Alternative splicing at neuroligin site A regulates glycan interaction and synaptogenic activity

  1. Shinichiro Oku
  2. Huijuan Feng
  3. Steven Connor
  4. Andrea Toledo
  5. Peng Zhang
  6. Yue Zhang
  7. Olivier Thoumine
  8. Chaolin Zhang
  9. Ann Marie Craig  Is a corresponding author
  1. Djavad Mowafaghian Centre for Brain Health and Department of Psychiatry, University of British Columbia, Canada
  2. Departments of Systems Biology and Biochemistry and Molecular Biophysics, Center for Motor Neuron Biology and Disease, Columbia University, United States
  3. Department of Biology, York University, Canada
  4. Interdisciplinary Institute for Neuroscience UMR 5297, CNRS and University of Bordeaux, France
9 figures, 1 table and 2 additional files

Figures

Figure 1 with 1 supplement
NLGN1 alternative splicing at site A regulates heparin/HS binding.

(A) The amino acid sequences of each human NLGN A splice insert highlighting positively charged (blue) and negatively charged (red) residues. Splice insert sequences in mouse NLGN1-3 are identical to those in human NLGN1-3 except for one residue in NLGN1 A2 (7th residue H in mouse and Q in human). (B) Structure of the NLGN1-NRXN1β LNS domain complex (PDB: 3VKF) (Tanaka et al., 2012) showing the position of the A1 insert relative to the constitutive HS binding site (yellow asterisk) (Zhang et al., 2018). The NLGN1 surface is colored according to the electrostatic potential from blue (+8 kbT/ec) to red (−8 kbT/ec), and the NRXN LNS domain is in cyan. (C) Structure of the NLGN1 A1 splice insert (PDB: 3VKF) highlighting the positively charged surface residues proposed to participate in HS interaction. Residues are numbered from the beginning of the A1 insert. (D) Elution profile of purified recombinant NLGN1 isoform ectodomain proteins from a heparin column. Elution at a higher concentration of salt indicates stronger binding.

Figure 1—figure supplement 1
NLGN gene structures.

Structure of the human and mouse NLGN1-3 gene coding region highlighting alternatively spliced exons. Exon (boxes) and intron sizes are listed in base pairs.

The NLGN1 A1 splice insert enhances neurexin recruitment in coculture.

(A–C) Each NLGN1 isoform with an extracellular Myc tag was expressed in COS7 cells which were cocultured with hippocampal neurons. NLGN1 induced recruitment of neurexin along tau-positive axons. Axon regions contacting expressing COS7 cells and lacking contact with MAP2-positive dendrites were assessed to exclude native synapses. The total intensity of native neurexin (B) per contact area was normalized to a baseline value of 1 measured from sister cocultures performed with the negative control protein Myc-tagged Amigo. Neurexin recruitment differed among NLGN1 splice variants, p<0.0005 by Kruskal-Wallis test and **p<0.01 compared with NLGN1 -A by post hoc Dunn's multiple comparisons test, n = 35–43 cells from four independent experiments. Although the value for NLGN1 +A1A2 was higher than that for NLGN1 +A1, this difference was not significant. COS7 cells were chosen for equal surface NLGN1 expression (C). Scale bar, 10 µm.

The NLGN1 A1 splice insert is high in GABAergic neuron cell types.

The fraction of each mouse NLGN1 site A splice isoform transcript is plotted in each upper graph and the fraction of NLGN1 transcript that contains the A1 insert, that is (+A1 plus +A1A2)/total, is plotted in each lower graph. Datasets are from Tasic et al., 2018 (A, B) and Furlanis et al., 2019 (C).

The NLGN1 A1 splice insert is regulated developmentally and by Rbfox splicing factors.

The fraction of each mouse NLGN1 site A splice isoform transcript is plotted in each upper graph and the fraction of NLGN1 transcript that contains the A1 insert, that is (+A1 plus +A1A2)/total, is plotted in each lower graph. In the developmental studies (A, B), ages are indicated in embryonic (E) or postnatal (P) days, or in postnatal weeks (W) or months (M). Panel (C) data are from spinal neuron cultures differentiated from Rbfox1,2,3 triple KO (TKO) embryonic stem cells and grown for the indicated number of days. Panel (D) data are from hippocampi from 2 to 3 month old Mbnl2 KO mice or frontal cortex from adult Mbnl1−/−Mbnl2loxP/loxP Nestin-Cre conditional double KO (cDKO) mice. Panel (E) data are from embryonic day 18 brain of Ptbp2loxP/loxP Nestin-Cre cKO mice or postnatal day 1 cortex of Ptbp2loxP/loxP Emx1-Cre cKO mice. Development datasets are from (A) (Yan et al., 2015) and (B) (Lister et al., 2013). Datasets from KO cells or mice lacking splice factors are from (C) (Jacko et al., 2018), (D) (Charizanis et al., 2012; Weyn-Vanhentenryck et al., 2018) and (E) (Li et al., 2014).

