Weight Loss During Obesity Provokes Asparaginase‐Associated Liver Steatosis and Endoplasmic Reticulum Stress

ABSTRACT

Asparaginase is an anti‐leukemic agent that triggers severe adverse metabolic events. Obesity is a known risk factor for asparaginase‐associated liver steatosis. To better understand why, we first compared the liver metabolome of lean versus diet‐induced obese (DIO) mice exposed to native asparaginase and observed a substantially altered liver metabolome in DIO mice only. To explore the basis for the altered liver metabolome in DIO mice, we designed experiments to clarify the relative contributions of obesity versus feeding excessive fat during asparaginase on liver triglycerides. Lean mice and DIO mice were fed a high‐fat, obesogenic diet (OD) or low fat, maintenance diet (MD) during exposure to pegylated (PEG)‐asparaginase. In lean mice, feeding OD during PEG‐asparaginase modestly (2‐fold) increased liver steatosis. Obese mice fed OD during PEG‐asparaginase showed the lowest food intake alongside the lowest liver triglyceride secretion rates, resulting in the largest (6‐fold) increase in liver triglycerides and emergent endoplasmic reticulum (ER) stress. Switching obese mice to a MD during PEG‐asparaginase did not rescue liver steatosis nor alleviate ER stress. In a separate study, DIO mice globally lacking albumin (AlbKO) were fed OD during exposure to PEG‐asparaginase to examine if loss of the major plasma free fatty acid carrier could lessen liver steatosis, but loss of circulating albumin did not mitigate elevated liver triglycerides. In total, the results revealed that body weight loss enables asparaginase‐associated liver steatosis and ER stress. Mitigating asparaginase‐induced weight loss may be a meaningful strategy in preventing liver stress during treatment.

Lean mice and DIO mice were fed either a low‐fat maintenance diet (MD) or a high‐fat obesogenic diet (OD) during pegylated asparaginase exposure. OD exacerbated asparaginase‐induced weight loss and modestly increased liver triglycerides and ER stress in lean mice. DIO mice, regardless of diet, lost significant weight and had severe hepatic steatosis and ER stress from asparaginase. Genetic deletion of albumin in mice reduced plasma free fatty acids (FFA), glycerol, and triglycerides but did not protect against asparaginase‐associated liver steatosis.

1. Introduction

Asparaginase is used in the treatment of acute lymphoblastic leukemia (ALL), the most common childhood cancer [1]. Asparaginase depletes circulating asparagine, an essential nutrient for leukemic lymphoblasts. The metabolic toxicity profile of asparaginase includes liver steatosis and dysfunction which can be difficult to manage [2, 3, 4, 5, 6]. A pegylated form (PEG‐asparaginase) lowers rates of anaphylaxis but not the spectrum of adverse metabolic events [2, 7, 8, 9, 10]. Critically, obese and overweight patients are at a significantly increased risk of developing liver steatosis during treatment with pegylated (PEG)‐asparaginase [11]. Furthermore, diets varying in fat content during treatment with asparaginase affect survival in a mouse model of ALL [12].

Previously we published that mice with diet‐induced obesity lose approximately 20% of their body weight and develop severe fatty liver and endoplasmic reticulum (ER) stress after 8 days of exposure to native asparaginase [13]. Our group and others have opined that asparaginase‐induced liver steatosis and ER stress are due to an accumulation of bioactive lipid molecules in the liver in combination with an influx of free fatty acids (FFAs) derived from adipose tissue lipolysis [13, 14]. As such, there is a need to understand how body fat versus dietary fat promotes asparaginase‐induced weight loss and liver stress. Reconciling this may reveal new options for preventing or treating steatosis and other adverse metabolic events related to liver health.

The objective of this study was two‐fold: first, to investigate the contribution of dietary fat content to the development of PEG‐asparaginase‐induced liver steatosis and ER stress, and second, to further elucidate the mechanism by which liver steatosis develops during exposure to PEG‐asparaginase. Our results support a model in which obesity promotes liver steatosis and ER stress during asparaginase exposure via an influx of adipose‐derived FFAs due to weight loss in combination with an impaired ability to secrete triglycerides from the liver. Loss of circulating albumin does not mitigate this outcome. We conclude that preventing weight loss during PEG‐asparaginase exposure may lessen asparaginase‐associated liver steatosis and stress, especially in obesity and overweight.

