Growth cone advance requires EB1 as revealed by genomic replacement with a light-sensitive variant
Curation statements for this article:-
Curated by eLife
eLife assessment
In their manuscript, Dema et al. showcase an important tool to study the role of the microtubule end-binding protein, EB1. This important study is the first to locally inactivate EB1 in human neurons, and while the authors have previously published the effects of replacing endogenous EB1 with a light-sensitive variant, the novelty in this current study is that they use a one-step gene editing replacement method in addition to using human neurons derived from iPSCs. The data is of high quality and the evidence supporting the conclusions is solid, although including more controls are needed to strengthen the study. The findings of this work will be of interest to cell biologists and neurobiologists, while the methods utilized will have an even broader general interest.
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Cell Biology (eLife)
Abstract
A challenge in analyzing dynamic intracellular cell biological processes is the dearth of methodologies that are sufficiently fast and specific to perturb intracellular protein activities. We previously developed a light-sensitive variant of the microtubule plus end-tracking protein EB1 by inserting a blue light-controlled protein dimerization module between functional domains. Here, we describe an advanced method to replace endogenous EB1 with this light-sensitive variant in a single genome editing step, thereby enabling this approach in human induced pluripotent stem cells (hiPSCs) and hiPSC-derived neurons. We demonstrate that acute and local optogenetic EB1 inactivation in developing cortical neurons induces microtubule depolymerization in the growth cone periphery and subsequent neurite retraction. In addition, advancing growth cones are repelled from areas of blue light exposure. These phenotypes were independent of the neuronal EB1 homolog EB3, revealing a direct dynamic role of EB1-mediated microtubule plus end interactions in neuron morphogenesis and neurite guidance.
Article activity feed
-
-
Author Response
Reviewer #1 (Public Review):
In this manuscript, the authors describe a one-step genome editing method to replace endogenous EB1 with their previously-developed light-sensitive variant, in order to examine the effect of acute and local optogenetic inactivation of EB1 in human neurons. They then attempt to assess the effects of EB1 inactivation on microtubule growth, F-actin dynamics, and growth cone advance and turning. They also perform these experiments in neurons that are lacking EB3, in order to determine whether EB1 can function in a direct and specific way without possible EB3 redundancy.
First, the experiments depicting the methodology are rigorous and compelling. Most previous studies of +TIP function use knockout or knockdown studies in which the proteins are inactivated over many hours or days in non-human …
Author Response
Reviewer #1 (Public Review):
In this manuscript, the authors describe a one-step genome editing method to replace endogenous EB1 with their previously-developed light-sensitive variant, in order to examine the effect of acute and local optogenetic inactivation of EB1 in human neurons. They then attempt to assess the effects of EB1 inactivation on microtubule growth, F-actin dynamics, and growth cone advance and turning. They also perform these experiments in neurons that are lacking EB3, in order to determine whether EB1 can function in a direct and specific way without possible EB3 redundancy.
First, the experiments depicting the methodology are rigorous and compelling. Most previous studies of +TIP function use knockout or knockdown studies in which the proteins are inactivated over many hours or days in non-human systems. This is the first study to acutely and locally inactivate a +TIP in human neurons. While this group previously published the effects of replacing endogenous EB1 with the light-sensitive variant, the novelty in this current study is that they use a one-step gene editing replacement method (using CRISPR/Cas9) along with using human neurons derived from iPSCs. After proving their new experimental system works, the authors next seek to test the effect that acutely inactivating EB1 (alongside chronic EB3 knockdown) has on microtubule dynamics, and they observe a marked reduction in MT growth and MT length. They then seek to investigate whether F-actin dynamics are immediately affected by EB1 inactivation.
While measured F-actin flow rates are not significantly affected, which leads the authors to conclude that EB1 inactivation does not have any immediate effect, the included figures and movies show a different phenotype, which is not discussed. Finally, they examine the effect of EB1 inactivation on growth cone advance and growth cone turning, and find that both are affected. However, the lack of certain controls in these final experiments (specifically for Figures 3, 4, and 5) reduces the strength of their findings.
