LAG3 is not expressed in human and murine neurons and does not modulate α‐synucleinopathies

  1. Marc Emmenegger
  2. Elena De Cecco
  3. Marian Hruska‐Plochan
  4. Timo Eninger
  5. Matthias M Schneider
  6. Melanie Barth
  7. Elena Tantardini
  8. Pierre de Rossi
  9. Mehtap Bacioglu
  10. Rebekah G Langston
  11. Alice Kaganovich
  12. Nora Bengoa‐Vergniory
  13. Andrès Gonzalez‐Guerra
  14. Merve Avar
  15. Daniel Heinzer
  16. Regina Reimann
  17. Lisa M Häsler
  18. Therese W Herling
  19. Naunehal S Matharu
  20. Natalie Landeck
  21. Kelvin Luk
  22. Ronald Melki
  23. Philipp J Kahle
  24. Simone Hornemann
  25. Tuomas P J Knowles
  26. Mark R Cookson
  27. Magdalini Polymenidou
  28. Mathias Jucker
  29. Adriano Aguzzi

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

No abstract available

Article activity feed

  1. Version published to 10.15252/emmm.202114745
  2. Review Commons

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

    Learn more at Review Commons

    Reply to the reviewers

    Reviewer #1 (Evidence, reproducibility and clarity (Required)): Very high evidence and clarity. Excellent scientific rigor. The findings are important and reported clearly. The experiments are conducted in a rigorous way by numerous participating laboratories. Reviewer #1 (Significance (Required)): Very high significance, both from a molecular biology and clinical standpoints. This is an important manuscript that challenges the findings and conclusions of a prior high-profile paper in Science by Ma et al 2016, claiming that LAG3 is a receptor for aggregation-prone species of alpha-synuclein and that deletion of LAG3 results in reduced cell to cell propagation of alpha-synuclein aggregates. The experiments in this paper are numerous and employ a variety of techniques. The overall conclusions are that LAG3 is not expressed by the relevant neurons and that LAG3 is not a receptor for alpha-synuclein fibrils (of different sizes). Therefore, the authors conclude that LAG3 is unlikely to play a role in the spread of alpha-synuclein pathology in Parkinson's disease and related disorders. There are, however, some weaknesses. For example, the Introduction contains passages that are not written in a stringent way: 1. "Histologically, PD is characterized by α-synuclein aggregates known as Lewy Bodies in neurons of the substantia nigra," That is not a good description of PD neuropathology. Lewy pathology is present in numerous areas of the CNS and PNS, and is not restricted to the substantia nigra.

    We have added a more detailed account:

    “Histologically, PD is characterized by α-synuclein inclusions known as Lewy Bodies whose accumulation is associated with neurodegeneration (Dickson, 2012; Mullin and Schapira, 2015; Corbillé et al., 2016). These inclusions affect the Substantia nigra and other mesencephalic regions as well as, in some cases, the amygdala and neocortex (Dickson, 2018).”

    1. "Growing evidence suggests that α-synuclein fibrils spread from cell to cell". While alpha-synuclein pathology can spread from cell to cell, it is not known if the fibrils are the species (alone or combined with other conformers) that cause the spreading of the pathology in a seeding fashion, or if smaller alpha-synuclein assemblies play that role.

    We have reformulated the sentence to credit the fact that we do not know which synuclein species is the one that is transmitted:

    “Growing evidence suggests that α-synuclein aggregates spread from cell to cell (Volpicelli-Daley et al., 2011; Volpicelli-Daley, Luk and Lee, 2014)… “

    1. "...by a "prionoid" process of templated conversion (Aguzzi, 2009; Aguzzi and Lakkaraju, 2016; Jucker and Walker, 2018; Kara, Marks and Aguzzi, 2018; Scheckel and Aguzzi, 2018; Uemura et al., 2020)." This sentence gives the impression that the corresponding author has led the field when it comes to alpha-synuclein's prionid properties. That is not really the case, and it would be appropriate to cite the literature in a more scholarly fashion that reflects how this part of the alpha-synuclein research field developed.