2. Materials and Methods

2.1. Animals and Experimental Design

All experiments were approved by the Rutgers University Institutional Animal Care and Use committee and in accordance with the ARRIVE 2.0 guidelines [15]. All mice were nongrowing adult C57BL/6J mice aged between 20 to 35 weeks and were housed in transparent, plastic shoebox cages with soft bedding and enrichment in a facility that was humidity‐controlled (40%–60%) and temperature‐controlled (23°C) with a 12 h:12 h light:dark cycle. Mice were bred on site and maintained with free access to purified water and food. Unless otherwise stated, injections of asparaginase or phosphate buffered saline (PBS) were administered at 10 am and mice were killed by decapitation between 3:00 to 4:00 p.m., tissues were rapidly dissected on ice and snap frozen in liquid nitrogen. Trunk blood was collected, centrifuged at 10 000× g at 4°C for 10 min, and serum collected. All tissues were stored at −80°C. An animal MRI (EchoMRI, Houston, TX) was used to measure body composition.

2.2. Experiment 1—Untargeted Metabolomics

Mice (n = 6 per group with n = 3 females and n = 3 males) were fed either a laboratory rodent diet (5001; LabDiet) or a 60% fat by kcal purified diet (Table S1) to induce obesity and then administered 8 daily injections of native asparaginase (Elspar) at 3 international units per gram body weight (IU/g BW) or PBS as previously described [13].

2.3. Experiment 2—Maintenance Diet Comparisons

A comparison between two low fat diets (Table S1) during exposure to PEG‐asparaginase or PBS was conducted in four cohorts of mice. A first cohort of lean mice (5 male, 1 female) was provided a purified low fat diet (LFD) and then a single intraperitoneal (i.p) injection of either PEG‐asparaginase (Oncaspar, 1.5 IU/g BW) (n = 3) or equivolume PBS (n = 3). A second cohort of lean mice (9 male, 9 female) was provided a laboratory rodent diet (Chow) (5001; LabDiet) while exposed to either PEG‐asparaginase (n = 9) or PBS (n = 9). A third cohort of diet‐induced obese mice (4 male, 2 female) was switched from a purified obesogenic diet (OD) to a LFD 1 day prior to single administration of PEG‐asparaginase (n = 3) or PBS (n = 3). A fourth cohort of diet‐induced obese mice (6 male, 8 female) was switched to Chow 1 day prior to a single administration of PEG‐asparaginase (n = 7) or PBS (n = 7). Results from these experiments validated combining Chow and LFD groups into a single low‐fat maintenance diet (MD) treatment group.

2.4. Experiment 3—Impact of OD in Lean Mice Exposed to PEG‐Asparaginase

Lean mice (14 male, 10 female) were either maintained on MD or switched to an OD (Table S1) 1 day prior to a single i.p. injection of either PEG‐asparaginase (Oncaspar) at 1.5 IU/g BW or equivolume PBS (n = 4–8 per sex per group). The study ended 5 days after injection.

2.5. Experiment 4—Impact of MD in Diet‐Induced Obese Mice Exposed to PEG‐Asparaginase

Lean mice (16 male, 18 female) were fed an OD for 15 weeks to induce obesity then switched to a MD (n = 20) or remained on an OD (n = 14) and underwent the same study protocol as in Experiment 3.

2.6. Experiment 5—Triglyceride Secretion Rate

Ten lean (6 male, 4 female), and eleven obese (5 male, 6 female) mice were assessed for liver triglyceride secretion rate as described previously [16]. Specifically, mice were injected once with either PEG‐asparaginase (Oncaspar) (1.5 IU/g BW) or PBS and then on Day 4, mice were placed in wire bottom cages and food was removed at 7:00 p.m., at 7:00 am the next day, mice were injected with P‐407 (500 mg/kg) and blood was sampled via tail snip for triglyceride concentrations at baseline and every 30 min post‐injection for a total of up to 180 min. Blood triglyceride measurements were taken using a handheld meter and test strips (Polymer Technology Systems; CardioChek, Indianapolis, IN). Mice were euthanized by CO₂ inhalation.

2.7. Experiment 6—Albumin KO Mice

Weanling wild‐type mice (4 male, 4 female) and albumin knockout mice (Alb ^−/−) on the C57BL6/J genetic background [17] (6 male, 6 female) were fed an OD for 12 weeks to induce obesity. On the first day of the experiment, body composition was measured and mice were i.p. injected either PEG‐asparaginase (Oncaspar) at 1.5 IU/g BW or an equivalent volume of PBS. Five days later, body composition was analyzed by MRI and mice were euthanized via CO₂ inhalation.

2.8. Metabolomics

Untargeted metabolomics were performed by Metabolon Inc., (Durham, NC, USA) to identify 663 compounds of known identity. Peaks derived from Ultrahigh Performance Liquid Chromatography‐Tandem Mass Spectroscopy were quantified using peak area, normalized for inter‐day variation, and expressed relative to the median of each metabolite (median equals 1.00).

2.9. SDS‐PAGE and Immunoblotting

SDS‐PAGE and immunoblotting were performed as previously described [18]. All antibodies used are found in Table S2.

