The presence of ac4C modification in PolyA RNA of mouse hippocampus.

(A, B) Verification of the purity of PolyA RNA through the bioanalyzer profiling. The x-axis represents the size of nucleotides, and the y-axis represents the fluorescent signal from the bioanalyzer. LM, Lower Marker. (A), total RNA. (B), polyA RNA. (C) Diminished levels of 18S rRNA in the polyA RNA samples, compared with total RNA samples, which was revealed by RT-qPCR analysis. The levels of 18S rRNA were normalized by that of Gapdh. *** P < 0.001, two-tailed t-test, n = 6 biological replicates. (D-G) The chromatograms of blank, standard and testing samples of ac4C (D), C (E), m6A (F) and A (G) in the LC-MS/MS analysis. Shown are the representative peaks of ac4C, C, m6A and A identified in the polyA RNA samples purified from the HOM of mouse hippocampus. (H) Reduction of ac4C levels by the chemical deacetylation protocol. The stoichiometry levels of ac4C in polyA RNA from control and NH2OH-treated samples were quantified. ** P = 0.008, two-tailed t-test, n = 3 biological replicates. (I) Unchanged m6A levels by the chemical deacetylation protocol. The stoichiometry levels of m6A in polyA RNA from control and NH2OH-treated samples were quantified. NS, not significant, two-tailed t-test, n = 3 biological replicates. Quantification data are expressed as mean ± SEM.

Increased mRNA acetylation in the SYN of mouse hippocampus after memory.

(A) The purified synaptosomes (SYN) of mouse hippocampus expressed high levels of PSD95 and synapsin 1 (SYN1), two synaptic protein markers, but very low levels of α-tubulin, a cytoplasmic protein marker. The homogenates (HOM), cytoplasm (S2) and SYN of mouse hippocampus were subjected to western blots with the indicated antibodies. Left panel, representative images of western blots; right panel, Ponceau staining to show equal protein loading. (B) Acetylation of polyA RNA was not altered in SYN of mouse hippocampus at day 1. The polyA RNA purified from the SYN of control and memory mice at day 1 after MWM training were subjected to ac4C dot-blots and methylene blue staining. Left, the representative images; right, quantification data. NS, not significant, two-tailed t-test, n = 6 biological replicates. Data were normalized by controls. (C) Acetylation of polyA RNA was increased in the SYN of mouse hippocampus at day 6. The polyA RNA purified from the SYN of control and memory mice at day 6 after MWM training were subjected to ac4C dot-blots and methylene blue staining. Left, the representative images; right, quantification data. *** P <0.001, two-tailed t-test, n = 6 biological replicates. Data were normalized by controls. (D) Acetylation of polyA RNA was not altered in the SYN of mouse hippocampus at day 20. The polyA RNA purified from the SYN of control and memory mice at day 20 after MWM training were subjected to ac4C dot-blots and methylene blue staining. Left, the representative images; right, quantification data. NS, not significant, two-tailed t-test, n = 6 biological replicates. Data were normalized by controls. (E) Acetylation of RNA in the SYN was increased during memory but returned to normal levels after forgetting. The stoichiometry levels of ac4C of total RNA purified from the SYN of memory (M) mice were shown at days 1, 6 and 20 after MWM training. ** P = 0.0023, one-way-ANOVA, n = 3 biological replicates. (F-H) Acetylation of mRNA was increased after memory but returned to normal levels after forgetting in the SYN of mouse hippocampus. Shown are cumulative distribution curves of FE of ac4C mRNA purified from the SYN of control (C) and memory (M) mice at days 1 (F), 6 (G), and 20 (H) after MWM training. The FE of ac4C mRNA was quantified in the quartile boxplots, where the solid line in the box is the median and the dashed line is the mean, where the maximum and minimum values are identified. NS, not significant, *** adj-P = 9.9814e-13, Kolmogorov-Smirnov test. (I-K) The total mRNA levels were not significantly altered in the SYN of mouse hippocampus during memory. Shown are cumulative distribution curves of normalized mRNA counts for SYN in control (C) and memory (M) mice at days 1 (I), 6 (J) and 20 (K) after MWM training. The normalized mRNA counts were quantified in the quartile boxplots, where the solid line in the box is the median and the dashed line is the mean, where the maximum and minimum values are identified. NS, not significant, Kolmogorov-Smirnov test. C, control; M, memory. FE, Fold Enrichment. Quantification data are expressed as mean ± SEM.

A representative MISA mRNA for Arc.

