p53 and TIGAR promote redox control to protect against metabolic dysfunction-associated steatohepatitis

TP53 is a potent tumour suppressor that coordinates diverse stress response programmes within the cell. The activity of p53 is frequently context and cell type-dependent, and ranges from pro-survival activities, including the implementation of transient cell cycle arrest and metabolic rewiring, through to cell death. In addition to tumour suppressor functions, p53 also has established roles in the pathological response to stress that occurs during tissue damage and repair, including within the liver. Metabolic dysfunction-associated steatohepatitis (MASH) is a major driver of hepatocellular carcinoma (HCC), but our understanding of the molecular determinants of MASH development remains incomplete. Here, using a p53 reporter mouse, we report early and sustained activation of hepatic p53 in response to an obesogenic high fat and high sugar diet. In this context, liver-specific loss of p53 accelerates the progression of benign fatty liver disease to MASH that is characterised by high levels of reactive oxygen species (ROS), extensive fibrosis, and chronic inflammation. Using an in vitro culture system, we show that p53 functions to control ROS and protect against the development of MASH, in part through induction of the antioxidant gene TP53-induced glycolysis and apoptosis regulator (TIGAR). Our work demonstrates an important role for the p53-TIGAR axis in protecting against MASH, and identifies redox control as an essential barrier against liver disease progression.


INTRODUCTION
Liver cancer (Hepatocellular carcinoma (HCC)) is a common and lethal disease with limited treatment options and increasing prevalence 1,2 .HCC arising from metabolic dysfunction-associated steatohepatitis (MASH, formerly called non-alcoholic steatohepatitis (NASH) 3 ) represents the fastest growing fraction of HCC mortality globally and has become the leading cause of HCC in some western countries, replacing viral hepatitis 2,4,5 .Liver disease progression from benign metabolic dysfunction-associated steatotic liver disease (MASLD, formerly called non-alcoholic fatty liver disease (NAFLD) 3 ) to diet-induced MASH is strongly linked with the consumption of a high fat and high sugar 'Western' diet as well as with features of metabolic syndrome (MetS), including obesity, insulin resistance, hypertension, and hyperlipidaemia [6][7][8][9] .Of concern, MASH has especially high prevalence in MetS patients with co-morbid type 2 diabetes mellitus (T2DM).Disease progression is particularly rapid in this cohort, and the development of HCC is more common 10 .
The p53 transcription factor coordinates a diverse cellular stress response to balance adaptation and cell survival against senescence and cell death [24][25][26][27][28] .As in many tissues, both unrestrained activation of p53 and the absence of p53 in the liver can be lethal 29,30 .Outside of these severe cases, however, context is an important mediator of p53 activity in the liver.On the one hand, undue p53 pathway activation has been shown to impede liver regeneration, promote fibrosis, and support HCC development following chronic liver damage [31][32][33][34][35] .p53-mediated apoptosis has also been reported to promote MASH in a nutritional stress-induced methionine and choline-deficient (MCD) diet model 36 .Even so, p53 can also exert clear hepatoprotective activity.During damage-induced liver regeneration, for example, p53 has also been shown to limit inflammation and fibrosis, protect against fatty-acid induced apoptosis, act to maintain mitotic fidelity, and promote detoxification of damaging lipid peroxidation ROS [37][38][39][40][41][42][43] .
Redox control is an important aspect of p53 function more broadly and is commonly coordinated by downstream antioxidant effector proteins including TP53-inducible glycolysis and apoptosis regulator (TIGAR), amongst many others 25 .TIGAR is a fructose-2,6bisphosphatase that acts in this capacity to limit accumulation of fructose-2,6-bisphosphate, a potent allosteric activator of phosphofructokinase-1 (PFK-1) and glycolytic activity more generally 44,45 .In doing so, TIGAR activation dampens PFK-1 activity, thereby reducing glycolytic flux, a response that has been shown to promote the generation of nicotinamide adenine dinucleotide phosphate (NADPH) via increased utilisation of the oxidative pentose phosphate pathway [44][45][46] .Although it is generally cytoplasmic, TIGAR can also localise to the mitochondria where it exerts additional antioxidant activity that is independent of its fructose-2,6-bisphophatase activity 47,48 .Functionally, TIGAR has been shown to support intestinal regeneration and promote tumourigenesis, in part by enhancing the detoxification of peroxidised lipids and keeping ROS levels low 49 .
Here, we investigate the importance of p53 activity as a mediator of diet-induced liver disease progression.Our work demonstrates an important role for p53-mediated redox control in protecting against the development of MASH in vivo.We have further identified TIGAR as an important mediator of protective p53 activity in this context and provide evidence that antioxidant therapy can ameliorate features of MASH.

