Distinct neurexin-cerebellin complexes control AMPA- and NMDA-receptor responses in a circuit-dependent manner

Curation statements for this article:
  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This manuscript is of broad interest to neuroscientists studying mechanisms regulating synapse formation and maintenance. Following up on the previous work by the authors on trans-synaptic signaling complexes involving neurexins and cerebellins, this study shows that the basic framework of the complexes operates broadly across different synapses in the brain albeit with subtle differences. The experiments are carefully executed, while some key conclusions could be better supported by additional data.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

This article has been Reviewed by the following groups

Read the full article

Abstract

At CA1→subiculum synapses, alternatively spliced neurexin-1 (Nrxn1 SS4+ ) and neurexin-3 (Nrxn3 SS4+ ) enhance NMDA-receptors and suppress AMPA-receptors, respectively, without affecting synapse formation. Nrxn1 SS4+ and Nrxn3 SS4+ act by binding to secreted cerebellin-2 (Cbln2) that in turn activates postsynaptic GluD1 receptors. Whether neurexin-Cbln2-GluD1 signaling has additional functions besides regulating NMDA- and AMPA-receptors, and whether such signaling performs similar roles at other synapses, however, remains unknown. Here, we demonstrate using constitutive Cbln2 deletions in mice that at CA1→subiculum synapses, Cbln2 performs no additional developmental roles besides regulating AMPA- and NMDA-receptors. Moreover, low-level expression of functionally redundant Cbln1 did not compensate for a possible synapse-formation function of Cbln2 at CA1→subiculum synapses. In exploring the generality of these findings, we examined the prefrontal cortex where Cbln2 was recently implicated in spinogenesis, and the cerebellum where Cbln1 is known to regulate parallel-fiber synapses. In the prefrontal cortex, Nrxn1 SS4+ -Cbln2 signaling selectively controlled NMDA-receptors without affecting spine or synapse numbers, whereas Nrxn3 SS4+ -Cbln2 signaling had no apparent role. In the cerebellum, conversely, Nrxn3 SS4+ -Cbln1 signaling regulated AMPA-receptors, whereas now Nrxn1 SS4+ -Cbln1 signaling had no manifest effect. Thus, Nrxn1 SS4+ - and Nrxn3 SS4+ -Cbln1/2 signaling complexes differentially control NMDA- and AMPA-receptors in different synapses in diverse neural circuits without regulating synapse or spine formation.

Article activity feed

  1. Author Response

    Reviewer #1 (Public Review):

    This study is a follow-up to the previous work by the authors in establishing a surprising role for the presynaptic adhesion molecules, neurexin (Nrxn) variants containing the SS4+ splice site, in differentially controlling postsynaptic NMDA and AMPA receptors by forming links through a shared system of extracellular cerebellins (Cbln) and postsynaptic GluD1. Here the authors show at CA1 to subiculum synapses, that the role for Clbn2 in mediating the effects of Nrxn1-SS4+ and Nrxn3-SS4+ in enhancing NMDAR and suppressing AMPAR, respectively, is redundant with that of Clbn1. Moreover, Clbns do not appear to play a role in synapse formation. Dai and colleagues extend their previous work also by highlighting the common function for Nrxn-Clbn signaling system across different synapses albeit with subtle differences and point to a lack of a role for Nrxn-Clbn signaling in morphological synapse development. Overall the data are solid, while the key findings are mostly incremental, and the basis for the selectivity in the observed differential regulation of AMPARs and NMDARs via the same trans-synaptic link through Clbns at various types of synapses remain to be clarified. Importantly, the authors make a definitive conclusion concerning the lack of a role for Nrxn-Cbln signaling complexes in synapse formation during development. Nevertheless, this is a contentious issue, and as such, the conclusions could be more compellingly supported with further experiments.

    We appreciate the reviewer’s positive assessment of our study.

