Comparative effects of viable Lactobacillus rhamnosus GG and its heat‐inactivated paraprobiotic in the prevention of high‐fat high‐fructose diet‐induced non‐alcoholic fatty liver disease in rats

Abstract

Nonalcoholic fatty liver disease (NAFLD) is one of the most prevalent chronic liver alterations worldwide, being gut microbiota dysbiosis one of the contributing factors to its development. The aim of this research is to compare the potential effects of a viable probiotic (Lactobacillus rhamnosus GG) with those exerted by its heat‐inactivated paraprobiotic counterpart in a dietary rodent model of NAFLD. The probiotic administration effectively prevented the hepatic lipid accumulation induced by a high‐fat high‐fructose diet feeding, as demonstrated by chemical (lower TG content) and histological (lower steatosis grade and lobular inflammation) analyses. This effect was mainly mediated by the downregulation of lipid uptake (FATP2 protein expression) and upregulating liver TG release to bloodstream (MTTP activity) in rats receiving the probiotic. By contrast, the effect of the paraprobiotic preventing diet‐induced liver lipid accumulation was milder, and mainly derived from the downregulation of hepatic de novo lipogenesis (SREBP‐1c protein expression and FAS activity) and TG assembly (DGAT2 and AQP9 protein expression). The obtained results demonstrate that under these experimental conditions, the effects induced by the administration of viable L. rhamnosus GG preventing liver lipid accumulation in rats fed a diet rich in saturated fat and fructose differ from those induced by its heat‐inactivated paraprobiotic counterpart.

Viable LGG effectively prevents diet‐induced hepatic triacylglycerides (TG) accumulation downregulating liver lipid uptake (decreased FATP2 expression) and enhancing TG release to bloodstream (increased MTTP activity). Whereas its heat‐inactivated paraprobiotic downregulates de novo lipogenesis (lower SREBP‐1c expression and FAS activity), TG assembly (DGAT2 and AQP9 expression) and intestinal lipid absorption, its less effective preventing diet‐induced hepatic steatosis.

Abbreviations

ACYL

ATP‐citrate lyase

ALT

alanine aminotransferase

AMPK

AMP‐activated proteinkinase

AQP9

aquaporin 9

AST

aspartate aminotransferase

ATGL

adipose triglyceridelipase

BSA

bovine serum albumin

CD36

cluster of differentiation 36

CPT‐1a

carnitine palmitoyl transferase‐1a

CS

Citrate synthase

DGAT2

diacylglycerol O‐acyltransferase2

FA

fatty acid

FAS

fatty‐acid synthase

FATP2

For fatty‐acidtransport protein 2

HCC

hepatocellular carcinoma

HSL

hormone‐sensitive lipase

LPL

Lipoprotein lipase

MTTP

microsomal triacylglycerol transfer protein

NADPH

nicotinamide adenine dinucleotide phosphate

NAFLD

Non‐alcoholic fatty liver disease

NAS

NAFLD Activity Score

NASH

non‐alcoholic steatohepatitis

PBS

phosphate buffered saline

PGC‐1α

peroxisomeproliferator‐activated receptor‐gamma coactivator

SCD1

stearoyl‐CoAdesaturase‐1

SIRT1

NAD‐dependentdeacetylase sirtuin‐1

SIRT3

NAD‐dependent deacetylase sirtuin‐3

SREBP‐1c

sterol regulatoryelement‐binding protein‐1c

TFAM

mitochondrialtranscription factor A

TG

triacylglycerides

VLDL

very low‐density lipoprotein

WD

WesternDiet

1. INTRODUCTION

Nonalcoholic fatty liver disease (NAFLD) stands as the predominant chronic liver condition worldwide, encompassing diverse stages that span from simple steatosis and nonalcoholic steatohepatitis (NASH) to cirrhosis and even hepatocellular carcinoma (HCC). ¹ One of the main features of NAFLD is the enlargement of hepatic lipid droplets, which contain triacylglycerides (TG) and cholesterol esters, among other lipid species. Liver steatosis may be diagnosed through chemical or histological analysis, and it is confirmed when lipid accumulation surpasses 5% of parenchyma without apparent hepatocyte injury. ²

Even though, NAFLD is a multifactorial disease, the consumption of the so‐called Western Diet (WD) emerges as a significant contributor to its development and progression. This dietary pattern is characterized by high intakes of fats (mainly saturated) and added sugars (especially fructose) that are rapidly absorbed in the small intestine, depriving colonic bacteria from nutrients and fostering gut microbiota dysbiosis. ³ The result is a decrease in microbial diversity along with a shift in bacterial communities toward pro‐inflammatory mediator‐producers. ³ Given that WD triggers gut barrier dysfunction (leading to a higher permeability), these gut‐derived metabolites leak into the portal circulation, leading to various assaults on the liver and playing a pivotal role in the pathogenesis of NAFLD. ³

Due to the intricate nature of NAFLD pathogenesis, there is a shortage of approved targeted medications, and efforts to address this condition through dietary interventions and lifestyle modifications have demonstrated effectiveness, albeit with limited adherence among the population. ¹ In light of the established connection between gut dysbiosis and liver steatosis, significant emphasis has been placed on treatments aimed at restoring gut microbiota composition, thereby addressing this liver alteration. In this context, probiotics have surfaced as a compelling therapeutic tool for managing NAFLD, as explained in a recent review article. ⁴ Nevertheless, the administration of viable bacteria to vulnerable individuals may pose risks, including the potential for systemic infection, excessive immune response stimulation and gene transfer. ⁵ Moreover, probiotics exhibit techno‐functional limitations, such as the necessity to maintain its viability during storage and industrial processing. Hence, there is a gradual shift in focus from viable probiotic bacteria to nonviable microorganisms, termed paraprobiotics. ⁶ Interestingly, and as occurs with probiotics, these nonviable microbial cells (intact or broken) or crude cell extracts may confer benefits to the host if consumed in enough quantities, emerging as a feasible alternative to probiotics. ⁶

In this scenario, the current research aims to compare the potential effects of a viable probiotic (Lactobacillus rhamnosus GG) with those exerted by its heat‐inactivated paraprobiotic counterpart, in the prevention of diet‐induced NAFLD in rats. To achieve this, metabolic pathways involved in liver lipid accumulation and metabolism have been addressed.

2. MATERIALS AND METHODS

2.1. Animals, diets, and experimental design

All experimental procedures involving animals were performed following the Guidelines for Care and Use of Laboratory Animals of the University of the Basque Country. The experimental protocol employed in this research study was approved by the Ethics Committee on Animal Experimentation of the University of the Basque Country (under reference M20/2021/214). The experiment was performed using 34 (8‐ to 9‐week‐old) male Wistar rats that were housed in conventional polycarbonate cages (two animals per cage) and maintained in an air‐conditioned room (22 ± 2°C) under a 12 h light–dark cycle. Following a 6‐day adaptation period, the animals were distributed into four experimental groups. The first group received a standard laboratory diet (C group, n = 8) (D10012G; Research Diets, New Brunswick, NJ, USA) while the remaining three groups were fed a diet rich in saturated fat and fructose (D21052401; Research Diets) (Table 1). For the animals subjected to the unbalanced diet, one group exclusively received the high‐fat high‐fructose diet (HFHF group, n = 8), while the other two groups were additionally administered a commercially available probiotic (L. rhamnosus GG, Ferring Pharmaceuticals, Switzerland) at a dosage of 1 × 10⁹ CFU/day (PRO group, n = 9), or its paraprobiotic counterpart (nonviable, heat‐inactivated probiotic) (PARA group, n = 9), also at the same dose. The inactivated bacteria were obtained by exposing L. rhamnosus GG to 80°C for 20 min, following guidelines outlined elsewhere. ⁷ Both the probiotic and the paraprobiotic were diluted in phosphate buffered saline (PBS) containing 5% sucrose, while the animals in the C and HFHF groups received sucrose‐enriched PBS as the vehicle. Throughout the 6‐week experimental period, all treatments and vehicle were administered daily via oral gavage. During the entire experimental period, animals were provided ad libitum access to food and water. Body weight and food intake were meticulously recorded on a daily basis. Cecum content was collected prior to sacrifice and immediately stored at −80°C until further analysis.

