FTO participates in hemorrhage-induced thalamic pain by stabilizing toll-like receptor 4 expression in thalamic neurons

Abstract

Background and Purpose:

Hemorrhage-caused gene changes in the thalamus likely contribute to thalamic pain genesis. RNA N⁶-methyladenosine (m⁶A) modification is an additional layer of gene regulation. Whether fat-mass and obesity-associated protein (FTO), an m⁶A demethylase, participates in hemorrhage-induced thalamic pain is unknown.

Methods:

Expression of Fto mRNA and protein was assessed in mouse thalamus after hemorrhage caused by microinjection of type IV collagenase (Coll IV) into unilateral thalamus. Effect of intraperitoneal administration of meclofenamic acid (MA; a FTO inhibitor) or microinjection of adeno-associated virus-5 expressing Cre (AAV5-Cre) into the thalamus of Fto^fl/fl mice on the Coll IV microinjection-induced Toll-like receptor 4 (TLR4) upregulation and nociceptive hypersensitivity was examined. Effect of thalamic microinjection of AAV5 expressing Fto (AAV5-Fto) on basal thalamic TLR4 expression and nociceptive thresholds was also analyzed. Additionally, level of m⁶A in Tlr4 mRNA and its binding to FTO or YTH N⁶-methyladenosine RNA binding protein 2 (YTHDF2) were observed.

Results:

FTO was detected in neuronal nuclei of thalamus. Level of FTO protein, but not mRNA, was time-dependently increased in the ipsilateral thalamus on days 1-14 after Coll IV microinjection. Intraperitoneal injection of MA or AAV5-Cre microinjection into Fto^fl/fl mouse thalamus attenuated the Coll IV microinjection-induced TLR4 upregulation and tissue damage in the ipsilateral thalamus and development and maintenance of nociceptive hypersensitivities on the contralateral side. Thalamic microinjection of AAV5-Fto increased TLR4 expression and elicited hypersensitivities to mechanical, heat and cold stimuli. Mechanistically, Coll IV microinjection produced an increase in FTO binding to Tlr4 mRNA, an FTO-dependent loss of m⁶A sites in Tlr4 mRNA and a reduction in the binding of YTHDF2 to Tlr4 mRNA in the ipsilateral thalamus.

Conclusion:

Our findings suggest that FTO participates in hemorrhage-induced thalamic pain by stabilizing TLR4 upregulation in thalamic neurons. FTO may be a potential target for the treatment of this disorder.

Introduction

Stroke is a leading cause of disability or death among aging population¹. It significantly impacts the patients’ quality of life and leads to a marked economic burden to the patients’ families. Although only about 8-18% of the strokes are caused by brain hemorrhage, the rate in mortality associated with hemorrhagic stroke is much higher than that after ischemic stroke². Besides functional losses in cognition and locomotor after stroke, approximately 8-14% of hemorrhagic victims suffer from refractory central poststroke pain (CPSP)³. The treatments for CPSP are very limited as current medications are ineffective and/or produce severe side effects⁴. Therefore, understanding the mechanisms of how CPSP is caused may open a new avenue in the management of this disorder.

Following stroke, there were a large number of differentially expressed genes in brain^{5, 6}. These dysregulated genes likely contribute to CPSP genesis. For example, toll-like receptor 4 (TLR4) was significantly increased in damaged brains after ischemia or hemorrhage⁷. Suppression or blockage of TLR4 improved brain damage and blocked mechanical allodynia after carotid artery occlusion^{7, 8}. However, how these CPSP-associated genes are increased after stroke remains incompletely understood.

Recent studies indicate that the mechanism for gene regulation involves N6-methyladenosine (m⁶A)-mediated RNA modification. The m⁶A on RNAs is reversibly catalyzed by a protein complex (writers) including methyltransferase-like 3 and 14 (METTL3 and METTL14) and Wilms’ tumor 1-associating protein (WTAP) and removed by two demethylases (erasers) fat-mass and obesity-associated protein (FTO) and AlkB homolog 5 (ALKBH5)⁹. RNA m⁶A is specifically recognized by m⁶A-bidning proteins (readers) such as YTH N⁶-Methyladenosine RNA Binding Proteins 1/2/3 (YTHDF1/2/3), which mediates divergent roles of m⁶A in the RNA metabolism^10-12. In particular, YTHDF2 rapidly degrades its binding RNAs¹³. We recently reported that increased FTO contributes to nerve injury-induced nociceptive hypersensitivity through stabilizing G9a expression in injured neurons of dorsal root ganglion¹¹. Whether FTO also participates in CPSP genesis is unknown. In the present study, we evaluated its role in hemorrhage-induced thalamic pain.

Methods

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Animal preparation

The usage of adult (7-8 weeks) male CD1 mice and Fto^fl/fl mice (C57BL/6J background) was approved by the Animal Care and Use Committee of Rutgers New Jersey Medical School (Protocol #: PROTO999900990). We adhered to the AHA Journals' implementation of the Transparency and Openness Promotion (TOP) Guidelines. Details are described in the Data Supplement.

