All About Multiple Sclerosis

More MS news articles for May 2003

A New Approach to Autoimmune Diseases

Instead of studying disorders in isolation, some investigators are seeking common denominators among them

May 5, 2003
By Eugene Russo
The Scientist
Volume 17, Issue 9, 30

Nearly one hundred years ago, two German scientists introduced the concept of autoimmune disease;1 it didn't catch on. Indeed, even the investigators themselves were skeptical. To most immunologists, the notion that people might develop immune reactions to their own bodies seemed counterintuitive, even preposterous. "People were very, very reluctant to accept the idea," says Noel Rose, director of the Johns Hopkins University Center for Autoimmune Disease Research and a long-time expert in the field.

Then, in the 1950s, two experimental models validated the autoimmune disease concept: one for thyroiditis, spearheaded by Rose, and another for lupus. In recent years, researchers have applied these models to a variety of diseases and conditions; in fact, it might have become a bit too popular. "As often happens in science, the pendulum swung the other way," says Rose. "People began to think of every disease in the book as autoimmune." Autism, Alzheimer disease, even eating disorders are suggested candidates. Significant immune components could be present, but whether such diseases are caused by an attack on one's cells and tissues remains unclear. To date, more than 80 clinically distinct autoimmune diseases have been identified, though there is no completely foolproof way to define what is autoimmune and what is not.

The detective story gets murkier. Even commonly accepted autoimmune diseases may consist of subclasses of tens or hundreds of disorders that have their own specific genetic and environmental causes. Elucidating etiologies leading to better treatments likely will require the more precise categorization of this set of very complex diseases.

But to discover different subclasses of autoimmune disease, researchers first need to find what, if any, genetic and environmental factors these disorders share. The search for common denominators, now underway, represents a major shift in research strategy.

"The people who studied autoimmune disease generally studied their own disease and really had blinders on," says Kevin Becker, who runs the DNA array unit at the National Institute on Aging. "If it is true that there are these common genetic elements, then the people who study type-1 diabetes should talk to the people who study multiple sclerosis, who should talk to the people who study thyroiditis, and they should work in a cooperative way more than they do."


The increased incidence of many autoimmune diseases appears to be on the rise, especially in industrialized nations. These disorders affect an estimated 5% to 8% of the US population, or 14 to 22 million people. Much improved recognition and diagnosis for some illnesses, along with a lack of epidemiological studies for others, however, makes the numbers hard to interpret. For example, data indicate increased incidence of type 1 diabetes and multiple sclerosis in the US.2

In January, the National Institutes of Health announced new priorities for more accurately determining incidence, prevalence, and severity of autoimmune diseases, for identifying the interaction among genetic and environmental factors, and for developing new clinical centers that encourage public-private partnerships. The plans include, for example, continued funding of immunology centers at institutions, and also the Immune Tolerance Network, a collaborative effort by more than 70 experts worldwide aimed at developing therapies to enhance disease tolerance. The recommendations also include continued work to uncover the commonalties among different autoimmune diseases.

The first major effort in this regard came in 1998.2 A meta-analysis conducted by Becker and Jeffrey Trent, now president and scientific director of the Translational Genomics Research Institute, compared the genetic linkage results from 23 published autoimmune or immune-mediated disease genome-wide scans. Diseases examined included multiple sclerosis, Crohn disease, familial psoriasis, asthma, type-I diabetes, murine lupus, and rat inflammatory arthritis. They found 18 overlapping clusters of loci, suggesting that these autoimmune diseases could be governed by a set of common susceptibility genes.

The results are merely suggestive, however. The overlapping regions are large, as in millions of base pairs. "All you know is that there are some genes within this large region that seem to be involved," says Peter Gregersen, director of the Center for Genomics and Human Genetics at the North Shore Long Island Jewish Research Institute. The presumption is that, for example, the gene families of juvenile diabetes and rheumatoid arthritis share regions because they share the same gene or genes in that region. But that presumption may be incorrect because the identified regions are so large. Thus far, confirming genetic linkages in autoimmune diseases has proven extremely difficult. The lone exception: the linking of Crohn disease to a region on chromosome 16 identified as nod2.4

Nevertheless, the Becker group's findings have piqued interest among researchers. "The challenge now is to take these 10 or 20 regions that keep coming up again and again and get down to what genes are there," says Gregersen.

The Multiple Autoimmune Disease Genetics Consortium (MADGC), funded by the NIH and led by Gregersen, will attempt to identify genes that several autoimmune diseases have in common. MADGC is collecting data on families whose members have several different autoimmune disease phenotypes. To qualify, a family must have at least two of the eight autoimmune diseases chosen for the study, namely rheumatoid arthritis, juvenile rheumatoid arthritis, systemic lupus, multiple sclerosis, autoimmune thyroid disease (also called Graves or Hashimoto disease), type I diabetes, psoriasis, and inflammatory bowel disease (Crohn disease or ulcerative colitis). The family data, including cellular components, blood serum components, DNA samples, and anonymous clinical data, then will be released to interested researchers so they can, according to Gregersen, "examine these families for their favorite candidate or their favorite hypothesis about what may explain a familial aggregation." MADGC, formed in 1999, recently released data on their first 65 families comprising 300-400 individuals.

