Pronucleating proteins in bile – a myth?
Article Outline
In the inexorable pursuit of biological truth, it is inevitable that new techniques be invented to examine old problems. Novel findings are discovered to shape our thoughts yet once again. The advent of gene cloning, mapping, a quantitative approach to genomics, and proteonomics will compel us to reexamine what we now know in epidemiology, biochemistry and pathophysiology anew. Amongst these methodologies is the ability to apply mouse genetics to explore gene function in terms of physiology, or phenotypic features expressed by the gene. Using inbred selected mouse strains possessing gallstone susceptible genes (C57L) and low susceptibility (AKR) strains, van Erpecum et al. [1] investigated that in the process of lithogenesis, whether soluble biliary proteins are involved in cholesterol crystal and stone formation. This impressive piece of work came at the heel of another elegant and instructive review of candidate genes involved in cholesterol gallstone formation [2] by the same group of investigators. They clarified a number of issues in biliary physiology. First, transporters rule, and they dictate what your liver puts into bile [2]. Second, in gallstone formation, gallstone susceptible (Lith) genes correlate with the phenotype (stones) without involving other proteins in a major way. Changes in soluble proteins when measured appeared to be either secondary or irrelevant [1]. Through the power of mice [3], they have refocused the theme of lithogenesis back to biliary lipid secretion and composition.
Hypotheses are important things in the evolution of medicine and science. They help to rationalize all the observations up to a point in time. New observations require a succession of revision of a hypothesis or paradigm shifts. The current hypothesis of pro- and anti-nucleating factors had arisen because of the lack of a good correlation between the calculated cholesterol solubilization capacity and the presence of crystals and stones in gallbladder bile [4]. This propensity was better measured by the time taken for spontaneous cholesterol crystals to precipitate out from the bile sample – ‘nucleation’ or crystal appearance time [5]. Further in vivo and in vitro experiments suggested protein-lipid interaction in bile might be important in inducing or inhibiting crystal formation. A large number of proteins have been implicated to be pro- and anti-nucleating agents [6]. Amongst the putative pronucleators are mucin, con-A binding protein, phospholipase C, anionic fraction peptide (calcium binding protein), α, acid glycoprotein, aminopeptidase N, IgG, IgM, haptoglobin, fibronectin, phospholipase A2, α1-antichymotrypsin [6]. These proteins all have the ability to bind lipid microaggregates in bile. The postulated mechanism is that specific hydrophobic domains bind to cholesterol-enriched microdomains in lipid (cholesterol–phospholipid) vesicles, leading to their aggregation, fusion and eventually phase separation of cholesterol. However, how exactly do these proteins truly effect nucleation (phase transition), precipitation, and organization – growth of crystals into stones, are not well understood. On the other side of the equation, anti-nucleators include various species of apolipoproteins, IgA, and various biliary protein fractions with lectin-binding properties [6]. Most of the data had been generated from in vitro experiments where protein fractions had been added or subtracted, without altering the lipid composition, to observe the effect on cholesterol crystal precipitation. As van Erpecum et al. have ably discussed, most of these studies had the shortcomings of being imperfect in vitro attempts to mimic the in vivo scenario. They often have employed proteins in unphysiological concentration as test compounds. Further, it was difficult or impossible to determine whether the alteration in biliary protein content was a primary event preceding stone formation, or a secondary phenomenon occurring as a result of the presence of stones. van Erpecum et al. found no relevant correlation of protein fractions (IgA, IgM, aminopeptidase-N) with stone formation in comparing the high and low gallstone susceptible (C57L vs. AKR) mice. They found that hydrophobic bile acids (taurodeoxycholic and taurochenodeoxycholic acids) rose, and together with cholesterol supersaturation and possibly a higher mucin concentration, it constituted the primary determinants of cholesterol gallstone formation. The strengths of this work include the use of genetically selected mouse models, the observation of pathophysiology in vivo (gene function), as well as a study on the kinetics of the lithogenic process. It has clearly shown that, apart from some changes in soluble mucin concentration, the lipid composition alone adequately explained gallstone formation without the necessity to invoke any involvement of soluble proteins. Does this mean that we need to here forth abandon the hypothesis of protein-lipid interaction in gallstone disease?