    Reviewer #2 (Public Review):

    In this manuscript Dai et al. investigated the role of Nrxn-Cbln complexes in regulating AMPA- and NMDA- receptor function in different brain regions. Using a combination of genetic manipulations, together with electrophysiological and biochemical assays, the authors showed that, at CA1-subiculum synapses, Cbln2 regulates NMDA- and AMPA- receptors via Nrxn1SS4+ -Cbln2 and Nrxn3SS4+-Cbln2 signaling complexes, respectively. In the prefrontal cortex, only Nrxn1SS4+-Cbln2 signaling-dependent regulation of NMDA receptors occurs, while in the cerebellum, only Nrxn3SS4+-Cbln1 signaling-dependent regulation of AMPA receptor occurs. This systematic investigation of the function of different Neurexin-Cerebellin signaling complexes contributes to our understanding of how different members of the same family, in combination pairs, regulate synaptic transmission with circuit specificity. This work adds to the authors' systemic investigation of molecular mechanisms regulating synaptogenesis, synaptic transmission and synaptic plasticity.

    We thank the reviewer for the positive and astute comments.

    Some suggestions for clarifications:

    1. Regarding expression of Cbln1 in the subiculum, in lines 271-273, the authors stated that "in these and earlier experiments we only studied Cbln2, but quantifications show that Cbln1 is also expressed in the subiculum, albeit at lower levels Figure S3)." However, Figure S3 does not include any quantifications, and the example image does not show visible Cbln1 expression. Thus, the above-mentioned statement is inconsistent with the data presented. Please revise. If the authors would like to keep the statement about quantifications of Cbln1, then quantification should be provided for all panels of this Figure, in order to give the readers some ideas about relative expression levels.

    We agree, and have addressed this issue as described above (introductory point 4).

    1. Does Cbln4, which is also broadly expressed in the brain, play a role in regulating AMPA- and NMDA-receptors at the synapses investigated? Does Cbln3 contribute to regulation of synaptic transmission in the cerebellum? Please discussion.

    Cbln4 is not expressed in the subiculum, but is expressed in the PFC. Specifically, Cbln1, Cbln2, and Cbln4 are broadly expressed in brain, whereas Cbln3 is restricted to cerebellar granule cells and requires Cbln1 or Cbln2 for secretion (Bao et al., 2006; Miura et al., 2006). Remarkably, Cbln1, Cbln2, and Cbln4 are not uniformly expressed in all neurons, but synthesized in restricted subsets of neurons (Seigneur and Südhof, 2017). For example, cerebellar granule cells express high levels of Cbln1 but only modest levels of Cbln2, excitatory entorhinal cortex (EC) neurons express predominantly Cbln4, and neurons in the medial habenula (mHb) express Cbln2 or Cbln4 (Seigneur and Südhof, 2017).

    Cbln4 is poorly studied, and Cbln3 has not been functionally studied at all. To the best of our knowledge, there are only four studies on Cbln4 function, three of which are from our lab. The Seigneur & Sudhof (2018) paper showed that the deletion of Cbln4 in a large number of brain regions caused no change in excitatory or inhibitory synapse numbers. Subsequently, the Seigneur et al. (2018) paper demonstrated that genetic deletion of Cbln4 in the mHb had no major effect on synapse numbers, but because of the limits of this preparation (synaptic transmission is hard to monitor in the mHB), no detailed synaptic studies were done. The Fossati et al. (2019) paper in Neuron shows that Cbln4 regulates inhibitory synapse numbers in the cortex by binding to GluD1, but this study depended on RNAi, not genetic manipulations. Its results are puzzling because structural biology studies have shown that Cbln4 does not bind to GluD2, which is highly homologous to GluD1 and has the same function as GluD1. Instead of binding to GluD’s, Cbln4 binds to another class of receptors, Neogenin-1 and DCC, making the Fossati et al. (2019) paper difficult to interpret. The Liakath-Ali et al. (2022) paper, finally, demonstrated that deletion of Cbln4 in the EC or deletion of Neo1 in the dentate gyrus (DG) blocks long-term potentiation at EC→DG synapses but does not change basal synaptic transmission or synapse numbers, again consistent with the notion that Cbln4 regulates synapse properties similar to Cbln1 and Cbln2.

