Activity-induced histone modifications govern Neurexin-1 mRNA splicing and memory preservation

Article metrics

Abstract

Epigenetic mechanisms regulate the formation, consolidation and reconsolidation of memories. However, the signaling path from neuronal activation to epigenetic modifications within the memory-related brain circuit remains unknown. We report that learning induces long-lasting histone modifications in hippocampal memory-activated neurons to regulate memory stability. Neuronal activity triggers a late-onset shift in Nrxn1 splice isoform choice at splicing site 4 by accumulating a repressive histone marker, H3K9me3, to modulate the splicing process. Activity-dependent phosphorylation of p66α via AMP-activated protein kinase recruits HDAC2 and Suv39h1 to establish repressive histone markers and changes the connectivity of the activated neurons. Removal of Suv39h1 abolished the activity-dependent shift in Nrxn1 splice isoform choice and reduced the stability of established memories. We uncover a cell-autonomous process for memory preservation in which memory-related neurons initiate a late-onset reduction of their rewiring capacities through activity-induced histone modifications.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Figure 1: Memory encoding engages a late-onset increase of Nrxn1 SS4 inclusion in memory-related neurons in DG.
Figure 2: Activity-dependent histone modifications regulate late-onset Nrxn1 SS4 inclusions.
Figure 3: Neuronal activation recruits HDAC2 to p66α through AMPK pathway.
Figure 4: Memory encoding induces Nrxn1 SS4 inclusion in the hippocampus through epigenetic regulation.
Figure 5: Suv39h1 regulates Nrxn1 SS4 inclusion for memory preservation.

Accession codes

Primary accessions

Gene Expression Omnibus

References

  1. 1

    Stevens, C.F. A million dollar question: does LTP = memory? Neuron 20, 1–2 (1998).

  2. 2

    Kandel, E.R., Dudai, Y. & Mayford, M.R. The molecular and systems biology of memory. Cell 157, 163–186 (2014).

  3. 3

    Martin, S.J., Grimwood, P.D. & Morris, R.G. Synaptic plasticity and memory: an evaluation of the hypothesis. Annu. Rev. Neurosci. 23, 649–711 (2000).

  4. 4

    Garner, A.R. et al. Generation of a synthetic memory trace. Science 335, 1513–1516 (2012).

  5. 5

    Liu, X. et al. Optogenetic stimulation of a hippocampal engram activates fear memory recall. Nature 484, 381–385 (2012).

  6. 6

    Milad, M.R. & Quirk, G.J. Neurons in medial prefrontal cortex signal memory for fear extinction. Nature 420, 70–74 (2002).

  7. 7

    Gräff, J. et al. Epigenetic priming of memory updating during reconsolidation to attenuate remote fear memories. Cell 156, 261–276 (2014).

  8. 8

    Guan, J.S., Xie, H. & Ding, X. The role of epigenetic regulation in learning and memory. Exp. Neurol. 268, 30–36 (2015).

  9. 9

    Guan, J.S. et al. HDAC2 negatively regulates memory formation and synaptic plasticity. Nature 459, 55–60 (2009).

  10. 10

    Levenson, J.M. et al. Regulation of histone acetylation during memory formation in the hippocampus. J. Biol. Chem. 279, 40545–40559 (2004).

  11. 11

    Feng, J. et al. Dnmt1 and Dnmt3a maintain DNA methylation and regulate synaptic function in adult forebrain neurons. Nat. Neurosci. 13, 423–430 (2010).

  12. 12

    Zovkic, I.B., Paulukaitis, B.S., Day, J.J., Etikala, D.M. & Sweatt, J.D. Histone H2A.Z subunit exchange controls consolidation of recent and remote memory. Nature 515, 582–586 (2014).

  13. 13

    Alarcón, J.M. et al. Chromatin acetylation, memory, and LTP are impaired in CBP+/- mice: a model for the cognitive deficit in Rubinstein-Taybi syndrome and its amelioration. Neuron 42, 947–959 (2004).

  14. 14

    Gräff, J. et al. An epigenetic blockade of cognitive functions in the neurodegenerating brain. Nature 483, 222–226 (2012).

  15. 15

    Aoto, J., Martinelli, D.C., Malenka, R.C., Tabuchi, K. & Südhof, T.C. Presynaptic neurexin-3 alternative splicing trans-synaptically controls postsynaptic AMPA receptor trafficking. Cell 154, 75–88 (2013).

  16. 16

    Dalva, M.B., McClelland, A.C. & Kayser, M.S. Cell adhesion molecules: signalling functions at the synapse. Nat. Rev. Neurosci. 8, 206–220 (2007).

