Influence of the 5’4Jntranslated Region of Ornithine Decarboxylase mRNA and Spermidine on Ornithine Decarboxylase Synthesis*

The effect of the 5’-untranslated region of ornithine decarboxylase (ODC) mRNA and of spermidine on the translation of ODC mRNA was studied in a rabbit reticulocyte cell-free system. The ODC mRNAs, pos- sessing different sizes of the 5’-untranslated region and named mODC241, mODC188, mODC85, and mODC24 according to the number of nucleotides in the 5’-untranslated region, were synthesized by in vitro transcription techniques through the use of plasmids containing various sizes of mouse ODC cDNA. As the size of the 5’-untranslated region of ODC mRNA in- creased, the efficiency of the translation decreased and the degree of stimulation of the translation by 0.2 mM spermidine increased. The inhibition of ODC mRNA translation by high concentrations of spermidine 1 mM) also increased with an increase in the size of the 5’-untranslated region. The spermidine effects were increased greatly when the size of the 5’-untranslated region of ODC mRNA was increased from 85 to 188 nucleotides. These results suggest that the nucleotides 70-220 upstream from the initiator AUG are involved in the decrease in translational efficiency. In addition, at least the nucleotides in the 70-170 region upstream from the initiator AUG are important in the strong stimulation of ODC synthesis by low spermidine concentrations and in the inhibition of ODC synthesis at high spermidine concentrations.

The effect of the 5'-untranslated region of ornithine decarboxylase (ODC) mRNA and of spermidine on the translation of ODC mRNA was studied in a rabbit reticulocyte cell-free system. The ODC mRNAs, possessing different sizes of the 5'-untranslated region and named mODC241, mODC188, mODC85, and mODC24 according to the number of nucleotides in the 5'-untranslated region, were synthesized by in vitro transcription techniques through the use of plasmids containing various sizes of mouse ODC cDNA. As the size of the 5'-untranslated region of ODC mRNA increased, the efficiency of the translation decreased and the degree of stimulation of the translation by 0.2 mM spermidine increased.
The inhibition of ODC mRNA translation by high concentrations of spermidine (0.6-1 mM) also increased with an increase in the size of the 5'-untranslated region. The spermidine effects were increased greatly when the size of the 5'-untranslated region of ODC mRNA was increased from 85 to 188 nucleotides.
These results suggest that the nucleotides 70-220 upstream from the initiator AUG are involved in the decrease in translational efficiency.
In addition, at least the nucleotides in the 70-170 region upstream from the initiator AUG are important in the strong stimulation of ODC synthesis by low spermidine concentrations and in the inhibition of ODC synthesis at high spermidine concentrations.
Polyamine biosynthesis is highly regulated in mammalian cells; polyamines themselves participate in this regulation (1,2). It is known that the amount of ornithine decarboxylase (ODC)' is decreased by polyamines increasing the degradation rate of ODC (3-7) and reducing the rate of ODC synthesis (6-11). The finding that the reduction in the rate of ODC synthesis does not involve a decrease in ODC mRNA content (9-11) suggests that polyamine regulation of ODC synthesis is at the translational level. In our studies of the polyamine effect on protein synthesis, we have shown that a low concentration of polyamines stimulates protein synthesis mainly at the level of initiation (12,13). However, the degree of polyamine stimulation differs with individual proteins (12)(13)(14)(15)(16) Spermidine can stimulate ODC synthesis at a concentration lower than that necessary for stimulating total protein synthesis. Higher concentrations of spermidine were inhibitory, and its inhibitory effect on ODC synthesis was greater than on protein synthesis in general (17)(18)(19). This indicates that at least part of the feedback regulation of ODC synthesis exerted by polyamines is due to their direct inhibition of ODC mRNA translation.
In this study, we have prepared various ODC mRNAs having different sizes of the 5'-untranslated region (5'-UTR) and examined the contribution of this region to ODC synthesis in the presence and absence of spermidine. The results suggest that the size of the 5'-UTR is inversely correlated to ODC synthetic activity and that at least the nucleotides in the 70-170 region upstream from the initiator AUG are important for polyamine stimulation and inhibition of ODC synthesis.  1). The 13-17 5'-terminal nucleotides were derived from the vector plasmid and linker used in plasmid construction (Fig.  2). The nucleotide sequences from the 5'-cap to the initiator AUG of the various ODC mRNAs used in these experiments are shown in Fig. 2. We named these mRNAs mODC241, mODC188, mODC85, and mODC24, respectively, according to the number of nucleotides in the 5'-UTR. Translation of ODC mRNA was examined using these mRNAs and a rabbit reticulocyte cell-free translation system (Figs. 3-6). ODC (Mr 51,000) synthesis in the absence of spermidine was highest when mODC24 and mODC85 were used as templates, and the optimal Mg2+ concentration was 2.5-3 mM (Figs. 3 and 4, and Table I). When mODC188 was used as a template, the amount of ODC synthesized was approximately 5% of that with mODC24 (Table I). With mODC241, the amount of ODC synthesized was less than 0.5% of that with mODC24 (Table  I). This could not be seen clearly by fluorography (Fig. 6). However, significant amounts of nascent polypeptide of ODC (&f= 32,000) were synthesized (Fig. 6). Since the exposure time for this fluorography was 150 h rather than 20 or 50 h, a nonspecific protein band (hfr 49,000) appeared in the figure.
