CNTN4 modulates neural elongation through interplay with APP

  1. Rosemary A. Bamford
  2. Amila Zuko
  3. Jan J. Sprengers
  4. Harm Post
  5. Renske L. R. E. Taggenbrock
  6. Annika Mehr
  7. Owen J. R. Jones
  8. Aurimas Kudzinskas
  9. Josan Gandawijaya
  10. Madeline Eve
  11. Ulrike C. Müller
  12. Martien J. Kas
  13. J. Peter H. Burbach
  14. Asami Oguro-Ando

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

The neuronal cell adhesion molecule contactin-4 (CNTN4) has been genetically linked to autism spectrum disorders (ASD) and other psychiatric disorders. The Cntn4 -deficient mouse model has previously shown that CNTN4 has important roles in axon guidance and synaptic plasticity in the hippocampus. However, the pathogenesis and functional role of CNTN4 in the cortex have not yet been investigated.

Using Nissl staining, immunohistochemistry and Golgi staining the motor cortex of Cntn4 -/- mice was analysed for abnormalities. Interacting partners of CNTN4 were identified by immunoprecipitation and mass spectrometry. Further analysis of the interaction between CNTN4 and APP utilised knockout human cells generated via CRISPR-Cas9 gene editing.

Our study newly identified reduced cortical thickness in the motor cortex of Cntn4 -/- mice, but cortical cell migration and differentiation were unaffected. Significant morphological changes were observed in neurons in the M1 region of the motor cortex, indicating that CNTN4 is also involved in the morphology and spine density of neurons in the motor cortex. Furthermore, mass spectrometry analysis identified an interaction partner for CNTN4, and we confirmed an interaction between CNTN4 and APP. Knockout human cells of CNTN4 and/or APP revealed a relationship between CNTN4 and APP.

This study demonstrates that CNTN4 contributes to cortical development, and that its binding and interplay with APP controls neural elongation. This is an important finding for understanding the function of APP, a target protein for Alzheimer’s disease. The binding between Cntn4 and APP, which is involved in neurodevelopment, is essential for healthy nerve outgrowth.

Article activity feed

  1. Review Commons

    Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    1. General Statements [optional]

    Thank you for the constructive comments and suggestions from the reviewers to further strengthen our manuscript (RC-2023-02156) entitled, “CNTN4 modulates neural elongation through interplay with APP”. In response to the reviewer’s comments, we have outlined the following revision plan. Please find the point-by-point responses to the reviewer comments in red. All the additions and changes in the manuscript are shown as track changes. We trust that the revised manuscript and revision plan will meet the approval of the editor and reviewers. We would also be glad to respond to any further questions and comments that you may have.

    2. Description of the planned revisions

    Insert here a point-by-point reply that explains what revisions, additional experimentations and analyses are planned to address the points raised by the referees.

    Reviewer 2 6) Figure 8 C-E shows a reduction of APP mRNA in SH-SY5Y knockdown of CNTN4 and a reduction CNTN4 mRNA in SH-SY5Y knockdown for APP. These data suggest might suggest that "interaction between CNTN4 and APP contributes to their gene expression". However, this observation needs to be proved mainly in the CNTN4 and APP KO mice.

    Thank you for your insightful comment. We recognize the importance of validating this observation in CNTN4 and APP knockout mice. In line with your suggestion, we are currently conducting these experiments and plan to incorporate the results into our revised manuscript.

    3. Description of the revisions that have already been incorporated in the transferred manuscript

    Please insert a point-by-point reply describing the revisions that were already carried out and included in the transferred manuscript. If no revisions have been carried out yet, please leave this section empty.

    Reviewer #1 Page 12, lines 7-9. The conclusion here is that CNTN4 KO and APP KO phenotypes are different. But a more interesting way to look at it is that the Scholl analysis of dendrites shows that they are almost exactly the reverse of each other with regard to near and distal morphologic differences. That could be interesting. Maybe CNTN4 binding inhibits APP signaling or binding to another partner, or APP binding inhibits CNTN4 signaling or binding to another partner, or both. Then loss of either one would hyper-activate the signal induced by the other and give the observed yin-yang phenotypic relationship. I note that this would not fit with the neuroblastoma phenotypes, which seem to be in the same direction; but a developing brain is different in many ways from a neuroblastoma cell in culture. The Discussion is also somewhat vague about possible interpretations of the two phenotypes in vivo.

