In Vivo Remodeling of Altered Autophagy-Lysosomal Pathway by a Phosphopeptide in Lupus

Round 1

Reviewer 1 Report

The study by Wang et al. reports the alterations in chaperone-mediated autophagy in lupus-prone mice treated with the phosphopeptide P140. There are several important issues requiring authors' attention:

1. Fig. 1: The authors’ statement that P140 exerted a “very efficient protective effect” on lysosomes in MRL/lpr mice is not supported by the results presented in Fig. 1, showing no significant effects on lysosomal volume in T cells, lysosomal pH in total splenocytes, or lysosomal degradation capacity in B and T cells. While there is a tendency towards beneficial changes in lysosomal volume/function in P140-treated mice, the differences did not reach statistical significance, probably because of the small size of experimental groups. This issue should be clarified by repeating the experiments with larger experimental groups. Also, why the lysosomal pH was measured in total splenocytes and not in B and T cells? Finally, the authors completely ignore the fact that both P140 and ScP140 significantly decrease TFEB expression in MRL/lpr mice.

2. Fig. 2. Again, there is a clear trend towards decrease in GAPDH, HSPA8, and LAMP1 in P140-treated MRL/lpr mice, but the differences did not reach statistical significance because of the insufficient size of experimental groups. Also, why the ScP140 control is missing?

3. Fig. 3: Again, ScP140-treated control group is missing, which makes interpretation of the results rather difficult.

4. The authors should be more consistent in displaying the p values - the p values > 0.05 are shown in some graphs, while in others they are omitted.

5. Finally, the authors provide no experimental not theoretical explanation as to the mechanisms presumably involved in CMA-dependent immunological deffects in MRL/lpr mice. Having in mind that the role of macroautophagy in B and T cell survival and function is better defined, one would expect that macroauotphagy modulation by P140 is more important in this context than that of CMA. This issue should be discussed.

Author Response

Comment 1

-Our previous data showed that autophagy alterations are more severe in B cells compared to T cells. In good agreement with these earlier findings, in figure 1, there is effectively no effect of P140 on the lysosomal volume in T cells, but there is an effect in B cells.

-The measurement of the lysosomal pH was done using splenocytes (mainly B cells) as in ref. 3. With regard to ref. 3 we introduced the missing control peptide ScP140. It was not possible to perform this experiment in T cells due to the number of T cells that can be collected from unique mice.

- Regarding the fact that both P140 and ScP140 significantly decrease TFEB expression in MRL/lpr mice, we found it a very interesting effect worth reporting and that we intend to investigate more in the future. However, because it does not directly relate with the effect on CMA, main focus of this work, we have not investigated in this study. Our working hypothesis is that normalization of the upregulated levels of TFEB by P140 can exert a correcting impact on lysosomal function (LysoTracker intensity; figure 1B) and in the observed abnormal expansion of the endolysosomal system. However, we anticipate that changes on TFEB may not directly affect CMA activity (figure 4E of ref. 22), as limiting components for CMA such as LAMP2A are not under TFEB regulation. The fact that ScP140 loses its beneficial effect on CMA even though it decreases TEFB level, further supports that the observed TFEB changes, albeit interesting, are not the determinants of the P140-dependent changes in CMA described in this work.  ScP140 is an “inactive” control of P140 in all the other experiments that we have performed so far (refs 9, 22, 23, 24, 27, 28). Further studies are required to investigate the mechanism behind the observed changes in TFEB.

Comment 2

-Effectively, the only statistically significant result was obtained with LAMP2A. Since this receptor has been shown the limiting component for CMA, we consider that the significant changes in LAMP2A weight more in the overall functional outcome on CMA of our intervention. Independently of changes in HSPA8, just the reduction in levels of LAMP2A alone observed with P140 is sufficient to restore CMA activity to control levels.   

-The ScP140 control peptide was not included in these experiments to limit the number of MRL/lpr that are used, according to the 3 Rs principle (Replacement, Reduction and Refinement) imposed by the European “Regulations and Ethical Considerations in Animal Experiments”. We recognize that scientifically speaking, it is objectionable.

Comment 3

Again, we agree with this remark. The ScP140 control peptide was not included in these experiments for the same reasons as above. As detailed in the legend of figure 3, each point in the graph represents a pooling of 2-3 livers, meaning that 35 mice have been sacrificed for these experiments. Any additional control doubles the number MRL/lpr mice. 

Comment 4

The authors should be more consistent in displaying the p values - the p values > 0.05 are shown in some graphs, while in others they are omitted.

