Eight week-old male C57BL/6 were purchased from Charles River Laboratories (France). Male and female Drd2-eGFP heterozygous mice (C57BL/6) were generated as described previously (Gong et al., 2003). Animals were housed under standardized conditions with a 12 h light/dark cycle, stable temperature (22 ± 1°C), controlled humidity (55 ± 10%), and food and water ad libitum. All experiments were in accordance with the guidelines of the French Agriculture and Forestry Ministry for handling animals and were approved by the local Ethic Committee (D34-172-13).

(+)-α-Methylphenethylamine [D-amphetamine (D-amph)] sulfate salt (10 mg/kg) from Sigma–Aldrich (St.-Quentin-Fallavier, France) was dissolved in 0.9% (w/v) NaCl (saline) and injected intraperitoneally (i.p) in a volume of 10 ml/kg, one injection per day during 5 days. Mice were habituated to handling and saline injection for 3 consecutive days before the day of the experiment.

After pharmacological manipulation, the striatum (including the nucleus accumbens and dorsal striatum) of one hemisphere was extracted as previously described (Puighermanal et al., 2016a), sonicated in 300 μl of 10% sodium dodecyl sulfate (SDS), and boiled at 100°C for 10 min. Protein quantification and western blots were performed as described (Biever et al., 2015). Primary antibodies against p-eIF2α (Ser51) (1:1000; Cell Signaling, #3398), eIF2α (1:1000; Cell Signaling, #5324), p-eEF2 (Thr56) (1:1000; Cell Signaling, #2331), p-p70S6K (Thr389) (1:1000; Cell Signaling, #9234), p-4EBP1 (Thr37/46) (1:500; Cell Signaling, #2855), 4EBP1 (1:500; Cell Signaling, #9644), OPHN1 (1:1000; Cell Signaling, #11939), ATF4 (1:1000; NeuroMab, #75-345), MAP2 (1:2000; Sigma, #M4403) from Sigma, CaMKIIa (1:1000; Millipore, #05-532), puromycin [1:1000; (David et al., 2012)], and β-actin (1:40000; Abcam, #AB6276) were used. The optical density of the relevant immunoreactive bands (or for all the bands for puromycin staining) was quantified after acquisition on a ChemiDoc XRS System (Bio-Rad) controlled by Image Lab software version 3.0 (Bio-Rad).

After pharmacological manipulation, tissue preparation and immunofluorescence were performed as described (Bertran-Gonzalez et al., 2008). Primary antibodies against GFP (1:1000; Life Technologies, #A10262), Calbindin-D28k (1:1000; Swant, #CB38), and Arc/Arg3.1 (1:500; Santa Cruz Biotechnology, #sc17839) were used.

Puromycin incorporation was performed as described previously (Biever et al., 2015). Briefly, striata were rapidly dissected on an ice-cooled dish and homogenized using 20 up-and-down strokes of a prechilled glass homogenizer with 800 μl of polysomal buffer containing 50 mM Tris pH 7.8, 240 mM KCl, 10 mM MgCl₂, 250 mM D-sucrose, 2% Triton X-100, 20 μl/ml emetine, 5 mM DTT, 100 U/ml RNasin (Promega), and protease inhibitor cocktail (Roche). Samples were centrifuged for 5 min at 16,000 × g at 4°C and supernatant was incubated with 100 μg/ml of puromycin for 10 min at 4°C and then boiled for 10 min at 100°C. Protein concentrations were determined using BCA protein assay (Pierce, Rockford, IL, USA) and samples were stored at -20°C for further western blot analyses.

The polysome profiling approach was performed as described previously (Biever et al., 2015). RNA from fractions <2 ribosomes (referred to as non-polysomal ‘NP’) and fractions with ≥2 ribosomes (referred to as polysomal ‘P’) was extracted using the TRIZOL (Thermo Fischer) protocol according to the manufacturer’s instructions. The carrier glycoblue was added before RNA precipitation step during the TRIZOL protocol. To remove potential DNA contamination, fractions were treated with DNAse (Ambion) according to the manufacturer’s instruction. RNA integrity was tested using Fragment Analyzer (Advanced Analytical).