    We have now described these studies in the introduction to the paper. Many synaptic proteins are associated with contentious studies in the literature, and we completely concur that it is essential to evenly discuss the issues in detail, even if this expands the size of a paper.

    Reviewer #3 (Public Review):

    In this study, Dai and colleagues used genetic models combined to electrophysiological recordings and behavior as well as immunostaining and immunoblotting to investigate the role of trans-synaptic complexes involving presynaptic neurexins and cerebellins in shaping the function of central synapses. The study extends previous findings from the same authors as well as other groups showing an important role of these complexes in regulating the function of central synapses. Here, the authors sought to achieve two main objectives: (1) investigating whether their previous findings obtained at mature CA1-> subiculum synapses (Aoto et al., 2013; Dai et al., Neuron 2019; Dai et al., Nature 2021) extend to different synapse subtypes in the subiculum as well as to other central synapses including cortical and cerebellar synapses and (2) investigating whether Nrx-Cbln-GluD trans-synaptic complexes play a role in synapse formation as previously proposed by other groups.

    Overall, the study provides interesting and solid electrophysiological data showing that different Nrxns and Cblns assemble trans-synaptic complexes that differently regulate AMPAR and NMDAmediated synaptic transmission across distinct synaptic circuits (most likely through binding to postsynaptic GluD receptors).

    We appreciate the reviewer’s accurate and positive assessment of our study.

    However, the study has several important weaknesses:

    1. The novelty of the findings appears limited. Indeed, previous studies from the same authors with similar experimental paradigms and readouts already demonstrated the role of Nrxn-CblnGluD complexes in regulating AMPARs versus NMDARs in mature neurons (Aoto et al., Cell 2013; Dai et al., Neuron 2019; Dai et al., Nature 2021). Moreover, the absence of role of Cblns and GluD receptors in synapse formation was already suggested in previous studies from the same authors (Seigneur and Sudhof, J Neurosci 2018; Seigneur et al., PNAS 2018; Dai et al., Nature 2021).

    Not surprisingly, we disagree with this comment. We do concur that our data are consistent with previous studies, but believe that this reproducibility is a strength since so many data in the literature are irreproducible.

    We do not agree, however, that our findings lack novelty. The novelty is admittedly limited, after all we like to be consistent, but our paper is the first to demonstrate that the Nrxn1/Cbln/GluD and Nrxn3/Cbln/GluD complexes are differentially active in different synapses, with the subiculum synapses having both, the mPFC synapses only the former, and the cerebellum only the latter. This is a very important innovation that illustrates the power of the Nrxn/Cbln/GluD signaling complex in shaping synapses. In addition, our paper is the first to analyze a possible developmental function of Cbln2 in depth, to analyze its differential role at the two dominant types of pyramidal neurons in the subiculum, regular- and burst-spiking neurons, to analyze conditional deletions of Cbln1 in the adult cerebellum, and to directly measure the effect of Cbln2 deletions in the PFC. Especially in view of the recent Nature paper that concluded that Cbln2 regulates spine numbers in the PFC, these findings are highly relevant.

    1. The conclusion made by the authors that the Nrxn-Cbln-GluD trans-synaptic complexes do not play a role in synapse formation/development is not sufficiently supported by their data, while previous studies suggest the opposite. Actually, this conclusion is essentially based on the two following measurements taken as a 'proxy' for synapse density: (1) 'the average vGluT1 intensity calculated from the entire area of subiculum' and (2) the 'synaptic proteins levels' assessed by immunoblotting. None of these measurements (only performed in the subiculum) allow to precisely assess synapse density on the neurons of interest. While the average vGluT1 intensity over large fields of view does not directly reflect the density of synapses and does not take into account the postsynaptic compartment, the immunoblotting data only reflects the overall expression of synaptic proteins without discriminating between intracellular, surface and synaptic pools and between cell types. In the subiculum from Cbln1+2 KO mice, the authors performed mEPSCs recordings and found an increase in frequency. However, this increase may reflect the unsilencing and/or potentiation of AMPAR-EPSCs above the detection threshold, irrespectively of the actual synapse number. Finally, the decrease in NMDAR-EPSCs is not discussed by the authors while it could actually reflect a decrease in synapse number.