TABLE 1.

Nutritional composition of the experimental diets (% energy).

	Total energy (kcal/g)	Carbohydrates	Fructose	Protein	Lipids
STD	3.9	63.9	‐	20.3	15.8
HFHF	4.5	40	10	20	40

Abbreviations: g, grams; HFHF, high‐fat high‐fructose diet; kcal, kilocalories; STD, standard diet.

At the end of the experimental period, the animals were anesthetized (chloral hydrate) and sacrificed following a fasting period (8–12 h), by cardiac exsanguination. Liver and skeletal muscle samples were dissected, weighed, and immediately frozen in liquid nitrogen. Blood samples were centrifuged (1000g for 10 min, at 4°C) for serum collection. All samples were stored at −80°C until further analysis.

2.2. Serum determinations

The determination of serum parameters, namely TG, alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels was carried out with commercially available spectrophotometric kits (Biosystems, Barcelona, Spain). Results were presented as mg TG per 100 ml and μmol (serum transaminase) min⁻¹ L⁻¹.

2.3. Cecum lipid content

Total lipids were extracted from fresh cecal samples following the method described by Kraus et al. ⁸ Briefly, 100–120 mg of cecum samples of each animal were weighted and powdered. Subsequently, equal volumes (500 μl) of normal saline and chloroform: methanol (2:1) were added, and samples were vigorously mixed. Following this, tubes were centrifuged at 1000g for 10 min at room temperature, and the lower liquid phase was collected and transferred to a pre‐weighted tube. Samples were kept under a fume hood until complete evaporation of all liquid components occurred. Finally, the sample‐containing tubes were weighted and the cecum lipid content was calculated by gravimetric analysis (mg of lipids of each sample). Results were expressed as % of lipids mg mg⁻¹ sample.

2.4. Liver TG content

Total lipids were extracted from liver samples following the method described by Folch et al. ⁹ Lipid extracts were then dissolved in isopropanol and the TG content was measured using a commercial kit (Spinreact, Barcelona, Spain). Results were expressed as % of TG mg g⁻¹ tissue.

2.5. Steatosis assessment

After sacrifice, a liver section from each animal (obtained from the same lobule) was immersed in formaldehyde buffer and subsequently embedded in paraffin. Afterward, samples were stained with hematoxylin and eosin, using standard techniques. The treatment group from each animal was deliberately concealed during the analysis of the sections.

The assessed characteristics included steatosis, lobular inflammation, ballooning degeneration, and fibrosis. Lobular inflammation was graded based on the number of inflammatory foci per 200 × field: 0 when there were no inflammatory foci, 1 when 1 to 2 inflammatory foci per 200 × field were present, and 2 when 2 to 4 inflammatory foci per 200 × were identified. Ballooning was evaluated based on numbers of hepatocytes exhibiting this lesion: 0 when there was no ballooning, 1 when few hepatocytes showed ballooning, and 2 when many cells displayed ballooning degeneration. Finally, the NAFLD Activity Score (NAS) was calculated using data from steatosis, lobular inflammation and ballooning scores.

2.6. Enzymatic activities in liver and skeletal muscle

2.6.1. Enzymatic activity of proteins involved in de novo lipogenesis, β‐oxidation, TG assembly, and release

ATP‐citrate lyase (ACYL) enzymatic activity was determined using a commercially available spectrophotometric kit (SolarBio, Beijing, China). Briefly, 100 mg of liver tissue were homogenized in 1 ml extraction reagent. Subsequently, the samples were centrifuged at 8000g for 10 min at 4°C, and the supernatants were transferred to separate tubes. Following this, one tube per sample containing a mixture of reagents provided in the commercial kit was incubated, and 50 μl of sample was added to the respective tube. Absorbance was recorded at 340 nm both in t = 0, and t = 2 min. ACYL activity was calculated using the following formula, and results were expressed as U enzyme mg⁻¹ protein:

$$\text{ACYL}\left(\mathrm{U}\text{enzyme}/\mathrm{mg}\text{prot}\right)=1607.7\mathrm{x}\mathrm{\Delta A}/\mathrm{Cpr}$$

$$\mathrm{\Delta A}={\mathrm{\Delta A}}_{\text{sample}}–{\mathrm{\Delta A}}_{\text{blank}},\mathrm{Cpr}:\text{protein concentration}\left(\mathrm{mg}/\mathrm{ml}\right)$$

For the determination of the activities of other lipogenic and oxidative enzymes, liver samples weighting 100–150 mg were homogenized in 1 ml homogenization buffer (250 mM saccharose, 1 mM EDTA, 10 mM Tris–HCl, pH 7.4) and centrifuged at 700g for 10 min at 4°C. Subsequently, the supernatants were collected and further centrifuged at 12,000g for 15 min at 4°C. The resulting supernatant corresponded to the cytosolic fraction of the hepatic cells. Pellets containing the mitochondrial fraction were resuspended in resuspension buffer (70 mM saccharose, 220 mM mannitol, 2 mM HEPES, and 1 mM EDTA, pH 7.4). The protein content of both fractions was determined by the Bradford method ¹⁰ using bovine serum albumin (BSA) as standard.

In regards to de novo lipogenesis, the activity of fatty‐acid synthase (FAS) was assessed in the cytosolic fraction following the method outlined by Lynen ¹¹ and by Nepokroeff et al. ¹² The activity of FAS was assayed by monitoring the absorbance changes of the sample resulting from nicotinamide adenine dinucleotide phosphate (NADPH) consumption during the reaction catalyzed by this enzyme. The results were expressed as consumed nmol min⁻¹ mg⁻¹ protein.

The activity of enzymes involved in fatty acid (FA) oxidation was assessed using the mitochondrial/peroxisomal fraction. Initially, the activity of carnitine palmitoyl transferase‐1a (CPT‐1a), responsible for the transport of long‐chain fatty acylCoA into the mitochondria for oxidation, was determined spectrophotometrically following the method described by Bieber et al. ¹³ This method is based on the measurement of the released CoA‐SH and results are expressed as nmol min⁻¹ mg⁻¹ tissue. Citrate synthase (CS) activity, often considered a marker of mitochondrial density, ¹⁴ was also evaluated spectrophotometrically by measuring the production of free CoA, according to the method described by Srere. ¹⁵ Briefly, after an initial 2‐min incubation at 30.1°C with a solution comprising acetyl‐CoA, 1.01 mM DTNB, 10 mM oxaloacetate, Triton X‐100 (10%) and distilled water, the absorbance of the samples was measured at 412 nm. After a second 5‐min incubation, the absorbance of the samples was again recorded at the same wavelength. CS activity was then quantified and expressed as nmol CoA min⁻¹ mg⁻¹ protein.

Finally, to determine the rate of very low‐density lipoprotein assembly and secretion from the liver into circulation, the activity of microsomal triacylglycerol transfer protein (MTTP) was assessed by fluorimetry using a commercial kit (Sigma‐Aldrich, St. Louis, MI, USA). In brief, 100–150 mg of liver samples were homogenized in 1 ml of extraction buffer (150 mM NaCl, 10 mM Tris–HCl, 1 mM EDTA) supplemented with protease inhibitors (100 mM phenylmethylsulfonyl fluoride). The homogenates were then centrifuged at 7500g for 30 min at 4°C, and the resulting supernatant was utilized for MTTP activity determination. The enzyme activity was expressed as % of activity mg⁻¹ protein h⁻¹.

