Gene expression profiling in neutrophils from children with polyarticular juvenile idiopathic arthritis

The emerging paradigm in pediatric rheumatology: harnessing the power of artificial intelligence

Artificial intelligence algorithms, with roots extending into the past but experiencing a resurgence and evolution in recent years due to their superiority over traditional methods and contributions to human capabilities, have begun to make their presence felt in the field of pediatric rheumatology. In the ever-evolving realm of pediatric rheumatology, there have been incremental advancements supported by artificial intelligence in understanding and stratifying diseases, developing biomarkers, refining visual analyses, and facilitating individualized treatment approaches. However, like in many other domains, these strides have yet to gain clinical applicability and validation, and ethical issues remain unresolved. Furthermore, mastering different and novel terminologies appears challenging for clinicians. This review aims to provide a comprehensive overview of the current literature, categorizing algorithms and their applications, thus offering a fresh perspective on the nascent relationship between pediatric rheumatology and artificial intelligence, highlighting both its advancements and constraints.

Expanded population of low-density neutrophils in juvenile idiopathic arthritis

Introduction

Juvenile idiopathic arthritis (JIA), a clinically variable disease characterized by autoimmune arthritis, affects children, and its immunopathology remains elusive. Alterations in neutrophil biology play an important role in this disease. In the present study, we aimed to explore the features of low-density neutrophils (LDNs) in patients with JIA.

Methods

Gene expression of peripheral blood mononuclear cells (PBMCs) from children with distinct subtypes of JIA was analyzed by NanoString Immunology panel. Presence of LDNs was ascertained by flow cytometry and the release of neutrophil-associated products were analyzed by LUMINEX.

Results

LDNs were detected in patients' peripheral blood mononuclear cells (PBMCs) after density gradient centrifugation. Transcriptomic analysis of JIA PBMCs revealed that genes related to neutrophil degranulation were markedly upregulated. The number of LDNs and level of their degranulation products increased in patients' PBMCs and correlated with serum calprotectin, but not with disease activity, sedimentation rate and C-reactive protein (CRP) levels. The phenotypes of LDNs varied from those of normal-density neutrophils and healthy donor LDNs. Phenotypical analysis revealed LDNs are immature and primed population with decreased suppressive capacity. A negative correlation between surface proteins CD62L, CD66b, and CD11b and the number of inflamed joints/JADAS was established.

Conclusion

Our results describe LDNs as primed, degranulated, immature cells with impaired suppressive activities. This work thus contributes to the increasing body of evidence that LDNs in JIA are altered and their role in the disease immunopathogenesis and possible clinical associations should be investigated further.

Analysis of chromatin data supports a role for CD14+ monocytes/macrophages in mediating genetic risk for juvenile idiopathic arthritis

Introduction

Genome wide association studies (GWAS) have identified multiple regions that confer genetic risk for the polyarticular/oligoarticular forms of juvenile idiopathic arthritis (JIA). However, genome-wide scans do not identify the cells impacted by genetic polymorphisms on the risk haplotypes or the genes impacted by those variants. We have shown that genetic variants driving JIA risk are likely to affect both innate and adaptive immune functions. We provide additional evidence that JIA risk variants impact innate immunity.

Materials and methods

We queried publicly available H3K4me1/H3K27ac ChIP-seq data in CD14+ monocytes to determine whether the linkage disequilibrium (LD) blocks incorporating the SNPs that tag JIA risk loci showed enrichment for these epigenetic marks. We also queried monocyte/macrophage GROseq data, a functional readout of active enhancers. We defined the topologically associated domains (TADs) encompassing enhancers on the risk haplotypes and identified genes within those TADs expressed in monocytes. We performed ontology analyses of these genes to identify cellular processes that may be impacted by these variants. We also used whole blood RNAseq data from the Genotype-Tissue Expression (GTEx) data base to determine whether SNPs lying within monocyte GROseq peaks influence plausible target genes within the TADs encompassing the JIA risk haplotypes.

Results

The LD blocks encompassing the JIA genetic risk regions were enriched for H3K4me1/H3K27ac ChIPseq peaks (p=0.00021 and p=0.022) when compared to genome background. Eleven and sixteen JIA were enriched for resting and activated macrophage GROseq peaks, respectively risk regions (p=0.04385 and p=0.00004). We identified 321 expressed genes within the TADs encompassing the JIA haplotypes in human monocytes. Ontological analysis of these genes showed enrichment for multiple immune functions. Finally, we found that SNPs lying within the GROseq peaks are strongly associated with expression levels of plausible target genes in human whole blood.

Conclusions

These findings support the idea that both innate and adaptive immunity are impacted by JIA genetic risk variants.

Distinct transcription factor networks control neutrophil-driven inflammation

Neutrophils display distinct gene expression patters depending on their developmental stage, activation state and tissue microenvironment. To determine the transcription factor networks that shape these responses in a mouse model, we integrated transcriptional and chromatin analyses of neutrophils during acute inflammation. We showed active chromatin remodeling at two transition stages: bone marrow-to-blood and blood-to-tissue. Analysis of differentially accessible regions revealed distinct sets of putative transcription factors associated with control of neutrophil inflammatory responses. Using ex vivo and in vivo approaches, we confirmed that RUNX1 and KLF6 modulate neutrophil maturation, whereas RELB, IRF5 and JUNB drive neutrophil effector responses and RFX2 and RELB promote survival. Interfering with neutrophil activation by targeting one of these factors, JUNB, reduced pathological inflammation in a mouse model of myocardial infarction. Therefore, our study represents a blueprint for transcriptional control of neutrophil responses in acute inflammation and opens possibilities for stage-specific therapeutic modulation of neutrophil function in disease.

Synovial fluid neutrophils in oligoarticular juvenile idiopathic arthritis have an altered phenotype and impaired effector functions

Background

Neutrophils are the most prevalent immune cells in the synovial fluid in inflamed joints of children with oligoarticular juvenile idiopathic arthritis (JIA). Despite this, little is known about neutrophil function at the site of inflammation in JIA and how local neutrophils contribute to disease pathogenesis. This study aimed to characterize the phenotype and function of synovial fluid neutrophils in oligoarticular JIA.

Methods

Neutrophils obtained from paired blood and synovial fluid from patients with active oligoarticular JIA were investigated phenotypically (n = 17) and functionally (phagocytosis and oxidative burst, n = 13) by flow cytometry. In a subset of patients (n = 6), blood samples were also obtained during inactive disease at a follow-up visit. The presence of CD206-expressing neutrophils was investigated in synovial biopsies from four patients by immunofluorescence.