  17. 17

    Yamagata, M., Sanes, J.R. & Weiner, J.A. Synaptic adhesion molecules. Curr. Opin. Cell Biol. 15, 621–632 (2003).

  18. 18

    Südhof, T.C. Neuroligins and neurexins link synaptic function to cognitive disease. Nature 455, 903–911 (2008).

  19. 19

    Kim, H.G. et al. Disruption of neurexin 1 associated with autism spectrum disorder. Am. J. Hum. Genet. 82, 199–207 (2008).

  20. 20

    Ullrich, B., Ushkaryov, Y.A. & Südhof, T.C. Cartography of neurexins: more than 1000 isoforms generated by alternative splicing and expressed in distinct subsets of neurons. Neuron 14, 497–507 (1995).

  21. 21

    Iijima, T. et al. SAM68 regulates neuronal activity-dependent alternative splicing of neurexin-1. Cell 147, 1601–1614 (2011).

  22. 22

    Tonegawa, S., Pignatelli, M., Roy, D.S. & Ryan, T.J. Memory engram storage and retrieval. Curr. Opin. Neurobiol. 35, 101–109 (2015).

  23. 23

    Ryan, T.J., Roy, D.S., Pignatelli, M., Arons, A. & Tonegawa, S. Memory. Engram cells retain memory under retrograde amnesia. Science 348, 1007–1013 (2015).

  24. 24

    Luco, R.F., Allo, M., Schor, I.E., Kornblihtt, A.R. & Misteli, T. Epigenetics in alternative pre-mRNA splicing. Cell 144, 16–26 (2011).

  25. 25

    Xie, H. et al. In vivo imaging of immediate early gene expression reveals layer-specific memory traces in the mammalian brain. Proc. Natl. Acad. Sci. USA 111, 2788–2793 (2014).

  26. 26

    Zhang, F., Wang, L.P., Boyden, E.S. & Deisseroth, K. Channelrhodopsin-2 and optical control of excitable cells. Nat. Methods 3, 785–792 (2006).

  27. 27

    Rea, S. et al. Regulation of chromatin structure by site-specific histone H3 methyltransferases. Nature 406, 593–599 (2000).

  28. 28

    Hathaway, N.A. et al. Dynamics and memory of heterochromatin in living cells. Cell 149, 1447–1460 (2012).

  29. 29

    Brackertz, M., Gong, Z., Leers, J. & Renkawitz, R. p66alpha and p66beta of the Mi-2/NuRD complex mediate MBD2 and histone interaction. Nucleic Acids Res. 34, 397–406 (2006).

  30. 30

    Xue, Y. et al. NURD, a novel complex with both ATP-dependent chromatin-remodeling and histone deacetylase activities. Mol. Cell 2, 851–861 (1998).

  31. 31

    Mairet-Coello, G. et al. The CAMKK2-AMPK kinase pathway mediates the synaptotoxic effects of Aβ oligomers through Tau phosphorylation. Neuron 78, 94–108 (2013).

  32. 32

    Green, M.F., Anderson, K.A. & Means, A.R. Characterization of the CaMKKβ-AMPK signaling complex. Cell. Signal. 23, 2005–2012 (2011).

  33. 33

    Reijmers, L.G., Perkins, B.L., Matsuo, N. & Mayford, M. Localization of a stable neural correlate of associative memory. Science 317, 1230–1233 (2007).

  34. 34

    Jaenisch, R. & Bird, A. Epigenetic regulation of gene expression: how the genome integrates intrinsic and environmental signals. Nat. Genet. 33 (Suppl. 1), 245–254 (2003).

  35. 35

    Gupta-Agarwal, S. et al. G9a/GLP histone lysine dimethyltransferase complex activity in the hippocampus and the entorhinal cortex is required for gene activation and silencing during memory consolidation. J. Neurosci. 32, 5440–5453 (2012).

  36. 36

    Chih, B., Gollan, L. & Scheiffele, P. Alternative splicing controls selective trans-synaptic interactions of the neuroligin-neurexin complex. Neuron 51, 171–178 (2006).

  37. 37

    Schreiner, D. et al. Targeted combinatorial alternative splicing generates brain region-specific repertoires of neurexins. Neuron 84, 386–398 (2014).

  38. 38

    Vargas, D.Y. et al. Single-molecule imaging of transcriptionally coupled and uncoupled splicing. Cell 147, 1054–1065 (2011).