Thus, the first part of this paper describing the new methodology is very compelling and should be of interest to a wide readership, while the second part describing the functional analysis is mostly solid, with very high-quality imaging data. However, additional analysis and controls would be needed to increase confidence in their conclusions.
- Analysis of F-actin dynamics is not thorough, and their claim is not completely supported by the data. Figure 3 only depicts F-actin dynamics data from growth cones of π-EB1 EB3-/- i3Neurons and does not [include] control growth cones (to compare dark and light conditions). While their conclusion is that F-actin dynamics are not affected, there do appear to be immediate changes in the F-actin images, other than flow rates. For example, the F-actin bundles do not appear to emanate straight out with the light condition, compared to the dark condition. There also appears to be more F-actin intensity in the transition domain of the growth cone, compared to the dark condition. If the reason is due to the effects of four minutes of blue light exposure, this would be made clear by doing this experiment with control growth cones as well.
In Figure 3, we wanted to specifically test if π-EB1 photoinactivation has an immediate effect on growth cone leading edge actin polymerization (for example because of rapid changes in Rho GTPase activity) by measuring F-actin retrograde flow. Because of photobleaching, these experiments are limited to relatively short time-lapse data sets, and within 4-5 min of blue light exposure, we found no significant difference between the dark and light conditions. As requested by this and another reviewer, we added a few more data points as well as a wild-type control. Statistical analysis by ANOVA shows no difference in retrograde flow between any of the four groups.
We did not see a consistent difference in overall F-actin organization after a few minutes of blue light, and we now include control and π-EB1 growth cones in Fig. 3 that are more similar to one another with the dark image shown more immediately before blue light exposure. The growth cone that we had in the original figure (and that remains in Video 5 to illustrate retrograde flow and how dynamic these growth cones are) was a poor choice for this figure as it undergoes quite dramatic F-actin reorganization before the blue light is turned on, and the morphology immediately before blue light exposure is much more similar to the growth cone during blue light compared with the -5 min time point that we had originally shown.
Lastly, the apparent relocalization of F-actin to the growth cone center is seen in both control and experimental conditions and we believe that has to do with photobleaching of the F-actin probe at the relatively high frame rates required to observe retrograde flow. We agree with the reviewer that it is important to know this, and we included a note in the figure legend.
- Analysis of the effect of EB1 inactivation on growth cone advance and growth cone turning. Figure 4C, showing the neurite unable to cross the blue light barrier, is potentially quite compelling data, but it would be even more convincing if there were also data showing that the blue light barrier has no effect on a control neurite. Given that a number of previous recent studies have shown a detrimental effect of blue light on neurons, it seems important to include these negative controls in this current study.
The experiment growing neurites on a micropatterned laminin surface in combination with photoinactivation in (now) Figure 4D is incredibly low throughput but serves to illustrate repeated retraction from blue light over many hours of imaging. To show that blue light barriers do not affect control cells we have instead included a quantification of the retraction response of control and π-EB1 neurites growing randomly on a laminin-coated surface (not micropatterned stripes) in new Fig. 4C. It is also worth noting that the dose of blue light used for π-EB1 photoinactivation is much lower than what is typically used for fluorescence imaging (we analyzed and discussed this in great detail in our original π-EB1 publication), and especially in experiments with a blue light barrier, cells are not exposed to any blue light before they hit the barrier.
- This concern also holds true for the final experiment, in which the authors examine whether localized blue light would lead to growth cone turning. The authors report difficulty with performing this technically challenging experiment of accurately targeting the light to only a localized region of the growth cone. Thus, the majority of the growth cones (72%) were completely retracted, and so only a small subset of growth cones showed turning. However, this data would be more compelling if there were also a control condition of blue light with neurons that are not expressing the light-inactivated EB1. Another useful control would be to examine whether precise region-of-interest blue light leads to localized loss of EGFP-Zdk1-EB1C on MT plus-ends within the growth cone, or if the loss extends throughout the growth cone. Either outcome would be helpful to potential readers.
We modified Fig. 5 to include control i3Neurons in this experiment. We also included a supplement to Fig. 5 showing that π-EB1 photodissociation remains localized to the blue light-exposed region. However, because in our π-EB1 line the C-terminal π-EB1 half is EGFP-tagged, we cannot show before and after images of local π-EB1 photodissociation.