(A) The biological processes (BP) in which MISA mRNAs are significantly overrepresented. The x-axis represents the negative logarithm of the adjusted p-value, reflecting the significance of the enrichment, while the y-axis shows the biological processes. (B) The protein-protein interaction (PPI) network of MISA mRNA implicated in the BP of learning or memory. P < 1.0e-16, Hypergeometric Test. (C) Increased ac4C modification of Arc in the SYN rather than in the HOM at day 6. Also note the unaltered ac4C Arc mRNA at days 1 or 20 in the SYN. Shown are the IGV maps of the ac4C peaks from the 3’ UTR to the 5’ UTR of Arc mRNA. The ac4C peaks were represented by acRIP/input (See details in the Method). (D, G) Arc mRNA levels from the input of SYN were not altered at days 1 (D) or 20 (G) after MWM training. The RNA purified from the SYN was subjected to the RT-qPCR analysis. NS, not significant, two-tailed t-test, n = 6 biological replicates. The levels of Arc mRNA were normalized by that of Gapdh in each sample, and then the ratio of Arc to Gapdh were normalized by controls. (E and H) The ac4C modification of Arc mRNA was not altered in the SYN at days 1 (E) or 20 (H) after MWM training. The RNA purified through the acRIP and IgG-IP products of SYN was subjected to the RT-qPCR analysis. NS, not significant, two-way-ANOVA followed by Tukey’s multiple comparison test, n = 6 biological replicates. Data were normalized by input. (F and I) The protein levels of ARC did not change in the SYN at days 1 (F) or 20 (I) after MWM training. Top, representative western blot images; bottom, quantification data. NS, not significant, two-tailed t-test, n = 6 biological replicates. The protein levels of ARC were normalized by that of SYN1 in each sample, and then the ratio of ARC to SYN1 were normalized by controls. (J) Arc mRNA levels from the input of SYN were not altered at day 6 after MWM training. The RNA purified from the SYN was subjected to the RT-qPCR analysis. NS, not significant, two-tailed t-test, n = 6 biological replicates. The levels of Arc mRNA were normalized by that of Gapdh in each sample, and then the ratio of Arc to Gapdh were normalized by controls. (K) The ac4C modification of Arc mRNA was significantly increased in the SYN at day 6 after MWM training. The RNA purified from the acRIP and IgG-IP products of SYN was subjected to the RT-qPCR analysis. *** P < 0.001, two-way-ANOVA followed by Tukey’s multiple comparison test, n = 6 biological replicates. Data were normalized by input. (L) The protein levels of ARC were significantly increased in the SYN at day 6 after MWM training. Top, representative western blot images; bottom, quantification data. *** P < 0.001, two-tailed t-test, n = 6 biological replicates. The protein levels of ARC were normalized by that of SYN1 in each sample, and then the ratio of ARC to SYN1 were normalized by controls. (M) Arc mRNA levels from the HOM were not altered at day 6 after MWM training. The Arc mRNA levels were revealed by the log2TPM from the next-generation sequencing of the HOM RNA. NS, not significant, two-tailed t-test, n = 3 biological replicates. (N) The protein levels of ARC did not change in the HOM at day 6 after MWM training. Top, representative western blot images; bottom, quantification data. NS, not significant, two-tailed t-test, n = 6 biological replicates. The protein levels of ARC were normalized by that of GAPDH in each sample, and then the ratio of ARC to GAPDH were normalized by controls. C, control; M, memory. Quantification data are expressed as mean ± SEM.

Expression of Nat10 in mature neurons of mouse hippocampus.