The NLGN1 A1 splice insert enhances presynaptic differentiation in coculture.

(A–C) Each NLGN1 isoform with an extracellular Myc tag was expressed in COS7 cells which were cocultured with hippocampal neurons. NLGN1 induced clustering of synapsin along tau-positive axons. Axon regions contacting expressing COS7 cells and lacking contact with MAP2-positive dendrites were assessed to exclude native synapses. The total intensity of synapsin (B) per contact area was normalized to a baseline value of 1 measured from sister cocultures performed with the negative control protein Myc-tagged Amigo. Synapsin clustering differed among NLGN1 splice variants, p<0.001 by Kruskal-Wallis test and *p<0.05 and **p<0.01 compared with NLGN1 -A by post hoc Dunn's multiple comparisons test, n = 47–56 cells from four independent experiments. COS7 cells were chosen for equal surface Myc-NLGN1 expression (C). Scale bar, 10 µm.

The NLGN1 A1 splice insert promotes structural synapse development.

(A–D) Cultured hippocampal neurons were transfected with U6-shNlgn1-hSyn-CFP and hSyn-Myc-NLGN1* -A or +A1 at DIV 5 and neurons were fixed at DIV 12. The density of VGlut1 clusters did not differ but the total fluorescence intensity of VGlut1 inputs was higher for neurons expressing the NLGN1* +A1 than the -A isoform, ***p=0.0006 by Mann Whitney test, n = 59–66 neurons from three independent experiments. Cells were chosen for equal intensity of surface Myc-NLGN1. Scale bar, top 10 µm, bottom 5 µm. (E–I) Cultured hippocampal neurons were electroporated at plating with shNlgn1-GFP, Homer1c-dsRed, BirAER, and hSyn-AP-NLGN1* -A or +A1 and neurons imaged at DIV 14 following live cell labeling with streptavidin-Alexa647. The synaptic enrichment of AP-NLGN1*, defined as the intensity of AP-NLGN1* in Homer1c-dsRed-positive clusters relative to the intensity in the local dendrite shaft, did not differ between the splice variants, nor was there any difference in the synaptic enrichment or density or mean area of Homer1c-dsRed clusters (all p>0.1 by Mann Whitney test, n > 14 neurons from two independent experiments). Scale bar, 10 µm.

The NLGN1 A1 splice insert promotes functional synapse development.

Cultured hippocampal neurons were transfected with U6-shNlgn1-hSyn-YFP, hSyn-DIO-YFP-P2A-HA-NLGN1* -A or +A1, and CAG-Cre at DIV 5. YFP positive neurons were selected for mEPSC recording at DIV 13 and 14. (A) Representative mEPSC traces from neurons expressing NLGN1* lacking (black trace) or containing (red trace) the A1 splice variant. (B) mEPSC frequency was significantly increased in NLGN1* +A1-expressing neurons (n = 26) relative to cells expressing NLGN1* -A (n = 26; *p=0.040 by Welch's t test), with a corresponding change in interevent interval. (C) mEPSC amplitude did not significantly differ between groups (p=0.93). Scale bar, 20 pA, 1 s.

Figure 8 with 1 supplement
A1 splice inclusion in the autism-linked NLGN1 P89L mutant enhances presynaptic differentiation in coculture.

(A–C) Myc-NLGN1 -A, Myc-NLGN1 P89L -A, or Myc-NLGN1 P89L +A1 was expressed in COS7 cells which were cocultured with hippocampal neurons. The total intensity of synapsin (B) per COS7 cell - axon contact area lacking contact with MAP2-positive dendrites was normalized to a baseline value of 1 measured from sister cocultures performed with the negative control protein Myc-tagged CD4. Synapsin clustering by NLGN1 -A was reduced by the P89L mutation and partially restored by A1 splice site inclusion, ****p<0.0001, **p<0.01, NS not significant by Kruskal-Wallis test with post hoc Dunn's multiple comparisons test. n = 18–29 cells from two independent experiments. COS7 cells were chosen for equal NLGN1 expression (C). Scale bar, 10 µm.

Figure 8—figure supplement 1
The NLGN1 P89L mutation reduces surface expression.

COS7 cells were co-transfected with expression vectors for CFP and for Myc-NLGN1 -A (WT) or Myc-NLGN1 P89L -A. After 48 hr, cells were fixed without permeabilization and immunolabelled for surface Myc. Cells were chosen for imaging only by CFP expression. Myc-NLGN1 P89L -A showed reduced surface expression relative to WT, ****p<0.0001 by Mann-Whitney test, n = 80 cells from two independent experiments. Scale bar, 10 µm.

