PREreview of dAux orchestrates the phosphorylation-dependent assembly of the lysosomal V-ATPase in glia and contributes to α-synuclein degradation

Review coordinated via ASAPbio’s crowd preprint review

This review reflects comments and contributions by Femi Arogudade, Teena Bajaj, Joseph Biggane, Vanessa Bijak, María Constanza Silvera, Ryan Cubero, Andreia Faria-Pereira, Luciana Gallo, Arpita Ghosh, Vaishnavi Nagesh, Morufu Olalekan Raimi, Marta Oliva-Santiago, Anna Oliveras, Queen Saikia, Pankaj Vilas Jadhav, and Gliday Yuka. Review synthesized by Joseph Biggane.

This study explores the function of dAuxilin (dAux), the Drosophila equivalent of the human Parkinson’s disease-related gene Cyclin G-associated kinase (GAK), in controlling the lysosomal breakdown of α-synuclein (α-syn) within glial cells. The authors reported that a lack of dAux in glia results in an increased number and size of lysosomes, disrupted lysosomal acidification, and impaired α-syn degradation. The mechanism involves dAux facilitating the phosphorylation-dependent assembly of the V-ATPase V1C subunit, Vha44, at serine 543, which is essential for proper lysosomal acidification and function. These malfunctions lead to the buildup of α-syn and other substances, along with DA neurodegeneration and locomotor deficits, which are characteristic of Parkinson’s disease. The findings imply that targeting this lysosomal pathway in glial cells might be a potential therapeutic strategy to slow down the progression of Parkinson’s disease.

Positive aspects of this study:

• The implementation of multiple model systems is great to see.

• The breadth of techniques utilized is impressive, and given the amount of data generated, the authors could consider dividing this preprint into multiple manuscripts.

• This study presents several novel conclusions that may help elucidate the mechanisms underlying the accumulation of protein inclusions within the nervous system.

Major comments:

• The analyses presented follow some level of statistical methods, but it is unclear why the authors chose a certain method and with what clarity. The statistical analyses statements in the methods section seem boilerplate and could use some clarity. 

• Many plots used to demonstrate the quantification of imaging across this study use different — sometimes vastly so — numbers of samples between groups that are being directly compared.  The authors should take caution with this approach, especially given the utilization of SEM to represent the variance within groups.  Uneven sampling paired with SEM makes raises skepticism about the reliability of the data and associated statistical analyses.  Some of the quantitative plots do have similar sample numbers, but the inclusion of different sample sizes between groups, without explanation, raises the possibility of improper inclusion/exclusion of data.  If this is not a valid concern, then the authors should consider how they can address and explain uneven sampling in high-variance groups within the text.

  • For example, in Figure 1G, "LacZ" condition has n=12 and "dAux-RNAi" condition has n=34.

• Caution is suggested when discussing knockout dAux results. The authors clearly stated that they are working with dAux RNAi strains and that upon RNAi expression, dAux levels were reduced 50% (lines 102-105). So, as authors are not working with a knockout fly strain, but with a reduction in the levels of dAux expression, maybe it would be wise to change ‘Lack of dAux’ to ‘Downregulation of dAux’ throughout the manuscript, to avoid further criticism.

• The conclusion “Taken together, these results suggest that dAux regulates the lysosome number in glia” may be the most likely explanation, but the authors may consider changing the statement to ‘Taken together, these results suggest that the downregulation of dAux negatively regulates the lysosome number in glia’. This statement may be more in line with the actual results and could prevent subsequent reviewers from asking for dAux overexpression experiments to corroborate whether overexpressing dAux also translates into lysosome number changes, which was not an objective of this work. If authors count with dAux overexpression data, it would be good to include it or reference it if this information is available in other publications.

Minor comments:

• General:

  • The introduction could use some clarification for readability.

    • For example, several reviewers were confused by the use of the phrase ‘double-edged sword’ and suggested that other phrases might be more appropriate to convey meaning to the reader.

• The grammar could use revision throughout the preprint.  Several comments suggested that inappropriate grammar significantly hindered readability.

• Several important references are seemingly missing to support some of the claims made within the preprint. 

• Regarding the introduction:

  • Some terms, such as α-synuclein, should be more thoroughly introduced to expand relevancy to a broader audience.  Many of these terms will be familiar within Parkinson’s research, but may be unfamiliar to other cell and molecular biology researchers.

• The authors should exercise caution with the initial statement 'Protein inclusions in the brain are the histological hallmarks of neurodegenerative diseases, and a strategic way to target them for elimination is urgently needed to resolve the pathology.' While protein inclusions are indeed a histological hallmark of neurodegenerative diseases, it is not the only one. The loss of specific types of neurons, for example, also histologically defines neurodegenerative processes. Additionally, since neurodegenerative processes are characterized by their multi-causal onset and conserved mechanisms of maintenance, it is advisable to soften the tone regarding the urgency of targeting only protein inclusions to resolve the pathology. Although the authors are correct that such strategies are highly needed, a more appropriate statement might be: 'Protein inclusions in the brain are one of the histological hallmarks of neurodegenerative diseases, and a strategic way to target them for elimination is urgently needed to gain insights that help resolve the pathology’.

