Edited by: Masaaki Murakami, Osaka University, Japan
Reviewed by: Sandra Sacre, University of Sussex, UK; Daisuke Kamimura, Osaka University, Japan
*Correspondence: Mayumi Fujita and Carl K. Edwards III, University of Colorado Denver, Anschutz Medical Campus, 12801 East 17th Avenue, RC-1 South, Aurora, CO 80045, USA. e-mail:
This article was submitted to Frontiers in Inflammation, a specialty of Frontiers in Immunology.
†These authors equally contributed to this work.
This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in other forums, provided the original authors and source are credited and subject to any copyright notices concerning any third-party graphics etc.
Periodic assessment of gene expression for diagnosis and monitoring in rheumatoid arthritis (RA) may provide a readily available and useful method to detect subclinical disease progression and follow responses to therapy with disease modifying anti-rheumatic agents (DMARDs) or anti-TNF-α therapy. We used quantitative real-time PCR to compare peripheral blood gene expression profiles in active (“unstable”) RA patients on DMARDs, stable RA patients on DMARDs, and stable RA patients treated with a combination of a disease-modifying anti-rheumatoid drug (DMARD) and an anti-TNF-α agent (infliximab or etanercept) to healthy human controls. The expression of 48 inflammatory genes were compared between healthy controls (
Rheumatoid arthritis (RA) is a chronic, progressive, autoimmune, inflammatory disorder that affects approximately 1% of the population in the United States (Lee and Weinblatt,
RA is characterized by chronic inflammation and hypertrophy of the synovial membranes. Inflammation of the joint occurs in response to production of growth factors, cytokines, and chemokines by many different cell types present in synovium and cartilage, in addition to infiltrating cells from the peripheral blood. Cartilage and bone destruction subsequently occur through the enhanced actions of prostaglandins, leukotrienes, and the matrix degrading metalloproteinases (MMPs). The importance of interleukin-1 (IL-1) and tumor necrosis factor-α (TNF-α) in animal models of RA is well documented (Saklatvala,
Although the most important actions of these proteins are likely to occur in the joint, the joint space is relatively inaccessible, prohibiting quantitative measurement of cytokines. Recent advances in the clinical application of pharmacogenomics suggest that biomarkers of disease activity and drug efficacy can be identified in blood to enable the identification of specific patient populations and to monitor subclinical changes in disease status and responses to treatment over time (Frank and Hargreaves,
We have previously shown that among healthy volunteers, expression of a number of inflammatory genes can be accurately measured by quantitative reverse transcriptase PCR (qRT-PCR) from peripheral blood samples collected over time, and that this expression is relatively stable (McLoughlin et al.,
Samples from blood donor subjects were collected for the current study with approval of the University of Colorado Institutional Review Board and after obtaining written consent from each volunteer. Whole blood samples were collected at a single time point from 122 apparently healthy blood donors at a local blood bank (Bonfils Blood Center, Denver, CO). Enrollment criteria for blood donors followed the American Red Cross donor standards. Subject age was normally distributed and ranged from 22 to 82 years, with an average age of 47.2 ± 13.1 years. Females (
Using the criteria described above, there were three experimental patient subpopulation groups: (1) RA patients who were on a systemic disease-modifying anti-rheumatoid drug (DMARD, either oral prednisone or oral methotrexate), with active disease and judged by their physician to require a change in therapy (“unstable”) (
