The idea that there is a viruslike agent or some kind of invisible infectious component in the tissues affected by arthritis had its start in work done at the Rockefeller Institute between 1936 and 1939. The results of that work were published first in a brief article in Science in 1939, and then elaborated some ten years later in the rheumatic disease section of Post Graduate Medicine and Surgery. A paper by Dr. Louis Dienes and Howard Weinberger entitled “Experimental Arthritis: Pleuropneumonia-like Organisms and their Possible Relation to Articular Disease” spoke of the L-form of bacteria as “an invisible, filterable form that may exist after the parent germ disappears.”
I had arrived at a similar understanding through my own observations in the late 1930s, but from a different point of view. I had gone to the Rockefeller Institute to try to disclose the presence of some virus or infection in rheumatic tissues, using embryonated hen’s eggs as the tissue on which to grow the organisms. All of my results had been negative for the first three or four months; I had inoculated hundreds of eggs, and nothing at all had shown up on the membrane. Finally, when something did appear, it turned out to be the so-called L-form, or what we thought was bacteria. (The L-form took its name from the Lister Institute in London, where it was first described by Dr. Emmy Klieneberger. It was not until some years later that a committee of the American Microbiological Society decided to give this class of organisms a more formidable name. Mycoplasma derives from the Latin base words for fungus and fluid.)
By isolating a strain of the L-form organism from a Bartholin cyst, which occurs on female labia, Dienes became the first researcher to observe it in humans. I was the first one to find an L-form strain in joints and to connect it to arthritis.
I was asked to write a discussion of the Dienes-Weinberger paper, and I reported that we had observed this organism in a number of clinical situations. Seventeen patients, representing a wide variety of rheumatic diseases including rheumatic arthritis, rheumatoid spondylitis, chorea, erythema nodosum, and rheumatic fever, were treated with Aureomycin, a tetracycline derivative, because it had an effect on these organisms. As an interesting corollary, we noted that gold salts produced approximately the same result. And we were also impressed that when we started treating these patients with a tetracycline product (the Aureomycin), they invariably got worse before they started getting better—a phenomenon in medicine known as the Jarisch-Herxheimer reaction.
This discussion was the first linking in the entire medical literature of the Jarisch-Herxheimer effect with a tetracycline derivative in a rheumatic problem. Presumably the effect was due to the dislodging and breaking up of the mycoplasma and the releasing of antigen to a sensitized field. The effect showed a number of important principles at work. It demonstrated that the disease was a hypersensitive reaction, not to the drug itself but to the toxins that a germ creates in response to the drug’s presence. It showed that the germ must indeed be invisible, as Dienes and Weinberger had suggested, because no other germs of such standard types as streptococcus and staphylococcus were isolable. And it opened the way to a chemical attack on the whole area of arthritis and rheumatic diseases.
After our original research results were published in Science in 1939, our observations quickly became news. The New York Times ran a full page on the subject, and there was a lot of excitement among the media. But after a very short time, the excitement died.
The problem was that no one else was able to duplicate our results. This was thoroughly predictable; after all, it had taken me three months before I had been able to achieve those results in the first place, and other researchers tried it once or twice and gave up. When I left the Rockefeller Institute in 1939 and went back to Johns Hopkins to become chief resident in medicine, the school gave me a laboratory and eventually a fellow to pursue the subject further. I remained convinced that the subject was worth more work, but I found myself in a fast-shrinking minority in my conviction.
These original observations gave us a basis for looking at the possibility that rheumatoid arthritis was due to a kind of infectious allergy, and that mycoplasmas, unlike other germs that worked by invading the host, operated instead by sitting still in one place and giving off, in a pulsating fashion, something to which the body reacted. Because those emissions were irregular, there were periods when the disease was thoroughly quiescent and the person with arthritis would feel reasonably healthy. Conversely, there were other times—often associated with periods when the patient worked too hard or was under great stress or suffered an injury—when the hidden organism released toxins and the disease would flare up.
