Potential conflict of interest: Nothing to report.
The prevalence of the metabolic syndrome and nonalcoholic fatty liver disease (NAFLD) in humans increases with age. It is unknown whether this association is secondary to the increased incidence of risk factors for NAFLD that occurs with aging, reflects the culmination of years of exposure to lifestyle factors such as a high-fat diet (HFD), or results from physiological changes that characterize aging. To examine this question, the development of NAFLD in response to a fixed period of HFD feeding was examined in mice of different ages. Mice aged 2, 8, and 18 months were fed 16 weeks of a low-fat diet or HFD. Increased body mass and insulin insensitivity occurred in response to HFD feeding irrespective of the age of the mice. The amount of HFD-induced hepatic steatosis as determined biochemically and histologically was also equivalent among the three ages. Liver injury occurred exclusively in the two older ages as reflected by increased serum alanine aminotransferase levels, positive terminal deoxynucleotide transferase–mediated deoxyuridine triphosphate nick end-labeling, and caspase activation. Older mice also had an elevated innate immune response with a more pronounced polarization of liver and adipose tissue macrophages into an M1 phenotype. Studies of cultured hepatocytes from young and old mice revealed that aged cells were selectively sensitized to the Fas death pathway. Conclusion: Aging does not promote the development of hepatic steatosis but leads to increased hepatocellular injury and inflammation that may be due in part to sensitization to the Fas death pathway and increased M1 macrophage polarization. (HEPATOLOGY 2013)
Nonalcoholic fatty liver disease (NAFLD) represents a continuum from simple hepatic steatosis alone to steatosis together with hepatitis and eventual fibrosis. 1 The mechanisms underlying the development of this disease remain unclear, but NAFLD is considered the hepatic manifestation of the metabolic syndrome and therefore linked to the development of obesity and insulin resistance. 2 The prevalence of the metabolic syndrome increases markedly with aging. 3 In 20- to 29-year-olds the prevalence is 7% but rises to 44% in individuals aged 60-69 years. 3 Similarly aging has also been shown to be a risk factor for the development of human NAFLD, 4, 5 although some studies have suggested that this effect is limited to females. 6, 7 More conclusive of an effect of aging on NAFLD is evidence that the prevalence of nonalcoholic steatohepatitis (NASH) and chronic liver disease is increased in the aged. Thus, steatohepatitis 8 and fibrosis 9, 10 are increased with aging leading to a higher mortality in aged individuals with NAFLD. 11, 12
Aging may promote the development of NAFLD through several mechanisms. First, aging promotes the onset of obesity and diabetes, and the increase in these known risk factors for NAFLD may account for the higher prevalence of NAFLD in the aged. Supportive of this possibility is that in parallel with steep rises in obesity and insulin resistance in the pediatric population, problems that had formerly been restricted to the elderly, there has been a dramatic increase in the incidence of NAFLD in young individuals. Second, the increase in disease with age may result from the cumulative effects over many years of lifestyle or environmental factors, such as the overconsumption of a Western diet. Finally, physiological changes inherent to the process of aging may trigger or promote the development of components of the metabolic syndrome later in life. For example, function of the lysosomal degradative pathway of autophagy decreases with aging, 13 and adequate levels of hepatic autophagy are critical to limit lipid accumulation in the liver. 14, 15 Another possibility is that aging is associated with increased oxidative stress, 16 which has also been implicated as a mediator of NAFLD. 17 Several of the known—or even currently unknown—physiological changes that occur with aging may therefore lead to NAFLD.
To begin to examine the potential effects of aging on NAFLD development, we challenged mice of different ages with a defined period of high-fat diet (HFD) feeding to induce the metabolic syndrome and NAFLD. This design was intended to ask whether the aging state altered the development of steatosis and hepatitis in response to a defined period of dietary factors that induce NAFLD. Age failed to affect the development of metabolic abnormalities such as weight gain and insulin resistance. The amount of steatosis was similarly unaffected by age. However, aged mice developed more liver injury and a greater inflammatory response from a HFD. Studies in cultured hepatocytes revealed that cells from aged mice were sensitized to death from stimulation of the Fas death pathway. Aging therefore does not affect the development of simple steatosis but promotes the progression to steatohepatitis.
