METTL14‐mediated upregulation of lncRNA HOTAIR represses PP1α expression by promoting H3K4me1 demethylation in oxycodone‐treated mice

Abstract N6‐methyladenosine (m6A) methylation is a vital epigenetic mechanism associated with drug addiction. However, the relationship between m6A modification and oxycodone rewarding is less well explored. Based on an open field test, the present study evaluated oxycodone rewarding using chromatin immunoprecipitation PCR, immunofluorescence, and RNA sequencing. A marked increase in METTL14 protein and a decrease in PP1α protein due to oxycodone abundance in the striatal neurons were observed in a dose‐ and time‐dependent manner. Oxycodone markedly increased LSD1 expression, and decreased H3K4me1 expression in the striatum. In the open field test, intra‐striatal injection of METTL14 siRNA, HOTAIR siRNA, or LSD1 shRNA blocked oxycodone‐induced increase in locomotor activity. The downregulation of PP1α was also inhibited after treatment with METTL14/HOTAIR siRNA and LSD1 shRNA. Enhanced binding of LSD1 with CoRest and of CoRest with the PP1α gene induced by oxycodone was also reversed by LSD1 shRNA. In addition, H3K4me1 demethylation was also blocked by the treatment. In summary, the investigation confirmed that METTL14‐mediated upregulation of HOTAIR resulted in the repression of PP1α, which in turn facilitated the recruitment of LSD1, thus catalyzing H3K4me1 demethylation and promoting oxycodone addiction.

0][11] In eukaryotic cells, m6A demethylation is catalyzed by demethylases ALKBH5 and FTO, while its methylation is catalyzed by a methyltransferase complex consisting of methyltransferase-like 3 (METTL3) and 14 (METTL14). 12cording to a previous study, aberrant chemical modifications of synaptic mRNAs have a role in the pathogenesis of neuropsychiatric disorders. 13In adult rodents, deficiency of FTO or METTL14 was reported to intervene with neurogenesis, memory, and dopaminergic signaling. 14,15Although there exists some evidence relating to m6A methylation and drug addiction, the role of m6A in oxycodone rewarding is still an open question.
Histone methylation is closely related to transcriptional regulation and is involved in processes that induce drug abuse.Previous studies have shown that chronic cocaine decreases global levels of histone 3 lysine 9 tri-methylation (H3K9me3) in the NAc. 16Repeated morphine downregulates NAc H3K9 di-methylation (H3K9me2). 17K4 demethylation is regulated by the lysine-specific demethylase (KDM) families LSD1 and KDM5, which demethylate H3K4me2/ me1 to H3K4me0 and H3K4me3/me2 to H3K4me1, respectively.18 However, little is known about how oxycodone affects the levels of H3K4 methylation.
In order to uncover the potential epitranscriptomic role of m6A methylation in oxycodone abuse, three different doses of oxycodone were administered to a group of mice for 9 days.Subsequently, after 4 days of oxycodone withdrawal, striatum samples from the experimental mice were collected for RNA-sequencing (RNA-seq) to detect m6A methylation-associated enzymes and then examine m6A-related epigenetic alterations.An upregulation in the levels of METTL14, lncRNA HOX transcript antisense RNA (lncRNA HOTAIR), and lysine demethylase LSD1 was detected in all oxycodone-treated mice.As confirmed by RNA-seq analyses, treatment with siRNA METTL14/HOTAIR or shRNA LSD1 alleviated oxycodone-induced increases in locomotor activity, throwing some light on the potential uses of this new approach in opioid deaddiction.

| Animals
Animal use protocol (2021PS221K) was approved by the Laboratory Animal Care Committee of Shengjing Hospital of China Medical University.For this study, 310 male C57BL/6J mice (18-22 g) were purchased from Beijing HFK Bioscience Co., Ltd.(China).All the mice were housed in an SPF room with a temperature of 23 ± 1°C, humidity maintained at 60 ± 15%, and a 12:12 h light/dark cycle, with access to sufficient food and water.For anesthetic purposes, 5.0% isoflurane for induction and 2.5% isoflurane for maintenance were used.The experimental design of the present study was shown in Figure 1.