2.10. RNA Isolation and Real‐Time Quantitative PCR

RNA was isolated from frozen liver tissue using TRI Reagent (TR 119, Molecular Research Center Inc.). RT‐qPCR was performed using the SYBR green system as previously described [19]. Fold change was calculated using the delta–delta Ct method. All primer sequences can be found in Table S3.

2.11. Liver Triglyceride Measurement

Approximately 40 mg of frozen, powdered liver tissue was homogenized in 10% (v/v) NP‐40 (J619; Amresco) using a handheld, motorized pestle. Triglyceride content was determined using a commercially available kit Triglyceride Assay Kit‐ Quantification (ab65336; Abcam) according to the manufacturer's instructions.

2.12. Plasma Lipid and Protein Measurements

Plasma lipids were measured using commercially available kits and performed according to the manufacturer's instructions: FFA (299–94 301, FUJIFLM Wako Pure Chemical Coporation), glycerol (F6428, Sigma‐Aldrich), triglyceride (T2449, Sigma‐Aldrich). Plasma protein was quantified using a Bradford assay (#5000006, Bio‐Rad).

2.13. Statistical Analyses

Statistical tests and graphs were performed and generated in GraphPad Prism. Where stated, statistical comparisons used a two‐way ANOVA with diet and drug as independent variables or a two‐way ANOVA with obesity status and drug as independent variables. A two‐way ANOVA allowed us to assess whether diet or PEG‐asparaginase independently influenced a given phenotype (defined as a Main Effect), or whether the effect of PEG‐asparaginase differed by diet or genotype or obesity status (defined as an Interaction Effect). A Shapiro–Wilk test was conducted to test for normality. If data were not normally distributed, values were log₁₀ transformed prior to conducting the ANOVA to approach normality. If a significant main effect (diet, drug, obesity, genotype) or significant interaction effect (diet × drug, or drug × obesity, or diet × obesity, or drug × genotype) was detected by the ANOVA, a post hoc test was done to pinpoint exactly where the differences between treatment groups lie. We selected the Tukey Honestly Significant Difference post hoc method to reduce the chance of a Type I error when performing multiple pairwise t‐tests. Results were deemed statistically significant if the adjusted p < 0.05. All bar graphs represent means +/− SD. Box and whisker plots represent the group median, the upper and lower edges of the box represent the 75th and 25th quartiles, respectively, and the top and bottom of the whiskers represent the maximum and minimum values, respectively.

3. Results

3.1. Asparaginase Induces a Unique Metabolic Signature in the Livers of Obese Mice

To discern the effect of obesity on the liver metabolome during asparaginase exposure, we performed untargeted metabolomics in lean versus diet‐induced obese mice. Asparaginase induced minimal changes to the liver metabolomes of lean mice fed MD, reducing only asparagine, N‐acetylasparagine, guanidinosuccinate, and 1‐(10enyl‐palmitoyl)‐2‐arachidonoyl‐GPE (P‐16:0/20:4) (Figure 1A and Table S4). In contrast to lean mice, asparaginase significantly altered 260 metabolites in the livers of diet‐induced obese mice consuming an OD (Figure 1B and Table S5). Of these 260 metabolites, 47 were increased upon asparaginase exposure with the following categories most impacted: lipid/phospholipid metabolism (phosphoethanolamine, 1‐oleoly‐GPG, 1‐linoleoyl‐GPG), TCA cycle (citrate, aconitate, succinate), and bile acid and steroid metabolism (glycocholate, 7‐hydrocholesterol, 4‐cholesten‐3‐one). Conversely, the 217 metabolites that were downregulated upon asparaginase exposure fell into the following categories: polyamine (spermine, spermidine, guanidinosuccinate), nucleotide metabolism, (adenosine, ADP, AMP, uracil), and amino acid metabolism (leucine, asparagine, valine). The list of metabolites that were significantly altered by obesity alone (i.e., comparing lean mice injected with PBS to obese mice injected with PBS) represented glycerolipid, carbohydrate, and amino acid metabolism consistent with published literature [20] (Figure 1C and Table S6). Venny analysis revealed these changes overlapped ~20% with asparaginase‐induced changes in obese mice [21], indicating the metabolic response to asparaginase differs from obesity alone (Figure 1D and Table S7). Overall, these data show that the combination of obesity plus a high fat diet significantly changes the metabolic response to asparaginase in the liver. However, the relative contribution of diet versus obesity to perturbed liver metabolism could not be determined. To further interrogate this point, we designed additional experiments to isolate the impact of diet on liver triglyceride content during asparaginase exposure.

FIGURE 1.