Male C57BL/6 mice aged 2, 8, and 18 months were obtained from the aging colony maintained by the National Institute of Aging (Bethesda, MD). Mice were maintained under 12-hour light/dark cycles with unlimited access to food and water. Mice were fed either a low-fat diet (LFD) (10% kcal supplied by fat; Research Diets #D12450B, New Brunswick, NJ) or a HFD (60% kcal supplied by fat; Research Diets #D12492) for 16 weeks prior to sacrifice. All studies were approved by the Animal Care and Use Committee of the Albert Einstein College of Medicine and followed the National Institutes of Health guidelines for animal care.
Livers were fixed in 10% neutral formalin, and 6 μm sections were stained with hematoxylin and eosin. Tissue sections were examined in a blinded fashion by a single pathologist and graded for the degree of hepatic steatosis. Hepatocellular steatosis was semiquantitatively graded on the following sliding scale: 0, absent; 0.5, minimal; 1, mild; 1.5, mild to moderate; 2, moderate; 2.5, moderate to marked; 3, marked.
Serum glucoses were assayed with an Ascensia Contour glucose meter (Bayer HealthCare, Mishawaka, IN). Serum insulin levels were measured by radioimmunoassay as described. 18 Relative insulin sensitivity was determined by the homeostasis model assessment of insulin resistance (HOMA) as described. 19 Alanine aminotransferase (ALT) levels were measured using a commercial kit (TECO Diagnostics, Anaheim, CA).
Hepatic triglyceride content was determined using a commercial kit according to the manufacturer's instructions (Sigma, St. Louis, MO). Triglyceride values were normalized to liver weight.
Terminal Deoxynucleotide Transferase–Mediated Deoxyuridine Triphosphate Nick End-Labeling Assay.
The numbers of terminal deoxynucleotide transferase–mediated deoxyuridine triphosphate nick end-labeling (TUNEL)-positive cells were determined in liver sections with the DeadEnd Colorimetric System (Promega, Madison, WI). Tissue sections were deparaffinized in xylene, rehydrated in decreasing concentrations of ethanol, and the assay performed according to the manufacturer's instructions. Under light microscopy, the number of TUNEL positive cells in 20 randomly selected high power fields was determined as a percentage of the total number of hepatocytes.
Total liver protein isolation and Western blotting was performed as previously described. 19 Immunoblots were probed with antibodies for caspase 3 and 7, Fas and tubulin (Cell Signaling, Beverly, MA).
RNA was isolated from livers or adipose tissue with RNeasy Plus (QIAGEN, Valencia, CA). Reverse transcription was performed with 1 μg of RNA in an Eppendorf Mastercycler (Hamburg, Germany) using a high-capacity complementary DNA reverse-transcription kit (ABI, Foster City, CA). Annealing of primers was performed at 25°C for 10 minutes, followed by elongation at 37°C for 2 hours and inactivation of the enzyme at 85°C for 5 minutes. Negative controls (no added transcriptase) were performed in parallel. Polymerase chain reaction (PCR) for tumor necrosis factor (TNF), monocyte chemoattractant protein 1 (MCP-1), F4/80, nitric oxide synthase 2 (NOS2), arginase 1 (Arg1), Fas, Fas ligand (FasL), and glyceraldehyde 3-phosphate dehydrogenase was performed in triplicate in a 7500 Fast Real-Time PCR System (ABI). Primers to these genes (Supporting Table 1) were purchased from Integrated DNA Technologies (Coralville, IA). PCR was performed using Power SYBR Green Master Mix (ABI). Taq polymerase was activated at 95°C for 10 minutes. The cycling parameters were denaturation at 95°C for 30 seconds and extension at 60°C for 1 minute (for 40 cycles). Data analysis was performed using the 2−ΔΔCT method for relative quantification. All samples were normalized to glyceraldehyde 3-phosphate dehydrogenase, which served as an endogenous control.