    Thank you for your valuable insights. In response to your comments, we have expanded our discussion (line 420) to include the following considerations: “Our hypothesis is that when CNTN4 is deficient there are two possibilities considered, 1) the function to which the binding of CNTN4 to APP contributes is lost, and the ability of CNTN4 to regulate dendritic spine formation diminishes which would cause abnormal neurite outgrowth (Figure 8); 2) the loss of CNTN4 would cause other proteins to alternately bind to the E1 domain of APP and affect neurite outgrowth and arborization. For example, the arborization trends of the near and distal Sholl apical and basal dendrites are the opposite of one another in the Cntn4- and App-deficient mice, respectively. Loss of either CNTN4 or APP may activate these opposing scenarios through inhibition of signaling, binding to another partner or both. However, further studies are needed to understand the interplay.”

    Reviewer #1 Where is APP normally expressed in the developing mouse brain? A few sentences in the introduction or discussion would be helpful. This information was provided for Cntn4 in the introduction.

    Thank you for this suggestion. We have now included important background information on APP expression in the Introduction (line 95): “Osterfield et al. has previously shown a direct binding between CNTN4 and transmembrane amyloid-beta precursor protein (APP). Expression of APP in mice has been observed early in development, and is ubiquitously expressed in adult mice (45).”

    Reviewer #1 Minor grammatical items.

    Page 3, line 6. "Over 1000 genes..."

    Page 4, line 17. "...battery to study Cntn4-deficient mice, which revealed subtle..."

    Page 5. First sentence in the Results section. "the cortical layer thickness is involved in migration" - that phrase needs to be reworked.

    Page 6, line 16. "total numbers of cells or of neurons"

    Page 8, line 12. "maturity morphologies" - that phrase needs to be reworked.

    Page 8, line 16. "In vitro primary cell culturing..." should be "Primary cell culturing...". After all, the only kind of cell culturing is in vitro.

    Thank you for pointing out these grammatical issues. We appreciate your attention to detail and have carefully revised each of the mentioned sections in the manuscript to ensure clarity and accuracy. The corrections have been implemented as follows:

    Page 3, line 6: The phrase "Over 1000 genes..." has been added into line 52

    Page 4, line 17: "...battery to study Cntn4-deficient mice, which revealed subtle..." has been added to line 88

    Page 5, first sentence in the Results section is now “In the cerebral cortex, the cortical layer thickness is related to migration and may be an indicator of neurodevelopment abnormalities.” (line 113)

    Page 6, line 16: "total numbers of cells or of neurons" has been added to line 138

    Page 8, line 12: The term "maturity morphologies" has been replaced by “spine morphologies” in line 183.

    Page 8, line 16: The phrase "In vitro primary cell culturing..." has been replaced by "Primary cell culturing..." (line 188)

    Reviewer #2 1) Figure 1A-E shows the organization of cortical layers in the CNTN4+/- and CNTN4-/- respect to WT mice. Looking at the NeuN staining the images in C show a reduction of the NeuN+ neurons of the upper layer in the CNTN4-/- mice with respect to WT mice, confirmed by the quantification and an increase of the NeuN+ neurons of the lower layer in the CNTN4-/-mice respect to WT mice, not confirmed by the quantification. Or upper layer thickness is reduced and the density of the NeuN+ is not changed. Also, surprisingly the Cux1/NeuN+ neurons are reduced in the CNTN4+/- compared to CNTN4-/- and WT mice, but images for the CNTN4+/- were not shown. These results need to be better clarified.

    Thank you for your insightful comments. We understand the importance of clearly presenting the staining differences among the layers however were faced with limitations due to space constraints in including the Cntn4+/- images initially. In response, we have revised Figure 1 to display the three phenotypes (Cntn4+/+, Cntn4+/-, and Cntn4-/-) side-by-side using smaller images for a more comprehensive comparison. Regarding the quantification presented in Figure 1D and 1E, our analysis indicates that while there is a reduction in the number of NeuN+ neurons in the upper layers of the Cntn4-/- mice, this reduction is not statistically significant when compared to the Cntn4+/+ mice. A similar pattern is observed in the lower layers. However, we did observe a significant reduction in the thickness of the upper layer in the Cntn4-/- mice, which suggests a change in cell density, even though the overall number of cells (DAPI+) remains consistent across phenotypes. This implies a shift in the proportion of cells in the Cntn4-/- mice. We have now clarified this in the results section, emphasizing the change in neuronal proportion rather than density (line 144), to better convey our findings.