Liver p53 is engaged in response to a high fat and high sugar 'Western' diet
In humans, increased expression of either TP53 or the p53 target gene CDKN1A/p21 (p21) is correlated with MASH, liver fibrosis, and T2DM 20,21 .These observations suggest a potential role for p53 in mediating the aetiology of diet-induced liver disease.We have previously utilised a p53 reporter mouse to non-invasively monitor the p53 pathway after total body gamma-irradiation and during hepatotoxin-mediated liver regeneration 50 .In p53 reporter mice, active p53 induces expression of the near-infrared fluorescent protein iRFP713 (iRFP) that is linked to the synthetic PG13 p53-responsive promoter (Fig. S1A).The resulting fluorescence can be monitored non-invasively and longitudinally 50 .
To explore the dynamics of p53 activity during diet-induced liver disease, male and female p53 reporter mice were shifted onto either a purified high fat and high sugar (HFHS) 'Western' diet containing 59% energy from fat, 15% energy from protein, 25% energy from carbohydrates, and rich in sucrose 51 or onto a purified control diet containing 10% energy from fat, 14% energy from protein, 76% energy from carbohydrates, and reduced sucrose content.Mice were maintained on either diet for a period of up to 300 days and regularly monitored for weight gain and iRFP expression (Figs. 1A-F and S1B).As expected, both male and female mice exhibited weight gain on the HFHS diet relative to mice on the control diet (chow) within the first 50 days (Fig. 1E/F).Consistent with reports linking p53 activity with MASH 20,21,36 , we observed robust signal from the p53 reporter in both male and female mice by the end of our study (Figs.1A-D and S1B).Interestingly, there was also significant and sustained activation of the p53 reporter within male mice beginning much earlier, starting at 100 days of the HFHS diet and increasing thereafter (Fig. 1A/B).At this timepoint, although HFHS-fed mice of both sexes had gained significant weight, they did not exhibit significant changes in oral glucose tolerance compared with mice on the control diet (Fig. S1C), suggesting that diet-induced MetS was still in the early stages.Within female HFHS-fed mice, even though initial weight gain occurred similarly rapidly, induction of the p53 reporter was a much later event than in male HFHS-fed mice-at 200 days of the HFHS diet-and occurred with significantly lower fold iRFP reporter induction compared with male mice (Figs.1C/D and S1D).
Although it is possible to differentiate the spatial distribution of iRFP expression on wholebody scans, for example between liver-and gut region-specific signals, organ overlap within a region of interest can sometimes obfuscate the source of p53 reporter signal within the two-dimensional images.Ex vivo analysis of organs is unambiguous, however, and confirmed that p53 induction was restricted to the liver in HFHS-fed mice sampled throughout the duration of the study (after 100d, 150d, and 300d of HFHS diet) (Fig. 1G).
Together, these findings reveal that p53 induction is an early and sustained feature of the hepatic response to chronic consumption of a high fat and high sugar 'Western' diet.Although male and female mice both induce p53, its induction is particularly rapid and robust in male mice in vivo.

Liver-specific loss of p53 accelerates 'Western' diet-induced MASH
Within the liver, p53 activity can be both protective and damaging, seemingly determined in part by the extent and duration of underlying liver damage 34 .To explore potential functional roles for p53 activity during diet-induced liver disease, we created cohorts of mice harbouring liver-specific deletion of Trp53 (p53) (Albumin-Cre+; Trp53 FL/FL mice) and proceeded to characterise the long-term response to the HFHS diet within this model.Importantly, Albumin-Cre+; Trp53 FL/FL mice retain wildtype p53 expression in all tissues except the liver.They are therefore not susceptible to developing cancer before approximately 2 years of age, as previously reported 43 , and allow for long-term studies into the effects of p53 loss on liver biology.Based on our findings in Figure 1, we focused on examining the response to the HFHS diet within male mice in these experiments to provide the best signal-to-noise resolution of potential aspects of p53 function.
In Albumin-Cre+; Trp53 FL/FL male mice, we observed clear signs of MASH after a period of one year on the HFHS diet that remained largely absent in Trp53 wildtype mice fed the HFHS diet (Figs.2A-E and S2A-C).Compared with either Albumin-Cre+; Trp53 FL/FL mice on the control diet or with Albumin-Cre+; Trp53 wildtype mice on the HFHS diet, Albumin-Cre+; p53 FL/FL HFHS diet fed mice exhibited histological features of MASH, including abundant steatosis, hepatocyte ballooning, the accumulation of intracytoplasmic proteins known as Mallory-Denk bodies, hepatic hypertrophy, and inflammatory infiltrates in the liver (Figs.2A and S2A).MASH within Albumin-Cre+; Trp53 FL/FL HFHS diet fed mice was further characterised by significantly increased levels of immunohistochemical (IHC) staining for malondialdehyde (MDA), a marker of lipid peroxidation, substantial staining for picrosirius red (PSR), a marker of fibrosis, and a clear increase in hepatic lipogranulomas, consistent with increased chronic inflammation (Figs.2A-E and S2A).This was not the case in Albumin-Cre+; Trp53 wildtype HFHS diet fed mice, where features of MASLD, such as lipid accumulation (Figs.2A/B and S2A) were also evident, but with significantly lower staining for further markers indicative of MASH, consistent with greater ongoing damage in Albumin-Cre+; Trp53 FL/FL HFHS diet fed mice compared with Albumin-Cre+; Trp53 wildtype HFHS fed mice.
Functionally, plasma levels of alanine transaminase (ALT) and aspartate aminotransferase (AST) enzyme activity were both elevated in Albumin-Cre+; Trp53 FL/FL HFHS diet fed mice, consistent with liver damage and compromised liver function in these mice compared to other cohorts at the conclusion of the study (Fig. S2B/C).Even so, overall survival within this time period was not significantly different between Albumin-Cre+; Trp53 wildtype and liver Trp53-deficient HFHS diet mice, suggesting a longer latency was necessary for consistent development of MASH-HCC within this HFHS diet model (Fig. S2D).Based on these findings, we concluded that p53 exerts a protective function to oppose MASLD to MASH progression in vivo.Although the presence of wildtype p53 did not alter lipid accumulation resulting from long-term HFHS diet consumption, downstream detrimental effects, including undue lipid peroxidation, fibrosis, and inflammation were significantly reduced-and liver function was maintained-in HFHS diet mice that retained p53 activity.