(A) Experimental design. Cre-reporting AAV was injected into the dorsal hippocampus of 7-week-old Nat10Cre/+ mice. Three weeks after AAV injection, the fluorescent protein EGFP was analyzed in the Nat10 reporter mice. (B) Co-localization of EGFP with NeuN in the hippocampus of Nat10 reporter mice. The hippocampal slices from Nat10 reporter mice were immunostained with anti-NeuN antibodies. so, stratum oriens; sp, stratum pyramidale; sr, stratum radiatum; slm, stratum lacunosum-moleculare; sm, stratum moleculare; gcl, granule cell layer; DG: dentate gyrus. Scale bar, 100 μm. (C) Co-localization of EGFP and NeuN or NG in the CA1 region of Nat10 reporter mice. The hippocampal slices from Nat10 reporter mice were immunostained with anti-NeuN or anti-NG antibodies. The yellow arrow indicates a neuron or excitatory pyramidal neuron that expresses Nat10, whereas the red arrow represents a neuron or excitatory pyramidal neuron that does not express Nat10. The green arrow indicates a non-neuronal cell that expresses Nat10. Scale bar, 25 μm. (D) Quantification of the percentage of NeuN-positive cells expressing Nat10 (Nat10+ / NeuN+ = 65.61 ± 3.245%) and the percentage of Nat10-positive cells expressing NeuN (NeuN+ / Nat10+ = 81.31 ± 2.029%). n = 9 slices from 3 mice. (E) Quantification of the percentage of NG-positive cells expressing Nat10 (Nat10+ / NG+ = 81.87 ± 2.765%) and the percentage of Nat10-positive cells expressing NG (NG+ / Nat10+ = 75.82 ± 1.882%). n = 9 slices from 3 mice. (F) Experimental design of generate Nat10 reporter mice. (G) Co-localization of tdTomato with NeuN in the hippocampus of Nat10 reporter mice. The hippocampal slices from Nat10 reporter mice were immunostained with anti-NeuN antibodies. Scale bar, 25 μm.

Increased NAT10 proteins in the SYN of mouse hippocampus after memory.

(A) Experimental design. The SYN fractions were purified from the hippocampus of control and memory mice at days 1, 6 and 20. The NAT10 protein levels in the SYN were then assessed by western blots. (B-D) Protein levels of NAT10 were increased in the SYN after memory but returned to normal levels after forgetting. The SYN proteins from control and memory mice at days 1 (B), 6 (C), and 20 (D) were subjected to western blots with the indicated antibodies. The protein levels of NAT10 were normalized by that of SYN1 in each sample, and then the ratio of NAT10 to SYN1 was normalized by controls. Top, representative western blot images; bottom, quantification data. NS, not significant, ***P < 0.001, two-tailed t-test, n = 6 biological replicates. (E) The fold change (FC) of NAT10 proteins in panel B-D was quantified. ***P < 0.001, one-way-ANOVA, n = 6 biological replicates. (F) Protein levels of NAT10 were not altered in the HOM at day 6. The HOM proteins from control and memory mice at day 6 were subjected to western blots with the indicated antibodies. Top, representative western blot images; bottom, quantification data. NS, not significant, two-tailed t-test. Data were normalized by controls, n = 6 biological replicates at each condition. The protein levels of NAT10 were normalized by that of α-tubulin in each sample, and then the ratio of NAT10 to α-tubulin was normalized by controls. (G) Experimental design. Proteins were extracted from nucleus (N), cytoplasm (S2), and postsynaptic density (PSD) of control and memory mice hippocampus at day 6, and then NAT10 protein levels were assessed by western blots. (H) Expression of different protein markers in the nuclear (N), cytoplasmic (S2), P2 (Crude synaptosomes fraction), SYN, and PSD fractions of mouse hippocampus. Shown are representative images of western blots. Protein samples purified from different subcellular fractions of mouse hippocampus were subjected to western blots with the indicated antibodies. (I) Protein levels of NAT10 were not altered in the nuclear fraction of mouse hippocampus after memory. The nuclear protein samples from control and memory mice at day 6 were subjected to western blots with the indicated antibodies. Top, representative western blot images; bottom, quantification data. NS, not significant, two-tailed t-test, n = 4 biological replicates. The protein levels of NAT10 were normalized by that of Histone H3 proteins in each sample, and then the ratio of NAT10 to Histone H3 was normalized by controls. (J) Protein levels of NAT10 were significantly reduced in the S2 fraction of mouse hippocampus after memory. The cytoplasmic protein samples from control and memory mice at day 6 were subjected to western blots with the indicated antibodies. Top, representative western blot images; bottom, quantification data. ** P = 0.0052, two-tailed t-test, n = 6 biological replicates. The protein levels of NAT10 were normalized by that of α-tubulin in each sample, and then the ratio of NAT10 to α-tubulin was normalized by controls. (K) Protein levels of NAT10 were significantly increased in the PSD fraction of mouse hippocampus after memory. The PSD protein samples from control and memory mice at day 6 were subjected to western blots with the indicated antibodies. Top, representative western blot images; bottom, quantification data. *** P = 0.0008, two-tailed t-test. n = 6 biological replicates. The protein levels of NAT10 were normalized by that of PSD95 in each sample, and then the ratio of NAT10 to PSD95 was normalized by controls. (L-N) Protein levels of PCAF were not altered in the nuclear (L), S2 (M), or PSD fraction (N) of mouse hippocampus after memory. The protein samples purified from the different subcellular fractions of the hippocampus of control and memory mice at day 6 were subjected to western blots with the indicated antibodies. Top, representative western blot images; bottom, quantification data. NS, not significant, two-tailed t-test, n = 6 biological replicates. The protein levels were normalized by the protein markers in different subcellular fractions and then were normalized by controls. C, control; M, memory. Quantification data are expressed as mean ± SEM.

