Author response:
The following is the authors’ response to the original reviews.
eLife assessment
This valuable study reports on the packing of molecules in cellular compartments, such as actin-based protrusions. The study provides solid evidence for parameters that enable the building of a biophysical model of filopodia, which is required to gain a complete understanding of these important actin-based structures. Some areas of the manuscript require further clarification.
Public Reviews:
Reviewer #1 (Public Review):
Summary:
The manuscript proposes an alternative method by SDS-PAGE calibration of Halo-Myo10 signals to quantify myosin molecules at specific subcellular locations, in this specific case filopodia, in epifluorescence datasets compared to the more laborious and troublesome single molecule approaches. Based on these preliminary estimates, the authors developed further their analysis and discussed different scenarios regarding myosin 10 working models to explain intracellular diffusion and targeting to filopodia.
Strengths:
Overall, the paper is elegantly written and the data analysis is appropriately presented.
Weaknesses:
While the methodology is intriguing in its descriptive potential and could be the beginning of an interesting story, a good portion of the paper is dedicated to the discussion of hypothetical working mechanisms to explain myosin diffusion, localization, and decoration of filopodial actin that is not accompanied by the mandatory gain/loss of function studies required to sustain these claims.
To be fair, the detailed mechanisms that we raise related to diffusion, localization, and decoration are based on extensive work by others. Many prior papers use domain deletions of Myo10 and fall in the category of gain/loss-of-function studies. It is true that we have not repeated those extensive studies, but it seems appropriate to connect with and cite their work where appropriate.
Reviewer #2 (Public Review):
Summary:
The paper sought to determine the number of myosin 10 molecules per cell and localized to filopodia, where they are known to be involved in formation, transport within, and dynamics of these important actin-based protrusions. The authors used a novel method to determine the number of molecules per cell. First, they expressed HALO tagged Myo10 in U20S cells and generated cell lysates of a certain number of cells and detected Myo10 after SDS-PAGE, with fluorescence and a stained free method. They used a purified HALO tagged standard protein to generate a standard curve which allowed for determining Myo10 concentration in cell lysates and thus an estimate of the number of Myo10 molecules per cell. They also examined the fluorescence intensity in fixed cell images to determine the average fluorescence intensity per Myo10 molecule, which allowed the number of Myo10 molecules per region of the cell to be determined. They found a relatively small fraction of Myo10 (6%) localizes to filopodia. There are hundreds of Myo10 in each filopodia, which suggests some filopodia have more Myo10 than actin binding sites. Thus, there may be crowding of Myo10 at the tips, which could impact transport, the morphology at the tips, and dynamics of the protrusions themselves. Overall, the study forms the basis for a novel technique to estimate the number of molecules per cell and their localization to actin-based structures. The implications are broad also for being able to understand the role of myosins in actin protrusions, which is important for cancer metastasis and wound healing.
Strengths:
The paper addresses an important fundamental biological question about how many molecular motors are localized to a specific cellular compartment and how that may relate to other aspects of the compartment such as the actin cytoskeleton and the membrane. The paper demonstrates a method of estimating the number of myosin molecules per cell using the fluorescently labeled HALO tag and SDS-PAGE analysis. There are several important conclusions from this work in that it estimates the number of Myo10 molecules localized to different regions of the filopodia and the minimum number required for filopodia formation. The authors also establish a correlation between number of Myo10 molecules filopodia localized and the number of filopodia in the cell. There is only a small % of Myo10 that tip localized relative to the total amount in the cell, suggesting Myo10 have to be activated to enter the filopodia compartment. The localization of Myo10 is log-normal, which suggest a clustering of Myo10 is a feature of this motor.
Weaknesses:
One main critique of this work is that the Myo10 was overexpressed. Thus, the amount in the cell body compared to the filopodia is difficult to compare to physiological conditions. The amount in the filopodia was relatively small - 100s of molecules per filopodia so this result is still interesting regardless of the overexpression. However, the overexpression should be addressed in the limitations.