Gender (% female) | 79 | 61 | 85 |
Age (years) | 54.1 | 55.8 | 57.7 |
Race (% white, non-Hispanic) | 67 | 50 | 90 |
Duration of disease (months) | 220 | 200 | 227 |
DMARD use (%) | 75 | 100 | 100 |
Prednisone use (%) | 64 | 75 | 25 |
MTX weekly dose (mg) | 15.5 | 9.95 | 14.5 |
Blood was collected from study subjects by standard phlebotomy methods into PAXgene™ tubes (PreAnalytiX, Valencia, CA) to stabilize mRNA levels. Samples were frozen at −70°C and shipped on dry ice in compliance with International Air Transport Association (IATA) shipping regulations. Total RNA was extracted as described previously using the PAXgene™ Blood RNA System (Rainen et al.,
Target gene products were analyzed by quantitative PCR of each cDNA preparation using 2X TaqMan® Universal PCR Master Mix (Applied Biosystems, #4305719, Foster City, CA) and Source Precision Medicine's proprietary primer/probe sets and adhering to previously described protocols (McLoughlin et al.,
B7 | CD80 | Regulatory protein that may be associated with lupus | 21.17 |
C1QA | Complement component 1, q subcomponent, A chain | Serum complement system component, forms C1 complex with pro-enzymes C1r and C1s | 21.64 |
CD3Z | CD247 molecule | T-cell surface glycoprotein | 15.20 |
CD4 | CD4 molecule | Helper T-cell marker; accessory protein in the MHC/T-cell receptor interaction | 15.84 |
CD8A | CD8a molecule | Cytotoxic/suppressor T-cell marker; binds MHC I; thought to play role T-cell mediated killing | 16.40 |
CD14 | CD14 molecule | Monocyte LPS receptor; cooperates with MD-2 and TLR-4 in response to LPS | 14.84 |
CD19 | CD19 molecule | B-cell growth factor; membrane protein that potentiates receptor-dependent activation | 18.65 |
CXCL1 | Chemokine (C-X-C motif) ligand 1 | Chemotactic for neutrophils; pro-inflammatory; modulates metalloproteinase activity | 19.78 |
CXCL2 | Chemokine (C-X-C motif) ligand 2 | Chemotactic for neutrophils and hematopoietic precursor cells | 23.32 |
CSF2 | Colony stimulating factor 2 | 23.59 | |
CSF3 | Colony stimulating factor 3 | 23.45 | |
F3 | Coagulation factor 3 (thromboplastin, tissue factor) | 23.63 | |
HLA-DRB1 | Major histocompatibility complex, class II, DR beta 1 | Binds antigen for presentation to CD4+ cells | 22.09 |
HMOX1 | Heme oxygenase (decycling) 1 | Essential for heme catabolism; cleaves heme to form biliverdin and CO; endotoxin inducible | 17.10 |
HSPA1A | Heat shock protein A1A | Molecular chaperone; stabilizes AU rich mRNA; hydrophobic peptide | 14.76 |
ICAM1 | Intercellular adhesion molecule 1 | Endothelial cell surface molecule; regulates cell adhesion and trafficking of leukocytes | 18.05 |
IFNA2 | Interferon, alpha 2 | Interferon produced by macrophages with antiviral effects | 22.20 |
IFNG | Interferon gamma | Pro- and anti-inflammatory activity, TH1 cytokine, inflammatory mediator of activated T-cells | 22.98 |
IL1A | Interleukin-1, alpha | Pro-inflammatory; generally cytosolic, released during severe inflammatory disease | 23.14 |
IL1B | Interleukin-1, beta | Pro-inflammatory; made by activated macrophages; endogenous pyrogen | 16.19 |
IL1RN | Interleukin-1 receptor antagonist | Anti-inflammatory; inhibits binding of IL-1 its receptor without stimulating IL-1 activity | 16.41 |
IL2 | Interleukin-2 | T-cell growth factor, expressed by activated T-cells, TH1 cytokine | 23.11 |
IL4 | Interleukin-4 | Anti-inflammatory; TH2; suppresses cytokines, increases IL-1RN expression | 23.25 |
IL5 | Interleukin-5 | Stimulates eosinophil expansion and B-cell differentiation | 22.70 |
IL6 | Interleukin-6 | Pro- and anti-inflammatory activity; TH2 cytokine; regulates hematopoiesis | 23.11 |
IL8 | Interleukin-8 | Pro-inflammatory CXC chemokine; major secondary inflammatory mediator | 21.04 |
IL10 | Interleukin-10 | Anti-inflammatory; TH2 cytokine; suppresses production of pro-inflammatory cytokines | 22.84 |
