ALR is an enigmatic molecule that was originally found to augment partial hepatectomy-induced liver regeneration and to prevent PCS-induced hepatic pathology in animal models (1). Remarkable homology between ALR and ERV1 (33,34), a protein expressed by Saccharomyces cerevisiae that is essential for the survival of the yeast, and the abundance of ALR in normal quiescent hepatocytes (2) indicated that ALR may have physiological functions. Indeed, depletion of ALR from hepatocytes resulted in their apoptotic/necrotic death (3). Although ALR is secreted by hepatocytes constitutively, its rapid increase in serum after partial hepatectomy (2) suggested that increase in circulating ALR may be an indicator of hepatocyte stress/injury. In the present study, we used in vivo and in vitro models of liver or hepatocyte injury to examine whether increased ALR levels could be an alarm for ongoing hepatocyte stress/injury.
PCS that causes hepatocellular atrophy is accompanied by disruption of the rough endoplasmic reticulum and mitochondria due to diversion of the nutrient- and hormone-rich blood from the liver to systemic circulation (35,36). In rats, we found that PCS-induced liver atrophy and hepatocyte apoptosis is significant on d 4 after operation (20), which is consistent with the peak of pathological changes observed in dogs (35,36). Notably, serum ALR increased on d 4 after PCS. These pathological changes occur despite an immediate increase in hepatic levels of ALR as well as the mitogens HGF and TGF-α (20).
We sought to extend these findings and to demonstrate that the roles of ALR in inflammation are not restricted to a single species or a single form of acute inflammatory challenge, but rather that ALR production or release during inflammation is a general phenomenon. Hence, we used both mice and rats subjected to diverse inflammation paradigms. We also sought to leverage in silico modeling to better help define the role, if any, of ALR in acute inflammation. We found that the mRNA expression of inflammatory cytokines IL-6 and TNF-α, which are known to prime hepatic regeneration (12), were increased in approximately the same time frame as or after an increase in hepatic ALR. These results suggest that hepatic ALR may be induced along with Kupffer cell IL-6 and TNF-α after PCS (9) and that these mediators then function to limit injury and promote liver regeneration in conjunction with HGF and TGF-α. Indeed, there was no further decrease in the liver size or liver-to-body weight ratio up to 60 d after the initial atrophy up to 7 d after PCS (20). However, hepatocyte apoptosis peaked at 10 d and declined by 30 d but stabilized thereafter (20). Hepatic regeneration is a complex phenomenon involving induction of several protooncogenes, growth mediators and cytokines, including IL-6. Whereas we found that ALR stimulates IL-6 synthesis in Kupffer cells (9), stimulation of Foxa2 (HNF-3β)-transfected HepG2 hepatoma cells with IL-6 was reported to activate the ALR promoter (37), suggesting that cross-talk between IL-6 and ALR may have important implications in hepatic regeneration. Taken together, these findings suggest that the hepatic system attains equilibrium after PCS, establishing a balance between cell death and regeneration, with a plausible significant role of ALR in the latter.
Low-level hepatocyte apoptosis and modest increases in serum ALT levels occur in the established rat model of endotoxemia (38). Here, we found a rapid decrease in hepatic ALR with a concomitant increase in serum ALR on LPS administration. Interestingly, whereas hepatic ALR tended to normalize, serum ALR was still significantly elevated versus the basal level, even at 24 h. It is noteworthy that serum ALR increased before an increase in ALT, suggesting that increased ALR may be considered an indicator of ongoing hepatocyte stress/injury. In this regard, circulating ALR levels increase in patients with acute and fluminant hepatitis (39).
Induction of gram-negative bacterial sepsis by implantation of a fibrin clotcontaining E. coli was the third model of liver injury in which we determined ALR levels at various times. In this experimental model, increases in serum ALR were observed early and remained elevated, even at 24 h. The levels of inflammatory cytokines and NO were elevated at approximately the same time (IL-6) or later (TNF-α and NO) than ALR, in a manner consistent with known differences between true gram-negative sepsis and endotoxemia (40). Our data suggest that serum ALR is an early and reliable indicator of inflammation and liver damage in the setting of sepsis, although these findings are yet to be validated in clinical studies.
Finally, we used a mouse model of hemorrhagic shock to examine whether ALR could be an indicator of the severity of organ damage in shock and also to see whether the dynamics of circulating ALR could match those of any of the variables included in a previously published mathematical model of trauma/hemorrhage-induced acute inflammation in mice (25,41,42). In our experimental sepsis model, as in the other experimental models we used, circulating ALR levels increased rapidly but returned to the basal value after peaking at approximately 4 h. We hypothesize that this result might be an effect of resuscitation that followed hemorrhagic shock. Application of our mathematical model of predicted organ damage after experimental trauma/hemorrhage (25,41,42) demonstrated a strong correlation of circulating ALR with the predicted dynamics of shock-induced organ damage, which we have previously suggested correlates with the release of DAMPs as well as acting as a proxy for the health status of the animal (25,41,42). This same mathematical model can account for quite different dynamics of classic inflammatory cytokines such as TNF-α, IL-6 and IL-10 (25,41,42); these cytokines are produced in the same general time frame as DAMPs such as high mobility group protein B1 (HMGB1) in the setting of trauma/hemorrhage (43,44).
These observations suggest that ALR may be a DAMP, a cytokine or a biomarker of inflammatory stress. Cytokines act via specific cytokine receptors that are generally coupled with serine/threonine or tyrosine kinases (45,46), whereas DAMPs often act via receptors that also sense pathogen-derived products (for example, toll-like receptors and nucleotide oligomerization domain [NOD]-like receptors) (19) that are often coupled with the same second-messenger signaling cascades as cytokine receptors. In this regard, a G-protein-coupled receptor for ALR was identified on Kupffer cells, stimulation of which modestly increased expression and release of TNF-α, IL-6 and NO (9). In addition, initial reports suggested that DAMPs were exclusively late mediators in settings such endotoxemia and sepsis (19,47,48), but, as mentioned above, more recent studies in trauma/hemorrhage have shown that DAMPs peak within ~2 h after injury, a time when classic cytokines are secreted as well (43,44). Thus, some ambiguity remains regarding the exact classification of ALR as a cytokine or a DAMP, and future studies may need to focus on the receptor for ALR to resolve this question.
Although ALR is expressed ubiquitously, its expression is much greater in the liver than in other organs except the testes (34,49). This observation suggests that the liver is the primary organ that contributes to the circulating ALR levels. It remains to be determined whether the liver is the only source or other organs contribute to increased serum ALR in the conditions studied herein. However, our in vitro experiments showed that LPS and TNF-α, which do not affect hepato-cyte viability, induced greater release of ALR from hepatocytes than the unstimulated cells. These data suggest that even a minor stress can manifest into increased ALR release, which would indicate increased sensitivity of hepatocytes to injury by other stimuli. In this context, LPS treatment conditions hepatocytes to undergo necrosis when challenged in vivo with otherwise innocuous doses of acetaminophen (32). Indeed, in the present study, we found that actinomycin D-induced hepatocyte injury is strongly augmented in the presence of LPS, which by itself did not affect hepatocyte viability. Because both these agents induced modest release of ALR from cultured hepatocytes, their effect being additive when added together, it can be rationalized that hepatocytes under stress increase their ALR release, which is augmented on injury/damage. This proposal is supported by the observation that the DNA-damaging agent MMS-induced increase in ALR secretion preceded an increase in the LDH release from and loss of viability of cultured hepatocytes. This characteristic of ALR—the release under submaximal stress—is similar to the behavior of other DAMPs such as HMGB1 (19).