  39. 39

    Leutgeb, J.K., Leutgeb, S., Moser, M.B. & Moser, E.I. Pattern separation in the dentate gyrus and CA3 of the hippocampus. Science 315, 961–966 (2007).

  40. 40

    Yassa, M.A. & Stark, C.E. Pattern separation in the hippocampus. Trends Neurosci. 34, 515–525 (2011).

  41. 41

    Peleg, S. et al. Altered histone acetylation is associated with age-dependent memory impairment in mice. Science 328, 753–756 (2010).

  42. 42

    Picelli, S. et al. Full-length RNA-seq from single cells using Smart-seq2. Nat. Protoc. 9, 171–181 (2014).

Download references

Acknowledgements

Y. Jun (Tsinghua University) kindly provided advice on CRISPR-Cas9 assays in primary culture. The work is supported by grants from the National Basic Research Program of China (2013CB835100), NSFC (31671104), NSFC (31371059), Brain Inspired Computing Research, Tsinghua University (20141080934) and Beijing Municipal Science & Technology Commission (Z161100000216126) to J.-S.G., and is supported by the Beijing NOVA program (2015B057). The work was partially funded by the National Natural Science Foundation of China and the German Research Foundation (DFG) in project Crossmodal Learning, NSFC (61621136008)/DGF TRR-169 to J.-S.G.

Author information

J.-S.G. designed and directed the project. X.D. initiated the splicing studies and the behavior experiments. S.L. generated the knockout mice and performed biochemical experiments. H.X. developed the cell sorting method. T.Z., M.T. and D.L. helped on the cellular studies. Y.J. helped generate knockout mice. J.W. helped with the kinase assay. H.D. helped with the mass-spectrum analysis. W.X. and W.Z. helped with the ChIP-seq analysis. The manuscript was written by J.-S.G., S.L. and X.D. and commented on by all the authors.

Correspondence to Ji-Song Guan.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 FACS purification of GFP-positive neurons from mouse DG.

(a) DG neurons were dissected from the adult EGR1-EGFP transgenic mice. The EGFP-positive neurons and EGFP-negative neurons were separated by a FACS system (BD FACS AriaIII). Neurons were stained with DAPI and PI dyes. Cell debits without DAPI staining were discarded. The isolated intact neurons were in the population with high DAPI signal and low PI signal. The neurons with different EGFP signal intensities were collected, according to the EGFP intensity via a FACS sorter. n=4 independent experiments. (b) Averaged freezing levels during training and recall trials. Unpaired t-test, t(6)=9.414, p<0.0001, n=4 mice from 3 litters. (c) Percentages of each fraction of sorted cells in training and control group. Unpaired t-test, p>0.05, n=4 independent experiments. (d) The example of the isolated neurons in each fraction. All neurons were stained with DAPI. Only the P6 fraction showed higher EGFP-positive signal. Scale bar, 25 um. n=4 independent experiments. (e) qPCR for mRNA level of Map2 in each fraction from the FACS experiment. EGR1-; unpaired t-test, t(22)=0.2531, p=0.8025; EGR1+; unpaired t-test, t(22)=1.623, p=0.1189; EGR1++; Unpaired t-test, t(22)=1.032, p=0.3135. n=4 independent experiments. Shown are mean value ± s.e.m (b), and median, 25th and 75th percentile, min and max value (c,e).

Supplementary Figure 2 FACS purification of GFP-positive neurons from mouse cortical layer 4.

(a) The expression of EGR1-EGFP in the whole brain. Red mask indicated cortical layer4 region collected for FACS. Scale bar, 1 mm. n=3 independent experiments. (b-c) Neurons with different EGFP signal intensities were collected, according to the EGFP intensity via a FACS sorter. n=3 independent experiments. (d) Semi-quantitative PCR results showed Nrxn1 SS4(+)/SS4(-) ratio in each fraction from homecage and training group. n=3 independent experiments. (e-f) qPCR for Nrxn1 SS4(+)/SS4(-) ratio (e) (EGR1-; unpaired t-test, t(16)=1.044, p=0.3121; EGR1+; unpaired t-test, t(16)=1.304, p=0.2105; EGR1++; unpaired t-test, t(16)=0.3726, p=0.7143. n=3 independent experiments.) or NeuN (f) (EGR1-; unpaired t-test, t(10)=1.938, p=0.0813; EGR1+; unpaired t-test, t(10)=0.8754, p=0.4019; EGR1++; unpaired t-test, t(10)=1.161, p=0.2726. n=3 independent experiments.) in each fraction. Shown are median, 25th and 75th percentile, min and max value (e,f). Full-length gels are presented in Supplementary Figure 14.