IL12B | Interleukin-12b | Pro-inflammatory, TH1 cytokine, requires co-stimulation with IL-18 to induce IFN-γ | 23.28 |
IL13 | Interleukin-13 | Inhibits inflammatory cytokine production | 23.09 |
IL15 | Interleukin-15 | Pro-inflammatory; T-cell activator; inhibits apoptosis; with IL-2 induces IFN-γ and TNF-α | 22.05 |
IL18 | Interleukin-18 | Pro-inflammatory; TH1 cytokine; promotes apoptosis; induces IFNγ ; blocked by IL18BP | 20.12 |
IL18BP | Interleukin-18 binding protein | Binds and inactivates IL18; implicated in inhibition of early TH1 cytokine responses | 17.59 |
MMP3 | Matrix metallopeptidase 3 | 23.71 | |
MMP9 | Matrix metallopeptidase 9 | Degrades extracellular matrix molecules; made by neutrophils; Role in arthritis and metastasis | 16.39 |
PLA2G7 | Phospholipase A2, group VII | Activates platelet activating factor (PF4) | 19.93 |
NOS2A | Nitric oxide synthase 2a, inducible | 23.53 | |
PLAUR | Plasminogen activator, urokinase receptor | Ligand-specific cell surface receptor for UPA; localizes and promotes plasmin formation | 15.59 |
PTGS2 | Prostaglandin-endoperoxide synthase 2 | Pro-inflammatory; regulates angiogenesis and cell migration | 17.92 |
PTPRC | Protein tyrosine phosphatase, receptor type, C | An essential regulator of T- and B-cell antigen receptor signaling; suppresses JAK kinases | 11.89 |
SERPINE1 | Serpin peptidase inhibitor, clade E | Interacts with tissue plasminogen activator to regulate fibrinolysis; inhibits PLAU | 22.95 |
TGFB1 | Transforming growth factor, beta 1 | Pro- and anti-inflammatory activity; anti-apoptotic | 13.55 |
TIMP1 | TIMP metallopeptidase inhibitor 1 | Inhibitors of matrix metalloproteinases; transcriptionally induced by cytokines and hormones | 15.11 |
TNF | Tumor necrosis factor | Pro-inflammatory TH1 cytokine, primary mediator of immune response and regulation | 20.65 |
TNFSF5 | CD40 ligand | Ligand for CD40; expressed on T-cells; regulates B-cell function by engaging CD40 | 17.61 |
TNFSF6 | Fas ligand (TNF superfamily, member 6) | Ligand for FAS antigen; critical in triggering apoptosis | 21.00 |
TNFSF13B | Tumor necrosis factor (ligand) superfamily, member 13b | B cell activating factor, TNF family | 15.46 |
TNFRSF13B | Tumor necrosis factor receptor superfamily, member 13b | Controls T-cell-dependent B-cell antibody responses | 20.81 |
VEGF | Vascular endothelial growth factor | Produced by monocytes | 23.07 |
Each PCR reaction contained primer/probe sets for the target gene and 18S RNA, used as the internal control. The difference between the fluorescence CT for the target and the internal endogenous control (18S) is presented as a ΔCT value. Increases or decreases in the target mRNA concentration correspond to lower or higher ΔCT values, respectively, at approximately 2-fold concentration change per ΔCT unit. The CT reporting system and estimation of relative gene expression is well described in the literature (Livak and Schmittgen,
Gene expression levels were measured in whole blood samples collected from 18 unstable patients with RA maintained on DMARD therapy. Of a total of 48 gene products analyzed, 13 demonstrated very low or undetectable levels among the study subjects. These genes (CSF2, CSF3, CXCL2, F3, IL1A, IL2, IL4, IL6, IL12B, IL13, MMP3, NOS2A, and PLAUR) were not included in further analysis. Of 35 inflammation-related genes examined, these unstable RA patients exhibited increased expression of 25 genes (B7, C1QA, CD14, CD19, CD4, CD8A, CXCL1, HMOX1, HSPA1A, ICAM1, IL10, IL15, IL18, IL18BP, IL1RN, IL1B, MMP9, PTGS2, PTPRC, SERPINE1, TGFB1, TIMP1, TNF, TNFSF13B, TNFSF6, and VEGF) and decreased expression of 1 gene (CD19) compared to healthy controls (
B7 | 1.24 | 1.07 | 0.75 | 0.5217 | ||
C1QA | 2.42 | 1.10 | 1.52 | < |
0.5567 | |
CD14 | 2.25 | 1.47 | 1.10 | < |
0.3615 | |
CD19 | 0.64 | 0.54 | 0.68 | < |
||
CD3Z | 1.01 | 0.81 | 0.89 | 0.9495 | 0.1683 | |
