All-Trans Retinoic Acid induces synaptic plasticity in human cortical neurons
Curation statements for this article:-
Curated by eLife
Summary: All three reviewers are highly enthusiastic about the study reporting the acute effects of retinoic acid on excitatory synaptic transmission and its underlying mechanisms. The experiments are well executed and the results convincing. Aside from some minor comments that require minimal additional experiments or further clarification, the reviewers expressed one major concern regarding the dentate gyrus LTP data. Although further experiments are required to clarify the concerns, the reviewers recommended removing the LTP figure from the present study as it is not well connected with the rest of the study.
This article has been Reviewed by the following groups
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
- Evaluated articles (eLife)
Abstract
A defining feature of the brain is its ability to adapt structural and functional properties of synaptic contacts in an experience-dependent manner. In the human cortex direct experimental evidence for synaptic plasticity is currently missing. Here, we probed plasticity in human cortical slices using the vitamin A derivative all-trans retinoic acid, which has been suggested as medication for the treatment of neuropsychiatric disorders, e.g., Alzheimer’s disease. Our experiments demonstrate coordinated structural and functional changes of excitatory synapses of superficial (layer 2/3) pyramidal neurons in the presence of all-trans retinoic acid. This synaptic adaptation is accompanied by ultrastructural remodeling of the calcium-storing spine apparatus organelle and requires mRNA-translation. We conclude that all-trans retinoic acid is a potent mediator of synaptic plasticity in the adult human cortex.
Article activity feed
-
Reviewer #3:
The manuscript by Dr. Vlachos group has demonstrated many important features as well as mechanisms of RA-induced synaptic plasticity. For example, they demonstrated that RA-induced plasticity happens in human neurons as well as in rodent neurons in vivo; discovery that synapodin as a critical mediator of RA plasticity as well as RA effect on the size of spine head, synaptopodin cluster and spine apparatus. Moreover, the effect of RA on in vivo LTP plasticity is very interesting. The data looks solid and supports the authors' conclusions.
However the manuscript can be significantly improved by discussion of their results, in the context with literature data as well as explaining the possible mechanism of their results.
- RA effect on AMPAR upregulation has been reported to not share the same SNARE mechanisms as electrical LTP …
Reviewer #3:
The manuscript by Dr. Vlachos group has demonstrated many important features as well as mechanisms of RA-induced synaptic plasticity. For example, they demonstrated that RA-induced plasticity happens in human neurons as well as in rodent neurons in vivo; discovery that synapodin as a critical mediator of RA plasticity as well as RA effect on the size of spine head, synaptopodin cluster and spine apparatus. Moreover, the effect of RA on in vivo LTP plasticity is very interesting. The data looks solid and supports the authors' conclusions.
However the manuscript can be significantly improved by discussion of their results, in the context with literature data as well as explaining the possible mechanism of their results.
- RA effect on AMPAR upregulation has been reported to not share the same SNARE mechanisms as electrical LTP (Synt1/7 independent vs dependent). How does RA have the extra effect on the LTP amplitude? Moreover, RA plasticity is recognized as a form of homeostatic synaptic plasticity, i.e., the effect takes hours to develop as shown by the authors of RA incubation of many hours in their experiment on human neurons. How does this compare with their RA manipulations in LTP exp (Is TA injected shortly before LTP stimulus? What do the author think that LTP stimulus does to RA signaling?)?
What about metaplasticity involves RA? any connections to the present study?
- The authors conclude that RA have effects on spines with or without spine apparatus, however, the authors' data suggest that RA-plasticity is blocked when spine apparatus is eliminated (with synaptopodin KO). Moreover, there is significant overlap of spine size for spines with or without spine apparatus... How do the authors interpret their results here? Is spine apparatus dynamic? can floating between spines quickly? Any literature on this? The authors need to discuss more on the possible ways, with supporting literature data, of how this spine apparatus can affect RA function.
In short, a discussion of the above points will add significance to the study.
-
Reviewer #2:
This paper explores the effect of all-trans retinoic acid (atRA) on synaptic plasticity in human and murine brain slices. The paper builds on previous work showing that atRA plays a key role in various forms of homeostatic and Hebbian plasticity, but extends our understanding in two very significant ways. First, the work convincingly shows that atRA enhances synaptic function in human layer 2/3 pyramidal neurons in intact cortical slices, and like previous studies using murine models and human ipSCs, this is critically dependent on new protein synthesis. Second, the studies show that atRA-mediated synaptic plasticity requires synaptopodin, a protein that is specifically localized to the spine apparatus.
Overall, the studies have been well-executed and the data are both rigorous and convincing. The paper is very clearly …
Reviewer #2:
This paper explores the effect of all-trans retinoic acid (atRA) on synaptic plasticity in human and murine brain slices. The paper builds on previous work showing that atRA plays a key role in various forms of homeostatic and Hebbian plasticity, but extends our understanding in two very significant ways. First, the work convincingly shows that atRA enhances synaptic function in human layer 2/3 pyramidal neurons in intact cortical slices, and like previous studies using murine models and human ipSCs, this is critically dependent on new protein synthesis. Second, the studies show that atRA-mediated synaptic plasticity requires synaptopodin, a protein that is specifically localized to the spine apparatus.
Overall, the studies have been well-executed and the data are both rigorous and convincing. The paper is very clearly written and the findings are significant. This is a very strong body of work that will be of broad interest.