The same band was also observed in Fig. 5. When [35S] methionine-labeled polypeptides were precipitated with ODCspecific immunoglobulin, the nonspecific protein band (Mr 49,000) disappeared and a faint band of ODC (Mr 51,000) was observed (Fig. 7). The overall results suggest that the size of the 5'-UTR is very important for the translational efficiency of ODC synthesis. The shorter sequences were much more effective than the longer ones.
Effect of Spermidine on Translation of Various ODC mRNAs-With all the mRNAs tested, the addition of spermidine lowered the optimal M2+ concentration for the translation of ODC mRNA. The optimal M$+ concentration in the presence of spermidine was 1.5-2 mM. When mODC24 was used as a template, 0.2 mM spermidine increased the ODC mRNA translation 9.1-fold compared with translations without added spermidine at 1.5 mM M$+ (Table I). The degree of spermidine stimulation was nearly equal to that observed for thymidine kinase and globin synthesis, and greater than that for protamine and bovine lymphocyte total protein synthesis in this translation system (data not shown). The inhibitory effect at high concentrations of spermidine (0.6-l mM) was weak (Figs. 3 and 8)  When mODC85 was used as a template, the degree of spermidine stimulation at low concentration and the inhibitory effect at high concentration both increased slightly in comparison to the results obtained with mODC24 (Figs. 3, 4, and 8, and Table I and inhibited at high spermidine concentration. When mODC188 was used as a template, the spermidine effect became greater. The degree of spermidine stimulation of ODC synthesis at 0.2 mM spermidine was 19.0-fold (Table  I). ODC synthesis at 0.8 mM spermidine was very low (approximately 5% of that at 0.2 mM spermidine) and that at 1 mM spermidine was negligible (Figs. 5 and 8). Similar results were obtained with the synthesis of the nascent polypeptides of ODC (Mr 32,000 and 39,000). With mODC241, only a small  (Figs. 6 and 7). The degree of spermidine stimulation at 0.2 mM spermidine was always more than l7-fold (Table I). The inhibition of ODC synthesis by high concentrations of spermidine was also observed, but the exact degree of inhibition could not be calculated. The synthesis of the nascent polypeptide of ODC (M, 32,000) was stimulated about H-fold by 0.2 mM spermidine at 1.5 mM Mg2' (data not shown), and the amount synthesized at 0.8 mM was approximately 7% of that at 0.2 mM (Fig. 6). The synthesis of the 39,000 protein was very low. These results suggest that the nucleotides 70-170 upstream from the initiator AUG are important both in the strong spermidine stimulation of ODC synthesis at low spermidine concentration and the inhibition of ODC synthesis by high spermidine concentration. Next, we tried to determine whether or not a specific  interaction between the 5'-side of ODC mRNA and the 3'side of mRNA is necessary for the decrease in translational efficiency and the spermidine effect. For this purpose, an ODC-TK fusion mRNA was synthesized from template DNA, in which the 0.92 kb of the 5'-side of pODC188 was fused to the 1.24 kb of the 3'-side of pTK11 through a Hind111 linker (Fig. 1). When protein synthesis was performed with this mRNA template, results similar to those obtained with mODC188 were obtained (Fig. 9). The translational efficiency was almost the same as with mODC188 (Table I). Spermidine stimulation at 0.2 mM was 17.1-fold, and the inhibitory effect by 0.8-l mM spermidine was clearly observed (Fig. 8). These results show that the nucleotide sequence in the 5'-UTR itself is involved in the decrease of translational efficiency and the spermidine effect.

DISCUSSION
Mouse ODC mRNA has an unusually long 5'-UTR, which consists of 313 nucleotides (33, 34). The predicted secondary structures of the 5'-UTR may explain why ODC mRNA is difficult to translate in cell-free systems (2). To clarify the role of the 5'-UTR in the translational efficiency of ODC synthesis and the spermidine effect on this synthesis, we synthesized various ODC mRNAs with different sizes of 5'-UTR and constructed an ODC-TK fusion mRNA. From a comparison of the results with mODC188 and mODC188-TK fusion mRNAs (Table I and Fig. 8), we could show that the nucleotide sequence in the 5'-UTR itself is involved in the decrease of translational efficiency and the spermidine regulatory effect. Our data support the previously proposed role of the 5'-UTR in the translational control of ODC synthesis ca.