What caution should we exercise in interpreting the results of van Erpecum et al? First, it is always a leap of faith in extrapolating animal experiments to human physiology. This is particularly so with lipid metabolism. Impressive species–species difference exists amongst the rabbit, rat, guinea pig, mouse and human in terms of bile acid and cholesterol metabolism and physiology. Within a species, there are also gender and other intra-species differences. For example, the response to a cholesterol-enriched lithogenic diet is more pronounced in male mice and less so in female mice in both C57L and AKR. Between the two inbred strains, the C57L exhibited a higher bile flow rate with or without a lithogenic diet. Unlike the C57L strain on the lithogenic diet, the AKR strain did not respond with a marked increase in hydrophobic (taurodeoxycholic and taurochenodeoxycholic) bile acid secretion, suggesting further disparities in intestinal transit, bile acid transport or microflora. The differences in biliary soluble protein response to the lithogenic diet between the C57L and AKR mice also cannot be readily explained except by intra-species differences between these two inbred strains other than simply the presence or absence of Lith genes. In addition, for reasons that have been mentioned by the authors, they have measured the soluble fraction of mucin and they did not quantify the gel phase, which is believed to be the more relevant component of mucin involved as a pro-nucleator [7]. Because of the restricted availability of specific antibodies, they have measured also only the immunoglobulins and aminopeptidase-N levels in bile, thus leaving a large number of putative pro- and anti-nucleators unexplored. Therefore, one is left to face the argument that ‘what you cannot see does not mean it is not there’. It would be fascinating to try to mirror the experiment of van Erpecum et al. and examine the propensity for cholesterol precipitation when the only known abnormality is with a biliary soluble protein. One can then probe, without any involvement of lipid transporters or biliary lipid metabolism, if one can alter the susceptibility of cholesterol gallstone formation. No such transgenic, knockout, or inbred selection model is available. The closest model may be the cystic fibrosis knockout (cftr−/−) mouse where airway and gastrointestinal mucin is hyperviscous [8]. There is, as yet, no published work on gallstone formation in the cftr−/− mice.
van Erpecum et al. have taught us valuable lessons in cholesterol gallstone formation. They have emphasized the shortcomings of artificial in vitro systems when studying complex in vivo physiology, and have redirected the focus to cholesterol supersaturation and hydrophobicity of bile salts. We need to follow these leads. At the same time, it is also useful to remind ourselves that gallstones form in the gallbladder and quite apart from the biliary lipid secretion, which is a hepatic effect, there is a gallbladder effect where the mucosal and motor functions also contribute to stone formation. And those protein components – are they a myth which is now dead? Perhaps not yet. The myth – nay, the adventure, continues.
Acknowledgements
Supported by a Merit Review Award from the Department of Veterans Affairs, USA.
References
- . Phenotypic characterization of Lith genes that determine susceptibility to cholesterol cholelithiasis in inbred mice: soluble pronucleating proteins in gallbladder and hepatic biles. J Hepatol. 2001;35:444–451
- . Chromosomal organization of candidate genes involved in cholesterol gallstone formation: a murine gallstone map. Gastroenterology. 2001;120:221–238
- . A miracle enough: the power of mice. Nat Med. 1995;1:215–220
- . Cholesterol solubility in bile. Evidence that supersaturated bile is frequent in healthy man. J Clin Invest. 1973;52:1467–1479
- . Nucleation time: a key factor in the pathogenesis of cholesterol gallstone disease. Gastroenterology. 1979;77:611–617
- . Gallstone formation: local factors. Gastroenterol Clin North Am. 1999;28:99–115
- . Bovine gallbladder mucin accelerates cholesterol monohydrate crystal growth in model bile. Gastroenterology. 1993;104:1515–1523
- . An animal model for cystic fibrosis made by gene targeting. Science. 1992;257:1083–1088
PII: S0168-8278(01)00209-4
© 2001 European Association for the Study of the Liver. Published by Elsevier Inc. All rights reserved.