This is a reasonable perspective and we now note this caveat in the Limitations section so that readers will take note. Our goal here was to understand a system in which Myo10 is the limiting reagent for filopodia, rather than a native system that expresses high Myo10 on its own. Because U2OS cells do not express detectable levels of Myo10 (see below), the natural perturbation here is overexpressing Myo10 to stimulate filopodial growth.
The authors have not addressed the potential for variability in transfection efficiency. The authors could examine the average fluorescence intensity per cell and if similar this may address this concern.
Indeed, cells are heterogenous and will naturally express different levels of Myo10 not only due to transfection efficiency, but also due to their state (cell cycle stage, motile behavior, and more). In fact, we measure the transfection efficiency of each bioreplicate and account for it in our calibration procedure. We also measure the fluorescence intensity per cell, which lets us calculate the total Myo10s per cell and the cell-to-cell variability. These Myo10 distributions across cells are shown in Fig. 1D-E.
We note here an error that we made in applying this transfection efficiency correction in the first submission. When we obtain the total Myo10 molecules by SDS-PAGE, we should divide by the total number of transfected cells. However, due to an operator precedence error, the transfection efficiency appeared in the numerator rather than the denominator. We have now corrected this error, which has the effect of increasing the number of molecules in all of our measurements. The effect of this correction has strengthened one of the paper’s main conclusions, that Myo10 is frequently overloaded at filopodial tips.
The SDS PAGE method of estimating the number of molecules is quite interesting. I really like this idea. However, I feel there are a few more things to consider. The fraction of HALO tag standard and Myo10 labeled with the HALO tagged ligand is not determined directly. It is suggested that since excess HALO tagged ligand was added we can assume nearly 100% labeling. If the HALO tag standard protein is purified it should be feasible to determine the fraction of HALO tagged standard that is labeled by examining the absorbance of the protein at 280 and fluorophore at its appropriate wavelength.
This is a fair point raised by the reviewer, and we have now measured a labeling efficiency of 90% in Supplementary Figure 2A-C. We have adjusted all values according to this labeling efficiency.
The fraction of HALO tagged Myo10 labeled may be more challenging to determine, since it is in a cell lysate, but there may be some potential approaches (e.g. mass spec, HPLC).
As noted, this value is considerably more challenging. Instead, we determined conditions under which labeling in cells is saturated. We have now stained with a concentration range for both fixed and live cell samples. Saturation occurs with ~0.5 μM HaloTag ligand-TMR in fixed/permeabilized cells and in live cells (Supplementary Figure 2D-E). This comparison of live cells vs. permeabilized cells allows us to say that the intact plasma membrane is not limiting labeling under these conditions.
In Figure 1B, the stain free gel bands look relatively clean. The Myo10 is from cell lysates so it is surprising that there are not more bands. I am not surprised that the bands in the TMR fluorescence gel are clean, and I agree the fluorescence is the best way to quantitate.
Figure 1B shows the focused view at high MW, and there is not much above Myo10. The full gel lanes shown in Supp. Fig. 1C show the expected number of bands from a cell lysate.
In Figure 3C, the number of Myo10 molecules needed to initiate a filopodium was estimated. I wonder if the authors could have looked at live cell movies to determine that these events started with a puncta of Myo10 at the edge of the cell, and then went on to form a filopodia that elongated from the cell. How was the number of Myo10 molecules that were involved in the initiation determined? Please clarify the assumptions in making this conclusion.
We thank the reviewer (and the other reviewers) for this excellent suggestion. We have now carried out these live cell experiments. These experiments were quite challenging, because we needed to collect snapshots of ~50 cells to measure the mean fluorescence intensity of transfected cells and then acquire movies of several cells for analysis. The U2OS cells were also highly temperature-sensitive and would retract their filopodia without objective heating.
We have now analyzed filopodial initiation events and measured considerably more Myo10 at the first signs of accumulation– in the 100s of molecules. The dimmer spots that we measured in the first draft were likely unrelated to filopodial initiation, and we have corrected the discussion on this point.
