Retinoic acid-induced protein 14 controls dendritic spine dynamics associated with depressive-like behaviors
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
In this manuscript, the authors discovered a new function of Rai14, an F-actin binding protein, in dendritic spine dynamics. They showed that Rai14 is localized at the spine neck and regulate spine density and function. Heterozygous Rai14 knockout mice showed impaired learning and memory and depressive-like behavior. Overall, this study provides novel insights into the molecular mechanisms underlying spine dynamics and depressive-like behavior. The main conclusions are supported by the 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. Reviewer #2 agreed to share their name with the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Neuroscience (eLife)
Abstract
Dendritic spines are the central postsynaptic machinery that determines synaptic function. The F-actin within dendritic spines regulates their dynamic formation and elimination. Rai14 is an F-actin-regulating protein with a membrane-shaping function. Here, we identified the roles of Rai14 for the regulation of dendritic spine dynamics associated with stress-induced depressive-like behaviors. Rai14-deficient neurons exhibit reduced dendritic spine density in the Rai14 +/- mouse brain, resulting in impaired functional synaptic activity. Rai14 was protected from degradation by complex formation with Tara, and accumulated in the dendritic spine neck, thereby enhancing spine maintenance. Concurrently, Rai14 deficiency in mice altered gene expression profile relevant to depressive conditions and increased depressive-like behaviors. Moreover, Rai14 expression was reduced in the prefrontal cortex of the mouse stress model, which was blocked by antidepressant treatment. Thus, we propose that Rai14-dependent regulation of dendritic spines may underlie the plastic changes of neuronal connections relevant to depressive-like behaviors.
Article activity feed
-
-
Author Response:
Reviewer #1 (Public Review):
The manuscript by Kim et al. identifies a new role for the F-actin binding protein Rai14 in dendritic spine dynamics. The authors demonstrate both in mice and in culture that Rai14-deficient neurons have decreased dendritic spine density, which corresponds with a reduction in excitatory synapse density and the frequency of miniature excitatory postsynaptic currents (mEPSCs). They also provide convincing evidence that Rai14 is protected from degradation through an interaction with another F-actin binding protein, Tara, and that the two proteins accumulate together in dendritic spines necks when overexpressed in neurons, resulting in enhanced spine maintenance. Characterization of Rai14+/- mice revealed that mice display learning and memory deficits and depressive-like behaviors, and that …
Author Response:
Reviewer #1 (Public Review):
The manuscript by Kim et al. identifies a new role for the F-actin binding protein Rai14 in dendritic spine dynamics. The authors demonstrate both in mice and in culture that Rai14-deficient neurons have decreased dendritic spine density, which corresponds with a reduction in excitatory synapse density and the frequency of miniature excitatory postsynaptic currents (mEPSCs). They also provide convincing evidence that Rai14 is protected from degradation through an interaction with another F-actin binding protein, Tara, and that the two proteins accumulate together in dendritic spines necks when overexpressed in neurons, resulting in enhanced spine maintenance. Characterization of Rai14+/- mice revealed that mice display learning and memory deficits and depressive-like behaviors, and that they have reduced expression of a number of genes identified in major depressive disorder gene set. Finally, the authors show that chronic restraint stress results in a decrease in mRNA and protein expression of Rai14, and that treatment with the antidepressant fluoxetine can rescue depressive-like behavior and reduced spine density in Rai14+/- mice as well as prevent a reduction in Rai14 expression following chronic restraint stress in wild-type (WT) mice. Together, these results identify Rai14 as a novel regulator of dendritic spine dynamics that may play a role in stress-induced depressive-like phenotypes. While the individual conclusions made by the authors are interesting and generally supported by the data (although in some cases missing important details/analyses), the evidence connecting the various findings together to provide proof that Rai14 is involved in regulating dendritic spine dynamics associated with depressive-like behaviors (as the title suggests) is still somewhat lacking and could be further strengthened.
- In Figure 1, the authors use Golgi staining of WT and Rai14+/- mouse brain slices as well as primary neuron cultures from WT and Rai14-/- mice and shRNA knockdown of Rai14 to demonstrate that Rai14 loss leads to a reduction in dendritic spine density in cortical and hippocampal neurons. From this data, the authors conclude that Rai14 is required to maintain a normal number of dendritic spines. However, some important details and analyses are missing in these experiments. For instance, in Figure 1A and 1B, the authors do not specify which hippocampal or cortical brain regions (or cell types) they are analyzing in the WT or Rai14+/- mice. In Figure 1C-E, the authors claim there is a reduction in mature dendritic spine density in Rai14-/- neurons compared to WT neurons, but they do not detect differences in spine length or spine head width. It would be useful if the authors could include a description of how they are defining "mature spines". The authors also claim that the reduction in spine density on Rai14-deficient neurons is due to a maintenance phenotype, rather than a formation phenotype, but they do not present evidence to differentiate between these two possibilities. Have the authors examined younger Rai14+/- mice (or Rai14-/- neurons) to determine when the spine phenotype is first detected (i.e. do spines form and then are lost, or do they fail to form correctly in the first place)? The authors attempt to address this question in Figure 3 with experiments in neurons overexpressing Rai14 and Tara, but it might also be useful to look at earlier timepoints in Rai14+/- mice and/or time-lapse imaging of Rai14-deficient neurons.