Loss of liver p53 is lethal in the Lep ob/ob leptin-deficient genetic model of obesity
Leptin-deficient Lep ob/ob (ob/ob) mice develop progressive liver disease with features of MASLD from an early age that progresses to include evidence of MASH within 20-30 weeks as a consequence of chronic hyperphagia 52,53 .To examine whether p53 also exerts protective activity against liver disease within this context, we created cohorts of male and female ob/ob mice harbouring liver-specific deletion of Trp53 (ob/ob; Albumin-Cre+; Trp53 FL/FL mice) and proceeded to assess liver disease development within this model compared with Trp53 wildtype ob/ob; Albumin-Cre+ mice.
Consistent with published reports 52 , we observed significant steatosis in our Trp53 wildtype ob/ob; Albumin-Cre+ mice but this did not appreciably progress to MASH within the study period (Fig. 3A-C).Unexpectedly, however, the loss of liver Trp53 was lethal in both male and female ob/ob; Albumin-Cre+; Trp53 FL/FL mice (Fig. 3A/B).In comparison to Trp53 wildtype ob/ob mice, where all cohort mice survived to the study endpoint, a median survival of just 319 days was observed in ob/ob; Albumin-Cre+; Trp53 FL/FL mice (Fig. 3A/B).Survival within this cohort appeared to be independent of sex, with both male and female mice exhibiting a median survival based on humane clinical endpoint of between 295 days of age (male) and 319 days (female).Histopathological analysis of end-point samples confirmed a preponderance of liver tumours within ob/ob; Albumin-Cre+; Trp53 FL/FL mice (8/10 with evidence of liver tumours at necropsy with 5/10 confirmed to have significant tumour burden by histopathology), consistent with a strong liver-related phenotype.
As with Albumin-Cre+; Trp53 FL/FL mice on the HFHS diet (Fig. 2), liver tissue sections from ob/ob; Albumin-Cre+; Trp53 FL/FL mice at endpoint were characterised by elevated IHC staining for MDA (Fig. 3C/D).Liver tumours from ob/ob; Albumin-Cre+; p53 FL/FL mice also exhibited extensive fibrosis (Figs.3C and 3E).Unlike in the HFHS diet model, however, ob/ob; Albumin-Cre+; Trp53 FL/FL tumour mice had fewer lipogranulomas than age-matched Trp53 wildtype ob/ob mice, although overall macrophage infiltration was significantly increased (Figs. 3C and 3F/G).These potentially related findings could be due to differences in macrophage response to MASLD/MASH in the ob/ob model compared with diet-induced disease, which is in accordance with previous studies 52 .It could also relate to the nature of the tumours observed in ob/ob; Albumin-Cre+; Trp53 FL/FL mice, which generally lacked visible steatosis (Fig. 3C).Collectively, these findings suggest that as in diet-induced liver disease, liver p53 acts to limit MASLD to MASH progression in the ob/ob leptin-deficient genetic model of obesity in vivo.

