Book Review

Experimental Models of Diabetes

N Engl J Med 2000; 342:63-64January 6, 2000

Article

Experimental Models of Diabetes
Edited by John H. McNeill. 418 pp., illustrated. Boca Raton, Fla., CRC Press, 1999. $149.95. ISBN: 0-8493-1667-7

Experimental models are conventions; the community of scientists designates, often informally, its subjects of study. Today, scientists not only breed their experimental animals; they also exercise the power to add to, delete, exchange, or mutate the animals' genes. Thus, experimental models are inventions as well as conventions. Like all inventions, they may be judged by the way they serve our needs. Books about experimental models, too, may be judged by the advantage gained by reading them.

We may divide experimental models into two sorts: analogue models and intrinsic models. Analogue models are useful as substitutes for some reality otherwise inaccessible to experimentation — a human disease, for example. Intrinsic models, unlike analogue models, do not have to mimic reality; they fascinate on their own.

Experimental Models of Diabetes tells us about diabetes in laboratory animals. Do these models serve as analogues of human diabetes, or are they models intrinsic to themselves? The book contains 14 chapters, 10 of which relate to diabetes induced experimentally by administration of a toxin called streptozotocin. Streptozotocin is a Trojan horse–like molecule. In essence, it is glucose linked to a reactive nitrosourea moiety, and as such it is internalized through the cell's glucose transporters. There is no gift here, however. Once the molecule is inside, the nitrosourea moiety is released and actively poisons the cell by cross-linking vital structures. Because the beta cells of the pancreas are more active than other cells in taking up glucose (they continuously sample the blood glucose concentration), they are also more sensitive than other cells to streptozotocin poisoning. That's why streptozotocin, at a given dose, preferentially kills beta cells and causes a type of diabetes.

But streptozotocin rarely is the cause of diabetes in humans, and so diabetes produced by this toxin is not an analogue of the pathogenesis of human diabetes, either type 1 or type 2. In Experimental Models of Diabetes, the 10 chapters on streptozotocin describe the complications of insulin insufficiency and glucose intolerance that arise in rodents after streptozotocin poisoning. An additional chapter describes the diabetes produced (usually in rabbits) by a different toxin, alloxan. The streptozotocin and alloxan models are intrinsic, but they may serve as intriguing analogues of the complications of uncontrolled diabetes.

One of the remaining three chapters describes the spontaneous autoimmune diabetes that develops in mice of the nonobese diabetic (NOD) strain, another tells us about the diabetes that affects the Bio Breeding (BB) rat, and the final chapter recounts various genetic models of non-insulin-dependent (type 2) diabetes in rodents. The BB rat is characterized by a severe form of immune deficiency, so it has little in common with most humans who have diabetes. The NOD-mouse model seems to mimic much more faithfully the autoimmunity associated with type 1 diabetes in many humans. Indeed, recent clinical trials of specific immune modulation of the diabetogenic process are based on studies of the NOD-mouse model. Unfortunately, the chapter on the NOD-mouse model does not discuss these trials, and it does not critically examine the analogy of the NOD mouse to the human with disease.

The final chapter of the book describes a series of single mutations associated with the appearance of obesity and non-insulin-dependent diabetes in mice. In few, if any, humans does obesity or diabetes arise because of a single mutation; these mice therefore exemplify intrinsic, not analogue, models of diabetes. Nevertheless, studies of mice with the ob and db mutations have uncovered a key molecule involved in the physiology of appetite and energy regulation, called leptin. An initially intrinsic model of diabetes, like the db mouse, has thus become an analogue model of appetite dysregulation. Intrinsic and analogue models can shift roles, once we learn the reality behind the appearances.

Readers who are in need of an encyclopedic collection of information about streptozotocin models of diabetes will be well served by this book. Today, however, the most current and comprehensive information can be accessed not in print but on the Internet and in its resources. To reach safe port in the deep sea of electronic information, however, we need new kinds of charts. Books might help us navigate by clarifying the meaning of the information we have, or that we lack.

Irun Cohen, M.D.
Weizmann Institute of Science, Rehovot 76100, Israel