Asparaginase induces a unique metabolic signature in the livers of obese mice. (A) Volcano plot illustrating significantly altered liver metabolites when comparing lean mice exposed to native asparaginase (ASNase) or PBS as vehicle control (B) Volcano plot illustrating significantly altered liver metabolites when comparing obese mice exposed to native asparaginase (ASNase) or PBS as vehicle control (C) Volcano plot illustrating significantly altered liver metabolites when comparing obese and lean PBS injected mice. X‐axis shows the log 2‐fold change to quantify the change in metabolite levels with a negative value representing a decrease and a positive value representing an increase in the ASNase exposed group relative to the phosphate buffered saline (PBS) injected group. Y‐axis shows the logarithm base 10 of the FDR and the horizontal gray line represents the statistical cut off. Dots in red are individual metabolites that are statistically different between treatments, FDR ≤ 0.05. (D) Venn diagram comparing significantly altered liver metabolites shown in B and C.

3.2. Acute Feeding of an Obesogenic Diet Modestly Amplifies Weight Loss and Liver Steatosis During Exposure to PEG‐Asparaginase

Pilot experiments first compared low‐fat commercial chow to low‐fat purified diet as the control maintenance diet (MD) (Figure S1). In these experiments, diet choice did not alter body weight loss (Figure S1A), liver triglyceride content (Figure S1B), or markers of endoplasmic reticulum (ER) stress (Figure S1C,D) following PEG‐asparaginase. We then assessed the contribution of dietary energy density to the physiological responses to PEG‐asparaginase by maintaining lean mice on a MD or switching to an OD during exposure to PEG‐asparaginase or PBS (Figure 2A). Body weight, lean mass, and fat mass were similar between groups at the start of the study (Figure S2A,B,C). Similar to our previous report, PEG‐asparaginase induced significant weight loss in mice fed a MD (Figure 2B) [18]. Interestingly, PEG‐asparaginase induced weight loss was greater in mice fed an OD (Figure 2B). PEG‐asparaginase reduced lean mass independent of diet (Figure 2C) but mice fed OD lost more body fat as compared to mice fed MD (Figure 2D). Measurement of food intake showed acute feeding of an OD exaggerated the anorexic effect of PEG‐asparaginase, leading to the lowest amount of food intake (Figure 2E). The livers of lean mice fed OD were significantly smaller than lean mice fed MD regardless of PEG‐asparaginase exposure (Figure 2F). Quantification of liver TG content showed PEG‐asparaginase increased lipid abundance overall, with a greater effect observed in OD fed mice (Figure 2G). In summary, acute consumption of an OD during exposure to PEG‐asparaginase decreases food intake and body weight and increases liver TG content.

FIGURE 2.

Acute consumption of an obesogenic diet during exposure to PEG‐asparaginase modestly induces liver steatosis in lean mice. (A) Schematic of experimental design for lean mice administered PEG‐asparaginase or PBS and fed at MD or OC diet. Created using Biorender.com. Change in body weight (B), lean mass (C), and fat mass (D) from Day 0 to Day 5 as determined by MRI. (E) Cumulative food intake from Day 0 to Day 5. (F) Liver weight at euthanasia represented as percent of body weight. (G) Quantification of hepatic triglyceride content (H) Representative immunoblots, and (I) subsequent quantification of p‐T980‐PERK normalized to total PERK represented as a fold change to PBS‐MD mice. Box plots show median values, top and bottom hinges refer to the first and third quartiles (25th and 75th percentiles), and the ends of the whiskers mark the smallest and largest values. Bar plots represent group mean ± SD. = females. = males. Hollow points in the “MD” group represent mice fed a LFD. Solid points in the “MD” group represent mice fed chow. Main effects of an independent variable after a two‐way ANOVA (diet, drug) are indicated on graph. If a statistical interaction occurred, groups not sharing a common letter indicated a statistically significant difference between groups after post hoc pairwise comparisons were conducted with a Tukey correction for multiple comparisons. n = 3–4 per sex per group.

To investigate the role of diet in the development of hepatic ER stress during exposure to PEG‐asparaginase, the phosphorylation status of PERK, the transcript expression of the pro‐apoptotic Ddit3, and splicing of Xbp1 were measured. Phosphorylation of PERK did not change across groups, although there was a trend toward an increase in PERK phosphorylation in the livers of the PEG‐asparaginase exposed mice fed OD (Figure 2H,I). PEG‐asparaginase increased eIF2α phosphorylation in MD‐ and OD‐fed groups similarly (Figure S3A–B). In agreement with the pattern of PERK phosphorylation, the hepatic transcript abundance of Ddit3 and Xbp1 splicing were highest in asparaginase exposed OD‐fed mice (Figure S3C–F). In summary, consumption of an OD during exposure to PEG‐asparaginase induced mild liver steatosis and modestly increased biomarkers of hepatic ER stress.