Supplementary Figure 3 Neuronal activity dependent inclusion of Nrxn1 SS4.

(a) Semi-quantitative PCR results showed Nrxn1 mRNA level 24 hr after high K+ stimulation. Unpaired t-test, t(16)=0.9947, p=0.3347. n=3 independent experiments. (b-c) qPCR for (b) Nrxn1 SS4(+)/SS4(-) ratio (unpaired t-test, t(14)=2.413, p=0.0301; unpaired t-test, t(16)=0.7685, p=0.4534.n=3 independent experiments.) and (c) mRNA level of EGR1 (unpaired t-test, t(4)=6.390, p=0.0031; unpaired t-test, t(4)=1.458, p=0.2186. n=3 independent experiments.). Neurons infected with LV-CamKII-ChR2 were activated by 30 Hz blue light (LED) for 15 min in presence or absence of the voltage-gated sodium channels tetrodotoxin (2 μm). (d) Semi-quantitative PCR result for high K+ stimulation induced inclusion of ‘GATA’ containing small exons from cortical neurons. Samples were gathered 24 hours after high K+ stimuli. n=3 independent experiments. (e) qPCR results of (d) (n=3). Gphn, Gephrin; Scn1a, sodium channel α1 subunit gene; Nrxn 1-3, Neurexin 1-3. Nrxn1; unpaired t-test, t(16)=5.054, p=0.0001; Nrxn2; unpaired t-test, t(16)=11.09, p<0.0001; Nrxn3; unpaired t-test, t(16)=5.806, p<0.0001; Gphn; unpaired t-test, t(16)=4.786, p=0.0002; Scn1a; unpaired t-test, t(16)=4.462, p=0.0004. n=3 independent experiments. (f) Semi-quantitative PCR results showed the splicing pattern of ‘GATA’ containing small exons in each fraction from homecage and training group. n=3 independent experiments. (g) qPCR result for inclusion of ‘GATA’ containing small exons in Nrxn2 and Nrxn3 in (f). Nrxn 2; unpaired t-test, t(10)=4.266, p=0.0016; Nrxn 3; unpaired t-test, t(10)=2.658, p=0.0240. n=3 independent experiments. Shown are median, 25th and 75th percentile, min and max value (a,b,c,e,g). *P<0.05; **P<0.01; ***P<0.001. Full-length gels are presented in Supplementary Figure 15.

Supplementary Figure 4 Activity-dependent Nrxn1 SS4 inclusion is regulated by epigenetic modifications.