| Drug delivery
Three siRNAs targeting METTL14 mRNA or HOTAIR as well as a scrambled nontargeting oligo (scRNA; 5′-TUCUCUTGCTUGUCAUACUTT-3′) were obtained from GenePharma (China).Their transfection efficiency in the primary neurons was assessed.Adeno-associated virus (AAV2) expressing LSD1 shRNA (AAV-LSD1 shRNA-EGFP, 4 μL of 10 12 vg/mL, 200 nL/min) and a negative control (AAV-LSD1 ncRNA-EGFP) were purchased from HANBIO (China).siRNA or shRNA was administered to the mice by intra-striatal injection.Catheterization was performed as follows: two 26-gauge stainless steel guide cannulas were inserted into the striatum bilaterally (AP +0.8 mm, ML ±1.6 mm, DV +3.0 mm) and stabilized on the skull using dental cement.Two dummy cannulas with the same dimensions as guide cannulas were inserted from the top to prevent clogging and infection.After allowing recovery for a week, a 31-gauge injector tube was inserted into the guide cannula, and siRNA or scRNA (1.0 μL per mouse) was delivered 30 min prior to open field tests.In our preliminary experiment, we observed that shRNA required at least 3 weeks to exhibit its functional effects.Conversely, the inhibition effects of siRNA on the expression of HOTAIR were notably faster.
Therefore, the administration paradigm for shRNA and siRNA differs significantly in this study.A previous study on the functional characteristics of siRNA and shRNA has also confirmed the difference in their action times. 19Prior to this, LSD1 shRNA or ncRNA was microinjected into the bilateral striatum (2.0 μL per mouse) 3 weeks before the open field tests.Saline or oxycodone (1.5, 3.0, and 6.0 mg/kg) was injected intraperitoneally (i.p., 0.1 mL/10 g) from day 2 to day 10.

| Open field test
To assess locomotor activity, each mouse from different groups (n = 8 animals in each group) was placed in a square arena (50 cm in length, width, and height) under dim light.Mice were dropped in the corner every day and allowed to move freely in the arena.The arena was cleaned thoroughly using 75% alcohol solution between each trial to remove odor cues.The movement of every individual mouse was recorded for 15 min and analyzed the travel distance of each mouse using video-tracking software (EthoVision XT, Noldus, the Netherlands).

| Co-immunoprecipitation (Co-IP)
After collection, the striatum (n = 4 animals in each group) was lysed using a lysis buffer (Cat# P0013; Beyotime, China), and the lysates were incubated with anti-CoREST antibody for 8 h followed by the addition of A/G Plus-Agarose (Cat# sc-2003; Santa Cruz, USA).The protein-antibody-bead mixtures were incubated at 4°C overnight with rotary agitation.All mixtures were washed using a lysis buffer followed by the addition of 5× SDS loading buffer (Cat# P0015; Beyotime, China).Western blotting was used to detect the expression of KDM1A and CoREST.An IB assay for CoREST was used as a control.

| RNA-sequencing (RNA-seq)
Total RNA was extracted from the striatum (n = 3 animals in each group), and the quantification of lncRNA, ncRNA, and mRNA followed.

For the construction of sequencing libraries, the MGISEQ-2000RS
High-Throughput Sequencing Reagent Kit (Cat#1000012554; MGI, China) was used.For quality checking, the Standard Sensitivity RNA Analysis Kit (DNF-471-0500) and the Fragment Analyzer (Agilent 5300) were employed.SOAPnuke (v1.5.6) was used for filtering the sequencing data.Quality-checked libraries were sequenced on the DNBseq platform using MGISEQ-2000 with 100PE sequencing.
Library construction, lncRNA-seq, ncRNA-seq, and RNA-seq, as well as data collection and mapping were outsourced to HuaDa Gene Biotechnology (Shenzhen, China).