    I cannot disagree, and in fact I suspect that the present paper may be my second and possibly last experimental contribution to the synuclein field! However, I do claim intellectual parenthood of the prionoid (not “prionid”) concept, which I first expounded in a 2009 Nature paper. Anyway, we now provide a more balanced citation:

    “…by a “prionoid” process of templated conversion (Aguzzi, 2009; Jucker and Walker, 2018; Kara, Marks and Aguzzi, 2018; Henderson, Trojanowski and Lee, 2019; Karpowicz, Trojanowski and Lee, 2019; Uemura et al., 2020; Kara et al., 2021).“

    1. "Interrupting transmission of a-synuclein may slow down or abrogate the disease course." This is a bold statement and far from certain. While one might propose that this is the case, it is still just a hypothesis and the Introduction should reflect that.

    We have rewritten the sentence in a more subdued manner:

    “It is thought that interrupting transmission of a-synuclein may slow down or abrogate the disease course.”

    **Referee Cross-commenting** I concur with reviewers 2 and 3, and the new comment from reviewer 2. This paper should be published as soon as possible.

    *********************************************

    Reviewer #2 (Evidence, reproducibility and clarity (Required)): This study conclusively shows that LAG3 is not the receptor for a-synuclein that underlies the spread of synucleinopathic damage in various PD-related conditions. The paper is done extremely carefully and comprehensively. My only suggestion is to indicate the significance level in Figure 5a, as it may turn out that LAG3 is actually protective.

    We have added the significance level in Fig. 5A, in the legend: “The survivals of ASYNA53T LAG3-/-, LAG3+/- and LAG3+/+ mice were similar (Mantel-Cox log-rank test, p-value = 0.165).”

    Reviewer #2 (Significance (Required)): This study is of extremely high significance - we need mechanisms to deal with spectacular results in the literature that should not have been published because they are were uncompelling to begin with, but were published for various sociological/political reasons. Science won't progress if we don't find correction mechanisms for wrong conclusions. **Referee Cross-commenting** I agree with reviewers 1 and 3, especially with the suggestions made by reviewer 1, which should be instituted. I think we all concur that the paper should be published without new experiments. I believe testing a-synuclein propagation in vivo in LAG3 KO mice would be useful, but given the complete lack of replication of LAG3 expression in brain and of a-synuclein binding to LAG3, this is not necessary.

    We considered running experiments in addition to those performed in vivo in ASYNA53T transgenic mice (including LAG3 KO) and ex vivo in organotypic slices, the latter using pre-formed fibrils. However, the outcome of these experiments, along with the absence of LAG3 expression in neurons and its unclear binding, convinced us that the usage of further animals and reagents would be unwarranted.

    *****************************************

    Reviewer #3 (Evidence, reproducibility and clarity (Required)): It was proposed that LAG3 is important in the treatment of PD and related disorders, because it functions as a receptor of pathogenic α-synuclein and the treatment with anti-LAG3 antibodies attenuated the spread of pathological α-synuclein and drastically lowered the aggregation in vitro (Mao et al, Science 2016). In this study, authors characterized 8 antibodies to LAG3 and investigated the presence of LAG3 in cultured cell lines, NSC-derived neural cultures, or organ homogenates for the presence of human or murine LAG3. But it was not detected in any of the neuronal samples tested. In addition, single cell (sc) RNAseq yielded only minimal counts for the LAG3 transcript in neurons, astrocytes, and mixed glial cells, and single-nucleus (sn) RNAseq human brain dataset for LAG3 expression across different cell types confirmed no LAG3 signals for any of 34 identified cell clusters, including 13 clusters of excitatory and 11 subtypes of inhibitory neurons, oligodendrocytes, oligodendrocyte precursor cells, microglia, astrocytes, and endothelial cells. Authors also analyzed the binding of LAG3 with α-synuclein in mouse and human model systems, and concluded that the affinity of LAG3 for α-synuclein fibrils, if any, is micromolar or less. Furthermore, authors studied the propagation of pre-formed fibrils (PFFs) of α-synuclein in neural stem cell (NSC)-derived neural cultures in the presence or absence of LAG3, and the impact of LAG3 on survival in ASYNA53T transgenic mice expressing wild-type LAG3 as well as hemizygous or homozygous deletions thereof. However, they were unable to see any significant role for LAG3 in these in vitro and in vivo models of α-synucleinopathies. In this connection, the reviewer would like to ask one question: Have you conducted any experiments of the propagation of PFFs of α-synuclein in LAG3-KO mice ? If they did, what were the results ?