(a) Up, positions of primers used for analysis are indicated. Down, qPCR results of H3K9ac-,H3K9me2-,H3K9me3-immunoprecipitated chromatin at the exon 21 region of Nrxn1 in the NaCl and KCl stimulated cortical neurons. H3K9me3; unpaired t-test, t(33)=0.6774, p=0.5029. n=6 independent experiments. (b) qPCR results of H3K9me3-immunoprecipitated chromatin at the Nrxn1 exon 6, exon 13, exon 23 and 3’UTR regions in the NaCl and KCl stimulated cortical neurons. Exon 6; unpaired t-test, t(10)=0.5931, p=0.5663; exon 13; unpaired t-test, t(10)=1.871, p=0.0909; exon 23; unpaired t-test, t(10)=0.7849, p=0.4507; 3’UTR; unpaired t-test, t(10)=0.02865, p=0.9777. n=3 independent experiments. (c-d) qPCR results of H3K27me3- (c) (H3K27me3; unpaired t-test, t(4)=3.079, p=0.0370. n=3 independent experiments.), H3K36me2- (d) (n=3 independent experiments.) immunoprecipitated chromatin at the exon 21 and exon 22 region of Nrxn1 in the NaCl and KCl stimulated cortical neurons. (e) qPCR results of H3K9me3-immunoprecipitated chromatin at the exon 22 and exon 21 region of Nrxn1 in the NaCl and KCl stimulated cortical neurons derived from wild type or Suv39h1+/- mouse. Exon 22; WT; unpaired t-test, t(16)=4.247, p=0.0006; Suv39h1+/-; unpaired t-test, t(16)=2.793, p=0.0130; unpaired t-test, t(16)=4.249, p=0.0006. Exon 21; WT; unpaired t-test, t(16)=1.212, p=0.2431; Suv39h1+/-; unpaired t-test, t(16)=0.6685, p=0.5135. n=3 independent experiments. (f) Quantitation of mRNA level of Suv39h1 in cortical neurons following infection with the LV-Suv39h1-RNAi. Doxcycline (Dox) (1ug/ml) was added to induce TurboRFP/shRNAmir expression. Unpaired t-test, t(12)=4.860, p=0.0004. n=3 independent experiments. (g) Representative western blot images and quantification of Suv39h1 in cortical neurons following infection with the LV-Suv39h1-RNAi. Paired t-test, t(2)=4.836, p=0.0402. n=3 independent experiments. (h) DIV11 cortical neurons were activated with 30 Hz blue light (LED) for 15 min and doxycycline (1 μg/ml) was added to induce the knockdown of Suv39h1 as indicated. Nrxn1 SS4(+)/SS4(-) ratio was monitored by qPCR. Unpaired t-test, t(10)=4.471, p=0.0012; unpaired t-test, t(10)=0.8390, p=0.4211; unpaired t-test, t(10)=0.6272, p=0.5441. n=3 independent experiments. (i) Left, Schematic diagram of nucleosome occupied splice reporter construct, positions of primers used for analysis are indicated by arrows. Right, qPCR results of H3K9me3-immunoprecipitated chromatin at the indicated region in the splice reporter construct. Unpaired t-test, t(16)=6.605, p<0.0001. n=3 independent experiments. (j-k) qPCR results of (j) (n=3 independent experiments.) HDAC2- and (k) (unpaired t-test, t(16)=2.243, p=0.0394. n=3 independent experiments.) RNAPII-immunoprecipitated chromatin at the exon 21 region of Nrxn1 in the NaCl and KCl stimulated cortical neurons. (l) Left, Schematic diagram shows the purification of Br-UTP incorporated NRO-RNAs. Right, semi-quantitative PCR results showed the specificity of BrdU antibody binding. n=3 independent experiments. (m) Relative Luciferase activity of cultured neurons transfected with Nrxn1 exon 22 (90bp) or ‘GATA’ to ‘GATG’ mutation (n=5). psi-GATA: original sequence of exon 22.psi-GATG: mutation from ‘GATA’ to 'GATG' in exon 22 (schematic of constructs shown above graph). Unpaired t-test, t(8)=2.965, p=0.0180. n=5 independent experiments. Shown are mean value ± s.e.m (c,d,g), and median, 25th and 75th percentile, min and max value (a,b,e,f,h,i,j,k,m). *P<0.05; **P<0.01; ***P<0.001. Full-length gels and blots are presented in Supplementary Figure 15.

Supplementary Figure 5 Targeted deletion of Suv39h1 in mouse.

(a) Overview schematic of the Cas9/sgRNA target site. The sgRNA-targeting sequence is underlined, and the PAM sequence is labeled in green. (b) The sequence of mutant allele in mouse #3. PAM sequence is labeled in red. (c) Genotyping of both WT and Mutant mice. Primers were designed to detect the WT and Mutant allele respectively. Primers for WT were shown in solid arrows, Primers for mutant were shown in dash arrows. Deleted 4 bps were shown in red. n=3 independent experiments. (d) RT-qPCR analysis for Suv39h1 in the mouse hippocampus and cortex. Cortex; +/-; unpaired t-test, t(13)=4.315, p=0.0008; -/-; unpaired t-test, t(7)=8.623, p<0.0001. n=6 mice for WT, 9 mice for +/-, 3 mice for -/-. Hippocampus; +/-; unpaired t-test, t(12)=3.295, p=0.0064; -/-; unpaired t-test, t(7)=5.936, p=0.0006. n=6 mice for WT, 8 mice for +/-, 3 mice for -/-. (e) Immunoblot analysis for Suv39h1 in the cortex from WT and Suv39h1 mutant mice. n=4 independent experiments. (f) Immunoblot analysis for H3K9me3, H3K9me2, H3K9ac in cortex from WT and Suv39h1+/- heterozygous mice. n=3 independent experiments. (g) qPCR results of H3K9me3-immunoprecipitated chromatin at exon 22 and exon 21 of Nrxn1 in cortex from Suv39h1+/- heterozygous mice compared to wild type group. Exon 22; unpaired t-test, t(22)=3.216, p=0.0040; Exon 21; unpaired t-test, t(22)=3.562, p=0.0017. n=4 independent experiments. Shown are median, 25th and 75th percentile, min and max value (d,g). *P<0.05; **P<0.01. Full-length gels and blots are presented in Supplementary Figure 15.

Supplementary Figure 6 Targeted deletion of Gatad2a (p66α) in mouse cortical neurons.