P53 engages TIGAR and supports redox control in response to nutrient excess
Comparison of our findings across diet-induced and genetic models of liver disease highlighted diminished redox control as one of the common connecting characteristics.We have previously reported on protective functions of p53-mediated redox control in the liver during regeneration 43 and sought to further interrogate the possibility of a similar paradigm operating to protect against MASH.To do so, we developed an in vitro medium formulation based on Plasmax, a culture medium that was developed to mimic human plasma 54 , that additionally matched the relative proportions of the major fatty acids (steric, palmitic, oleic, linoleic, and linolenic acids) and carbohydrate sources (glucose and fructose) found within our murine HFHS diet.In order to ensure their bioavailability, the fatty acids in our formulation were conjugated to bovine serum albumin (BSA) to improve solubility within medium.For this reason, we utilised Plasmax supplemented with an equivalent amount of fatty acid-free BSA as the control 'baseline' formulation for in vitro experiments.
To explore the role of p53 in hepatocytes during the response to diet stress more fully in vitro, we turned to murine (Hep53.4)and human (HepG2) HCC cell lines that maintain wildtype p53 43,55 .Consistent with the observations in HFHS diet fed mice, treatment of either cell line with HFHS Plasmax medium for a period of 48 hours increased expression of the potent p53 target genes p21 and MDM2 56 , as well as elevated the levels of TP53inducible glycolysis and apoptosis regulator (TIGAR)-a key mediator of p53 redox regulation (Fig. 4A-D) 25,26 .This response was abrogated in HepG2 cells with CRISPRmediated knock-out of TP53 (p53 KO), confirming the role of p53 in the upregulation of TIGAR expression in response to HFHS medium (Fig. 4C/D).
Functionally, although p53 WT and p53 KO HepG2 cells accumulated similar amounts of lipid after 48hrs of growth within HFHS medium, HepG2 p53 KO cells exhibited significantly increased levels of ROS that were not observed within HFHS-treated HepG2 p53 WT cells (Fig. 4E/F).This response was also observed after siRNA-mediated depletion of TP53 in the HepG2 cells-although basal p53 protein levels were difficult to detect compared with downstream targets in the HepG2 cells, as has been previously reported 57 -or following siRNA-mediated depletion of Trp53 within the Hep53.4 cells (Figs. 4G-J and S3A-E).Treatment with the anti-oxidant N-acetyl-cysteine (NAC) was sufficient to restore redox control to HFHS-treated TP53 KO HepG2 cells without altering lipid accumulation, consistent with p53-mediated redox control supporting this process (Fig. S3F/G).Conversely, siRNA-mediated knockdown of TIGAR/Tigar in both p53 WT HepG2 and Hep53.4 cells phenocopied CRISPR-mediated p53 depletion, resulting in significantly increased ROS in HFHS conditions without altering lipid accumulation (Fig. 4G-J).
TIGAR has been reported to exert antioxidant activities both in the cytoplasm as a consequence of its fructose-2,6-bisphosphatase activity and at the mitochondria independent of this function 47,48 .In both HepG2 and Hep53.4 cells we observed a clear double band for TIGAR on western blots (Figs.4A/C and S3A/D).Interestingly, these bands were found to preferentially localise to either cytoplasmic or mitochondriallyenriched fractions and the clearest TIGAR accumulation we observed was within the mitochondrial-enriched cell fraction in HepG2 cells treated with HFHS medium (Fig. 4K/L).Of note, these findings did not generalise to other TP53 wildtype human cancer cell lines we examined, including the SK-Hep-1 liver cancer cell line or the colorectal carcinoma HCT116 cell line.The majority of TIGAR in these cells was cytoplasmic, the p53 pathway was not induced by HFHS diet treatment, and perturbing p53 with siRNA in SK-Hep-1 cells did not alter cellular ROS levels (Fig. S3H-J).Although derived from the liver of a patient with adenocarcinoma, SK-Hep-1 cells are endothelial in origin, with gene expression and morphological characteristics more consistent with liver sinusoidal endothelial cells (LSECs) rather than with an epithelial hepatocyte origin 58 .We believe this difference explains the observed lack of a p53-TIGAR response to HFHS conditions within SK-Hep-1 cells, as in the colorectal HCT116 cells, compared with the hepatocellular HepG2 and Hep53.4 cell lines.
Based on these findings, we concluded that p53-mediated induction of TIGAR represents an important aspect of redox control within the hepatocellular compartment during the response to HFHS conditions in vitro.

Loss of TIGAR increases ROS and accelerates liver disease in leptin receptor-deficient (Lepr db/db ) mice
Based on our in vitro data identifying TIGAR as an important mediator for managing the redox response to HFHS diet conditions, we sought to validate its importance in liver disease more broadly by returning to an in vivo model.Since the leptin-deficient genetic model of obesity yielded a rapid and potent example of repercussions for liver p53 loss (Fig. 3), we decided to examine the consequences of Tigar loss in a similar setting.Owing to the fact that Tigar and leptin (Lep) share murine chromosome 6, instead of attempting to create a Tigar-deficient; Lep ob compound mouse, we instead generated leptin receptordeficient (Lepr db/db (db/db)) mice harbouring whole body deletion of Tigar (Tigar -/-; db/db mice).As with ob/ob mice, db/db mice develop progressive obesity due to hyperphagia alongside hyperglycaemia and frank diabetes 53 .The diabetic condition of db/db mice is generally more severe than that observed in ob/ob mice and lifespan is markedly decreased 53,59 .This is especially true within male db/db mice, where mean survival has been reported to be only 14 months, compared with 19 months for female db/db mice 59 .Considering the already rapid lethality that we observed in ob/ob; Albumin-Cre; Trp53 FL/FL mice (Fig. 3A), we decided to focus on liver disease development within cohorts of female Tigar -/-; db/db mice compared with Tigar WT; db/db control mice to ameliorate the potentially confounding effects of accelerated lethality expected in db/db male mice.Young (100-day-old) Tigar -/-; db/db female mice exhibited histopathological features of MASLD including steatosis and hepatocyte ballooning alongside significantly increased IHC staining for MDA-indicative of elevated lipid peroxidation (Fig. 5A/B).In addition, Tigar -/-; db/db female mice also had greater ALT and AST serum enzyme activity than Tigar WT; db/db, consistent with established liver damage in these mice (Fig. 5C/D).These findings are consistent with our data from ob/ob; Albumin-Cre; Trp53 FL/FL mice, albeit from a cohort examined at a much earlier time point and with a differing genetic driver of obesity.In contrast with ob/ob; Albumin-Cre; Trp53 FL/FL mice, 100-day-old Tigar -/-; db/db female mice did not exhibit increased levels of picrosirius red staining for fibrosis or have altered levels of lipogranulomas formation compared with age-matched Tigar WT; db/db mice (Fig. 5E/F).These differences could be a consequence of the earlier timepoint examined in Db mice (~100 days vs ~300 days), but could also be related to reported differences in the manifestation of MASLD/MASH between ob/ob and db/db mouse strains 60 .
To test whether redox control, as opposed to other functions of TIGAR, underpinned our findings in vivo, we supplemented the drinking water of Tigar -/-; db/db and Tigar WT; db/db female mice with the anti-oxidant NAC beginning at 6 weeks of age and assessed liver disease development within these cohorts when the mice reached 100 days of age.In NAC-treated mice, the histology of Tigar -/-; db/db and Tigar WT; db/db mice were similar and IHC staining for MDA was low for both, suggesting that NAC supplementation could prevent the increased lipid peroxidation and histopathological features of MASLD evident within Tigar -/-; db/db mice-at least at the timepoint examined (Fig. 5G/H).Staining for fibrosis was similarly low in both NAC-treated cohorts, although we did observe a small increase in the number of lipogranulomas per field within Tigar -/-; db/db mice treated with NAC compared with NAC-treated Tigar WT; db/db controls (Fig. S4A-C).
Together, these findings suggest that Tigar is important for limiting liver disease progression within the db/db genetic model and that targeting ROS directly with an antioxidant can alleviate detrimental effects of Tigar loss in vivo.

