Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

In this paper from Delport A et al, the authors make a nice presentation about the role of APP and beta-CTF in different tissues after a HFD. Although the experiments are well performed, several concerns have arisen after reading carefully the manuscript. 

 

Major concerns:

1. The authors indicate that, after HFD, APPP and beta-CTF appear in different tissues including kidney and several white adipose tissue compartments. How this material have accesed all of these tissues?, is this through the bloodstream in a free form or using extracellular vesicles connecting the brain with these recipient tissues?
2. The authors indicate that these changes occurs after a HFD challenge. However, is it possible to find these molecules (APP and beta-CTF) under a normal diet in the tissues where they found it under HFD?. Why the authors suggest there is no accumulation of these molecules in the liver, where it is the main organ with capacity of retention?.
3. The authors suggest that the accumulation promotes an alteration of complex I and compromise electron chain transport. Did the authors measure either ATP or ROS production?
4. Since, this alteration in the complex I is very important, it should be very interesting to analyze this capacity in the pancreas and, more especifically in beta cells, which insulin secretion capacity depends on ATP synthesis through the aerobic metabolism. 

Minor concerns:

1. What is the meaning of post-mitochondria found in the figures?

Author Response

Major concerns

Comment 1. The authors indicate that, after HFD, APPP and beta-CTF appear in different tissues including kidney and several white adipose tissue compartments. How this material have accesed all of these tissues?, is this through the bloodstream in a free form or using extracellular vesicles connecting the brain with these recipient tissues?

Response 1. We would like to thank the reviewer for the interesting query. APP is present in the different tissue through protein expression pathway induced by the presence of high-fat intake. Our study suggests that this may be through a cytokine-induced gene expression pathway upregulated by the presence of high‑fat foods. We identify IL-4, IL-13, TNF-α and IL-1β as key cytokines upregulated after high-fat diet that have the potential to induce APP gene expression. This would suggest that it is not the APP or the beta-CTF that moves between tissues through the blood stream but rather the activation of immune cells that release cytokines into the bloodstream that activate APP gene expression – in certain tissue - through unidentified cascades. The APP within the tissue is then processed by a non-conical secretase (likely BACE1) to produce the betaCTF fragment. Why APP is overexpressed during high-fat intake remains to be determined and is a key question that should be addressed in future work.

We felt that the manuscript covered this subject in a clear manner already and so no addition to the manuscript was made.

Comment 2. The authors indicate that these changes occurs after a HFD challenge. However, is it possible to find these molecules (APP and beta-CTF) under a normal diet in the tissues where they found it under HFD?. Why the authors suggest there is no accumulation of these molecules in the liver, where it is the main organ with capacity of retention?.

Response 2. We would like to thank the reviewer for the questions. APP and CTF fragments are found in various tissues under physiological conditions. This can be seen in our study in Figure 2, where we detect APP and betaCTF in ND challenged mice in all tissue screened. HFD challenge induced the overexpression and accumulation of APP and betaCTF in adipose tissue (shown previously) and now in our study the kidney. We did not observe this increase in liver or brain. This would suggest – as outlined in our discussion – that these two tissue types may lack the activation mechanism required for HFD challenge to induce APP overexpression or for the brain, be region specific. These observations are more consistent with tissue-specific regulation of APP and betaCTF expression in response to HFD, rather than accumulation driven by systemic transport through the bloodstream.

We felt the manuscript showed the presence of APP and CTF under normal conditions (Figure 2) as well as hypothesises that the overexpression is inflammatory driven through transcription rather than systemic transport so no changes to the manuscript were made.

Comment 3. The authors suggest that the accumulation promotes an alteration of complex I and compromise electron chain transport. Did the authors measure either ATP or ROS production?

Response 3. We would like to thank the reviewer for their comment. We, unfortunately, did not measure either ATP or ROS production in our study. Given prior reports that APP overexpression and mitochondrial accumulation are associated with reduced oxygen consumption rate, ATP production and ROS levels, we considered that repeating these assays may offer limited additional insight beyond existing studies (ref 5, 8, 9, 11). Mitochondrial dysfunction in response to APP overexpression seems to be a consistent outcome in vitro and in vivo, in all cases tested (ref 5, 8, 9, 11). Therefore, we sought to determine whether the inevitable mitochondrial impairment was associated with alterations in specific electron transport chain enzyme activities, particularly complex I and IV, which have not routinely been performed in prior studies.