Primary Hepatocyte Isolation and Culture.
Primary hepatocytes from C57BL/6 mice 2 or 18 months of age were obtained by collagenase perfusion as described. 20 The cells were purified by Percoll (Sigma) density centrifugation and cultured in Roswell Park Memorial Institute 1640 medium with 5% fetal bovine serum, 1.7 μM insulin, 1 μM dexamethasone, and antibiotics on Biocoat plates (Becton Dickinson, Franklin Lakes, NJ). The medium was changed 3 hours after plating and again at 18 hours, at which time the experiments were started. Some of the cells were treated with 40-60 μM menadione (Sigma); 15 ng/mL TNF (R&D Systems, Minneapolis, MN) alone or after a 1-hour pretreatment with 50 nM actinomycin D (Sigma); 0.25 or 0.5 mM palmitate conjugated with bovine serum albumin (Sigma) as described; 21 or 25 or 100 ng/mL of Jo2 antibody (BD Pharmingen, San Diego, CA).
The percentage cell death in cultured hepatocytes was quantified by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. 22 At 24 hours after treatment with a death stimulus, the cell culture medium was replaced by an equal volume of a 1 mg/mL MTT solution (pH 7.4) in Roswell Park Memorial Institute 1640 medium. After incubation at 37°C for 1 hour, the MTT solution was discarded, the formazan product was solubilized with 1-propanol, and the absorbance of this compound was measured at 560 nm in a plate reader. The percentage cell death was determined by dividing the optical density of the treated group by the optical density for untreated control cells, multiplying by 100, and subtracting that number from 100.
Numbers of apoptotic and necrotic hepatocytes were determined by fluorescence microscopy of acridine orange and ethidium bromide–costained cells as described. 23 Cells with fragmented or condensed nuclei and a shrunken cytoplasm by acridine orange staining were considered apoptotic. Necrotic cells were detected by positive staining with ethidium bromide. A minimum of 300 cells per sample were examined, and numbers of apoptotic and necrotic cells expressed as a percentage of the total number of cells.
Numerical results are reported as the mean ± SE and represent at least three independent experiments. Control and experimental groups were compared using an unpaired Student t test with statistical significance defined as P < 0.05.
Effects of Aging on Body and Fat Mass and Insulin Sensitivity.
C57BL/6 mice were started on either a LFD or HFD at 2, 8, and 18 months of age. These ages were selected to reflect the normal physiology of young, middle-aged, and old mice. The mice were sacrificed for analysis after 16 weeks on their respective diets at the ages of 6, 12, and 22 months. Modest weight gain (19%-31%) occurred in LFD-fed mice, with the greatest increase occurring in the 6-month-old animals (Fig. 1A). The HFD induced a highly significant increase in body mass (71-86%) compared with LFD feeding in all three age groups (P<0.0002), with the largest increase again occurring in the young mice (Fig. 1A). The increase in body weight in 6-month-old mice on either diet was significantly greater than that in the 12-month-old mice, but did not reach statistical significance compared with the 22-month-old mice. All three groups of mice therefore had a marked weight gain that was not significantly affected by aging.
The effect of aging on white adipose tissue (WAT) mass was examined as WAT-produced factors contribute to NAFLD development. 24 Epididymal WAT mass in LFD-fed mice increased with age, but the differences did not reach statistical significance (Fig. 1B). Significant increases in WAT mass did occur with HFD feeding in all three age groups (P < 0.001), and the increase was significantly greater in the 22-month-old animals (Fig. 1B). Highly metabolic brown adipose tissue, which may be protective against obesity and its complications, 25 was unaffected by aging in LFD-fed animals (Fig. 1C). Brown adipose tissue mass increased significantly in all age groups with HFD feeding (P < 0.005), with the increase in 12-month-old mice being significantly greater (Fig. 1C).