Hemorrhage-Induced Thalamic Pain Model

Hemorrhage-induced thalamic pain was established as described in our previous reports¹⁴. Details are described in the Data Supplement.

Behavioral Tests

Mechanical, heat and cold tests as well as locomotor function test were carried out as described previously¹⁵. There was a one-hour interval between two tests. Details are described in the Data Supplement.

Cell Line Culture and Transfection

The CAD cells were cultured and transfected by the plasmids as described previously¹⁵. Details are described in the Data Supplement.

TUNEL Staining

Cell death was detected using an in situ cell death detection kit for terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end labeling (TUNEL, EMD Millipore). Details are described in the Data Supplement.

Real-Time Polymerase Chain Reaction

The levels of the RNAs were measured with the SYBR-green method. Single-cell real-time RT-PCR assay was carried out as described previously¹⁵. Details are described in the Data Supplement.

Plasmid Constructs and Virus Production

The pro-viral plasmids were constructed and the AAV5 particles were packaged as described previously¹⁵. Details are described in the Data Supplement.

Western Blotting and Immunofluorescence Analyses

Western Blotting and immunofluorescence were evaluated, as described previously^{11, 14, 15}. The specificity of the primary antibodies used has been reported previously^{11, 14, 16-18}. Details are described in the Data Supplement.

RNA Immunoprecipitation (RIP) Assay

The RIP assay was performed using the Magna RIP Kit or the Magna MeRIP™ m6A Kit (Upstate/ EMD Millipore). Details are described in the Data Supplement.

Statistical Analysis

Statistical analysis was performed by using Graph Pad Prism 8.0 or Microcal Origin 2019 software. Normality and variance homogeneity of the data were assessed across all groups by the Shapiro-Wilk test/D'Agostino K squared test and the Levene's test, respectively. Parametric data were presented as mean ± SD and nonparametric data were reported as median and interquartile range from at least 3 independent experiments. Parametric data were analyzed by two-tailed, unpaired Student’s t test or a one-way or two-way ANOVA followed by the post hoc Tukey test for multiple comparisons, while nonparametric data were analyzed by Scheirer-Ray-Hare test. Differences were considered significant with a P < 0.05. Details are described in the Supplemental materials.

RESULTS

FTO is increased in the ipsilateral thalamus after thalamic hemorrhage

Unilateral microinjection of Coll IV, but not saline, produced significant and sustained mechanical allodynia, heat hyperalgesia and cold hyperalgesia, which occurred 1 day after microinjection and persisted for at least 14 days after microinjection on the contralateral (not ipsilateral) side (Supplemental Fig. IA-D). This microinjection increased the expression of FTO protein, but not METTL3, METTL14, WTAP, ALKBH5, YTHDF1 and YTHDF2 proteins, in the whole ipsilateral thalamus (Fig. 1A-C). Interestingly, basal expression of Fto mRNA in the ipsilateral thalamus was not significantly altered after Coll IV or saline microinjection during observation period (Fig. 1D).

Figure 1.

Thalamic FTO protein was increased in collagenase IV (Coll IV)-induced thalamic pain model in male mice. (A) Level of FTO protein in the ipsilateral thalamus at different days after saline or Coll IV microinjection. n = 5 mice/group/time point. Statistical significance was calculated with two-way ANOVA (model × time) followed by post hoc Tukey test. F_{model (1, 40)} = 111.597, p < 0.001; F_{time (4, 40)} = 6.031, p < 0.001; F_{model × time (4, 40)} = 7.403, p < 0.001. (B, C) Expression of YTHDF1, YTHDF2, METTL3, METTLE14, WTAP and ALKBH5 in the ipsilateral thalamus at different days after Coll IV microinjection. n = 5 mice/time point. Statistical significance was calculated with one-way ANOVA followed by post hoc Tukey test. YTHDF1: F _{(4, 20)} = 0.0694, p = 0.991. YTHDF2: F _{(4, 20)} = 0.344, p = 0.845. METTL3: F _{(4, 20)} = 1.881, p = 0.153. METTL14: F _{(4, 20)} = 0.425, p = 0.789. WTAP: F _{(4, 20)} = 0.503, p = 0.734. ALKBH5: F _{(4, 20)} = 0.1, p = 0.981. (D) Level of Fto mRNA in the ipsilateral thalamus at different days after Coll IV or saline microinjection. n = 5 mice/group/time point. Statistical significance was calculated with two-way ANOVA (model × time) followed by post hoc Tukey test. F_{model (1, 40)} = 1.664, p = 0.205; F_{time (4, 40)} = 0.429, p = 0.787; F_{model × time (4, 40)} = 0.103, p = 0.981. (E) Number of the cells labeled by FTO in the ipsilateral thalamus 7 days after Coll IV or saline microinjection. n = 5 mice/group. Statistical significance was calculated with two-tailed unpaired Student’s t-test. t ₍₁₀₎ = 10.035. Scale bar: 50 μm.