Reviewer #3 (Public Review):
The major strength of the study was the approach of using photosensitive protein variants to replace endogenous protein with the 1-step Crispr-based gene editing, which not only allowed acute manipulation of protein function but also mimicked the endogenous targeted protein. However, the same strategy has been used by the same first author previously in dividing cells, somewhat reducing the novelty of the current study. In addition, the results obtained from the study were the same as those from previous studies using different approaches. In other words, the current study only confirmed the known findings without any novel or unexpected results. As a result, the study did not provide strong evidence regarding the advantage of the new experimental approach in our understanding of the function of EB1. Some specific comments are listed below.
- In Figure 1, to show that the photosensitive EB1 variant did not affect stem cell properties and their neuronal differentiation, Oct4 staining and western blot of KIF2C and EB3 were not strong evidence. Some new experiments more specifically related to stem cell properties or iPSC-derived neurons are necessary.
While we did not attempt to fully characterize stemness in our π-EB1 edited i3N lines, we believe, most importantly, we show that π-EB1 i3N hiPSCs differentiate normally into i3Neurons. We show this morphologically as well as by immunoblotting and RT-qPCR experiments looking at marker proteins also including DCX, a well-established neuronal differentiation marker. Although not directly related to stemness, we included one additional RT-qPCR experiment more carefully analyzing the expression level of π-EB1 in the edited lines compared with EB1 in control i3N hiPSCs (new Fig. 1E).
In addition, the effect of EB1 inactivation on microtubule growth was quantified in stem cells but not in differentiated neurons, which supposed to be the focus of the study.
Quantification of MT dynamics in the hiPSCs parallels our previous experiments in cancer cell lines to demonstrate that π-EB1 photoinactivation had a similar inhibitory effect on MT growth in interphase cells. This serves as an additional control that our new system works as expected. Because of our inability to efficiently transfect i3Neurons, we could not measure MT growth in i3Neurons with the same method (i.e. automated EB1N tracking). However, as further outlined below we have added a quantification of MT growth rates in i3Neuron growth cones by additional manual tracking of SPY555-tubulin-labelled growth cone MTs after at least one minute of blue light exposure.
In Figure S2D, quantification is needed to show the effect of blue light-induced EB1 inactivation in growth cones.
Fig. 1 – supplement 2D (together with Video 3, and Fig. 2A) is simply to illustrate that the C-terminal π-EB1 half dissociates in blue light as expected. We previously characterized the kinetics of π-EB1 photodissociation and do not think redoing this would add substantially to the current manuscript. The remainder of the manuscript, however, examines the functional consequences of π-EB1 photoinactivation in i3Neurons.
- In Figure 2, the effect of blue light on microtubule retraction in the control cells was examined, showing little effect. However, it is still unclear if the blue light per se would have any effect on microtubule plus end dynamics, a more sensitive behavior than that of retraction. In Figure 2C, the length of individual microtubules in different growth cones was presented, showing microtubule retraction after blue light. Quantification and statistical analysis are necessary to draw a strong conclusion.
Figure 2 shows that growth cone MTs in π-EB1 lines shorten in response to blue light and we did this by analyzing MTs that were visible in a short time window before and after blue light exposure. In response to another reviewer’s comment, we have redesigned this figure to better illustrate this result. We have now included statistical analysis comparing relative MT length 20 s before and during blue light exposure. In control cells that was not statistically significantly different. We also report statistical difference between control and π-EB1 lines at the 20 s by ANOVA in the text. Lastly, we also measured MT growth rates after at least one minute of blue light exposure showing that MT growth is greatly attenuated in π-EB1 lines (new Fig. 2D).
The results showed that EB3 did not seem to contribute to stabilizing microtubules in growth cones. It was discussed that EB3 might have a different function from that of EB1 in the growth cone, although they are markedly up-regulated in neurons. In the differentiated neuronal growth cones examined in the study, does EB3 actually bind to the microtubule plus ends? In the EB3 knockout cells without the blue light, the microtubules were stable, indicating that EB3 had no microtubule stabilization function in these cells. Is such a result consistent with previous studies? If not, some explanation and discussion are needed.