2.6.2. Enzymatic activity of lipoprotein lipase in skeletal muscle

Lipoprotein lipase (LPL) enzymatic activity in skeletal muscle samples was determined following the methodology described by Del Prado et al. ¹⁶ Briefly, 200–250 mg were incubated for 15 min at 37°C in 1 ml of Krebs‐Ringer‐phosphate buffer (100 vol. NaCl, 4 vol. KCl, 3 vol. CaCl₂, 1 vol. Mg SO₄ 7H₂O and 21 vol. buffer phosphate without NaCl) with the addition of 2 μg/ml of sodium heparin (pH 7.4). Subsequently, two tubes per sample were prepared containing phosphate buffer (3 mM NaH₂PO₄, 50 mM Na₂HPO₄ with or without 2.5 M NaCl), ethylene‐glycol monomethyl ether, and fluorescein solution (2 mg fluorescein, 9.5 ml NaCl‐containing phosphate buffer and 500 μ ethylene‐glycol monomethyl ether). Finally, 100 μl of the sample were added to each of these tubes. Following incubation for 5 min at 37°C, reaction was stopped by placing the tubes on ice, and fluorescence was recorded at excitation and emission wavelengths of 490 and 530 nm, respectively. The enzyme activity results were expressed as nmol fluorescein min⁻¹ g⁻¹ tissue and are provided in the supplementary material (Supp. Material S1).

2.7. Western blot

For fatty‐acid transport protein 2 (FATP2), aquaporin 9 (AQP9), AMP‐activated protein kinase (AMPK), cluster of differentiation 36 (CD36), stearoyl‐CoA desaturase‐1 (SCD1), diacylglycerol O‐acyltransferase 2 (DGAT2), adipose triglyceride lipase (ATGL), NAD‐dependent deacetylase sirtuin‐1 (SIRT1), NAD‐dependent deacetylase sirtuin‐3 (SIRT3), hormone‐sensitive lipase (HSL) and mitochondrial transcription factor A (TFAM), 100–200 mg of liver samples or 150 mg skeletal muscle were homogenized in 1 ml of cellular PBS containing protease inhibitors (100 mM phenylmethylsulfonyl fluoride and 100 mM iodoacetamide). Homogenates were then centrifuged at 800g for 10 min at 4°C. Protein quantification was carried out following the Bradford method, ¹⁰ using BSA as the standard reference. For the sterol regulatory element‐binding protein‐1c (SREBP‐1c) and peroxisome proliferator‐activated receptor‐gamma coactivator 1α (PGC‐1α), nuclear protein extraction was conducted using 150–200 mg of hepatic tissue, following established procedures outlined in the literature. ¹⁷

Immunoblot analyses were performed using 60 μg of protein from total/nuclear liver or skeletal muscle extracts resolved via electrophoresis on 4–15% SDS‐polyacrylamide gradient gels (Bio‐Rad, Hercules, CA, USA) and subsequently transferred onto PVDF membranes (Merck, Darmstadt, Germany). Then, the membranes were blocked with 4% BSA PBS‐ Tween buffer for 2 h at room temperature. Following this, they were blotted with the corresponding antibodies overnight at 4°C. Protein levels were detected via specific antibodies for FATP2 (1:1000), AQP9 (1:1000), SIRT3 (1:1000) and TFAM (1:1000) (SantaCruz Biotech, Dallas, TX, USA); CD36 (1:1000), AMPK (1:1000), ATGL (1:1000), HSL (1:1000), β‐actin (1:1000) and α‐tubulin (1:1000) (Cell Signaling Technology, Danvers, MA, USA); DGAT2 (1:1000), SCD1 (1:1000), SIRT1 (1:1000) and SREBP‐1c (1:1000) (Abcam, Cambridge, UK); PGC‐1α (1:1000) (BioTechne, Minneapolis, Minnesota, USA). Subsequently, polyclonal anti‐rabbit for ATGL, CD36, β‐actin, α‐tubulin, SCD1, SREBP‐1c and PGC‐1α, anti‐mouse targeting AQP9, SIRT1 and SIRT3, and anti‐goat for FATP2, TFAM and DGAT2 were incubated for 2 h at room temperature and quantified. After antibody stripping, the membranes underwent a second blocking step before being incubated with antibodies targeting phosphorylated AMPK (threonine 172, 1:1000), phosphorylated HSL (serine 660, 1:1000) and acetylated lysine (1:1000) (Cell Signaling Technology, Danvers, MA, USA). Finally, polyclonal anti‐rabbit for phosphorylated AMPK, phosphorylated HSL and acetylated lysine were incubated for 2 h at room temperature, and quantified. The bound antibodies were visualized using an ECL system (Thermo Fisher Scientific Inc., Rockford, IL, USA) and quantified with a ChemiDoc MP Imaging System (Bio‐Rad). The measurements were normalized by β‐actin and α‐tubulin in total protein extractions. The results of SIRT3 protein expression in skeletal muscle can be found in the supplementary material (Supp. Material S2).

2.8. Statistical analysis

Results are presented as mean ± SEM. Statistical analysis was performed using SPSS 24.0 (SPSS, Chicago, IL, USA). The normal distribution of data was assessed by Shapiro–Wilks test. Normally distributed parameters were analyzed by one‐way ANOVA followed by the Newman–Keuls post hoc test. Significance was assessed at the p < 0.05 level. In cases in which nonstatistical trends (p = 0.1) were observed, the effect size of the treatment was determined using Cohen's d, with a value greater than 0.8 indicating a large effect size.

3. RESULTS

3.1. Body and liver weights

After 6 weeks of high‐fat high‐fructose feeding, animals in the HFHF group exhibited significantly higher body weight compared to those receiving the standard laboratory diet (C group). In this regard, the animals in the HFHF group consumed significantly more energy daily than the animals in the C group, despite the amount of diet consumed was similar. ¹⁸ , ¹⁹ Although animals treated with the probiotic and the paraprobiotic did not display significantly lower body weights compared to rats in the HFHF group, a partial reduction in this parameter was observed in both cases (p = 0.1 and p = 0.09 PRO and PARA vs. HFHF, respectively) (Figure 1A). Noteworthy, this slight body‐weight decrease occurred in absence of significant differences in food intake. ¹⁸ , ¹⁹ A similar pattern was found regarding liver weights (Figure 1B), with a significant increase noted in the non‐treated high‐fat high‐fructose fed animals (HFHF group) compared to the C group. Additionally, there was a nonsignificant trend toward lower liver weights in the animals treated with the viable and heat‐inactivated bacteria compared to those in the HFHF group (p = 0.1 and p = 0.07 PRO and PARA vs. HFHF, respectively).

FIGURE 1.

Final body weight (g) (A) and liver weight (g) (B) of rats fed a standard diet (C) or a high‐fat high‐fructose diet alone (HFHF), supplemented or not with viable or heat‐inactivated L. rhamnosus GG (PRO and PARA, respectively) for 6 weeks. Values are presented as mean ± SEM. Differences among the groups were determined using one‐way ANOVA followed by the Newman–Keuls post hoc test. Bars not sharing common letters are significantly different (p < 0.05).

3.2. Liver TG content and serum determinations

In terms of hepatic TG content, the group receiving the high‐fat high‐fructose diet alone exhibited significantly higher levels of TG (%) compared to the animals receiving the standard laboratory diet (p < 0.001). As for the effects of the treatments, while probiotic administration effectively prevented hepatic TG accumulation (p < 0.01), a nonsignificant trend but with a large effect size was observed in the group treated with the heat‐inactivated bacteria, compared to the non‐treated group (p = 0.1 PARA vs. HFHF; Cohen's d = +1.9) (Table 2). It is noteworthy that none of the treatments resulted in lowering TG hepatic content to levels comparable to those observed in animals fed with the standard diet (C group).

