|Home | About | Journals | Submit | Contact Us | Français|
Alcohol abuse is the leading etiologic factor of pancreatitis, although many heavy drinkers do not develop pancreatic damage. Alcohol promotes pancreatitis through a combination of remote (e.g., increased gut permeability to bacterial products such as lipopolysaccharide) and more proximal effects (e.g., altered pancreatic cholinergic inputs), including oxidative damage at the level of the pancreatic acinar cell. Recent evidence indicates that alcohol exposure to rodents disturbs proteostasis in the exocrine pancreas, an effect counterbalanced by homeostatic processes that include both the unfolded protein response (UPR) and autophagy. A corollary to this notion is that pancreatitis results when adaptive responses are insufficiently robust to alleviate the cellular stress caused by alcohol.
Disturbances in endoplasmic reticulum function, known as ER stress, cause accumulation of unfolded/misfolded proteins and trigger UPR as a protective cellular response. The UPR comprises a complex signaling system that coordinately regulates ER protein folding and trafficking, ER protein degradation (ERAD) and programmed cell death. A key regulator of UPR in the exocrine pancreas is X-Box binding protein-1 (XBP1), a transcription factor that regulates gene products involved in oxidative protein folding, ERAD, lipid synthesis and expansion of the ER-Golgi network.
We recently investigated ethanol effects on ER homeostasis and UPR activation in the pancreatic acinar cell, a professional protein secretory cell inherently susceptible to ER stress. We fed wild-type and constitutive XBP1 deficient (Xbp1+/−) mice either a control diet or an isocaloric ethanol- containing diet for four weeks using the Tsukamoto-French model. We found that although in wild-type mice ethanol feeding induces aberrant oxidative changes and mild ER stress, it does not cause pancreatic pathology. Under these stress conditions, we found marked activation of XBP1. In Xbp1+/− mice we found increased sensitivity to ethanol-induced pancreatic ER stress, marked activation of the PERK-eIF2α UPR branch and induction of ATF4 and CHOP relative to wild-type mice. Ethanol feeding selectively reduces in Xbp1+/− mice pancreatic expression levels of Bcl-2, ER oxidoreductases and the ERAD regulator EDEM1. Also present are histological features commonly found in pancreatitis: fewer secretory zymogen granules, accumulation of autophagic vacuoles in acinar cells and increased acinar cell death. Exploiting circumstances in which XBP1 activity is insufficient to reestablish ER function, our model established conditions in which ethanol feeding revealed the early stages of pathological change. These observations provide the first evidence that ethanol contributes to the onset of pancreatitis via altered UPR.
Much evidence supports a crosstalk between the UPR and autophagic pathways, with macroautophagy acting as a protein degradation system for misfolded proteins and damaged ER. Human acute pancreatitis is associated with profound accumulation of autophagic vacuoles in acinar cells. Moreover, ethanol feeding to rats promotes the appearance of autophagic vacuoles in pancreatitis models, an effect attributable, at least in part, to lysosomal dysfunction and LAMP-2 depletion. Whether this autophagic activity is linked to ER stress has not been established. Short-term ethanol treatment potentiated autophagy induced by ER stressors in isolated acinar cells (our unpublished observations). We evaluated the extent of autophagy in our mouse model of ethanol-induced ER stress by electron microscopy, LC3 immunostaining of pancreatic tissue sections and western blot analysis of autophagic markers. Autophagic vacuoles are rare in pancreatic acinar cells from wild-type or Xbp1+/− mice on control diet, and in ethanol-fed wild-type mice. In contrast, pancreas from ethanol-fed Xbp1+/− mice exhibit a marked increase in LC3-labeled autophagosomes, confirmed by LC3-II/LC3-I ratios and electron microscopy studies. That ethanol feeding reduces pancreatic LAMP-2 levels to a similar degree in both wild-type and Xbp1+/− mice indicates that LAMP-2 depletion by itself is insufficient to account for the observed accumulation of autophagic vacuoles in ethanol-fed Xbp1+/− mice.
Figure 1 depicts a model accounting for the engagement of autophagy in our mouse model of ethanol-induced acinar cell pathology. In wild-type mice (left side), ethanol-induced oxidative ER stress is alleviated. Activation of the IRE1-XBP1 pathway induces gene expression to control ER redox status, maintain efficient protein folding and clear unfolded/misfolded proteins by ERAD and, likely, basal autophagy. In ethanol-fed Xbp1+/− mice (right side), autophagic vacuoles accumulate due to: (1) increase in autophagic activity due to ER disturbances and impaired ERAD secondary to XBP1 deficiency and (2) blockade of the autophagic flux linked to ethanol-induced LAMP-2 depletion and lysosomal dysfunction. It appears plausible that the combination of ethanol feeding and XBP1 deficiency promotes intermediate stages of autophagy, yet simultaneously impedes its resolution. As yet, it remains uncertain whether vesicular compartments containing LC3-II puncta are unable to fuse with lysosomes or successfully fuse with functionally impaired lysosomal compartments.
This model is consistent with previous reports showing that EDEM depletion secondary to XBP1 deficiency and impairment of the ERAD and ubiquitin-proteasome systems (as likely occur in ethanol-fed Xbp1+/− mice) are robust stimuli for autophagy. In this respect, the more numerous autophagic vacuoles found in ethanol-fed Xbp1+/− mice may reflect a compensatory response to assist the defective ERAD in removing aberrant proteins and dysfunctional ER. Future studies to examine whether autophagy is selectively engaging damaged ER or involves nonselective autophagy in our in vivo model are warranted.
Complementary mechanisms stimulating autophagy may include reduced Bcl-2 levels and activation of PERK-ATF4 programs in ethanol-fed Xbp1+/− mice. Under normal conditions, Bcl-2 binds to Beclin 1 in an autophagy-inhibitory complex. Thus, excess Bcl-2 turnover in ethanol-fed Xbp1+/− mice releases Beclin 1 to engage in autophagy induction. ATF4 in our model may facilitate autophagy by upregulating LC3 expression through binding to the LC3 promoter, as shown previously.
Also to be resolved is how long-term ethanol feeding alters lysosomal function. In vitro studies indicate that fatty acid ethyl esters, nonoxidative ethanol metabolites, weaken lysosomal membranes, possibly compromising autolysosome formation. Alternatively, ethanol-induced ER stress may further hinder lysosomal function by altering the folding and trafficking of lysosomal proteases and membrane proteins such as LAMP-2. In ethanol-fed animals, elevated blood lipopolysaccharide may also promote LAMP-2 depletion. Moreover, the endotoxemic effect of alcohol abuse may be exacerbated in Xbp1+/− mice, since XBP1 deficiency has been linked to impaired intestinal mucosal defenses.
Future studies must determine whether autophagic activity in ethanol-fed Xbp1+/− mice is beneficial or contributes to the acinar cell pathology. Although we currently envision autophagy as a protective adaptation to ER dysfunction, the death of acinar cells unable to efficiently resolve autophagy eventually leads to tissue injury and pancreatitis.
We gratefully thank L. Glimcher and A.H. Lee for providing the Xbp1+/− mice. We acknowledge the Animal and Morphology Cores of the Research Center for ALPD & Cirrhosis (P50-A11999). Supported by NIAAA (R21AA016010) and NCCAM (1P01AT003960).
Punctum to: Lugea A, Tischler D, Nguyen J, Gong J, Gukovsky I, French SW, et al. Adaptive unfolded protein response attenuates alcohol-induced pancreatic damage. Gastroenterology. 2011;140:987–997. doi: 10.1053/j.gastro.2010.11.038.