Cracking the Code: The Rheumatologist’s Role in the Human Microbiome Project
The circumstances surrounding the Human Microbiome Project illustrate one of the hallmarks of scientific advancement: a giant leap forward, followed by a great deal of dirty work. Although the Project itself is a notable achievement, for many rheumatologists, the 2013 data set published in eLife by Scher and colleagues demonstrating the link between Prevotella bacteria and rheumatoid arthritis was the first meaningful step for the field. Similar studies have followed, with research groups as close as Connecticut and as far as Japan trying to make sense of the vast body of information the NIH undertaking yielded.
In 2016, Healio Rheumatology reported on the first explosion of new data from the Project, detailing the potentially significant influences of the microbiome on rheumatic diseases; it is now time to check in to see how far the experts have come as they delve into mouse models, microbial profiling and the types of disease-specific associations that may ultimately lead to targeted therapies.
As yet, breakthrough discoveries have only begun to trickle in. That does not mean, however, that there will not eventually be a deluge, according to James T. Rosenbaum, MD, chair of the division of arthritis and rheumatic diseases, and Edward E. Rosenbaum Professor of Inflammation Research at Oregon Health & Science University.
“This whole field is in its infancy,” he said. “The initial charge of the Human Microbiome Project was just to create a catalog. With a few exceptions, we are still in the cataloging phase.”
Renuka R. Nayak, MD, PhD, assistant adjunct professor in the division of rheumatology at the University of California, San Francisco, laid out the pathway from cataloging to treatment. “In order to translate our knowledge of the microbiome into therapies for patients with rheumatic disease, we need to conduct studies looking at the mechanisms by which microbes influence rheumatic disease progression and treatment,” she said. “Mechanistic studies of mouse models of disease coupled with longitudinal studies of patient cohorts will provide a foundation for conducting clinical trials in patients. Once we understand the mechanisms, we can develop targeted therapies.”
The mouse models are currently underway, as is investigation of specific microbes, along with analysis of specific disease states. In the aftermath of the deluge of microbiome data, researchers are examining the gut and the mouth and considering how antibiotic therapies can also fit into the equation.
Gregg J. Silverman, MD, professor of medicine and pathology at NYU School of Medicine, discussed the overall impact of this research as it pertains to rheumatic disease. “There is an emerging consensus that points to a microbial-specific role in determining both immune responsiveness in an individual and as triggers of inflammation,” he said.
While the evidence begins to mount that the microbial populations of patients with RA, psoriatic arthritis, systemic lupus erythematosus and gout are different from those found in healthy controls, Rosenbaum stressed that the clinical community is still in the dark about how this information will lead to interventions. “That is our challenge,” he said. “We haven’t solved it yet.”
Lachnospiracaea, Enterococcus gallinarum Data
During the 2017 American College of Rheumatology annual meeting, Silverman and colleagues presented findings demonstrating an association between lupus nephritis and intestinal commensal Lachnospiracaea species. Patients with SLE were reported to have a fivefold greater representation of a species in the Blautia genus of the Lachnospiracaea family of obligate anaerobic Gram-positive cocci, according to the findings. Additionally, patients with active nephritis at the time of sampling had higher levels of IgG anti-strain antibodies when compared with SLE patients without renal activity.
“We saw an expansion of a single strain of bacteria that was associated with lupus nephritis,” Silverman said. “We are developing an antibody test from those studies, learning lessons from mouse models as well as health and disease models that will lead to the clinic.”
Additional findings from the study showed that a class of anaerobes identified in all lupus patients were most expanded in those with most active lupus disease, and especially those with active kidney disease. “The more active the lupus was, the more of these bugs we saw — since lupus is a disease of hyper-activation of antigens, we wonder whether antigens released from these bacteria in the intestine are central disease drivers,” Silverman added.
In a recent study published in Science, Vieira and colleagues reported that translocation of Enterococcus gallinarum to the liver and other systemic tissues was linked to autoimmune response in mice with a genetic predisposition to autoimmunity. Moreover, E. gallinarum–specific DNA was found in liver biopsy results of patients with autoimmune diseases. “These discoveries show that a gut pathobiont can translocate and promote autoimmunity in genetically predisposed hosts,” the researchers wrote.
“The E. gallinarum story is the closest we have come to finding a specific bug that is implicated in lupus-related liver disease, and that we can target,” Rosenbaum said, noting, however, that some challenges remain. “First off, someone needs to replicate this, of course. Also, is there a single bacterium responsible? What does it mean to kidney disease, skin disease or lupus in other organs?”
Despite these questions, Rosenbaum remains undaunted. “I am not dismayed that we haven’t solved this yet,” he said. “We are just getting started. If there was a single bug we could immunize against, that would be unbelievable. We are going to solve it — it is just a matter of time and technology.”