TABLE 2.

Hepatic TG, serum TG, alanine aminotransferase (ALT), and aspartate aminotransferase (AST) levels on rats fed a standard diet (C) or a high‐fat high‐fructose diet alone (HFHF) supplemented or not with viable or heat‐inactivated Lactobacillus rhamnosus GG (PRO and PARA, respectively) for 6 weeks.

	C	HFHF	PRO	PARA	p value
Hepatic TG (%)	4.3 ± 0.16 c	13.1 ± 0.46 a	10.8 ± 0.44b	11.9 ± 0.46ab	p < 0.01
Serum TG (mg/100 ml)	49.2 ± 4.9c	81.5 ± 6.9a	63.7 ± 1.9b	69.4 ± 3.0ab	p < 0.05
ALT (μmol/min L−1)	13.6 ± 3.1 b	46.6 ± 11.5 a	32.3 ± 2.6 a	27.9 ± 3.2a	p < 0.05
AST (μmol/min L−1)	43.9 ± 1.8b	101.3 ± 21.6a	56.5 ± 6.5ab	61.0 ± 5.0a	p < 0.05

Note: Values are presented as mean ± SEM. Differences among groups were determined using one‐way ANOVA followed by the Newman–Keuls post hoc test. Values not sharing a common letter are significantly different (p < 0.05).

Concerning serum parameters, animals fed the high‐fat high‐fructose diet alone showed a significant boost in serum TG levels compared to the C group. Probiotic administration effectively mitigated this increase (p < 0.05 PRO vs. HFHF) albeit not reaching the levels observed in the C group. A partial reduction was also observed in the animals on the paraprobiotic compared to the HFHF group (p = 0.1 PARA vs. HFHF). Regarding serum transaminase levels, high‐fat high‐fructose feeding led to a significant increase in ALT and AST compared to the standard diet feeding. Even though none of the treatments prevented this effect, it is worth mentioning that ALT levels were 23.4 and 40.1% lower in the treated groups (PRO and PARA vs. HFHF, respectively). Additionally, both treatments elicited a notable but nonsignificant decrease in serum AST levels (−44.2%, p = 0.081 and −39.8%, p = 0.1, PRO and PARA vs. HFHF, respectively) (Table 2).

3.3. Cecum lipid content

Regarding the analysis of lipid content in feces, animals fed the high‐fat high‐fructose diet excreted a higher proportion of lipids compared to the C group. As for the effects of the treatments, animals receiving the probiotic exhibited no changes in this parameter compared to the HFHF group, whereas a significant increase in cecum lipid content was found in the PARA group in comparison with non‐treated animals subjected to the high‐fat high‐fructose diet (p < 0.05) (Figure 2). In addition, a trend toward higher cecum lipid content was also observed in the PARA group compared to the PRO group (p = 0.071).

FIGURE 2.

Cecum lipid content of rats fed a standard diet (C) or a high‐fat high‐fructose diet alone (HFHF) supplemented or not with viable or heat‐inactivated L. rhamnosus GG (PRO and PARA, respectively) for 6 weeks. Bars not sharing common letters are significantly different (p < 0.05).

3.4. Histopathological assessment

The evaluation of liver samples by microscopy allowed the assessment of steatosis, inflammation and ballooning, and thus, the calculation of NAS score (Figures 3 and 4). As expected, 100% of the animals fed the standard diet showed a grade 0 steatosis. As for the animals under the high‐fat high‐fructose diet alone, 87.5% of them displayed a grade 1 steatosis, and 12.5% a grade 2. Noteworthy, the majority of the animals (77.8%) fed the same high‐fat high‐fructose diet supplemented with the probiotic exhibited grade 0 steatosis, whereas 22.2% of the animals manifested grade 1, with none progressing to grade 2. A similar trend was also observed in the PARA group, were 44.3% of rats showed grade 0 steatosis, 55.6% developed grade 1 and none of the animals reached grade 2 (Figure 3A). Regarding hepatocyte inflammation, 62.5% of the rats in the C group exhibited grade 0, whereas 37.5% showed grade 1. None of the animals subjected solely to the high‐fat high‐fructose diet (HFHF group) demonstrated grade 0 inflammation; instead, 50% displayed grade 1 and the remaining 50% presented grade 2. Remarkably, among the animals treated with the probiotic, 22.2% showed grade 0 inflammation, while 44.4% displayed grade 1 and 33.3% presented grade 2. As occurred in steatosis, the effect of the paraprobiotic on hepatocyte inflammation was similar but generally less pronounced than that observed in the probiotic group. Specifically, none of the animals in the PARA group exhibited grade 0 inflammation, while 66.6% presented grade 1, and 33.3% displayed grade 2 (Figure 3B). As for hepatocyte ballooning, 100% of the animals in the C group showed grade 0 on this parameter. By contrast, the high‐fat high‐fructose feeding increased hepatocyte ballooning grade to 2 in 100% of the animals (HFHF group), and effect that remained unaltered by any of the tested interventions (Figure 3C). Finally, concerning total NAS score calculation, high‐fat high‐fructose feeding significantly elevated NAS score compared to the C group (Figure 3D). While none of the treatments succeeded in reducing NAS scores to the levels observed in the C group, probiotic administration effectively prevented the increase in the NAS score induced by high‐fat high‐fructose feeding (p < 0.01 PRO vs. HFHF). Additionally, a nonsignificant trend toward a reduction in this metric was observed with paraprobiotic treatment (p = 0.065 PARA vs. HFHF).

FIGURE 3.

Grade of steatosis (A), lobular inflammation (B), ballooning degeneration (C), and NAS score (D) in livers of rats fed a standard diet (C) or a high‐fat high‐fructose diet alone (HFHF) supplemented or not with viable or heat‐inactivated L. rhamnosus GG (PRO and PARA, respectively) for 6 weeks. Values are presented as mean ± SEM. Differences among the groups were determined using one‐way ANOVA followed by the Newman–Keuls post hoc test. Bars not sharing common letters are significantly different (p < 0.05).

FIGURE 4.

Hepatic histological study using hematoxylin and eosin staining of animals receiving a standard laboratory diet (C), a high‐fat high‐fructose diet alone (HFHF), a high‐fat high‐fructose diet supplemented with L. rhamnosus GG (PRO), or a high‐fat high‐fructose diet supplemented with the heat‐inactivated L. rhamnosus GG (PARA) for 6 weeks. Two images with 100× and 200× magnification are included per group.

Figure 4 presents a collection of images captured during the histopathological assessment of liver samples using hematoxylin and eosin staining. Animals fed the standard diet (C group) showed intact liver parenchyma with physiological sinusoid dilation. In contrast, high‐fat high‐fructose feeding resulted in the development of moderate microvesicular‐predominant steatosis (HFHF group). As for the effects of the treatments, liver tissues from animals treated with both, the probiotic and the parabiotic, exhibited no abnormalities in hepatocytes surrounding the central veins.

3.5. Enzymatic activities and expression of proteins involved in hepatic lipid metabolism

3.5.1. Proteins involved in liver lipid and glycerol uptake

Regarding the expression of proteins involved in hepatic lipid uptake, the results indicated that animals under the high‐fat high‐fructose diet alone exhibited a significant increase in FATP2, with no changes observed in CD36 protein expression compared to the C group. As for the effects of the treatments, probiotic administration effectively prevented the high‐fat high‐fructose diet‐induced increase in FATP2 (p < 0.05 PRO vs. HFHF) restoring levels akin to those in the C group. However, no discernible effect on CD36 protein expression was observed compared to the HFHF group. By contrast, paraprobiotic administration did not affect FATP2 protein expression compared to the HFHF group, whereas it significantly decreased CD36 expression (p < 0.05 PARA vs. HFHF) (Figure 5A,B). Considering the necessity of glycerol for TG assembly in hepatocytes, the protein expression of AQP9 was also studied (Figure 5C). In this regard, non‐treated animals fed the high‐fat high‐fructose diet showed a tendency toward lower AQP9 protein compared to the C group (p = 0.063 HFHF vs. C). As for the effects of the treatments, neither of the groups displayed significant changes compared to the non‐treated animals (HFHF group). However, both treatments resulted in significant reductions in this parameter in comparison with the C group. Noteworthy, compared to the HFHF group, paraprobiotic treatment demonstrated a substantial effect size (p = 0.1; Cohen's d = 1.87) toward lower AQP9 protein expression.