    We agree that additional data on synapse numbers are helpful for our paper. We have now performed these studies as described in detail in our response to introductory point 1 above. However, we would also like to refer to the already existing body of evidence on the role of neurexin-based complexes in synapse numbers. We have shown in papers published over the last two decades that deletions of individual neurexins or of multiple neurexins, as well as blocking cerebellin binding to neurexins by ablating splicing site #4 (SS4) in neurexins, have NO effect on synapse numbers. The most important of these papers are:

    1. Missler, M., Zhang, W., Rohlmann, A., Kattenstroth, G., Hammer, R.E., Gottmann, K., and Südhof, T.C. (2003) α-Neurexins Couple Ca2+-Channels to Synaptic Vesicle Exocytosis. Nature 423, 939948.
    2. Kattenstroth, G., Tantalaki, E., Südhof, T.C., Gottmann, K., and Missler, M. (2004) Postsynaptic Nmethyl-D-aspartate receptor function requires α-neurexins. Proc. Natl. Acad. Sci. U.S.A. 101, 2607-2612.
    3. Dudanova, I., Tabuchi, K., Rohlmann, A., Südhof, T.C., and Missler, M. (2007) Deletion of α-Neurexins Does Not Cause a Major Impairment of Axonal Pathfinding or Synapse Formation. J. Comp. Neurol. 502, 261-274.
    4. Etherton, M.R., Blaiss, C., Powell, C.M., and Südhof, T.C. (2009) Mouse neurexin-1α deletion causes correlated electrophysiological and behavioral changes consistent with cognitive impairments. Proc. Natl. Acad. Sci. U.S.A. 106, 17998-18003.
    5. Soler-Llavina, G.J., Fuccillo, M.V., Ko, J., Südhof, T.C., and Malenka, R.C. (2011) The neurexin ligands, neuroligins and LRRTMs, perform convergent and divergent synaptic functions in vivo. Proc. Natl. Acad. Sci. U.S.A. 108, 16502-16509.
    6. Aoto, J., Martinelli, D.C., Malenka, R.C., Tabuchi, K., and Südhof, T.C. (2013) Presynaptic Neurexin-3 Alternative Splicing Trans-Synaptically Controls Postsynaptic AMPA-Receptor Trafficking. Cell 154, 75-88. PMCID: PMC3756801.
    7. Aoto, J., Földy, C., Ilcus, S.M., Tabuchi, K., and Südhof, T.C. (2015) Distinct circuit-dependent functions of presynaptic neurexin-3 at GABAergic and glutamatergic synapses. Nat Neurosci. 18, 997-1007.
    8. Anderson, G.R., Aoto, J., Tabuchi, K., Földy, F., Covy, J., Yee, A.X., Wu, D., Lee, S.-J., Chen, L., Malenka, R.C., Südhof, T.C. (2015) α-Neurexins Control Neural Circuit Dynamics by Regulating Endocannabinoid Signaling at Excitatory Synapses. Cell 162, 593-606. PMCID: PMC4709013
    9. Chen, L.Y., Jiang, M., Zhang, B., Gokce, O., and Südhof, T.C. (2017) Conditional Deletion of All Neurexins Defines Diversity of Essential Synaptic Organizer Functions for Neurexins. Neuron 94, 611-625. PMCID: PMC5501922
    10. Dai, J., Aoto, J., and Südhof, T.C. (2019) Alternative Splicing of Presynaptic Neurexins Differentially Controls Postsynaptic NMDA- and AMPA-Receptor Responses. Neuron 102, 993-1008. PMCID: PMC6554035
    11. Luo, F., Sclip, A., Jiang, M., and Südhof, T.C. (2020) Neurexins Cluster Ca2+ Channels within presynaptic Active Zone. EMBO J. 39, e103208. PMCID: PMC7110102
    12. Khajal, A.J., Sterky, F.H., Sclip, A., Schwenk, J., Brunger, A.T., Fakler, B., and Südhof, T.C. (2020) Deorphanizing FAM19A Proteins as Pan-Neurexin Ligands with an Unusual Biosynthetic Binding Mechanism. J. Cell Biol. 219, e202004164
    13. Luo, F., Sclip, A., and Südhof, T.C. (2021) Universal role of neurexins in regulating presynaptic GABAB-receptors. Nature Comm. 12, 2380. PMCID: PMC8062527
    14. Wang, C.Y., Trotter, J.H., Liakath-Ali, K., Lee, S.J., Liu, X., and Südhof, T.C. (2021) Molecular SelfAvoidance in Synaptic Neurexin Complexes. Science Advances 7, eabk1924. PMCID: PMC8682996
    15. Dai, J., Patzke, C., Liakath-Ali, K., Seigneur, E., and Südhof, T.C. (2021) GluD1, A signal transduction machine disguised as an ionotropic receptor. Nature 595, 261-265. PMCID: PMC8776294