Nayak put all these findings in perspective. “While initial studies showed correlations between microbes and disease, more recently, we are also seeing cases of microbes playing causal roles in mouse models of autoimmune disease,” she said, but warned that significantly more research will be necessary. “Demonstration of a causal role of the microbiome in human disease is much more difficult to demonstrate without randomized controlled trial data.”
Many experts acknowledge that the work being done in the lab of Allen C. Steere, MD, professor of medicine at Harvard Medical School and director of translational research in rheumatology at Massachusetts General Hospital, is critical in moving the ball forward from Scher’s data. “In addition to Scher’s identification of over-expansion of Prevotella copri in stool samples in RA, he subsequently reported gut dysbiosis with prominence of certain other organisms in some patients with ankylosing spondylitis or psoriatic arthritis,” he said in an interview. “However, it was not clear whether these changes in the bowel microbiota had relevance regarding immune responses in joints.”
In 2017, Steere’s group published two data sets that provided initial evidence for immune relevance of Prevotella copri in a subgroup of patients with RA. In one study, Wang and colleagues identified a T cell epitope of a 27-kD protein of P. copri based on the identification of HLA-DR–presented peptides directly from patients’ peripheral blood mononuclear cells, according to Steere. In another study led by Pianta, the group showed that 42% of patients with new-onset RA had Th1 responses to this P. copri protein.
“In addition, 32% of RA patients had IgG or IgA P. copri antibody responses to this 27-kD protein or the whole organism,” Steere told Healio Rheumatology. “In contrast, P. copri antibody responses were rarely found in patients with other rheumatic diseases or in healthy controls, suggesting that antibody responses to this organism were specific for RA.”
The subgroup of RA patients with IgA antibody responses to P. copri often had Th17 cytokine responses in serum and frequent antibodies to citrullinated protein antigens, suggestive of a mucosal immune response, according to Steere.
“In contrast, the subgroup with IgG antibodies to the organism frequently had P. copri DNA in joint fluid, P. copri-specific Th1 responses, and less frequent ACPAs, suggestive of a systemic immune response,” Steere said. “Moreover, the positive PCR test for P. copri DNA in joint fluid suggested that the organism may sometimes spread systemically to joints, presumably within host immune cells, thereby providing a possible link between this gut microbe and joint disease in RA.”
“We have known for a number of years that certain specific bugs affect our immune response in terms of Th17,” Rosenbaum said. “What the researchers did was look at 53 different bugs in germ-free mice and evaluated them one at a time to see how they perturbed the immune system. Most had a sizeable effect. This both clarifies certain aspects of the challenge and shows us how complicated the system is.”
A key finding from Steere’s group is evidence of mimicry as a cause of RA, according to Rosenbaum. “In mimicry of this type, a component of a microbe would look enough like a human cell that the body gets collateral damage when the immune system responds,” he said. “We think mimicry causes rheumatic fever. What Steere and colleagues found is that there were specific bacteria inducing autoantibodies associated with RA but not the classic antibodies of RA.”
In the study from Wang and colleagues published in the Journal of Proteome Research, based on the identification of HLA-DR–presented self-peptides directly from RA patients’ synovial tissue, the two autoantigens identified were N-acetylglucosamine-6-sulfatase (GNS) and filamin A (FLNA), which were targets of T-cell and B-cell responses in more than half of patients with RA, primarily those with P. copri immune responses, according to Steere.
“Both self-proteins were highly expressed in the joints of RA patients,” he said. “The T cell epitope of GNS identified in synovial tissue had marked sequence homology with epitopes of sulfatase proteins of Prevotella species and Parabacteroides species, whereas the T cell epitope of FLNA identified in synovial tissue had marked sequence homology with epitopes of Prevotella speciesand Buyticimonas species. Moreover, patients with T-cell reactivity with each self-peptide also had responses to the corresponding microbial peptides, and the levels correlated directly. Thus, sequence homology between T cell epitopes of two self-proteins and a related order of gut microbes may be another possible mechanism linking mucosal and joint immunity in patients with RA.”
However, Steere is not the only researcher investigating these associations. Maeda and Takeda used gnotobiotic experiments in the guts of mice to demonstrate that dysbiosis in patients with RA was a contributing factor in developing Th17 cell-dependent arthritis. However, other findings showed that Prevotella histicola from human gut microbiota had a suppressive impact on arthritis development.
““We have made progress for sure,” Martin Alexander Kriegel, MD, PhD, of the department of immunobiology at Yale University School of Medicine, said in an interview. “This is a fast-moving field, with many recent discoveries including work from Japan showing gut bacteria-induced Th17 cells are implicated in aggravating or worsening an experimental model of inflammatory arthritis.”