It was one thing to infer the presence of mycoplasmas by observing their effects, but it was often quite another to study them at first hand. To begin with, they proved to be both extremely fragile and very good at hiding, not only on the host cell but within it. And if the research effort did anything to damage the area in which these tiny organisms were located, the mycoplasmas would quickly break up. When that happened, it was similar to an egg breaking in the refrigerator, with the remnants running everywhere.
Over the years, we studied what happens when the mycoplasmic egg is broken. We learned, eventually, that the toxin leaving the organism is not, in itself, particularly vicious; if it were, we realized that it would make the patient terribly sick, since a little bit of the toxin is escaping from the mycoplasma all the time. Instead, we came to understand that when the toxin is released over a period of months or years, it creates a reception by “fixed tissue antibodies” that are ready to react to these toxins every time they appear.
The way in which the body becomes sensitized to mycoplasmas is similar to the way in which it learns to react to poison ivy. First-time visitors to the United States could take a shower in poison ivy resins, and the only thing most of them would get from it would be sticky. Until the body produces an antibody or allergen, there’s no reaction. By the same token, one could expect that it would take prolonged exposure to mycoplasma toxins before the stage would be properly set for a reaction to occur. Many of us could be carrying mycoplasmas from childhood—from an old infection, perhaps from viral pneumonia—that are ticking away inside our bodies, just waiting for that process to play itself out.
We also learned early on that mycoplasmas have a unique affinity for certain parts of the body, and they attach themselves there like no other germ. Most of what we know in that respect comes from work performed by other scientists on animals such as swine, chickens, and turkeys, particularly the research conducted by Albert Sabin on laboratory mice in which mycoplasmas were isolated, injected into the animals’ bloodstreams, and headed unfailingly to the joints.
The history of science is filled with stories of major discoveries which seem so obvious in retrospect that one marvels at how long they remained unknown. The development of penicillin in 1928, for example, was greeted with universal delight, followed almost immediately by chagrin among several generations of physicians who, like Alexander Fleming, had observed the effects of mold on cultures growing in a petri dish but unlike him had inferred nothing useful from what they saw.
It is a lot easier to recognize a cure than to understand the mechanism of a disease, especially when the disease is as complex as rheumatoid arthritis. That principle may help us to understand why some of the most elemental properties of arthritis went unrecognized for decades after they should have become perfectly obvious. And it may help as well in explaining why arthritis has such a sorry history of “cures,” many of which have proven more destructive than the disease itself.
The affinity of mycoplasmas for the tissue and fluid of the joints is just one such uniform factor that no one seemed to have noticed before, despite ample opportunities going back at least to the 1890s. At that time, a devastating epidemic of arthritis and pneumonia, called pleuropneumonia bovis, swept through the cattle herds of Europe, causing incredible mortality and raging across to the eastern end of Siberia. It was caused by a germ called Streptobacillis moniliformis, which grows in great chains in the culture media—and produces L-forms that mimic mycoplasmas. Scientists cultured the larger germ and, at a hundred magnifications on a microscope stage, they observed colonies of what looked like dozens and even hundreds of tiny fried eggs, unquestionably satellites to the parent organism. While they recognized that the satellite derived from the host, much as the moon originated in the area of the earth now occupied by the Pacific Ocean, they failed to understand that the fried eggs were another form of the same lethal agent and were at least as potent as the long strands of ropelike streptobacilli.
The satellite concept is useful in another way in appreciating how the mechanism of arthritis has remained so elusive. Let’s suppose that a visitor from Mars were to visit the moon and ask the moon where it came from. If the moon could point to the earth and say, “I came from there,” the Martian visitor shouldn’t have much difficulty with that answer. But now suppose that something cataclysmic had happened after the moon was formed, and the earth was no longer there. It becomes a lot harder for the visitor to say, “I see,” when the thing that he might have seen has disappeared.
This phenomenon of a kind of metamorphosis or variation from the streptobacillus into the L-form or mycoplasma—coupled with the disappearance of the original form—led to a tremendous debate between Dienes in the United States and Klieneberger in London. Dienes said that the mycoplasma derived from the streptobacillus with which it was originally intermingled and was another phase of the same creature. Klieneberger said the two phases were separate organisms, much as a pony is separate from a horse. The argument took on the aspects of a theological debate, Klieneberger representing the redoubtable Lister Institute, where the L-form had first been identified and after which it was named, and Dienes, on the other side of the ocean, releasing his arrows from behind the mighty ramparts of Harvard University.
