“For me a picture must be an amiable thing, joyous and pretty—yes, pretty! There are enough troublesome things in life,” the French impressionist Pierre-Auguste Renoir wrote, barely hinting at his own misfortune. Renoir, whose luminous, rosy-cheeked nudes are considered among the most amiable images in the history of art, struggled with one of the most debilitating and disfiguring of diseases, rheumatoid arthritis. The disorder, which struck when he was 50, would slowly erode the cartilage and bone of his hands, shoulder, and feet and lay waste to other parts of his body. According to his grandson, Renoir would wake in pain and call for his paintbrush to be affixed to his gauze-wrapped, balled-up hands.
The great master had only his art—and the occasional visit to a spa—to see him through. A century later, there are various treatments for rheumatoid arthritis and yet they do not always alleviate patients’ suffering. Most are designed to quell the immune system’s attack on the tissue lining the joints, a defining event in the disease. Yet this attack is but one chapter in a complex, often variable, and still mysterious disease. “Rheumatoid arthritis is not a single disease, it’s a spectrum, a syndrome. So if you take a single treatment, only about half of patients respond,” said David Lee, HMS assistant professor of medicine at Brigham and Women’s Hospital.
In a surprising new twist, it now appears that, once attacked, the joint lining, or synovium, may ramp up inflammation in the joint as well as bring about erosion of the cartilage. Lee, working with Michael Brenner, the Theodore Bevier Bayles professor of medicine at BWH, and colleagues, has identified a protein that serves as a molecular ringleader for the synovium’s destructive activities. The findings, published in the Feb. 16 Science, could open up a new avenue for treating the joint-wasting disease.
The story of this discovery goes back several years when Lee, Brenner, and colleagues identified a protein in the synovium, a delicate but dense wisp of cells, that makes the cells of the tissue stick together and form a compact, tightly organized structure. But their experiments were carried out in vitro. Lee set out to explore the role of the protein, cadherin-11, in living animals, comparing cadherin-11–knockout mice with wild types. In contrast to the wild-type synovium, with its neat rows of cells, that of the knockouts appeared sparse and desultory, confirming cadherin’s role in shaping the synovial tissue.
Lee and colleagues tried inducing rheumatoid arthritis, using autoantibodies in the cadherin-11–deprived mice. The animals barely showed signs of disease. Not only did they suffer less cartilage loss, levels of inflammation were about half that found in wild-type animals who had been given the same autoantibodies.
In another experiment reported in the current paper, Lee and colleagues tried blocking cadherin-11 in the wild types before administering the autoantibodies. The mice resisted disease, much like the knockouts.
“This is a protein expressed on tissue, not on the inflammatory cells, and yet we see a decline in inflammation when it is not present,” said Brenner. “We interpret this as a role for the synovium in inflammation. It’s not just a target of inflammation, it’s a participant in generating the inflammation.”
The findings in mice hold obvious implications for treating humans; Brenner and Lee have patented and recently licensed the discovery. What especially excites the researchers is the possibility that the principle underlying their approach—that tissues, once attacked by the immune system, may actually stoke inflammation—could have broader applicability.
“This is a concept we’ve shown in this disease model, but it probably applies to inflammatory-related tissue injury that occurs all over the body, in all kinds of diseases,” Brenner said. The concept will likely lead to a revision of the traditional story of how rheumatoid arthritis develops.
The disease is thought to begin when the immune system attacks the synovium, though it is still not clear what causes this attack. Once inflamed, the veil-like synovium balloons to many times its original size, forming a thick, shroudlike structure, or pannus, that begins to creep along the cartilage-covered bone, chewing it up (see figure). At this point, the disease enters its destructive, bone-wasting phase. Five years ago, researchers discovered that this phase is carried out, at least in part, by osteoclasts, cells normally involved in remodeling bone. The finding set off a search for ways to control these bone-eating cells.
“We knew a lot about the inflammatory component,” Brenner said. “We knew that the osteoclast is being targeted to prevent bone damage.” It was the middle part of the story and its main character, the synovium, that intrigued Brenner the most. Curiously, little was known about the synovial structure. Though delicate in appearance, the membranous tissue is designed to pack a punch. “The synovium is thin and lacy, but in addition to fibroblasts, it is loaded with mast cells, macrophages, and other cells that can respond very quickly to pathogenic infection. And they do. Infected joints blow up in minutes, they just explode,” said Lee.
All membranes contain cell adhesion molecules such as integrins, selectins, and cadherins. Brenner was especially interested in the cadherins, which bind to one another instead of to different kinds of proteins. He set out to find which of the cadherins was working in the synovium and came up with cadherin-11. In vitro work showed that the protein was responsible for getting the cells of the synovium, or synoviocytes, to stick to one another and to line up in neat rows.
“Then David studied the knockout mouse and asked, ‘Does cadherin-11 direct development of the synovium in real life?’” said Brenner. Their current paper shows that it does. But the study goes further. The researchers found that just as cadherin-11 helps to organize the normal synovium, it plays a critical role in the formation of the pannus, not just its bloating but its subsequent crawl along the bones. In the knockouts, the pannus, rather than being a dense shroud, is a filmier, skimpier structure.
When cultured, the cells of the pannus exhibited much lower levels of migration than those of wild-type mice, a telling finding. In fact, the observation could help explain why cadherin-11 knockouts appear to resist cartilage damage. Lee and Brenner believe that cadherin-11, in addition to allowing synoviocytes to bind to one another, directs the cells to migrate across and eventually invade and destroy the cartilage.
“What we show here is that the synovium, when it attaches to and erodes cartilage, seems to be the dominant mode of cartilage damage,” Brenner said. “The synovium appears incapable of turning into this cartilage-destroying pannus without cadherin-11.”
It is this double-identity—instigator of both inflammation and cartilage damage—that makes cadherin-11 such an appealing drug target, either alone or, more likely, in combination with existing therapies.
“People have tried to combine several of the anti-inflammatory drugs, but the problem is, you get infection,” said Brenner. “The idea here might be you take an anti-inflammatory and you get a partial effect from it, but it’s not completely stopping disease. You could combine that treatment with a treatment that has a different toxicity profile that is targeting the synoviocytes. It might be an appealing combination, and the treatments also might be synergistic.”
The new study could raise the profile of the cadherins, which have been, according to Brenner, “low-liers” in the medical world. “Can we take what we have learned about cadherin-11’s role in rheumatoid arthritis and think about other autoimmune and inflammatory diseases?” he asked. “The liver, the kidney, the heart, brain, and skin—to what degree do those tissues and the cell types in those tissues, and the cadherins on the cell types, influence the way those tissues respond to inflammation? And is this a model for looking at disease in other contexts?”