FIGURE 5.

Protein expression of fatty‐acid transporter 2 (FATP2) (A), cluster of differentiation 36 (CD36) (B), and aquaporin 9 (AQP9) (C) in livers of rats fed a standard diet (C) or a high‐fat high‐fructose diet (HFHF), supplemented or not with viable or heat‐inactivated L. rhamnosus GG (PRO and PARA, respectively) for 6 weeks. Values are presented as mean ± SEM. Differences among the groups were determined using one‐way ANOVA followed by the Newman–Keuls post hoc test. Bars not sharing common letters are significantly different (p < 0.05).

3.5.2. Enzymatic activity and expression of proteins and transcription factors involved in de novo lipogenesis

The activity of ACYL was significantly increased in the non‐treated group under the high‐fat high‐fructose diet compared to the animals fed the standard diet (p < 0.001). As for the treatment effects, probiotic administration markedly reduced the activity of this enzyme compared to the HFHF group, although the difference did not reach statistical significance (−29%, p = 0.1 PRO vs. HFHF). Conversely, no differences were observed in ACYL activity in the PARA group compared to the HFHF group (Figure 6A). Concerning FAS enzymatic activity, animals subjected to the high‐fat high‐fructose diet alone showed a significant increase in this parameter compared to the animals in the C group (p < 0.01). The administration of the probiotic induced a substantial but nonsignificant decrease on this enzymatic activity (−43.75%, p = 0.1) with a large effect size (Cohen's d = 1.85), reaching values similar to those found in the C group. By contrast, paraprobiotic treatment effectively prevented the effect of the diet on FAS activity compared to the HFHF group (p < 0.05), reaching values similar to those observed in the C group (Figure 6B). No differences were observed between the two treated groups on this parameter.

FIGURE 6.

Enzymatic activity of ATP‐citrate lyase (ACYL) (A) and fatty‐acid synthase (FAS) (B) and nuclear protein expression of sterol regulatory element‐binding protein‐1c (SREBP1c) (C) in livers of rats fed a standard diet (C) or a high‐fat high‐fructose diet (HFHF), supplemented or not with viable or heat‐inactivated L. rhamnosus GG (PRO and PARA, respectively) for 6 weeks. Values are presented as mean ± SEM. Differences among the groups were determined using one‐way ANOVA followed by the Newman–Keuls post hoc test. Bars not sharing common letters are significantly different (p < 0.05).

Additionally, SREBP‐1c was also quantified as it serves as a pivotal transcriptional factor regulating FAS activity. In this context, no significant changes in SREBP‐1c protein expression were observed between the C and HFHF groups. While probiotic administration resulted in a nonsignificant but sharp decrease in the expression of this transcriptional factor (−36.2%, p = 0.081 PRO vs. HFHF), the reduction found in the PARA group reached statistical significance compared to the HFHF group (p < 0.01) (Figure 6C).

3.5.3. Expression of proteins involved in TG assembly

To assess lipid assembly in the liver, the protein expression of SCD1 was measured. Animals under the high‐fat high‐fructose diet alone exhibited a nonsignificant trend toward lower SCD1 protein expression compared to the C group (p = 0.075). Regarding the treatment effects, although no changes were observed in the animals treated with the probiotic, a partial prevention with a large effect size was achieved in the animals receiving the paraprobiotic compared to the non‐treated animals (p = 0.1; Cohen's d = −2.25 PARA vs. HFHF) (Figure 7A). As for DGAT2, a sharp but not significant decrease was observed in the animals fed the high‐fat high‐fructose diet alone compared to those under the standard diet (−29%, p = 0.1 HFHF vs. C). Although a reduction in this parameter was observed in treated animals compared to the C group (p < 0.05 and p < 0.001 PRO and PARA vs. C, respectively), neither of them showed significant differences compared to the HFHF group (Figure 7B).

FIGURE 7.

Protein expression of stearoyl‐CoA desaturase‐1 (SCD1) (A) and diacylglycerol O‐acyltransferase 2 (DGAT2) (B) in livers of rats fed a standard diet (C) or a high‐fat high‐fructose diet (HFHF), supplemented or not with viable or heat‐inactivated L. rhamnosus GG (PRO and PARA, respectively) for 6 weeks. Values are presented as mean ± SEM. Differences among the groups were determined using one‐way ANOVA followed by the Newman–Keuls post hoc test. Bars not sharing common letters are significantly different (p < 0.05). $ = p < 0.1 (vs. C), # = p < 0.1 (vs. HFHF).

3.5.4. Enzymatic activities of markers of fatty‐acid oxidation and mitochondrial density

Regarding CPT‐1a enzymatic activity, a significant decrease was found in the HFHF group compared to the C group (p < 0.01). No significant changes were found in the treated groups compared to the HFHF group (Figure 8A). As for CS activity, no differences were found among the four experimental groups (Figure 8B).

FIGURE 8.

Activity of carnitine palmitoyl transferase‐1a (CPT‐1a) (A) and citrate synthase (CS) (B) in livers of rats fed a standard diet (C) or a high‐fat high‐fructose diet (HFHF), supplemented or not with viable or heat‐inactivated L. rhamnosus GG (PRO and PARA, respectively) for 6 weeks. Values are presented as mean ± SEM. Differences among the groups were determined using one‐way ANOVA followed by the Newman–Keuls post hoc test. Bars not sharing common letters are significantly different (p < 0.05).

3.5.5. Expression and activation rates of proteins involved in energy metabolism and mitochondrial biogenesis

The expression of proteins and transcriptional factors involved in energy metabolism and mitochondrial biogenesis was also experimentally assessed. In this context, SIRT1 protein expression was significantly decreased in the HFHF group compared to the C group, an effect that, neither the probiotic nor the paraprobiotic were able to prevent (Figure 9A). Regarding AMPK activation, no changes were observed in the non‐treated animals fed the high‐fat high‐fructose diet alone compared to those in the C group. It is noteworthy that both treatments induced a nonsignificant albeit marked increase in AMPK phosphorylation in comparison with the HFHF group (+ 33.8% and +31.1% PRO and PARA vs. HFHF, respectively) (Figure 9B). As for markers of mitochondrial biogenesis, no changes were observed in PGC‐1α acetylation among the four experimental groups (Figure 9C). By contrast, high‐fat high‐fructose feeding led to a significant reduction in TFAM protein expression compared to the animals on the standard diet, an effect that was not prevented by any of the tested interventions (Figure 9D).

FIGURE 9.

Protein expression of NAD‐dependent deacetylase sirtuin‐1 (SIRT1) (A), phosphorylated AMPK (threonine 172)/total AMPK ratio (B), nuclear protein expression of acetylated peroxisome proliferator‐activated receptor‐gamma coactivator 1 alpha (PGC‐1α)/total PGC‐1α (C) and protein expression of mitochondrial transcription factor A (TFAM) (D) in livers of rats fed a standard diet (C) or a high‐fat high‐fructose diet alone (HFHF), supplemented or not with viable or heat‐inactivated L. rhamnosus GG (PRO and PARA, respectively) for 6 weeks. Values are presented as mean ± SEM. Differences among the groups were determined using one‐way ANOVA followed by the Newman–Keuls post hoc test. Bars not sharing common letters are significantly different (p < 0.05).