Unhappily, as this battle raged, I found myself somewhere near the metaphoric equivalent of Bermuda, right in the middle. In my work at Johns Hopkins, we had recently isolated some streptobacillus from the joint fluid of a man who had been bitten by a rat. Streptobacillus was known to be carried by rats, and I recognized that the outcome of the debate would have profound implications for my work, and possibly vice versa.
In due course, Jack Nunemaker joined me and we were able to demonstrate that the L-form was a variant of the original streptobacillus organism, and we had reason to believe that all bacteria could indeed go into separate invisible forms. We showed that a germ could enter a state in which it could become invisible, pass through a filter, and then return to the parent form. It was a whole new world of subclinical phenomena, a world in which things could appear and disappear, take new shapes, be at once different and the same—the biochemical equivalent of solid ice changing into water or steam and back again as circumstances require. Was it possible, from the long persistence of the forms in tissues, that bacterial allergy would evolve?
This view changed the way we looked at certain kinds of disease. A person could have osteomyelitis, for example, and the staphylococcus that caused it could clear up, leaving the patient fully healed, with no sign of the infectious agent no matter how many times it was cultured. But if that person then were to be given a lot of cortisone—enough to interfere with his immune mechanism—all of a sudden the old staphylococcus would come back to life again with a vengeance, returning to the visible world from out of nowhere, from out of the persisting invisible L-form.
When one accepts that Lamont Cranston is also the Shadow and can move back and forth between at least two states and maybe among several, it is easy to see implications for our understanding the mechanism of other diseases than arthritis. Multiple sclerosis, for example, and amyotrophic lateral sclerosis (Lou Gehrig’s disease) have all kinds of invisible but seemingly infectious states.
From that point on, we continued looking for the L-form, or mycoplasma, separate from any other organism. I realized that we couldn’t do this by inoculating tissues of animals or tissue cultures, because that procedure would leave us open to the suspicion of contamination. So instead we worked with exudates from the body and the blood directly into culture media, without using animals, eliminating the risk that we would pick up anything along the way. For thirty years, we tried to grow mycoplasmas from joint fluid, knowing from the start that they are extremely fragile structures and very difficult to cultivate outside the body.
Mycoplasmas are not viruses. True viruses require living cells in which to grow. Mycoplasmas are somewhere between a virus and a bacterium, but the real difference is that they contain RNA and DNA and are self-contained, living units. That difference is important, but their similarities to some viral forms were to prove instructive as well, especially when we began looking for mycoplasmas inside the human body and within living cells. A couple of years ago, armed with new, state-of-the-art equipment for electron microscopy and protein fractionation, we set about doing just that.
We were also armed with a formidable new director of research. Millicent Coker-Vann, with a Ph.D. in biochemistry in viral immunology, came to the Arthritis Institute from the National Institutes of Health, where she had been working in particular with the slow virus group. Although she had been connected with the NIH for more than a decade, along the way she taught for two years at the University of Ife in Nigeria, spent seven months as a fellow at the Max Planck Institute in West Germany, and received a research grant from the Rockefeller Foundation.
Breaking down the proteins of the body is somewhat like unwrapping a cocoon. Under Dr. Coker-Vann’s guidance, we soon found that by fractionating the protein, we could easily layer off the mycoplasma antigen fraction, draw it off, put it in a rabbit, and develop antibodies against it which would match those of the patient. In a very short time, we were able to complete the circle back to the patient without having to isolate any of the elements from animal intermediaries.
Now we had both the mycoplasma antigen and the antibody in relation to the disease. We have found higher concentrations of both the antigen and the antibody in the joint fluid than in the blood. And the levels of both antigen and antibody have been noted to decrease following intravenous antimycoplasma therapy.