(a) Overview schematic of the Cas9/sgRNA target sites. The sgRNA-targeting sequence is underlined, and the PAM sequence is labeled in green. (b) Immunoblot and quantitative analysis for p66α protein level in cortical neurons infected by lentivirus expressing Cas9 and mock or p66α #2 sgRNA. Paired t-test, t(3)=4.788, p=0.0173. n=4 independent experiments. (c) RT-qPCR analysis for p66α in cortical neurons infected by lentivirus expressing Cas9 and mock or p66α #2 sgRNA. Unpaired t-test, t(4)=5.032, p=0.0073. n=3 independent experiments. (d) Determining genome targeting efficiency using T7 endonuclease I in neurons infected by lentivirus expressing Cas9 and mock or p66α #2 sgRNA. In this case, the uncut band was 670 bp and the cut bands were approximately 415 bp and 255 bp. n=3 independent experiments. (e) Sequence analysis of the target gene region in cortical neurons infected by lentivirus expressing Cas9 and mock or p66α #2 sgRNA, PAM sequence is labeled in red. n=30 clones from 3 independent experiments. Shown are mean value ± s.e.m (b), and median, 25th and 75th percentile, min and max value (c). **P<0.01. Full-length blots and gels are presented in Supplementary Figure 15.

Supplementary Figure 7 p66α mediates the activity-dependent recruitment of repressive histone markers.

(a) qPCR results of p66α-immunoprecipitated chromatin at the ‘GATA’ containing small exons from cortical neurons. Nrxn2; unpaired t-test, t(28)=2.338, p=0.0268, n=5 independent experiments; Nrxn3; unpaired t-test, t(10)=3.830, p=0.0033, n=3 independent experiments; Gphn; unpaired t-test, t(32)=2.660, p=0.0121, n=6 independent experiments; Scn1a; unpaired t-test, t(13)=4.732, p=0.0004, n=3 independent experiments. (b) qPCR results of H3K9me3-immunoprecipitated chromatin at in the NaCl and KCl stimulated cortical neurons. Nrxn2; unpaired t-test, t(16)=3.775, p=0.0017, n=3 independent experiments; Nrxn3; unpaired t-test, t(16)=2.903, p=0.0104, n=3 independent experiments; Gphn; unpaired t-test, t(16)=2.826, p=0.0122, n=4 independent experiments; Scn1a; unpaired t-test, t(22)=2.337, p=0.0289, n=5 independent experiments. (c) qPCR results of the abundance of H3K9me3 at the same region in mouse hippocampus 1week after EE. Nrxn2; unpaired t-test, t(28)=2.109, p=0.0441, n=5 independent experiments; Nrxn3; unpaired t-test, t(25)=2.091, p=0.0468, n=5 independent experiments; Gphn; unpaired t-test, t(22)=3.068, p=0.0056, n=4 independent experiments; Scn1a; unpaired t-test, t(22)=1.925, p=0.0673, n=4 independent experiments. Shown are median, 25th and 75th percentile, min and max value (a,b,c). *P<0.05; **P<0.01; ***P<0.001.

Supplementary Figure 8 Neural activity phosphorylates p66α via AMPK and recruits HDAC2 to the Nrxn1 SS4 site.

(a) After AICAR treatment for 1 hour, MYC-p66α overexpressed in NG cell was immunoprecipitated by MYC antibody and analyzed by mass spectrometry. Upper panel showed the mass spectrometry of the dephosphorylated sample. (b) A summary of the major phosphorylated fragments of p66α in the mass-spectrum analysis. The detected phosphorylated sites were shown in red. Number of peptide spectrum match (PSM) for phosphorylated ones and total was listed below. (c) Quantitation showed phosphor-serine signal of wild-type GST-p66α (1-200aa), and the S96A mutant in Fig. 3e. p66α; 1-min; paired t-test, t(6)=3.052, p=0.0225; 2-min; paired t-test, t(6)=4.035, p=0.0068; 5-min; paired t-test, t(6)=4.311, p=0.0050. n=7 independent experiments. (d) qPCR results of HDAC2-immunoprecipitated chromatin at the exon 22 and exon 21 region of Nrxn1 in cortical culture. Neurons were incubated with AICAR (1mM) as indicated. Exon 22; unpaired t-test, t(16)=2.294, p=0.0357; Exon 21; unpaired t-test, t(16)=0.1481, p=0.8841. n=3 independent experiments. (e) Quantitation showed normalized co-IP efficiency of p66α-HDAC2 in Fig. 3j. Paired t-test, t(4)=2.875, p=0.0452; paired t-test, t(4)=2.865, p=0.0457. n=5 independent experiments. (f) Schematic model for neuronal activity induced epigenetic dependent Nrxn1 SS4 inclusion. AMPK pathway activation caused by neuronal high frequency action potential phosphorylated p66α at Ser96, which thereby recruited HDAC2 and Suv39h1 near ‘TGATAA’ element on exon 22 of Nrxn1 and then mediated the inclusion of SS4. Shown are mean value ± s.e.m (c,e), and median, 25th and 75th percentile, min and max value (d). *P<0.05; **P<0.01.