-> In response to the reviewer’s concern, we added information on brain regions, cell types, time points, and spine classification criteria analyzed in both figure legends and the Materials and Method section.
->Regarding the formation vs. maintenance issue, we have tried to address the role of Rai14 in dendritic spine maintenance by observing spine dynamics under naïve condition and spine elimination-induced condition. Spines containing the Rai14 cluster at their neck rarely disappeared during the imaging period (Figure 3D and 3E). Newly formed spines in which Rai14 recruited became stable, whereas newly formed dendritic protrusions in which Rai14 did not gather gradually disappeared (Figure 3F). In addition, dendritic spines from neurons overexpressing Rai14 and Tara were more resistant to spine loss caused by LatA treatment (Figure 3G and 3H), suggesting that Rai14 would take part in dendritic spine maintenance.
-> However, we agree that our data cannot exclude the possibility that the formation phenotype also affected dendritic spine loss in Rai14 deficient neurons. Therefore, we modified some expressions from the text as follows:
- Abstract, line 18: “Rai14-deficient neurons failed to maintain a proper dendritic spine density in the Rai14+/- mouse brain,” -> “Rai14-deficient neurons exhibit reduced dendritic spine density in the Rai14+/- mouse brain,”
- Result, line 64: “Rai14-depleted neurons fail to maintain a normal number of dendritic spines” -> “Rai14-depleted neurons exhibit decreased dendritic spine density”
- Figure 1 title: “Rai14-depleted neurons fail to maintain a normal number of dendritic spines” -> “Rai14-depleted neurons exhibit decreased dendritic spine density”
- In Figure 3, the authors report the interesting observation that overexpressed Rai14 and Tara accumulate in the necks of dendritic spines, which requires Rai14's ankyrin repeat domains, and that spines containing overexpressed Rai14 are less likely to be eliminated than spines lacking Rai14 clusters, and that neurons overexpressing Rai14 and Tara are resistant to spine loss caused by treatment with the actin destabilizer, latruculin A. Based on these results, the authors suggest in their model (Figure 6) that Rai14 regulates dendritic spine maintenance by stabilizing F-actin in the spine neck. While this is an interesting and feasible possibility, the authors do not directly assess how Rai14 affects F-actin dynamics. They do use RFP-LifeAct in Figure 3G, but only as a neuron fill and not to monitor F-actin dynamics. To better understand how Rai14 might be regulating dendritic spine dynamics, it would be beneficial to assess actin dynamics and/or organization in Rai14-deficient neurons.
-> For the concern regarding the F-actin dynamics, we admit that we did not provide the data on the direct link between Rai14 and F-actin dynamics within the neck of dendritic spines. Therefore, we modified some expressions in the text as follows:
- Result, line 144: “indicating that stabilized Rai14 protects F-actin from destruction in dendritic spines.” -> “indicating that Rai14 protects dendritic spines from the pressure of elimination by actin destabilization.”
- Fig 6 legend, line 1147: “The Rai14 cluster at the spine neck contributes to maintaining spines, probably by stabilizing F-actin, thereby upregulating dendritic spine density.” -> “The Rai14 cluster at the spine neck contributes to maintaining spines, thereby upregulating dendritic spine density.” -> There are multiple reports that Rai14 stabilizes F-actin (Peng et al., 2000; Qian et al., 2013a; Qian et al., 2013b). Tara is also known to stabilize F-actin (Seipel et al., 2001; Woo et al., 2019). Moreover, we showed that dendritic spines overexpressing Rai14 and Tara were more resistant to spine elimination caused by F-actin destabilizer (Figure 3G and 3H). Therefore, we believe it is plausible that Rai14 and Tara be related to F-actin stabilization in the dendritic spine necks. Therefore, it would be of immediate interest to investigate the direct mechanistic link between Rai14 and F-actin dynamics within the dendritic spine neck for spine stabilization with higher resolution imaging approaches.