NAT10-dependent ac4C modification of memory-related mRNAs in the SYN of memory mice.

(A) Experimental design. Three weeks after AAV injection, the control (Ctrl) and Nat10 conditional knockout (cKO) mice were subjected to MWM training and probe test, and then the SYN fractions were purified from hippocampal tissues for low-input acRIP-seq. (B) The total mRNA levels were not significantly altered in the SYN of Nat10 cKO mice, compared to controls. Shown are cumulative distribution curves of normalized mRNA counts for SYN in control and Nat10 cKO mice at day 6 after MWM training. The normalized mRNA counts were quantified in the quartile boxplots, where the solid line in the box is the median and the dashed line is the mean, where the maximum and minimum values are identified. NS, not significant, Kolmogorov-Smirnov test. (C) Acetylation of mRNA in the SYN of Nat10 cKO mice was significantly reduced, compared with controls. Shown are cumulative distribution curves of FE of ac4C mRNA purified from the SYN of control and Nat10 cKO mice at day 6 after MWM training. The FE of ac4C mRNA was quantified in the quartile boxplots, where the solid line in the box is the median and the dashed line is the mean, where the maximum and minimum values are identified. *** adj-P = 1.20076e-12, Kolmogorov-Smirnov test. (D) Venn diagrams show overlapping between MISA mRNA and NASA mRNA. Distinct consensus motifs distinguish NAT10-dependent and NAT10-independent MISA mRNAs. Sequence motif analysis reveals fundamentally different consensus sequence patterns between the two MISA mRNA populations identified in the Venn diagram. (E-G) Reduced ac4C modification of Arc (E), Camk2α (F), and Grin2b (G) mRNA in the SYN of Nat10 cKO mice at day 6, compared with controls. Shown are the IGV maps of the ac4C peaks from the 3’ UTR to the 5’ UTR of transcripts. The ac4C peaks were represented by acRIP/input (See details in the Method). (H) The mRNA levels of Arc, Camk2α and Grin2b from the input of SYN were similar between control and Nat10 cKO mice. The RNA purified from the SYN was subjected to the RT-qPCR analysis. NS, not significant, two-tailed t-test, n = 6 biological replicates. The levels of mRNA were normalized by that of Gapdh in each sample, and then the ratio of mRNA to Gapdh were normalized by controls. (I) The ac4C modification of Arc, Camk2α, and Grin2b mRNA was significantly reduced in the SYN of Nat10 cKO mice at day 6, compared with controls. The RNA purified from the acRIP and IgG-IP products of SYN was subjected to the RT-qPCR analysis. *** P < 0.001, two-way-ANOVA followed by Tukey’s multiple comparison test, n = 6 biological replicates. Data were normalized by input. (J) The protein levels of ARC, CaMKIIα and GluN2B were significantly decreased in the SYN of Nat10 cKO mice at day 6, compared with controls. Left, representative western blot images; right, quantification data. *** P < 0.001, two-tailed t-test, n = 6 biological replicates. The protein levels were normalized by that of PSD95 in each sample, and then the ratio of protein to PSD95 were normalized by controls. Quantification data are expressed as mean ± SEM.

Impaired LTP as well as memory in the Nat10 cKO mice.

