The Genetic I.D. of Lupus

By Jason Socrates Bardi

"Know Thyself"

—A saying often attributed to Socrates and inscribed at the entrance of the ancient temple of Apollo at Delphi

The patient is ten times more likely to be female, often a young adult in her 20s. She has a slightly higher chance of being of African, Native American, or Asian descent and often no one in her immediate family has had a history of medical problems. She herself may have enjoyed perfect health until, suddenly, there was any one of a thousand things wrong with her. Perhaps she was hospitalized with seizures. Perhaps she has painful swelling in the small joints of her hands. Perhaps she has red rashes on her hands and face, or anemia...

If her doctor knows enough to refer her to a Rheumatologist who defines her illness as lupus, she is one of the lucky ones. Lupus is often misdiagnosed.

Even the name "lupus" is a mischaracterization of sorts. The word lupus, which means "wolf" in Latin, was first used in the Middle Ages to describe a chronic rash on the skin. The name may have been chosen because the rash on the skin resembled the effects of a bite from one of these wild animals. Or, some believe the name arises from the fact that the rash was common about the cheeks, giving lupus victims a werewolf-like appearance.

Whichever the case, the disease is not the bite of a Canis lupus, but the bite of a person's own immune system.

Lupus is a chronic, inflammatory autoimmune disease caused by multiple genetic, environmental and other factors, most of which are unknown. It is a complicated disease that may appear a thousand different ways in a thousand different people. The Lupus Foundation of America estimates that approximately 1,400,000 Americans have a form of lupus, a disease that ranges widely case-by-case, has a long list of symptoms, and affects a wide variety of tissues—especially the skin, joints, blood, and kidneys.

Lupus occurs when a person's own B cells produce antibodies that are directed against "self" tissue. These antibodies—secreted proteins also called "immunoglobins" that help the body clear infections—normally target foreign pathogens in the bloodstream or those displayed on infected cells. But in lupus, the antibodies target the body's own molecules instead.

For instance, many people who have lupus produce an antibody that targets red blood cells, which are a vital oxygen-transporting component of blood. The antibodies coat the red blood cells, and these are then taken up and destroyed by macrophages. This can lead to a deficit of red blood cells and anemia.

Two researchers at The Scripps Research Institute (TSRI) are investigating the causes of lupus and looking for possible targets for intervention. Immunology Professor Argyrios Theofilopoulos has studied lupus for over 25 years—since soon after he came to TSRI as a postdoctoral fellow in the laboratory of Frank Dixon, then the director of the institute, in 1972. Now Theofilopoulos works closely with his colleague, Associate Professor Dwight Kono, who arrived at TSRI in 1989.

Treatment Largely Unchanged Since the 50s

The victims of lupus have yet to reap the benefits of the genomic revolution. There are tests that will determine if you have the disease, but there are no genetic screens to tell you if you have an increased susceptibility. In fact, the genes that contribute to lupus are not even known.

"It's clear genes play a big part and it's complex," says Kono. "We don't know what they are yet. We're still mapping the genome to find out where they are located."

Not all genes contribute to the disease equally. Many act in concert with one another and it is the sum effect that causes the disease. "In these diseases you have not only multiple genes but environmental and other stochastic influences," says Theofilopoulos. "Knowing one of them may not be sufficient to change the disease process. Yet some genes may have a greater effect than others, and if we can identify them, we may be able to intervene."

Instead of viewing diseases like lupus as one disease, one might be able to identify a sub-group of patients with a particular genetic defect that contributes to their particular form of the disease.

"Then we would be much better off designing treatments," says Theofilopoulos. Such an accomplishment would be a major advance, he adds, because no new treatments for lupus have been found since the 1950s, and these treatments have a high incidence of side effects.

The hope is that instead of giving non-specific drugs as is the case now, one could design specific drugs to treat a specific form of lupus. Additionally, if scientists know the genes that lead to the formation of the disease, they may be able to predict who is susceptible.

Lupus was rigorously described and defined as a medical condition in the early 1800s, but it was not until the latter half of that century that real progress was made in defining the full clinical spectrum of this disease. The first positive step was when doctors recognized that the disease could be systemic and could cause damage to the kidneys and other internal organs distinct from and sometimes in the absence of its defining rashes.

More importantly, at the turn of the 20th century, doctors discovered the effectiveness of anti-inflammatory and anti-malaria drugs as a treatment for lupus. But the real breakthrough came decades later, in the early 1950s, when doctors began using corticosteroids to treat lupus with even greater success.

Although a few other treatments have been developed in the last 50 years, corticosteroids, anti-inflammatories, and anti-malarials are still the main therapies for the disease today.

Models of the Disease

One of the key tools that has allowed so much research to be carried out at TSRI and elsewhere has been the identification and availability of murine models for lupus.

Dixon and Theofilopoulos characterized several murine models in the 1970s that developed symptoms resembling lupus. They spent the next several years analyzing these models, trying to understand the basic pathogenesis of lupus and writing some of the first detailed descriptions of the serological, cellular, and histological characteristics of the disease. Theofilopoulos was interested both in the basic immunological assessment of the disease—its relationship to T cells, B cells, organs like the thymus, antibodies—and in finding specific molecules important for its pathogenesis.