3.5.6. Expression, activation rate, and enzymatic activities of proteins involved in liver lipid mobilization and release

Regarding lipolysis, animals fed the high‐fat high‐fructose diet alone exhibited a tendency toward a lower hepatic ATGL protein expression compared to those in the C group (p = 0.1). In addition, statistically significant reductions in this parameter were observed in both treated groups compared to animals fed the standard diet (p < 0.05 PRO and PARA vs. C) (Figure 10A). Concerning HSL, no significant changes were observed in its activity (phosphorylation ratio) among the four experimental groups (Figure 10B).

FIGURE 10.

Protein expression of adipose triglyceride lipase (ATGL) (A) and phosphorylated HSL (serine 660)/total HSL ratio (B) in livers of rats fed a standard diet (C) or a high‐fat high‐fructose diet alone (HFHF), supplemented or not with viable or heat‐inactivated L. rhamnosus GG (PRO and PARA, respectively) for 6 weeks. Values are presented as mean ± SEM. Differences among the groups were determined using one‐way ANOVA followed by the Newman–Keuls post hoc test. Bars not sharing common letters are significantly different (p < 0.05).

Finally, with regard to lipid release from liver to systemic circulation, no changes were observed in MTTP enzymatic activity in animals receiving the high‐fat high‐fructose diet alone compared to the C group. However, a significant increase was noted in this parameter in the group treated with viable bacteria compared to animals in the HFHF group (p < 0.01 PRO vs. HFHF) (Figure 11). It is noteworthy that the increase found in MTTP enzymatic activity in animals receiving the probiotic was also significantly higher than that seen in animals receiving the heat‐inactivated bacteria (p < 0.001 PRO vs. PARA). As for the paraprobiotic administration, this intervention did not induce significant changes in MTTP activity compared to the C or HFHF groups.

FIGURE 11.

Microsomal triacylglycerol transfer protein (MTTP) activity in livers of rats fed a standard diet (C) or a high‐fat high‐fructose diet (HFHF), supplemented or not with viable or heat‐inactivated L. rhamnosus GG (PRO and PARA, respectively) for 6 weeks. Values are presented as mean ± SEM. Differences among the groups were determined using one‐way ANOVA followed by the Newman–Keuls post hoc test. Bars not sharing common letters are significantly different (p < 0.05).

4. DISCUSSION

NAFLD, a metabolic dysfunction often characterized by its silent nature, is influenced by several factors. Among them, the consumption of WD (rich in fats and sugars) frequently acts as a trigger for its development and progression. ²⁰ In particular, the excessive consumption of added fructose, often present in processed foods, has garnered considerable attention due to its notable lipogenic effects, serving as both, substrate and inducer, of lipogenic pathways in the liver. ²¹ Besides its detrimental effects on liver lipid metabolism, WD also contributes to gut microbiota dysbiosis and other intestinal affections (e.g., gut barrier dysfunction), which are present in most patients featuring NAFLD, highlighting the complex interaction between gut, microbiota and liver in the progression of this pathology. ²¹

In order to analyze the effectiveness of the tested treatments, the initial objective of the current research was to develop a dietary model of NAFLD. As expected, feeding rats a high‐fat high‐fructose diet for 6 weeks resulted in significantly higher body and liver weights. Indeed, alongside elevated liver weights, the animals exhibited liver steatosis and inflammation, as confirmed through TG quantification and histological analysis, and assessed by using the scoring system (namely NAS score) described by Kleiner et al. ²² In this regard, the grade 1 hepatocyte inflammation observed in some animals in the C group could be interpreted as physiological, since inflammation plays an essential role in the maintenance of liver homeostasis. ²³ In contrast, the inflammation found in the animals from the HFHF group not only was higher (reaching grade 2 in half of the animals) than that found in the C group, but also coincided with enhanced hepatocyte ballooning degeneration, which is considered a marker of NAFLD progression to more harmful stages. ²⁴ In addition, animals displayed elevated serum transaminase levels, a parameter recognized as an indicator of liver injury and commonly measured when studying NAFLD. ²⁵ Collectively, these findings provide compelling evidence that a dietary NAFLD model was successfully established.

Based on the obtained results, the administration of the tested treatments (viable and heat‐inactivated L. rhamnosus GG) partially prevented the increase in body weight, without affecting food intake, as reported previously in this same cohort of rats. ¹⁹ These results are in accordance with those reported by other authors in both murine models and human studies. ²⁶ , ²⁷ , ²⁸ As for the absence of change induced by either treatment in liver weight, it is worth mentioning that there is no consensus regarding the effects of probiotic or paraprobiotic administration on this parameter. While some studies have reported a decrease in liver weight and TG content following probiotic and paraprobiotic administration, ²⁹ other authors have only corroborated the effects of such treatments in preventing liver lipid accumulation, without affecting organ weight, ³⁰ , ³¹ as observed in the current study. In addition, the administration of the probiotic also resulted in the amelioration of further markers of liver alteration, as demonstrated by the histologic analysis (such as grade of steatosis and inflammation). These results align with findings reported by other authors, who described a reduction in the degree of liver steatosis and/or inflammatory cell infiltration in histologic examination using similar experimental models involving rodents fed steatotic diets. ³² , ³³ , ³⁴ By contrast, the ballooning degeneration, characterized by swelling and rounding up of hepatocytes, observed in the HFHF group, ³⁵ was not prevented by probiotic administration. Unfortunately, to the best of our knowledge, no studies have reported the impact of probiotic administration on this marker of liver injury. Regarding inactivated probiotic administration, although a partial prevention of the aforementioned markers was observed with this intervention, it did not reach statistical significance. In this context, the obtained results are consistent with those published by Jang et al., ³⁶ who observed that the preventive effect of formalin‐killed L. rhamnosus GG on diet‐induced steatosis (determined by NAS score calculation) was lower compared to the effect exerted by its viable counterpart. It is indeed interesting how the effects of both treatments in preventing diet‐induced excessive liver lipid accumulation contrast with the previously obtained results in this same cohort of rats regarding diet‐induced liver oxidative stress, which showed that the administration of the paraprobiotic resulted more effective than that of the probiotic. ¹⁹ Regarding liver transaminases, the elevation in serum AST and ALT levels induced by the high‐fat high‐fructose diet was not influenced by either of the tested treatments. These results are in line with those reported in previous studies using different probiotic strains (including L. rhamnosus GG), where no changes in these markers were reported following probiotic administration. ⁴ , ³⁷ Considering the absence of significant differences among the groups fed with the high‐fat high‐fructose diet (HFHF, PRO, and PARA groups) regarding food intake, the cecum lipid content was also assessed to determine whether the tested treatments might influence dietary lipid uptake at the intestinal level. As expected, the elevated consumption of dietary fat resulted in an inability of the animals to absorb it correctly, ³⁸ and thus, the cecum lipid content found in the non‐treated animals subjected to the high‐fat high‐fructose diet alone surpassed that of rats on the standard diet. The lack of effect observed in the animals administered the probiotic on this parameter is in line with findings from other studies available in the literature using other probiotic strains. ³⁹ , ⁴⁰ By contrast, paraprobiotic administration resulted in a notable increase in cecum lipid content, reaching statistical significance in comparison to the HFHF group. These results are in good accordance with existing literature on the impact of inactivated bacteria on intestinal lipid absorption. In this context, a recent study (based on the INFOGEST in vitro method) demonstrated that heat‐inactivated L. casei effectively decreased the intestinal lipid digestion, ⁴¹ thus increasing their excretion. Moreover, Ting et al. ⁴² observed a higher lipid cecum content (TGs and cholesterol) in animals receiving heat‐killed L. reuteri GMNL‐263.