Distribution pattern of FTO in the thalamus was also examined. The triple labeling assay revealed that FTO co-expressed with NeuN (a specific neuronal marker) in the nuclei (labeled by DAPI) of individual cells and was not detected in the cellular nuclei of GFAP (a marker for astrocyte)- or Iba1 (a marker for microglia)-labeled cells in the region around/ adjacent to the core of hemorrhagic lesions in VPM and VPL 7 days after Coll IV microinjection (Supplemental Fig. II and Fig. III). The number of FTO-labeled neurons in this thalamic region on day 7 post-Coll IV microinjection was increased by 2.66-fold as compared to that after saline microinjection (Fig. 1E).

Effect of pharmacological inhibition of FTO on the hemorrhage-induced thalamic pain and damage

Meclofenamic acid (MA; 1, 5 or 10 mg/kg; dissolved in PBS), a specific inhibitor of FTO^{19, 20} or vehicle (PBS) were intraperitoneally administered 30 min before Coll IV or saline microinjection and once a day for five days in male CD1 mice. Mechanical allodynia, heat hyperalgesia and cold hyperalgesia were developed on the contralateral (but not ipsilateral) side of the vehicle plus Coll IV-treated group on days 1-5 after microinjection (Fig. 2A-E). These pain hypersensitivities were markedly ameliorated in the MA (10 mg/kg) plus Coll IV-treated mice (Fig. 2A-C). MA’s effects were dose-dependent (Supplemental Fig. IVA-D). Moreover, intraperitoneal administration of MA (10 mg/kg), but not vehicle, once a day for 5 days starting 1 day after Coll IV microinjection significantly mitigated these pain hypersensitivities on the contralateral (not ipsilateral) side on days 3 and 5 after Coll IV microinjection (Supplemental Fig. VA-D). All treated mice displayed normal locomotor activities (Supplementary Table I).

Figure 2.

Effect of intraperitoneal administration of meclofenamic acid (MA) on Coll IV microinjection-induced thalamic pain development in male mice. MA (10 mg/kg) or vehicle was given 30 min before microinjection and once daily for 5 days post-microinjection. Paw withdrawal frequencies to 0.07 g and 0.4 g von Frey filaments (A, D) and paw withdrawal latencies to heat (B, E) and cold (C) stimuli were examined on days 1, 3 and 5 after microinjection on the contralateral (A-C) and ipsilateral (D, E) sides. n = 10 mice/group. The p values on top of blue triangles indicate a difference between the vehicle and MA groups in Coll IV mice at the corresponding time points. Statistical significance was calculated with two-way ANOVA (time × treatment) with repeated measures followed by post hoc Tukey test. (A) For 0.07 g, F_{time (3, 144)} = 43.958, p < 0.001; F_{treatment (3, 144)} = 150.666, p < 0.001; F_{time × treatment (9, 144)} = 20.398, p < 0.001. For 0.4 g, F_{time (3, 144)} = 15.887, p < 0.001; F_{treatment (3, 144)} = 57.782, p < 0.001; F_{time × treatment (9, 144)} = 7.951, p < 0.001. (B) F_{time (3, 144)} = 160.778, p < 0.001; F_{treatment (3, 144)} = 419.651, p < 0.001; F_{time × treatment (9, 144)} = 48.742, p < 0.001. (C) F_{time (3, 144)} = 120.155, p < 0.001; F_{treatment (3, 144)} = 307.617, p < 0.001; F_{time × treatment (9, 144)} = 46.073, p < 0.001. (D) For 0.07 g, F_{time (3, 144)} = 0.63, p = 0.597. F_{treatment (3, 144)} = 0.185, p = 0.906; F_{time × treatment (9, 144)} = 0.704, p = 0.705. For 0.4 g, F_{time (3, 144)} = 0.0652, p = 0.978; F_{treatment (3, 144)} = 0.0652, p = 0.978; F_{time × treatment (9, 144)} = 0.261, p = 0.984. (E) F_{time (3, 144)} = 0.373, p = 0.772. F_{treatment (3, 144)} = 0.351, p = 0.788; F_{time × treatment (9, 144)} = 0.568, p = 0.822.

TUNEL staining were carried out for the hemorrhage-induced cell death after behavioral tests on day 5 post-microinjection. As shown in Supplementary Fig. VI, many TUNEL-positive cells in the ipsilateral VPL and VPM were seen in the Coll IV plus vehicle-treated group and none of them in the saline plus vehicle- or MA-treated groups. Number of TUNEL-positive cells was dramatically reduced in the ipsilateral VPL and VPM of the Coll IV plus MA-treated group as compared to that in the Coll IV plus vehicle-treated group (Supplementary Fig. VI). Many cells that are positive for Iba1, GFAP, NeuN or CD68 (a marker for microphages and monocytes) were detected in the ipsilateral VPL and VPM of the Coll IV plus MA-treated group (Supplementary Fig. VII).