Results

Neutrophils in synovial fluid had an activated phenotype, characterized by increased CD66b and CD11b levels, and most neutrophils had a CD16^hi CD62L^lowaged phenotype. A large proportion of the synovial fluid neutrophils expressed CD206, a mannose receptor not commonly expressed by neutrophils but by monocytes, macrophages, and dendritic cells. CD206-expressing neutrophils were also found in synovial tissue biopsies. The synovial fluid neutrophil phenotype was not dependent on transmigration alone. Functionally, synovial fluid neutrophils had reduced phagocytic capacity and a trend towards impaired oxidative burst compared to blood neutrophils. In addition, the effector functions of the synovial fluid neutrophils correlated negatively with the proportion of CD206⁺ neutrophils.

Conclusions

Neutrophils in the inflamed joint in oligoarticular JIA were altered, both regarding phenotype and function. Neutrophils in the synovial fluid were activated, had an aged phenotype, had gained monocyte-like features, and had impaired phagocytic capacity. The impairment in phagocytosis and oxidative burst was associated with the phenotype shift. We speculate that these neutrophil alterations might play a role in the sustained joint inflammation seen in JIA.

Synovial Fluid Neutrophils From Patients With Juvenile Idiopathic Arthritis Display a Hyperactivated Phenotype

OBJECTIVE

Juvenile idiopathic arthritis (JIA) is the most common rheumatic disease in childhood. The predominant subtypes, oligoarticular and polyarticular JIA, are traditionally considered to be autoimmune diseases with a central role for T cells and autoantibodies. Mounting evidence suggests an important role for neutrophils in JIA pathogenesis. We undertook this study to investigate the phenotypic features of neutrophils present in the blood and inflamed joints of patients.

METHODS

JIA synovial fluid (SF) and parallel blood samples from JIA patients and healthy children were collected. SF-treated neutrophils from healthy donors and pleural neutrophils from patients with pleural effusion were investigated as controls for SF exposure and extravasation. Multicolor flow cytometry panels allowed for in-depth phenotypic analysis of neutrophils, focusing on the expression of adhesion molecules, activation, and maturation markers and chemoattractant receptors. Multiplex technology was used to quantify cytokines in plasma and SF.

RESULTS

SF neutrophils displayed an activated, hypersegmented phenotype with decreased CD62L expression, up-regulation of adhesion molecules CD66b, CD11b, and CD15, and down-regulation of CXCR1/2. An elevated percentage of CXCR4-positive neutrophils was detected in SF from patients. Pleural neutrophils showed less pronounced maturation differences. Strikingly, significant percentages of SF neutrophils showed a profound up-regulation of atypical neutrophil markers, including CXCR3, intercellular adhesion molecule 1, and HLA-DR.

CONCLUSION

Our data show that neutrophils in inflamed joints of JIA patients have an activated phenotype. This detailed molecular analysis supports the notion that a complex intertwining between these innate immune cells and adaptive immune events drives JIA.

Transcriptional regulation of neutrophil differentiation and function during inflammation

Neutrophils are the most abundant leukocytes in innate immunity where they elicit powerful effector functions to eliminate invading pathogens and modulate the adaptive as well as the innate immune response. Neutrophil function must be tightly regulated during inflammation and infection to avoid additional tissue damage. Increasing evidence suggests that transcription factors (TFs) function as key regulators to modulate transcriptional output, thereby controlling cell fate decision and the inflammatory responses. However, the molecular mechanisms underlying neutrophil differentiation and function during inflammation remain largely uncharacterized. Here, we provide a comprehensive overview of TFs known to be crucial for neutrophil maturation and in the signaling pathways that control neutrophil differentiation and activation. We also outline how emerging genomic and single-cell technologies may facilitate further discovery of neutrophil transcriptional regulators.

Adjunctive vitamin D for the treatment of active juvenile idiopathic arthritis: An open-label, prospective, randomized controlled trial

Vitamin D has an important immunomodulatory effect, but no trial has examined the effect of boosting serum levels of 25-hydroxyvitamin D (25OHD) in individuals with juvenile idiopathic arthritis (JIA). The aim of the present study was to assess whether vitamin D supplementation reduced disease activity and adjusted/maintained bone mass in patients with active JIA. A 24-week randomized trial was undertaken at Children's Hospital of Chongqing Medical University. Treatment-naive patients with JIA were randomly assigned (1:1) to one of two groups: Standard treatment with high dose oral cholecalciferol [2,000 IU per day; experimental group (EG)] or without supplementation [control group (CG)]. The primary outcomes were the 27-joint Juvenile Arthritis Disease Activity Score (JADAS-27 score), the Z-score for bone mineral density (BMD), and serum levels of 25OHD. A per-protocol analysis set approach was used. The Mann-Whitney U test was the main tool used for data analysis. A total of 42 participants were assigned randomly to the EG (n=20) or the CG (n=22); of these, 36 (n=18 and n=18, respectively) were included in per-protocol analysis. After 24 weeks, the mean level of 25OHD in the EG was higher than that in the CG (P<0.05). At the end of the intervention, there were no clear differences between the two groups in terms of BMD or JADAS-27 score (both P>0.05). Cholecalciferol supplementation (2000 IU/day) for 24 weeks raised serum levels of 25OHD in JIA patients but did not reduce disease activity or improve BMD (registration no. ChiCTR-INR-16009235; Date of Registration: 2016-10-12).

The feasibility of developing biomarkers from peripheral blood mononuclear cell RNAseq data in children with juvenile idiopathic arthritis using machine learning approaches

Background

The response to treatment for juvenile idiopathic arthritis (JIA) can be staged using clinical features. However, objective laboratory biomarkers of remission are still lacking. In this study, we used machine learning to predict JIA activity from transcriptomes from peripheral blood mononuclear cells (PBMCs). We included samples from children with Native American ancestry to determine whether the model maintained validity in an ethnically heterogeneous population.

Methods

Our dataset consisted of 50 samples, 23 from children in remission and 27 from children with an active disease on therapy. Nine of these samples were from children with mixed European/Native American ancestry. We used 4 different machine learning methods to create predictive models in 2 populations: the whole dataset and then the samples from children with exclusively European ancestry.