Based on the results regarding the efficacy of both treatments in preventing liver lipid accumulation and improving histological markers (with the probiotic proving more effective than the paraprobiotic), alongside the observed enhancement of cecum lipid content by the paraprobiotic, the next step of this research sought to elucidate the potential mechanisms of action in terms of hepatic lipid metabolism. Thus, these mechanisms of action were clustered in two groups: those involved in hepatic lipid “input” (including hepatic lipid uptake, de novo lipogenesis, and TG assembly) and those implicated in hepatic lipid “output” (including FA oxidation, lipolysis, and hepatic TG release).

Regarding liver lipid “input,” it was observed that animals solely fed the high‐fat high‐fructose diet exhibited increased circulating TG levels and liver FATP2 expression, without changes in CD36 expression, suggesting that the higher hepatic fat accumulation found in these animals may be derived (at least partially) from a greater FA uptake from the bloodstream. In this regard, it is important to note that while FATP2 is considered a major FA transporter in the liver for its high expression in this organ, ⁴³ the role of CD36 in liver FA uptake appears to be relevant only in situations of obesity. ⁴⁴ Interestingly, probiotic administration significantly reduced hepatic FATP2 protein expression, an effect that could explain at least partially the lower TG content found in the livers of these animals compared to the rats in the HFHF group. In fact, it has been reported that the deletion of FATP2 in mice leads to an improvement of diet‐induced hepatic steatosis. ⁴⁵ However, there are also studies where no such reduction in liver FATP2 protein expression was reported following probiotic administration. In any case, this apparent discrepancy could potentially arise from differences in study design, such as the experimental period length (6 weeks vs. 12 weeks), the selected probiotic strain (L. rhamnosus GG vs. L. plantarum ZJUIDS14) or the animal model employed (Wistar rats vs. C57BL/6 mice). ⁴⁶ In the case of the animals receiving the paraprobiotic, although no changes were found in FATP2, a significant reduction in liver CD36 protein expression was observed. In this scenario, recent studies have highlighted some controversy regarding the role of CD36 in the development and progression of NAFLD. Thus, while higher CD36 levels have frequently been reported in patients with NAFLD, ⁴⁷ increased hepatic steatosis has also been described in mice with CD36 deletion. ⁴⁸ , ⁴⁹ , ⁵⁰ Moreover, individuals with CD36 deficiency, particularly those with Asian and African descent, often exhibit a predisposition to various metabolic disturbances, including NAFLD. ⁵¹ Therefore, considering that the most substantial reduction in liver lipid content observed in the current study (induced by probiotic administration) occurred concomitantly with a significant decrease in liver FATP2 expression, it appears that the significance of CD36 is not as relevant regarding liver lipid uptake.

Besides liver lipid uptake, de novo lipogenesis and hepatic TG assembly can also be considered as processes involved in liver lipid “input.” Notably, the utilization of a high‐fat high‐fructose diet resulted in an elevated consumption of fructose throughout the whole experimental period, a sugar recognized for its pronounced lipogenic properties. ⁵² In this context, the high‐fat high‐fructose diet induced a significant increase in the activity of key enzymes involved in de novo lipogenesis (ACYL and FAS), which is in line with the results reported by other authors. ⁵³ Regarding probiotic administration, this intervention did not induce significant changes in the activities of the aforementioned enzymes, although a sharp decrease in FAS activity (−43.75%) was observed. Interestingly, paraprobiotic administration resulted more effective than its probiotic of origin in reducing FAS activity, reaching values similar to those found in the C group. A similar pattern was also observed regarding SREBP‐1c protein (the transcriptional factor regulating FAS), which exhibited partial or complete reduction following probiotic and paraprobiotic administration, respectively. Some authors have reported no changes on hepatic ACYL mRNA expression after probiotic (a blend of B. subtilis mixture and L. plantarum LP1) or paraprobiotic (heat‐killed Enterococcus faecalis FK‐23, 110°C for 10 min) administration. However, no data regarding its activity were provided. ⁵⁴ , ⁵⁵ Significant reductions in the protein expression of FAS, along with decreased protein expression of SREBP‐1c, have been consistently reported in various studies using murine models undergoing probiotic administration. ³⁷ , ⁵⁶ , ⁵⁷ Despite the apparent discrepancy between the results reported in the current research and those in the literature, it is noteworthy that the studies cited above did not measure FAS activity, which is a more accurate indicator of its functionality than protein expression alone. As for the effects observed in the group treated with the paraprobiotic, the findings from this study regarding FAS activity are in good accordance with the still limited available experimental evidence using inactivated bacteria. ⁵⁸ , ⁵⁹ , ⁶⁰

The last process that could be considered within hepatic lipid “input” involves TG assembly. In this investigation, feeding animals a high‐fat high‐fructose diet tended to decrease DGAT2 protein expression compared to the C group. Previous studies using a similar animal model have also demonstrated a trend for hepatic DGAT2 protein expression to decrease in animals fed diets rich in fructose, which seems to promote the accumulation of lipids in the liver primarily as diglycerides. ⁶¹ In the present study, none of the tested treatments were found to prevent the diminishment of DGAT2, although other treatments based on bioactive compounds, such as polyphenols, which prevent liver steatosis, have proven effective. ⁶² Noteworthy, a similar trend was observed in AQP9 protein expression, which facilitates glycerol uptake in the liver. These results suggest the possibility of a hepatic compensatory mechanism aimed at reducing liver TG assembly in animals receiving the high‐fat high‐fructose diet in order to avoid an even greater TG accumulation in the liver. Indeed, similar findings have been previously reported in both animal and human studies. ⁶³ , ⁶⁴

The obtained results indicate that, although both treatments prevented alterations in key markers related to hepatic lipid “input,” the mechanisms of action seem to be different. While the probiotic reduced fatty‐acid import into the liver (FATP2) along with milder effects on de novo lipogenesis markers (FAS and SREBP‐1c), the paraprobiotic exerted a stronger effect on the latter, and was able to decrease intestinal fat absorption (as indicated by increased lipid cecum content).

Concerning liver lipid “output,” the consumption of fructose within a high‐fat diet has been shown to reduce the oxidation of dietary fat through different mechanisms, ⁶⁵ which could contribute to higher intrahepatic fat accumulation derived from this dietary pattern. In the present study, the high‐fat high‐fructose feeding led to a notable reduction in the markers associated with the liver's ability to cope with an increased lipid load (FA β‐oxidation), such as CPT‐1a and AMPK. Moreover, the high‐fat high‐fructose diet also resulted in the downregulation of key mitochondrial biogenesis markers such as SIRT1, PGC‐1α, and TFAM. These results suggest that the lower liver lipid oxidation observed in animals fed the high‐fat high‐fructose diet may be related to an impairment in the hepatic mitochondrial synthesis machinery. In fact, similar findings have been documented by other authors who fed rodents with diets rich in fat and/or fructose, observing decreased liver FA oxidation and mitochondrial dysfunction. ⁶⁶ , ⁶⁷ , ⁶⁸ The effects induced by the high‐fat high‐fructose diet feeding on the aforementioned parameters where not prevented by any of the tested interventions. Specifically, the lack of effect observed following probiotic administration contrasts with findings from other authors, who described upregulation of energy‐yielding pathways (mainly increased CPT‐1a protein expression/activity and AMPK phosphorylation) in the livers of rodent models of NAFLD subjected to similar interventions. ⁵⁶ , ⁶⁹ , ⁷⁰ , ⁷¹ This apparent discrepancy between the results obtained in this study and those previously published, may be attributed to differences in the experimental models used (C57BL/6N mice in some cases), the selected probiotic (some studies used a mixture of probiotic strains), the administered dose (up to 1 × 10¹⁰ CFU/day), and the duration of the experimental period (up to 42 weeks of treatment in some studies). As for the paraprobiotic administration, although the available literature in this field is still limited, current results indicate that the administration of heat‐killed bacteria to rodents with diet‐induced NAFLD leads to greater activation of AMPK. However, it must be noted that these results have been observed in mice (which tend to respond better to certain treatments than rats) fed diets rich in fat (but not fructose), receiving different paraprobiotics (E. faecalis EF‐2001 and Lactiplantibacillus plantarum K8), and using different experimental durations (up to 14 weeks). ⁷² , ⁷³ Overall, the results obtained in the present study regarding the analyzed mechanisms of action suggest that, under these experimental conditions, the effects of probiotic or paraprobiotic administration preventing liver lipid accumulation are not mediated by changes induced in pathways related to FA oxidation or mitochondrial biogenesis.

