Caki-1 Spheroids as a Renal Model for Studying Free Fatty Acid-Induced Lipotoxicity

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

General Comments

 

The manuscript by Battle et al uses a 3D model of renal proximal tubule cells, Caki-1 spheroids, to study the mechanism of palmitate-induced lipotoxicity in this nephron segment.   Their studies indicated that palmitate (up to 300 µM) was cytotoxic unlike oleic, consistent with the observations of other labs. And that this was associated with PARP cleavage, indicative of apoptosis.         Furthermore, palmitate toxicity was associated with ER stress (CHOP upregulation) and defective autophagy. These effects were mitigated by Oleic and the long chain fatty acyl synthetase inhibitor triacsin C (which prevents the conversion of palmitate to palmitoyl-CoA.   The formation of lipid droplets was evaluated as a possible means of protection from cytotoxicity.  Overall, the conducted studies were quite interesting. However, a number of issues still need to be resolved, as described below.

 

Specific Comments

 

Methods

 

Line 109. What type (brand) of microscope?

Line 120. Were different lysates equalized with respect to protein and if so, how?

 

Results

 

Line 148-9. Were the spheroids solid or hollow?  In Fig 1c, does the data differ significantly between Day 4 and Day 12? It appears so from the plot.  Similarly, in Fig 1d Days 4 and 8 appear to be significantly different, as well as da4 vis day 12.  Thus, there may very well be significant variation in these parameters over time.

 

Figure 2A. The different conditions in the ordinate should be labelled.  Also, there should be quantitative data associated with Fig. 2a and b (which are representative spheroids only). For example, the percent of spheroids disrupted by PA/ the number of spheroids scored; similarly, the percent of dead cells per spheroid section (given the section area is constant) should be quantified. In 2b, the number of cells dissociated from the spheroid in a particular section should also be quantified.

 

Figure 3a. More information needs to be included about the meaning behind the error analysis in the figure.

 

Figure 3b. The different conditions in the ordinate need to be labelled. Also, in Fig 3b, in the PA300 condition it appears that only dead cells are labeled on the exterior of the spheroid. This suggests a lack of penetration of the antibody against caspase 3 into the center of the spheroid.

 

Line 218. The data in Fig 4b suggest that LCB-II is not significantly elevated with PA compared with the control (especially as 1 of the control bands is higher than the other). Similarly, the result is unclear in Fig. 4d.

 

Figure 4a. The abscissa should be labelled with regards to the different conditions.

 

Figure 5a. The abscissa should have appropriate labels e.g. live, dead, composite. Also, in Figure 5a. there should be an accompanied graph with quantitation.

 

Lines 242-244. It is unclear that oleate prevented PARP cleavage and reduced p62 and CHOP protein expression in palmitate-overloaded spheroids.  This is because there is not a control with OA alone included in the experimental results.  Also, in Fig 5d, it is unclear from the western blot that LC3BII is elevated relative to control as indicated in the graph.

 

Lines 267-8. It is unclear in Fig 6b that FFA-treated spheroids show an increase in lipid vesicles. The green staining appears saturated in all conditions. It is unclear that lipid accumulation was more predominant in oleate treated spheroids. For some reason the stained region was of a smaller area with PA300 than the other 3 conditions, including controls without lipid. It is unclear that this specific spheroid is typical of other PA spheroids.

 

Figure 6a. There should be a clear explanation of the meaning of the variation and error analysis in the different conditions in the graph.

 

Figure 6b. Labels should be included on the ordinate.

 

Figure 7a. Does the PAR cleavage yield an 89 KD fragment distinctive of caspase activation (also Fig 4,5)?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

 

Overall, the authors showed that upon palmitate treatment, Caki-1 spheroids showed impaired morphology, and increased markers of apoptosis and cell death. However, the authors failed to characterize the model sufficiently and, more importantly, failed to show that their spheroid model can respond to palmitate at a metabolic level. Metabolomics analysis, gene expression, and protein expression of key players in the lipid metabolism are missing in the present study, which is crucial information to ensure their metabolism is active and functioning correctly. Additionally, good negative controls are lacking, making it difficult to differentiate whether the toxiticy effect observed in the spheroids is due to palmitate treatment and consequent lipid toxicity or due to something else.

 

 

Major comments:

 

1) A thorough characterization of the spheroids model is missing, as the “model is suitable” criteria is based on solely morphological observations. The authors should at least include immunofluorescent stainings of the spheroids showing their approriate polarization, and the gene expression levels and functionality of key transporters and enzymes involved in palmitate (and overall lipid) metabolism. This will ensure that the morphological changes also match a change in gene expression and/or protein expression patterns, and this information should be added to figure 1 and 2.

 

2) Despite Caki-1 cell line has been used for proximal tubule modeling, they lack crucial characteristics that might make lipotoxicity studies difficult to interpret, and might lead to misinterpretation of the proposed mechanisms, and overall results. For instance, Caki-1 cell lines have shown to have high levels of mannitol permeation, amongst other major limitations (doi: 10.3390/ph14090908), which is crucial in lipid metabolism. I would suggest the authors to elaborate on the limitations of this cell line when studying lipotoxiticy, and any other lipid metabolism-related mechanisms.