Results

In both populations, models were able to predict JIA status well, with training accuracies > 74% and testing accuracies > 78%. Performance was better in the whole dataset model. We note a high degree of overlap between genes identified in both populations. Using ingenuity pathway analysis, genes from the whole dataset associated with cell-to-cell signaling and interactions, cell morphology, organismal injury and abnormalities, and protein synthesis.

Conclusions

This study demonstrates it is feasible to use machine learning in conjunction with RNA sequencing of PBMCs to predict JIA stage. Thus, developing objective biomarkers from easy to obtain clinical samples remains an achievable goal.

Immunomics in Pediatric Rheumatic Diseases

The inherent complexity in the immune landscape of pediatric rheumatic disease necessitates a holistic system approach. Uncertainty in the mechanistic workings and etiological driving forces presents difficulty in personalized treatments. The development and progression of immunomics are well suited to deal with this complexity. Immunomics encompasses a spectrum of biological processes that entail genomics, transcriptomics, epigenomics, proteomics, and cytomics. In this review, we will discuss how various high dimensional technologies in immunomics have helped to grow a wealth of data that provide salient clues and biological insights into the pathogenesis of autoimmunity. Interfaced with critical unresolved clinical questions and unmet medical needs, these platforms have helped to identify candidate immune targets, refine patient stratification, and understand treatment response or resistance. Yet the unprecedented growth in data has presented both opportunities and challenges. Researchers are now facing huge heterogeneous data sets from different origins that need to be integrated and exploited for further data mining. We believe that the utilization and integration of these platforms will help unravel the complexities and expedite both discovery and validation of clinical targets.

Neutrophils From Children With Systemic Juvenile Idiopathic Arthritis Exhibit Persistent Proinflammatory Activation Despite Long-Standing Clinically Inactive Disease

Background: Systemic juvenile idiopathic arthritis (SJIA) is a chronic childhood arthropathy with features of autoinflammation. Early inflammatory SJIA is associated with expansion and activation of neutrophils with a sepsis-like phenotype, but neutrophil phenotypes present in longstanding and clinically inactive disease (CID) are unknown. The objective of this study was to examine activated neutrophil subsets, S100 alarmin release, and gene expression signatures in children with a spectrum of SJIA disease activity.

Methods: Highly-purified neutrophils were isolated using a two-step procedure of density-gradient centrifugation followed by magnetic-bead based negative selection prior to flow cytometry or cell culture to quantify S100 protein release. Whole transcriptome gene expression profiles were compared in neutrophils from children with both active SJIA and CID.

Results: Patients with SJIA and active systemic features demonstrated a higher proportion of CD16⁺CD62L^lo neutrophil population compared to controls. This neutrophil subset was not seen in patients with CID or patients with active arthritis not exhibiting systemic features. Using imaging flow cytometry, CD16⁺CD62L^lo neutrophils from patients with active SJIA and features of macrophage activation syndrome (MAS) had increased nuclear hypersegmentation compared to CD16⁺CD62L⁺ neutrophils. Serum levels of S100A8/A9 and S100A12 were strongly correlated with peripheral blood neutrophil counts. Neutrophils from active SJIA patients did not show enhanced resting S100 protein release; however, regardless of disease activity, neutrophils from SJIA patients did show enhanced S100A8/A9 release upon PMA stimulation compared to control neutrophils. Furthermore, whole transcriptome analysis of highly purified neutrophils from children with active SJIA identified 214 differentially expressed genes (DEG) compared to neutrophils from healthy controls. The most significantly upregulated gene pathway was Immune System Process, including AIM2, IL18RAP, and NLRC4. Interestingly, this gene set showed intermediate levels of expression in neutrophils from patients with long-standing CID yet persistent serum IL-18 elevation. Indeed, all patient samples regardless of disease activity demonstrated elevated inflammatory gene expression, including inflammasome components and S100A8.

Conclusion: We identify features of neutrophil activation in SJIA patients with both active disease and CID, including a proinflammatory gene expression signature, reflecting persistent innate immune activation. Taken together, these studies expand understanding of neutrophil function in chronic autoinflammatory disorders such as SJIA.

Using Chromatin Architecture to Understand the Genetics and Transcriptomics of Juvenile Idiopathic Arthritis

The presence of abnormal gene expression signatures is a well-described feature of the oligoarticular and polyarticular forms of juvenile idiopathic arthritis. In this review, we discuss how new insights into genetic risk for JIA and the three dimensional architecture of the genome may be used to develop a better understanding of the mechanisms driving these gene expression patterns.

Evaluation of plasma microRNA expressions in patients with juvenile idiopathic arthritis

Juvenile idiopathic arthritis (JIA) is the most common chronic rheumatic disease of childhood, yet its etiology is unknown. It is known that microribonucleic acids (miRNAs) play a role in immunoregulation. We aimed to evaluate the plasma expression of some candidate miRNAs that are associated with the pathogenesis of autoimmunity. Thirty-one patients diagnosed with JIA and age-sex-matched 31 healthy children were enrolled for the study. The plasma levels of four candidate miRNAs (miRNA-16, miRNA-155, miRNA-204, and miRNA-451), which are known to be associated with autoimmunity, were examined in all the subjects. The plasma levels of miRNAs were measured with real-time PCR in the patients in active and inactive periods and in the healthy controls. The groups were compared with each other. The plasma miRNA-155 levels were found to increase in the JIA patients compared to the healthy controls, and it was statistically more significant in the inactive period. We found that the JIA patients had the higher levels of miRNA-16 and the lower levels of miRNA-204/miRNA-451 expressions compare with the control group, but there was no statistically significant difference. A statistically significant decrease in the plasma levels of miRNA-204 was found in the patients that were in inactive disease with only methotrexate therapy. The plasma miRNA expressions were compared in the JIA subtypes, and it was observed that miRNA-204 levels were higher in polyarticular JIA and miRNA-451 levels were higher in enthesitis-related arthritis without statistical significance. The significant alterations in the plasma expression of miRNA-155 and miRNA-204 suggest to us that these molecules may be related to the pathogenesis of JIA. More comprehensive and functional researches about the role of these molecules are needed in this regard.