Comments:
While the authors rightly point out in the introduction that no previous studies have assessed atRA effects in human cortical circuits, the Zhang et al. (2018) paper did elegantly show synaptic plasticity effects in human neurons (derived from ipSCs). This is noted in the discussion, but should also be pointed out in the introduction as it bears directly on the rationale for the studies described in the paper.
Figure 1C illustrates responses of layer 2/3 pyramidal neurons to intracellular current injection. While the passive membrane properties are quantified and similar regardless of atRA exposure, it is not clear if atRA affects intrinsic excitability of these neurons (i.e., the number of spikes elicited by different levels of injected current). These data should be included.
The legend for Figure 1 C-E is too vague and does not describe the specific measures that are shown in the figure.
For the mouse studies shown in Figure 3A and 3B, did wild-type littermates serve as controls (the gold standard)? Data from wild-type neurons is described in the text but it is not clear if these were collected from a different cohort of animals or from the WT littermates of the Synpo-deficient mice. Also, the authors should state whether the deficient allele is null.
The Synpo-deficient mice have basal sEPSC amplitudes that are noticeably larger than WT mice (as reported in the text). Some discussion of this observation is warranted.
The cumulative frequency plots shown throughout the paper show a curious trend where the smallest events appear to be at least 10 pA or larger. This is somewhat atypical, as most studies find a large number of events between 5 and 10 pA (and many lower still). Does this reflect events only larger than 10 pA being included in the analysis? If so, the points to the left of 10 pA should probably be removed from these plots as including them implies that this data range was adequately sampled.
The schematic shown in Fig4B refers to early-phase and late-phase LTP, but the recordings appear to be limited to 60 min post-LTP induction (i.e well before the late-phase). These terms should be replaced with the actual times post-LTP induction.
The discussion is quite on point, but is rather brief. The paper would benefit from a more detailed discussion of the link between the spine-apparatus and translation-dependent forms of synaptic plasticity.
-
Reviewer #1:
The study by Lenz et al. explores the acute action of retinoic acid (RA) in adult human cortical neurons. The main findings are:
Consistent with previous findings in mouse neurons, the authors reported enhanced excitatory synaptic transmission in RA-treated cortical layer 2/3 neurons.
Also consistent with previous findings, this enhancement is independent of gene transcription, but requires protein synthesis.
RA's effect on EPSC requires expression of an actin-modulating protein called synaptopodin. In the Synaptopodin deficient mouse mPFC neurons, RA's effect on EPSC is eliminated. Moreover, in synaptopodin deficient hippocampal dentate gyrus neurons, enhancement of LTP by RA is also reversed.
Overall, this study demonstrates RA-induced synaptic plasticity in acute human cortical neurons, thus expanding the previous findings …
Reviewer #1:
The study by Lenz et al. explores the acute action of retinoic acid (RA) in adult human cortical neurons. The main findings are:
Consistent with previous findings in mouse neurons, the authors reported enhanced excitatory synaptic transmission in RA-treated cortical layer 2/3 neurons.
Also consistent with previous findings, this enhancement is independent of gene transcription, but requires protein synthesis.
RA's effect on EPSC requires expression of an actin-modulating protein called synaptopodin. In the Synaptopodin deficient mouse mPFC neurons, RA's effect on EPSC is eliminated. Moreover, in synaptopodin deficient hippocampal dentate gyrus neurons, enhancement of LTP by RA is also reversed.
Overall, this study demonstrates RA-induced synaptic plasticity in acute human cortical neurons, thus expanding the previous findings from mouse neurons and immature human neurons induced from iPS cells to adult human cortical neurons.
Specific Comments:
Figure 3 shows that in synaptopodin deficient mouse neurons, RA no longer increases sEPSC amplitudes. The rescue experiments are very nice. However, in both WT neurons (stated in main text, not in figure) and rescue neurons (Fig. 3B), the baseline sEPSC amplitudes are significantly smaller than those of the KO neurons. Can the authors speculate why deletion of synaptopodin may lead to enhanced basal excitatory synaptic transmission?
The LTP experiments are a bit problematic. First of all, it was done in mouse hippocampal DG neurons, not cortical neurons. The effect of RA may be different in different neuronal types, as has been shown in previous mouse studies. It will be nice to examine whether RA changes basal synaptic transmission in these neurons in acute slices. Without knowing the effect on basal transmission, it is hard to interpret the LTP results. Second, why did WT DG show no LTP? Third, previous work by Arendt et al. (2015) showed that RA enhances hippocampal CA1 neuron basal EPSCs, and occludes further LTP. The observation here in the DG with RA treatment points the opposite direction. Can the authors offer some explanation (i.e. RA alters LTP threshold through some kind of priming)? Again, knowing the effect of RA on basal transmission specifically in the DG neurons would be informative toward understanding the effect on LTP.
The pharmacological treatments (ActD, anisomycin etc.) in this study are in general very long (6 hr) compared to conventional methods (less than 2 hr). To control for potential toxicity associated with prolonged treatment, vehicle control should be added in both Fig 5 and Fig 6.
-
Summary: All three reviewers are highly enthusiastic about the study reporting the acute effects of retinoic acid on excitatory synaptic transmission and its underlying mechanisms. The experiments are well executed and the results convincing. Aside from some minor comments that require minimal additional experiments or further clarification, the reviewers expressed one major concern regarding the dentate gyrus LTP data. Although further experiments are required to clarify the concerns, the reviewers recommended removing the LTP figure from the present study as it is not well connected with the rest of the study.
-