(A) A representative image showing recording LTP at the SC-CA1 synapses from Nat10 cKO mice using the protocol of high-frequency stimulation (HFS). Scale bar, 500 μm. (B) The representative fEPSP traces from baseline (BSL), post-tetanus potentiation (PTP), early phase of LTP (E-LTP, 55-60 min after HFS) and late phase of LTP (L-LTP, 115-120 min after HFS). (C) Impaired maintenance of L-LTP at the SC-CA1 synapses in Nat10 cKO mice, compared with controls. Normalized fEPSP amplitudes were plotted every 1 min for hippocampal slices from control and Nat10 cKO mice. *** P < 0.001, two-way ANOVA, n = 8 slices from 4 mice for each group. (D) Normal LTP induction at the SC-CA1 synapses in Nat10 cKO mice, compared to controls. The PTP in panel C was quantified. NS, not significant, two-tailed t-test. n = 8 slices from 4 mice for each group. (E) Unaltered E-LTP at the SC-CA1 synapses in Nat10 cKO mice, compared to controls. The E-LTP in panel C was quantified. NS, not significant, two-tailed t-test. n = 8 slices from 4 mice for each group. (F) Impaired L-LTP at the SC-CA1 synapses in Nat10 cKO mice, compared with controls. The L-LTP in panel C was quantified. ** P = 0.008, two-tailed t-test. n = 8 slices from 4 mice for each group. (G) Experimental design. The control and Nat10 cKO mice were subjected to fEPSP recording at the SC-CA1 synapses after training and probe test in the MWM. (H, I) Memory-induced enhancement of glutamatergic transmission at SC-CA1 synapses was significantly impaired in Nat10 cKO mice, compared with controls. (H) Representative fEPSP traces from control and Nat10 cKO mice. (I) Quantification of the fEPSP amplitudes in panel H. *** Genotype F (3, 680) = 128.2, P < 0.001, two-way ANOVA followed by Sidak’s multiple comparison test, n = 18 slices from 5 mice each group. (J) Impaired memory process in Nat10 cKO mice, compared with controls. Shown are the memory curves of control and Nat10 cKO mice in the MWM. n = 14 mice per group, ** Genotype F (1, 130) = 24.08, ** P = 0.0014, two-way ANOVA followed by Sidak’s multiple comparison test. (K) Similar distance travelled during probe tests between control and Nat10 cKO mice. NS, not significant, two-tailed t-test, n = 14 mice per group. (L) Representative swimming traces of control and Nat10 cKO mice during probe tests. (M) Time spent in the target quadrant was significant reduced in Nat10 cKO groups during probe tests, compared with controls. * P = 0.0198, two-way ANOVA followed by Sidak’s multiple comparison with post hoc t-test, n = 14 mice per group. (N) Number of platform crossing was significantly reduced in Nat10 cKO mice during probe tests, compared with controls. *** P < 0.001, two-tailed t-test, n = 14 mice per group. Quantification data are expressed as mean ± SEM.

Protocols of MWM for control and memory mice.

(A) MWM protocol for control mice. The control mice were placed in MWM with a visible platform at day 1 and 2, after which they returned to home cage and were subjected to probe tests at day 6. (B) Control mice were able to discern the visible platform above the water surface and consequently locate it rapidly. Shown are the latency to find the visible platform at day 1 and 2. (C) Control mice were not able to remember the target quadrant during probe tests on day 6. Shown are the percentage of time staying in each quadrant during probe tests. n = 9 mice, NS, not significant, one-way ANOVA. (D) MWM protocol for memory mice. The memory mice were trained in MWM with a hidden platform from day 1 to 5, and then were subjected to probe tests at day 6 and 20. (E) Memory mice showed improvement in their ability to find the hidden platform during the training process. Shown are the memory curves. (F) Memory mice could remember the target quadrant during probe tests at day 6. Shown are the percentage of time staying in each quadrant during probe tests. n = 9 mice, *** P < 0.001, one-way ANOVA. (G) Memory mice could not remember the target quadrant during probe tests at day 20. Shown are the percentage of time staying in each quadrant during probe tests. n = 6 mice, NS, not significant, one-way ANOVA. (H) Schematic summary of memory score of control and memory mice at day 1, 6, and 20 after MWM training. Memories were recalled at day 6 but were forgotten at day 20 in memory mice. The memory score indicates spatial memory performance in the MWM test.

No significant change of ac4C mRNA in the homogenates of mouse hippocampus after memory.