This has been corrected in all figures in the revised version.

Comment 5

Finally, the authors provide no experimental not theoretical explanation as to the mechanisms presumably involved in CMA-dependent immunological deffects in MRL/lpr mice. Having in mind that the role of macroautophagy in B and T cell survival and function is better defined, one would expect that macroauotphagy modulation by P140 is more important in this context than that of CMA. This issue should be discussed.

Our investigations, both in vitro and in vivo (refs 3, 9, 22, 23, 24, 28, 29, 54, 55), have led to the conclusion that P140 peptide displays its protective effects via a mechanism that involves autophagy. After treatment of MRL/lpr mice with P140, autophagy markers p62/SQSTM1 and MAP1LC3 accumulate in B cells, consistent with a down-regulation of autophagic flux. Using cell lines stably expressing two different photoactivable CMA reporters, it was shown further that P140 directly interferes with this process. Expression of HSPA8/HSC70, to which P140 readily interacts, and of LAMP2A, limiting component of CMA, which are both over-expressed in MRL/lpr B cells, is down-regulated, also supporting the view that CMA represents a privileged target for P140. P140 uses the clathrin-dependent endo-lysosomal pathway to enter into MRL/lpr B lymphocytes where it accumulates in the lysosomal lumen. By purifying lysosomes from the liver and spleen of MRL/lpr mice pre-treated or not with P140, and incubating these lysosomes with CMA substrate, we observed a decrease of intra-lysosome uptake of substrate, confirming for the very fist time in vivo the inhibition effect of P140 on CMA.

As a matter of consequences of these upstream events, our hypothesis, substantiated with experimental evidence, is that P140 binds to lysosomal HSPA8 in the lysosomal lumen, and that this interaction directly hampers its chaperoning functions by altering the composition of HSPA8 heterocomplexes and/or also destabilizing LAMP2A, as a result of its effect on HSP90AA1. This effect likely interferes with the endogenous autoantigen processing and loading to MHCII molecules, thereby contributing to a much weaker priming of autoreactive CD4+ T cells, a result we described both in mouse and human settings.

Yet, we recognized that nowadays, it is not known why in lupus (as well as in other inflammatory, autoimmune or non-autoimmune settings), the CMA process is over activated. This eminent question deserves specific studies.

Although we agree on the contribution of macroautophagy to B and T survival, and in fact we have contributed to that literature, we have demonstrated that P140 has no effect on mitochondria and mitophagy (Bendorius et al. Front. Immunol. 2018; 9:2158). This motivated, in the first place, our interest on analyzing CMA that, indeed, we demonstrate is one of the primary mediators of the effect of P140.

Reviewer 2 Report

In this work, Wang et al. have shown the effect of P140/Lupuzor on CMA by reducing the CMA proteins, LAMP2A, and HSPA8 in vivo in MRL/lpr mouse model, its possible pathophysiological role in many diseases. They observed the lysosomal dysfunction by P140 administration in mice, and measured the reduced protein levels (LAMP2A, HSPA8), and altered the uptake of Tau as a potential CMA substrate using WB and quantification.

The work is interesting as there is increasing evidence of the role of P141 on (auto)inflammatory disease and claimed that the target of P141 might be CMA machinery, esp, the early steps of CMA process. However, the molecular mechanism of P141 needs to be addressed by which it regulates the process inside of autophagy-lysosomal vesicles. And certain questions need to be clarified.

1. The quantification analysis would be necessary to show the difference between the blots of with/without P140 in Fig 1, Fig 2. This is important because the conclusions were from the blots that authors showed. Another quantitative analysis is needed - other than “Densitometric quantification (LAS3000)-Image J/based on Ponceau staining.

2. the working hypothesis of P140 on the scheme in Figure 4. The authors need to show more data of the (co)locations between P141 and LAMP2A, and HSPA8, etc. for clarifying the working mechanism inside (lumen of lysosomal lumen).

Author Response

Comment 1

The quantifications have been done and the results are presented by the respective histograms (Fig. 1c and 2b). We have used total protein stain free method to quantify the blot, a method of choice that is highly recommended when different cell fractions are involved (ref. 40). We would like to clarify that two different analysis were used to determine amount of protein per line. The amount of protein loaded in the gels was quantified prior to loading but to account for differences in efficiency of transferring or running, densitometric values were normalized for total abundance of protein determined by Ponceau staining. When working with subcellular fractions, the use of the so called “housekeeping” proteins is not possible as there is not a protein that maintains equal levels across all cellular compartments. In addition, normalizing for levels of integral lyososomal proteins in this case could introduce a bias due to the observed changes in overall size and abundance of the lysosomal compartments and the anticipate cargo to lysosomal protein ratio.