RNA from non-polysomal and polysomal fractions was reverse transcribed to first strand cDNA using the SuperScript^® VILO^TM cDNA synthesis kit (Invitrogen). Resulting cDNA was used for quantitative real-time PCR (qRT-PCR), using 2X SYBR Green Mix and LC480 Real-Time PCR System (Roche) as described (Puighermanal et al., 2016b). Analysis was performed using LightCycler^® 480 Software (Roche). Results are presented as linearized Cp-values normalized to the stably-expressed genes β-actin or gapdh and the ΔCP method was used to give the fold change. The primer sequences used in this study are detailed in Table 1.

All statistical analyses were performed using one-way analysis of variance (ANOVA) for multiple comparisons, followed by Bonferroni post hoc test. Student t-test with equal variances was used for groups of two, when relevant. Statistical significance was determined as p < 0.05. Prism 5.0 software was used to perform statistical analyses.

To investigate whether repeated exposure to D-amphetamine (10 mg/kg, once daily for 5 days) could alter global mRNA translation in the striatum, we performed polysome profile analysis on striatal lysates at 60 min following the last injection of D-amphetamine. We observed an increase in the amplitude of the ‘vacant’ 80S monosome peak along with a reduction in the polysome population in mice treated with D-amphetamine compared to saline-treated mice (Figure 1A). Consequently, the polysome to monosome ratio was significantly decreased in D-amphetamine-treated mice (Figure 1A, inset).

Polysome profiling is a representation of the steady-state ribosomes engaged in translation. To determine whether repeated exposure to D-amphetamine could modulate de novo protein synthesis, we performed an assay adapted from the ribopuromycylation method (David et al., 2012; Biever et al., 2015). We found a transient decrease in puromycin incorporation in striatal lysates of mice treated with D-amphetamine (Figure 1B). Altogether, these results indicate that, in the striatum, global mRNA translation was decreased in mice repeatedly exposed to D-amphetamine.

The regulation of mRNA translation is tightly controlled by the phosphorylation of translation initiation and elongation factors (Buffington et al., 2014). We therefore analyzed the phosphorylation state of the initiation factor eIF2α and the elongation factor eEF2 in the striatum of mice repeatedly exposed to D-amphetamine. Western blot analysis of whole striatal lysates at different time points (15, 30, 60, and 120 min) following the last D-amphetamine administration revealed a robust enhancement of pS51-eIF2α and pT56-eEF2 at 30 and 60 min post-injection, respectively (Figure 1C). The increase in pS51-eIF2α and pT56-eEF2, which was not observed following a single D-amphetamine exposure (Table 2), does not result from an accumulation of phosphorylation over the 5 days of the injection protocol as no change was found at earlier time points (e.g., 15 min post-injection for pS51-eIF2α; 15 and 30 min post-injection for pT56-eEF2; Figure 1C). The mTORC1 pathway, known to regulate the initiation of translation (Ma and Blenis, 2009), was not activated following repeated D-amphetamine exposure (Figure 1D). Together, these results revealed that the decrease in global mRNA translation observed in mice repeatedly administered with D-amphetamine is paralleled by an enhanced eIF2α and eEF2 phosphorylation.

Enhanced eIF2α phosphorylation is observed in response to ER stress (Walter and Ron, 2011). To test whether repeated D-amphetamine exposure promoted ER stress in the striatum, the levels of ER stress markers including PERK (PKR-like ER kinase), PDI (protein disulfide isomerase), IRE1α (inositol requiring enzyme 1 α), and BiP (Binding immunoglobulin protein) were analyzed 30, 60, or 120 min following the last D-amphetamine administration. While striatal levels of PERK, PDI, and IRE1α were not altered by D-amphetamine (Table 3), BiP levels were transiently decreased only 60 min after the last D-amphetamine administration (Table 3).

To determine whether repeated D-amphetamine exposure produced apoptosis or gliosis, we analyzed the levels of PARP [Poly (ADP-ribose) polymerase], cleaved caspase 3, cleaved lamin A/C, and GFAP (glial fibrillary acid protein). As summarized in Table 3, none of these markers were changed in mice repeatedly treated with D-amphetamine. Moreover, no persistent damage of dopamine terminals was observed as suggested by the stable levels of tyrosine hydroxylase between saline- (100 ± 4) and D-amphetamine-treated (30 min, 95 ± 2; 60 min, 102 ± 16; 120 min, 93 ± 4) mice. Altogether, these results indicate that the increased eIF2α phosphorylation observed in the striatum following repeated D-amphetamine exposure is not accompanied by ER stress or neurotoxicity.

