tRNA epitranscriptomic alterations associated with opioid-induced reward-seeking and long-term opioid withdrawal in male mice

Abstract

DNA cytosine methylation has been documented as a potential epigenetic mechanism of transcriptional regulation underlying opioid use disorder. However, methylation of RNA cytosine residues, which would drive another level of biological influence as an epitranscriptomic mechanism of gene and protein regulation has not been studied in the context of addiction. Here, we probed whether chronic morphine exposure could alter tRNA cytosine methylation (m⁵C) and resulting expression levels in the medial prefrontal cortex (mPFC), a brain region crucial for reward processing and executive function that exhibits opioid-induced molecular restructuring. We identified dynamic changes in glycine tRNA (tRNA^Gly_GCC) cytosine methylation, corresponding to altered expression levels of this tRNA at multiple timepoints following 15 days of daily morphine. Additionally, a robust increase in methylation, coupled with decreased expression, was present after 30 days of withdrawal, suggesting that repeated opioid administration produces changes to the tRNA regulome long after discontinuation. Furthermore, forebrain-wide knockout of neuronal Nsun2, a tRNA methyltransferase, was associated with disruption of opioid conditioned place preference, and this effect was recapitulated by regional mPFC Nsun2 knockout. Taken together, these studies provide a foundational link between the regulation of tRNA cytosine methylation and opioid reward and highlight the tRNA machinery as a potential therapeutic target in addiction.

Fig. 1. Dynamic tRNA^Gly expression and methylation following acute morphine and withdrawal in wild-type mice.

a Adult wild-type male mice were injected IP with 10 mg/kg morphine daily for 15 days, followed by 2 h, 24 h, or 30 days of withdrawal and collection of medial prefrontal cortex (mPFC) tissue. See overlap of mPFC region of interest and viral Cre injections used in behavioral experiments. b Schematic of tRNA^Gly_GCC depicting methylated cytosines. c qPCR for tRNA^Gly_GCC expression after a single acute morphine injection (n = 8 saline/6 morphine) or morphine withdrawal (2 h: n = 7 saline/8 morphine, 24 h: n = 7 saline/9 morphine, 30d: n = 8 saline/8 morphine; *p < 0.05). d Bisulfite amplicon sequencing for tRNA^Gly_GCC after 2 h, 24 h, and 30 days of withdrawal. Note the increase in methylation for morphine-treated mice at cytosine 39 at 30 days (t-test; ***p < 0.001). e Left, schematic representation of tRNA^Glu_CTC and cytosines methylated by Nsun2. Right, Bisulfite amplicon sequencing for tRNA^Glu_CTC showed a trending increase in methylation for morphine-treated mice at cytosine 47 (t-test #p < 0.057; n = 8/group). f Left, schematic representation of tRNA^Asp_GTC and cytosines methylated by NSUN2 and DNMT2. Right, Bisulfite amplicon sequencing for tRNA^Asp_GTC showed no changes in methylation at any sites between saline and morphine-treated mice (n = 8/group). g Cytosine 39 methylation at each timepoint shown as a fold change relative to the average of each group’s saline control (***p < 0.001). h qPCR for Nsun2 mRNA levels at all 3 withdrawal timepoints.

Fig. 2. Dynamic changes in morphine-induced tRNA hypomethylation are associated with shifts in Nsun2 and tRNA abundance.

Correlations between tRNA^Gly_GCC methylation and (a) Nsun2 mRNA levels and (b) tRNA^Gly_GCC expression levels. Left, saline groups show no significant correlation or anticorrelation, while the morphine groups (right) show significant relationships between expression levels of tRNA and Nsun2 and C39 methylation.

Fig. 3. Morphine conditioned place preference in Nsun2 conditional knockout mice.

a For CPP, mice received a 20-min pre-test wherein all compartments of the chamber could be explored. Subsequently, mice underwent three 45-min conditioning sessions in which saline was injected in the morning restricted to one side of the chamber (“unpaired”), and 10 mg/kg morphine was injected in the afternoon on the other side (“paired”). Conditioning to morphine was assessed in a subsequent 20-min session with access to both sides of the chamber, with positive conditioning indicated by a higher preference on the morphine-paired side. b Camk2a-Cre Nsun2 KO mice (n = 7 WT, 8 KO) and c PFC AAV-Cre Nsun2 KO (n = 6 AAV-GFP, 7 AAV-Cre KO; see Fig. 1a for AAV localization schematic) mice both show a reduction in time spent in the morphine-paired chamber compared to the unpaired chamber, indicating a blunted reward response to morphine (t-test *p < 0.05).