Identification of enhanced IFN-γ signaling in polyarticular juvenile idiopathic arthritis with mass cytometry

Polyarticular juvenile idiopathic arthritis (JIA) is among the most challenging of the JIA subtypes to treat. Even with current biologic therapies, the disease remains difficult to control in a substantial subset of patients, highlighting the need for new therapies. The aim of this study was to use the high dimensionality afforded by mass cytometry with phospho-specific antibodies to delineate signaling abnormalities in immune cells from treatment-naive polyarticular JIA patients. Peripheral blood mononuclear cells were isolated from 17 treatment-naive polyarticular JIA patients, 10 of the patients after achieving clinical remission, and 19 healthy controls. Samples were stimulated for 15 minutes with IL-6 or IFN-γ and analyzed by mass cytometry. Following IFN-γ stimulation, increased STAT1 and/or STAT3 phosphorylation was observed in subsets of CD4 T cells and classical monocytes from treatment-naive patients. The enhanced IFN-γ signaling was associated with increased expression of JAK1 and SOCS1 in CD4 T cells. Furthermore, substantial heterogeneity in surface marker expression was observed among the subsets of CD4 T cells and classical monocytes with increased IFN-γ responsiveness. The identification of enhanced IFN-γ signaling in CD4 T cells and classical monocytes from treatment-naive polyarticular JIA patients provides mechanistic support for investigations into therapies that attenuate IFN-γ signaling in this disease.

Modeling Transcriptional Rewiring in Neutrophils Through the Course of Treated Juvenile Idiopathic Arthritis

Neutrophils in children with the polyarticular form of juvenile idiopathic arthritis (JIA) display abnormal transcriptional patterns linked to fundamental metabolic derangements. In this study, we sought to determine the effects of therapy on mRNA and miRNA expression networks in polyarticular JIA. Using exon and miRNA microarrays, we studied children with untreated active JIA (ADU, n = 35), children with active disease on therapy with methotrexate ± etanercept (ADT, n = 26), and children with inactive disease also on therapy (ID, n = 14). We compared the results to findings from healthy control children (HC, n = 35). We found substantial re-ordering of mRNA and miRNA expression networks after the initiation of therapy. Each disease state was associated with a distinct transcriptional profile, with the ADT state differing the most from HC, and ID more strongly resembling HC. Changes at the mRNA level were mirrored in changes in miRNA expression patterns. The analysis of the expression dynamics from differentially expressed genes across three disease states indicated that therapeutic response is a complex process. This process does not simply involve genes slowly correcting in a linear fashion over time. Computational modeling of miRNA and transcription factor (TF) co-regulatory networks demonstrated that combinational regulation of miRNA and TF might play an important role in dynamic transcriptome changes.

Chromatin landscapes and genetic risk for juvenile idiopathic arthritis

Background

The transcriptomes of peripheral blood cells in children with juvenile idiopathic arthritis (JIA) have distinct transcriptional aberrations that suggest impairment of transcriptional regulation. To gain a better understanding of this phenomenon, we studied known JIA genetic risk loci, the majority of which are located in non-coding regions, where transcription is regulated and coordinated on a genome-wide basis. We examined human neutrophils and CD4 primary T cells to identify genes and functional elements located within those risk loci.

Methods

We analyzed RNA sequencing (RNA-Seq) data, H3K27ac and H3K4me1 chromatin immunoprecipitation-sequencing (ChIP-Seq) data, and previously published chromatin interaction analysis by paired-end tag sequencing (ChIA-PET) data to characterize the chromatin landscapes within the known JIA-associated risk loci.

Results

In both neutrophils and primary CD4+ T cells, the majority of the JIA-associated linkage disequilibrium (LD) blocks contained H3K27ac and/or H3K4me1 marks. These LD blocks were also binding sites for a small group of transcription factors, particularly in neutrophils. Furthermore, these regions showed abundant intronic and intergenic transcription in neutrophils. In neutrophils, none of the genes that were differentially expressed between untreated patients with JIA and healthy children were located within the JIA-risk LD blocks. In CD4+ T cells, multiple genes, including HLA-DQA1, HLA-DQB2, TRAF1, and IRF1 were associated with the long-distance interacting regions within the LD regions as determined from ChIA-PET data.

Conclusions

These findings suggest that genetic risk contributes to the aberrant transcriptional control observed in JIA. Furthermore, these findings demonstrate the challenges of identifying the actual causal variants within complex genomic/chromatin landscapes.

Electronic supplementary material

The online version of this article (doi:10.1186/s13075-017-1260-x) contains supplementary material, which is available to authorized users.

Immune and inflammatory gene expressions are different in Behçet's disease compared to those in Familial Mediterranean Fever

OBJECTIVE

The immune classification of Behçet's disease (BD) is still controversial. In this study, we aimed to compare the immune/inflammatory gene expressions in BD with those in familial Mediterranean fever (FMF), an autoinflammatory disorder with innate immune activation.

MATERIAL AND METHODS

CD4+ T cells and CD14+ monocytes were isolated from the peripheral blood mononuclear cells of Behçet's disease patients (n=10), FMF (n=6) patients, and healthy controls (n=4) with microbeads, and then, the mRNA was isolated. The expressions of 440 genes associated with immune and inflammatory responses were studied with a focused DNA microarray using a chemiluminescent tagging system. Changes above 1.5-fold and below 0.8-fold were accepted to be significant.

RESULTS

In BD patients, in the CD4+ T-lymphocyte subset, interleukin 18 receptor accessory protein (1.7-fold), IL-7 receptor (1.9-fold), and prokineticin 2 (2.5-fold) were all increased compared to those in FMF patients, whereas chemokine (C-X3-C motif ) receptor-1 (CX3CR1) (0.7-fold) and endothelial cell growth factor-1 (0.6-fold) were decreased. In the CD14+ monocyte population, the V-fos FBJ murine osteosarcoma viral oncogene homolog (1.5-fold), Interleukin-8 (IL-8) (2.1-fold), and Tumor Necrosis Factor alpha (TNF-α) (1.8-fold) were all increased, whereas the chemokine (C-C motif ) ligand 5 (CCL5) (0.6-fold), C-C chemokine receptor type 7 (0.6-fold), and CX3CR1 (0.7-fold) were decreased, again when compared to those in FMF. Compared to healthy controls in the CD4+ T-lymphocyte population, in both BD and FMF patients, pro-platelet basic protein and CD27 had elevated expression. In BD and FMF patients, 24 and 19 genes, respectively, were downregulated, with 15 overlapping genes between both disorders. In the CD14+ monocytes population, chemokine (C-C motif ) receptor-1 (CCR1) was upregulated both in BD and FMF patients compared to that in the controls, whereas CCL5 was downregulated.