    Reviewer #2 2) Figure 3 shows the quantification of dendritic spines but there are no images to support the quantification, in particular, it's unclear how "abnormal spines" were morphologically defined.

    Thank you for your valuable feedback regarding Figure 3. In response to your comment, we have now incorporated representative images of the apical dendritic spines into Figure 3. These images feature white arrows pointing to specific examples of spine morphology, thereby visually supporting our quantification. To further clarify how the dendritic spines were morphologically categorized, we have updated both the Materials and Methods section under 'Golgi Staining' and the legend of Figure 3. Furthermore, we have referenced a pertinent study in the Methods section where similar categorization of spine morphology has been undertaken. This citation provides a methodological context and validation for our approach in spine classification. Additionally, the reader can refer to Figure 3A schematic for examples.

    Reviewer #2 5) Figure 6 shows a nice characterization of dendrite arborizations in the APP-/- mice. However, these results are not really related to the function of CNTN4. Indeed, the minimal alteration in the number of apical dendrite tips that have been described in the APP-/- mice might be due to a function of APP unrelated to the interaction with CNTN4.

    Thank you for highlighting this aspect of our study. We acknowledge that the findings in the APP-/- mice, as presented in Figure 6, might initially seem tangential to the primary focus on CNTN4. However, our intention in examining the App-/- mice was guided by prior studies indicating a potential link between APP and the pyramidal neuron phenotypes observed in cortical neurons. This exploration was aimed at broadening our understanding of APP's role in neuronal development, which, while not exclusively tied to its interaction with CNTN4, is nevertheless relevant to the overarching context of our research. To address your concern and enhance clarity, we have made an explicit statement in the manuscript's discussion section (line 383): “Results in the App-/- mice cannot be attributed solely to any interaction APP may have with CNTN4.”

    Reviewer #2 4) Figures 6 B and C should show the colocalization of CNTN4 with APP, however, it's difficult to see any colocalization with images at this low magnification. Please provide images at higher magnification.

    Thank you for the comment. We have adjusted Figure 6B and 6C to include zoomed in versions of the existing image to highlight regions of colocalization. It should also be noted from the description in the main text results that the expression pattern isn’t just colocalization.

    Reviewer #2 3) The results of Figure 4 do not really provide a significant clue regarding the function of CNTN4 in relation to all the other data presented in this paper. Also, the staining of CNTN4 should be shown.

    Thank you for your feedback regarding Figure 4. We understand your concerns about the relevance of these results to the overall function of CNTN4 as explored in our study. Our objective with Figure 4 was to contrast the effects of CNTN4 overexpression in primary cultured neurons with the phenotypes observed in the CNTN4 knockout detailed elsewhere in the manuscript. This comparison was intended to provide a more comprehensive understanding of CNTN4's role in neuronal development and function. To address your point about the visualization of CNTN4, we have now included more explicit details in the legend of Figure 4 (line 1210). Both the full-length Cntn4 construct and the empty pcDNA3.1 control vector used in our experiments are tagged with GFP.

    Reviewer #1 Page 12, bottom. One would expect that the effect of CRISPR KO would be a complete elimination of the western blot band. It appears from the Figure and text that there is a little bit of residual signal. Could that band simply be cross-reactivity with another protein? Could it be protein contamination from serum? Or is the cell line not clonally pure?

    Thank you for your comment and for pointing out the ambiguity in our manuscript. We have now revised the relevant sections to describe mRNA and protein expression levels more accurately.

    To clarify, in our study, CRISPR knockout effectively eliminated CNTN4 protein expression in the CNTN4 knockout cell line, and similarly, APP protein expression was completely diminished in the APP knockout cell line. This complete reduction aligns with the expected outcomes of successful CRISPR knockout. Furthermore, we observed that the level of CNTN4 protein expression in the APP knockout cell lines showed a reduction of approximately 50%. Similarly, the level of APP protein expression in the CNTN4 knockout cell lines was reduced by about 50%. We believe these findings suggest an interdependent regulatory mechanism between CNTN4 and APP, which we have now elaborated upon in the revised manuscript (line 285).