Supplementary Figure 9 Genome-wide mapping of neural-activity-induced alternative splicing and the H3K9me3 markers.

(a) Activity-induced AS events in neural culture at 24 hours after stimuli. Cultured primary neurons (DIV 11) were stimulated by 50 mM KCl for 10 min. RNA was extracted at 24 hr after the stimuli and subjected to RNA-seq. Alternative splicing events (including 6 types of alternative splicing) with FDR<0.01 and over 2 fold-changes in splicing index (FC SI) in either of the two RNA-seq replicates were marked in red. Highly up-regulated Nrxn1 SS4 and Nrxn2 SS4 are indicated in arrow. n=2 replicates of RNA-seq. (b) A summary of the proportion of neural activity-induced alternative spliced (AS) genes within each key functional GO entity. (c) Validation of the whole genome screening. Experimental assessment of variant exons by reverse transcription–polymerase chain reaction (RT-PCR) in the control and KCl stimulated cortical neurons; blue, alternative exons. n=3 independent experiments. (d) Training induced enrichment of H3K9me3 on the variant exons with activity-induced inclusion in hippocampus. Mice were trainied in enriched envrionment for 24 hours and hippocampal samples were examined 7 days after training. Representitive tracks showed the UCSC genome browser snapshots of RNA-seq. Mice underwent environmental enrichment (red) or homecage controls (black) were compared at selected AS exons (blue) in the activity-innduced inclusion group, activity-induced exclusion group and no change group. Y axis, signal intensity; the gene locations are indicated on top. n=2 replicates of ChIP-seq. (e) RPKM profiles of H3K9me3 within a 1 kb window at alternative spliced exons (<300 bp), from activity-induced inclusion group (n=95) (FDR<0.01, FC SI>1.5), activity-induced exclusion group (n=145) (FDR<0.01, 0<FC SI<0.67) and no change group (n=103) (FDR<0.01, 0.95<FC SI<1.05). Black line indicates the mean RPKM of H3K9me3 signals under control conditions, whereas the red line depicts the mean RPKM of H3K9me3 following EE treatment. Included (from -200 bp to +200 bp); two-way ANOVA, group, F(1,188)=8.431, p=0.0041, n=95 genes; excluded (from -200 bp to +200 bp); two-way ANOVA, group, F(1,288)=0.7739, p=0.3798, n=145 genes; unchanged (from -200 bp to +200 bp); two-way ANOVA, group, F(1,204)=0.8105, p=0.3690, n=103 genes. Shadows are mean value ± s.e.m (e). **P<0.01. Full-length gels are presented in Supplementary Figure 15.

Supplementary Figure 10 Neuronal hyperactivity induces suppression of synaptophysin cluster assembly between neurons and NLGN1B-expressing cells.

(a) Neurons were activated by KCl (50 mM) for 10 minutes and analyzed 36 hours later. NaCl treatment or no treatment were conducted as negative controls. Quantification showed intensity of presynaptic marker SYP on NLGN1B+ HEK293T cells. Scale bar, 10 μm. Untreated versus KCl; unpaired t-test, t(89)=2.368, p=0.0201, n=46 cells for untreated, 45 cells for KCl from 3 independent experiments; Untreated versus NaCl; unpaired t-test, t(90)=2.248, p=0.0270, n=46 cells for untreated, 47 cells for NaCl from 3 independent experiments. (b) Overexpression of SS4(-) isoform of Nrxn1β attenuated activity induced repression of the assembly of synaptophysin on NLGN1B-expressing cells. Neurons were infected with LV-dsRED-Nrxn1β SS4(-).Scale bar, 10 μm. Unpaired t-test, p>0.05, n=40 cells for untreated, 41 cells for NaCl, 40 cells for KCl from 3 independent experiments. (c) Inclusion of SS4 were suppressed by the transfection of antisense oligonucleotides targeting on Nrxn1 SS4 (U6-AO 5’SS4) (schematically shown above the graph) in NG cells. Nrxn1, 2, 3 SS4(+)/SS4(-) ratio was monitored by quantitative PCR. Nrxn1; unpaired t-test, t(28)=22.76, p<0.0001, n=5 independent experiments; Nrxn2; unpaired t-test, t(26)=1.600, p=0.1217, n=5 independent experiments; Nrxn3; unpaired t-test, t(21)=3.188, p=0.0044, n=4 independent experiments. (d) Neurons transfected with antisense oligonucleotides targeting on Nrxn1 SS4 (U6-AO 5’SS4) or mock plasmid were co-cultured with HEK293T cells expressing NLGN1B-EGFP. The ratio of synaptophysin formed on GFP+ axon region or uncovered region on the same HEK293 cell was monitored. Scale bar, 10 μm. n=3 independent experiments. (e) Quantification of (d). KCl; unpaired t-test, t(61)=2.205, p=0.0312. n= 24 cells for mock, 39 cells for U6-AO from 3 independent experiments. Shown are mean value ± s.e.m (a,b,e), and median, 25th and 75th percentile, min and max value (c). *P<0.05; **P<0.01; ***P<0.001.