The ultimate goal is to collect 400 families. The biggest challenge has been confirming the data before releasing it to the scientific community; diagnoses are rarely obvious or clear-cut. "Hopefully we're getting enough diversity in these families that they'll tell us something when we study them, but not so much diversity that every family's unique and we have no statistical power to say anything," says Gregersen. "That's the tradeoff."

The short-term goal is identifying these subclasses based on genetic and environmental factors. Frederick Miller, chief of the Environmental Autoimmunity Group at the National Institute of Environmental Health Sciences, calls this motivating rationale the "elemental disorder hypothesis." Essentially, reads the hypothesis, diseases labeled as lupus or diabetes actually could represent dozens of diagnosable elemental disorders with unique clinical manifestations. The unique pathogenesis of each disorder comes about after the necessary and sufficient genetic and environmental risk factors have interacted. Miller likes to use a movie theater analogy: Genetic and environmental elements common to several autoimmune diseases allow entry into a "theater" of autoimmunity, and then other genetic risk factors and different environmental exposures "usher" them into different subcategories or different "seats" in that theater.


Miller's group recently started recruiting for a large-scale pedigree study; the intended focus is to isolate both genetic and environmental risk factors. They plan to enroll 400 pairs of twins and siblings, the latter having one sibling who is within four years of developing one of a group of autoimmune diseases--namely rheumatoid arthritis/juvenile rheumatoid arthritis, systemic lupus erythematosus, systemic sclerosis or myositis. Parents and other family members also will be enrolled for genetic analysis. By carefully following subjects for five years and taking detailed histories through mailings and office visits, researchers hope to get some idea of pertinent environmental factors, and which ones appear to be most relevant to which genetic profiles. Factors likely to be scrutinized include infectious and noninfectious agents, drugs, vaccines, food, dietary supplements, organic solvents, ultraviolet light, stress and stressful life events, and occupational exposures (see UV Radiation, Autoimmunity, and Questions Galore) to substances like silica.

Ideally, Miller's group will break down these autoimmune disorders into unique subsets with nearly identical genetic and environmental risk factors, so that highly specific treatments and/or preventative measures can be developed--it's a move toward the much-anticipated era of individualized medicine. Current, more generalized treatments have mixed success; according to Gregersen, they typically work for about a third of patients, evoke partial responses for another third, and do absolutely nothing for the remainder. "It may be paradoxical that the environmental components will be easier to find once we understand the genetics," says Gregersen. "You won't have to study tens and tens of thousands of people to see an association [with an environmental factor] if you know what subgroup you should be looking at for that association."

"We're finding out that more and more diseases have immune processes involved," says Miller. "Whether they're autoimmune or not, at least there are immune-mediated processes that are important in the development of the pathology."

With numerous diseases nominated as autoimmune candidates-from autism to obesity to schizophrenia to Alzheimer's to obsessive compulsive disorders-weeding out the all-but-proven from the potentials or partials is a formidable task. Noel Rose, director of the Johns Hopkins University Center for Autoimmune Disease Research, has helped establish oft-cited criteria for classifying evidence of autoimmune disease-that is, disorders in which autoimmunity plays a causal role rather than occurring after disease onset.

Rose calls the first type of evidence "direct." By definition, it must have the ability to transfer the disease from someone who has it, to someone who doesn't. Researchers know of only six to 10 diseases where this proves true. Most are cases in which mothers transfer autoantibodies to their children; one such disorder is transient myasthenia gravis, a disease of the peripheral nervous system. A few additional diseases can be reproduced by transferring human serum to another animal, as in the case of pemphigus vulgaris, a rare, autoimmune skin disease.

A second autoimmune disease category can be reproduced only in animal models-that is, by "indirect evidence." Researchers might, for example, replicate the human disorder in animals by immunizing them with culprit antigens. Examples include several murine lupus models. These are useful but limiting, because it's often unclear how well they imitate the human disease.

Altering an animal's immune system to spontaneously reproduce disease constitutes a third means of creating autoimmune disease; it's become a popular method. Researchers often change the animal's cytokine profile. For example, murine knockouts for the cytokine IL-10 sometimes develop an inflammatory bowel disease that looks very much like Crohn in humans. 

The final disease category comprises those with circumstantial evidence that support a causal role for autoimmunity. There may be autoantibody production, but no evidence that antibodies cause the disease; such is the case with autism. Says Rose: "Just finding autoantibodies might be the opening clue, but it doesn't really establish that the disease is caused by autoimmunity."

Eugene Russo is a freelance writer in Takoma Park, Md.

Multiple Autoimmune Disease Genetics Consortium

Autoimmune Diseases Research Plan


1. J. Donath, K. Landsteiner, "Uber paroxysmale haemoglobinurie," Munchen Medicine Wochenschr, 51:1590-3, 1904.

2. D.L. Jacobson et al., "Epidemiology and estimated population burden of selected autoimmune diseases in the United States," Clin Immunol Immunopathol, 84:223-43, 1997.

3. K.G. Becker et al., "Clustering of non-major histocompatibility complex susceptibility candidate loci in human autoimmune diseases," Proc Natl Acad Sci, 95:9979-84, 1998.

4. M. Anderson, "Crohn's: An autoimmune or bacterial-related disease?" The Scientist, 15[16]:22, Aug. 20, 2001.

©2003, The Scientist Inc.