A comparison of ODC synthetic activity with mODC241, mODC188, mODC85, or mODC24 shows that the decrease in the translational efficiency of ODC synthesis is proportional to the size of the 5'-UTR. The free energy of the potential secondary structures in the first 300 nucleotides are almost the same for the various mRNAs except that AG for mODC241 is only slightly lower (Fig. 2). However, when the free energy of the potential secondary structures in the region between the cap and the initiator AUG was calculated, AG correlated with ODC synthetic activity ( Fig. 2 and Table I). We found that ODC synthetic activity decreases drastically if AG was lower than -50 kcal/mol. This is in good accordance with the results obtained with preproinsulin synthesis (35). These results suggest that the potential secondary structures may be a considerable obstacle for the scanning of 40 S ribosomal subunits from the cap to the initiator AUG, but not much of an obstacle for the running of 80 S ribosomes moving from the initiator AUG to the termination codon during protein synthesis. The same idea has been proposed recently from experimental results regarding the translational efficiency of aldehyde dehydrogenase mRNA (36).
Our results are similar to previous reports that the 5'-UTR region of the thymidine kinase mRNA of the herpes simplex virus (37) and of the porcine pro-opiomelanocortin mRNA (38) reduces translational efficiency but are contrary to the recent observation that deletion of the 5'-UTR region of phosphoglycerate kinase mRNA in yeast decreases translational efficiency (39).
When we used mODC241 as a template, ODC synthetic activity was extremely low, as reported in the experiments with intact ODC mRNA (2, 40). A factor which is necessary for the translation of ODC mRNA may be lacking from the rabbit reticulocyte lysate.
We have defined the M, 32,000 and 39,000 polypeptides observed in fluorography as the nascent polypeptides of ODC for the following reasons. 1) The polypeptides reacted with the polyclonal antibody for ODC. 2) If the polypeptides had been synthesized from the internal AUG instead of the initiator AUG, the translational efficiency would have been extremely low since, as mentioned above, the potential secondary structures from the cap to the internal AUG are thought to form an obstacle for the scanning of 40 S ribosomal subunits. The ratio of M, 39,000 to M, 32,000 polypeptides synthesized increased with the decrease in the size of 5'-UTR. This also indicates that the decrease in the translational efficiency of ODC synthesis is proportional to the size of 5'-UTR and is consistent with the finding that the ratio of nascent polypeptide to intact polypeptide synthesized becomes smaller when protein synthetic activity is high.
Although the inhibition of ODC synthesis by polyamines at the translational level has been well documented (17)(18)(19), strong stimulation of ODC synthesis by a low concentration of spermidine (0.2 mM) has not been a consistent finding. We have examined the spermidine stimulatory effects on the synthesis of protamine, globin, thymidine kinase, ODC, and the total proteins of bovine lymphocytes in a rabbit reticulocyte cell-free system. Among these protein syntheses, the degree of spermidine stimulation of ODC synthesis was greatest. From our overall results, it can be concluded that spermidine regulates ODC synthesis positively at a low concentration (0.2 mM) and negatively at a high concentration (0.6-1 mM). Other investigators recently reported the stimulation of ODC synthesis by a low concentration of spermidine (19,41).
When the size of the 5'-UTR of ODC mRNA increased from 85 to 188 nucleotides, both spermidine stimulation and inhibition were increased greatly (Figs. 4 and 5). A small open reading frame can be seen in this region (Met-Gly-Gln-Ala-Ser-Arg-Ala-Thr-Val-Leu) (Fig. 2). Such a potential leader peptide has been suggested as decreasing the translational efficiency of mRNA in another system (42). There is also a GC-rich region located in the 128-174 nucleotide sequence upstream from the initiator AUG. We do not yet know the precise mechanism(s) of the polyamine effect, but either the open reading frame or the GC-rich region is thought to be involved. A GC-rich region also exists in the 5' terminus of the 5'-UTR of intact ODC mRNA. If we can construct a more efficient translation system for ODC synthesis, we may be able to clarify whether or not the GC-rich region is really related to the polyamine effect. In this regard, we have reported previously that spermine binds more readily to GCrich DNA than to AT-rich DNA (43).
Using a cell culture system, Coffin0 and his co-workers (44) have reported that there is little or no translational regulation of ODC by polyamines. However, in a cell-free system, we have shown that high concentrations of spermidine (0.8-l mM) directly inhibit protein synthesis. Our ODC-TK fusion protein (M, 51,000) encoded by mODC188-TK does not contain PEST sequences, which are involved in the rapid degradation of ODC (5). The effect of spermidine on mODC188and mODClS&TK-directed protein synthesis was almost the same, suggesting that spermidine is involved in the translation of ODC mRNA rather than the degradation of the ODC protein in this cell-free system. More detailed studies will be required to clearly determine how much of the polyamine effect on ODC synthesis is involved in the regulation of the ODC amount in cells, in addition to the polyamine effect on ODC degradation.