CONCLUSION

Immune and inflammatory gene expressions seem to be variable in both the innate (CD14+) and adaptive (CD4+) immune responses in BD and FMF patients compared to those in controls, suggesting differences in immune regulation between the two disorders.

Limits of Peripheral Blood Mononuclear Cells for Gene Expression-Based Biomarkers in Juvenile Idiopathic Arthritis

Juvenile Idiopathic Arthritis (JIA) is one of the most common chronic disease conditions affecting children in the USA. As with many rheumatic diseases, there is growing interest in using genomic technologies to develop biomarkers for either diagnosis or to guide treatment ("personalized medicine"). Here, we explore the use of gene expression patterns in peripheral blood mononuclear cells (PBMC) as a first step approach to developing such biomarkers. Although PBMC carry many theoretical advantages for translational research, we have found that sample heterogeneity makes RNASeq on PBMC unsuitable as a first-step method for screening biomarker candidates in JIA. RNASeq studies of homogeneous cell populations are more likely to be useful and informative.

Whole blood expression profiling from the TREAT trial: insights for the pathogenesis of polyarticular juvenile idiopathic arthritis

Background

The Trial of Early Aggressive Therapy in Juvenile Idiopathic Arthritis (TREAT trial) was accompanied by a once-in-a-generation sample collection for translational research. In this paper, we report the results of whole blood gene expression analyses and genomic data-mining designed to cast light on the immunopathogenesis of polyarticular juvenile idiopathic arthritis (JIA).

Methods

TREAT samples and samples from an independent cohort were analyzed on Affymetrix microarrays and compared to healthy controls. Data from the independent cohort were used to validate the TREAT data. Pathways analysis was used to characterize gene expression profiles. Furthermore, we correlated differential gene expression with new information about functional regulatory elements within the genome to develop models of aberrant gene expression in JIA.

Results

There was a strong concordance in gene expression between TREAT samples and the independent cohort. In addition, rheumatoid factor (RF)-positive and RF-negative patients showed only small differences on whole blood expression profiles. Analysis of the combined samples showed 158 genes represented by 176 probes that showed differential expression between TREAT subjects at baseline and healthy controls. None of the differentially expressed genes were encoded within linkage disequilibrium blocks containing single nucleotide polymorphisms known to be associated with risk for JIA. Functional analysis of these genes showed functional associations with multiple processes associated with innate and adaptive immunity, and appeared to reflect overall suppression of STAT1–3/interferon response factor-mediated pathways.

Conclusions

Despite their limitations, whole blood expression profiles clearly distinguish children with polyarticular JIA from healthy controls. Whole blood expression profiles identify several immunologic pathways of biologic relevance that will need to be pursued in homogeneous cell populations in order to clarify mechanisms of pathogenesis.

Trial registration

ClinicalTrials.gov registry #NCT00443430, originally registered 2 March 2007 and last updated 30 May 2013.

Complexity and Specificity of the Neutrophil Transcriptomes in Juvenile Idiopathic Arthritis

NIH projects such as ENCODE and Roadmap Epigenomics have revealed surprising complexity in the transcriptomes of mammalian cells. In this study, we explored transcriptional complexity in human neutrophils, cells generally regarded as nonspecific in their functions and responses. We studied distinct human disease phenotypes and found that, at the gene, gene isoform, and miRNA level, neutrophils exhibit considerable specificity in their transcriptomes. Thus, even cells whose responses are considered non-specific show tailoring of their transcriptional repertoire toward specific physiologic or pathologic contexts. We also found that miRNAs had a global impact on neutrophil transcriptome and are associated with innate immunity in juvenile idiopathic arthritis (JIA). These findings have important implications for our understanding of the link between genes, non-coding transcripts and disease phenotypes.

Gene Expression Deconvolution for Uncovering Molecular Signatures in Response to Therapy in Juvenile Idiopathic Arthritis

Gene expression-based signatures help identify pathways relevant to diseases and treatments, but are challenging to construct when there is a diversity of disease mechanisms and treatments in patients with complex diseases. To overcome this challenge, we present a new application of an in silico gene expression deconvolution method, ISOpure-S1, and apply it to identify a common gene expression signature corresponding to response to treatment in 33 juvenile idiopathic arthritis (JIA) patients. Using pre- and post-treatment gene expression profiles only, we found a gene expression signature that significantly correlated with a reduction in the number of joints with active arthritis, a measure of clinical outcome (Spearman rho = 0.44, p = 0.040, Bonferroni correction). This signature may be associated with a decrease in T-cells, monocytes, neutrophils and platelets. The products of most differentially expressed genes include known biomarkers for JIA such as major histocompatibility complexes and interleukins, as well as novel biomarkers including α-defensins. This method is readily applicable to expression datasets of other complex diseases to uncover shared mechanistic patterns in heterogeneous samples.

Mapping of Variable DNA Methylation Across Multiple Cell Types Defines a Dynamic Regulatory Landscape of the Human Genome

DNA methylation is an important epigenetic modification involved in many biological processes and diseases. Many studies have mapped DNA methylation changes associated with embryogenesis, cell differentiation, and cancer at a genome-wide scale. Our understanding of genome-wide DNA methylation changes in a developmental or disease-related context has been steadily growing. However, the investigation of which CpGs are variably methylated in different normal cell or tissue types is still limited. Here, we present an in-depth analysis of 54 single-CpG-resolution DNA methylomes of normal human cell types by integrating high-throughput sequencing-based methylation data. We found that the ratio of methylated to unmethylated CpGs is relatively constant regardless of cell type. However, which CpGs made up the unmethylated complement was cell-type specific. We categorized the 26,000,000 human autosomal CpGs based on their methylation levels across multiple cell types to identify variably methylated CpGs and found that 22.6% exhibited variable DNA methylation. These variably methylated CpGs formed 660,000 variably methylated regions (VMRs), encompassing 11% of the genome. By integrating a multitude of genomic data, we found that VMRs enrich for histone modifications indicative of enhancers, suggesting their role as regulatory elements marking cell type specificity. VMRs enriched for transcription factor binding sites in a tissue-dependent manner. Importantly, they enriched for GWAS variants, suggesting that VMRs could potentially be implicated in disease and complex traits. Taken together, our results highlight the link between CpG methylation variation, genetic variation, and disease risk for many human cell types.