Hyperglycemia and hyperinsulinemia occurred with HFD feeding in all three age groups. Serum glucoses were significantly increased (P < 0.03) in HFD-fed compared with LFD-fed mice at all ages (Fig. 1D). Although serum insulins were significantly elevated in each HFD-fed age group, the difference from the LFD-fed mice was only significant at 12 months (P < 0.004) (Fig. 1E). Relative insulin sensitivity was determined via HOMA. HOMA values were increased in all HFD-fed groups of mice, although none reached statistical significance over LFD-fed mice (Fig. 1F). Thus, the degree of metabolic abnormalities such as obesity and insulin insensitivity resulting from HFD feeding were relatively uniform among young, middle-aged, and old mice.
Steatohepatitis but Not Simple Steatosis Is Increased with Aging.
The effects of aging were examined on the development of NAFLD, the hepatic manifestation of the metabolic syndrome. Liver weight increased significantly (P < 0.02) in all age groups with HFD feeding, and the increase was significantly greater in 12- and 22-month-old mice (Fig. 2A). However, the increases in liver weight with a HFD paralleled those in whole body mass, and the liver/body weight ratios were not significantly changed with diet or aging (Fig. 2B). Hepatic triglyceride content increased significantly (P < 0.02), but to an equivalent extent in all three age groups with HFD feeding (Fig. 2C). This biochemical finding was confirmed by blinded grading of the histological degree of steatosis that similarly demonstrated significantly increased steatosis with HFD feeding for all ages (P < 0.01), but no significant difference among the three age groups (Fig. 2D). Thus, levels of hepatic steatosis induced by 16 weeks of HFD were equivalent irrespective of age, consistent with the metabolic findings of equal body weight gain and levels of insulin insensitivity in these animals.
The effect of aging on steatohepatitis was also assessed. Serum ALT levels were significantly elevated by HFD feeding in the 12- and 22-month-old mice (P < 0.02), but not in the 6-month-old group (Fig. 3A). ALT levels in the HFD-fed older mice were also significantly increased compared with 6-month-old HFD-fed animals (Fig. 3A). As an additional measure of hepatic injury, the numbers of TUNEL-positive cells in the livers were determined. With HFD feeding, the numbers of TUNEL-positive cells increased significantly in all three age groups (P < 0.001) (Fig. 3B). However, the numbers of TUNEL-positive cells in the 12- and 22-month age groups were significantly increased compared with the levels in the 6-month-old HFD-fed animals (Fig. 3B,C-F). Additional evidence of increased apoptosis was provided by the finding of the active, cleaved forms of the effector caspases 3 and 7 in the livers of 12- and 22-month-old HFD-fed mice via immunoblotting (Fig. 4). Thus, despite similar metabolic abnormalities and an equivalent degree of steatosis in the three age groups, significant liver injury was restricted to the older mice.
Aging Promotes a Proinflammatory Immune Response to HFD Feeding.
Inflammation in both liver and adipose tissue mediates NAFLD development, 26 suggesting that the ability of aging to promote a hyperresponsive inflammatory reaction may underlie the liver injury in older mice. The effects of aging on hepatic inflammation were assessed by relative levels of proinflammatory cytokine gene expression in whole liver. Hepatic levels of TNF messenger RNA (mRNA) were not significantly increased with HFD feeding compared with age-matched, LFD-fed mice because of increased TNF levels in older animals on a LFD (Fig. 5A). However, TNF levels were significantly increased in the HFD-fed 12- and 22-month-old mice compared with 6-month-old HFD-fed mice (Fig. 5A). Levels of MCP-1, a second proinflammatory cytokine implicated in NASH pathogenesis, 27 increased with HFD feeding in all ages but were not significantly different from LFD-fed mice or among ages (Fig. 5B).
To assess the effect of aging on activation of the hepatic macrophage innate immune response by HFD feeding, levels of the macrophage-specific marker F4/80 were determined in whole liver. Similar to the TNF findings, no significant change occurred between LFD- and HFD-fed mice of the same age, but levels of F4/80 were significantly higher in HFD-fed, older mice compared with young mice fed the HFD (Fig. 5C). The macrophage response in older mice was also shifted to a more proinflammatory M1 polarized phenotype as levels of the M1 marker NOS2 28 were significantly increased in the 12- and 22-month-old HFD-fed mice (Fig. 5D). In contrast, the M2 marker Arg1 28 was unaffected by diet or age (Fig. 5E). Along with the presence of liver injury, aged mice also had an increased M1 proinflammatory macrophage response in the liver.