CD4 | 1.47 | 1.13 | 1.04 | 0.2490 | < |
|
CD8A | 1.54 | 0.87 | 0.81 | 0.3299 | 0.1707 | |
CXCL1 | 1.89 | 1.43 | 0.75 | < |
||
HLA-DRB1 | 0.69 | 0.32 | 0.92 | 0.4900 | 0.9067 | |
HMOX1 | 2.25 | 1.49 | 1.30 | < |
||
HSPA1A | 2.62 | 1.77 | 1.15 | < |
< |
0.2548 |
ICAM1 | 2.25 | 1.44 | 1.07 | < |
0.5007 | |
IFNA2 | 1.25 | 1.06 | 0.62 | 0.1295 | 0.7230 | |
IFNG | 1.24 | 0.84 | ND | 0.0614 | 0.0795 | |
IL10 | 1.34 | 0.93 | 0.61 | 0.5005 | < |
|
IL15 | 1.43 | 0.94 | 1.33 | 0.6242 | ||
IL18 | 2.14 | 1.41 | 1.09 | < |
0.4815 | |
IL18BP | 1.60 | 1.18 | 0.78 | < |
0.0644 | |
IL1RN | 2.48 | 1.84 | 0.68 | < |
< |
|
IL1B | 2.24 | 1.69 | 0.93 | < |
< |
0.5218 |
IL5 | 1.26 | 0.97 | 0.61 | 0.0857 | 0.8539 | < |
IL8 | 1.01 | 1.06 | 2.69 | 0.9721 | 0.7469 | < |
MMP9 | 3.45 | 2.04 | 1.19 | < |
0.3469 | |
PLA2G7 | 1.14 | 0.92 | 0.92 | 0.4591 | 0.5298 | 0.6262 |
PTGS2 | 2.23 | 1.26 | 0.78 | < |
||
PTPRC | 1.49 | 1.25 | 0.67 | < |
0.6788 | |
SERPINE1 | 2.56 | 1.21 | 1.12 | < |
0.1680 | 0.4726 |
TGFB1 | 2.52 | 1.59 | 0.96 | < |
< |
0.5757 |
TIMP1 | 2.02 | 1.35 | 1.08 | < |
0.4304 | |
TNF | 1.90 | 1.11 | 0.88 | < |
0.3426 | 0.3158 |
TNFRSF13B | 0.80 | 0.61 | 0.60 | 0.1674 | ||
TNFSF13B | 1.97 | 1.35 | 1.47 | < |
||
TNFSF5 | 1.17 | 0.97 | 0.62 | 0.1691 | 0.7831 | < |
TNFSF6 | 1.70 | 0.79 | 0.93 | 0.0839 | 0.6686 | |
VEGF | 1.38 | 0.84 | ND |
Whole blood was collected from 26 RA patients with stable clinical examinations after at least 12 weeks of DMARD therapy. Serum levels of CRP in these patients ranged from 0.05 to 2.7 mg/L (average 0.35 mg/L), consistent with low systemic levels of inflammation. Of 35 inflammation-related genes examined, stable RA patients maintained on DMARD therapy exhibited increased expression of 14 genes (CD14, CXCL1, HMOX1, HSPA1A, ICAM1, IL18, IL1RN, IL1B, MMP9, PTGS2, PTPRC, TGFB1, TIMP1, and TNFSF13B) and decreased expression of 5 genes (CD19, CD3Z, HLA-DRB1, TNFRSF13B, and VEGF) compared to healthy controls (
Whole blood was collected from 20 RA patients with stable clinical examinations after 12 weeks of treatment with combination therapy (a DMARD and an anti-TNF-α agent). Of 35 inflammation-related genes examined, 6 genes (C1QA, CD4, HMOX1, IL15, IL8, and TNFSF13B) exhibited increased expression and 11 genes (B7, CD19, CXCL1, IFNA2, IL10, IL18BP, IL1RN, IL5, PTGS2, TNFRSF13B, and TNFSF5) demonstrated decreased expression compared to healthy controls (
Among 20 patients receiving combination therapy, patients receiving infliximab (
B7 | 0.67 | 0.92 | 0.4999 | |
C1QA | 2.02 | 0.99 | 0.9696 | |
CD14 | 1.14 | 1.04 | 0.3689 | 0.7978 |
CD3Z | 0.89 | 0.94 | 0.2948 | 0.5865 |
CD4 | 1.03 | 1.06 | 0.8423 | 0.6788 |
CD8A | 1.04 | 0.58 | 0.8418 | |
CXCL1 | 0.68 | 0.79 | 0.1727 | |
HMOX1 | 1.26 | 1.40 | 0.1278 | |
HSPA1A | 1.15 | 1.09 | 0.3865 | 0.5708 |
ICAM1 | 0.98 | 1.13 | 0.8911 | 0.3179 |
IL10 | 0.68 | 0.64 | ||
IL15 | 1.43 | 1.20 | 0.0555 | 0.3005 |
IL18 | 0.90 | 1.25 | 0.4656 | 0.0925 |
IL18BP | 0.82 | 0.73 | 0.0526 | |
IL1RN | 0.61 | 0.73 | ||
IL1B | 0.78 | 1.07 | 0.078 | 0.6039 |
IL5 | 0.66 | 0.46 | < |
|
IL8 | 1.50 | 5.70 | 0.1585 | < |
MMP9 | 1.20 | 1.06 | 0.461 | 0.7911 |
PLA2G7 | 0.86 | 1.07 | 0.4403 | 0.7478 |
PTGS2 | 0.82 | 0.71 | 0.1617 | |
PTPRC | 0.75 | 0.63 | < |
|
SERPINE1 | 1.21 | 0.94 | 0.286 | 0.7372 |
TGFB1 | 1.02 | 0.88 | 0.8391 | 0.1836 |
TIMP1 | 1.02 | 1.10 | 0.8459 | 0.4071 |
TNF | 0.83 | 0.81 | 0.2149 | 0.1582 |
TNFRSF13B | 0.76 | 0.51 | 0.1389 | |
TNFSF13B | 1.37 | 1.49 | 0.0592 | |
TNFSF5 | 0.59 | 0.62 | ||
TNFSF6 | 1.09 | 0.75 | 0.6164 | 0.102 |
Identification of easily accessible and reliable biomarkers of inflammatory disease activity to diagnose and monitor disease progression in individual patients over time is an attractive therapeutic goal. Here, we demonstrate that gene expression analysis of whole blood using qRT-PCR can be used to assess disease activity of RA patients.