Comment 4. Since, this alteration in the complex I is very important, it should be very interesting to analyze this capacity in the pancreas and, more especifically in beta cells, which insulin secretion capacity depends on ATP synthesis through the aerobic metabolism. 

Response 4. We would like to thank the reviewer for the interesting comment. We agree that the pancreas is an important organ to examine in the context of HFD and APP overexpression, given its susceptibility to chronic inflammation and oxidative stress. Moreover, APP is detected at the protein level in pancreas tissue of humans (https://www.proteinatlas.org/ENSG00000142192-APP/tissue). Our study was designed to examine highly metabolically active organs (brain, liver, and kidney), and therefore did not include the pancreas. We also examined adipose tissue – although not as metabolically active – is known to have a consistent APP overexpression response after HFD challenge. We do think that investigating APP in the pancreas is a worthwhile future direction and may help explain the improved glucose tolerance and insulin sensitivity observed in APP knockout mice, alongside effects in adipose tissue. 

The following was added to the manuscript to highlight this (Section 4, line 499-501): “To further characterize this dysregulation, the pancreas is a key organ for future studies due to its sensitivity to HFD-induced inflammation and its baseline APP expression [35, 36, 49].”

Minor concerns:  

Comment 1. What is the meaning of post-mitochondria found in the figures?

Response 1. We would like to thank the reviewer for the question. The post-mitochondrial fraction is the fraction that is left after the mitochondria is isolated as described in the methods and materials (section 2.7 and 2.9). This contains cytoplasm, plasma membrane (broken up during initial rupture), Golgi, and endosomes.  

In the manuscript this is outlined in Section 3.3, line 361-362: “containing plasma membrane, Golgi, and endosomes”  

We do note that we are missing the cytoplasm in this statement, so it was revised to: “…(containing cytoplasm, plasma membrane, Golgi, and endosomes).” in the manuscript.

Reviewer 2 Report

Comments and Suggestions for Authors

This study provides valuable insights into how HFD-induced APP and βCTF accumulation in peripheral mitochondria contributes to metabolic dysfunction in adipose and kidney tissues; however, certain technical details mechanistic discussions could be further refined to strengthen the overall narrative.

1. In section 2.11, regarding the enzymatic activity assays in Section 2.11, while the methodology is correctly cited, the manuscript would benefit from a brief description of the biochemical principles used for the Complex I and IV assays.
2. In sections 2.12 2.13 2.15, please correct the cell density formatting. '0.5 x 105' should be written either using superscript or in E-notation to ensure clarity.
3. The finding that APP/βCTF expression is higher in adipose and kidney tissues rather than in the liver and brain is intriguing. Since the liver is a central metabolic hub and the brain is the typical site of APP study, the authors should discuss the potential mechanisms behind this tissue-specific resistance. 
4. The authors proposed a causal relationship between APP levels and mitochondrial function. To further strengthen the proposed causal link, the authors could discuss whether the magnitude of mitochondrial impairment scales with the degree of APP accumulation across these different tissues. If the data permits, a correlation analysis would be highly informative.
5. The discussion regarding IL-4/IL-13 as drivers of APP dysregulation is novel but requires more depth. Given that IL-4 is traditionally viewed as an anti-inflammatory cytokine, the authors should expand on why this specific axis leads to an increase in APP.

Author Response

Comment 1. In section 2.11, regarding the enzymatic activity assays in Section 2.11, while the methodology is correctly cited, the manuscript would benefit from a brief description of the biochemical principles used for the Complex I and IV assays.

Response 1. We would like to thank the reviewer for the suggestion. To improve the clarity of our methods section we have made the following changes to the manuscript:

To section 2.11, line 212-221: “Briefly, complex IV activity was assessed with 1 mM reduced cytochrome c with and without 10 mM KCN using 10 µg of tissue lysate or 50 µg of cell lysate by reading absorbance at 550 nm for 2 minutes. KCN inhibited the reaction in all cases. Citrate synthase activity was measured using 1 mM 5,5′-Dithiobis(2-nitrobenzoic acid) and 10 mM Acetyl CoA with 10 µg of tissue lysate or 50 µg of cell lysate by recording the absorbance at 412 nm for 2 minutes. Complex I activity was measured using 10 mM NADH and 10 mM ubiquinone with and without 1 mM rotenone reacted with 25 µg of isolated mitochondria by measuring absorbance at 340 nm for 2 minutes. Complex I activity was taken as rotenone-sensitive activity by subtracting rotenone-resistant activity from total activity.”