Adipose tissue was similarly examined for an effect of aging on the immune response to the HFD. TNF levels in epididymal WAT were significantly increased (P < 0.05) in all three age groups with HFD feeding, but there was no significant difference among HFD-fed mice of different ages (Fig. 6A). Adipose tissue MCP-1 mRNA levels also increased significantly (P < 0.01) but equivalently in all three ages with HFD feeding (Fig. 6B). Consistent with the findings for these two proinflammatory cytokines, F4/80 mRNA levels were significantly and equally increased (P<0.01) in all HFD-fed groups (Fig. 6C). However, significant increases in NOS2 with HFD feeding only occurred in the aged mice (Fig. 6D), whereas significant increases in adipose Arg1 levels were present in all ages (Fig. 6E). These findings indicate that, similar to the liver, WAT in aged mice had a more pronounced M1 macrophage phenotype.
Aged Hepatocytes are Selectively Sensitized to Fas-Induced Cell Death.
In vivo studies demonstrated that aging sensitized hepatocytes to liver injury in the setting of steatosis in association with an increased inflammatory response. Prominent factors implicated in steatotic liver injury and inflammation include oxidant stress, 17 TNF, 29 saturated free fatty acids 30 and FasL. 31, 32 To determine whether aging may affect hepatocellular injury from any of these factors, we examined the sensitivity of cultured primary hepatocytes from young (2-month-old) and aged (18-month-old) mice to cell death from the oxidant menadione, TNF, the saturated fatty acid palmitate, and the Fas agonist antibody Jo2. Primary hepatocytes from young and old mice had equivalent amounts of cell death in response to superoxide-mediated oxidant stress from menadione (Fig. 7A). Hepatocytes from both ages of mice were resistant to cytotoxicity from TNF alone and were equally sensitive to death from actinomycin D and TNF cotreatment (Fig. 7B). Cell death from the toxic saturated free fatty acid palmitate was also equivalent between the two ages (Fig. 7C). In contrast to the findings with these three agents, aged hepatocytes were significantly more sensitive to death from the Fas agonist antibody Jo2 (Fig. 7D). To confirm that aged cells were sensitized to Fas-mediated death, the steady-state levels of apoptosis and necrosis after Jo2 treatment were determined by fluorescence microscopy of acridine orange/ethidium bromide–costained cells. Death was primarily apoptotic, and aged hepatocytes had significantly increased numbers of apoptotic and necrotic cells in response to Jo2 (Fig. 7E). To determine whether aging may have additionally sensitized the livers to Fas-mediated injury by up-regulating the levels of FasL or its receptor, levels of FasL and Fas mRNA in whole liver were examined using real-time PCR in HFD-fed mice. No significant change occurred in FasL or Fas mRNA content in liver with aging (Fig. 8A,B). Levels of Fas protein did increase in the liver with HFD feeding, as reported for mice fed a high carbohydrate diet, 31 but to a similar extent among the three ages of mice (Fig. 4). A possible mechanism for the development of hepatitis in older animals is therefore increased sensitivity of aged hepatocytes to Fas-dependent cell death.
Prevalence of the metabolic syndrome increases dramatically with aging. 3 Studies have also suggested that aging is associated with an increased prevalence of NAFLD, and in particular that the presence of steatohepatitis and mortality are increased in aged individuals with this disease. 4-12 This association may result from age-related increases in obesity and diabetes, cumulative effects on the liver of lifestyle factors such as diet, or the physiological changes that are part of aging. The present study examined these possibilities by determining whether aging affected NAFLD development in mice caused by a defined period of HFD feeding. Interestingly, features of the metabolic syndrome such as increased body mass and insulin insensitivity that developed from HFD feeding were equivalent in young, middle-aged and old mice. Thus, the physiological changes of aging failed to accelerate or worsen the development of two critical components of the metabolic syndrome, obesity and insulin resistance. These findings suggest that rather than age-related changes promoting the development of these metabolic abnormalities, it may be the cumulative length of time or frequency of exposure to factors such as a HFD that increase the prevalence of obesity and insulin insensitivity with aging.
