidelity of mRNA was significantly reduced by its oxidation despite the fact that the oxidized mRNA exhibited an affinity for polysomes that was similar to its nonoxidized form. This loss of translational efficiency could be due to the fact that the oxidized ribonucleotides causes (i) premature termination and (ii) generation of modified full-length enzymes/proteins because of translation errors, and these modified proteins have partial or no enzymic activity. The modified full-length proteins, particularly those that exhibit low or no enzymic activity, are more susceptible to proteolytic degradation, which may contribute to some of the SPs observed. Contrary to the supposition that RNA oxidation leads to suppression of protein synthesis, our results indicated that oxidized mRNA was translated to form nonfunctional full-length or full-length proteins with reduced activities and short peptides. While these modified protein derivatives and SPs are vulnerable to proteolytic degradation, under conditions of reduced proteolytic activity these protein derivatives generated by oxidized mRNA may, in part, contribute to the accumulation of protein aggregates observed in neurodegenerative diseases (7, 8, 10) and in age-related changes in protein modification and thermostability (14).

Materials and Methods

Cell Culture and Transfection.

Human embryonic kidney (HEK) 293 cells purchased from the American Type Culture Collection (Manassas, VA) were cultured in DMEM supplemented with 10% fetal bovine serum (FBS) (Invitrogen, Carlsbad, CA) under a humidified 5% CO₂/5% O₂/90% N₂ environment. One day before transfection, cells were distributed onto poly(d-lysine)-coated six-well tissue culture plates. After rinsing the cells once with Opti-MEM (Invitrogen), 1 ml of Opti-MEM containing 5 μg of in vitro synthesized mRNA (see below) and 8 μl of DMRIE-C (Invitrogen) was added to each well of the six-well plate. After 4-h incubation at 37°C, 2 ml of preheated DMEM plus 10% FBS was added and the cells were cultured for a total of 8–10 h in medium, during which the cells were treated with 10 μM MG132, together with 3 μM lactacystin, 100 μM pepstatin A, 100 μM leupeptin, or 100 μM zVAD-fmk 6 h before harvesting.

Preparation of DNA Template for mRNA Synthesis.

Poly(deoxyadenylic acid)·poly(thymidylic acid) homopolymer (Sigma, St. Louis, MO) was partially digested with RNase mix (Ambion, Austin, TX) to give double-stranded poly(dA)·poly(T) oligonucleotides. The oligonucleotides were ligated into the pGEM-4Z vector (Promega, Madison, WI) at the SmaI site to generate a plasmid containing 45-mers of the polyadenylate sequence, pGEM-4Z A₄₅.

On the other hand, a DNA fragment of synthetic F-luciferase gene, luc2, was excised from the pGL4.14[luc2/Hygro] vector (Promega) and inserted in-frame with the triple FLAG tag into the P×3FLAGCMV10 vector (Sigma) between the BglII and HindIII sites. The FLAG-tagged luc2 gene was then subcloned into the pGEM-4Z A₄₅ vector at the HincII and XbaI sites (designated pGEM-4Z ×3FLAGLuc2 A₄₅). The 3′-terminal coding region of luc2 gene between HpaI and XbaI was replaced by a PCR fragment containing the primers GCGGCTACGTTAACAACCCCGAGGCTA (HpaI site is underlined) and AAAAATCTAGAAACTCGAGCACGGCGATCTTGCCGCCCTTCTT (XhoI site is underlined, XbaI site is boldfaced) to yield a mutant that provides the XhoI and XbaI sites at the 3′ end instead of at the stop codon of luc2 gene. A triple Myc-tagged cDNA was prepared by PCR amplification using primers TACCGTCGACAACTCGAGGAGCA and AAAAATCTAGAGGTACCGGGCCCTTACAGAT with pCMV-3Tag-4 vector (Stratagene, La Jolla, CA) as a template. The PCR product was cleaved by XhoI and XbaI, then inserted between the XhoI and XbaI sites at the 3′ end of the luc2 gene to yield a plasmid designated pGEM-4Z ×3FLAGLuc2×3Myc A₄₅. All of the cloned plasmids were verified by DNA sequencing.

In Vitro Synthesis of mRNA.