Supplementary Figure 11 Learning-induced synaptic remodeling on mossy fibers is regulated by Suv39h1.

(a) Schematic diagram of the mossy fiber projection of EGR1 positive dentate gyrus (DG) granule cells. (b) The expression of EGFP in dentate gyrus. Scale bar, 10 um. (c) Mossy fiber axons of EGR1 positive DG neurons were imaged in CA3. Scale bar, 20 um. (d,e) Representative images of mossy fiber axons and boutons from EGR1 positive DG neurons in the homecage (d) (n=13 slices from 3 mice) and recall groups (e) (n=15 slices from 3 mice). Mossy fiber axons were traced by yellow arrows and boutons with different size were marked in filled and open arrowheads respectively. Scale bar, 10 um. (f) Learning associated activity decreased bouton area in the mossy fiber axon of memory related cells. Unpaired t-test, t(26)=4.778, p=0.0041. n= 13 slices for HC, 15 slices for CFC from 3 mice. (g) The distribution of bouton area within each slice from homecage and recall groups with a bin width of 2 μm2 2-μm2 bin; unpaired t-test, t(26)=3.935, p=0.0006;10-μm2 bin; unpaired t-test, t(26)=3.067, p=0.0050. n= 13 slices for HC, 15 slices for CFC from 3 mice. (h,i) Representative images of mossy fiber axons and boutons from EGR1 positive DG neurons of Suv39h1+/-/EGR1-EGFP mice in the homecage (h) (n=15 slices from 3 mice) and recall groups (i) (n=17 slices from 3 mice). Scale bar, 10 um. (j) Suv39h1+/- mice blocked the reduction of bouton area in the mossy fiber axon. Shown are mean value ± s.e.m (f,g,j,k). **P<0.01; ***P<0.001.

Supplementary Figure 12 Schematic model for transcriptionally coupled inclusion of Nrxn1 SS4 24 h after neuronal activity.

(a) Left, schematic diagram showing the locations of primers used to measure the levels of co-transcriptionally spliced mRNA. Primer In.21-In.22 detected the unspliced pre-mRNA containing the upstream and downstream introns that flank exon 22. Primer In.22-Ex.23 detected the total nascent transcripts. Once Nrxn1 exon 22 in the pre-mRNA underwent the transcriptional coupled splicing, the preceding intron 21 was removed when RNAPII was transcribing intron 22. Right, semi-quantitative PCR results showed lower level of In.21-In.22 transcript, indicating the transient neuronal activity induced a transcriptionally coupled inclusion of Nrxn1 SS4 24 hr later. Paired t-test, t(3)=7.968, p=0.0041. n=4 independent experiments. (b) Illustration of the two-step regulation of activity-dependent Nrxn1 splicing. Shown are median, 25th and 75th percentile, min and max value (a). **P<0.01. Full-length gels are presented in Supplementary Figure 15.

Supplementary Figure 13 Full-length pictures of the blots and gels presented in Figures 1 and 2.

Supplementary Figure 14 Full-length pictures of the blots and gels presented in Figures 3,4,5 and Supplementary Figures 1 and 2.

Supplementary Figure 15 Full-length pictures of the blots and gels presented in Supplementary Figures 3,4,5,6,7,8,9,10,11,12.

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Ding, X., Liu, S., Tian, M. et al. Activity-induced histone modifications govern Neurexin-1 mRNA splicing and memory preservation. Nat Neurosci 20, 690–699 (2017) doi:10.1038/nn.4536

Download citation

Further reading