(A) Experimental design. The RNA samples were divided into input and acRIP. Among them, input was used to eliminate background and acRIP was performed with ac4C antibodies. The process of sequencing peak calling identifies reliable ac4C peaks and calculates FE for each ac4C peak based on the sequencing information of the two-part sample. FE, Fold Enrichment. (B and C) Venn diagrams demonstrating substantial overlap of ac4C peaks and ac4C-modified mRNAs across biological replicates in control mice. (B) Venn diagram showing the intersection of ac4C peaks among three biological replicates from hippocampal homogenates (HOM) of control mice on day 6. The numbers indicate unique and shared peaks between replicates, with 8303 peaks common to all three replicates. (C) Venn diagram showing the intersection of ac4C-modified mRNAs among the same three biological replicates, with 8107 mRNAs common to all three replicates. (D) Consensus motif analysis reveals consistent ac4C modification patterns across biological replicates in both control and memory groups. HOMER motif analysis identified the identical consensus sequence AGCAGCTG in ac4C peak regions across all biological replicates. For control mice, the motif shows highly significant p-values with target percentages at 99% confidence intervals of 57.20% (Rep 1), 54.23% (Rep 2), and 55.32% (Rep 3). For memory mice, similarly significant p-values were observed with target percentages of 56.14 (Rep 1), 54.32% (Rep 2), and 52.44% (Rep 3). The consistency of this motif across both conditions and all replicates demonstrates the specificity of ac4C modification sites. (E and F) Distribution of ac4C peak density along mRNA transcripts in control and memory groups. Metagene profiles display the relative density of ac4C peak coverage across different regions of acetylated mRNAs. Peak density distribution in HOM samples shows enrichment primarily within the CDS region, with lower density in the 5’ and 3’ UTRs. The analysis encompasses all three biological replicates from control (E) and memory groups (F). (G-H) Upset plots illustrating the overlap patterns of ac4C mRNAs across biological replicates. Because many genes express multiple splice isoforms, ac4C calls were summarized at the gene level (a gene is counted as present if ≥1 isoform is ac4C-positive); isoform-specific differences between replicates may therefore not be visible at the transcript level. (G) Upset plot displaying the intersection of ac4C mRNAs among three biological replicates from control HOM samples. The analysis reveals 4494 peaks common to all three replicates, with varying numbers of peaks shared between pairs of replicates or unique to individual samples. (H) Upset plot showing the intersection of ac4C mRNAs among three biological replicates from memory HOM samples. The analysis identifies 4110 peaks common to all three replicates. The distribution patterns and overlap characteristics remain comparable between control and memory conditions, further supporting the stability of ac4C modifications after memory formation. (I) The number of ac4C peaks for each ac4C mRNA is comparable across the three biological replicates in both the HOM control and memory groups. (J) Acetylation of polyA RNA was not altered in HOM of mouse hippocampus after memory. The polyA RNA samples used for each ac4C dot-blot were 0.5 μg that falls into the linear range. The HOM polyA RNA samples of control and memory mice on day 6 after MWM training were subjected to ac4C dot-blots and methylene blue staining. Left, representative images of ac4C dot-blots and methylene blue staining; right, quantification data. n = 6 biological replicates at each condition, NS, not significant, two-tailed t-test. Data were normalized by controls. (K) Acetylation levels of ac4C mRNA were not significantly changed after memory in HOM of mouse hippocampus. Shown are cumulative distribution curves of ac4C levels for HOM in control (C) and memory (M) mice on day 6 after MWM training. The FE of ac4C mRNA was quantified in the quartile boxplots, where the solid line in the box is the median and the dashed line is the mean, where the maximum and minimum values are identified. NS, not significant, Kolmogorov-Smirnov test. FE, Fold Enrichment. (L) The normalized mRNA counts were not significantly changed in HOM of mouse hippocampus on day 6 after MWM training. Shown are cumulative distribution curves of normalized mRNA counts for HOM in C and M mice on day 6 after MWM training. The normalized mRNA counts of each group were quantified in the quartile boxplots, where the solid line in the box is the median and the dashed line is the mean, where the maximum and minimum values are identified. NS, not significant, Kolmogorov-Smirnov test.

Memory-induced synaptic ac4C mRNAs (MISA) in mouse hippocampus.

(A-B and E-F) The transcripts and ac4C mRNAs are largely overlapping across the SYN control group (A-B) and memory group (E-F) at days 1, 6, and 20. The upset plot shows the number of mRNAs (A and E) and ac4C mRNAs (B and F) between samples. (C and G) The same consensus motif, AGCAGCTG, was identified in ac4C peak regions across the SYN control (C) and memory group (G) at days 1, 6, and 20. Consensus motif and its adjusted p-values generated by HOMER of the SYN group at days 1, 6, and 20 in acRIP-seq. (D and H) Density of ac4C peaks across different mRNA regions in the SYN control group (D) and SYN memory group (H) at days 1, 6, and 20. The line plot illustrates the density distribution of ac4C peaks in 5’UTR, CDS, and 3’UTR. The black line represents day 1, the red line represents day 6, and the blue line represents day 20. (I) PCA to demonstrate that the sample SYN memory-day6 separates from the other five samples. (J-L) Scatter plot showing the correlation between changes in mRNA expression and acetylation levels on days 1, 6, and 20. The x-axis represents the fold change in acetylation levels, while the y-axis shows the fold change in mRNA expression. (M) The ac4C levels of MISA mRNA are upregulated following memory and return to control levels after memory retrieval. At days 1, 6, and 20, there are no significant changes in MISA mRNA levels. The black line represents the trend of mRNA expression changes, and the red line represents the trend of ac4C level changes.