A1 splice inclusion in the autism-linked NLGN1 P89L mutant promotes synapse development.

(A–D) Cultured hippocampal neurons were transfected with U6-shNlgn1-hSyn-YFP and hSyn-CFP-P2A-Myc-NLGN1* P89L -A or P89L +A1 at DIV 5 and neurons were fixed at DIV 13. The density of VGlut1 clusters did not differ (C) but the total fluorescence intensity of VGlut1 inputs (B) was higher for neurons expressing NLGN1* P89L +A1 than the -A isoform, ***p=0.0008 by Mann Whitney test, n = 97–99 neurons from three independent experiments. Cells were chosen by equal surface Myc-NLGN1 immunofluorescence (D). Scale bar, top 10 µm, bottom 5 µm.

Tables

Key resources table
Reagent type
(species) or resource
DesignationSource or referenceIdentifiersAdditional
information
Strain, strain background (Rattus norvegicus)Sprague-DawleyCharles RiverCD-IGSFor primary hippocampal neuron cultures
Strain, strain background (Mus musculus)C57BL/6JJackson LaboratoryCat# 00064For RNA for RT-PCR for cloning
Cell line (Homo sapiens)HEK293ATCCCat# CRL-1573
Cell line (Cercopithecus aethiops)COS7ATCCCat# CRL-1651
Biological sample (Rattus norvegicus)Primary hippocampal neuronsThis paperFreshly isolated from Rattus norvegicus
Antibodyanti-Myc (rabbit polyclonal)SigmaCat# C3956; RRID:AB_439680(ICC 1:40000)
Antibodyanti-neurexin (rabbit polyclonal)MilliporeCat# ABN161; RRID:AB_10917110(ICC 1:1000)
Antibodyanti-MAP2 (chicken polyclonal)AbcamCat# ab5392; RRID:AB_2138153(ICC 1:80000)
Antibodyanti-VGlut1 (mouse monoclonal)NeuroMabCat# N28/9; RRID:AB_2187693(ICC 1:4000)
Antibodyanti-synapsin I (mouse monoclonal)Synaptic SystemsCat# 106 011; RRID:AB_2619772(ICC 1:20000)
Antibodyanti-tau (mouse monoclonal)MilliporeCat# MAB3420; RRID:AB_94855(ICC 1: 2000)
Antibodyanti-HA (rat monoclonal)RocheCat# 1186743100; RRID:AB_390919(WB 1: 2000)
Recombinant DNA reagentpNICE-HA-NLGN1Scheiffele et al., 2000To subclone NLGN1 isoforms
Recombinant DNA reagentpCAGGS-Myc-NLGN1(-A+B)WTThis paperTo express Myc-NLGN1 -A
Recombinant DNA reagentpCAGGS-Myc-NLGN1(+A1+B)WTThis paperTo express Myc-NLGN1 +A1
Recombinant DNA reagentpCAGGS-Myc-NLGN1(+A2+B)WTThis paperTo express Myc-NLGN1 +A2
Recombinant DNA reagentpCAGGS-Myc-NLGN1(+A1A2+B)WTThis paperTo express Myc-NLGN1 +A1A2
Recombinant DNA reagentpCAGGS-Myc-NLGN1(-A+B)P89LThis paperTo express Myc-NLGN1 P89L -A
Recombinant DNA reagentpCAGGS-CFP-P2A-Myc-NLGN1(-A+B)WTThis paperTo express Myc-NLGN1 -A
Recombinant DNA reagentpCAGGS-CFP-P2A-Myc-NLGN1(-A+B)P89LThis paperTo express Myc-NLGN1 P89L -A
Recombinant DNA reagentpCAGGS-CFP-P2A-Myc-NLGN1(+A1+B)P89LThis paperTo express Myc-NLGN1 P89L +A1
Recombinant DNA reagentpcDNA4-HA-ecto-NLGN1(-A+B)-HisThis paperTo express HA-NLGN1-His -A ectodomain
Recombinant DNA reagentpcDNA4-HA-ecto-NLGN1(+A1+B)-HisThis paperTo express HA-NLGN1-His +A1 ectodomain
Recombinant DNA reagentpcDNA4-HA-ecto-NLGN1(+A2+B)-HisZhang et al., 2018To express HA-NLGN1-His +A2 ectodomain
Recombinant DNA reagentpcDNA4-HA-ecto-NLGN1(+A1A2+B)-HisThis paperTo express HA-NLGN1-His +A1A2 ectodomain
Recombinant DNA reagentpCAGGS-Myc-AmigoThis paperTo express Myc-Amigo
Recombinant DNA reagentAmigo-CFPSiddiqui et al., 2010To subclone Myc-Amigo
Recombinant DNA reagentMyc-CD4This paperTo express Myc-CD4