• In reference to the statement “In addition, glia (astrocytes and microglia) are double-edged swords in PD and respond to signals from neuronally-secreted α-syn by releasing pro- or anti-inflammatory factors to enhance or ameliorate PD progression, respectively”, It is unclear if the authors are excluding other types of glia as "double-edged swords" or if they are being excluded altogether as glia. In either case, this seems odd, since the accumulation of α-syn in oligodendrocytes is discussed in the previous sentence.

• In reference to the text “V0 domain with six subunits (a, c, c”, d, e, and c’ in yeast or Ac45 in higher eukaryotes)”, instead of yeast, the authors could just talk about drosophila since yeast is not used as a model in this study.

• Regarding the Results:

  • In subsection ‘Lack of dAux increases the lysosome number in adult fly glia’:

    • The statement “Yet, the numbers of total glial cells, Rab5-positive early endosomes, and Rab7-positive late endosomes remain largely unaffected in the absence of glial dAux” should be modified to clarify that this finding is shown elsewhere (in another preprint) and not in this study. This statement could be moved to the discussion since it doesn't directly pertain to the conclusions of this line of experiment.

• Figures 1F and 1F’ should be discussed in results subsection ‘Lack of dAux disrupts lysosomal acidification in adult fly glia’ rather than ‘Lack of dAux increases the lysosome number in adult fly glia’ because these results are discussed within the context of pH.

• The statement “Yet, the numbers of total glial cells, Rab5-positive early endosomes, and Rab7-positive late endosomes remain largely unaffected in the absence of glial dAux” should be modified to clarify that this finding is shown elsewhere (in another preprint) and not in this study. This statement could be moved to the discussion since it doesn't directly pertain to the conclusions of this line of experiment.

• The authors describe glia dAux knockdown does not affect the total number of glial cells in Drosophila. However, this does not appear to be the case in the mouse model. In Figures S2C, 7A, and 7B, the number of glial cells (IBA1+) seems decreased in GAK cKO mice.

The authors should provide a quantification of the percentage of IBA1+ cells in control and cKO brains. If significant differences are found, the authors should discuss these findings in detail.

• In subsection ‘Lack of glial dAux causes accumulation of Ref(2)P, Ubi, and α-syn’:

  • From the statement “Further in-depth analysis showed that the accumulated α-syn in adult fly brains were enriched in glial lysosomes as both the α-syn intensities inside the Lamp1-positive lysosomes and α-syn-Lamp1 colocalization increased upon dAux-RNAi expression in glia (Figures 2E and 2F). Taken together, these results suggest that dAux mediates the lysosomal degradation of α-syn and other substrates in glia”, it is not clear whether α-syn accumulation occurs specifically in glial lysosomes or also in neuronal lysosomes. To address whether dAux-driven α-syn accumulation is occurring only in glial cells, the authors should perform cell-specific staining (neurons and glia) and provide quantification.

• In subsection ‘dAux regulates lysosomal acidification and hydrolase activity via Vha44 S543 phosphorylation in glia’:

  • Regarding the statement “Taken together, these results suggest that Vha44 acts downstream of dAux and S543 phosphorylation is required for dAux-mediated lysosomal acidification in glia”, Myc-Vha44 expression almost completely restores LTG and MTR puncta compared to the control as seen in 4F and 4H, and dAux-RNAi expression does not seem to have any effect. If Vha44 was to act downstream of dAux, there should have been partial restoration compared to the control. The data does suggest that S543 phosphorylation is required, but this could be independent of dAux, or there could be other proteins including dAux which can be involved in S543 phosphorylation. The authors could repeat the experiment with Myc-Vha44 and Myc-Vha44S543A expression without expressing dAux-RNAi and compare the restoration level.

• In subsection ‘dAux-mediated Vha44 S543 phosphorylation regulates V-ATPase assembly in glia’:

  • The statement “As EGFP-Vha44 expression in glia revealed signals mainly distributed on the glial processes, mCherry-Vha100-2 expression in glia exhibited as puncta, localizing on the lysosomal membrane (Figure 5)” seems to be an assumption based on the presentation of puncta by both mCherry signals (See 5C and 5E).  This seems reasonable, but no direct evidence is presented here.

• Regarding the statement “...live-cell imaging analysis showed that the percentage of EGFP-positive puncta (Vha44) being trafficked to the vicinity of mCherry-positive puncta (Vha100-2) within a radius of 0.42 μm decreased...”, the authors should exercise caution when interpreting v-ATPase assembly through live-imaging colocalization within a 0.42 µm radius. Given that the average diameter of a lysosome is approximately 400 nm, the observed proximity of Vha44 and Vha100-2 within 420 nm could be coincidental rather than indicative of true colocalization.

To more accurately evaluate v-ATPase assembly, the authors should consider employing high-resolution techniques, such as proximity-driven fluorescent probes (e.g., PLA), which can detect protein proximity within ~40 nm. Alternatively, the authors could use Blue Native PAGE followed by probing with antibodies against the V1 and V0 subunits to further investigate the assembly.