Comment 2

- We would like to clarify that the proposed model was generated by combining our previous findings on the effect of P140 on CMA in culture cells and the data in vivo generated in the current study. In fact, the co-localization of P140 and HSPA8/HSC70 requested by this reviewer, was already performed in our previous studies published in 2011 (ref. 24) and 2018 (ref. 27). It has been demonstrated by fluorescence and electron microscopy. It was not possible however to definitively determine that both co-localize into lysosomes (the vesicles we identified in ref 27 resembled to endosomes). P140 interaction with HSPA8 was also demonstrated using surface plasmon resonance experiments and fishing experiments (ref. 23) at cell membrane.

-The interaction of P140 with HSP90AA1 has also been measured using surface plasmon resonance experiments (manuscript in preparation).

-The destabilization of HSPA8-HSP90AA1 complex by P140 was described in ref. 22.

- We place the effect of P140 in the lysosomal lumen because we have previously shown that incubation of intact lysosomes with P140 in vitro does not modify CMA activity. These findings are in agreement with P140 being internalized by endocytosis and reaching the lysosomal lumen through endo-lysosome fusion.

-The co-localization of P140 and LAMP2A has been published in 2015 (ref. 24) using fluorescence microscopy experiments. Again due to the resolution of images, it was not possible to definitively determine that that both co-localize within lysosomes. However, the results obtained in this work with a highly purified fraction of lysosomes active for CMA provide strong support of this compartment being the target of P140 effect. 

 

Round 2

Reviewer 1 Report

1) While following the 3Rs guidelines might justify the omission of controls in the experiment in Fig. 3, I am not sure that the same applies to Fig. 2, where 3 additional mice (as in Fig. 1) would have ensured the validity of experimental design. Also, it is unclear, at least to this reviewer, how the inclusion of one additional (and smaller, like in Fig. 1) group would double the number of mice in a 4-group experiment?

2) It is not possible to compare more than two different groups using Kruskal Wallis test alone. The authors should explain the procedure for the post-hoc comparison.

3) Finally, I am surprised that none of the explanations provided by the authors were incorporated in the manuscript. A reviewer is no different from a reader, so the authors should provide their readers with the same information, unless they find the reviewer's requests unreasonable. Therefore, at least the discussion of the TFEB data and the explanation for the missing ScP140 controls should be included in the manuscript.

Author Response

1) If we were to include the ScP140 control group in figure 2, that would require actually 6 mice, as in the other groups, to obtain significant statistics. It was indeed unfortunate that we did not include this control but because animals are treated in France and organelles isolated in USA, the current COVID restrictions and uncertainty of where are they going to be lifted, prevent us to add that additional control in a timely manner. A sentence has been added in page 11 to make this point clear and justify it without ambiguity for the reader.

2) We compared only two different groups at a time using Kruskal Wallis, meaning that we compare CBA/J vs MRL/lpr, MRL/lpr vs MRL/lpr+P140, MRL/lpr vs MRL/lpr+ScP140, for example. We have modified the text in the manuscript accordingly (page 4).

3) We have included the discussion dealing with TFEB (page 10-11) and the explanation of excluding ScP140 controls from some experiments in page 11.

Reviewer 2 Report

I think the authors did the best for this revised manuscript.

 

Because the conclusions of the effect of P140 on CMA might from the different amount of CMA players, the quantification analysis should be essential to tell the effect of P140 on TFEB or other players (in the WB blots in Fig 1, and Fig 2). With the poor quality or without repeated blots (for the quantification, or statistics with mean, SEM, etc), it is still unclear how authors could make the conclusion strongly, at least to one of the potential readers or a reviewer.

Author Response

As indicated, we show a representative western blot image in the Fig. 1C. In the legend of the figure 1C (page 6) of the revised version 2, we have now clearly specified the number of mice used for the quantification in each group, namely 7 CBA/J mice, 9 MRL/lpr mice, 6 MRL/lpr mice+P140 and 3 MRL/lpr mice+ScP140. The mean value (right panel) was calculated and the error bars represent the SEM. The number of mice was indicated in the legend of Fig. 2 (n=6 mice per group) with three western blot replicates for each sample.

 

Round 3

Reviewer 1 Report

All comments have been addressed.

Reviewer 2 Report

O.K. but really not easy to follow the author's conclusion with the current version of the MS.