DISCUSSION
During the transition from MASLD to MASH, the liver undergoes successive alterations that are reminiscent of key steps of tumourigenesis.Progressive rewiring of hepatic metabolism, significant inflammation, development of tissue damage, and engagement of cancer-associated signalling programmes all contribute to the multi-faceted disease state that defines advanced MASH 8,9 .Many of these attributes are also important hallmarks of cancer 11 .With this overlap in mind, we have interrogated the activity of the tumour suppressor protein p53 throughout MASLD and MASH development.Our findings identify p53-mediated redox control and the induction of TIGAR as important protective features of the hepatic p53 response both in vivo and in human and murine HCC cell lines in vitro.
We report that p53 activation is an early feature of MASLD pathogenesis that increases alongside disease progression in p53 reporter mice.Across experimental models, we have found that loss of hepatic p53 is associated with worsening liver disease that is characterised by liver damage, chronic inflammation, fibrosis, and high levels of ROS.In Trp53/TP53 wildtype murine and human HCC cell lines, we show that the antioxidant gene Tigar/TIGAR is induced in a p53-dependent manner after HFHS treatment, localises to the mitochondria, and supports redox control.In response to HFHS medium, loss of either p53 or TIGAR exacerbates redox stress and this can be rescued with N-acetyl-cysteine supplementation, confirming the role of TIGAR in supporting p53-mediated ROS control in this setting.Consistent with these observations, in the db/db genetic model of obesity, we show that Tigar-deficient mice exhibit evidence of enhanced liver disease, including increased ROS and liver damage from an early age that can also be rescued by N-acetylcysteine supplementation.Our findings support the conclusion that TIGAR-mediated redox control, likely alongside additional functions of p53, protects the liver against deleterious consequences of chronic consumption of a high fat and high sugar 'Western' diet.This protection is lost in the absence of liver p53 or TIGAR but can be partially restored with extended antioxidant treatment.
Our results add nuance to reports in mouse models and human patients where p53 induction and/or increased p53 pathway activity are associated with MASLD to MASH progression 20,21,36,61 .We also observe greater p53 activity in advanced vs early stages of diet-induced liver disease.Nevertheless, induction of p53 in our reporter mice is clearly a much earlier feature of adaptation to HFHS diet stress, at least in male mice, and occurs prior to evidence of, for example, significant glucose intolerance that would indicate advanced metabolic syndrome, T2DM, or the onset of MASH.Whether this is also true within female mice is not yet clear, and it would be interesting to examine whether noted sex differences in the p53 response that we observed reflect different underlying biology in future studies.Nevertheless, combined with earlier reports, our findings reported here suggest a dynamic interaction may exist between diet-induced stress and p53 activity.In this model, the modest activation of p53 we observe early in MASLD disease aetiology may be sufficient to coordinate redox protection, but this could eventually switch into a prodisease progression paradigm later when p53 signalling increases above a critical threshold and p53-mediated apoptosis, for example, is achieved 36 .Such a model would be consistent with more general literature on the threshold mechanisms inherent in determining the p53 response 62 .
Focusing on this point further, it is tempting to propose metabolic zonation as a potential mediator of the p53 response during MASLD to MASH progression.Diet-induced steatosis begins pericentrally, as we observed within our HFHS diet model and as is reported in human patients 63 .This region of the liver maintains low oxygen tension compared with periportal regions of the liver 63 .Even mild hypoxia is known to alter p53 activity, including changing the regulation of p53-mediated apoptosis and altering the p53 response to ROS 64,65 .One of the main features of MASLD progression to MASH is the overall expansion of steatosis from a zonally-restricted pericentral localisation to a global distribution, with lipid accumulation, inflammation, and fibrosis occurring across the liver lobule, including within more oxygenated areas of the liver where p53 activity is normalised.Future work could examine this possibility, for example by mapping the spatiotemporal distribution of p53 activation throughout MASLD progression to MASH.This would provide insight into the molecular mechanisms that underpin p53 activity during liver disease.
Taken together, our results underscore the importance of p53 and TIGAR for protecting the liver against damage associated with diet-induced liver disease.They also suggest that antioxidant interventions may have efficacy in preventing some of the deleterious aspects of MASH.Interestingly, considering that metabolic syndrome is a multi-organ condition, our findings also suggest that nuances of the p53 response to diet stress are disease-state specific and potentially not shared across all metabolic organs.