Mitochondrial and oxidative stress genes are differentially expressed in neutrophils of sJIA patients treated with tocilizumab: a pilot microarray study

BACKGROUND

Various pathways involved in the pathogenesis of sJIA have been identified through gene expression profiling in peripheral blood mononuclear cells (PBMC), but not in neutrophils. Since neutrophils are important in tissue damage during inflammation, and are elevated as part of the acute phase response, we hypothesised that neutrophil pathways could also be important in the pathogenesis of sJIA. We therefore studied the gene profile in both PBMC and neutrophils of sJIA patients treated with tocilizumab.

METHODS

We studied the transcriptomes of peripheral blood mononuclear cells (PBMC) and neutrophils from eight paired samples obtained from 4 sJIA patients taken before and after treatment, selected on the basis that they achieved ACR90 responses within 12 weeks of therapy initiation with tocilizumab. RNA was extracted and gene expression profiling was performed using Affymetrix GeneChip Human Genome U133 Plus 2.0 microarray platform. A longitudinal analysis using paired t-test (p < 0.05 and FC ≥ 1.5) was applied to identify differentially expressed genes (DEGs) between the two time points followed by ingenuity pathway analysis. Gene Set Enrichment Analysis (GSEA) and quantitative real-time PCR were then performed to verify the microarray results.

RESULTS

Gene ontology analysis in neutrophils revealed that response to tocilizumab significantly altered genes regulating mitochondrial dysfunction and oxidative stress (p = 4.6E-05). This was independently verified with GSEA, by identifying a set of oxidative genes whose expression correlated with response to tocilizumab. In PBMC, treatment of sJIA with tocilizumab appeared to affect genes in Oncostatin M signalling and B cell pathways.

CONCLUSIONS

For the first time we demonstrate that neutrophils from sJIA patients responding to tocilizumab showed significantly different changes in gene expression. These data could highlight the importance of mitochondrial genes that modulate oxidative stress in the pathogenesis of sJIA.

Dynamic tracking of functional gene modules in treated juvenile idiopathic arthritis

BACKGROUND

We have previously shown that childhood-onset rheumatic diseases show aberrant patterns of gene expression that reflect pathology-associated co-expression networks. In this study, we used novel computational approaches to examine how disease-associated networks are altered in one of the most common rheumatic diseases of childhood, juvenile idiopathic arthritis (JIA).

METHODS

Using whole blood gene expression profiles derived from children in a pediatric rheumatology clinical trial, we used a network approach to understanding the impact of therapy and the underlying biology of response/non-response to therapy.

RESULTS

We demonstrate that therapy for JIA is associated with extensive re-ordering of gene expression networks, even in children who respond inadequately to therapy. Furthermore, we observe distinct differences in the evolution of specific network properties when we compare children who have been treated successfully with those who have inadequate treatment response.

CONCLUSIONS

Despite the inherent noisiness of whole blood gene expression data, our findings demonstrate how therapeutic response might be mapped and understood in pathologically informative cells in a broad range of human inflammatory diseases.

Disease-Associated Single-Nucleotide Polymorphisms From Noncoding Regions in Juvenile Idiopathic Arthritis Are Located Within or Adjacent to Functional Genomic Elements of Human Neutrophils and CD4+ T Cells

OBJECTIVE

Juvenile idiopathic arthritis (JIA) is considered a complex disease in which the environment interacts with inherited genes to produce a phenotype that shows broad interindividual variability. Twenty-four regions of genetic risk for JIA were identified in a recent genome-wide association study (GWAS); however, as is typical of the results of GWAS, most of the regions of genetic risk (22 of 24) were in noncoding regions of the genome. This study was undertaken to identify functional elements (other than genes) that might be located within the regions of genetic risk.

METHODS

We used paired-end RNA sequencing to identify noncoding RNAs (ncRNAs) located within 5 kb of disease-associated single-nucleotide polymorphisms (SNPs). In addition, we used chromatin immunoprecipitation (ChIP) followed by sequencing to identify epigenetic marks associated with enhancer function (H3K4me1 and H3K27ac) in human neutrophils to determine whether enhancer-associated histone marks were enriched in the linkage disequilibrium blocks that encompassed the 22 SNPs identified in the GWAS.

RESULTS

In human neutrophils, we identified H3K4me1 and/or H3K27ac marks in 15 of the 22 regions previously identified as risk loci for JIA. In CD4+ T cells, 18 regions had H3K4me1 and/or H3K27ac marks. In addition, we identified ncRNA transcripts at the rs4705862 and rs6894249 loci in human neutrophils.

CONCLUSION

Much of the genetic risk for JIA lies within or adjacent to regions of neutrophil and CD4+ T cell genomes that carry epigenetic marks associated with enhancer function and/or ncRNA transcripts. These findings are consistent with the hypothesis that JIA is fundamentally a disorder of gene regulation that includes both the innate and the adaptive immune system. Elucidating the specific roles of these noncoding elements within leukocyte genomes will be critical to our understanding of JIA pathogenesis.

RNA sequencing from human neutrophils reveals distinct transcriptional differences associated with chronic inflammatory states

BACKGROUND

The transcriptional complexity of mammalian cells suggests that they have broad abilities to respond to specific environmental stimuli and physiologic contexts. These abilities were not apparent a priori from the structure of mammalian genomes, but have been identified through detailed transcriptome analyses. In this study, we examined the transcriptomes of cells of the innate immune system, human neutrophils, using RNA sequencing (RNAseq).

METHODS

We sequenced poly-A RNA from nine individual samples corresponding to specific phenotypes: three children with active, untreated juvenile idiopathic arthritis (JIA)(AD), three children with the same disease whose disease was inactive on medication (CRM), and three children with cystic fibrosis (CF).

RESULTS

We demonstrate that transcriptomes of neutrophils, typically considered non-specific in their responses and functions, display considerable specificity in their transcriptional repertoires dependent on the pathologic context, and included genes, gene isoforms, and long non-coding RNA transcripts. Furthermore, despite the small sample numbers, these findings demonstrate the potential of RNAseq approaches to biomarker development in rheumatic diseases.

CONCLUSIONS

These data demonstrate the capacity of cells previously considered non-specific in function to adapt their transcriptomes to specific biologic contexts. These data also provide insight into previously unrecognized pathological pathways and show considerable promise for elucidating disease and disease-state specific regulatory networks.