3.3. Acute Feeding of an Obesogenic Diet Does Not Change Hepatic Expression of Genes Involved in β‐Oxidation or de Novo Lipogenesis

To investigate mechanisms underlying the PEG‐asparaginase‐induced increase in liver TG content during OD feeding, mRNA expression was measured in the liver of genes which regulate β‐oxidation and de novo lipogenesis. PEG‐asparaginase did not change transcript abundance of peroxisome proliferator‐activated receptor α (Pparα) or acyl‐CoA oxidase 1 (Acox1) (Figure 3A,B). However, PEG‐asparaginase significantly increased expression of carnitine palmoyltransferase 1 (Cpt1a) (Figure 3C). Hepatic mRNA expression of stearoyl‐desaturase 1 (Scd1) and fatty acid synthase (Fasn) was downregulated by PEG‐asparaginase (Figure 3E). In mice consuming the MD, PEG‐asparaginase decreased hepatic acetyl‐CoA carboxylase 1 (Acc1), but this was not observed in mice fed the OD (Figure 3F). These data indicate that mice consuming an OD do not show signs of increased de novo lipogenesis or decreased β‐oxidation during asparaginase exposure.

FIGURE 3.

Acute feeding of an obesogenic diet to lean mice during PEG‐asparaginase exposure does not uniquely alter hepatic expression of lipid metabolism genes. Hepatic mRNA abundance of Pparα (A), Acox1 (B), Cpt1α (C), Scd1 (D), Fasn (E), and Acc (F) in lean mice exposed to PEG‐asparaginase or PBS control represented as a fold change to PBS‐MD mice. Box plots show median values, top and bottom hinges refer to the first and third quartiles (25th and 75th percentiles), and the ends of the whiskers mark the smallest and largest values. = females. = males. Main effects of an independent variable after a two‐way ANOVA (diet, drug) are indicated on graph. n.s., no significant effects. If a statistical interaction occurred, groups not sharing a common letter indicated a statistically significant difference.

3.4. Acute Feeding of a Low‐Fat Maintenance Diet Is Not Sufficient to Protect Obese Mice From Severe PEG‐Asparaginase‐Induced Liver Steatosis

To assess the contribution of diet to hepatic stress in obese mice during exposure to PEG‐asparaginase, obese mice remained on OD or switched to a MD during exposure to PEG‐asparaginase or PBS (Figure 4A). Prior to the diet switch, there were no differences in starting body weight, body lean mass, or body fat mass between the groups (Figure S4A–C). PEG‐asparaginase induced significant losses in body weight, lean mass, and fat mass when compared to PBS injected mice, regardless of diet (Figure 4B–D). Transition to a MD in obese mice reduced body weight and fat mass (Figure 4B&D). Separate experiments confirmed MD diet choice did not alter body weight loss (Figure S1E), liver triglyceride content (Figure S1F), or hepatic PERK phosphorylation (Figure S1G,H) in mice exposed to PEG‐asparaginase. PEG‐asparaginase provoked anorexia in obese mice (Figure 4E). Obese mice that were switched to the MD ate fewer calories during exposure to PEG‐asparaginase (Figure 4E). PEG‐asparaginase increased liver weight, independent of diet (Figure 4F). Quantification of liver TG content showed PEG‐asparaginase induced severe liver steatosis in obese mice that was not rescued by switching to an MD (Figure 4G). Similarly, phosphorylation of PERK in the liver was substantially elevated by PEG‐asparaginase in obese mice with no significant effect of diet (Figure 4H,I). PEG‐asparaginase increased eIF2α phosphorylation and Ddit3 mRNA expression, independent of diet (Figure S5A–C). PEG‐asparaginase did not change hepatic mRNA expression of spliced Xbp1 (Figure S5D) but decreased total Xbp1 mRNA (Figure S5E). These effects resulted in an increased ratio of spliced/unspliced Xbp1 mRNA in the liver (Figure S5F). In summary, a switch to a low‐fat MD was not sufficient to protect obese mice from PEG‐asparaginase‐associated liver steatosis and ER stress.

FIGURE 4.

Feeding obese mice a low‐fat maintenance diet does not rescue asparaginase‐induced liver steatosis or ER stress. (A) Schematic of experimental design for obese mice administered PEG‐asparaginase or PBS and fed at MD or OC diet for 6 days. Created using Biorender.com. Change in body weight (B), lean mass (C), and fat mass (D) from Day 0 to Day 5 as determined by MRI. (E) Cumulative food intake from Day 0 to Day 5. (F) Liver weight at euthanasia represented as percent of body weight. (G) Quantification of hepatic triglyceride content. (H) Representative immunoblots and (I) subsequent quantification of p‐T980‐PERK normalized to total PERK represented as a fold change to PBS‐MD mice. Box plots show median values, top and bottom hinges refer to the first and third quartiles (25th and 75th percentiles), and the ends of the whiskers mark the smallest and largest values. Bar plots represent group mean ± SD. = females. = males. Hollow points in the “MD” group represent mice fed a LFD. Solid points in the “MD” group represent mice fed chow. Main effects of an independent variable after a two‐way ANOVA (diet, drug) are indicated on graph. If a statistical interaction occurred, groups not sharing a common letter indicated a statistically significant difference between groups after post hoc pairwise comparisons were conducted with a Tukey correction for multiple comparisons. n = 3–4 per sex per group.