We now also track further growth from a stable filopodial tip (the phased-elongation mechanism from Ikebe and coworkers) and find approximately 500 molecules bud off in those events. We also track filopodial elongation rates as a function of Myo10 numbers. We have added additional live cell imaging sections that include these results.
It is stated in the discussion that the amount of Myo10 in the filopodia exceeds the number of actin binding sites. However, since Myo10 contains membrane binding motifs and has been shown to interact with the membrane it should be pointed that the excess Myo10 at the tips may be interacting with the membrane and not actin, which may prevent traffic jams.
This is also an excellent point to consider, and we have expanded the relevant discussion along these lines. We agree that the Myo10 at the filopodial tip is likely membrane-bound. We now estimate the 2D membrane area occupied by Myo10, and find that it reaches nearly full packing in many cases (under a number of assumptions that we spell out more fully in the manuscript).
Reviewer #3 (Public Review):
Summary:
The unconventional myosin Myo10 (aka myosin X) is essential for filopodia formation in a number of mammalian cells. There is a good deal of interest in its role in filopodia formation and function. The manuscript describes a careful, quantitative analysis of Myo10 molecules in U2OS cells, a widely used model for studying filopodia, how many are present in the cytosol versus filopodia and the distribution of filopodia and molecules along the cell edge. Rigorous quantification of Myo10 protein amounts in a cell and cellular compartment are critical for ultimately deciphering the cellular mechanism of Myo10 action as well as understand the molecular composition of a Myo10-generated filopodium.
Consistent with what is seen in images of Myo10 localization in many papers, the vast majority of Myo10 is in the cell body with only a small percentage (appr 5%) present in filopodia puncta. Interestingly, Myo10 is not uniformly distributed along the cell edge, but rather it is unevenly localized along the cell edge with one region preferentially extending filopodia, presumably via localized activation of Myo10 motors. Calculation of total molecules present in puncta based on measurement of puncta size and measured Halo-Myo10 signal intensity shows that the concentration of motor present can vary from 3 - 225 uM. Based on an estimation of available actin binding sites, it is possible that Myo10 can be present in excess over these binding sites.
Strengths:
The work represents an important first step towards defining the molecular stoichiometry of filopodial tip proteins. The observed range of Myo10 molecules at the tip suggests that it can accommodate a fairly wide range of Myo10 motors. There is great value in studies such as this and the approach taken by the authors gives one good confidence that the numbers obtained are in the right range.
Weaknesses:
One caveat (see below) is that these numbers are obtained for overexpressing cells and the relevance to native levels of Myo10 in a cell is unclear.
A similar concern was raised by Reviewer 2; please see above.
An interesting aspect of the work is quantification of the fraction of Myo10 molecules in the cytosol versus in filopodia tips showing that the vast majority of motors are inactive in the cytosol, as is seen in images of cells. This has implications for thinking about how cells maintain this large population in the off-state and what is the mechanism of motor activation. One question raised by this work is the distinction between cytosolic Myo10 and the population found at the ‘cell edge’ and the filopodia tip. The cortical population of Myo10 is partially activated, so to speak, as it is targeted to the cortex/membrane and presumably ready to go. Providing quantification of this population of motors, that one might think of as being in a waiting room, could provide additional insight into a potential step-by-step pathway where recruitment or binding to the cortical region/plasma membrane is not by itself sufficient for activation.
As mentioned in our response to Reviewer 2, we have now carried out quantitation in live cells to capture Myo10 transitions from cell body into filopodial movement. We attempted to identify this membrane-bound population of motors in our new live cell experiments but were unable to make convincing measurements. Notably, we see no noticeable enrichment of Myo10 at the cortex relative to the cytosol. Although we believe there is a membrane-bound waiting room (akin to the 3D-2D-1D mechanism of Molloy and Peckham), we suspect that the 2D population is diffusing too rapidly to be detected under our imaging conditions.
Specific comments:
(1) It is not obvious whether the analysis of numbers of Myo10 molecules in a cell that is ectopically overexpressing Myo10 is relevant for wild type cells. It would appear to be a significant excess based on the total protein stained blot shown in Fig S1E where a prominent band the size of tagged Myo10 seen in the transfected sample is almost absent in the WT control lane.