| Quantification of RNA m6A
The EpiQuik m6A RNA Methylation Quantification Kit (Cat# P9005-48; Epigenetek, USA) was used to quantify overall m6A methylation as per manufacturer's instructions.Briefly, 200 ng RNA (n = 4 animals in each group) was transferred to 96 wells followed by the addition of antibody solution.The level of m6A absorbance in each well at a wavelength of 450 nm was evaluated.All computations were performed based on the standard curve.

| Statistical analysis
Data were expressed as mean ± standard error of the mean (SEM).All analyses were carried out using SPSS 13.0 software (SPSS Inc., USA).
The Shapiro-Wilk test was used to test the normality of data variance, and data homogeneity was assessed with Levene's test.When analyzing the results of molecular studies, between-group differences were tested using one-way analysis of variance (ANOVA) with Tukey's post hoc test for multiple comparisons or using Student's unpaired t-test.Results of open field tests were analyzed using twoway ANOVA with repeated measurements followed by Tukey's post hoc test for differences over time.Individual Tukey's post hoc test between groups was only run when the F-value was p < 0.05.The level of statistical significance was set at p < 0.05.

| Oxycodone altered mRNAs, ncRNAs, and lncRNAs levels in the striatum
RNA-seq was used to investigate the pathology of oxycodone abuse.
According to our results, 15 mRNAs, 8 ncRNAs, and 39 lncRNAs were markedly downregulated, while 91 mRNAs, 20 ncRNAs, and 50 lncR-NAs were upregulated after oxycodone injection (Figure 2A).The details of differential gene expression were displayed in Table S1.The Gene Ontology (GO) study suggested the enrichment of the most number of differentially expressed genes in the extracellular region, notably those in secretory granule, synapse, and cholinergic synapse.
The highest rich ratio of GO enrichment was in the catenin-TCF7L2 complex and tethering complex (Figure 2B).The BGI online platform (www.bgi.com) was used to analyze the KEGG (Kyoto Encyclopedia of Genes and Genomes) pathways of different genes.Accordingly, differentially expressed genes were enriched in the chemokine signaling pathway and the hippo signaling pathway.The highest number of differential genes was enriched for the human papillomavirus infection (Figure 2C).It was noted that the expression of METTL14 and PP1α in the striatum was affected by oxycodone injection.Thus, we explored the involvement of m6A methylation in the regulation of PP1α expression under conditions of oxycodone abuse.

| Oxycodone downregulated striatal PP1α underlying m6A methylation
Oxycodone at dosages of 1.5, 3.0, and 6.0 mg/kg was administered to all experimental mice.Compared to the SAL group, the movement of the OXY group was marked elevated (OXY_3.0:p < 0.05; OXY_6.0:p < 0.001, Figure 3A).Thus, it was decided to use oxycodone at a dosage of 6.0 mg/kg in subsequent experiments.The results of Western blotting and RT-qPCR show a significant increase in METTL14 expression and a significant decrease in PP1α expres-

| Blockade of METTL14 or HOTAIR decreased locomotor activity in oxycodone-treated mice
Identifying the role of METTL14 in oxycodone abuse is the next course of our work.To that end, METTL14 siRNAs were prepared and investigated, whose results indicated that siRNA1 (si1) significantly inhibited METTL14 expression (p < 0.001, Figure 4A).Therefore, fur-