Effect of genetic knockdown of thalamic FTO on the hemorrhage-induced thalamic pain and damage

A major limitation of systemic MA administration is the lack of anatomical and pharmacological specificity. To this end, we knocked down thalamic FTO through microinjection of AAV5-Cre into unilateral thalamic VPL and VPM of male Fto^fl/fl mice 5 weeks before Coll IV or saline microinjection into same region (Supplementary Fig. VIIIA). AAV5-Gfp was used as a negative control. Consistent with the previous study¹⁵, GFP-labeled AAV5 transfected predominantly into thalamic neurons, as GFP co-existed with NeuN, but not Iba1 and GFAP, in the region around/adjacent to the core of hemorrhagic lesions in the ipsilateral thalamus on day 7 post-Coll IV microinjection (Supplementary Fig. VIIIB). The Coll IV microinjection-induced increase in the level of FTO protein was significantly blocked in whole ipsilateral thalamus of the AAV5-Cre-pre-microinjected male Fto^fl/fl mice 7 days after Coll IV microinjection (Fig. 3A). Unexpectedly, male Fto^fl/fl mice pre-microinjected with AAV5-Cre failed to alter basal level of FTO protein in the ipsilateral thalamus 7 days after saline microinjection (Fig. 3A). In line with behavioral observations above, significant mechanical allodynia, heat hyperalgesia and cold hyperalgesia were seen on days 1, 3 and 7 post-Coll IV microinjection on the contralateral (not ipsilateral) side of the AAV5-Gfp-premicroinjected male Fto^fl/fl mice (Fig. 3B-F). However, these pain hypersensitivities were markedly reduced in the Coll IV male Fto^fl/fl mice pre-microinjected with AAV5-Cre (Fig. 3B-D). Similar behavioral responses were observed in the female Fto^fl/fl mice (Supplementary Fig. IXA-G).

Figure 3.

Effect of thalamic pre-microinjection of AAV5-Cre on Coll IV microinjection-induced FTO protein expression and thalamic pain genesis in male Fto^fl/fl mice. (A) Representative Western blots (left) and statistical summary of the densitometric analysis (right) on day 7 post-Coll IV or saline microinjection. n = 5 mice/group. Statistical significance was calculated with two-way ANOVA (treatment × model) followed by post hoc Tukey test. F_{treatment (1, 16)} = 45.678, p < 0.001; F_{model (1, 16)} = 46.617, p < 0.001; F_{treatment × model (1, 16)} = 32.469, p < 0.001. (B-F) Paw withdrawal frequencies to 0.07 g and 0.4 g von Frey filaments (B, E) and paw withdrawal latencies to heat (C, F) and cold (D) stimuli on the contralateral (B-D) and ipsilateral (E, F) sides on days −35, 1, 3 and 7 after Coll IV or saline microinjection. n = 10 mice/group. The p values on top of blue triangles indicate a difference between the AAV5-Gfp and AAV5-Cre groups in Coll IV-microinjected mice at the corresponding time points. Statistical significance was calculated with two-way ANOVA (treatment × time) with repeated measures followed by post hoc Tukey test. (B) For 0.07 g, F_{treatment (3, 180)} = 140.742, p < 0.001; F_{time (4, 180)} = 81.826, p < 0.001; F_{treatment × time (12, 180)} = 27.861, p < 0.001. For 0.4 g, F_{treatment (3, 180)} = 110.388, p < 0.001; F_{time (4, 180)} = 65.624, p < 0.001; F_{treatment × time (12, 180)} = 22.932, p < 0.001. (C) F_{treatment (3, 180)} = 410.233, p < 0.001; F_{time (4, 180)} = 198.524, p < 0.001; F_{treatment × time (12, 180)} = 68.038, p < 0.001. (D) F_{treatment (3, 180)} = 201.154, p < 0.001; F_{time (4, 180)} =117.912, p < 0.001; F_{treatment × time (12, 180)} = 44.077, p < 0.001. (E) For 0.07 g, F_{treatment (3, 180)} = 0.063, p = 0.979; F_{time (4, 180)} = 0.155, p = 0.961; F_{treatment × time (12, 180)} = 0.202, p = 0.998. For 0.4 g, F_{treatment (3, 180)} = 0.322, p = 0.809; F_{time (4, 180)} = 1.021, p = 0.398; F_{treatment × time (12, 180)} = 0.23, p = 0.997. (F) F_{treatment (3, 180)} = 1.694, p = 0.17; F_{time (4, 180)} = 1.489, p = 0.207; F_{treatment × time (12, 180)} = 0.503, p = 0.911.

The effect of thalamic FTO knockdown on the Coll IV-induced hemorrhagic lesion was also examined. As expected, many TUNEL-positive cells were seen in the ipsilateral VPL and VPM of the AAV5-Gfp-treated Coll IV mice, whereas none of TUNEL-positive cells were detected in the same region of the saline plus AAV5-Gfp- or AAV5-Cre-treated groups (Supplementary Fig. X). Number of TUNEL-positive cells was markedly decreased in the ipsilateral VPL and VPM of the AAV5-Cre plus Coll IV-treated Fto^fl/fl mice as compared to that in the AAV5-Gfp plus Coll IV-treated Fto^fl/fl mice (Supplementary Fig. X).