Early in the course of pursuing the new fractionation technology, we found that it wasn’t necessary to have the whole living mycoplasmic organism to provoke the reaction in a joint or tissue. Our studies proved what we had previously been able to assume only by inference—that just the fragments of mycoplasmas were sufficiently potent to create a powerful antigenic reaction, causing the host to produce antibodies in reaction to their presence.
Over the course of time, we accumulated a larger and larger body of evidence that mycoplasmas were indeed the mechanism for rheumatoid arthritis.
First of all, only antimycoplasma drugs had any impact on rheumatoid disease, except in those instances where there was a strong streptococcal connection. Ampicillin may be needed to lower the streptococcal antigen level and reduce symptoms. Tetracyclines were the only group of antibiotics we could find that suppress mycoplasma growth in the laboratory, and they were the only ones that seemed to improve the arthritis patients’ conditions—unless there were other bacterial complications which added to the sensitizing process.
A further body of data was derived from studies of the other medicines, potions, and old wives’ cures that have been used over the years to combat arthritis. Quinine is a long-time remedy for arthritis and a ubiquitous component of patent medicines that goes back centuries; it was used by our great-grandparents for backaches and rheumatism—so much so that many of our ancestors developed serious quinine allergies. We demonstrated, not surprisingly, that quinine has a specific effect on mycoplasmas, as does Plaquenil (hydroxychloroquine sulfate), another antimalarial substance to which quinine is closely related. Gold has also been found effective in suppressing mycoplasma growth, explaining its effectiveness in arthritis. We demonstrated the antimycoplasmic action of gold nearly forty years ago. Copper salts were first used by J. Forestier at about the same time he discovered the value of gold in rheumatoid disease, and he particularly noted their beneficial effect on rheumatoid arthritis; again, we demonstrated the ability of this substance to suppress mycoplasma growth. By the same token, it has been observed that copper bracelets react with perspiration to form copper salts, and work in Australia recently proved that those copper salts can penetrate the skin, which would explain the popularity of this preventive talisman from as long ago as the time of the ancient Romans.
Collectively, our observations confirmed a pattern in which mycoplasma could be seen as the consistent offender in the process by which rheumatoid arthritis occurs, and that pattern more than justified our efforts to isolate it.
We are also trying to “tag” the mycoplasma organism so we can see where it combines with human tissue. Tagging is a procedure by which the element under study is labeled with a tracer which signals its presence in microscopy. The reason for this approach is that without tagging, once the mycoplasma enters tissue, it is lost to view and can’t be identified with any certainty, even with the electron microscope.
Regardless of what we learn about the final pieces of the rheumatoid arthritis puzzle in the time just ahead, I believe we already know enough about the mechanism—inferentially and from direct observation in our laboratory research, and from forty years of clinical experience—to state with certainty that mycoplasmas are the primary infectious agent, and that tetracycline therapy is the only effective therapy to reach toward the cure of rheumatoid arthritis.
The cause of rheumatoid arthritis is an antigenic substance which operates not as an invader, in the way of germs, but as the trigger for an internal allergic response, releasing toxins intermittently to a sensitized area, subsiding then reappearing We know that antibodies are transported throughout the body by white blood cells and platelets, much as smoke jumpers are carried by an airplane from one brush fire to the next. It is by this means that arthritis migrates, in the knee one day and in the shoulder or the back the next: the antigen builds up in the shoulder, and the antibody that was busy counteracting it in the knee moves on to the new battlefield, leaving the knee to a brief season of peace.
Consider the alternatives. If one is to believe that rheumatoid arthritis is just a hereditary problem, or that it is due to stress or nutritional deficits or aging or a deficiency in the body’s natural supply of cortisone, how is it possible to explain, for example, that the condition flares up the day before a storm comes?