All of the reactions were carried out under anaerobic conditions. The solutions and enzymes used were preequilibrated in an anaerobic chamber for 30 min or longer in a precooled box. The in vitro mRNA synthesis was started by adding a T7 RNA polymerase (obtained from Ambion) to the reaction mixture containing pGEM-4Z ×3FLAGLuc2×3Myc A₄₅ linearized with EcoRI cap analogue and all of the required nucleotides. The reaction mixture was incubated at 37°C for 1.5 h. The reaction was then terminated by DNase treatment for 15 min. For the preparation of radiolabeled mRNA, 40 μCi (1 μCi = 37 kBq) of [³H]GTP (Amersham Biosciences, Piscataway, NJ) was added to the reaction mixture to yield ≈80,000 cpm per μg of mRNA. The mRNA was then purified by using an RNeasy kit (Qiagen, Valencia, CA). This purified mRNA was used for the in vitro translation experiments. To further stabilize the in vitro synthesized mRNA in HEK293 cells, the reaction mixture described above was incubated with 16 units of polyadenylate polymerase (Ambion) in a buffer containing 0.8 mM ATP for 1 h to elongate the polyadenylation tail. Subsequently, the mixture was treated with EDTA (final concentration 2 mM) for 30 min at 4°C. The mRNA was then purified as describe above. For preparation of the truncated mRNA, the AgeI-cleaved pGEM-4Z ×3FLAGLuc2×3Myc A₄₅ was used for an in vitro RNA synthesis as described above. The RNA contained the first 1,132 bases of the luc2 coding region (1,653 bases) without the poly(A) tail. All of the synthesized mRNA was stored at −80°C.

Modification of Synthesized mRNA.

Twenty-five micrograms of luciferase mRNA were incubated for 1 h at 37°C with the indicated concentration of ascorbic acid, ferrous chloride, and/or hydrogen peroxide in 100 μl of 10 mM Chelex 100-pretreated Hepes buffer, pH 7.4. When required, the reaction was carried out under anaerobic conditions after bubbling the RNA sample and stock solutions with nitrogen gas and preequilibrating in an anaerobic chamber for 30 min. The oxidative reaction was terminated by adding 1 mM EDTA and maintained on ice. The modified RNA was stored under anaerobic conditions at −80°C. Partially degraded mRNA was prepared by treating the mRNA with 0.5 ng/ml RNase A at pH 7.4 in 10 mM sodium phosphate-buffered solution for 10 min at 37°C. After reaction, the RNA was purified.

Immunoblotting.

Transfected cells were lysed in pH 7.2 PBS containing 10 mM phosphate, 0.138 M NaCl, 2.7 mM KCl, 1% Triton X-100, and 1% protease inhibitor mix (Calbiochem, San Diego, CA) for 30 min on ice. The cell lysate was centrifuged at 13,000 × g for 5 min to remove the insoluble fraction. Thirty micrograms of each cell lysate was subjected to SDS/PAGE and transferred onto a poly(vinylidene difluoride) (PVDF) membrane. An aliquot of cell lysate was also used for the luciferase activity assay (Promega) and protein determination by BCA (bicinchoninic acid; Pierce, Rockford, IL) using BSA as a standard. The PVDF membrane was blocked with PBS containing 0.1% Tween 20 (PBST) and 5% skim milk (Bio-Rad, Hercules, CA) for 1 h. The blocked PVDF membrane was then probed with the primary antibody against FLAG M2-horseradish peroxidase (HRP) conjugate (Sigma) at 1:1,000 dilution, Myc-tag (Cell Signaling, Beverly, MA) at 1:1,000 dilution, or anti-β-actin-HRP conjugate (Santa Cruz Biotechnology, Santa Cruz, CA) at 1:500 dilution for 2 h at room temperature or overnight at 4°C, followed by washing three times with PBST.

For quantitating the band intensity on the transferred membrane, ODYSSEY (LI-COR Bioscience, Lincoln, NE) was used. For these experiments, the primary antibody, such as anti-FLAG mouse monoclonal antibody (Sigma), was probed with a secondary antibody, IRDye 680-conjugated goat polyclonal anti-mouse IgG (Molecular Probes, Eugene, OR).

In Vitro Translation.