Enhanced eIF2α phosphorylation represses global translation but coincidently promotes the translation of mRNAs containing uORFs in their 5′UTRs (Dever, 2002; Buffington et al., 2014). To determine whether repeated exposure to D-amphetamine could increase the translation of specific uORF-bearing mRNAs, we performed polysome profiling combined with qRT-PCR to analyze mRNA levels in the non-polysomal fraction (poorly or not translated mRNAs) and in the polysomal fraction (actively translated mRNAs) in saline and D-amphetamine-treated mice. To correct for the change in the mRNA abundance in the monosomal fraction, mRNA levels in polysomal fractions were systemically normalized to levels in non-polysomal fractions (P/NP ratio). We first analyzed the level of Atf4 and Ophn1, two mRNAs encoding for the activating transcription factor 4 (ATF4) and oligophrenin 1 (OPHN1), respectively, in polysomal fractions from whole striatal lysates. As shown in Figures 2A,B, Atf4 and Ophn1 mRNAs were not enriched in polysomal fractions suggesting that repeated D-amphetamine exposure did not enhance the translation of Atf4 and Ophn1 within the striatum. In line with these results, ATF4 and OPHN1 protein levels, analyzed 60 and 120 min following the last administration of D-amphetamine, remained unchanged (Figures 2C,D). The analysis of Ppp1r15a and Ddit3 mRNAs, two other uORF-bearing mRNAs encoding for the protein phosphatase regulatory subunit 15A and the DNA damage-inducible transcript 3, respectively, revealed an enrichment in polysomal fractions of D-amphetamine-treated mice indicating that these two mRNAs were translationally upregulated (Figures 2E,G). However, while Ppp1r15a mRNA was also enriched in non-polysomal fractions suggesting that Ppp1r15a was both transcriptionally and translationally enhanced (Figure 2F), Ddit3 mRNA was only enriched in polysomal fractions (Figure 2H). Altogether, these results indicate that repeated D-amphetamine exposure represses global protein synthesis but concomitantly upregulates the translational efficiency of selective uORF-bearing mRNAs.

Besides the slowing of elongation, eEF2 phosphorylation at T56 increases the translation of dendrite-localized mRNAs such as Camk2a and Arc/Arg3.1 (Scheetz et al., 2000; Park et al., 2008; Kenney et al., 2016). We therefore tested whether the enhanced eEF2 phosphorylation induced by D-amphetamine was accompanied by increased binding of Camk2a, Map2, and Arc/Arg3.1 mRNAs to polysomes. No changes were observed in the abundance of Map2 mRNA in non-polysomal and polysomal fractions between saline and D-amphetamine-treated mice (Figure 3A). In addition, striatal MAP2 protein levels remained unchanged at 60 and 120 min following the last administration of D-amphetamine (Figure 3B). The analysis of Camk2a mRNAs revealed an enrichment in both non-polysomal and polysomal fractions from mice repeatedly treated with D-amphetamine (Figure 3C). However, no significant changes were found in the Camk2a P/NP ratio (saline = 1.8 ± 0.58, D-amphetamine = 1.6 ± 0.07; p = 0.14) as well as at the protein level (Figure 3D). On the other hand, although Arc/Arg3.1 mRNAs were also enriched in both non-polysomal and polysomal fractions (Figure 3E), we observed a threefold increase in P/NP ratio of Arc/Arg3.1 mRNA levels (Figure 3F). To determine whether the enrichment of Arc/Arg3.1 mRNAs in the polysomal fractions was accompanied by an increase at the protein level, we monitored Arc/Arg3.1-positive cells 60 min after the last administration of saline or D-amphetamine. Immunofluorescence analysis revealed a robust increase in Arc/Arg3.1 immunoreactivity in the dorsal striatum and the nucleus accumbens corresponding to both Arc/Arg3.1-positive neurons and neuropil (most presumably dendritic processes; Figures 3G,H). This increase occurred preferentially in striosomes/patches (identified as calbindin-D28k-poor zones) and only sparsely in the matrix compartment (Figure 3I). Finally, double immunofluorescence analysis of Arc/Arg3.1 and GFP in Drd2-eGFP mice showed that D-amphetamine evoked Arc/Arg3.1 expression predominantly in GFP-negative neurons, presumably accounting for D1R-expressing neurons (Figures 3J,K). Together, our results suggest that repeated D-amphetamine administration enhances Arc/Arg3.1 mRNA translation preferentially in a subset of striosomal D1R-expressing neurons.