    Reviewer #2 7) The discussion is too long and needs to be more concise.

    Thank you for your feedback regarding the length of our discussion section. We appreciate your guidance on enhancing the manuscript's clarity and focus. In response to your comment, we have thoroughly reviewed and condensed the discussion. Our aim was to streamline the content without compromising the coverage of our broad study.

    4. Description of analyses that authors prefer not to carry out

    Please include a point-by-point response explaining why some of the requested data or additional analyses might not be necessary or cannot be provided within the scope of a revision. This can be due to time or resource limitations or in case of disagreement about the necessity of such additional data given the scope of the study. Please leave empty if not applicable.

    Reviewer #1 Page 11. It isn't clear whether the binding of a soluble protein ligand to a cell-surface protein is measuring cis or trans binding configurations. It could be either or both, depending on the geometry of the interaction. Demonstrating a bone fide cis interaction is not easy - that requires a FRET experiment with tagged cell-surface proteins or a cryoEM structure.

    Thank you for your insightful suggestion regarding the experimental approach to differentiate between cis and trans binding configurations. We acknowledge the importance of distinguishing these interactions and the potential insights that such experiments, like FRET with tagged proteins or cryoEM structure analysis, could provide. However, after careful consideration, we have concluded that incorporating these specific methodologies would extend beyond the current scope of our paper. While the suggestion of a more detailed examination through FRET or cryoEM is undoubtedly valuable, it would necessitate a separate set of experimental conditions and analyses, potentially forming the basis for future research.

  2. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    In this paper, Bamford et al. showed a reduced cortical thickness in the motor cortex of Cntn4-/- mice, but cortical cell migration and differentiation were unaffected. They also found morphological changes in neurons in the M1 region of the motor cortex, indicating that CNTN4 is also involved in the morphology and spine density of neurons in the motor cortex. With mass spectrometry analysis they identified a number of interaction partners for CNTN4, among then they confirmed a previously demonstrated interaction between CNTN4 and APP. Thus, this study demonstrates that CNTN4 contributes to cortical development and that its binding and interplay with APP might also be important for neural elongation.

    The paper shows nice and convincing results, but the following points should be fully addressed in order to improve the significance of these findings.

    1. Figure 1A-E shows the organization of cortical layers in the CNTN4+/- and CNTN4-/- respect to WT mice. Looking at the NeuN staining the images in C show a reduction of the NeuN+ neurons of the upper layer in the CNTN4-/- mice with respect to WT mice, confirmed by the quantification and an increase of the NeuN+ neurons of the lower layer in the CNTN4-/-mice respect to WT mice, not confirmed by the quantification. Or upper layer thickness is reduced and the density of the NeuN+ is not changed. Also, surprisingly the Cux1/NeuN+ neurons are reduced in the CNTN4+/- compared to CNTN4-/- and WT mice, but images for the CNTN4+/- were not shown. These results need to be better clarified.
    2. Figure 3 shows the quantification of dendritic spines but there are no images to support the quantification, in particular, it's unclear how "abnormal spines" were morphologically defined.
    3. The results of Figure 4 do not really provide a significant clue regarding the function of CNTN4 in relation to all the other data presented in this paper. Also, the staining of CNTN4 should be shown.
    4. Figures 6 B and C should show the colocalization of CNTN4 with APP, however, it's difficult to see any colocalization with images at this low magnification. Please provide images at higher magnification.
    5. Figure 6 shows a nice characterization of dendrite arborizations in the APP-/- mice. However, these results are not really related to the function of CNTN4. Indeed, the minimal alteration in the number of apical dendrite tips that have been described in the APP-/- mice might be due to a function of APP unrelated to the interaction with CNTN4.
    6. Figure 8 C-E shows a reduction of APP mRNA in SH-SY5Y knockdown of CNTN4 and a reduction CNTN4 mRNA in SH-SY5Y knockdown for APP. These data suggest might suggest that "interaction between CNTN4 and APP contributes to their gene expression". However, this observation needs to be proved mainly in the CNTN4 and APP KO mice.
    7. The discussion is too long and needs to be more concise.

    Significance

    This study demonstrates that CNTN4 contributes to cortical development and that its binding and interplay with APP might also be important for neural elongation. However, the role of APP interaction on CNTN4 function was not well demonstrated.