After 10 min of preincubation at 30°C, 33 μl of rabbit reticulocyte lysate (Promega) was mixed with 17 μl of a reaction mixture containing each amino acid at 20 μM, 80 mM potassium chloride, 40 units of RNase inhibitor, 0.5 μl of a protease inhibitor set (Calbiochem; contains 100 mM 4-(2-aminoethyl)benzenesulfonyl fluoride (AEBSF), 80 μM aprotinin, 5 mM bastatin, 1.5 mM E-64 protease inhibitor, 2 mM leupeptin, and 1 mM pepstatin A), 100 μM zVAD-fmk, 5 μM lactacystin, and 2 μg of mRNA oxidized in vitro. After incubation for the indicated time, aliquots of the reaction mixture were subjected to SDS/PAGE. The SDS/PAGE-separated proteins were transferred onto a PVDF membrane. The samples were then subjected to Western blot analysis using FLAG M2 antibody or c-Myc antibody (9E10 monoclonal) coupled with the corresponding secondary antibody. The signal intensity of the Western blot was measured by using the Personal Densitometer (Amersham Biosciences) and subtracting the intensity of the blank obtained with a reaction mixture containing no mRNA template. The blank corrected signal intensity was converted to relative protein concentration by using a standard curve obtained with serial dilutions of the full-length tagged luciferase. The relative quantity of N terminus- and C terminus-containing SPs was expressed as a ratio to the total N-terminal or C-terminal tags.

Quantification of Oxidized Luciferase mRNA.

HEK293 cells transfected with tritium-radiolabeled mRNA were treated with 20 μM paraquat for 6 h before harvest. Cells were suspended in cell lysis buffer provided by Qiagen plus 25 mM EDTA and passed through a QIA shredder (Qiagen) to reduce viscosity. Total cellular RNA was isolated by using an RNeasy spin column (Qiagen). Aliquots of radioactive RNA (≈6 × 10³ cpm) were incubated with 3 μl of anti-8-oxoguanosine antibody solution obtained from QED Bioscience (San Diego, CA) for 2 h at 4°C. Fifteen microliters of washed protein G-agarose (Amersham Biosciences) was then added to the reaction mixture, which was shaken for 2 h at 4°C. The mixture was then centrifuged at 10,000 × g for 1 min and the supernatant was carefully transferred to a new tube. The pellet was further washed three times with cold PBST. The radioactivity in the final washed pellet and the first supernatant were determined as the bound and the unbound fraction, respectively.

Sucrose Density Gradient Fractionation.

HEK293 cells transfected with nonoxidized and oxidized radiolabeled mRNA in a 6-cm dish were incubated in DMEM for 10 h. Where applicable, these cells were incubated in the presence of 100 μg/ml puromycin or cycloheximide for an additional 15 min. After rinsing with chilled PBS containing 100 μg/ml cycloheximide, cells were lysed with 200 μl of cold lysis buffer containing 10 mM Hepes (pH 7.4), 15 mM MgCl₂, 75 mM KCl, 5 mM DTT, 100 μg/ml cycloheximide, and 0.5% Nonidet P-40. The cell lysate was centrifuged at 20,000 × g for 5 min to remove debris before its application onto a polyallomer centrifuge tube filled with a 10–45% linear sucrose gradient in the presence of 10 mM Hepes (pH 7.4), 15 mM MgCl₂, and 75 mM KCl prepared by a gradient maker (ISCO, Lincoln, NE). The lysate-containing sucrose gradient samples were further centrifuged in a Beckman (Fullerton, CA) SW41Ti rotor at 164,000 × g for 1 h or 1.5 h at 4°C. The centrifuged sample was fractionated with a density gradient fractionator (model BR184; Brandel, Gaithersburg, MD) equipped with UV flow cell monitor, recorder, and a Foxy Jr. fraction collector (ISCO). Each fraction was resuspended in distilled water after precipitation with ethanol to remove high-density sucrose, and the amount of radioactivity was measured.

Luciferase Assay.

Luciferase assays were performed by using the Dual-Luciferase Reporter Assay System (Promega) as previously described (15). The oxidized and nonoxidized F-luciferase and nonoxidized R-luciferase were cotransfected into HEK293 cells at a 95:5 ratio. After 12 h of incubation, transfected cells were washed twice with cold PBS and incubated in 250 μl of Passive Lysis Buffer (Promega) for 15 min at room temperature. Five microliters of cell lysate was mixed with 50 μl of F-luciferase substrate, and the luminescence intensity was integrated for 10 sec by using a Turner Designs (Sunnyvale, CA) model TD-20/20 luminometer. A similar procedures with R-luciferase substrate (Stop & Glo buffer; Promega) was used to monitor the R-luciferase activity. The relative luciferase activity was expressed as a ratio of F-luciferase to R-luciferase.

Abbreviations

AD

Alzheimer's disease

F-luciferase

firefly luciferase

R-luciferase

Renilla luciferase

ROS

reactive oxygen species

SPs

short polypeptides

zVAD-fmk

benzyloxycarbonyl-Val-Ala-Asp(OMe) fluoromethyl ketone.