Translational regulation of GluR2 mRNAs in rat hippocampus by alternative 3' untranslated regions

Abstract

The glutamate receptor 2 (GluR2) subunit determines many of the functional properties of the alpha-amino-3-hydroxy-5-methylisoxazole-4-propionate subtype of glutamate receptor. The roles of untranslated regions (UTRs) in mRNA stability, transport, or translation are increasingly recognized. The 3' end of the GluR2 transcripts are alternatively processed to form a short and long 3'UTR, giving rise to two pools of GluR2 mRNA of 4 and 6 kb in length, respectively, in the mammalian brain. However, the role of these alternative 3'UTRs in GluR2 expression has not been reported. We demonstrate that in the cytoplasm of rat hippocampus, native GluR2 mRNAs bearing the long 3'UTR are mostly retained in translationally dormant complexes of ribosome-free messenger ribonucleoprotein (mRNP), whereas GluR2 transcripts bearing the short 3'UTR are predominantly associated with actively translating ribosomes. One day after pilocarpine-induced status epilepticus (SE), the levels of both long and short GluR2 transcripts were markedly decreased in rat hippocampus. However, GluR2 mRNAs bearing the long 3'-UTRs were shifted from untranslating mRNP complexes to ribosome-containing complexes after SE, pointing to a selective translational derepression of GluR2 mRNA mediated by the long 3'UTR. In Xenopus oocytes, expression of firefly luciferase reporters bearing alternative GluR2 3'UTRs confirmed that the long 3'UTR is sufficient to suppress translation. The stability of reporter mRNAs in oocytes was not significantly influenced by alternative 5' or 3'UTRs of GluR2 over the time period examined. Overall, our findings that the long 3'UTR of GluR2 mRNA alone is sufficient to suppress translation, and the evidence for seizure-induced derepression of translation of GluR2 via the long 3'UTR strongly suggests that a regulatory signaling mechanism exists that differentially targets GluR2 transcripts with alternative 3'UTRs.

Figure 1

Characterization of GluR2 3′UTRs. A. Schematic of known transcriptional and translational regulation domains of GluR2. B, Northern blots of rat hippocampus (H) and cortex (C) mRNAs hybridized to 3′UTR-specific probes that recognize total GluR2 mRNA (open circle ∼4kb) and GluR2 mRNA with long 3′UTR (dark filled circle ∼6kb). C, Subcellular distribution of native GluR2 transcripts in rat hippocampus; supernate and pellet fractions of detergent-treated hippocampal homogenates were analyzed by Q-RT-PCR for their native GluR2 mRNA content and compared to that of untreated total lysate from the hippocampus of same animals.

Figure 2

Differential association of native GluR2 transcripts with ribosomes. A, Fractionation of cytoplasmic extracts of rat hippocampus on a linear sucrose gradient revealed ribosome-free mRNPs (fractions 1 and 2), 80S monosome peak (fraction 3) and translating polyribosomes (fractions 4 to 10). B, The majority of total GluR2 transcripts are detected in ribosomal fractions (both monosomes and polysomes) (fractions 3 to 10) whereas the majority of GluR2 transcripts bearing the longer 3′UTRs are associated mainly with free mRNPs (fractions 1 and 2). C, Treatment of the lysate with EDTA, a Mg²⁺ chelator that disrupts polysome formation, dissociated translating polyribosomes into 40S and 60S ribosomal subunits in the gradient, as expected. D, A major shift of mRNAs (including GAPDH as internal control) from ribosomal to mRNP fractions was observed in EDTA-treated lysates.

Figure 3

Cellular distribution of native GluR2 mRNAs in rat hippocampus. In situ hybridization of digoxigenin-labeled RNA probes specific to GluR2 coding region (pan GluR2) or the long 3′UTR reveal the tissue distribution of GluR2 transcripts in CA1, CA3 and DG regions of hippocampus in control rats (A and B), and pilocarpine-treated animals that experienced status epilepticus 24 hr before (E and F). The arrows in panels E and F indicate increased signal associated with the long 3′UTR of GluR2 in the CA2 region after pilocarpine. The sense probes hybridized to the tissues from the control (C and D) and SE animals (G and H) show only background staining.

Figure 4

Translation of reporter mRNAs bearing alternative GluR2 UTRs. A. Schematic of firefly luciferase reporters bearing alternative combinations of GluR2 5′ and 3′ UTRs. Gray boxes indicate the position of GU repeats on the long 5′UTRs (Myers et al., 2004), and the firefly coding region (ff) common to all constructs. The primers (p) used in Q-RT-PCR recognize all four species of reporter mRNA. B. The quality and amount of each in vitro transcribed reporter mRNA were evaluated with an Agilent bioanalyzer. The expected size of the transcripts is SS: 2760; SL: 4710; LS: 3140; and LL: 5110 bp. C and D, Luciferase activity was proportional to both mRNA amount and time after injection into Xenopus oocytes, for the SS and SL populations of mRNA.

Figure 5

Expression profile of firefly reporter mRNAs bearing alternative GluR2 UTRs in Xenopus oocytes. A, Individual Xenopus oocytes were microinjected with reporter mRNAs (5 fmol/oocyte) and incubated at 17°C. At indicated time points after injection, the oocytes were individually homogenized and firefly reporter expression was detected as recorded luminescence units (RLU). *p<0.05 ANOVA, post-hoc Bonferroni, comparing SS with SL at 16h, 24h and 40h. B, Recovered reporter mRNAs from microinjected-oocytes were quantified by Q-RT-PCR using UTR-specific primers and known quantities of cDNA standards. The apparent increase in LS mRNA levels between 24 and 40 hours was not statistically significant (ANOVA, post-hoc Bonferroni). C, Expression of firefly luciferase protein presented as luciferase activity per fmol mRNA recovered from oocytes [n=10-15 oocytes for each time point from each of five different animals, *p< 0.05, **p< 0.001 comparing SS with SL, ANOVA, pot-hoc Bonferroni. D, Rate of luciferase expression from 4 to 16 hours after injection presented as percent of SS expression, **p< 0.001 ANOVA, post-hoc Bonferroni, comparing SS vs SL, ns (no significant difference) compared for LS vs LL.

Figure 6

Effects of pilocarpine-induced status epilepticus (SE) on native GluR2 transcripts. A. The levels of mRNAs are inversely proportional to the average number of PCR cycles needed to reach detection threshold (CT). The difference between the CT values (ΔCT) of GluR2 transcripts and that of the GAPDH mRNA in both control and SE animals indicates that the levels of both pan-GluR2 mRNA, and GluR2 bearing the long 5′UTR, are reduced by pilocarpine-induced SE (N=8, *P<0.01, ANOVA, Post hoc Bonferroni). B-D. To determine the effects of pilocarpine-induced SE on the association of GluR2 mRNAs with ribosomes in rat hippocampus, polyribosome association of pan GluR2 (B) and GluR2 bearing long 3′UTRs (D) was examined over a sucrose gradient assay. (C) native GluR2 transcripts recovered from free mRNP (fraction 1 and 2) and active ribosomes (fraction 3 to 10) are quantified. The ratio of RNA levels in ribosome-free mRNP and ribosome-containing fractions is shown (n=6, * p< 0.01, ANOVA, post-hoc Bonferroni).