My association with Tony Hugli, long-term editor of Immunopharmacology and International Immunopharmacology, came about by a specific and long-standing problem in inflammation research. What is the trigger mechanism of inflammation in physiological shock? This is an important clinical problem due to the high mortality associated with physiological shock. We joined forces in the search of the answer to this question for more than a decade. Our journey eventually led to development of the hypothesis that shock may be associated with pancreatic enzymes, a set of powerful digestive enzymes that are an integral part of human digestion. The digestive enzymes need to be compartmentalized in the lumen of the intestine where they break down a broad spectrum of biological molecules into their building blocks, suitable for molecular transport across the mucosal epithelium into the circulation. The mucosal epithelial barrier is the key element for compartmentalization of the digestive enzymes. But under conditions when the mucosal barrier is compromised, the fully activated digestive enzymes in the lumen of the intestine are transported into the wall of the intestine, starting an auto-digestion process. In the process several classes of mediators are generated that by themselves have inflammatory activity and upon entry into the central circulation generate the hallmarks of inflammation and eventually cause multi-organ failure. Thus, our journey led to a new hypothesis, which is potentially of fundamental importance for death by multi-organ failure. The auto-digestion hypothesis is in line with the century old observation that the intestine plays a special role on shock - indeed it is the organ for digestion. Auto-digestion may be the prize to pay for life-long nutrition.

My encounter with Tony Hugli was deliberate and planned. Having recognized from the short distance between the University of California San Diego and The Scripps Research Institute his pioneering work on complement, the isolation and identification of biological and pathophysiological properties of complement fragments [1], our collaboration started in the early 1990^th at a luncheon. Tony's interest and generosity to consult was a natural growing ground to focus on an important medical problem and generate new ideas. His wit, quick reasoning and all encompassing knowledge of protein chemistry were a refreshing inspiration and source of ideas on the journey we went on to embark. We have reached a new vista on one of the most pressing problems in medicine. It has been a roller coaster with few high points and plenty of disappointments. Bioengineering, my field, and biochemistry, Tony's field, played a complimentary role; either discipline by itself would have been inadequate to take us to the level of understanding we have reached today.

Nowhere is the lack of firm knowledge about the trigger mechanisms more visible than in the severe forms of inflammation encountered in physiological shock - a condition with extraordinary high mortality. Shock is accompanied by high levels of inflammatory mediators in plasma and in lymph fluid. In experimental forms of hemorrhagic shock we detect significantly elevated levels of inflammatory markers already within one hour after central blood pressure reduction [19,20]. The markers can be detected by exposure of plasma to naïve leukocytes from a donor animal. These inflammatory mediators have been reported repeatedly in the past and have received various designations, e.g. leukocyte activating factor, clastogenic factor, myocardial depressing factor, T-cell proliferation depressing factor, and others [21]. None of these designations fully embrace the spectrum of activity that is associated with plasma from individuals with physiological forms of shock. In general, shock plasma depresses cell functions irrespective of the particular cell type under investigation. In-vivo the appearance of inflammatory mediators in plasma is accompanied by multi-organ failure often in relatively rapid succession following the initial insult that precipitates the shock.

This opportunity arose when Professor Alan Lefer of Thomas Jefferson University spend a sabbatical in my laboratory in 1995. He shared his decade long experience in attempting to isolate an inflammatory mediator encountered in shock which he had designated as the “Myocardial Depressing Factor” [30,31]. We determined that the same plasma samples generate the characteristic depression of cardiac muscle contraction as well as the typical activation in circulating leukocytes. Professor Lefer and Dr. Erik Kistler suggested to look for a possible source of inflammatory mediators by use of selected organ homogenates. While this approach does not directly identify the actual inflammatory mediator in plasma, it may still be helpful for this problem since it may suggest which organ or cells may be a source of mediators under conditions of shock.

Pancreatic enzymes are released into the lumen of the intestine as part of normal digestion. An ischemic intestine is associated with elevated permeability [37,38,39,40,41] and therefore the usual compartmentalization of digestive enzymes in the lumen of the intestine may no longer be preserved. The question arises whether products generated during inadvertent entry of digestive enzymes into the wall of the intestine may cause an inflammatory reaction. While normal intestinal tissue by itself generates minimal levels of inflammatory mediators, intestine after ischemia generates significant inflammatory mediators [42]. Highly cytotoxic inflammatory mediators may be generated in the intestinal wall by the mixture of digestive enzymes from the lumen of the intestine or even by individual proteases (e.g. trypsin, chymotrypsin, elastase) [43]. In fact, in the rat, regular commercial chow will generate a powerful cytotoxic mediator if incubated with the mixture of pancreatic digestive enzymes [43]. This is a surprising result. Food may contain cytotoxic components once it enters into the intestine and is being digested by pancreatic enzymes? This is an important issue that requires a great deal more investigations.

In the meantime, the question arises whether blockade of the pancreatic digestive enzymes in the lumen of the intestine may serve to reduce the production of inflammatory mediators in shock. Indeed, blockade of the pancreatic digestive enzymes in the lumen of the ileum serves to reduce the level of inflammatory mediators in shock produced by occlusion of the superior mesentery artery [44], irrespective of the particular choice of the inhibitors [45], in hemorrhagic shock [46], even after some delay in the enzyme blockade [47], or in endotoxic shock [48]. The protection provided by the protease inhibition is not enhanced by supplementation with oxygen free radical blockers [49]. Blockade of digestive enzymes in the lumen of the intestine during ischemia serves to prevent formation of inflammation in remote organs [50]. Blockade of digestive enzymes in the lumen of the intestine also reduces the need for fluid required to maintain blood pressure after a severe form of ischemia and eventually also reduces the level of inflammatory mediators in the plasma [51].

For most living creatures, shock and multiorgan failure is a problem second-to-none. Unless death is caused by an abrupt event (e.g. trauma, fibrillation), multi-organ failure is likely to play a central role when the circulation come to a stand-still. Understanding the origin of the inflammatory cascade that leads to the progressive failure of vital organs is the basis for rational and optimal treatment. It has become evident that there are several classes of inflammatory mediators which in the simplest of all analysis need to be divided into two classes: (a) the group of inflammatory mediators that are actually involved in breakdown of the tissue after an insult (e.g. protein and lipid fragments generated by pancreatic digestive enzymes, complements), and (b) the inflammatory mediators — perhaps better designated as signaling or repair molecules — that are part of the resolution associated with inflammation (e.g. cytokines). The evidence for the first group has come from studies of organ homogenates generated in the presence of digestive enzymes and from studies with blockade of digestive enzymes as an in-vivo intervention. While the evidence for the second group is already well documented, the evidence for the first group is new and less well understood. Clinical interventions need to be directed at the first group of mediators. But for now a new hypothesis for the origin of inflammation in shock has been born: the Auto-digestion hypothesis (Figure 1) [52], i.e. a breakdown of a variety of autologous molecules by one's own digestive enzymes. So far, the evidence supporting this hypothesis is derived from preclinical studies [53]. The evidence in man is still missing but the idea will spark clinical studies. We celebrate Tony Hugli and hope he will continue to be an integral part of the effort.