Increase of NAT10 proteins and ac4C modification of Arc mRNA in the SYN of mouse hippocampal neurons and slices after cLTP stimulation.

(A) Experimental design. (B and C) Distribution of NAT10 in the PSD were increased in cultured hippocampal neurons after cLTP stimulation. Shown are the fluorescent images of NAT10 were significantly increased in the PSD fraction of primary neuronal of mouse hippocampal after cLTP stimulated. Scale bar, 25 μm and 10 μm. (D) Quantification of PSD95 puncta colocalized with NAT10 per neuron. *** P < 0.001, two-tailed t-test. n = 6 independent experiments. (E) cLTP stimulation slightly increased Arc mRNA levels in the synaptosomes of mouse hippocampal slices. RNA purified from input samples was subjected to RT-qPCR analysis. Arc mRNA levels were normalized by Gapdh and then normalized to controls. * P = 0.0242, two-tailed t-test, n = 6 independent experiments. (F) ac4C modification of Arc mRNA was significantly increased after cLTP stimulation. Mouse hippocampal slices were subjected to vehicle treatment or glycine-induced cLTP stimulation, and 1 h after that, followed by RNA purification from acRIP and IgG-IP products and RT-qPCR analysis. Data were normalized by input. *** P < 0.001, two-way-ANOVA followed by Tukey’s multiple comparison test, n = 6 independent experiments. (G) cLTP increases ARC protein in synaptosomes. Quantification data for representative western blot images. ** P = 0.0011, two-tailed t-test, n = 6 independent experiments.

NAT10-dependent ac4C modification of polyA RNA in mouse hippocampus.

(A) Experimental design. Three weeks after AAV injection, Ctrl and Nat10 cKO mice underwent MWM training and probe test, followed by hippocampal tissue collection for biochemical assays. (B) Infection of AAV in the dorsal mouse hippocampus. Fluorescent images of the mouse hippocampus expressing EGFP are shown. Scale bar, 500 μm. (C) Representative western blot images. Protein samples from Ctrl and Nat10 cKO mouse hippocampus were probed with antibodies against NAT10, Ac-α-tubulin, α-tubulin, Ac-H3, and Histone H3. (D) Reduction of NAT10 protein levels in Nat10 cKO mice compared to Ctrl. Quantification data for panel C are shown. *** P < 0.001, two-tailed t-test, n = 6 biological replicates per condition, Data were normalized by controls. (E) Similar ac-H3 levels between Ctrl and Nat10 cKO mice. Quantification data for panel C are shown. NS, not significant, two-tailed t-test, n = 6 biological replicates per condition, Data were normalized by controls. (F) Similar Ac-α-tubulin levels in Ctrl and Nat10 cKO mice. Quantification data for panel C are shown. NS, not significant, two-tailed t-test, n = 6 biological replicates per condition, Data were normalized by controls. (G and H) Reduced ac4C modification of polyA RNA in the hippocampus of Nat10 cKO mice, compared with controls. The polyA RNA samples purified from the HOM of mouse hippocampus were subjected to ac4C dot-blots and methylene blue staining. (G) Representative images. (H) Quantification data. *** P < 0.001, two-tailed t-test, n = 6 biological replicates per condition. Data were normalized by controls. (I and J) Similar m6A modification of polyA RNA in the hippocampus of control and Nat10 cKO mice. The polyA RNA samples purified from the HOM of mouse hippocampus were subjected to m6A dot-blots and methylene blue staining. (I) Representative images. (J) Quantification data. NS, not significant, two-tailed t-test, n = 6 biological replicates per condition. (K) Reduced ac4C modification of polyA RNA in the hippocampus of Nat10 cKO mice, compared with controls. The polyA RNA samples purified from the HOM of mouse hippocampus were subjected to LC-MS/MS analysis. The stoichiometry levels of ac4C from LC-MS/MS analysis were quantified. ** P = 0.0077, two-tailed t-test, n = 3 biological replicates per condition. (L) Similar m6A modification of polyA RNA in the hippocampus of control and Nat10 cKO mice. The polyA RNA samples purified from the HOM of mouse hippocampus were subjected to LC-MS/MS analysis. The stoichiometry levels of m6A from LC-MS/MS analysis were quantified. NS, not significant, two-tailed t-test, n = 3 biological replicates per condition.