Other papers have shown that EB3 localizes to growth cone MT ends; for example, in rat cortical neurons (Poobalasingam et al., 2022). We did not test if endogenous EB3 is present on MT ends in i3Neurons, but transfected EB3 certainly is. Interestingly, it was reported by multiple groups that EB1 and EB3 do not bind to the exact same place near MT ends. EB3 trails behind EB1, which would be consistent with functional differences especially in controlling MT growth. We have expanded the discussion of such differences in the text, and thank Phillip Gordon-Weeks, who reminded us of this in a comment on the bioRxiv preprint.
- In Figure 3, for the potential roles of EB1 on actin organization and dynamics, only the rates of retrograde flow were measured for 5 min. and no change was observed. However, based on the images presented, it seemed that there was a reduced number of actin bundles after blue light and the actin structure was somewhat disrupted. Some additional examination and measurement of actin organization are necessary to get a clear result.
This point was also raised by reviewer #1, and we now include images and quantification of retrograde flow in control growth cones and we increased the number of data points. We still see no difference in retrograde flow between all these groups. The original π-EB1 growth cone in Fig. 3A was a poor example because it underwent large morphological changes before the blue light was even turned on and just before light exposure is a lot more like the end point image. We therefore replaced this image with a different growth cone that is more similar to the wild-type growth cone shown, and also show images more immediately before blue light exposure. The bottomline is that we do not see a consistent difference in overall F-actin organization after a few minutes of blue light.
- In Figure 4, the effect of blue light and EB1 inactivation on neurite extension need to be quantified in some way, such as the neurite length changes in a fixed time period, and the % of growth cones passing the blue light barrier compared with growth cones of the control cells.
We have included a statistical comparison (by ANOVA) at the 15 min time point, and a quantification of neurite retraction of growth cones encountering a blue light barrier.
- For the quantification of growth cone turning, a control condition is needed to show that blue light itself has no effect on turning.
We have also added a control experiment to Fig. 5.
-
eLife assessment
In their manuscript, Dema et al. showcase an important tool to study the role of the microtubule end-binding protein, EB1. This important study is the first to locally inactivate EB1 in human neurons, and while the authors have previously published the effects of replacing endogenous EB1 with a light-sensitive variant, the novelty in this current study is that they use a one-step gene editing replacement method in addition to using human neurons derived from iPSCs. The data is of high quality and the evidence supporting the conclusions is solid, although including more controls are needed to strengthen the study. The findings of this work will be of interest to cell biologists and neurobiologists, while the methods utilized will have an even broader general interest.
-
Reviewer #1 (Public Review):
In this manuscript, the authors describe a one-step genome editing method to replace endogenous EB1 with their previously-developed light-sensitive variant, in order to examine the effect of acute and local optogenetic inactivation of EB1 in human neurons. They then attempt to assess the effects of EB1 inactivation on microtubule growth, F-actin dynamics, and growth cone advance and turning. They also perform these experiments in neurons that are lacking EB3, in order to determine whether EB1 can function in a direct and specific way without possible EB3 redundancy.
First, the experiments depicting the methodology are rigorous and compelling. Most previous studies of +TIP function use knockout or knockdown studies in which the proteins are inactivated over many hours or days in non-human systems. This is the …
Reviewer #1 (Public Review):
In this manuscript, the authors describe a one-step genome editing method to replace endogenous EB1 with their previously-developed light-sensitive variant, in order to examine the effect of acute and local optogenetic inactivation of EB1 in human neurons. They then attempt to assess the effects of EB1 inactivation on microtubule growth, F-actin dynamics, and growth cone advance and turning. They also perform these experiments in neurons that are lacking EB3, in order to determine whether EB1 can function in a direct and specific way without possible EB3 redundancy.