  3. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    9-17-2023

    Bamford et al present a very thorough series of experiments to explore the role of CNTN4 and its interacting partner APP in the development of the mouse cerebral cortex. The manuscript presents the following analyses/experiments: (1) a thickness analysis of different cortical areas in the CNTN4 KO mouse, finding reduced cortical thickness and reduced cell numbers in motor cortex; (2) a pyramidal dendrite and spine morphologic analysis in motor cortex in the CNTN4 KO mouse, finding defects in proximal dendritic morphology and altered spine distribution; (3) a biochemical (mass spectrometry) search for CNTN binding partners in 293 cells, finding APP among others and confirming a previous finding; (4) a cell adhesion analysis in transfected 293 cells showing a trans interactions between the extracellular domains of CNTN4 and APP; (5) an analysis of pyramidal dendritic morphology in APP KO mouse motor cortex that looks like the reciprocal of the Scholl analysis phenotype observed with CNTN4 KO mouse cortex, and (6) a demonstration that neurite outgrowth is reduced in cultured SH-SY5Y neuroblastoma cells with CRISPR-mediated deletion mutations of either CNTN4, APP, or both. The experiments look to be technically well done, and they data are interpreted with an appropriate level of caution.

    This work will be of substantial interest to scientists working on CNS development in general and cortical development in particular. Genetic variation in CNTN4 has been implicated in developmental disorders, such as autism spectrum disorder, and the present work represents an important step forward in defining the mechanistic underpinning of that phenotype. This work will also be of interest to those studying Alzheimer disease who wonder (as many have) what the normal function of APP is.

    Minor comments:

    Page 11. It isn't clear whether the binding of a soluble protein ligand to a cell-surface protein is measuring cis or trans binding configurations. It could be either or both, depending on the geometry of the interaction. Demonstrating a bone fide cis interaction is not easy - that requires a FRET experiment with tagged cell-surface proteins or a cryoEM structure.

    Page 12, lines 7-9. The conclusion here is that CNTN4 KO and APP KO phenotypes are different. But a more interesting way to look at it is that the Scholl analysis of dendrites shows that they are almost exactly the reverse of each other with regard to near and distal morphologic differences. That could be interesting. Maybe CNTN4 binding inhibits APP signaling or binding to another partner, or APP binding inhibits CNTN4 signaling or binding to another partner, or both. Then loss of either one would hyper-activate the signal induced by the other and give the observed yin-yang phenotypic relationship. I note that this would not fit with the neuroblastoma phenotypes, which seem to be in the same direction; but a developing brain is different in many ways from a neuroblastoma cell in culture. The Discussion is also somewhat vague about possible interpretations of the two phenotypes in vivo.

    Page 12, bottom. One would expect that the effect of CRISPR KO would be a complete elimination of the western blot band. It appears from the Figure and text that there is a little bit of residual signal. Could that band simply be cross-reactivity with another protein? Could it be protein contamination from serum? Or is the cell line not clonally pure?

    Where is APP normally expressed in the developing mouse brain? A few sentences in the introduction or discussion would be helpful. This information was provided for Cntn4 in the introduction.

    Minor grammatical items.

    Page 3, line 6. "Over 1000 genes..."

    Page 4, line 17. "...battery to study Cntn4-deficient mice, which revealed subtle..."

    Page 5. First sentence in the Results section. "the cortical layer thickness is involved in migration" - that phrase needs to be reworked.

    Page 6, line 16. "total numbers of cells or of neurons"

    Page 8, line 12. "maturity morphologies" - that phrase needs to be reworked.

    Page 8, line 16. "In vitro primary cell culturing..." should be "Primary cell culturing...". After all, the only kind of cell culturing is in vitro.

    Significance

    This work will be of substantial interest to scientists working on CNS development in general and cortical development in particular. Genetic variation in CNTN4 has been implicated in developmental disorders, such as autism spectrum disorder, and the present work represents an important step forward in defining the mechanistic underpinning of that phenotype. This work will also be of interest to those studying Alzheimer disease who wonder (as many have) what the normal function of APP is.

  4. Version published to 10.1101/2023.08.25.554833v2 on bioRxiv
  5. Version published to 10.1101/2023.08.25.554833v1 on bioRxiv