Thalamic overexpression of FTO led to thalamic pain symptoms

The effect of thalamic FTO overexpression through thalamic microinjection of AAV5 expressing full-length Fto mRNA (AAV5-Fto) into unilateral VPL and VPM on basal nociceptive thresholds in naive adult male mice was observed. AAV5-Gfp was used as a control. Dramatic increases in the amounts of FTO mRNA and protein were detected in the injected thalamus 5 weeks after AAV5-Fto microinjection as compared to those after AAV5-Gfp microinjection (Fig. 4A-B). Mice microinjected with AAV5-Fto, but not AAV5-Gfp, displayed significant increases in paw withdrawal frequencies in response to 0.07 g and 0.4 g von Frey filament stimuli and marked decreases in paw withdrawal latencies in response to thermal and cold stimuli on the contralateral (not ipsilateral) side (Fig. 4C-F; Supplementary Fig. XI). These nociceptive hypersensitivities occurred 3 weeks post-microinjection and persisted for at least 5 weeks (Fig. 4C-F), consistent with the 3–4 weeks lag period of AAV5 expression¹⁵. Intraperitoneally injection of MA at 10 mg/kg (not vehicle) once a day for 5 days starting 5 weeks after viral microinjection significantly blocked these nociceptive hypersensitivities on days 3 and 5 post-MA injection (Supplementary Fig. XII). As expected, systemic MA administration did not affect locomotor function (Supplementary Table I).

Figure 4.

Effect of thalamic FTO overexpression through AAV5-Fto microinjection into unilateral thalamus on nociceptive thresholds in naïve male mice. (A-B) Levels of Fto mRNA (A) and FTO protein (B) in the ipsilateral thalamus 5 weeks after viral microinjection. n = 5 mice/group/assay. Statistical significance was calculated with two-tailed, independent Student’s t test. t ₍₈₎ = −10.509 in A and −8.993 in B. (C-F) Paw withdrawal frequencies to 0.07 g (C) and 0.4 g (D) von Frey filament stimuli and paw withdrawal latencies to thermal (E) and cold stimuli (F) on both ipsilateral (Ipsi) and contralateral (Con) sides at the different weeks after microinjection. n = 10 mice/group. The p values on top of red diamonds indicate a difference from the baseline (0 w) of the corresponding group. Statistical significance was calculated with two-way ANOVA (treatment × time) with repeated measures followed by post hoc Tukey test. (C) Con: F_{treatment (1, 90)} = 105.313, p < 0.001; F_{time (4, 90)} = 24.397, p < 0.001; F_{treatment × time (4, 90)} = 20.603, p < 0.001. Ipsi: F_{treatment (1, 90)} = 0.0545, p = 0.816; F_{time (4, 90)} = 0.0545, p = 0.994; F_{treatment × time (4, 90)} = 0.0545, p = 0.994. (D) Con: F_{treatment (1, 90)} = 109.442, p < 0.001; F_{time (4, 90)} = 22.753, p < 0.001; F_{treatment × time (4, 90)} = 18.662, p < 0.001. Ipsi: F_{treatment (1, 90)} = 0, p = 1; F_{time (4, 90)} = 0.563, p = 0.69; F_{treatment × time (4, 90)} = 0.256, p = 0.905. (E) Con: F_{treatment (1, 90)} = 1045.843, p < 0.001; F_{time (4, 90)} = 199.553, p < 0.001; F_{treatment × time (4, 90)} = 213.138, p < 0.001. Ipsi: F_{treatment (1, 90)} = 0.381, p = 0.539; F_{time (4, 90)} = 0.759, p = 0.555; F_{treatment × time (4, 90)} = 0.736, p = 0.57. (F) F_{treatment (1, 90)} = 396.824, p < 0.001; F_{time (4, 90)} = 113.187, p < 0.001; F_{treatment × time (4, 90)} = 104.324, p < 0.001.

Increased FTO involves in the hemorrhage-induced TLR4 upregulation in the thalamus

Given that upregulated TLR4 in hemorrhagic brains contributed to mechanical allodynia after stroke^{7, 8}, we predicted that increased thalamic FTO might involve in this upregulation. Indeed, pre-microinjection of AAV5-Cre, but not AAV5-Gfp, significantly blocked the Coll IV microinjection-induced increases in the level of TLR4 protein in the ipsilateral thalamus of Fto^fl/fl mice on day 7 post-Coll IV microinjection (Fig. 5A). Neither virus affected basal expression of TLR4 protein in the ipsilateral thalamus of Fto^fl/fl mice on day 7 after saline microinjection (Fig. 5A). Consistently, thalamic overexpression of FTO in the AAV5-Fto-microinjected mice markedly increased the amount of TLR4 protein in the injected thalamus, as compared to that in the AAV5-Gfp-microinjected mice (Fig. 5B). Moreover, this increase could be abolished after intraperitoneal injection of MA (but not PBS) (Fig. 5C). Our single-cell RT-PCR assay showed co-expression of Fto mRNA and Tlr4 mRNA in individual thalamic neurons (Fig. 5D). Collectively, these findings suggest that increased FTO directly participates in the hemorrhage-induced TLR4 upregulation in thalamic neurons.