First, the experiments depicting the methodology are rigorous and compelling. Most previous studies of +TIP function use knockout or knockdown studies in which the proteins are inactivated over many hours or days in non-human systems. This is the first study to acutely and locally inactivate a +TIP in human neurons. While this group previously published the effects of replacing endogenous EB1 with the light-sensitive variant, the novelty in this current study is that they use a one-step gene editing replacement method (using CRISPR/Cas9) along with using human neurons derived from iPSCs. After proving their new experimental system works, the authors next seek to test the effect that acutely inactivating EB1 (alongside chronic EB3 knockdown) has on microtubule dynamics, and they observe a marked reduction in MT growth and MT length. They then seek to investigate whether F-actin dynamics are immediately affected by EB1 inactivation. While measured F-actin flow rates are not significantly affected, which leads the authors to conclude that EB1 inactivation does not have any immediate effect, the included figures and movies show a different phenotype, which is not discussed. Finally, they examine the effect of EB1 inactivation on growth cone advance and growth cone turning, and find that both are affected. However, the lack of certain controls in these final experiments (specifically for Figures 3, 4, and 5) reduces the strength of their findings.
Thus, the first part of this paper describing the new methodology is very compelling and should be of interest to a wide readership, while the second part describing the functional analysis is mostly solid, with very high-quality imaging data. However, additional analysis and controls would be needed to increase confidence in their conclusions.
Analysis of F-actin dynamics is not thorough and their claim is not completely supported by the data. Figure 3 only depicts F-actin dynamics data from growth cones of π-EB1 EB3-/- i3Neurons and does not control growth cones (to compare dark and light conditions). While their conclusion is that F-actin dynamics are not affected, there do appear to be immediate changes in the F-actin images, other than flow rates. For example, the F-actin bundles do not appear to emanate straight out with the light condition, compared to the dark condition. There also appears to be more F-actin intensity in the transition domain of the growth cone, compared to the dark condition. If the reason is due to the effects of four minutes of blue light exposure, this would be made clear by doing this experiment with control growth cones as well.
Analysis of the effect of EB1 inactivation on growth cone advance and growth cone turning. Figure 4C, showing the neurite unable to cross the blue light barrier, is potentially quite compelling data, but it would be even more convincing if there were also data showing that the blue light barrier has no effect on a control neurite. Given that a number of previous recent studies have shown a detrimental effect of blue light on neurons, it seems important to include these negative controls in this current study.
This concern also holds true for the final experiment, in which the authors examine whether localized blue light would lead to growth cone turning. The authors report difficulty with performing this technically challenging experiment of accurately targeting the light to only a localized region of the growth cone. Thus, the majority of the growth cones (72%) were completely retracted, and so only a small subset of growth cones showed turning. However, this data would be more compelling if there were also a control condition of blue light with neurons that are not expressing the light-inactivated EB1. Another useful control would be to examine whether precise region-of-interest blue light leads to localized loss of EGFP-Zdk1-EB1C on MT plus-ends within the growth cone, or if the loss extends throughout the growth cone. Either outcome would be helpful to potential readers.
-
Reviewer #2 (Public Review):
In their manuscript "Growth cone advance requires EB1 as revealed by genomic replacement with a light-sensitive variant", Dema et al. showcase a CRISPR-based strategy to introduce a photo activatable EB1 variant into cultured cells in a single genome engineering step. Upon photoactivation this EB1 variant, which they term π-EB1, dissociates, thus severing the connection between the microtubule tip and +TIP proteins. They demonstrate this technique in human induced pluripotent stem cells, verifying that this genetic engineering procedure neither influences the cells beyond the EB1 gene nor hinders their ability to differentiate into neurons. Subsequently, they nicely verify that dissociation of π-EB1 leads to hindered microtubule growth, which subsequently leads to growth cone retraction. Accordingly, π-EB1 …
Reviewer #2 (Public Review):
In their manuscript "Growth cone advance requires EB1 as revealed by genomic replacement with a light-sensitive variant", Dema et al. showcase a CRISPR-based strategy to introduce a photo activatable EB1 variant into cultured cells in a single genome engineering step. Upon photoactivation this EB1 variant, which they term π-EB1, dissociates, thus severing the connection between the microtubule tip and +TIP proteins. They demonstrate this technique in human induced pluripotent stem cells, verifying that this genetic engineering procedure neither influences the cells beyond the EB1 gene nor hinders their ability to differentiate into neurons. Subsequently, they nicely verify that dissociation of π-EB1 leads to hindered microtubule growth, which subsequently leads to growth cone retraction. Accordingly, π-EB1 expressing axons cannot grow into an area illuminated with blue light, demonstrating the system's usefulness in circuit engineering. Finally, the authors try to specifically redirect growth cones by illuminating defined sections of the growth cone. This however leads mainly to growth cone retraction, in 70% of axons as the authors note, but succeeds in the remaining axons. Sadly, the authors do not further investigate the mechanism at the bottom of the observed axon retraction. Nevertheless, this study adds a valuable tool to the optogenetic toolbox of neurobiologists in the axon growth as well as circuit engineering fields.