    We did consider the possibility of replicating the experiments using PFFs in LAG3 KO mice. However, as stated above, we felt that our experiments – including the survival study in vivo in ASYNA53T transgenic mice – were unambiguous. After critical consideration, we remained unconvinced that this additional experiment would change the weight of our evidence in a substantial manner that would justify the inoculation of other animals and the utilisation of more resources.

    **Minor point** In Page 10, I think it's a typo: ASYYN mice must be ASYN mice.

    Thank you for pointing this out. We corrected it.

    Reviewer #3 (Significance (Required)): These negative findings about the LAG in α-synucleinopathies shown in this manuscript do not provide any new insight into the mechanisms of α-synuclein propagation. However, it is clear that LAG3 is not expressed in neuronal cells and the binding of LAG3 to α-synuclein fibrils appears limited. Overexpression of LAG3 in cultured human neural cells did not cause any worsening of α-synuclein pathology ex vivo. The overall survival of A53T α- synuclein transgenic mice was unaffected by LAG3 depletion and the seeded induction of α-synuclein lesions in hippocampal slice cultures was unaffected by LAG3 knockout. These data shown in this manuscript are convincing and the information is very important in terms of correcting the direction of disease treatment and research. **Referee Cross-commenting** I agree with reviewers 1 and 2. This paper should be published as soon as possible.

  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 #3

    Evidence, reproducibility and clarity

    It was proposed that LAG3 is important in the treatment of PD and related disorders, because it functions as a receptor of pathogenic α-synuclein and the treatment with anti-LAG3 antibodies attenuated the spread of pathological α-synuclein and drastically lowered the aggregation in vitro (Mao et al, Science 2016).

    In this study, authors characterized 8 antibodies to LAG3 and investigated the presence of LAG3 in cultured cell lines, NSC-derived neural cultures, or organ homogenates for the presence of human or murine LAG3. But it was not detected in any of the neuronal samples tested. In addition, single cell (sc) RNAseq yielded only minimal counts for the LAG3 transcript in neurons, astrocytes, and mixed glial cells, and single-nucleus (sn) RNAseq human brain dataset for LAG3 expression across different cell types confirmed no LAG3 signals for any of 34 identified cell clusters, including 13 clusters of excitatory and 11 subtypes of inhibitory neurons, oligodendrocytes, oligodendrocyte precursor cells, microglia, astrocytes, and endothelial cells.

    Authors also analyzed the binding of LAG3 with α-synuclein in mouse and human model systems, and concluded that the affinity of LAG3 for α-synuclein fibrils, if any, is micromolar or less.

    Furthermore, authors studied the propagation of pre-formed fibrils (PFFs) of α-synuclein in neural stem cell (NSC)-derived neural cultures in the presence or absence of LAG3, and the impact of LAG3 on survival in ASYNA53T transgenic mice expressing wild-type LAG3 as well as hemizygous or homozygous deletions thereof. However, they were unable to see any significant role for LAG3 in these in vitro and in vivo models of α-synucleinopathies.

    In this connection, the reviewer would like to ask one question: Have you conducted any experiments of the propagation of PFFs of α-synuclein in LAG3-KO mice ? If they did, what were the results ?

    Minor point

    In Page 10, I think it's a typo: ASYYN mice must be ASYN mice.