3.5. Acute Feeding of an Obesogenic Diet Does Not Change Hepatic Expression of Genes Involved in β‐Oxidation or de Novo Lipogenesis

To further interrogate the mechanisms underlying PEG‐asparaginase‐induced liver steatosis in obese mice, mRNA expression levels of genes which regulate β‐oxidation and de novo lipogenesis were measured in the liver. PEG‐asparaginase lowered hepatic transcript abundance of Pparα and Acox1 (Figure 5A,B). However, PEG‐asparaginase significantly increased expression of Cpt1a (Figure 5C). Hepatic mRNA expression of Scd1 and Fasn was downregulated by PEG‐asparaginase (Figure 5D,E). There were no effects of diet or PEG‐asparaginase exposure on expression of Acc1 (Figure 5F). Thus, PEG‐asparaginase decreased expression of some genes involved in β‐oxidation while genes involved in de novo lipogenesis were either lowered or unchanged in obese mice.

FIGURE 5.

Feeding obese mice a low‐fat maintenance diet does not modulate PEG‐asparaginase‐induced changes in hepatic expression of genes involved in lipid homeostasis. Hepatic mRNA abundance of Pparα (A), Acox1 (B), Cpt1α (C), Scd1 (D), Fasn (E), and Acc (F) in lean mice exposed to PEG‐asparaginase or PBS control represented as a fold change to PBS‐MD mice. Box plots show median values, top and bottom hinges refer to the first and third quartiles (25th and 75th percentiles), and the ends of the whiskers mark the smallest and largest values. = females. = males. Main effects of an independent variable after a two‐way ANOVA (diet, drug) are indicated on graph. n.s., no significant effects. If a statistical interaction occurred, groups not sharing a common letter indicated a statistically significant difference between groups after post hoc pairwise comparisons were conducted with a Tukey correction for multiple comparisons. n = 3–4 per sex per group.

3.6. Obesity Decreases Liver Triglyceride Secretion Rate

To examine if PEG‐asparaginase compromises the liver's ability to export lipids, mice were injected with P‐407, an inhibitor of peripheral TG uptake, after 5 days of PEG‐asparaginase or PBS. Obese mice showed significantly lower rates of liver TG secretion overall (Figure 6A), indicating that obese mice have a compromised ability to export TGs from the liver. Notably, obese mice receiving asparaginase showed a statistical trend toward the lowest TG secretion (Obesity × Drug Interaction (p = 0.0857)). Our lab previously showed that in response to native asparaginase, the loss of GCN2 in obese mice decreases hepatic protein abundance of ApoB100, a lipoprotein essential for VLDL production and secretion [13]. Given the decrease in TG secretion observed in obesity, ApoB100 levels were measured. Interestingly, hepatic ApoB100 and ApoB48 protein abundances were elevated in obese mice relative to their lean counterparts (Figure 6B–D). PEG‐asparaginase had no significant effects on ApoB100 or ApoB48 protein levels (Figure 6B–D). These findings indicate obese mice have a compromised capacity to export TGs from the liver despite elevated hepatic ApoB100 levels.

FIGURE 6.

Obese mice have impaired liver TG secretion despite elevated ApoB100 abundance in the liver. (A) TG secretion rates from the liver as determined by the slope of TG increase in the blood after administration of the LPL inhibitor, P‐407. Main effects of independent variables (obesity status, drug) were determined by a two‐way ANOVA. n = 4–6 per group. (B) Representative immunoblot of ApoB100 and ApoB48 proteins and subsequent quantification of each normalized to total protein, expressed as fold change of lean, MD, PBS mice. n = 3–4 per sex per group. Box plots show median values, top and bottom hinges refer to the first and third quartiles (25th and 75th percentiles), and the ends of the whiskers mark the smallest and largest values. = females. = males. Main effects of independent variables (obesity status, diet, drug) were determined by a three‐way ANOVA.