Besides a few small writing and figure-editing faux pas, the study is well-written and robustly designed. The conclusions drawn by the authors are well supported by the data, which itself is technically well-prepared and controlled.
-
Reviewer #3 (Public Review):
The major strength of the study was the approach of using photosensitive protein variants to replace endogenous protein with the 1-step Crispr-based gene editing, which not only allowed acute manipulation of protein function but also mimicked the endogenous targeted protein. However, the same strategy has been used by the same first author previously in dividing cells, somewhat reducing the novelty of the current study. In addition, the results obtained from the study were the same as those from previous studies using different approaches. In other words, the current study only confirmed the known findings without any novel or unexpected results. As a result, the study did not provide strong evidence regarding the advantage of the new experimental approach in our understanding of the function of EB1. Some …
Reviewer #3 (Public Review):
The major strength of the study was the approach of using photosensitive protein variants to replace endogenous protein with the 1-step Crispr-based gene editing, which not only allowed acute manipulation of protein function but also mimicked the endogenous targeted protein. However, the same strategy has been used by the same first author previously in dividing cells, somewhat reducing the novelty of the current study. In addition, the results obtained from the study were the same as those from previous studies using different approaches. In other words, the current study only confirmed the known findings without any novel or unexpected results. As a result, the study did not provide strong evidence regarding the advantage of the new experimental approach in our understanding of the function of EB1. Some specific comments are listed below.
1. In Figure 1, to show that the photosensitive EB1 variant did not affect stem cell properties and their neuronal differentiation, Oct4 staining and western blot of KIF2C and EB3 were not strong evidence. Some new experiments more specifically related to stem cell properties or iPSC-derived neurons are necessary. In addition, the effect of EB1 inactivation on microtubule growth was quantified in stem cells but not in differentiated neurons, which supposed to be the focus of the study. In Figure S2D, quantification is needed to show the effect of blue light-induced EB1 inactivation in growth cones.
2. In Figure 2, the effect of blue light on microtubule retraction in the control cells was examined, showing little effect. However, it is still unclear if the blue light per se would have any effect on microtubule plus end dynamics, a more sensitive behavior than that of retraction. In Figure 2C, the length of individual microtubules in different growth cones was presented, showing microtubule retraction after blue light. Quantification and statistical analysis are necessary to draw a strong conclusion.
The results showed that EB3 did not seem to contribute to stabilizing microtubules in growth cones. It was discussed that EB3 might have a different function from that of EB1 in the growth cone, although they are markedly up-regulated in neurons. In the differentiated neuronal growth cones examined in the study, does EB3 actually bind to the microtubule plus ends? In the EB3 knockout cells without the blue light, the microtubules were stable, indicating that EB3 had no microtubule stabilization function in these cells. Is such a result consistent with previous studies? If not, some explanation and discussion are needed.
3. In Figure 3, for the potential roles of EB1 on actin organization and dynamics, only the rates of retrograde flow were measured for 5 min. and no change was observed. However, based on the images presented, it seemed that there was a reduced number of actin bundles after blue light and the actin structure was somewhat disrupted. Some additional examination and measurement of actin organization are necessary to get a clear result.
4. In Figure 4, the effect of blue light and EB1 inactivation on neurite extension need to be quantified in some way, such as the neurite length changes in a fixed time period, and the % of growth cones passing the blue light barrier compared with growth cones of the control cells.
5. For the quantification of growth cone turning, a control condition is needed to show that blue light itself has no effect on turning.
-