    Individual papers may not convince the reviewer, but the cumulative evidence seems to us to be hopefully persuasive. We have published less evidence on the lack of a role of cerebellins and GluD’s in synapse numbers than on neurexins, but the only in-depth studies of these molecules by others are in the cerebellum. Here, deletions of Cbln1 and GluD2 indeed cause a significant, albeit partial, loss of synapses. However, this loss may not be due a lack of synapse formation, but to an elimination of synapses that have been formed, as demonstrated by many beautiful papers from leading investigators. It is regrettable that reviews and textbooks continue to state that cerebellins mediate synapse formation as an established fact because as far as we can see, there is limited evidence for that conclusion, but it keeps coming back again and again.

    1. The authors do not provide sufficient data in order to interpret the increase in AMPAR-EPSCs and decrease in NMDAR-EPSCs amplitudes. Are the changes in AMPARs and NMDARs occurring at pre-existing synapses or do they result from alterations in the number of physical synapses and/or active synapses (see point#2)? In particular, the increase in AMPAR/NMDAR ratio accompanied by the increase in mEPSCs frequency might be well explained by the unsilencing of some synapses and/or by the fact that the available pool of AMPARs is distributed over a smaller number of synapses, resulting in higher quantal size. These effects could explain the blockade of LTP, i.e., through an occlusion mechanism.

    We addressed these points in previous studies (Aoto et al., 2013; Dai et al., 2019 and 2021), as discussed and cited in the present paper, and expanded on these points in the present paper.

    In a nutshell, we showed by surface AMPAR staining that presynaptic Nrxn3-SS4+ decreases postsynaptic AMPAR levels, and by direct application of AMPA that it decreases the functional surface levels of AMPARs, whereas presynaptic Nrxn1-SS4+ increases the functional surface levels of NMDARs. We also demonstrated the opposite effects for the GluD1 KO, and furthermore showed by minimal stimulation experiments that the Cbln2 deletion does not alter the number of silent synapses. In the present manuscript, we performed a detailed analysis of the miniature quantal size for AMPAR- and NMDAREPSCs.

    Finally, we have demonstrated in a large number of papers, including this one, that genetic manipulations of neurexins, cerebellins, and GluD’s do not alter synapse numbers with a few exceptions in which synapses are secondarily eliminated, like in the cerebellum. Together, these data show that the observed changes are mediated by a regulation of postsynaptic functional AMPARs and NMDARs, not by alterations in synapse numbers or by synapse silencing/unsilencing.

    1. The authors did not demonstrate (or did not cite relevant studies) that the deletion of Cbln1 and/or Cbln2 does not affect the expression of the remaining Cblns isoforms (Cbln2 and/or Cbln4) or Nrxns1/3 and GluD1/2. This verification is important to preclude the emergence of any compensatory effect.