In a study from Kriegel’s lab, Greiling and colleagues studied Ro60 orthologs from the skin, mouth, and gut bacterial species in patients with lupus as well as healthy controls. Working in conjunction with Sandra L. Wolin, MD, PhD, from the RNA Biology Laboratory at the National Cancer Institute, the researchers found that Ro60 autoantigen-specific CD4 memory T-cell clones were activated by Ro60-containing bacteria in the lupus group. Other findings showed that in germ-free mice, anti-human Ro60 T-cell and B-cell responses occurred, which the researchers suggested links anti-Ro60 commensal responses in vivo with the production of human Ro60 autoantibodies and signs of autoimmunity.
“The study from Dr. Kriegel’s group showed an imbalance of a bacteria present in excess escaping the intestine,” Silverman said. “This could be relevant to a subset of patients with autoimmune diseases. We must consider that these studies are very early.”
Translating Advances into Treatment
Despite research progress, Nayak highlighted some unexplored territory. “We have yet to pinpoint the mechanisms by which specific bacteria or the proteins or metabolites they produce lead to disease or modulation of treatment,” she said. “Without a better understanding of precise mechanisms, it is hard to develop targeted microbiome-based therapies to improve patient care. Thus, few microbiome-based discoveries have led to concrete improvements in patient care thus far.”
Nayak then took the discussion one step further, suggesting that the research must move beyond simply finding treatments that work — understanding who would likely respond is the next level. “Associations or correlations between microbes and their products, such as proteins, metabolites, and disease pathogenesis and treatment may be useful in predicting patient disease progression or response to treatment,” she said. “These interactions are mediated by the bacteria themselves or by the metabolites they produce.”
For Steere, these findings may someday have an impact on adjunctive treatment strategies in RA. First, targeted antibiotic therapy may be of value. Second, “identification of pathogenic T cell epitopes in synovial tissue — as was done in our second study — may make it possible to engineer blocking peptides, which would limit autoimmune stimulation and ameliorate these presumably disadvantageous autoimmune responses,” he said.
Although Silverman acknowledged that antibiotic therapies may ultimately play a role in the rheumatic disease arena, he offered a broad warning about their overuse, focusing on early life as a predictor of autoimmune disease. “When we are born, we are essentially sterile,” he said. “The first bacteria that colonize the gastrointestinal tract are in the birth canal. From there, exposure to family members and friends, and the world beyond are the training ground for our immune system to deal with the dirty world outside.”
An unintended consequence of an overly sterilized world, a world where courses of antibiotics are distributed frequently and easily, is that the immune system loses some of the dynamic capabilities that should begin in the birth canal and build during very early life, according to Silverman. “How many courses of antibiotics do children receive in their first few years? Our increasingly ‘clean’ world has shifted the environment we live in,” he said. “In short, it may be weakening our immune system; for our purposes, if you inherit genes that make you susceptible to autoimmune diseases, it may be that much easier for imbalances in the distribution of bowel bacteria to trigger their onset.”
Kriegel’s take is that the era of personalized medicine, particularly in terms of the human genome, is dovetailing with advances in understanding of the microbiome. “The microbiome is all about personalized medicine because everyone has a unique microbiome,” he said. He acknowledged certain gut microbes play a role across disease states, but pointed to subsets of patients. “Some patients have antibody positive RA, others are negative. Similarly, some lupus patients have certain clinical features different from others. These can be associated with different microbial signatures.”
It may then be a logical leap to apply lessons about both the genome and the microbiome to lupus, psoriatic arthritis, ankylosing spondylitis and other rheumatic diseases, Kriegel suggested. “To make this complex field simpler, it is not hard to imagine a time when we will combine genetic studies and microbiome studies to target individual patients,” he said.
Beyond Traditional Therapeutic Options
While the microbe- and disease-specific studies have pointed toward therapeutic directions, many experts, including Rosenbaum, are getting back to basics. “As a practicing rheumatologist, I am now more sympathetic to the idea that diet impacts my patients’ disease,” he said. “I will admit I used to think diet was predominantly a placebo effect; however, now we know that what you have growing in you depends on the substrates going in to you. We may ultimately find that diet has a much more profound effect on our patients and their microbial populations than we originally anticipated.”
Rosenbaum cautioned that studying diet brings a host of challenges for those used to the controlled atmosphere of a clinical trial. “It is difficult to get a group of subjects to adhere to a specific diet,” he said. “It’s difficult to control one variable at a time. If the ecosystem in our body is complicated, diet is more so. There is no doubt that diet is critical, but it will take some time to sort out the variables.”