3.7. Loss of Albumin Does Not Protect Mice From Asparaginase‐Induced Liver Steatosis During Obesity

We then hypothesized that loss of albumin, the major plasma FFA carrier [22, 23], would protect obese mice from liver steatosis during asparaginase exposure. Previous reports from our group showed Alb ^−/− mice display reduced plasma FFA and are also protected from liver steatosis during diet‐induced obesity [24, 25]. Therefore, we fed both wild‐type mice and Alb ^−/− mice an OD to induce obesity then injected mice with PEG‐asparaginase or PBS (Figure 7A). Systemic deletion of Alb was confirmed via serum analysis (Figure S6A,B). PEG‐asparaginase reduced body weight, lean mass, and fat mass in both WT and Alb ^−/− mice (Figure 7B–D). Notably, Alb ^−/− mice had an overall significant increase in fat mass relative to wild‐type mice (Figure 7D). There were no changes in liver weight between groups (Figure 7E). Surprisingly, both WT mice and Alb ^−/− mice showed increased liver TG levels (Figure 7F). Moreover, PEG‐asparaginase reduced plasma FFA, glycerol, and triglycerides (Figure 7G–I). Alb ^−/− mice, similar to our previous report [25], had a significant reduction in plasma FFA and glycerol relative to WT mice (Figure 7G,H). Alb ^−/− mice also had a significant reduction in circulating plasma proteins which was exacerbated by PEG‐asparaginase (Figure S6C). Overall, these results indicate that albumin is dispensable for PEG‐asparaginase‐induced liver steatosis.

FIGURE 7.

Loss of albumin does not protect against asparaginase‐induced liver steatosis. (A) Schematic of experimental design for WT or Alb ^−/− mice administered PEG‐asparaginase or PBS. Created using Biorender.com. Change in (B) body weight (C) lean mass (D) fat mass. (E) Liver weight as percentage of body weight at euthanasia. (F) Quantification of liver triglyceride content. Plasma levels of (G) free fatty acids (FFA) (H) glycerol and (I) triglycerides. Box plots show median values, top and bottom hinges refer to the first and third quartiles (25th and 75th percentiles), and the ends of the whiskers mark the smallest and largest values. Bar plots represent group mean ± SD. = females. = males. Main effects of an independent variable after a two‐way ANOVA (drug, genotype) are indicated on graph. If a statistical interaction occurred, groups not sharing a common letter indicated a statistically significant difference between groups after post hoc pairwise comparisons were conducted with a Tukey correction for multiple comparisons. Liver triglyceride content was analyzed by two independent t‐tests comparing each PBS group to PEG‐asparaginase group within each genotype. n = 2–3 per sex per group.

4. Discussion

Obesity is a risk factor for liver steatosis and toxicity in patients with ALL receiving asparaginase [2, 5, 6]. Obese mice exposed to native asparaginase develop severe fatty liver and ER stress [13] and modulation of dietary fat content alters survival in mouse models of ALL [12]. Based on these observations, we wondered to what extent diet affects liver outcomes during PEG‐asparaginase? Our results showed that a high fat diet does not directly provoke liver steatosis and ER stress during exposure to PEG‐asparaginase. Rather, it is weight loss that features prominently and corresponds with an accumulation of lipid species in the liver. Obesity further contributes to liver steatosis by reducing TG secretion from the liver. Changes in liver β‐oxidation, fatty acid synthesis, and lipid export do not fully account for the development of fatty liver in this model. Despite lower plasma FFA, obese albumin knockout mice were not protected from PEG‐asparaginase‐associated liver steatosis. In sum, our results support a model in which PEG‐asparaginase‐associated weight loss provokes a substantial influx of adipose tissue‐derived FFAs to the liver [18], in a manner that does not require albumin. Obesity further interferes with efficient liver TG export, exacerbating PEG‐associated liver steatosis and ER stress.

The mechanism by which PEG‐asparaginase reduces body weight is not fully clear. Recently we reported that PEG‐asparaginase substantially reduces food intake alongside modest increases in energy expenditure [18]. The reduction in food intake was observed in parallel with increased circulating growth differentiation factor 15 (GDF15), a target gene of the integrated stress response. GDF15 is increased in mice during chronic high fat feeding and in response to feeding an amino acid imbalanced diet [26]. Administration of halofuginone, an anticoccidial agent that mimics amino acid depletion by inhibiting prolyl tRNA charging, reduces food intake by a GDF15‐dependent mechanism [27]. Considering the recent emergence of GDF15 blockade drugs for use in cancer therapy, future experiments directly testing the role of GDF15 in managing PEG‐asparaginase‐associated weight loss are warranted.