Even “wildtype” cells vary considerably in their Myo10 expression levels. For example, melanoma cells often heavily upregulate Myo10, while these U2OS cells produce nearly none (Supplementary Figure 1E). Thus, there is no single, widely acceptable target for Myo10 expression in wildtype cells.
Please note that the new Supplementary Figure 1E is a Myo10 Western blot, not total protein staining as before.
Ideally, and ultimately an important approach, would be to work with a cell line expressing endogenously tagged Myo10 via genome engineering. This can be complicated in transformed cells that often have chromosomal duplications.
Indeed, we chose U2OS cells for this work because they do not express detectable levels of Myo10, and thus we can avoid all of these complications. Here we can examine how Myo10 levels control filopodial production through ectopic expression.
However, even though there is an excess of Myo10 it would appear that activation is still under some type of control as the cytosolic pool is quite large and its localization to the cell edge is not uniform. But it is difficult to gauge whether the number of molecules in the filopodium is the same as would be seen in untransfected cells. Myo10 can readily walk up a filopodium and if excess numbers of this motor are activated they would accumulate in the tip in large numbers, possibly creating a bulge as and indeed it does appear that some tips are unusually large. Then how would that relate to the normal condition?
As noted above, the normal condition depends on the cellular system. However, endogenous Myo10 also accumulates in bulges at filopodial tips, so this is not a phenotype unique to Myo10 overexpression. For example, the images from Figure 1 of the Berg and Cheney (2002) citation show bulges from endogenous Myo10 in endothelial cells.
(2) Measurements of the localization of Myo10 focuses in large part on ‘Myo10 punctae’. While it seems reasonable to presume that these are filopodia tips, the authors should provide readers with a clear definition of a puncta. Is it only filopodia tips, which seems to be the case? Does it include initiation sites at the cell membrane that often appear as punctae?
We define puncta as any clusters/spots of Myo10 signal detected by segmentation, not limited to any location within the surface-attached filopodia. We exclude puncta that appear in the cell interior (~5 of which appear in Fig. 1A). These are likely dorsal filopodia, but there are few of these compared to the surface attached filopodia of U2OS cells. In Figure 2, “puncta” includes all Myo10 clusters along the filopodia shaft, though a majority happen to be tip-localized (please see Supplementary Figure 4B). We have edited the main text for clarification.
Along those lines, the position of dim punctae along the length of a filopodium is measured (Fig 3D). The findings suggest that a given filopodium can have more than one puncta which seems at odds if a puncta is a filopodia tip. How frequently is a filopodium with two puncta seen? It would be helpful if the authors provided an example image showing the dim puncta that are not present at the tip.
We have now provided an example image of dim puncta along filopodia in Supplementary Figure 4C.
(3) The concentration of actin available to Myo10 is calculated based on the deduction from Nagy et al (2010) that only 4/13 of the actin monomers in a helical turn are accessible to the Myo10 motor (discussion on pg 9; Fig S4). Subsequent work (Ropars et al, 2016) has shown that the heads of the antiparallel Myo10 dimer are flattened, but the neck is rather flexible, meaning that the motor can a variable reach (36 - 52 nm). Wouldn’t this mean that more actin could be accessible to the Myo10 motor than is calculated here?
Although we see why the reviewer might believe otherwise, the 4/13 fraction of accessible actin holds. This fraction is obtained from consideration of the fascin-actin bundle structure alone, independent of the reach of any particular myosin motor. Every repeating layer of 13 actin subunits (or 36 nm) has 4 accessible myosin binding-sites. The remaining 9 sites are rejected because a single myosin motor domain will have a steric clash with a neighboring actin filament in the bundle. A myosin with an exceptionally long reach might reach the next 13 subunit layer, but that layer also has only 4 binding sites. Thus, we can calculate the number of binding sites per unit length along the filopodium. This number would hold for a dimeric myosin with any reach, including myosin-5 or myosin-2.