| DISCUSS ION
This study on chronic oxycodone exposure has demonstrated that METTL14-mediated lncRNA HOTAIR methylation promoted the Existing evidence shows the connections of m6A methylation and its regulatory proteins with learning and memory. 21,22Although narcotics can lead to profound alterations in the expressions of lncRNAs, mRNAs, and ncRNAs underlying both histone and DNA methylation, 4,23 the potential role of m6A methylation in oxycodone PP1α pathway, marked by m6A methylation.Similarly, studies have demonstrated the impact of FTO on the CREB pathway, confirming its involvement in the locomotor-activating effects of 15,24 The role of METTL14 in the development of major neuropsychiatric diseases such as Huntington's disease and Alzheimer's disease has been well researched. 25,26Therefore, the mechanism of action of METTL14 in regulating the PP1α expression is worth investigating.
Our RNA-seq analysis shows sufficient proof of elevation in the expression of lncRNA HOTAIR in the striatum due to oxycodone addiction.HOTAIR plays a coordinating role in gene expression and activating chromatin-modifying enzymes.It plays a part in gene silencing by facilitating H3K4 demethylation and/or H3K27 methylation via directing LSD1 and/or PRC2 complex to the target genes. 27HOTAIR is also connected with epigenetic regulation, poor survival, and immune escape. 28,29In a recent study, HOTAIR inhibition reduced the expression of a range of key dopamine neuron specification factors, proving the functional role of HOTAIR in dopamine neuron development and differentiation. 30In this study, treatment with HOTAIR siRNA significantly blocked both increases in locomotor activity and decreases in PP1α expression induced by oxycodone, suggesting that HOTAIR is a vital epigenetic regulator of the PP1α protein.In addition, treatment with LSD1 shRNA blocked oxycodone-induced increases in locomotor activity and decreases in PP1α expression.Meanwhile, oxycodone administration significantly elevated the expression of CoRest and reduced the expression of H3K4me1 at the same time, suggesting that LSD1-mediated H3K4me1 demethylation triggered some gene expression via chromatin remodeling.It was reported that H3K4me1 and H3K36me3 have the strongest association with splicing indicating they play a significant role in alternative splicing in brain reward tissue. 31In the study, oxycodone also reduced the expression of KDM5A, which induced the upregulation of H3K4me3/me2 level.
In line with our results, a previous study found the opposite trend of LSD1 and KDM5A in patients with alopecia areata.It shows that alopecia areata increased KDM5A, MLL, SETD7, and G9A expression, as well as reduced LSD1, KDM4A, and KDM4B expression. 32We speculated that H3K4 methylation maintains an equilibrium state, where oxycodone induces demethylation of H3K4me1 and methylation of H3K4me3/me2.Although we can not make the conclusion on which enzyme (LSD1 or KDM5A) plays the leading role, our results showed the status of H3K4 methylation at this time point during oxycodone rewarding.The involvement of the LSD1/CoRest complex in the repression of gene transcription has been recently reported. 33From that perspective, a conclusion can be drawn from our results that oxycodone promoted the binding of CoRest with LSD1 in the striatum.However, these oxycodone-mediated effects were blocked by LSD1 shRNA administration.Now the question that arises: Whether the LSD1/CoRest complex promoted H3K4me1 demethylation and, in turn, repressed PP1α expression?
According to one study, H3K4me2 demethylation at the FosB gene induced morphine-conditioned place preference in rats under chronic stress. 34Similarly, another study suggests LSD1-mediated demethylation of H3K9me2 in the amygdala also resulted in alcohol abuse in adolescents. 35We hypothesize that oxycodone-induced upregulation of the LSD1-CoRest complex through H3K4me1 demethylation in the striatum may induce the binding of the complex with the PP1α gene, triggering the expression of PP1α protein.We ran the ChIP-PCR experiment to measure the binding of the LSD1-CoRest complex with the PP1α gene, which demonstrated that oxycodone administration led to a significant increase in the interaction of the LSD1-CoRest complex with the PP1α gene.Such binding can be blocked by treatment with LSD1 shRNA, indicating that LSD1 is a potent source of defense against oxycodone-induced interaction of the LSD1-CoRest complex with the PP1α gene.

| CON CLUS IONS
The present study brings to light evidence suggesting that METTL14-mediated upregulation of HOTAIR has a potency to repress the expression of PP1α and facilitate the recruitment of LSD1, thus catalyzing H3K4me1 demethylation and promoting oxycodone addiction.