- The authors observe both learning and memory deficits and depressive-like behaviors in Rai14+/- mice compared to WT mice. Treatment with the antidepressant fluoxetine rescues Rai14+/- mouse behavior in the forced swim test to WT levels (i.e. decreases immobility time). Likewise, fluoxetine treatment rescues dendritic spine density in the prefrontal cortex of Rai14+/- mice to a level seen in WT saline-treated mice. Moreover, chronic restraint stress causes downregulation of Rai14 mRNA and protein expression in the prefrontal cortex, which is blocked by fluoxetine treatment. From these data, the authors conclude that Rai14 is important for the remodeling of synaptic connections relevant to depressive-like behaviors (and to the cognitive deficits possibly related to the depressive-like behavior). However, the link the authors are proposing between Rai14's role in regulating spine dynamics and stress-induced depression may be a bit premature. For instance, the analyses done to determine Rai14's role in regulating dendritic spine density and behavior were done using Rai14+/- mice, where Rai14 was deleted throughout development. Thus, it is not clear whether the behavior and spine defects in Rai14+/- mice are developmental, or whether they would arise from Rai14 loss in adulthood (such as in response to chronic stress). The results with fluoxetine treatment are encouraging, but the authors do not show whether fluoxetine treatment would have similar effects on WT mice (they only treated Rai14+/- mice with fluoxetine in their experiments). Since Rai14 is downregulated in the prefrontal cortex of chronic restraint stressed mice, would stabilized Rai14 (i.e. Rai14 948-967) rescue spine loss and/or depressive-like behavior in stressed mice?
-> In response to the reviewer’s suggestion, we modified the expression from the text as follows:
- Result, line 191-192: “Taken together, these results support the importance of Rai14 in the plastic changes of neuronal connections relevant to depressive-like behaviors” -> “Taken together, these results support the link between the Rai14-controlled dendritic spine dynamics and depressive-like behaviors.”
-> As the reviewer pointed out, we cannot exclude the possibility that the defects in behavior and dendritic spines in Rai14+/- mice are developmental phenotypes. On the other hand, we would like to note that; 1) Chronic stress reduced the Rai14 expression in adult mice (Figure 5J and 5K), 2) Fluoxetine treatment rescued dendritic spine and behavioral defects in adult Rai14+/- mice, 3) Knockdown of Rai14 starting from DIV15 also led to spine density decrease in primary cultured neurons. These results support the notion that Rai14 can participate in the events related to spine dynamics in the adult brain. We believe that this important question will be better addressed when a mouse model with conditional expression or KO of Rai14 gene becomes available.
-
Evaluation Summary:
In this manuscript, the authors discovered a new function of Rai14, an F-actin binding protein, in dendritic spine dynamics. They showed that Rai14 is localized at the spine neck and regulate spine density and function. Heterozygous Rai14 knockout mice showed impaired learning and memory and depressive-like behavior. Overall, this study provides novel insights into the molecular mechanisms underlying spine dynamics and depressive-like behavior. The main conclusions are supported by the 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. Reviewer #2 agreed to share their name with the authors.)
-
Reviewer #1 (Public Review):
The manuscript by Kim et al. identifies a new role for the F-actin binding protein Rai14 in dendritic spine dynamics. The authors demonstrate both in mice and in culture that Rai14-deficient neurons have decreased dendritic spine density, which corresponds with a reduction in excitatory synapse density and the frequency of miniature excitatory postsynaptic currents (mEPSCs). They also provide convincing evidence that Rai14 is protected from degradation through an interaction with another F-actin binding protein, Tara, and that the two proteins accumulate together in dendritic spines necks when overexpressed in neurons, resulting in enhanced spine maintenance. Characterization of Rai14+/- mice revealed that mice display learning and memory deficits and depressive-like behaviors, and that they have reduced …
Reviewer #1 (Public Review):
The manuscript by Kim et al. identifies a new role for the F-actin binding protein Rai14 in dendritic spine dynamics. The authors demonstrate both in mice and in culture that Rai14-deficient neurons have decreased dendritic spine density, which corresponds with a reduction in excitatory synapse density and the frequency of miniature excitatory postsynaptic currents (mEPSCs). They also provide convincing evidence that Rai14 is protected from degradation through an interaction with another F-actin binding protein, Tara, and that the two proteins accumulate together in dendritic spines necks when overexpressed in neurons, resulting in enhanced spine maintenance. Characterization of Rai14+/- mice revealed that mice display learning and memory deficits and depressive-like behaviors, and that they have reduced expression of a number of genes identified in major depressive disorder gene set. Finally, the authors show that chronic restraint stress results in a decrease in mRNA and protein expression of Rai14, and that treatment with the antidepressant fluoxetine can rescue depressive-like behavior and reduced spine density in Rai14+/- mice as well as prevent a reduction in Rai14 expression following chronic restraint stress in wild-type (WT) mice. Together, these results identify Rai14 as a novel regulator of dendritic spine dynamics that may play a role in stress-induced depressive-like phenotypes. While the individual conclusions made by the authors are interesting and generally supported by the data (although in some cases missing important details/analyses), the evidence connecting the various findings together to provide proof that Rai14 is involved in regulating dendritic spine dynamics associated with depressive-like behaviors (as the title suggests) is still somewhat lacking and could be further strengthened.