Comment 2. In sections 2.12 2.13 2.15, please correct the cell density formatting. '0.5 x 105' should be written either using superscript or in E-notation to ensure clarity.

Response 2. We would like to thank the reviewer for spotting this error in our manuscript. We have corrected all instances where a superscript is required.

Changes to the manuscript were made as follows:
Line 231 - 1×10⁵ cells/ml
Line 244 - 0.5 × 10⁵ cells/ml
Line 280 - 0.75 × 10⁵ cells/ml

Comment 3. The finding that APP/βCTF expression is higher in adipose and kidney tissues rather than in the liver and brain is intriguing. Since the liver is a central metabolic hub and the brain is the typical site of APP study, the authors should discuss the potential mechanisms behind this tissue-specific resistance.

Response 3.  We would like to thank the reviewer for the comment and we agree, the tissue specificity of the response is intriguing and was somewhat unexpected.

The following was added to the discussion to further suggest potential mechanisms behind this tissue-specific resistance:

Section 4, line 487-490: “For instance, although an acute HFD does not alter total brain APP mRNA [5] – a finding we extend to protein levels after chronic feeding – hippocampal APP can be elevated by prolonged HFD exposure [7], indicating regional neuronal specificity. The hippocampus – rich in microglia and astrocytes – is particularly susceptible to HFD-induced neuroinflammation, which may account for this specificity [42-44]. Similarly, despite HFD-induced hepatic dysfunction [42-44].”

Section 4, line 492-497: “The differential hepatic overexpression of APP may reflect the nature of the underlying stimulus, as alcohol and HFD induce distinct inflammatory responses. Chronic alcohol exposure activates Kupffer cells via the liver–gut axis, promoting a neutrophil-rich response [47], whereas HFD induces lipotoxic hepatocyte stress and a macro-phage-dominated profile [48]. This suggests that HFD-associated inflammation in the liver may be less effective in driving APP overexpression.”

Comment 4. The authors proposed a causal relationship between APP levels and mitochondrial function. To further strengthen the proposed causal link, the authors could discuss whether the magnitude of mitochondrial impairment scales with the degree of APP accumulation across these different tissues. If the data permits, a correlation analysis would be highly informative.

Response 4. We would like to thank the reviewer for the suggestion to improve our manuscript and agree, discussion on whether APP level directly correlated with mitochondrial impairment would make our study more robust. To address this, we compared the magnitude of decrease in activity vs the level of APP in the mitochondria in each respective HFD exposed tissue.

We made the following changes to the manuscript to incorporate the new result:

Section 2.11, line 223: “To determine the decrease in complex I/CS or IV/CS ratio for vWAT, sqWAT and kidney tissue after HFD exposure, the I/CS or IV/CS ratio was subtracted from the mean I/CS or IV/CS for each tissue after ND exposure.”

Section 2.16, line 301: “Correlation was calculated using Pearson coefficient.”

Section 3.3, line 372: “Moreover, a positive correlation between the decrease in complex I (r = 0.6712, p < 0.05) and IV (r = 0.8946, p < 0.0001) activities and the level of mitochondrial APP observed in HFD exposed vWAT, sqWAT and kidney tissues was also apparent (Figure S3).”

Figure S3 was added to the supplementary data showcasing the correlation between the decrease in complex I/CS and complex IV/CS ratio compared to mitochondrial APP level in vWAT, sqWAT and kidney tissue after HFD exposure in each individual mouse.

Comment 5. The discussion regarding IL-4/IL-13 as drivers of APP dysregulation is novel but requires more depth. Given that IL-4 is traditionally viewed as an anti-inflammatory cytokine, the authors should expand on why this specific axis leads to an increase in APP.