Figure 5.

Increased FTO stabilizes the Coll IV microinjection-induced TLR4 upregulation in the thalamus. (A) Effect of thalamic pre-microinjection of AAV5-Cre or AAV5-Gfp into unilateral thalamus of Fto^fl/fl male mice 35 days before Coll IV or saline microinjection into the same regions on the expression of TLR4 protien in thalamus. n = 5 mice/group. Statistical significance was calculated with two-way ANOVA (treatment × model) followed by post hoc Tukey test. F_{treatment (1, 16)} = 65.941, p < 0.001; F_{model (1, 16)} = 60.536, p < 0.001; F_{treatment × treatment (1, 16)} = 52.71, p < 0.001. (B) Level of TLR4 protein in the ipsilateral thalamus 5 weeks after microinjection of AAV5-Fto or AAV5-Gfp into unilateral thalamus. n = 5 mice/treatment. Statistical significance was calculated with two-tailed unpaired Student’s t-test. t ₍₈₎ = −6.442. (C) Effect of intraperitoneal administration of meclofenamic acid (MA; 10 mg/kg) once daily for 5 days starting 35 days post-viral microinjection on the levels of FTO and TLR4 proteins in the ipsilateral thalamus of naïve male mice 40 days after microinjection. n = 5 mice/treatment. Statistical significance was calculated with two-way ANOVA (treatment × model) followed by post hoc Tukey test. FTO: F_{model (1, 16)} = 98.045, p < 0.001; F_{treatment (1, 16)} = 0.015, p = 0.904; F_{treatment × treatment (1, 16)} = 0.038, p = 0.848. TLR4: F_{model (1, 16)} = 38.893, p < 0.001; F_{treatment (1, 16)} = 27.386, p < 0.001; F_{treatment × model (1, 16)} = 27.313, p < 0.001. (D) Co-expression of Fto mRNA and Tlr4 mRNA in individual thalamic neurons. n = 3 mice. Tuba-1a was used a positive control.

Increased FTO is responsible for a loss of m⁶A sites in Tlr4 mRNA in the hemorrhagic thalamus

Finally, we examined how increased FTO participated in TLR4 upregulation in hemorrhagic thalamus. RIP assay showed that binding activity between FTO and 3’-untranslated region (UTR) of Tlr4 mRNA after Coll IV microinjection was increased by 2.59-fold as compared to that after saline microinjection in the ipsilateral thalamus 7 days post-microinjection (Fig. 6A). As FTO functions as an “eraser” to remove m⁶A on RNAs⁹, we detected that Coll IV microinjection produced a decrease of m⁶A sites in this Tlr4 mRNA region by 40% of the value after saline microinjection in the ipsilateral thalamus of the AAV5-Gfp-treated Fto^fl/fl mice 7 days after microinjection (Fig. 6B). This decrease was completely rescued by blocking increased FTO through thalamic microinjection of AAV5-Cre in the ipsilateral thalamus of the Fto^fl/fl mice 7 days after Coll IV microinjection (Fig. 6B). These findings suggest that FTO is required for the hemorrhage-induced loss of m⁶A in Tlr4 mRNA. Expectedly, the binding activity between YTHDF2 and Tlr4 mRNA’s 3-UTR from the ipsilateral thalamus on day 7 post-Coll IV microinjection was significantly reduced by 54% as compared to that post-saline microinjection (Fig. 6C). This reduction caused by increased FTO-induced loss of m⁶A sites may promote the stabilization of Tlr4 mRNA. To further support this conclusion, we examined the effect of increased FTO on the decay time of Tlr4 mRNA in in vitro cell lines. When the transcription was halted with actinomycin D, the decay rate of Tlr4 mRNA was much slower in the FTO-overexpressing cells as compared to that in naive cells (Fig. 6D-E). This effect could be diminished by blocking FTO overexpression in the cells co-transfected with the vector expressing Fto shRNA (Fig. 6D-E). Our results indicate that increased FTO may be responsible for a loss of m6A sites in Tlr4 mRNA and then stabilize TLR4 upregulation in the thalamus under hemorrhagic conditions.

Figure 6.

