Research in The New England Journal of Medicine has opened new avenues for exploring the pathophysiology of disease flares in rheumatoid arthritis.1 Through longitudinal genomic analysis, researchers have identified a naive B cell signature prior to rheumatoid arthritis flares, as well as a type of mesenchymal cell, that may play an important role in flare pathways.
Background
The first author of the paper, “RNA Identification of PRIME Cells Predicting Rheumatoid Arthritis Flares,” Dana E. Orange, MD, assistant professor of clinical investigation at the Rockefeller University and assistant attending rheumatologist at the Hospital for Special Surgery, New York, explains, “We started this study because we noticed a gap in understanding how rheumatoid arthritis develops—a gap between the known risk factors for developing a diagnosis of rheumatoid arthritis and the actual patient experience of flares.”1
Although researchers have identified certain genes and environmental factors, such as smoking, that increase the risk of rheumatoid arthritis, much about the immediate cause of disease flares remains poorly understood. To help address these questions, Dr. Orange et al. collected longitudinal genomic data from a cohort of patients, allowing the authors to compare differences in gene expression prior to flares, during flares and during quiescence.
As pointed out in an accompanying editorial, previous analyses utilizing micro-array technology had not yielded a great deal of helpful information about differences in gene expression associated with levels of disease activity.2 These earlier studies had only been able to sample sparsely—every three months or less frequently. Dr. Orange points out that with such data sets, it’s difficult to look at flares in a granular way. The researchers attempted to collect a much greater swath of data, hoping to look at differences in genetic expression that might be occurring just prior to flare.
Longitudinal RNA Sequencing Analysis
The study authors used RNA sequencing to identify patterns of gene expression from one index patient with rheumatoid arthritis. Self-administered fingerstick blood specimens from 364 time points encompassed eight flares over four years. They eventually added information from an additional 235 time points from three more rheumatoid arthritis patients.
The patients assessed their disease activity using the Routine Assessment of Patient Index Data 3 (RAPID3) questionnaire, which compiles scores of patient-perceived pain, functional impairment and a global health impression. Physician assessments of disease activity were also performed at monthly clinic visits using both RAPID3 and the Disease Activity Score 28 (DAS28), as well as with standard blood count tests.
First, the team looked at gene expression differences between baseline and flare. They noticed an increase in the expression of genes related to inflammation during the flare, such as those reflecting an increased white blood cell count, which corresponded to flares as assessed by the patient (via the RAPID3).
Researchers have identified a naive B cell signature prior to rheumatoid arthr itis flares, as well as a type of mesenchymal cell, that may play an important role in flare pathways.
But they noticed other patterns as well. “What we weren’t expecting was that we saw certain genes go down during flare, representing things that we wouldn’t have expected blood cells to make,” Dr Orange notes. “It was a little surprising to see in the blood all these things going down that represented pathways of cartilage and bone development.”
To dig deeper, the group used a time-series longitudinal analysis to help identify genetic expression patterns prior to flare. They identified a cluster of genes that became activated two weeks before a flare, which they termed antecedent cluster 2 (A2). They also identified a different set of genes, antecedent cluster 3 (A3), that became highly activated one week before the flare and then were downregulated during the flare itself. Dr. Orange explains, “That group of genes really overlapped with what we had seen previously: There were these cartilage and bone and extracellular matrix genes.”
Single-Cell RNA Sequencing of Synovial Cells & Flow Cytometry Techniques
To try and make sense of these results, the team turned to a previously published data set. Zhang et al. analyzed synovial tissue samples from patients with rheumatoid arthritis or osteoarthritis.3 They used bulk RNA sequencing and single-cell RNA sequencing of individual T cells, B cells, monocytes and fibroblasts. Unlike standard RNA sequencing, which looks at aggregated genetic expression from multiple cells, single-cell RNA sequencing allows scientists to analyze genetic expression in a single, specific cell.
Using this and flow cytometry, which can separate cells into subtypes based on various cell surface proteins, the researchers had identified multiple distinct cell populations, with different patterns of genetic expression and cell surface proteins. These included three distinct types of sublining fibroblasts, which live in the synovial sublining layer of the joint synovium. Some of these sublining fibroblasts are present in greater numbers in the inflamed synovium of patients with rheumatoid arthritis.3
Combining these existing data with their findings, Dr. Orange et al. performed an analysis of the A2 and A3 gene clusters. The A2 cluster was enriched with naive B cell genes. The A3 cluster was heavily enriched with the specific genes that Zhang et al. had demonstrated were prominent in sublining fibroblasts. When they looked at the expression of genes that were common to both these synovial sublining fibroblasts and the A3 gene cluster, they again found increased gene expression one week prior to flare and decreased expression during the flare itself.
All solid tissues have fibroblasts residing in them that are responsible for making and maintaining the tissue scaffolding. Traditionally these cells are defined as being of mesenchymal origin, and they are generally not thought to circulate in the blood. But this is exactly where Dr. Orange et al. had found this genetic signature.
“It has been becoming clear for a while that fibroblasts [in the synovium] in rheumatoid arthritis are quite activated. They express inflammatory cytokines and metalloproteinases and are thought to be playing a pathogenic role in rheumatoid arthritis,” says Dr. Orange. However, no role had previously been established for a fibroblast-like cell in the blood in rheumatoid arthritis pathogenesis.
PRIME Cells: Fibroblast-like Cells in the Blood
To further confirm the potential source of this genetic pattern, Dr. Orange’s group performed fluorescence-activated flow sorting—a type of flow cytometry—on a group of blood samples. These were collected from an additional 19 patients with rheumatoid arthritis, irrespective of disease activity, as well as 18 healthy controls.
