PREreview del Hepatocyte Circadian Clocks Control Cholesterol Metabolism and Protect From Metabolic Dysfunction-Associated Steatohepatitis (MASH)

In this preprint, Monteiro de Assis et. al; aimed to investigate the role of the liver circadian clock in metabolic dysfunction-associated steatohepatitis (MASH). While several studies in the literature emphasize the role of circadian disruption in metabolic disorders, including MASH, a gap remains in understanding how the circadian clock may be directly involved. Thus, the authors characterized the role of the circadian gene Bmal1 in MASH progression, based on preliminary analyses indicating dysregulated cholesterol metabolism in the liver. This was done by utilizing transcriptomics/lipidomics in Bmal1 knockout mice on a normal and high-fat diet. Overall, this study reports that Bmal1absence in hepatocytes disrupts cholesterol-related gene expression in mouse livers. While cholesterol was not impacted in the livers of human patients with MASH in this study, the authors report that these patients showed indications of circadian phase shifts. Together, these conclusions provide insight as to how clock regulation of cholesterol-related metabolism could be translatable between mice and humans. While the data does highlight that a Bmal1 knockout in the liver does contribute to metabolic-related changes related to rhythmicity, the data presented do not indicate what is causing the metabolic changes and the direct connection to cholesterol. Thus, this study could be further improved by additional clarification and experiments to support the authors' claims about the link between Bmal1 and MASH.

Major Comments:

• 1. The methods indicate that female and male mice were used for this study, but it is not clear whether equal numbers of males and females were combined at each timepoint for animal collections. The authors could consider including this information to clarify the results and improve the ability of others to replicate the data.

• 2. To aid interpretation, it would be helpful if the statistical analyses used for each assay were addressed in the figure legends. Multiple circadian tools were used to identify rhythmicity (JTK_cycle, Metacycle, CircaN, and DryR) based on what was written in the results section, but the manuscript relies heavily on CircaCompare to define rhythmicity. It may be impactful to highlight how these circadian analysis tools differed statistically across this study. A supplemental figure that highlights differences between the number of rhythmic genes identified for each circadian analysis approach, with the appropriate p-value cutoffs, would suffice, if necessary. Providing this information would be helpful for readers to assess the magnitude of significance and replicate the data in this study.

• 3. Figures 2 and 4 use the ChEA3 algorithm to predict candidate transcription factors associated with differentially regulated genes (DRGs). Given that the ChEA3 analysis primarily identifies candidate genes, further experiments that provide mechanistic validation of the identified targets would strengthen claims about the role of Bmal1 in lipid and cholesterol metabolism. The same rationale may also be considered for the ChEA3 analysis conducted for Figure 4, which has similar limitations and challenges. For instance, Chrebp/Mixipl was identified as a candidate transcription factor (TF) in the ChEA3 analysis, but justification for hypothesizing its role in an interaction with Bmal1 for MASLD/MASH progression can be further developed. Explanation for the Bmal1 and Chrebp double knockdowns in this context may help readers justify the rationale behind the experiments in the figure.

• 4. The paragraph above Figure 3 highlights the use of the peak time, ZT 6, in reference to previous literature from the lab (de Asis et. al; 2024; http://doi.org/10.7554/eLife.79405.), which indicated peaks in triglyceride and cholesterol levels. If this is in reference to Figure 5 of the cited study, this time point doesn’t appear to be justified, given that the peaks appear earlier during the day. In addition, other literature outside of the laboratory can corroborate the use of this time point to assess triglyceride and cholesterol levels during the day, such as Adamovich et. al; 2014 (https://doi.org/10.1016/j.cmet.2013.12.016) and Lu et. al; 2023 (https://doi.org/10.3390/nu15112547).

• 5. The results from the translational experiments in Figure 6 did not indicate a significant association between circadian rhythmicity and MASH. Conversely, the authors reported a correlation between disrupted circadian rhythms and MASH due to their non-significant internal phase measurements and positive correlation between hepatic cholesterol and serum HDL levels (the statistical analysis was not stated). Based on these results, the authors could consider rephrasing the text in the results to clarify the level of association between MASH and circadian rhythmicity.

Minor Comments:

• 1. In the second-to-last paragraph of the introduction, the authors state “the absence of a hepatocyte clock.” However, the previous sentence highlights that a Bmal1 deletion in the liver results in the ‘rewiring’ of the liver transcriptome overall clock function. Can the authors comment on this discrepancy in the text to provide readers with a more nuanced understanding?

• 2. Figure 3C assesses the total hepatic cholesterol across genotype and diet. Given that the animals in this study were collected across multiple timepoints to reflect circadian rhythmicity, does the data reflect the average total cholesterol across all time points or cholesterol expression at that given time of day? Providing this information would be helpful for replication purposes and increase the insight into the magnitude of the result and overall claims in this study.

• 3. The rationale for the use of PER3 and NR1D1 in the analysis for an internal phase index is not clear, especially given that the literature supports more rigorous phase estimators to identify gene expression levels across time of day, such as Anafi et. al; 2017 (https://doi.org/10.1073/pnas.1619320114) and Talamanca et. al; 2023 (https://doi.org/10.1126/science.add0846). References to additional studies that implement phase estimation (such as the literature mentioned) could improve reader understanding and justify the use of the phase estimator approach in this study.

Competing interests

The authors declare that they have no competing interests.

Use of Artificial Intelligence (AI)

The authors declare that they did not use generative AI to come up with new ideas for their review.