Mice
Procedures involving mice were performed under Home Office licence numbers 70/8645, 70/8468, PP6345023 and PP1389725 as previously described 50 .Experiments were conducted in accordance with the Animals (Scientific Procedures) Act 1986 and the EU Directive 2010 and were sanctioned by Local Ethical Review Process (University of Glasgow) and The Francis Crick Institute.Mice were housed on a 12-hour light/12-hour dark cycle in groups of 3-5 as much as possible.Mice were provided with environmental enrichment in the form of Sizzle-Nest bedding and polycarbonate tunnels, normal chow diet, and water ad libitum.Non-aversive handling techniques were utilised throughout each experiment.Mice were genotyped by Transnetyx (Cordova, TN).
Within experimental cohorts, mice were age and littermate-matched as much as possible, and all were started on experimental procedures and/or imaging prior to 6 months of age.In the imaging studies, a subset of the mice were sampled at either 100 days (N=5 male chow fed mice, N=3 male HFHS fed mice, N=3 chow fed female mice, or N=2 HFHS fed female mice) or 150 days into the experiment (N=3 HFHS fed male mice).In addition, two female mice on the HFHS diet developed skin conditions that required them to be euthanised.These animals were included until the point of euthanasia, and removed from the longitudinal data after these instances.
The presented weight measurements in Figure 1 were obtained from the reported imaging cohorts, but timepoints prior to 150 days reflect only a subset of these mice, as these data were not initially collected from the first groups of experimental animals.
Imaging cohorts (Fig. 1) included both male and female mice.These were either hemizygous (males) or homozygous (females) for the PG13-iRFP allele.Based on imaging results that indicated a preponderance of p53 activity in males during the response to HFHS diet, the functional consequences of liver Trp53 deletion were examined in only male cohorts (Fig. 2).Experiments involving Lep ob mice (Fig. 3) were undertaken in both male and female mice.Considering the rapid sex-independent lethality observed within Lep ob/ob ; Albumin-Cre; Trp53 FL/FL mice, experiments examining Tigardeficiency within the more aggressive Lepr db model were undertaken in only females, where the normal disease latency is significantly longer 59 .Downstream analyses were performed on a random order of samples by researchers blinded to the genotype and treatment regime of a given sample until the summation of results.

Long-term diet experiments
For long-term diet experiments, male 65-75 day old Trp53 wildtype; Albumin-Cre and Trp53 FL/FL ; Albumin-Cre mice were shifted onto either an obesogenic high-fat and highsugar (HFHS) diet or left on normal mouse chow (control diet).HFHS diet (AIN-76A, 58R3) was purchased from TestDiet and contained 59% energy from fat, 15% energy from protein, 25% energy from carbohydrates, was rich in sucrose, and did not contain cholesterol.The control normal mouse chow (DS801752G10R, Special Diet Services/SDS) contained 9% energy from fat, 21% energy from protein, 70% energy from carbohydrates and reduced sucrose.
For imaging long-term diet experiments, male and female 65-75-day old p53 reporter (PG13-iRFP) mice were all shifted onto a purified control maintenance diet suitable for imaging mice (TestDiet, AIN-93M, 58M1) rather than normal mouse chow.This diet change was done at least 7 days prior to imaging mice in order to eliminate non-specific fluorescence in the gut prior to obtaining baseline p53 levels.Mice in imaging experiments continued to be maintained on either control imaging diet or were shifted onto the HFHS diet for the duration of the experiment thereafter.As with the purified control maintenance diet, the HFHS diet was also confirmed to not cause non-specific fluorescence in the gut.

In vivo imaging
Mice were imaged longitudinally on a Pearl Impulse Small Animal Imaging System (LI-COR) as previously described 68 .For quantification, the PG13-iRFP signal registered within the liver region of each mouse was normalised to the individual initial baseline liver-region signal for each mouse, taken at the start of the experiment.For a small number of experimental mice, the observed iRFP signal of the initial baseline scan was either transiently very low (N=3/42) or very high (N=1/42) for one imaging session compared with subsequent scans.In these cases, the liver region signal observed one week later was instead used to define the baseline for longitudinal analysis.Scan images are presented with false colour LUTs.Imaging parameters were kept constant across time points and mice within an experiment.For quantification, Image Studio software (LI-COR, V5.5) was used.

Ex vivo imaging of tissues
Harvested tissues were fixed overnight in 10% neutral buffered formalin (Solmedia), assembled onto a 10 cm petri dish, and imaged on a Pearl Impulse Small Animal Imaging System (LI-COR) as for in vivo imaging described above.Imaging parameters were kept constant across samples and scans were analysed using Image Studio software (LI-COR, V5.5).Scan images are presented with false colour LUTs which were held constant across all images shown.

Oral Glucose Tolerance Test
Prior to the oral glucose tolerance test, mice were fasted for a period of 4 hours.Sterile glucose (60% w/v in water) was administered in a bolus of 3 g/kg via oral gavage per mouse, based on weights obtained at the start of the fasting period.Blood glucose levels were measured with an Accu-Chek blood glucose meter (Accu-Chek UK) using blood obtained from the tail vein before oral gavage (time 0), and at 15 mins, 30 mins, 45 mins, 60 mins, and 120 mins after administration of the glucose bolus.