One way in which myths are preserved is by the denial of evidence that would disprove them. Advocates of the genetic, nutritional, stress, aging, and cortisone-deficiency schools of arthritis simply dismiss the indications that arthritis is barometrically poised by calling them old wives’ tales. But Dr. Joseph Hollander, the head rheumatologist at the University of Pennsylvania in Philadelphia, did a study some years ago of the effects on rheumatoid disease of barometric pressure, along with such other factors as temperature, humidity, and oxygen. He got volunteers to stay in sealed, climatically controlled rooms, writing their impressions of how they felt from hour to hour as he worked various subtle changes in their environment. It can be assumed that some of the changes, such as in humidity, might have been detectable through normal sensory means, but most of the changes were not. Hollander concluded, on the basis of excellent clinical evidence, that environment does indeed make a measurable difference in one’s sensitivity to rheumatic disorders. Two factors were found to cause flare-ups: a sudden drop in barometric pressure, and the presence of high humidity in conjunction with such a drop. Both of them are common atmospheric characteristics prior to a storm. The barometric effect fits our concept perfectly by promoting a sudden release of antigen to a sensitized area.
I have always asked new patients about their experience with changes in weather, and also about their sensitivity to changes in the seasons. It has been well known for many years that patients with rheumatoid arthritis experience a higher incidence of flare-ups in the late spring and early fall. Very little work has been done in understanding the impact of seasonal changes on disease, although the increased incidence of depression that comes with the arrival of fall has been widely noted for a number of years, perhaps because it is supported by a measurable indicator in the form of a corresponding rise in the suicide rate. Our clinical experience with rheumatoid arthritis has shown that it, too, is subject to seasonality.
One other old wives’ tale is worth examining for the clues it might provide to the arthritis mechanism, and that is the ancient belief that the pain in arthritic joints can be relieved by bee venom. This rather painful therapy is commonly practiced by the residents of New Hampshire, as indeed it is by rural folk throughout the world. New Hampshire happens to be favored by its proximity to Boston, and many members of the immense medical community in that city spend their weekends among the rustics in the Granite State, observing their ways.
As might be expected, visiting doctors who hear about beesting therapy warn its practitioners that they are risking a fatal reaction to the venom—which is indeed the case—and they usually dismiss the claims for its efficacy out of hand. But the claims persist. Following a hunch, several years ago Dr. Harold Clark in our laboratory sent away for some bee venom and our microbiologist, Jack Bailey, tested it against various strains of mycoplasma. It proved to be a very potent suppressor of their growth.
I am happy to report that the bee venom story isn’t likely to end there. Some researchers in Great Britain have been studying its effects on arthritis and attempting to dissect it chemically, aiming toward the possibility that they will be able to separate the curative component from the part that creates the sting. It would be my own guess that the cure and the sting are one and the same thing, but I am delighted to see more research that is directed toward the mechanism by which arthritis occurs.
In the main, the history of arthritis research over the past four decades has been one of frustration, suffering, or serial disappointment as one miracle cure after another failed to live up to its early promise. The reason for those failures always comes down to one basic flaw: no major disease has ever been cured before science has first developed an understanding of how it works.
At long last, the main parts of the puzzle are falling into place, and our understanding of the fundamental mechanism of rheumatoid arthritis is now nearly complete. There are still many aspects that need to be explored, but this new framework allows us to approach them in a different light, no longer dismissing the disease as hopeless and without a known cause.
This is the framework within which arthritis occurs. The process taking place within that framework is often far more complicated than I have shown, and some of its parts are still only dimly seen. There may well be other bacteria that contribute to the process of arthritis once it starts. Many people with tonsillitis or bad teeth or kidney infections may experience flares in their arthritis from these additional sources of antigen, just as people with noses that are sensitized to ragweed will find that they have also become sensitive to such other stimuli as house dust and feathers. But as surely as the primary cause of hay fever is ragweed, the starting point for arthritis is mycoplasmas.
Once a person’s joints have become sensitized to mycoplasmas, which move in and settle there, other antigens can move into the same neighborhood and compound the problem. If there is a strong history of streptococcal infection (rheumatic fever, scarlet fever, frequent strep throats with ear and sinus complications) or a strongly positive ASO (streptococcal antibody level), additional treatment with appropriate antistreptococcal medication as a therapeutic probe is often indicated. Regardless of these other factors, however, the basis for the initial sensitivity is still mycoplasma, and if the mycoplasma can be removed, the sensitivity level of the rest of the body can be lowered as well and the other factors become more manageable. This framework provides the conceptual core for at last proving the cause of rheumatoid arthritis and finding its cure.