(4) Quantification of numbers of Myo10 molecules in filopodial puncta (Fig 3C) leads the authors to conclude that ‘only ten or fewer Myo10 molecules are necessary for filopodia initiation’ (pg 7, top). While this is a reasonable based on the assumption that the formation of a puncta ultimately results from an initiation event, little is known about initiation events and without direct observation of coalescence of Myo10 at the cell edge that leads to formation of a filopodium, this seems rather speculative.
As noted above, we have now performed the necessary live cell imaging of filopodial nucleation events and have updated our conclusions accordingly.
Recommendations for the authors:
Reviewer #1 (Recommendations For The Authors):
I have made a series of comments that might help the authors improve their manuscript:
- A full calibration of the methodology would require testing a wider range of protein amounts, to exhaustively detect the dynamic range of the technique. The authors acknowledge in the discussion that “Furthermore, our estimates of molecules are predicated on the calibration curve of the Halo Standard Protein on the SDS-PAGE gels, which is likely the highest source of error on our molecule counts”. A good way of convincing a nasty reviewer is to provide a calibration with more than 3 reference points. At least this will help exclude from the analysis cells where Myo10 estimates are not in the linear regime of detection.
We completely agree with the reviewer’s suggestion to build a robust calibration curve. The SDS gel shown in Figure 1C originally contained 4 reference points, but the highest HaloTag standard protein point oversaturated the detector at the set exposure in the TMR channel and was omitted. We have now re-run the SDS gel to include a HaloTag standard protein curve comprising 5 points, alongside all three bioreplicates from the fixed cell experiments and all three bioreplicates from the live cell experiments (updated in Figure 1B-C). We had saved frozen lysates from the original fixed cell work, so we were able to reanalyze our data with the new set of standards. The Myo10 quantities are consistent, but with much tighter CIs from the standard curve.
- As already said this methodology is intriguing, however, a correlative validation with a conventional SMLM approach to address the bona-fide of the method would be ideal.
Unfortunately, single molecule approaches for validation are impractical for us. Due to the relatively high magnification of our TIRF microscope and the large spread area of the U2OS cells, single cells typically extend beyond the field of view. We acknowledge the benefits of SMLM quantitative techniques and other approaches cited in the introduction section. To avoid use of special tools/instruments, we offer our methodology, based off Pollard group’s quantitative Western blotting of GFP, as a simpler alternative accessible to anyone.
- TMR is a small ligand likely interacting also with Halo in its denatured state. However, to clear any doubts a parallel Native-PAGE investigation should be included, or if existing a specific reference should be provided.
Perhaps there is a misunderstanding here. One of the key advantages of the HaloTag labeling system is that the engineered dehalogenase is covalently modified by the ligand (the TMR-ligand is a suicide substrate). This means that the TMR remains bound even under denaturing conditions, which allows its detection in SDS-PAGE. Native gels are unnecessary here.
- Moreover, SDS-PAGE is run at alkaline pH, have the authors considered these points when designing the methodology? Fluorescence images were taken in PBS, which has a different pH. Could the authors, or the literature, exclude these aspects as potential pitfalls in the methodology? Also temperature is affecting fluorescence emission, but it is easier to control with certain tolerance in the room-temperature regime.
Our method does not compare fluorescence values that cross the experimental systems (SDS-PAGE vs. microscopy). Cellular proteins and HaloTag protein standards are compared in a single setting of SDS-PAGE to obtain the average number of Myo10s per transfected cell. Likewise, all measurements on intact (live or fixed) cells are conducted in that single setting to obtain average fluorescence per cell. Thus, there is no issue with the different buffers or temperatures affecting fluorescence emission.
- The authors should test their approach also with truncation variants of Myosin10 (for instance lacking the PH or motor domain). This is a classical approach that might prove the potential of the technique when altering the capacity of the protein to interact with a main binding partner. Also, treatments that induced filopodia formation might prove useful (i.e., hypotonic media induce filopodia formation in some fibroblast cell lines in our hands).