    Significance

    These negative findings about the LAG in α-synucleinopathies shown in this manuscript do not provide any new insight into the mechanisms of α-synuclein propagation. However, it is clear that LAG3 is not expressed in neuronal cells and the binding of LAG3 to α-synuclein fibrils appears limited. Overexpression of LAG3 in cultured human neural cells did not cause any worsening of α-synuclein pathology ex vivo. The overall survival of A53T α- synuclein transgenic mice was unaffected by LAG3 depletion and the seeded induction of α-synuclein lesions in hippocampal slice cultures was unaffected by LAG3 knockout. These data shown in this manuscript are convincing and the information is very important in terms of correcting the direction of disease treatment and research.

    Referee Cross-commenting

    I agree with reviewers 1 and 2. This paper should be published as soon as possible.

  4. 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

    This study conclusively shows that LAG3 is not the receptor for a-synuclein that underlies the spread of synucleinopathic damage in various PD-related conditions. The paper is done extremely carefully and comprehensively. My only suggestion is to indicate the significance level in Figure 5a, as it may turn out that LAG3 is actually protective.

    Significance

    This study is of extremely high significance - we need mechanisms to deal with spectacular results in the literature that should not have been published because they are were uncompelling to begin with, but were published for various sociological/political reasons. Science won't progress if we don't find correction mechanisms for wrong conclusions.

    Referee Cross-commenting

    I agree with reviewers 1 and 3, especially with the suggestions made by reviewer 1, which should be instituted. I think we all concur that the paper should be published without new experiments. I believe testing a-synuclein propagation in vivo in LAG3 KO mice would be useful, but given the complete lack of replication of LAG3 expression in brain and of a-synuclein binding to LAG3, this is not necessary.

  5. 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

    Very high evidence and clarity. Excellent scientific rigor.

    The findings are important and reported clearly. The experiments are conducted in a rigorous way by numerous participating laboratories.

    Significance

    Very high significance, both from a molecular biology and clinical standpoints. This is an important manuscript that challenges the findings and conclusions of a prior high-profile paper in Science by Ma et al 2016, claiming that LAG3 is a receptor for aggregation-prone species of alpha-synuclein and that deletion of LAG3 results in reduced cell to cell propagation of alpha-synuclein aggregates.

    The experiments in this paper are numerous and employ a variety of techniques. The overall conclusions are that LAG3 is not expressed by the relevant neurons and that LAG3 is not a receptor for alpha-synuclein fibrils (of different sizes). Therefore, the authors conclude that LAG3 is unlikely to play a role in the spread of alpha-synuclein pathology in Parkinson's disease and related disorders.

    There are, however, some weaknesses. For example, the Introduction contains passages that are not written in a stringent way:

    1. "Histologically, PD is characterized by α-synuclein aggregates known as Lewy Bodies in neurons of the substantia nigra," That is not a good description of PD neuropathology. Lewy pathology is present in numerous areas of the CNS and PNS, and is not restricted to the substantia nigra.
    2. "Growing evidence suggests that α-synuclein fibrils spread from cell to cell". While alpha-synuclein pathology can spread from cell to cell, it is not known if the fibrils are the species (alone or combined with other conformers) that cause the spreading of the pathology in a seeding fashion, or if smaller alpha-synuclein assemblies play that role.
    3. "...by a "prionoid" process of templated conversion (Aguzzi, 2009; Aguzzi and Lakkaraju, 2016; Jucker and Walker, 2018; Kara, Marks and Aguzzi, 2018; Scheckel and Aguzzi, 2018; Uemura et al., 2020)." This sentence gives the impression that the corresponding author has led the field when it comes to alpha-synuclein's prionid properties. That is not really the case, and it would be appropriate to cite the literature in a more scholarly fashion that reflects how this part of the alpha-synuclein research field developed.
    4. "Interrupting transmission of a-synuclein may slow down or abrogate the disease course." This is a bold statement and far from certain. While one might propose that this is the case, it is still just a hypothesis and the Introduction should reflect that.

    Referee Cross-commenting

    I concur with reviewers 2 and 3, and the new comment from reviewer 2. This paper should be published as soon as possible.

  6. Version published to 10.1101/2021.04.25.441302v2 on bioRxiv
  7. Version published to 10.1101/2021.04.25.441302v1 on bioRxiv