NAT10-dependent of ac4C mapping in the SYN of memory mice.

(A) Experimental design. Three weeks after AAV injection, the control (Ctrl) and Nat10 conditional knockout (cKO) mice were subjected to MWM training and probe test, and then the SYN fractions were purified from hippocampal tissues for low-input acRIP-seq. (B) Reduction of ac4C peak number in the SYN of Nat10 cKO mice, compared with controls. 7097 and 1933 ac4C peaks were identified in the SYN of control and Nat10 cKO mice, respectively. (C) Definition of NAT10-dependent ac4C peaks through acRIP summits displaying higher FE values in control relative to Nat10 cKO mice. Shown are the density scatter plot of ac4C peaks. Each dot represents an ac4C peak. 6331 ac4C peaks are specifically identified in the SYN of control but not Nat10 cKO mice. The FE of 408 ac4C peaks was higher in the SYN of control mice than Nat10 cKO mice. (D) NAT10-dependent ac4C peaks are predominantly located in the CDS region of mRNA. The plot shows the density of ac4C peaks along various mRNA regions, including the 5’UTR, CDS, and 3’UTR. (E) Reduction of ac4C mRNA number in the SYN of Nat10 cKO mice, compared with controls. 3778 and 1680 ac4C mRNAs were identified in the SYN of control and Nat10 cKO groups, respectively. (F) Similar mRNA numbers identified in the SYN of control and Nat10 cKO mice. 6041 and 6047 mRNAs were identified in the SYN of control and Nat10 cKO mice, respectively. (G) Definition of NAT10-dependent synaptic ac4C (NASA) mRNA through acRIP summits displaying higher FE values in control relative to Nat10 cKO mice. Shown are the density scatter plot of ac4C mRNA. Each dot represents an ac4C mRNA. 2507 ac4C mRNAs are specifically identified in the SYN of control but not Nat10 cKO mice. The FE of 839 ac4C mRNA was higher in the SYN of control mice than Nat10 cKO mice. (H) The expression of NASA mRNA is mostly enriched in excitatory neurons of mouse hippocampus. Shown are the percentage of NASA mRNA whose expression was relatively enriched in different cell types of mouse hippocampus through analysis of the single-cell RNA-sequencing dataset (http://dropviz.org). (I) A shared consensus motif, AGCAGCTG, was consistently enriched within ac4C peak regions in both the MISA and NASA groups, as identified by HOMER from acRIP-seq data. Notably, in Nat10 cKO samples, the ac4C-associated motifs exhibited a marked loss of the CXX sequence feature, suggesting that Nat10 depletion alters the sequence preference for ac4C deposition. (J) Analysis of the cell-type enrichment using the method of CIBERSORT reveals that both Nat10-dependent and independent MISA mRNAs are highly expressed in neurons. (K) Gene Ontology BPs associated with the 717 common mRNAs from Figure 6D are significantly enriched. The x-axis indicates both the included mRNAs and the negative logarithm of the adjusted p-values, representing enrichment significance, while the y-axis displays the corresponding biological processes.

Downregulation of NAT10 in mouse hippocampus does not affect body or brain weight, neuronal number or intrinsic excitability.

(A) Similar body weight between 10-week-old Ctrl and Nat10 cKO mice. Left, representative images of the mice, right, quantification of body weight. n = 12 mice for each group, NS, not significant, two-tailed t-test. (B) Similar brain weight between 10-week-old Ctrl and Nat10 cKO mice. Left, representative images of mouse brain, right, quantification of brain weight. n = 12 mice for each group, NS, not significant, two-tailed t-test. (C) Similar number of neurons expressing EGFP in the CA1 region of 10-week-old Ctrl and Nat10 cKO mice. Left, representative images of CA1 neurons expressing EGFP, right, quantification of EGFP-expressing cells in CA1 region. n = 8 slices from 4 mice, NS, not significant, two-tailed t-test. Data were normalized by Ctrl. (D) Similar intrinsic excitability of CA1 pyramidal neurons between 10-week-old Ctrl and Nat10 cKO mice. Left, representative traces of action potentials (AP) of CA1 pyramidal neurons; right, quantification of AP number after injection of different currents into CA1 pyramidal neurons. n = 20 cells from 4 mice per group, NS, not significant, two-way ANOVA followed by Sidak’s multiple comparison test.