Another process that may be considered within liver lipid “output” is lipolysis. In this regard, although no differences were found between the HFHF and C groups, probiotic administration tended to increase the activation of HSL compared to the non‐treated group, potentially leading to enhanced TG mobilization from lipid droplets. While there is limited literature on the effects of probiotics on the activation of hepatic HSL, its mRNA expression has been shown to increase under probiotic administration. ⁷⁴ , ⁷⁵ By contrast, no changes were observed in this parameter in the group treated with inactivated bacteria compared to the HFHF group. In this regard, the absence of studies thus far measuring HSL activation after administration of nonviable bacteria, the results of the current study are in line with those reported by Caimari et al., ⁷⁶ where no changes in HSL mRNA expression were found in rats (Wistar) receiving a different strain of heat‐killed bacteria (B. animalis subsp. lactis CECT 8145) alongside an obesogenic diet (cafeteria diet).

Finally, the activity of MTTP, an enzyme involved in releasing liver TG into the bloodstream, was measured. The administration of viable bacteria induced a higher enzymatic activity of MTTP compared to the other experimental groups (C, HFHF, and PARA). This elevation in MTTP activity could potentially account, at least in part, for the observed effect on the prevention of liver lipid accumulation in the animals of the PRO group. Indeed, an enhanced release of lipids from the liver into the bloodstream cannot be discarded as a potential mechanism for alleviating intrahepatic fat accumulation, as described elsewhere. ⁷⁷ Unfortunately, to the best of our knowledge, no studies have measured MTTP enzymatic activity following the administration of viable or inactivated bacteria. Therefore, further research is required to elucidate the real impact of these kinds of treatments on liver TG release.

Based on the discussed results, it appears that under these experimental conditions, the capacity of probiotic administration to prevent diet‐induced excessive liver lipid accumulation is primarily mediated by reduced lipid uptake (decreased FATP2 expression) and enhanced TG release from the liver into circulation (increased MTTP activity). In contrast, the involvement of further metabolic pathways, such as de novo lipogenesis or liver lipid oxidation, does not seem to be as relevant (Figure 12). In the case of the paraprobiotic administration, not only was the overall effect on liver lipid accumulation lower compared to animals receiving the probiotic, but the involved mechanisms of action also appear to differ. In this regard, the effects stemming from paraprobiotic administration are associated with the downregulation of de novo lipogenesis (lower SREBP‐1c expression and FAS activity) and TG assembly (DGAT2 and AQP9 expression), coupled with decreased intestinal lipid absorption (Figure 12).

FIGURE 12.

Schematic representation of the mechanisms of action regulated by probiotic or paraprobiotic administration in the prevention of diet‐induced hepatic lipid accumulation. ALT: alanine aminotransferase; AST: aspartate aminotransferase; FAS: fatty‐acid synthase; FATP2: fatty‐acid transport protein 2; HFHF: high‐fat high‐fructose; LPL: lipoprotein lipase; MTTP: microsomal triglyceride transfer protein; SIRT3: NAD‐dependent deacetylase sirtuin‐3; PARA: paraprobiotic; PRO: probiotic; SREBP‐1c: sterol regulatory element‐binding protein 1; TG: triglyceride; VLDL: very‐low‐density lipoprotein.

Interestingly, the higher MTTP activity observed in the PRO group was not correlated with higher TG serum levels compared to the non‐treated group fed with the same high‐fat high‐fructose diet, suggesting potential involvement of extrahepatic tissues in their clearance from bloodstream. To gain a better understanding of this phenomenon, different analyses were conducted on skeletal muscle tissue samples. In this regard, a trend towards decreased LPL activity was observed in the HFHF group compared to the C group. This effect was partially prevented by probiotic administration (Supp. Figure 1), suggesting that the rate of FA incorporation into the muscle in animals receiving the viable bacteria tended to be higher than in the non‐treated animals. These FAs may be directed to oxidative pathways, especially considering findings from a previous study conducted in this precise cohort of rats, where probiotic administration moderately increased CPT‐1b activity in skeletal muscle compared to animals in the HFHF group. ¹⁸ Nevertheless, this moderate increase in oxidation was not associated with a reduction in skeletal muscle TG content, which was similar between the PRO and the HFHF groups. Based on these observations, it seems that the lower serum TG levels found in animals receiving the probiotic, (despite a greater lipid release from liver to circulation) compared to the HFHF group, could be mediated by a higher FA uptake by skeletal muscle tissue. For animals receiving the paraprobiotic, no changes were observed in skeletal muscle LPL activity compared to the HFHF group, which could explain the lack of differences in serum TG levels between these two groups. On the contrary, animals receiving the paraprobiotic exhibited a tendency toward higher SIRT3 protein expression (Supp. Figure 2), a mitochondrial deacetylase that regulates the activity of enzymes in the Krebs cycle and FA oxidation. ⁷⁸ Additionally, higher CPT‐1b activity (as reported elsewhere) was observed in the animals receiving the paraprobiotic compared to the HFHF group in the same tissue. ¹⁸ These events may explain the decreased skeletal muscle TG content found in the PARA group in comparison to the HFHF group. These findings suggest that the mechanisms of action underlying TG uptake from the bloodstream and its management in skeletal muscle seem to differ among the treated groups.

5. CONCLUSIONS

In conclusion, the findings of this study demonstrate that, under the present experimental conditions, the effects induced by the administration of viable L. rhamnosus GG preventing liver lipid accumulation in rats fed a high‐fat high‐fructose diet differs from those induced by its heat‐inactivated paraprobiotic counterpart. Indeed, the magnitude of the overall effect resulting from the probiotic administration in preventing liver lipid accumulation was bigger than that induced by its heat‐inactivated paraprobiotic. Moreover, the involved mechanisms of action seem to differ between both treatments. These findings, coupled with prior observations in this cohort of rats regarding hepatic oxidative stress suggest that, depending on the nature of the liver injury, the viability of the bacteria appears to be a parameter that may significantly influence the potential outcome of the intervention. Nevertheless, as data regarding the effects of the tested interventions (particularly in the case of paraprobiotic administration), on certain mechanisms of action discussed in this article are yet to be published, further research is warranted to validate the findings presented in this study.

AUTHOR CONTRIBUTIONS

Iñaki Milton‐Laskibar, J. Alfredo Martínez, and María P. Portillo: Conceptualization. Laura Isabel Arellano‐García, Iñaki Milton‐Laskibar, and Miguel Arán‐González: Investigation. Laura Isabel Arellano‐García and Iñaki Milton‐Laskibar: Formal analysis. Laura Isabel Arellano‐García, Iñaki Milton‐Laskibar, J. Alfredo Martínez, and María P. Portillo wrote and edited the original paper; and all the authors reviewed and edited the original paper.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

Supporting information

DATA S1: Supporting Information.

Arellano‐García LI, Milton‐Laskibar I, Martínez JA, Arán‐González M, Portillo MP. Comparative effects of viable Lactobacillus rhamnosus GG and its heat‐inactivated paraprobiotic in the prevention of high‐fat high‐fructose diet‐induced non‐alcoholic fatty liver disease in rats. BioFactors. 2025;51(1):e2116. 10.1002/biof.2116