Patients with unstable RA demonstrated increased peripheral blood expression of numerous proinflammatory cytokines, including IL1B, TNF, and IL18, and increased expression of genes whose protein products have been shown to contribute to synovial deterioration (including MMP) compared to healthy human control subjects. These cytokines have been previously shown to be upregulated in the synovium or serum in RA (Brennan et al.,
Unstable RA patients in the current study also exhibited increased expression of several cytokines which are generally thought to have an anti-inflammatory effect, including IL10, IL1RN, and TGFB1 (Katsikis et al.,
Our data, however, also demonstrate that despite stable clinical assessments, RA patients on DMARD therapy alone continue to exhibit increased expression of some inflammatory genes, suggesting that subclinical inflammation is still present in these patients. Stable RA patients on combination therapy demonstrate a greater reduction in inflammatory gene expression compared to both unstable and stable RA patients on DMARD alone. Specifically, patients on combination therapy exhibit fewer increases in proinflammatory markers compared to patients on DMARD alone, and decreased expression of a number of proinflammatory genes even compared to healthy controls. Our findings suggest that even when chronic inflammatory disease is clinically stable, inflammatory pathways may still be active. These pathways could contribute to long-term development of comorbidities associated with RA. Further, variations in the inflammatory profile differ in patients on DMARD alone versus combined DMARD and anti-TNF-α therapy, and combination therapy appears to confer a greater overall anti-inflammatory effect compared to DMARD therapy. Whether changes in the expression of particular genes may be useful biomarkers to assess overall disease activity remains to be determined, and is an area of ongoing research.
Changes in gene expression with anti-TNF-α therapy were largely similar between patients receiving infliximab and etanercept, with the notable exception of significantly increased expression of the chemokine IL8 in patients on etanercept compared to healthy controls and to RA patients on DMARD or infliximab therapies (Table
Despite variable disease activity and treatment modalities, the expression of several proinflammatory genes was increased among all three experimental groups compared to healthy controls, including HMOX1 and TNFSF13B (Table
Assessment of peripheral blood gene expression provides an easily accessible population of inflammatory cells in which to study relative changes in cytokine expression over time. However, it is important to consider that in any given individual, relative proportions of each blood cell type may vary markedly. In this setting, overall changes in peripheral blood gene expression may be significantly influenced by changes in the proportion of blood cell types and their corresponding transcription profiles. For example, we observed upregulation of CD14, a monocyte-specific marker, in unstable RA patients treated with DMARD (Table
IL-6 is a proinflammatory cytokine that stimulates differentiation of B-cells into antibody-producing plasma cells and contributes to the release of metalloproteases from tissue fibroblasts (Badolato and Oppenheim,
Herein, we show that reduction of inflammatory gene expression levels differ between stable patients on DMARD and combined DMARD and anti-TNF-α therapies. Since most genes in healthy subjects exhibit limited dynamic ranges of expression, we believe that in our study populations, differences in gene expression levels may reflect disease activity. The current findings suggest that these peripheral blood biomarkers correlate with clinical status, and may therefore provide adjunctive information about the efficacy of various treatments for RA. These biomarkers may also allow for assessment of the efficacy of particular RA treatments, including anti-TNF-α therapies. Future studies to determine peripheral blood biomarkers that may predict individual patient responses to a particular systemic therapy may provide assistance with clinical decision making.
The authors have read the journal's policy and have the following conflicts. John Cheronis, David Trollinger, Danute Bankaitis-Davis, and Michael Bevilacqua are employees of Source MDx. Source MDx only helped to carry out high-throughput qRT-PCR analysis of whole blood samples. The company was not involved in sample collection and data interpretation, decision to publish, or preparation of the manuscript. This does not alter the authors' adherence to all the
The authors have no special acknowledgements to make. Brian Kotzin is currently an employee of Amgen, Inc. but did not take up this position during the completion of the study.