Increased FTO is responsible for a loss of m⁶A sites in Tlr4 mRNA from the thalamus after Coll IV microinjection. M: ladder marker. NC: negative control (distilled H₂O). Input: total purified Tlr4 mRNA fragments from the ipsilateral thalamus. (A) Level of Tlr4 mRNA fragment immunoprecipitated by rabbit anti-FTO in the ipsilateral thalamus 7 days after Coll IV or saline microinjection. n = 3 independent repeats (10 mice/repeat)/group. Saline group: 1.07 ± 0.429. Coll IV group: 2.766 ± 0.29. (B) Level of m⁶A in the Tlr4 mRNA fragment immunoprecipitated by rabbit anti-m⁶A in the ipsilateral thalamus of male Fto^fl/fl mice pre-microinjected with AAV5-Cre (Cre) or AAV5-Gfp (GFP) 7 days after Coll IV or saline microinjection. n = 3 independent repeats (10 mice/repeat)/group. Saline plus AAV5-Gfp group: 1.008 ± 0.155. Coll IV plus AAV5-Gfp group: 0.605 ± 0.074. Coll IV plus AAV5-Cre group: 0.978 ± 0.169. (C) Level of Tlr4 mRNA fragment immunoprecipitated by rabbit anti-YTHDF2 in the ipsilateral thalamus 7 days after Coll IV or saline microinjection. n = 3 independent repeats (10 mice/repeat)/group. Saline group: 1.025 ± 0.283. Coll IV group: 0.475 ± 0.117. (D) Level of Fto mRNA in CAD cells transfected by the vectors expressing FTO, GFP or FTO shRNA 24 h after transfection. n = 5 independent repeats/treatment. Statistical significance was calculated with one-way ANOVA followed by post hoc Tukey test. F_{(3, 16)} = 49.569, p < 0.001. (E) Expression of Tlr4 mRNA in the CAD cells transfected by the vectors as shown in D and treated with actinomycin D (ActD; 5 μg/ml) for the indicated times. n = 5 independent repeats/treatment/time point. The p values on top of red diamonds indicate a difference from the baseline (0 h) of the corresponding group. The p values on top of blue triangles indicate a difference between the FTO plus scrambled siRNA group and FTO plus FTO siRNA group at the corresponding time points. Statistical significance was calculated with two-way ANOVA (treatment × time) followed by post hoc Tukey test. F_{treatment (3, 64)} = 47.694, p < 0.001; F_{time (3, 64)} = 308.162, p < 0.001; F_{treatment × time (9, 64)} = 6.805, p < 0.001.

4. Discussion

Hemorrhage-induced thalamic pain is one of chronic, refractory clinical diseases. During the past decades, a numerous number of animal studies as well as clinical observations on this disorder have been carried out²¹, but the mechanisms of how thalamic hemorrhage causes pain hypersensitivities are still elusive. The present study demonstrated that thalamic microinjection of Coll IV led to an increase in FTO protein expression in the ipsilateral thalamus and that this increase contributed to Coll IV microinjection-induced development and maintenance of pain hypersensitivity. These effects were mediated likely by stabilizing TLR4 upregulation in thalamic neurons. Our findings indicate that FTO may play a key role in hemorrhage-induced thalamic pain.

The present study revealed that FTO was expressed predominantly in the nuclei of thalamic neurons. This cellular distribution was consistent with previous reports that showed the localization of FTO mainly in the neuronal nuclei of dorsal root ganglion¹¹ and cortex²². Moreover, we reported that FTO protein increased in the region around/adjacent to the core of hemorrhagic lesion of the ipsilateral thalamus after Coll IV microinjection. Consistently, FTO was detected in a great number of the neurons in this region after Coll IV microinjection. Inconsistent with our observation, a previous study reported that FTO expression significantly decreased in the peri-infarct cortex after transient middle cerebral artery occlusion²². The reason for the difference in FTO expression between present study and this previous work is still unclear, but it might be related to distinct molecular pathophysiology between ischemia and hemorrhage^23-25. Interestingly, Fto mRNA expression in thalamus was not altered after Coll IV microinjection during the observation period. This result suggests that thalamic Fto gene is activated at the translational level, but not at the transcriptional level, in response to hemorrhagic insults. How hemorrhage causes an increase in FTO protein in thalamus remains unknown, but this increase is likely related to the elevations in the stability and translational efficiency of Fto mRNA.