Response 5. We would like to thank the reviewer for the suggestion. Although IL4 is generally classified as an anti-inflammatory cytokine when activating immune cells, it is better classified as a pleiotropic cytokine meaning it can have act as both depending on the environment.  

The following was added to the manuscript to improve the clarity of the discussion:

Section 4, line 530-538: “Although the activation of immune cells, via IL-4 type I receptor, by IL-4 is generally classified as an anti-inflammatory response for the promotion of cell proliferation and survival, the role of IL-4 in non-hematopoietic cells is pro-inflammatory [55, 56]. This, via the activation of the IL-4 type II receptor by either IL-4 or IL-13, has a prominent role in allergic inflammation [55]. The precise mechanisms by which IL-4 and IL-13 may regulate APP expression have yet to be fully elucidated. However, given that IL-4 and IL-13 can activate the Akt/NF-κB and Ras/MAPK pathways—both of which are known to regulate APP transcription—these signalling cascades represent promising avenues for future investigation [53, 56].”

Reviewer 3 Report

Comments and Suggestions for Authors

Delport et al. report on amyloid precursor protein and its regulation in peripheral tissues. They report that high-fat diet in mice increases their body weight and visceral and subcutaneous fat and affects kidney function. They show that APP transfection increases cytokines in cell culture and that certain cytokines increase APP expression in kidney and fatty tissue in vivo. During high-fat diet, cytokines are increased in vivo, and APP and its C-terminal peptide accumulate in mitochondria and reduce mitochondrial function. APP may therefore link fat-induced inflammation to mitochondrial dysfunction in fatty tissue but also in kidney.

The manuscript is written in good English. Methods are well explained, the experimental part makes sense. The manuscript complements a previous study on APP induction in fatty tissue (ref. 5) and provides additional data on cytokines and kidney function. The discussion is commendably brief and to the point.

I have only a few minor issues. In some cases, the small animal number (N=3-4) is problematic, I prefer N=6 per group. The experiment with phenserine is very descriptive as the drug has several other actions including, of course, inhibiting esterases. While complex IV is documented in Fig. 5, why is complex I activity not shown ?

Author Response

In some cases, the small animal number (N=3-4) is problematic, I prefer N=6 per group. The experiment with phenserine is very descriptive as the drug has several other actions including, of course, inhibiting esterases. While complex IV is documented in Fig. 5, why is complex I activity not shown ?

Response: We would like to thank reviewer 3 for their comments and suggestions.

With regards to the number of animals used:

Unfortunately, although n=6 per group would have improved the study robustness, this would not have been approved by the University of KwaZulu-Natal Animal Ethics Committee which stringently adheres to the three Rs for animal research design - Replacement, Reduction and Refinement. With the use of power calculation based on similar studies in the literature, n=4 per group was determined to hold enough statistical power for this study. This is outlined in the methods in section 2.1, line: 79-81.

With regards to phenserine other actions:

While it is well-known that phenserine’s main action is to inhibit acetylcholinesterase, there is little to no acetylcholinesterase found in the HEK293 cell line. Additionally, phenserine can also reduce α‑synuclein and huntingtin protein in a similar mechanism to APP. However, the levels of these two proteins are low to undetectable in the HEK293 cell line (21.2 and 14.4 nTPM) especially when compared to that of APP (236.7 nTPM) (https://www.proteinatlas.org/). While an accumulated effect of reduction of all three proteins to restore mitochondrial function is conceivable, since the cell line has such low level of α‑synuclein and huntingtin protein already, removal was expected too negligible.

The following was added to the manuscript to clarify the use of phenserine:

Section 3.4, line 404-406: “The HEK293 cell line – known to have high-levels of endogenous APP – in comparison, express little to none of the other described phenserine targets: acetylcholinesterase, α-synuclein and huntingtin [35-39].”

With regards to no complex I activity in fig5:

Complex I activity was not provided for the in vitro assays. Several attempts to obtain specific complex I activity were made from isolated mitochondria from cell culture however little rotenone-sensitive activity could be detected with the majority of the activity being rotenone-resistant and therefore would not constitute only complex I activity but other dehydrogenase activity found within the mitochondria. Therefore, as an additional orthogonal assay for mitochondrial function, the TMRE assay was used which gave better more consistent results in mammalian cell culture. This problem was not found for complex I activity from mitochondria isolated from tissue.