The biologic basis of clinical heterogeneity in juvenile idiopathic arthritis

Objective

Childhood arthritis encompasses a heterogeneous family of diseases. Significant variation in clinical presentation remains despite consensus-driven diagnostic classifications. Developments in data analysis provide powerful tools for interrogating large heterogeneous data sets. We report a novel approach to integrating biologic and clinical data toward a new classification for childhood arthritis, using computational biology for data-driven pattern recognition.

Methods

Probabilistic principal components analysis was used to transform a large set of data into 4 interpretable indicators or composite variables on which patients were grouped by cluster analysis. Sensitivity analysis was conducted to determine key variables in determining indicators and cluster assignment. Results were validated against an independent validation cohort.

Results

Meaningful biologic and clinical characteristics, including levels of proinflammatory cytokines and measures of disease activity, defined axes/indicators that identified homogeneous patient subgroups by cluster analysis. The new patient classifications resolved major differences between patient subpopulations better than International League of Associations for Rheumatology subtypes. Fourteen variables were identified by sensitivity analysis to crucially determine indicators and clusters. This new schema was conserved in an independent validation cohort.

Conclusion

Data-driven unsupervised machine learning is a powerful approach for interrogating clinical and biologic data toward disease classification, providing insight into the biology underlying clinical heterogeneity in childhood arthritis. Our analytical framework enabled the recovery of unique patterns from small cohorts and addresses a major challenge, patient numbers, in studying rare diseases.

Anti-type II collagen immune complex-induced granulocyte reactivity is associated with joint erosions in RA patients with anti-collagen antibodies

Introduction

Rheumatoid arthritis (RA) patients with autoantibodies against collagen type II (CII) are characterized by acute RA onset with elevated inflammatory measures and early joint erosions as well as increased production of tumor necrosis factor-α (ΤΝF-α) by peripheral blood mononuclear cells (PBMC) stimulated by anti-CII immune complexes (IC) in vitro. Polymorphonuclear granulocytes (PMN) are abundant in RA synovial fluids, where they might interact directly with anti-CII IC in the articular cartilage, but no studies have investigated PMN responses towards anti-CII IC. The aim was to investigate whether PMN react towards anti-CII IC, and to what extent such reactivity might relate to the clinical acute onset RA phenotype associated with elevated levels of anti-CII.

Methods

PMN and PBMC isolated from healthy donors were stimulated with IC made with a set of 72 baseline patient sera (24 anti-CII positive, 48 anti-CII negative) chosen from a clinically well-characterized RA cohort with two-year radiological follow-up with Larsen scoring. PMN expression of cluster of differentiation (CD)11b, CD66b, CD16 and CD32 was measured by flow cytometry, whereas PMN production of myeloperoxidase (MPO) and interleukin (IL)-17, and PBMC production of ΤΝF-α was measured with enzyme linked immunosorbent assay.

Results

PMN expression of CD11b, CD66b and MPO, and PBMC production of ΤΝF-α were upregulated whereas PMN expression of CD16 and CD32 were downregulated by anti-CII IC. CD16, CD66b, and MPO production correlated to serum anti-CII levels (Spearman’s ρ = 0.315, 0.675 and 0.253, respectively). CD16 was associated with early joint erosions (P = 0.024, 0.034, 0.046 at baseline, one and two years) and CD66b was associated with changes in joint erosions (P = 0.017 and 0.016, at one and two years compared to baseline, respectively). CD66b was associated with baseline C-reactive protein and PBMC production of ΤΝF-α was associated with baseline erythrocyte sedimentation rate, in accordance with our earlier findings. No clinical associations were observed for MPO or IL-17.

Conclusion

PMN responses against anti-CII IC are more closely associated with early joint erosions than are PBMC cytokine responses. PMN reactivity against anti-CII IC may contribute to joint destruction in newly diagnosed RA patients with high levels of anti-CII.

Gene expression during the generation and activation of mouse neutrophils: implication of novel functional and regulatory pathways

As part of the Immunological Genome Project (ImmGen), gene expression was determined in unstimulated (circulating) mouse neutrophils and three populations of neutrophils activated in vivo, with comparison among these populations and to other leukocytes. Activation conditions included serum-transfer arthritis (mediated by immune complexes), thioglycollate-induced peritonitis, and uric acid-induced peritonitis. Neutrophils expressed fewer genes than any other leukocyte population studied in ImmGen, and down-regulation of genes related to translation was particularly striking. However, genes with expression relatively specific to neutrophils were also identified, particularly three genes of unknown function: Stfa2l1, Mrgpr2a and Mrgpr2b. Comparison of genes up-regulated in activated neutrophils led to several novel findings: increased expression of genes related to synthesis and use of glutathione and of genes related to uptake and metabolism of modified lipoproteins, particularly in neutrophils elicited by thioglycollate; increased expression of genes for transcription factors in the Nr4a family, only in neutrophils elicited by serum-transfer arthritis; and increased expression of genes important in synthesis of prostaglandins and response to leukotrienes, particularly in neutrophils elicited by uric acid. Up-regulation of genes related to apoptosis, response to microbial products, NFkB family members and their regulators, and MHC class II expression was also seen, in agreement with previous studies. A regulatory model developed from the ImmGen data was used to infer regulatory genes involved in the changes in gene expression during neutrophil activation. Among 64, mostly novel, regulatory genes predicted to influence these changes in gene expression, Irf5 was shown to be important for optimal secretion of IL-10, IP-10, MIP-1α, MIP-1β, and TNF-α by mouse neutrophils in vitro after stimulation through TLR9. This data-set and its analysis using the ImmGen regulatory model provide a basis for additional hypothesis-based research on the importance of changes in gene expression in neutrophils in different conditions.

Genomic characterization of remission in juvenile idiopathic arthritis

Introduction

The attainment of remission has become an important end point for clinical trials in juvenile idiopathic arthritis (JIA), although we do not yet have a full understanding of what remission is at the cell and molecular level.

Methods

Two independent cohorts of patients with JIA and healthy child controls were studied. RNA was prepared separately from peripheral blood mononuclear cells (PBMC) and granulocytes to identify differentially expressed genes using whole genome microarrays. Expression profiling results for selected genes were confirmed by quantitative, real-time polymerase chain reaction (RT-PCR).