Increased FTO in the ipsilateral thalamus is required for the development and maintenance of Coll IV microinjection-induced pain hypersensitivities. This conclusion was demonstrated not only by pharmacological FTO inhibition through intraperitoneal MA injection but also by blocking increased thalamic FTO through microinjection of AAV5-Cre into the ipsilateral VPM and VPL of the Fto^fl/fl mice. Systemic administration of MA likely acts directly on thalamic neurons, because the blood-brain barrier was disrupted or dysfunctional after the hemorrhage²⁶. However, MA’s effect on the neurons/cells of other tissues could not be ruled out. We also observed that Coll IV microinjection-caused thalamic damage was improved by thalamic FTO inhibition or knockdown, as less TUNEL-positive cells were detected and surviving neurons, microglia, astrocytes and microphages were identified in the VPM and VPL after systemic pre-administration of MA or thalamic pre-microinjection of AAV5-Cre in the Fto^fl/fl mice. This observation indicates that FTO inhibition or knockdown protects partial neurons from the Coll IV microinjection-induced neuronal damage in the VPM and VPL. The detected microglia, astrocytes and microphages might be associated with hemorrhage-caused inflammatory responses^{14, 27}. We noticed that FTO genetic knockdown effect was less strong than systemic MA’s effect. This may be attributed to the limitation in the microinjected volume of virus (0.5 μl). Additionally, both pharmacological inhibition and genetic knockdown of thalamic FTO profoundly attenuated, but not abolished, the Coll IV microinjection-induced pain hypersensitivities and thalamic damage. These partial effects may be related to the complicated and multiple mechanisms underlying the induction and maintenance of hemorrhage-induced thalamic pain. It is well documented that the excessive activation of excitatory postsynaptic glutamate receptors by increased synaptic release of glutamate and subsequently extreme Ca²⁺ influx into thalamic neurons through these receptors after hemorrhage contribute to thalamic pain genesis²¹. We recently reported that the expression of Fgr, a member of the Src family nonreceptor tyrosine kinases, was increased in thalamic microglia after hemorrhage and that this increase was involved in hemorrhage-induced thalamic pain¹⁴. Thus, combined strategies that target distinct mechanisms underlying thalamic hemorrhage-induced pain hypersensitivities may achieve great benefits in the management of this disorder.

Upregulated TLR4 in thalamic neurons may mediate the role of thalamic increased FTO in hemorrhage-induced thalamic pain. TLR4 was upregulated in ischemic or hemorrhagic brains⁷. Suppression or blockage of TLR4 improved brain damage and blocked mechanical allodynia after carotid artery occlusion⁷. TLR4 deficient mice displayed impaired peripheral neuropathic pain²⁸. Activation of TLR4 led to the release of pro-inflammatory cytokines²⁹. The latter activated microglia and vascular endothelial cells, increased cyclooxygenase, triggered apoptotic signaling pathways, contributed to intracerebral hemorrhagic development and provoked pain²⁹. The evidence suggests that TLR4 is a critical initiator in the hemorrhage-induced thalamic pain. The present study found that increased FTO removed m⁶A sites in Tlr4 mRNA, leading to a loss in the binding of Tlr4 mRNA to YTHDF2 in the ipsilateral thalamus after Coll IV microinjection. Given that YTHDF2 rapidly degrades its binding RNAs³⁰, increased FTO likely stabilized the hemorrhage-induced TLR4 upregulation in thalamic neurons. Indeed, FTO overexpression delayed the degradation of Tlr4 mRNA in in vitro cell line. Blocking the Coll IV microinjection-induced thalamic FTO increase abolished the Coll IV microinjection-induced TLR4 upregulation in the ipsilateral thalamus. Mimicking this increase elevated the level of TLR4 in the thalamus. Moreover, this elevation could be attenuated by FTO pharmacological inhibition. Given that Fto mRNA co-expressed with Tlr4 mRNA in individual thalamic neurons, the anti-nociception caused by pharmacological inhibition or genetic knockdown of increased thalamic FTO likely results from destabilizing the Coll IV microinjection-induced TLR4 upregulation in the ipsilateral thalamic neurons. However, other potential mechanisms by which FTO participates in hemorrhage-induced thalamic pain cannot be ruled out. Using m⁶A-eCLIP assay, our previous study revealed a loss and/or gain of m⁶A sites in thousands of transcripts in injured dorsal root ganglion after peripheral nerve injury¹¹. Thus, FTO contributes to hemorrhage-induced thalamic pain likely through its triggered multiple mechanisms, which merit to be further investigated in the future.

In conclusion, this study, to our knowledge, is the first time in providing the evidence that pharmacological inhibition or genetic knockdown of thalamic increased FTO attenuates the hemorrhage-induced thalamic pain without altering locomotor function and basal or acute pain. Considering that MA was approved by FDA in humans for use as a nonsteroidal anti-inflammatory drug and displays a favorable safety profile in patients²⁰, FTO may be a promising therapeutic target in the management of hemorrhage-induced thalamic pain.

Non-standard Abbreviations and Acronyms

AAV5-Cre

adeno-associated virus 5 expressing Cre

AAV5-Fto

adeno-associated virus 5 expressing FTO

AAV5-Gfp

adeno-associated virus 5 expressing GFP

ALKBH5

AlkB homolog 5

Coll IV

type IV collagenase

CPSP

central poststroke pain

FTO

fat-mass and obesity-associated protein

GFP

green fluorescent protein

M⁶A

N⁶-methyladenosine

MA

meclofenamic acid

METTL3

methyltransferase-like 3

METTL14

methyltransferase-like 14

PCR

polymerase chain reaction

RIP

RNA immunoprecipitation

RT-PCR

real-time polymerase chain reaction

TLR4

toll-like receptor 4

TUNEL

terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end labeling

VPM

ventral posterior medial nuclei

VPL

ventral posterior lateral nuclei

WTAP

Wilms’ tumor 1-associating protein

YTHDF1

YTH N⁶-methyladenosine RNA binding protein 1

YTHDF2

YTH N⁶-methyladenosine RNA binding protein 2