    To address this point, we have now measured the mRNA expression levels of Nrxns, Cblns, and GluDs in both the subiculum and the prefrontal cortex in littermate P35-42 Cbln2 WT and KO mice. The result show that the constitutive Cbln2 deletion causes no compensatory expression effects (new suppl Fig. S5). Please note that compensatory expression effects are often raised as a possibility for explaining genetically induced changes (or the lack thereof), but such effects are virtually never found.

  2. Evaluation Summary:

    This manuscript is of broad interest to neuroscientists studying mechanisms regulating synapse formation and maintenance. Following up on the previous work by the authors on trans-synaptic signaling complexes involving neurexins and cerebellins, this study shows that the basic framework of the complexes operates broadly across different synapses in the brain albeit with subtle differences. The experiments are carefully executed, while some key conclusions could be better supported by additional data.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

  3. Reviewer #1 (Public Review):

    This study is a follow-up to the previous work by the authors in establishing a surprising role for the presynaptic adhesion molecules, neurexin (Nrxn) variants containing the SS4+ splice site, in differentially controlling postsynaptic NMDA and AMPA receptors by forming links through a shared system of extracellular cerebellins (Cbln) and postsynaptic GluD1. Here the authors show at CA1 to subiculum synapses, that the role for Clbn2 in mediating the effects of Nrxn1-SS4+ and Nrxn3-SS4+ in enhancing NMDAR and suppressing AMPAR, respectively, is redundant with that of Clbn1. Moreover, Clbns do not appear to play a role in synapse formation. Dai and colleagues extend their previous work also by highlighting the common function for Nrxn-Clbn signaling system across different synapses albeit with subtle differences and point to a lack of a role for Nrxn-Clbn signaling in morphological synapse development. Overall the data are solid, while the key findings are mostly incremental, and the basis for the selectivity in the observed differential regulation of AMPARs and NMDARs via the same trans-synaptic link through Clbns at various types of synapses remain to be clarified. Importantly, the authors make a definitive conclusion concerning the lack of a role for Nrxn-Cbln signaling complexes in synapse formation during development. Nevertheless, this is a contentious issue, and as such, the conclusions could be more compellingly supported with further experiments.

  4. Reviewer #2 (Public Review):

    In this manuscript Dai et al. investigated the role of Nrxn-Cbln complexes in regulating AMPA- and NMDA- receptor function in different brain regions. Using a combination of genetic manipulations, together with electrophysiological and biochemical assays, the authors showed that, at CA1-subiculum synapses, Cbln2 regulates NMDA- and AMPA- receptors via Nrxn1SS4+ -Cbln2 and Nrxn3SS4+-Cbln2 signaling complexes, respectively. In the prefrontal cortex, only Nrxn1SS4+-Cbln2 signaling-dependent regulation of NMDA receptors occurs, while in the cerebellum, only Nrxn3SS4+-Cbln1 signaling-dependent regulation of AMPA receptor occurs. This systematic investigation of the function of different Neurexin-Cerebellin signaling complexes contributes to our understanding of how different members of the same family, in combination pairs, regulate synaptic transmission with circuit specificity. This work adds to the authors' systemic investigation of molecular mechanisms regulating synaptogenesis, synaptic transmission and synaptic plasticity.

    Some suggestions for clarifications:
    1. Regarding expression of Cbln1 in the subiculum, in lines 271-273, the authors stated that "in these and earlier experiments we only studied Cbln2, but quantifications show that Cbln1 is also expressed in the subiculum, albeit at lower levels Figure S3)." However, Figure S3 does not include any quantifications, and the example image does not show visible Cbln1 expression. Thus, the above-mentioned statement is inconsistent with the data presented. Please revise. If the authors would like to keep the statement about quantifications of Cbln1, then quantification should be provided for all panels of this Figure, in order to give the readers some ideas about relative expression levels.
    2. Does Cbln4, which is also broadly expressed in the brain, play a role in regulating AMPA- and NMDA-receptors at the synapses investigated? Does Cbln3 contribute to regulation of synaptic transmission in the cerebellum? Please discussion.