1. In Figure 1, the authors use Golgi staining of WT and Rai14+/- mouse brain slices as well as primary neuron cultures from WT and Rai14-/- mice and shRNA knockdown of Rai14 to demonstrate that Rai14 loss leads to a reduction in dendritic spine density in cortical and hippocampal neurons. From this data, the authors conclude that Rai14 is required to maintain a normal number of dendritic spines. However, some important details and analyses are missing in these experiments. For instance, in Figure 1A and 1B, the authors do not specify which hippocampal or cortical brain regions (or cell types) they are analyzing in the WT or Rai14+/- mice. In Figure 1C-E, the authors claim there is a reduction in mature dendritic spine density in Rai14-/- neurons compared to WT neurons, but they do not detect differences in spine length or spine head width. It would be useful if the authors could include a description of how they are defining "mature spines". The authors also claim that the reduction in spine density on Rai14-deficient neurons is due to a maintenance phenotype, rather than a formation phenotype, but they do not present evidence to differentiate between these two possibilities. Have the authors examined younger Rai14+/- mice (or Rai14-/- neurons) to determine when the spine phenotype is first detected (i.e. do spines form and then are lost, or do they fail to form correctly in the first place)? The authors attempt to address this question in Figure 3 with experiments in neurons overexpressing Rai14 and Tara, but it might also be useful to look at earlier timepoints in Rai14+/- mice and/or time-lapse imaging of Rai14-deficient neurons.
2. In Figure 3, the authors report the interesting observation that overexpressed Rai14 and Tara accumulate in the necks of dendritic spines, which requires Rai14's ankyrin repeat domains, and that spines containing overexpressed Rai14 are less likely to be eliminated than spines lacking Rai14 clusters, and that neurons overexpressing Rai14 and Tara are resistant to spine loss caused by treatment with the actin destabilizer, latruculin A. Based on these results, the authors suggest in their model (Figure 6) that Rai14 regulates dendritic spine maintenance by stabilizing F-actin in the spine neck. While this is an interesting and feasible possibility, the authors do not directly assess how Rai14 affects F-actin dynamics. They do use RFP-LifeAct in Figure 3G, but only as a neuron fill and not to monitor F-actin dynamics. To better understand how Rai14 might be regulating dendritic spine dynamics, it would be beneficial to assess actin dynamics and/or organization in Rai14-deficient neurons.
3. The authors observe both learning and memory deficits and depressive-like behaviors in Rai14+/- mice compared to WT mice. Treatment with the antidepressant fluoxetine rescues Rai14+/- mouse behavior in the forced swim test to WT levels (i.e. decreases immobility time). Likewise, fluoxetine treatment rescues dendritic spine density in the prefrontal cortex of Rai14+/- mice to a level seen in WT saline-treated mice. Moreover, chronic restraint stress causes downregulation of Rai14 mRNA and protein expression in the prefrontal cortex, which is blocked by fluoxetine treatment. From these data, the authors conclude that Rai14 is important for the remodeling of synaptic connections relevant to depressive-like behaviors (and to the cognitive deficits possibly related to the depressive-like behavior). However, the link the authors are proposing between Rai14's role in regulating spine dynamics and stress-induced depression may be a bit premature. For instance, the analyses done to determine Rai14's role in regulating dendritic spine density and behavior were done using Rai14+/- mice, where Rai14 was deleted throughout development. Thus, it is not clear whether the behavior and spine defects in Rai14+/- mice are developmental, or whether they would arise from Rai14 loss in adulthood (such as in response to chronic stress). The results with fluoxetine treatment are encouraging, but the authors do not show whether fluoxetine treatment would have similar effects on WT mice (they only treated Rai14+/- mice with fluoxetine in their experiments). Since Rai14 is downregulated in the prefrontal cortex of chronic restraint stressed mice, would stabilized Rai14 (i.e. Rai14 948-967) rescue spine loss and/or depressive-like behavior in stressed mice?
-
Reviewer #2 (Public Review):
The manuscript describes the role of Rai14 in the brain in a comprehensive manner. Authors have studied the role of Rai14 in both dissociated primary neurons and in mice. They have also studied the Rai14 interactions with other proteins or effects of altered Rai14 expression to other protein expression by methods of biochemistry. Results carried out with different methods support well each other suggesting that driven conclusions are correct. There are still many open questions with mechanistic details but these can be the topic of the next study.
-