Liver function tests
Alanine transaminase (ALT) and aspartate aminotransferase (AST) activity were measured in EDTA-treated plasma using kits from Abcam (ab105134 and ab138878) according to the manufacturer's instructions.Samples were run together, analysed in duplicate wells per mouse, and the mean value of these technical replicates was used for subsequent analysis.
The following antibodies were stained on a Leica Bond Rx autostainer: F4/80 (ab6640, Abcam) and malondialdehyde (MDA) (ab243066, Abcam).All FFPE sections underwent on-board dewaxing (AR9222, Leica) and epitope retrieval using appropriate retrieval solutions.Sections for F4/80 staining underwent epitope retrieval using enzyme 1 solution (AR9551, Leica) for 10 minutes at 37°C.Sections for MDA staining underwent epitope retrieval with ER2 solution (AR9640, Leica) for 20 minutes at 95°C.Sections were washed with Leica wash buffer (AR9590, Leica) before peroxidase block was performed using an Intense R kit (DS9263, Leica).MDA sections had an additional Mouse Ig blocking step (MKB-2213, Vector Labs) applied for 20 minutes.All sections were then rinsed with wash buffer before primary antibody application at the optimal dilution (F4/80: 1/200; MDA: 1/250).The sections were rinsed with wash buffer and appropriate secondary antibody applied for 30 minutes.MDA sections had Mouse Envision (K4001, Agilent) applied and sections for F4/80 had Rat ImmPRESS secondary solution (MP-7404, Vector Labs) applied.Sections were rinsed with wash buffer before being visualised using DAB and counterstained with haematoxylin in the Intense R kit.
H&E staining was performed on a Leica autostainer (ST5020).Sections were dewaxed in xylene, taken through graded ethanol solutions and stained with Haem Z (RBA-4201-00A, CellPath) for 13 mins.Sections were washed in water, differentiated in 1% acid alcohol, and washed.Nuclei were blued using Scott's Tap Water Substitute (in-house).After washing with tap water, sections were placed in Putt's Eosin (in-house) for 3 minutes to complete the staining.
Staining for PSR and ORO was performed as previously described 43 .
All stained sections except for those from experiments involving Lep ob mice (Fig. 3) were scanned at 20x magnification using a Leica Aperio AT2 slide scanner (Leica Microsystems, UK) prior to subsequent analysis of scanned slides as described below.

IHC and Staining Quantification
All quantification was undertaken on a random order of samples blinded to the genotype and treatment regime of a given sample until the summation of results.
Quantification for MDA, F4/80-positive lipogranulomas and PSR staining in Figs. 1, 2, and 5 were carried out using QuPath software (v0.5.1) 69 .The same parameters per stain were used for all samples.For PSR and MDA staining, random 20x magnification images were selected taking care to avoid any vessels where these stains were excessively high.The average percentage of positively stained area was calculated for each mouse.F4/80 staining attributable to hepatocyte-engulfing macrophages was scored by hand from five random 20x magnification images per slide in QuPath and averaged per mouse sample.Quantifications in Fig. 3 were performed as previously described 43 .

Cell culture
HepG2 (HB-8065), SK-Hep-1 (HTB-52), and HCT116 (CCL-247) cells were obtained from ATCC but were not authenticated for this study.Hep53.4 cells (Cellosaurus CVCL_5765) were obtained from Cytion (Product number 400200) but were not authenticated for this study.HepG2 TP53 knockout (p53 KO) cells were generated through CRISPR-mediated targeting of TP53 as previously described 70 .Mycoplasma testing was performed on all cells when they were thawed and semi-regularly thereafter using the MycoAlert PLUS Mycoplasma Detection Kit (Lonza LT07-703).Independent experiments were performed on cells treated with siRNA and/or compounds from separate passages of each cell line, with technical replicates used to confirm the effect of treatment conditions and/or siRNAmediated gene knockdown.Stock flasks were maintained in DMEM, low glucose, pyruvate, no glutamine, no phenol red (Gibco cat# 11880036) supplemented with 2 mM L-Glutamine (Gibco cat# 25030032), penicillin/streptomycin (Gibco cat# 15070063), and 5% FBS (Merck cat# F7524).Cells were cultured at 37°C in a humidified atmosphere of 5% CO2.

In vitro HFHS medium formulation
To create the HFHS medium formulation, a concentrated lipid mixture stock solution of fatty acids conjugated to fatty acid-free BSA (10% solution in PBS) (Merck cat# A1595) was created as previously described 71 .The stock solution was added at 1% v/v to Plasmax cell culture medium (CancerTools cat# 156371) supplemented with 5% FBS (Merck cat# F7524) and penicillin/streptomycin (Gibco cat# 15070063).The resulting high fat medium was further supplemented with an additional 20 mM D-Glucose (Merck cat# G7528) and 25mM D-Fructose (Thermo Scientific cat# A17718.30) to provide the high sugar component of the HFHS medium.The final concentrations of fatty acids used in the HFHS medium were: 150 µM palmitate (Merck cat# P5585), 150 µM stearic acid (Thermo Scientific cat# A12244.06), 25 µM linoleic acid (Merck cat# L9530), 2.8 µM linolenic acid (Thermo Scientific cat# 215040050), 0.14 µM oleic acid (Thermo Scientific cat# 270290050) and 2 µM arachidonic acid (Sigma, cat# 10931).These amounts were chosen to broadly match the lipid distribution found in the murine HFHS diet.For control-treated cells, Plasmax supplemented with 1% v/v of the 10% fatty acid-free BSA solution in PBS (Merck cat# A1595) supplemented with 10% Ethanol (Fisher Scientific cat# 10680993) (0.1% v/v final concentration) was used to mirror the ethanol used to dissolve fatty acids for BSA conjugation in HFHS medium formulation.
For experiments utilising HFHS medium, cells were plated at an assay-dependent dilution in the maintenance DMEM medium as described above.For flow cytometry and western blotting, 1.25*10 5 cells/well were seeded in 6-well plates (Thermo Scientific cat# 140675) and for mitochondrial enrichment, 3.0*10 6 cells were seeded in 150 mm dishes (Thermo Scientific cat# 168381).24-48 hours after seeding, cells were treated with either HFHS Plasmax or Plasmax supplemented with BSA (control) and left for a further period of 48 hours in treatment medium prior to analysis via Flow Cytometry, Western Blotting, or mitochondrial enrichment.In cells transfected with siRNA, reverse transfection was undertaken when cells were seeded in 6-well plates, and the transfection mixture was left on the cells until treatment medium was added for each condition.