Recombinant DNA reagentHA-CD4Takahashi et al., 2011To subclone Myc-CD4
Recombinant DNA reagentpLL3.7(hSyn)-Myc-NLGN1*(-A+B)WTThis paperTo express Myc-NLGN1* -A
Recombinant DNA reagentpLL3.7(hSyn)-Myc-NLGN1*(+A1+B)WTThis paperTo express Myc-NLGN1* +A1
Recombinant DNA reagentpLL3.7(hSyn)-CFP-P2A-Myc-NLGN1*(-A+B)P89LThis paperTo express Myc-NLGN1* -A P89L
Recombinant DNA reagentpLL3.7(hSyn)-CFP-P2A-Myc-NLGN1*(+A1+B)P89LThis paperTo express Myc-NLGN1* +A1 P89L
Recombinant DNA reagentpFB-hSyn-DIO-YFP-P2A-HA-NLGN1*(+A2+B)Zhang et al., 2018To subclone NLGN1* isoforms
Recombinant DNA reagentpFB-hSyn-DIO-YFP-P2A-HA-NLGN1*(-A+B)This paperTo express Cre-dependent HA-NLGN1* -A
Recombinant DNA reagentpFB-hSyn-DIO-YFP-P2A-HA-NLGN1*(+A1+B)This paperTo express Cre-dependent HA-NLGN1* +A1
Recombinant DNA reagentpCAG-CreAddgeneAddgene# 13775To express Cre
Recombinant DNA reagentpLL3.7U6-shNlgn1-hSyn-YFPZhang et al., 2018To express shNlgn1
Recombinant DNA reagentpLL3.7U6-shNlgn1-hSyn-CFPThis paperTo express shNlgn1
Recombinant DNA reagentpLL3.7(hSyn)-MCSThis paperTo maintain uniform DNA amounts for transfections
Recombinant DNA reagentpECFP-N1ClontechTo express CFP
Recombinant DNA reagentAP-NLGN1*(-A)This paperTo express AP-NLGN1* -A
Recombinant DNA reagentAP-NLGN1*(+A1)This paperTo express AP-NLGN1* +A1
Recombinant DNA reagentshNlgn1-GFPChamma et al., 2016To express shNlgn1
Recombinant DNA reagentdsRed-Homer1CMondin et al., 2011To express dsRed-Homer1c
Recombinant DNA reagentBirAERHowarth et al., 2006To express BirAER
Chemical compound, drugBicuculline methiodideAbcamCat# ab120109
Chemical compound, drugTetrodotoxin (TTX)AbcamCat# ab120054
OtherNi-NTA agarose beadsQIAGENCa# 30210
OtherHeparin agaroseGE HealthcareCat# 17-0406-01
Software, algorithmWinLTPWinLTPRRID:SCR_008590
Software, algorithmMiniAnalysisSynaptosoftRRID:SCR_002184
Software, algorithmFiji 64-bithttps://doi.org/10.1038/nmeth.2019RRID:SCR_002285
Software, algorithmPymolPymolRRID:SCR_000305
Software, algorithmOLego v1.1.2Wu et al., 2013RRID:SCR_005811
Software, algorithmRR FoundationRRID:SCR_001905
Software, algorithmGraphPad PrismGraphPadRRID:SCR_002798; SCR_000306
Software, algorithmGraphPad InStatGraphPadRRID:SCR_000306
Software, algorithmGraphPad SigmaPlotGraphPadRRID:SCR_000306

Additional files

Supplementary file 1

Summary of RNA-seq datasets analysed in this study.

Accession numbers refer to the Gene Expression Omnibus or Sequence Read Archive. Sample description ages indicate embryonic days (E) or postnatal days (P) or weeks (W). Genotypes refer to single (KO), double (DKO) and triple (TKO) knockout ESC-derived neurons or mice. Library type indicates paired end (PE) or single end (SE).

https://cdn.elifesciences.org/articles/58668/elife-58668-supp1-v2.docx
Transparent reporting form
https://cdn.elifesciences.org/articles/58668/elife-58668-transrepform-v2.pdf

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Shinichiro Oku
  2. Huijuan Feng
  3. Steven Connor
  4. Andrea Toledo
  5. Peng Zhang
  6. Yue Zhang
  7. Olivier Thoumine
  8. Chaolin Zhang
  9. Ann Marie Craig
(2020)
Alternative splicing at neuroligin site A regulates glycan interaction and synaptogenic activity
eLife 9:e58668.
https://doi.org/10.7554/eLife.58668