"Then," says Theofilopoulos, "when molecular cloning began, we started defining the structural characteristics of the autoantibodies implicated in lupus."

He discovered that the genes that encode these pathogenic antibodies were not very different from those encoding regular immunoglobins against foreign antigens. His goal became identifying all the other genetic components of the disease, something Theofilopoulos and Kono have been working on for the last several years.

Specifically, they became interested in identifying the "effector" genes that predispose a person to getting lupus or that lead normal genes to be involved in the disease process. To be important for therapy, these genes could be a root cause of lupus or simply a contributing factor. Some genes are not necessarily malfunctioning in lupus patients, but may nevertheless contribute to the pathogenesis of the disease.

"If you block these effector genes, you will still have a good therapeutic effect," says Kono.

The Contemplative Process of Identifying Predisposing Loci and Genes

In order to identify the genetic components of any disease when there are no obvious candidates, one must identify the chromosomal regions containing the genes and then screen virtually all the genes in the region for their involvement. This long and complicated process is known as positional cloning.

The process starts with one of the murine lupus models that Theofilopoulos has characterized. These mice develop lupus-like symptoms spontaneously, in much the same way as the disease is manifested in afflicted people. One of these lupus strains is crossed with a nonautoimmune strain and then those offspring are interbred to generate "F2" offspring that contain diverse random combinations of genetic material from the original parents. The general chromosomal locations of predisposing genes can be then deduced by identifying, in a large number of afflicted F2 models, what chromosomal regions are inherited from the lupus-prone strain with a greater frequency than chance alone. These intervals or loci are generally resolved to around 20 to 40 cMorgans or about 40 to 80 million base pairs, and can contain hundreds of genes, a size generally too large to screen for specific genetic alterations.

To narrow the interval, a new model, called a congenic, is developed that consists of one background strain containing only the locus region from the other strain. This permits direct testing of the effect of a single locus on the development of lupus, either through the replacement of the predisposing locus from the lupus-prone strain with the normal counterpart, or by the addition of the predisposing interval to the normal strain. Breeding of this initial congenic is a time-consuming process requiring backcrossing of the interval for several generations, on average requiring two years or more. Further narrowing of the region down to about a million base pairs can then be accomplished by generating additional congenics containing smaller and smaller intervals.

"Once you accomplish that, then you can mine the data from the genome projects and start to make intelligent conclusions about genes that are known," says Theofilopoulos. A significant effort, however, is still needed to screen candidate genes, a process that involves extensive cloning, sequencing and analysis of gene expression.

A general problem, however, with positional cloning of genes involved in lupus is that the disease is made up of many phenotypes—in fact, lupus is diagnosed clinically when a patient has four or more of 11 possible defining conditions. Mathematically, that gives several thousand possible combinations of symptoms that can be defined as lupus, although, practically, the number of common clinical scenarios are much fewer.

However, there are only three model strains that develop three specific types of lupus, which is more like having three individuals with the disease. The three strains are not a perfect representation of all the different manifestations of the disease in the 1.4 million people believed to be afflicted with lupus in the United States.

"Because of the diversity of the disease in the human population," says Theofilopoulos, "it is appropriate to focus our initial studies on these models."

Another complication is the fact that lupus involves an array of interacting genes. So identifying one gene may give you some of the answers, but certainly not all.

"You want to get at the root cause of the disease," says Kono. "But there's no particular factor that's so overwhelmingly contributory that you can eliminate all the others."

The two remain positive that, through this approach and concurrent studies in humans, at least some of the predisposing genes will be identified.

Another approach they are taking to find therapeutic targets is to identify genes that may be normal and unmutated, but which encode products that may be involved in the disease pathogenesis. These genes can be identified by deleting them in the lupus models.

"We have already selected genes that we believe will have an important effect on the disease process," says Theofilopoulos.

For instance, they have identified one possible target for therapy, an inhibitor of cyclin-dependent kinases that is overexpressed in T cells during lupus that may be responsible for one symptom of advanced lupus—a flood of helper T cells that are resistant to proliferation and apoptosis. Targeting this inhibitor and promoting cell proliferation might make severe lupus mild.

By deleting the genes encoding this inhibitor, Theofilopoulos and Kono have prevented the helper T cells from accumulating and reduced all the serologic and histologic characteristics of the disease.

Another class of genes that have been found to contribute to the disease are the Type 1 and Type 2 interferons, pro-inflammatory molecules referred to as IFN-a and IFN-g respectively. Based on these findings, Theofilopoulos and Kono have reported that using naked cDNA encoding the receptor for IFN-g could block the activity of the IFN-g and cure the lupus in these models.

Developing recombinant receptors for IFN-a and IFN-g should provide better means to reduce the severity of the disease.


Go back to News & Views Index

"We have selected genes that we believe will have an important effect on the disease process," says Argyrios Theofilopoulos (right), whose work with Dwight Kono seeks to address the problem of finding new treatments for lupus. Photo by Jason Bardi.