• Regarding figures:

  • Figures are nicely described in the text, which makes them easy to follow, however, detailed figure legends are only provided for the supplemental figures.

• Some figures could be broken into two separate figures for clarity and better visibility.

  • For example, Figures 1H-1S are not mentioned until results subsection ’Lack of dAux disrupts lysosomal acidification in adult fly glia’.  The authors may consider creating a new figure panel since the rest of Figure 1 corresponds to results subsection ‘Lack of dAux increases the lysosome number in adult fly glia’.

• It is confusing to see the quantification plots for EM mixed into the middle of the quantification plots for fluorescence microscopy within Figure 1G.

• Supplemental Figure S2 panel should not be supplemental since these are the only data presented to support this results subsection ‘Lack of Gak increases the lysosome number in immortalized microglia and mouse primary microglia’.

• The font size for the heat map color scale in Figure 3C is far to small to be readable.

• Regarding the Discussion:

  • The conclusion that ‘The disruption in lysosomal pH is caused by a phosphorylation-dependent regulation on the V-ATPase assembly, controlling the pumping of protons into the lysosomal lumen to construct the acidic milieu for substrate degradation.’ may be too non-specific.  The authors should specify that the mechanism that they are describing depends on dAux and is not general, which seems implied.

• Regarding the Materials and Methods:

  • Subsection ‘Mouse genetics’:

    • It's not clear when tamoxifen (or whichever was used to induce Cre recombination) was injected. Microglia being involved (if not crucial) during embryonic and postnatal brain development would potentially require a functioning lysosomal degradation system and could potentially confound the paper's results.

    • if the transgenic mice were gifted from a certain lab, the authors can cite the particular paper from that lab in which the methodology of creating these mice models are detailed.

• Subsection ‘Molecular biology’:

  • The authors could cite Livak and Schmittgen's paper that first reported the ΔΔCT method of quantification.

• Subsection ‘Confocal microscopy and statistical analysis’:

  • The statement ‘Shapiro-Wilk normality test was first used to check the data distribution., when the data were normally distributed, two-tailed unpaired t-test, or ordinary one-way ANOVA followed by Tukey’s multiple comparisons test was used., Otherwise, Mann Whitney test or Kruskal-Wallis tests followed by Dunn’s multiple comparisons test were used.’ is confusing.  The authors tested for normality, used parametric tests, yet speak of non-parametric tests in the same sentence. Greater detail clarity on the statistical method used would be helpful.

• Regarding the references:

  • Some article titles are in italics and some are not. The citation format should be assessed for consistency.

Comments on reporting: 

• The analyses presented follow some level of statistical methods, but it is unclear why the authors chose a certain method and with what clarity. The statistical analyses statements in the methods section seem broiler plate and could use some clarity. 

• The description of the process for selecting imaging fields used for quantification is too vague to allow the reader any real confidence about the measures the researchers took to avoid sampling bias.  Specifically, the authors mention picking representative fields for imaging and data collection in confocal imaging experiments, but there is no apparent mention of methods for the mitigation of selection bias.

• dAux RNAi knockdown consistently decreased the EGFP-Vha44 (see 5C and 5E, rows 2 and 4).  This decreased signal may have created the decreased colocalization as an artifact.  Specifically, the colocalization was identified based on both EGFP and mCherry fluorescence intensity.

The authors address this concern at the end of the "dAux-mediated Vha44 S543 phosphorylation regulates V-ATPase assembly in glia" results section by stating "It is noteworthy to mention that Manders’ Correlation was carefully chosen to analyze the colocalization of EGFP-Vha44 or EGFP-Vha44S543A with Vha100-2 or Lamp1 as EGFP levels were affected by dAux".  However, it’s not clear that Manders' correlation actually reduces potential skew introduced from the uneven channel intensities.  While it does ignore pixels with values of zero, it still relies upon pixel intensity.  Thus, just because the low signals are ignored, this still leaves the problem of there being more intense signals in any given area in the brighter channel.

• The authors might consider caution in their interpretation of this data.  The degree of overlap presented in the representative images just doesn't look all that strikingly different in any of the groups.  Row 3 images in 5C and 5C' are the most confusing.  In the Row 3 5C' merge they show an example of less overlapped punta, but if you look closely in 5C there are actually several puncta overlapping seemingly to the same degree as in Row 1, but were not selected to be magnified for Row 3 5C'.  The argument could be made that this is the point of the representative image, but seeing potentially more highly overlapped puncta within the broader viewing field is odd when the quantitative data suggests that there is less colocalization.  It is worth noting that the real-time imaging (Figure 5F-I) seems to demonstrate the conclusions of this set of experiments more definitively. The only potential issue is that the representative images in 5F only shows the merge, so it's not possible to tell the dAux RNAi channel is just too dim to show significant overlap between the channels.

Conflicts of interest:

• None declared

Competing interests

The authors declare that they have no competing interests.