Results

We found that remission in JIA induced by either methotrexate (MTX) or MTX plus a TNF inhibitor (etanercept, Et) (MTX + Et) is characterized by numerous differences in gene expression in peripheral blood mononuclear cells and in granulocytes compared with healthy control children; that is, remission is not a restoration of immunologic normalcy. Network analysis of the differentially expressed genes demonstrated that the steroid hormone receptor superfamily member hepatocyte nuclear factor 4 alpha (HNF4α) is a hub in several of the gene networks that distinguished children with arthritis from controls. Confocal microscopy revealed that HNF4a is present in both T lymphocytes and granulocytes, suggesting a previously unsuspected role for this transcription factor in regulating leukocyte function and therapeutic response in JIA.

Conclusions

These findings provide a framework from which to understand therapeutic response in JIA and, furthermore, may be used to develop strategies to increase the frequency with which remission is achieved in adult forms of rheumatoid arthritis.

Correlation analyses of clinical and molecular findings identify candidate biological pathways in systemic juvenile idiopathic arthritis

BACKGROUND

Clinicians have long appreciated the distinct phenotype of systemic juvenile idiopathic arthritis (SJIA) compared to polyarticular juvenile idiopathic arthritis (POLY). We hypothesized that gene expression profiles of peripheral blood mononuclear cells (PBMC) from children with each disease would reveal distinct biological pathways when analyzed for significant associations with elevations in two markers of JIA activity, erythrocyte sedimentation rate (ESR) and number of affected joints (joint count, JC).

METHODS

PBMC RNA from SJIA and POLY patients was profiled by kinetic PCR to analyze expression of 181 genes, selected for relevance to immune response pathways. Pearson correlation and Student's t-test analyses were performed to identify transcripts significantly associated with clinical parameters (ESR and JC) in SJIA or POLY samples. These transcripts were used to find related biological pathways.

RESULTS

Combining Pearson and t-test analyses, we found 91 ESR-related and 92 JC-related genes in SJIA. For POLY, 20 ESR-related and 0 JC-related genes were found. Using Ingenuity Systems Pathways Analysis, we identified SJIA ESR-related and JC-related pathways. The two sets of pathways are strongly correlated. In contrast, there is a weaker correlation between SJIA and POLY ESR-related pathways. Notably, distinct biological processes were found to correlate with JC in samples from the earlier systemic plus arthritic phase (SAF) of SJIA compared to samples from the later arthritis-predominant phase (AF). Within the SJIA SAF group, IL-10 expression was related to JC, whereas lack of IL-4 appeared to characterize the chronic arthritis (AF) subgroup.

CONCLUSIONS

The strong correlation between pathways implicated in elevations of both ESR and JC in SJIA argues that the systemic and arthritic components of the disease are related mechanistically. Inflammatory pathways in SJIA are distinct from those in POLY course JIA, consistent with differences in clinically appreciated target organs. The limited number of ESR-related SJIA genes that also are associated with elevations of ESR in POLY implies that the SJIA associations are specific for SJIA, at least to some degree. The distinct pathways associated with arthritis in early and late SJIA raise the possibility that different immunobiology underlies arthritis over the course of SJIA.

Juvenile idiopathic arthritis

Juvenile idiopathic arthritis is a heterogeneous group of diseases characterised by arthritis of unknown origin with onset before age of 16 years. Pivotal studies in the past 5 years have led to substantial progress in various areas, ranging from disease classification to new treatments. Gene expression profiling studies have identified different immune mechanisms in distinct subtypes of the disease, and can help to redefine disease classification criteria. Moreover, immunological studies have shown that systemic juvenile idiopathic arthritis is an acquired autoinflammatory disease, and have led to successful studies of both interleukin-1 and interleukin-6 blockade. In other forms of the disease, synovial inflammation is the consequence of a disturbed balance between proinflammatory effector cells (such as T-helper-17 cells), and anti-inflammatory regulatory cells (such as FOXP3-positive regulatory T cells). Moreover, specific soluble biomarkers (S100 proteins) can guide individual treatment. Altogether these new developments in genetics, immunology, and imaging are instrumental to better define, classify, and treat patients with juvenile idiopathic arthritis.

Functional genomics and rheumatoid arthritis: where have we been and where should we go?

Studies in model organisms and humans have begun to reveal the complexity of the transcriptome. In addition to serving as passive templates from which genes are translated, RNA molecules are active, functional elements of the cell whose products can detect, interact with, and modify other transcripts. Gene expression profiling is the method most commonly used thus far to enrich our understanding of the molecular basis of rheumatoid arthritis in adults and juvenile idiopathic arthritis in children. The feasibility of this approach for patient classification (for example, active versus inactive disease, disease subsets) and improving prognosis (for example, response to therapy) has been demonstrated over the past 7 years. Mechanistic understanding of disease-related differences in gene expression must be interpreted in the context of interactions with transcriptional regulatory molecules and epigenetic alterations of the genome. Ongoing work regarding such functional complexities in the human genome will likely bring both insight and surprise to our understanding of rheumatoid arthritis.

Oligoarticular and polyarticular JIA: epidemiology and pathogenesis

Juvenile idiopathic arthritis (JIA) refers to a group of chronic childhood arthropathies of unknown etiology, currently classified into subtypes primarily on the basis of clinical features. Research has focused on the hypothesis that these subtypes arise through distinct etiologic pathways. In this Review, we discuss four subtypes of JIA: persistent oligoarticular, extended oligoarticular, rheumatoid-factor-positive polyarticular and rheumatoid-factor-negative polyarticular. These subtypes differ in prevalence between ethnic groups and are associated with different HLA alleles. Non-HLA genetic risk factors have also been identified, some of which reveal further molecular differences between these subtypes, while others suggest mechanistic overlap. Investigations of immunophenotypes also provide insights into subtype differences: adaptive immunity seems to have a prominent role in both polyarticular and oligoarticular JIA, and the more-limited arthritis observed in persistent oligoarticular JIA as compared with extended oligoarticular JIA may reflect more-potent immunoregulatory T-cell activity in the former. Tumor necrosis factor seems to be a key mediator of both polyarticular and oligoarticular JIA, especially in the extended oligoarticular subtype, although elevated levels of other cytokines are also observed. Limited data on monocytes, dendritic cells, B cells, natural killer T cells and neutrophils suggest that the contributions of these cells differ across subtypes of JIA. Within each subtype, however, common pathways seem to drive joint damage.