Escribir un comentario

The preprint “Targeting ZNRF3 and RNF43 to Restore Regeneration and Reverse Metabolic Dysfunction-Associated Steatotic Liver Disease” addresses an important therapeutic gap in metabolic dysfunction–associated steatotic liver disease (MASLD). The authors propose that transient inhibition of the WNT/β-catenin pathway repressors Zinc and Ring Finger 3 (ZNRF3) and Ring Finger protein 43 (RNF43) can enhance regeneration and improve disease outcomes. They begin by showing that WNT/β- catenin signaling is impaired across the human MASLD spectrum and hypothesize that restoring pathway activity by silencing ZNRF3/RNF43 may regress the disease.

Prior studies, including the authors’ own, demonstrated that ZNRF3/RNF43 deletion in healthy animals disrupts lipid metabolism and increases tumor risk. Here, however, the authors explore whether the enhanced proliferation resulting from β-catenin activation could be harnessed in damaged livers to promote regeneration within a controlled therapeutic window.

This preprint presents compelling data that ZNRF3/RNF43 silencing enhances regenerative capacity and reduces disease features in MASH/MASLD mouse models. The study provides valuable mechanistic insights into lipid mobilization by activation of alternative bile synthesis pathway and its effective clearance from hepatocytes. This is supported by transcriptomics and metabolomics data suggesting enhancing such pathway could be a promising step toward translational applications. Additionally, the study reports improved histological markers of inflammation and fibrosis, while also enhancing regeneration. Importantly, transient silencing using LNP-siRNA provided partial recovery from established disease, representing a step toward translational feasibility. The figures are visually clear and engaging, and the integration of histology, transcriptomics, and metabolomics strengthens the mechanistic narrative. At the same time, several claims should be moderated to reflect early-stage findings. Additional experiments are needed to confirm that the beneficial effects are specifically due to WNT/β-catenin activation and the alternative bile acid pathway, and to establish that long-term ZNRF3/RNF43 deletion is not tumorigenic. Below, I outlined areas where additional clarification or evidence could be helpful.

Major Comments

1.       The manuscript (Abstract, Introduction) states that ZNRF3/RNF43 deletion restores liver function. While bile acid synthesis and excretion improve (Figs. 4–5) and ALT remains stable, broader hepatic functions—synthetic, metabolic, and detoxification—are not assessed. Extended CYP2E1 expression (Supplementary Fig. 1, Fig. 1C–D) also suggests altered zonation. To strengthen this claim, the authors could include functional assays (e.g., carbohydrate metabolism, urea clearance) to show the metabolic zonation is also restored from Portal vein to Central vein. Otherwise, the language should be revised to specify improvements in bile acid excretion rather than global liver function.

2.       The authors conclude that there is a “safe therapeutic window” (Abstract, Discussion). No tumors were observed at 4–12 weeks post-deletion, which is encouraging, but hepatocellular carcinoma has been reported at 7–12 months in other contexts (https://doi.org/10.1016/j.stem.2021.05.013, https://doi.org/10.1038/s41467-021-27923-z). As the authors did not conclude on exact timing of the possible therapeutic window, tumor risk due to suggested repeated or long term ZNRF3/RNF43 silencing cannot be excluded from <3 months of data. A more precise phrasing could be “no adverse effects observed within the experimental timeframe.”

3.       The manuscript describes ZNRF3/RNF43 deletion as a “viable and tractable therapeutic target” (Abstract, Results). Current evidence supports an early proof-of- concept rather than a tractable therapy. The siRNA experiment is promising, showing reduced steatosis and inflammation, but fibrosis remained unchanged. Additionally, optimization of dose/duration is still needed to ensure optimal beneficial effects. To avoid overstating conclusions, the authors could frame their approach as having potential therapeutic value. Similarly, the statement of “MASH regression” (Introduction, Fig. 6) should be toned down to “partial regression of steatosis and inflammation.”

4.       The Introduction suggests beneficial effects are “mediated through” activation of the alternative bile acid pathway. Transcriptomics and metabolomics clearly show pathway engagement, but causality is not established. Unless pathway inhibition (CYP27A1 and/or CYP7B1) experiments are performed, the phrasing could be revised to “associated with activation of the alternative bile acid pathway.”

5.       CYP2E1 coverage is used throughout as a surrogate for β-catenin activation. However, Supplementary Fig. 1 shows decreased CYPs in advanced human disease, while Fig. 1C–D shows >20% CYP2E1+ hepatocytes. The use of CYP2E1 for immunofluorescence and CYP1A2 for western blot further complicates interpretation. Controls also appear nearly fully CYP2E1+ (Supplementary Fig. 2B, K). To strengthen this argument, the authors could (i) present total and activated β- catenin staining by immunofluorescence and/or (ii) quantify CYP2E1 expression (intensity) levels rather than area coverage.

6.       The study could be strengthened by direct evidence of efficient ZNRF3/RNF43 deletion. Given that lipid accumulation could impair AAV delivery, showing mRNA depletion across liver zones (from PV to PC) using FISH or similar approaches could confirm uniform knockout. Adding this to Supplementary Fig. 2 will address this concern.

7.       Claims of disease “reversal” are based on improvements from diseased states but do not show whether knockout livers approximate healthy controls (with no perturbations). Clarification of how bile acid profiles and metabolic markers compare to healthy controls (Figs. 2–6) could make the conclusions more robust.

8.       Gubra Amylin NASH (GAN) diet durations vary: 35 weeks (Fig. 2), 16 weeks (Figs. 3–5), 55 weeks (siRNA), and 28 weeks (Supplementary Fig. 2S). The rationale for these choices is unclear. To improve reader understanding the authors should justify why different timepoints were used and clarify whether the observed mechanisms (β-catenin activation, bile acid pathway engagement) depend on timing of gene deletion.

Minor Comments

1.       The penultimate sentence of the Introduction’s first paragraph is incomplete. To improve reader understanding, please revise.

2.       Supplementary Fig. 1 shows ZNRF3 expression in human liver, but RNF43 is not shown. Including RNF43 analysis could clarify whether both ligases are downregulated in terminal MASLD. The authors in the preprint describe ZNRF3/RNF43 deletion results in WNT/β-catenin activation indicated by increased CYP2E1 % area. To improve reader understanding please discuss how ZNRF3 downregulation and increased CYP2E1 area (Fig. 1C–D) do not necessarily reflect β- catenin activation in MASH Stage II–III.

3.       The claim about early activation of regeneration and metabolic changes (p. 8) based on EM data (Suppl. Fig. 2S–T) is not well contextualized in this preprint. For example, the described ER and mitochondrial morphology is not shown across the periportal (PV) to pericentral (PC) axis, which would be needed to assess zonated metabolic changes. The EM data presents very small volume within a cell. In addition, the authors do not cite prior studies (for example, https://doi.org/10.1038/s41467-024-48272-7 ) that link such morphological alterations to metabolic function. Since the figure lacks information on ER and mitochondrial remodeling across PV–CV regions and does not change the overall conclusions, the authors could consider removing this dataset.

4.       FAT-MASH experiments used male mice while MASLD experiments used both sexes. To improve reader understanding, please comment on this design choice and indicate sex in figure legends or datapoints.

5.       The Methods assume data follow a normal distribution. Please explain this assumption, particularly in the context of sex-based differences.

This review was drafted by the author and revised with the assistance of ChatGPT 5 to improve grammar, conciseness, and tone.

Competing interests

The authors declare that they have no competing interests.