The reviewer raises interesting suggestions that we aim to address in future experiments, but truncation variants and environmental perturbations are beyond the focus of the current manuscript. Here, we report on the otherwise unperturbed state when we add exogenous full-length Myo10 to the U2OS cells. But indeed, experiments with Myo10 domain truncations, PI3K and PTEN inhibition, and cargo protein / activating cofactor knock-downs (among others) are on our drawing board.
- Most of the mechanisms hypothesized in the discussion are sound and plausible. However, the authors have chosen an experimental model where transient transfection of exogenous Myo10 in U2OS is performed. This approach poses two main and fundamental questions that are not resolved by the data provided:
A) how do different expression levels affect the Myo10 counting?
Our counting procedure does not assume uniform expression across a population of cells– quite the opposite, in fact. We directly measure Myo10 expression levels on a cell-by-cell basis with microscopy, once we know the number of molecules in our total pool (see the Methods for details). As an example of the final output, Figs. 1D and 1E show the total number of Myo10 molecules per cell for fixed and live cells, respectively.
B) how does endogenous and unlabeled Myo10 hamper the bonafide of counts? The authors claimed “U2OS cells express low levels of Myo10, so there is a small population of unlabeled endogenous Myo10 unaddressed by this paper”. As presented, the low levels of endogenous Myo10 sound an arbitrary parameter, and there are no data presented that can limit if not exclude this bias in the analysis. To produce data in a genetically modified cell line with Halo-tag on the endogenous protein will represent a much cleaner system. Alternatively, the authors should look for Myo10 KO cell lines where they can back-transfect their Halo-Tagged Myo10 construct in a more consistent framework, focusing on cells with low-to-mid levels of expression.
We agree, this is an important point to nail down (and is often neglected in the literature). We have now measured the endogenous Myo10 levels in U2OS cells by Western blotting and found that it is undetectable compared to our HaloTagged construct expression. Please see Supp. Fig 1E. Thus, for all intents and purposes, every Myo10 molecule in these experiments came from our expression plasmid. Accordingly, we have removed this caveat from the paper.
Minor points
- Figure 1B. To help the reader SDS-PAGE gels annotations should be clearer already from the figure.
We have updated the annotations for clarity.
- Methods should be organized in sessions. As it stands, it is hard for the reader to look for technical details.
We have expanded and added subsections to the Methods as requested.
- The good practice of indicating the gene and transcript entry numbers and the primer used to amplify and clone into the backbone vectors is getting lost in many papers. I would strongly encourage the authors to add this information to the methods.
We have included the gene entries to the methods and will include a full FASTA file of the coding sequence as supplementary information to avoid any ambiguity here.
The authors write “It is unclear how myosins navigate to the right place at the right time, but our results support an important interplay between Myo10 and the actin network.” It is a bit scholastic to say that Myo10 and actin have an important interplay, they are major binding partners. What is the new knowledge contained in this sentence?
Agreed– we have deleted the sentence in question.
Reviewer #2 (Recommendations For The Authors):
The authors should address all the weaknesses indicated in the public review.
There were a few other places that require clarification.
On page 4, the last paragraph. It is stated that the targeting of Myo10 was reported/proposed based on previous work (ref 31). The next few sentences are not referenced and thus likely refer to ref 31. The authors did not measure the parameters discussed in these sentences, so it is important to clarify that they are referring to previous work and not the current study.
Indeed, the next few sentences still refer to old reference 31, so we have now edited the paragraph for clarity.
On page 7, the reference to Figure 3A indicates that the trend of higher Myo10 correlating with more filopodia. However, the reference to Figure 3B indicates total intracellular Myo10 weakly correlates with more filopodia. However, the x-axis on Figure 3B is filopodia molecules not the intracellular Myo10. Please clarify.
We appreciate the reviewer for catching our mistake. Those plots are now in Fig. 2 and have been edited accordingly.
Reviewer #3 (Recommendations For The Authors):
The Discussion of results at the end of each section is rather brief and could be expanded on a bit more.
Before we were operating under the constraints of an eLife Short Report. We have now expanded the discussion for a full article.