Our study findings are supported by Kumar and colleagues who show that defects in β‐oxidation or increases in de novo lipogenesis do not account for increased liver TG abundance in mice exposed to PEG‐asparaginase [14]. We also note that obese mice had lower levels of TG secretion than lean mice, a finding that differs from models of metabolic dysfunction‐associated steatotic liver disease (MASLD) which show an increase in TG secretion is a consequence of increased hepatic lipid abundance [28, 29]. Limited data from other studies in C57BL/6J mice show obesity causes a decrease in TG secretion from the intestines after a high‐fat meal [30]. Perhaps the acute, rapid influx of adipose‐derived FFAs to the livers of these mice during exposure to PEG‐asparaginase does not provide a sufficient length of time for adaptation in TG secretion to occur as it does in a chronic model of fatty liver such as MASLD. As noted in our metabolomic data, livers from obese mice exposed to asparaginase show alterations in multiple lipid species including diacylglycerols, triacylglycerols, and phospholipids. Consistent with these findings, asparaginase treatment in human patients is observed to alter plasma lipid species such as phosphatidylethanolamines and specific TG species from pre‐to‐post ALL therapy [31].

Loss of ApoB100 could help explain decreased TG secretion with exposure to PEG‐asparaginase [13]. Obese mice showed significantly higher levels of hepatic ApoB100 protein abundance which can be indicative of a decrease in ApoB100 secretion. Work from Ota and colleagues demonstrated that an oleic acid infusion in mice leads to an initial compensatory increase in ApoB secretion that dissipates with a prolonged infusion [32]. Moreover, oleic acid infusion increases ER stress in the liver and inhibition of ER stress restored ApoB secretion [32]. Asparaginase induces activation of PERK, a sensor of ER stress, in the livers of obese mice fed an OD [13]. Previous studies also show that short‐term consumption of a high‐fat diet in mice is sufficient to induce ER stress [33]. Consistent with this, hepatic PERK phosphorylation trended upwards in lean mice fed an OD during PEG‐asparaginase and directionally mirrored changes in liver TG concentrations. On the other hand, PERK activation was not blunted in PEG‐asparaginase‐exposed obese mice consuming MD because they lost significant weight. Our study supports a model where an increase in FFAs from adipose tissue of obese mice during asparaginase exposure triggers hepatic ER stress, inhibiting ApoB100 secretion, which in turn exacerbates liver steatosis.

Albumin is the major plasma FFA carrier in the bloodstream [22, 23] and albumin knockout mice display significantly attenuated levels of plasma FFA and are protected from HFD‐induced hepatic steatosis [24, 25]. Based on this we surmised that loss of albumin might rescue PEG‐asparaginase induced liver steatosis. Notably, and contrary to Kumar and colleagues [14], plasma FFA decreased in PEG‐asparaginase treated WT mice. We speculate that the reduction of plasma FFA is most likely due to increased oxidation of FFA from peripheral tissues such as muscle, brown adipose tissue, and white adipose tissue, and the liver concomitantly acts as a sink for FFA, thus leading to the steatosis. Indeed, mice fed a leucine devoid diet, another model of an amino acid insufficiency, display increased markers of beta oxidation in white adipose tissue and brown adipose tissue and have a reduction in plasma FFA [34]. Further, obese albumin knockout mice exposed to asparaginase were not protected from liver steatosis, indicating that other plasma FFA carriers [35] were likely bringing adipose‐derived FFA to the liver and leading to the steatosis. Interestingly, total plasma proteins were lowest in albumin deficient animals treated with PEG‐asparaginase, suggesting greater stress in these animals. Clinically, low plasma albumin levels are common among patients receiving asparaginase and have been shown to be associated with lower event free survival and to impair efficacy of concurrent chemotherapies such as methotrexate [36, 37].

In summary, our findings indicate that preventing weight loss alongside increasing TG secretion from the liver may prevent or mitigate asparaginase‐associated metabolic adverse events. These important insights into the development of liver steatosis and ER stress during PEG‐asparaginase treatment may be useful to help improve treatment outcomes.

Author Contributions

Conceptualization (T.G.A.); Data curation (C.T.B., B.A.Z., E.M.L., E.T.M., K.T., Y.Z.); Formal analysis (C.T.B., B.A.Z., K.T., Y.Z.); Funding acquisition (T.G.A., R.C.W., G.C.H.); Investigation (C.T.B., B.A.Z., K.T., Y.Z., E.T.M., E.M.L., J.L.D); Methodology (J.L.D.); Project administration (T.G.A.); Resources (J.L.D., G.C.H., T.G.A.); Supervision (G.C.H., T.G.A.); Validation (C.T.B., B.A.Z., G.C.H., T.G.A.); Visualization (C.T.B., B.A.Z.); Writing, original draft preparation (C.T.B.); Writing, review and editing (C.T.B., B.A.Z., K.T., Y.Z., E.M.L., E.T.M., J.L.D., G.C.H., R.C.W., T.G.A.).

Funding

This work was funded by National Institute of Health: DK109714 (TGA, RCW); a Pilot Award from the New Jersey Institute for Food, Nutrition, and Health (TGA); and the McKinley Educational Initiative (GCH).

Disclosure

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: supporting Information.

Data Availability Statement

The data that support the findings of this study are available in the Materials and Methods, Results, and/or Supporting Information of this article.