  5. Reviewer #3 (Public Review):

    In this study, Dai and colleagues used genetic models combined to electrophysiological recordings and behavior as well as immunostaining and immunoblotting to investigate the role of trans-synaptic complexes involving presynaptic neurexins and cerebellins in shaping the function of central synapses. The study extends previous findings from the same authors as well as other groups showing an important role of these complexes in regulating the function of central synapses. Here, the authors sought to achieve two main objectives: (1) investigating whether their previous findings obtained at mature CA1-> subiculum synapses (Aoto et al., 2013; Dai et al., Neuron 2019; Dai et al., Nature 2021) extend to different synapse subtypes in the subiculum as well as to other central synapses including cortical and cerebellar synapses and (2) investigating whether Nrx-Cbln-GluD trans-synaptic complexes play a role in synapse formation as previously proposed by other groups.

    Overall, the study provides interesting and solid electrophysiological data showing that different Nrxns and Cblns assemble trans-synaptic complexes that differently regulate AMPAR and NMDA-mediated synaptic transmission across distinct synaptic circuits (most likely through binding to postsynaptic GluD receptors).

    However, the study has several important weaknesses:

    (1) The novelty of the findings appears limited. Indeed, previous studies from the same authors with similar experimental paradigms and readouts already demonstrated the role of Nrxn-Cbln-GluD complexes in regulating AMPARs versus NMDARs in mature neurons (Aoto et al., Cell 2013; Dai et al., Neuron 2019; Dai et al., Nature 2021). Moreover, the absence of role of Cblns and GluD receptors in synapse formation was already suggested in previous studies from the same authors (Seigneur and Sudhof, J Neurosci 2018; Seigneur et al., PNAS 2018; Dai et al., Nature 2021).

    (2) The conclusion made by the authors that the Nrxn-Cbln-GluD trans-synaptic complexes do not play a role in synapse formation/development is not sufficiently supported by their data, while previous studies suggest the opposite. Actually, this conclusion is essentially based on the two following measurements taken as a 'proxy' for synapse density: (1) 'the average vGluT1 intensity calculated from the entire area of subiculum' and (2) the 'synaptic proteins levels' assessed by immunoblotting. None of these measurements (only performed in the subiculum) allow to precisely assess synapse density on the neurons of interest. While the average vGluT1 intensity over large fields of view does not directly reflect the density of synapses and does not take into account the postsynaptic compartment, the immunoblotting data only reflects the overall expression of synaptic proteins without discriminating between intracellular, surface and synaptic pools and between cell types. In the subiculum from Cbln1+2 KO mice, the authors performed mEPSCs recordings and found an increase in frequency. However, this increase may reflect the unsilencing and/or potentiation of AMPAR-EPSCs above the detection threshold, irrespectively of the actual synapse number. Finally, the decrease in NMDAR-EPSCs is not discussed by the authors while it could actually reflect a decrease in synapse number.

    (3) The authors do not provide sufficient data in order to interpret the increase in AMPAR-EPSCs and decrease in NMDAR-EPSCs amplitudes. Are the changes in AMPARs and NMDARs occurring at pre-existing synapses or do they result from alterations in the number of physical synapses and/or active synapses (see point#2)? In particular, the increase in AMPAR/NMDAR ratio accompanied by the increase in mEPSCs frequency might be well explained by the unsilencing of some synapses and/or by the fact that the available pool of AMPARs is distributed over a smaller number of synapses, resulting in higher quantal size. These effects could explain the blockade of LTP, i.e., through an occlusion mechanism.

    (4) The authors did not demonstrate (or did not cite relevant studies) that the deletion of Cbln1 and/or Cbln2 does not affect the expression of the remaining Cblns isoforms (Cbln2 and/or Cbln4) or Nrxns1/3 and GluD1/2. This verification is important to preclude the emergence of any compensatory effect.