“Dietary interventions might be successful in correcting gut dysbiosis,” Steere said. “A more targeted approach might be fecal matter transplants, in which strains of gut commensal bacteria that become pathogenic are replaced by harmless commensal bacteria. However, any new adjunctive therapy will need to undergo systematic testing in clinical trials before it can be recommended for treatment of patients with a particular rheumatic disease.”
Silverman carried the conversation about fecal transplants to its natural conclusion in personalized medicine. “Patients are now banking their own fecal microbiome before entering bone marrow transplantation for which they receive broad-spectrum antibiotics that demolish the communities in their bowel, and put them at risk for developing a condition called Clostridia difficile colitis,” he said. “The doctors at Sloan-Kettering, for example, are looking at the imbalances and inconsistencies in the microbiome to determine if the patient should then later undergo a fecal transplant to reestablish a healthy balance in the bowel. While there is a lot of interest, it may be premature to use this as a therapy for our patients.”
For now, experts like Steere must continue to chip away at the edges. “Knowledge about gut dysbiosis and resulting immune responses in RA or other rheumatic diseases is not yet ready for translation into effective treatments for these diseases,” he said. “However, there is great interest in whether such knowledge may lead to new adjunctive therapies that could greatly improve the prognosis of these diseases.”
The first step will be to reproduce our findings in other patient populations, according to Steere. “Second, it is not yet known whether these microbial and related autoimmune responses precede the onset of joint inflammation in RA, implying that they may cause the joint disease, or whether gut dysbiosis results from joint inflammation, implying that they are not a primary cause of the joint disease,” he said. “Third, P. copri is likely not the only immunogenic gut microbe; other related organisms could well be involved too. Fourth, microbes in the periodontia, bronchi, or lungs could also be microbial triggers in other patient subgroups.”
For Kriegel, combining mechanistic research with profiling of patient cohorts is critical, but with an additional component. “We need to look at patients not at a single timepoint, but at multiple timepoints: during flares, remission, and before they develop disease,” he said. “This will help find targets in the microbiome.”
This type of research will entail animal models besides human studies. “Of course, there is always the caveat that much of what we find in mice will not translate to humans,” Kriegel said. “But mechanisms of how bugs affect the body often translate. In addition, we can get a closer look at some of the candidates from human studies and examine them in more detail using so- called ‘germ-free’ mice that are devoid of all other commensal bacteria.”
Despite all of these challenges, Rosenbaum believes in the power of research. “Based on the studies we are seeing in germ-free mice and others, I have to believe that in 20 years, we will have strategies based on the microbiome to treat RA, lupus, ankylosing spondylitis and PsA,” he said. “If we can extrapolate the importance in mice to importance in people, we will be able to re-educate our immune systems in such a way that these diseases will become quite rare. But first, we have to understand that the microbiome is not slightly important, it’s incredibly important.” – by Rob Volansky
- Greiling TM, et al. Sci Transl Med. 2018;doi:10.1126/scitranslmed.aan2306.
- Maeda Y, Takeda K. J Clin Med. 2017;doi:10.3390/jcm6060060.
- Pianta A, et al. Arthritis Rheumatol. 2018;doi:10.1002/art.40003.
- Pianta A, et al., J Clin Invest. 2017;127:2946-2956.
- Scher JU, et al. eLife. 2013;doi:10.7554/eLife.01202.001.
- Silverman, GJ, et al. Abstract #1786. Presented at the American College of Rheumatology Annual Meeting, Nov. 4, 2017; San Diego.
- Vieira SM, et al. Science. 2018;doi:10.1126/science.aar7201.
- Wang Q, et al. J Proteome Res. 2017;16:122-136.
- For more information:
- Martin Alexander Kriegel, MD, PhD, can be reached at 10 Amistad Street, New Haven, CT 06509; email: email@example.com.
- Renuka Nayak, MD, PhD, can be reached at 513 Parnassus Avenue, HSW 1501 / Hooper Foundation San Francisco, CA 94143; email: Renuka.Nayak@ucsf.edu.
- James T. Rosenbaum, MD, can be reached at 3181 SW Sam Jackson Park Rd., Portland, OR 97239; email: firstname.lastname@example.org.
- Gregg J. Silverman, MD, can be reached at 435 East 30th St., New York, NY 10016; email: email@example.com.
- Allen C. Steere, MD, can be reached at CNY 149-8, 149 13th Street, Charlestown, MA 02129; email: firstname.lastname@example.org.
Disclosures: Kriegel reports being an employee of Roche. Nayak reports consulting for Senti Biosciences. Rosenbaum reports a pending collaboration with Open Biome. Silverman reports he is a consultant for Lilly, Onyx, Pfizer, Quest, and Roche and receives grant support from the American College of Rheumatology Research Foundation, the Lupus Research Institute and NIH. Steere reports no relevant financial disclosures.