 

3) What solvent was used for palmitate? Where the negative controls treated with the same amount of reagent to control for the solvent toxicity? Are the doses of palmitate chosen relevant? The effect on cell death and apoptosis at the 300 µM dose of palmitate may be due to the toxicity of either the solvent used to reconstitute palmitate or the extremely high dose of palmitate itself. Authors should explain their choice of doses, the solvent used, the treatment their negative controls were exposed to, etc. As it is presented now, there is no way to differentiate between the effects of palmitate treatment or the above-mentioned factors.

 

4) Despite in the introduction section oleate is briefly introduced, there is no further mention to the reason why oleate was chosen as a treatment next to palmitate in figure 3. Was OA introduced as a control FAA and what did the authors expect to see upon treatment? The spheroids treated with OA seem to be dead, or at least severely affected by OA treatment. Authors should explain in this section why the OA treatment was taken along and what the brightfield changes observed in figure 3 panel b indicate.

 

5) In section 3.1 in the results the authors state that they treated “in the present study” with triacsin C (4 µM) for 1 h. Please clarify whether “in the present study” means the present experiment. Also, specify whether triacsin C treatment was also continued when adding palmitate after 1h or only during the 1 h pre-incubation. Based on the choice of treatment regime, authors should explain their reasoning and show some gene expression pattern to confirm that the transporters and enzymes involved in this process are expressed in their spheroid models.

 

6) In section 3.4, line 237-239, the sentence reads “Our preliminary study showed that simultaneous treatment with oleate (50 - 300 µM) prevented palmitate (300 µM)-mediated PARP cleavage in Caki-1 2D culture (data not shown)”. I would suggest adding this data in supplementary material and also compare it to the experimental data in the spheroids to strengthen the point that their 3D cultures show a better response than their 2D cultures, if so.

 

7) In Figure 6b, the BODIPY 493/503 staining seems mostly background and authors did not quantify it. To claim that lipid droplets were more abundant in the treated conditions, authors should provide better images that can be measured with image analysis software. Also, as mentioned earlier, the OA-treated spheroid seems impaired, and the authors should comment on that as well.

 

8) The supplementary materials provided are not well organized, making it difficult to trace result back. Some Western blot gels are only consisting of the beta-actin control, or they are cut and the beta-actin control is missing. Additionally, from the images of the spheroids the authors provided it can be observed that spheroids for the same experiment were in the same plate at the same time, thus, a technical replicate instead of a biological replicate. Authors should state how many biologicals and technical replicates were done for each experiment, and whether they considered each spheroid a biological replicate or a technical one. Lastly, it seems they used a normal fluorescent microscope to measure their fluorescent readouts. With 3D structures a confocal microscope is crucial and it would help the authors to quantify the immunostainings correctly.

 

Minor comments:

 

1) Line 64-65: The sentence reads “How far spheroids mimic in vivo lipid metabolism, however, remains enigmatic.” Maybe use another word instead of enigmatic (e.g. poorly understood)

 

2) Line 66: The sentence reads “To this end, the main goal of this work was to evaluate the usefulness of Caki-1 3D…”. Maybe use another word instead of usefulness (e.g. potential).

 

3) Line 190-191: The sentence reads “As expected, very few caspase3-positive cells were detected in the spheroids in both normal and BSA controls.”. It is understandable that you expect certain outcomes in your results, however, I would rephrase this as it shows a biased opinion about your results.

 

4) I would strongly suggest including the individual dots in the plots corresponding to each biological replicate, as well as all images used for analysis in the supplementary data.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript by Battle et al. has been substantially revised in response to the reviewer’s comment, and is acceptable for publication in its present form.

Author Response

Comments: A thorough characterization of the spheroids model is missing, as the “model is suitable” criteria is based on solely morphological observations. The authors should at least include immunofluorescent stainings of the spheroids showing their appropriate polarization, and the gene expression levels and functionality of key transporters and enzymes involved in palmitate (and overall lipid) metabolism. This will ensure that the morphological changes also match a change in gene expression and/or protein expression patterns, and this information should be added to figure 1 and 2.

Response: We appreciate the reviewer’s suggestion to further characterize the spheroid model beyond morphological observations. While we agree that additional functional characterization is interesting, the main goal of this study is not to demonstrate the superiority of tissue structure but to develop a useful renal lipotoxicity assay model. In our revised version of the manuscript, we have included additional cytotoxicity data (Figure 2c-e) to further support the suitability of our model for lipotoxicity studies. Additionally, we re-emphasize the importance of ability to study lipotoxicity and de-emphasize the potential molecular and structural benefits which we have not analyzed yet.

A full gene and protein expression analysis, while interesting, is beyond the intended scope of this manuscript. We have modified our manuscript to address potential limitations of Caki-1 spheroids to mimic normal PT while also expanding on the well-known relevance of Caki-1 spheroids to mimic RCC, a benefit we previously did not mention. Unlike for the liver and heart tissues where there are many reports and studies of lipotoxicity, studies on the effect of metabolic conditions and lipotoxicity on kidney tissues are lacking. This manuscript seeks to fill this need.