PREreview of Acidic nanoparticles restore lysosomal acidification and rescue metabolic dysfunction in pancreatic β-cells under lipotoxic condition

This review reflects comments and contributions by Marina Schernthanner, Elena Sena, Erika Monelli, Aude Angelini, Shaunak Deota and Pablo Ranea-Robles. Review synthesized by Jonny Coates.

In this study, Hung Lo et al developed a series of biodegradable polyesters, consisting of tetrafluorosuccinic acid (TFSA) and succinic acid (SA) to engineer lysosome acidifying nanoparticles. The authors utilized pre-existing datasets of type 2 diabetes (T2D) in human samples to reveal differentially expressed genes focussed on lysosomal functions, specifically in genes responsible for maintaining lysosomal acidification. The authors used a cellular model of T2D to demonstrate that their nanoparticles were capable of restoring lysosomal acidification and insulin secretion. Furthermore, these nanoparticles restored mitochondrial functions. Finally, the authors administered the nanoparticles to T2D mice on a high fat diet. Intravenous administration of the nanoparticles lowered blood insulin and glucose levels in the mice when on a high-fat diet. Overall this is a strong study although the in vivo aspects could be expanded. 

Major comments:

• The results on glucose tolerance are interesting, but they will be better supported by data of how this treatment affects BW. In the methods the author state they followed BW and food intake with injections. Was BW changed after the treatment? Did the improvement observed in glucose tolerance happen with the same BW as HFD-control animals?

• Since this manuscript describes from a preclinical point of view a new tool with potential for patient treatment, it will be interesting and useful for the community to know the in vivo effect on the measured parameters in control lean mice.

• It would be informative if authors perform autophagy flux analysis using Bafilomycin to indicate how the NPs improve autophagy function.

Minor comments:

• In general, figure legends are incomplete, they should indicate the n for each experiment and the test used for statistical analysis. It would also be great to show data in a column graph with points so the reader can see the n just looking at the graph. 

• Throughout the figures, for the IF images it would be useful to show a color legend so that the figure became self-explicative. For example, in Fig 4C I assume the more orange color illustrates higher levels of acidification/lower pH, but this should be clarified somewhere.

• For figure legends, the inclusion of detailed comments on statistical tests used and corresponding p values for the stars is necessary.

• Figure 5C (palmitate panel) looks like a magnified version of Figure 5A (iv)...however, the authors state images in panel A reflect mCherry reporter signal, while images in panel C illustrate MitoTracker signal? Can the authors please clarify?

• Authors should ensure the cropping of the western blot panels also show the molecular weight as this is something that many journals currently require for publication. I am not sure GAPDH is the best loading control when working on T2D conditions.

• Additional details should be provided regarding the animal studies. 

• Furthermore, it would seem that at times graphs reflect a ratio to the control (value is 1, and different for the other conditions), which should be clarified in the legend and/or axis title. How fold changes were derived should be added to plots or figure captions and the methods. For example, in Fig 4E, it should be noted if the data corresponds to ratios - if they're "true" activity levels, units should be indicated.

• For graphs where statistical significance of NP treatment is reported in comparison to the palmitate-treated cells, it could also be useful to include the statistical significance of the NP-treated cells to the non-palmitate-treated controls to showcase to which degree the treatment returns cells to a near-physiological state

• Figure 2F demonstrates a very important experiment. What about cellular specificity of NPs? Have the authors tested whether different cell types respond differently to the administration of NPs? Presumably, in vivo administration (as done intravenously later in this manuscript) will not only target beta-cells? Within that context, have the authors observed any off-target effects in f.e. other tissues?

• It is better if the authors perform separate pathway analysis for the upregulated and downregulated genes. Taking the complete list of DE genes does not provide information if the pathways are up or down

• The ratio of LC3I/II is a better indicator than LC3II/GAPDH

• My understanding is that mCherry in this construct is fused to the outer mitochondrial layer via FIS1, correct? If that is the case, how do the authors explain the differences in mCherry signal f.e. more 'dotty' signal in panel (i) versus smooth red fluorescence in (iii)?

• For the mCherry-GFP-FIS construct experiments - is the GFP signal supposed to overlap in cellular localization with mCherry? And does pH-dependent quenching disrupt that co-localization, which is why GFP localizes very differently in mitochondria compared to mCherry in panel A (i and ii)

• Could the authors detail in the legend what do the white arrows in Fig 5A(ii) refer to? Moreover, in the text they refer to the white puncta in (ii) but in the figure they do not use arrows to mark the purple puncta, also described in the text.

• Fig 6 C. Although the basal insulin release seems visibly equal in all cases considered, a statistical test should be carried out and noted somewhere, since otherwise it could be more appropriate to compare ratio of stimulated to basal release in the different conditions rather than the absolute values of the insulin release upon stimulation alone.

Comments on reporting:

• It is not correct to utilize an Unpaired t-test when you compare more than 2 groups; so statistics should be revised. Some analyses require a 1-way ANOVA while others will require a 2-way ANOVA. 

• Please provide GSEA statistics for Fig. 3F pathways -  normalized enrichment score (NES), P-value or FDR

• Generally a P-value based cutoff is not ideal, and a cutoff based on adjusted P-value or False discovery rate (FDR) is used for DEG analysis since it takes care of multiple testing.

Suggestions for future studies:

• The authors should also perform Insulin tolerance test (ITT) to check if the short term NP treatment improves whole organism insulin sensitivity.

• Since the main principle of using the NPs is to improve the defective insulin secretion from pancreas, authors should also measure fed state insulin levels after fasting and glucose injection.

• The administration route of NPs here was done intravenously - do the authors intend to try/optimize NPs for other routes, which would probably be more practical and popular for actual treatment options down the road?

Competing interests

The author declares that they have no competing interests.