The authors mention that actin filaments at the tips of filopodia could be frayed, citing Medalia et al, 2007 (ref 40). That paper describes an early cryoEM analysis of filopodia from the amoeba Dictyostelium. EM images of mammalian filopodia tips, e.g. Svitkina et al, 2003, JCB, do not show quite the same organization of actin as seen in the Dictyostelium filopodia tips. However, recent work from the Bershadsky lab, Li et al, 2023, presents a few cryoEM images of tips of left-bent filopodia that are tightly adhered to a substrate and there it looks like actin filaments become disorganized in tips, along with membrane bulging. The authors should consider expanding discussion of the filopodia tips to take into account what is known for mammalian filopodia.
We thank the reviewer for bringing these enlightening papers to our attention. We have now included these citations in the discussion.
Fig 1D - The x-axis is a bit odd, it goes from 0 then to 2.5e+06 with no indication of the bin size. Can this be re-labelled or the scale displayed a bit differently?
We have double-checked the axis breaks, which are large because the underlying values are large. We have also provided the bin size as requested for all histograms.
Fig 4A - What is the bin size for the histogram?
As above, we have now updated the figure legends (now in Fig. 3) to include the bin size.
Methods -
- Please provide an accession number for the Myo10 nucleotide sequence used for this work as there are at least two known isoforms.
Thank you for noting this. We are using the full-length, not the headless isoform. We have now updated the Methods accordingly.
- No mention is made of the SDS sample buffer used, was that also added to the sample?
We have now updated the Methods accordingly.
- How are samples boiled at 70 deg C? Do the authors actually mean ‘heated’?
Indeed. We have now corrected “boiled” to “heated.”
- Could the authors please briefly explain the connected component analysis used to identify filopodia?
We have now updated the Methods accordingly.
- The intensity of filopodia was determined by dividing tip intensity by the total bioreplicate sum of intensities then multiplying it by the total pool, if this reviewer understands correctly. It sounds like intensities are being averaged across a whole cell population instead of cell-by-cell. Is that correct? If so, can the authors please provide the underlying rationale for this? If not, then please better describe what was actually done.
We apologize for the confusion. Intensities are being averaged (summed) across a whole cell population, but importantly that step is only used to obtain a scale factor that converts the fluorescence signal at the microscope to the number of molecules. We then use that scale factor for all cells imaged in the bioreplicate, to both 1) find the total Myo10 in that cell, and 2) find the total amount of that Myo10 in any given location within that cell.
To further clarify, each bioreplicate has a known total number of Myo10 molecules associated with the number of cells loaded onto the SDS gel. From the SDS gel, we have an average number of Myo10 molecules per positively transfected cell. If 50 cell images are analyzed, then there is a Myo10 ‘total pool’ of (50 cells) * (average Myo10 molecules/cell). The fluorescence signal intensities in microscopy were summed for all cells within the bioreplicate (50 cells in this example). However, due to variation in expression, not every cell has the same signal intensity when imaged under the same conditions. It would be inaccurate to assume each cell contains the average Myo10 molecules/cell. Therefore, to get the number of molecules within a given Myo10 cell (or punctum), the summed cell (punctum) intensity was divided by the bioreplicate fluorescence signal intensity sum and multiplied by ‘total pool.’
- The authors quantify Myo10 protein amounts by western blotting using Halo tag fluorescence, a method that should provide good accuracy. The results depend on the transfection efficiency and it is rarely the case that it is 100%. The authors state that they use a ‘value correction for positively transfected cells’ (pg 11). It is likely that there was a range of expression levels in the cells, how was a cut-off for classifying a cell as non-expressing determined or set?
As described in the Methods, “microscopy was used to count the percentage of transfected cells from ~105-190 randomly surveyed cells per bioreplicate.” Cells were labeled and located with DAPI. If no TMR signal could be visually detected by microscopy, then the cell was deemed to be non-Myo10 expressing. We did not set a cutoff fluorescence value, as untransfected cells have no detectable signal. Please see Supplementary Figure 1F for examples.
- “In-house Python scripts” are used for image analysis. Will these be made publicly available?
Yes, we will package these up on GitHub.