Similar to the findings for obesity and insulin insensitivity, the degree of hepatic steatosis induced by the HFD was unaffected by age. This finding is not surprising in light of the similar extent of weight gain and insulin resistance, both of which are major risk factors for the development of steatosis. These two factors are thought to play a critical role in the excessive hepatic lipid accumulation that characterizes NAFLD by delivering increased amounts of serum free fatty acids to the liver that are then converted into excessive lipid stores. In contrast, significant steatohepatitis only occurred in older mice, demonstrating that aging does play a prominent role in liver injury. Significant hepatocyte injury and cell death as determined by increased ALTs and TUNEL positivity only occurred in the 12- and 22-month-old mice. Thus, factors unrelated to the mechanisms of excessive lipid accumulation with aging did promote the development of hepatocellular injury in the setting of steatosis.
Accompanying the liver injury in aged mice was an increased inflammatory response to HFD feeding. This finding is surprising in light of the well-established association of a depressed immune response with aging termed immunosenescence. 33 The increase in inflammation may be a secondary response to the liver injury. However, findings of a heightened immune response in adipose as well as liver tissue suggest a primary increase in inflammation in response to the HFD in the aged mice. Selective effects were noted with an increase in hepatic TNF but not MCP-1 levels with aging, and a generalized increase in both cytokines in the adipose tissue of all HFD-fed mice irrespective of age. A consistent finding in both tissues was an increased macrophage polarization to a proinflammatory M1 phenotype. This skewed polarization may explain in part the increased inflammatory response in aged animals and may be a mechanism of liver injury, as an unstrained M1 phenotype has been shown to mediate tissue injury in other organs. 34
The failure of young mice to progress to steatohepatitis within 16 weeks of HFD feeding contrasts with the increasing prevalence of NASH found in the human pediatric population. Although this inconsistency may simply reflect species differences between mice and humans, alternatively the finding supports the concept that additional factors unrelated to the direct effects of a HFD are necessary to trigger NASH in the young. For example, specific genetic factors may play a critical role in the progression to NASH in young individuals. Increasing evidence suggests that prenatal influences such as maternal imprinting from the effects of obesity and maternal diet promote NAFLD development in young rodents. 35, 36 The mice employed in the present investigations were the offspring of lean parents fed a normal chow diet and therefore lacked prenatal influences that cause the epigenetic effects necessary for NASH development at a young age. In contrast, in adults other factors may become more critical to promote the progression of HFD-induced steatosis to NASH, such as the propensity to develop an adipose tissue inflammatory response.
Cell culture studies of aged hepatocytes suggest that increased hepatocellular injury and death from stimulation of the Fas death pathway may be a mechanism of increased liver injury in aged mice. Aged hepatocytes did not exhibit a generalized susceptibility to injurious stimuli. Rather, aged cells were specifically sensitized to Fas death, as they had no increase in death from other possible mediators of liver injury in NAFLD, including oxidant stress, TNF, or the saturated free fatty acid palmitic acid. Hepatic Fas and FasL mRNA and Fas protein levels were not altered by aging, despite the increase in inflammation, indicating that receptor or ligand up-regulation did not contribute to the increase in liver injury and inflammation in aged mice. To our knowledge, these findings are the first to identify aging as a modulator of cellular sensitivity to Fas killing, and future studies must attempt to determine the mechanism of this aging effect.
The findings in this study demonstrate that the development of simple steatosis is unaffected by aging presumably because of similar metabolic changes in terms of weight gain and insulin sensitivity that ensue from a HFD irrespective of age. However, physiological changes inherent to aging may promote the development of liver injury and inflammation, and therefore the progression to chronic liver disease and mortality. A further understanding of the mechanisms of these effects may provide new insights into the pathogenesis of NAFLD and lead to specific therapies targeted for elderly individuals with this disease.