The team identified the cells based on the same membrane protein markers that Zhang et al. had used to sort synovial fibroblast cells, sorting for cells that did not express CD45 (indicating cells not of hematopoietic origin), did not express CD31 (indicating non-endothelial cells) and did express podoplanin (PDPN; a marker associated with fibroblasts). Cells expressing these specific characteristics of synovial fibroblasts were more common in the blood of patients with rheumatoid arthritis than they were in controls.
The team then performed RNA sequencing on these cells and found that they too were expressing large amounts of genes from the A3 gene cluster, including many extracellular matrix genes, as well as characteristic synovial fibroblast genes and proteins.
Dr. Orange explains, “Because it was so strange to think of a fibroblast in the blood, we didn’t want to call them that. So we called them PRIME cells, standing for preinflammatory mesenchymal, meaning they are not hematopoietic but mesenchymal in origin (because they are CD45 negative).”
Interpretation & Disease Model
Other data from humans demonstrate that some types of synovial fibroblasts play a key role in the pathogenesis of rheumatoid arthritis, furthering inflammation and causing joint degradation.2 Inflammatory sublining fibroblasts have previously been found next to the blood vessels in inflamed synovium from patients with rheumatoid arthritis, where they secrete proinflammatory cytokines.4 PRIME cells also appear to be phenotypically similar to a fibroblast-like cell that exacerbated inflammatory arthritis after passive cell transfer in a mouse model.5
Under the model proposed by Dr. Orange et al., B cells activate these PRIME cells just prior to flares, as indicated by upregulation of group A2 gene clusters and later the A3 gene clusters. “Since PRIME cells overlap with these cells that we see in rheumatoid arthritis synovium, we now hypothesize that they left the blood and trafficked to the synovium,” explains Dr. Orange. This is inferred from the drop in expression of A3 gene clusters in blood samples noted during the flare itself. In the synovium these PRIME cells—or perhaps their descendant cells—may promote joint inflammation.
Dr. Orange adds, “We’re not sure at this point whether they are a precursor to a sublining synovial fibroblast. They did express some mesenchymal stem cell-like genes, so that’s one possibility. Or it could be that the synovial fibroblasts somehow get out of the tissue and go into the blood. It’s not clear if they are upstream or downstream, but they are very closely related.”
The study authors suggest that B cell activation may have triggered these PRIME cells, based on what we know of the timing of gene expression and B cells’ general roles as potential immune activators. But currently the causation is not completely clear. Dr. Orange says, “It could be that there is some upstream activator that activates B cells and also activates PRIME cells, but the PRIME cells just take longer to get going.”
This work, suggesting one pathophysiological pathway for B cell activation, streams into an existing body of research showing that B cells play a key role in driving autoimmune responses, working as antigen-presenting cells and producing cytokines and antibodies.2 “There are lot of reasons to think that B cells play some role in arthritis,” notes Dr. Orange. “Some of the genetic risk factors for rheumatoid arthritis are expressed by B cells; patients have autoantibodies which are made by B cells; and B cell depletion therapy is approved for rheumatoid arthritis.”
Moving Forward
It is not clear how generalizable these results are, based on the small number of patients in this study. Dr. Orange et al. are currently working on another analysis with a larger patient set, which will examine whether background therapy affects this pattern of immune activation.
If the results are generalizable, it might be possible to intervene prior to flare onset, if patients were regularly monitoring their blood for detectable changes. This might not be practical in the end, but the insights into pathophysiology and immune responses could have long-term impacts on therapy development. Dr. Orange also speculates, “It does also suggest that if these cell types are activated just before a flare—they are sort of a smoking gun. Maybe if you targeted these PRIME cells, maybe we could prevent flares.”
The broad use of such an approach—using longitudinal genomics to identify key precursors of disease flares—may be applicable for multiple types of rheumatic disease. Dr. Orange points out that such an approach requires an outcome measure, similar to the RAPID3, which can reliably measure disease activity from home. “I think this approach could be used for a lot of chronic inflammatory diseases that wax and wane over time,” she adds.
One of the authors of the accompanying editorial is the ACR’s immediate past president, Ellen M. Gravallese, MD, chief of the Division of Rheumatology, Inflammation and Immunity at Brigham and Women’s Hospital and the Theodore Bevier Bayles Professor of Medicine at Harvard Medical School, Boston.2 She concludes, “This study is important from both the technical standpoint, utilizing serial blood samples from patients in their homes to perform detailed molecular and cellular analyses, and from the standpoint of the information gained about the mechanisms of rheumatoid arthritis flares.”
Ruth Jessen Hickman, MD, is a graduate of the Indiana University School of Medicine. She is a freelance medical and science writer living in Bloomington, Ind.
References
- Orange DE, Yao V, Sawicka K, et al. RNA identification of PRIME cells predicting rheumatoid arthritis flares. N Engl J Med. 2020 Jul 16;383(3):218¬–228.
- Gravallese EM, Robinson WH. PRIME time in rheumatoid arthritis. N Engl J Med. 2020 Jul 16;383(3):278–279.
- Zhang F, Wei K, Slowikowski K, et al. Defining inflammatory cell states in rheumatoid arthritis joint synovial tissues by integrating single-cell transcriptomics and mass cytometry. Nat Immunol. 2019 Jul;20(7):928–942.
- Mizoguchi F, Slowikowski K, Wei K, et al. Functionally distinct disease-associated fibroblast subsets in rheumatoid arthritis. Nat Commun. 2018 Feb 23;9(1):789.
- Croft AP, Campos J, Jansen K, et al. Distinct fibroblast subsets drive inflammation and damage in arthritis. Nature. 2019 Jun;570(7760):246–251.