Antioxidant treatment with N-acetyl-cysteine
For experiments involving the antioxidant N-Acetyl-cysteine (NAC), cells were treated with 2 mM NAC (Sigma cat# A7250) dissolved in PBS for a period of 48 hours.NAC treatment was added directly to the HFHS medium used for treated samples.

Flow Cytometry
Adherent cells were detached with Trypsin-EDTA solution (Merck cat# T4174), dissociated into single cells, and resuspended in phenol red-free DMEM (Gibco cat# 11880036).Cells were labelled with CellROX Deep Red reagent (7.5 μM, Thermo Fisher Scientific cat# C10422) and BODIPY 493/503 (5 μM, Thermo Fisher Scientific cat# D3922) for 10 minutes at room temperature in a 1:1 mixture of phenol red-free DMEM and a staining solution containing the dyes in PBS without calcium and magnesium (Corning cat# 21-040-CV) supplemented with 2.5% w/v BSA (Merck cat# 810533).4′,6-Diamidino-2phenylindole dihydrochloride (DAPI, Merck cat# D9542) was added to a final concentration of 1 μg/ml to each sample and was used to identify viable cells for analysis.Single cells were analysed on a MACSQuant Analyzer 10 (Miltenyi Biotec) using unstained and hydrogen peroxide-treated cells as controls.At least 10,000 events were collected for each sample.Data were analysed using FlowJo 10.10.0 (BD Biosciences).Unless otherwise stated, median fluorescence intensity values were obtained and compared across samples.

Mitochondrial Enrichment
The isolation of cytoplasmic and mitochondrial-enriched fractions from cells was performed as described previously 72 .Briefly, cells were scraped in isotonic HIM buffer (200 mM mannitol, 1 mM EGTA, 70 mM sucrose, 10 mM HEPES), passed through a cell homogeniser with a 16 µm bead and pelleted by centrifugation at various speeds to collect individual enriched fractions.The cytoplasmic fraction was retain in HIM buffer and the mitochondrial-enriched pellet was resuspended in NP-40 lysis buffer (Fisher Scientific cat# 15403489) supplemented with PhosSTOP (Merck cat# 4906845001) and the cOmplete Ultra EDTA-free protease inhibitor cocktail (Merck cat# 05892791001).

Western Blotting
Western blotting was performed as previously described 70 with the following modifications: Protein lysates were prepared using NP-40 lysis buffer as in mitochondrial enrichment experiments.Protein concentration was determined in each sample using a Pierce BCA Protein Assay kit (cat# 23227), equivalent concentrations were loaded in each well, and total protein was confirmed following transfer using a Pierce Reversible Protein Stain Kit (Fisher cat# PI24580).Blocking was performed in a TBS solution containing 5% BSA (Merck cat# 810533) and 0.1% Tween-20 (Merck cat# P1379), and wash steps were performed in TBS with 0.1% Tween-20 (TBS-T).
Primary and Secondary antibodies were used as indicated in the tables below.Proteins were detected using either an Odyssey DLx or Odyssey XF Imaging System (LI-COR Biosciences) and analysed using ImageStudio 5.5 software (LI-COR Biosciences).

Figure 4 :
Figure 4: p53 engages TIGAR and supports redox control in response to overfeeding in vitro

Liver p53 is engaged in response to an obesogenic high fat high sugar diet
43antitative RT-PCR analysis of HepG2 cells was undertaken as previously described43, using Taqman Fast Advanced Master Mix and the Taqman Gene Expression Assays indicated in the table below.Data were plotted using GraphPad Prism 10 (GraphPad).The statistical analysis for each experiment was performed using the test indicated in the relevant figure legend and multiplicity-adjusted p-values using the built-in analysis tools of Prism 10.Statistical tests were chosen based on standard tests utilised in the field and the nature of the comparison being made.Underlying assumptions for these tests were assumed to be met although not explicitly examined.Figures were prepared using Adobe Illustrator 2024 (Adobe).Data are presented as mean with min/max floating bars and individual data points (Figs.1 and S1) and mean +/-SEM with individual data points (Fig.2 onwards).P-values are denoted as follows: ns-not significant, *p<0.05,**p<0.01,***p<0.001,****p<0.0001.