Paroxysmal nocturnal hemoglobinuria (PNH) is a rare acquired disorder originating from hematopoietic stem cells and is a life-threating disease characterized by intravascular hemolysis, bone marrow (BM) failure, and venous thrombosis. The etiology of PNH is a somatic mutation in the phosphatidylinositol glycan class A gene (PIG-A) on the X chromosome, which blocks synthesis of the glycolipid moiety and causes deficiency in GPI-anchored proteins. PNH is closely related to aplastic anemia, in which T cells mediate destruction of BM. To identify aberrant molecular mechanisms involved in immune targeting of hematopoietic stem cells in BM, we applied RNA-seq to examine the transcriptome of T cell subsets (CD4+ naive, CD4+ memory, CD8+ naive, and CD8+ memory) from PNH patients and healthy control subjects. Differentially expressed gene analysis in four different T cell subsets from PNH and healthy control subjects showed distinct transcriptional profiles, depending on the T cell subsets. By pathway analysis, we identified novel signaling pathways in T cell subsets from PNH, including increased gene expression involved in TNFR, IGF1, NOTCH, AP-1, and ATF2 pathways. Dysregulation of several candidate genes (JUN, TNFAIP3, TOB1, GIMAP4, GIMAP6, TRMT112, NR4A2, CD69, and TNFSF8) was validated by quantitative real-time RT-PCR and flow cytometry. We have demonstrated molecular signatures associated with positive and negative regulators in T cells, suggesting novel pathophysiologic mechanisms in PNH. These pathways may be targets for new strategies to modulate T cell immune responses in BM failure.

Paroxysmal nocturnal hemoglobinuria (PNH) is a life-threatening rare blood disease, characterized by hemolytic anemia, bone marrow (BM) failure and venous thrombosis (1, 2). PNH originates in somatic mutations in hematopoietic stem cells (HSCs) of the phosphatidylinositol glycan class A gene (PIG-A), which block synthesis of the glycolipid moiety and cause global deficiency in GPI-anchored proteins (GPI-APs) (3).

PNH is clinically closely related to aplastic anemia (AA) (4). In most AA, the pathophysiology is immunity to HSCs mediated by activated cytotoxic T cells and destruction of marrow hematopoietic cells (5, 6). CD8+ cytotoxic T cells with restricted TCR diversity (oligoclonal T cells) are expanded in AA, and increased production of proinflammatory cytokines, such as IFN-γ, induces apoptosis of HSCs (7). We recently reported high frequency of circulating CD8+ memory stem T cells in AA, suggesting CD8+ memory stem T cells may be a biomarker and a therapeutic target in AA (8). Hematologic recovery after immunosuppressive therapy (IST) with anti-thymocyte globulin (ATG) and cyclosporine (CsA) occurs in 60–75% of AA patients (9, 10), and correlates to decreased expanded T cell clones (8, 11).

The involvement of T cells in PNH is strongly supported by the following evidence: clinical overlap between PNH and AA (4, 12), the presence of GPI-AP defective cells in AA associated with favorable response to IST (1315), and an oligoclonal T cell repertoire in PNH patients (16, 17). By flow cytometry, T cells from PNH patients show an altered CD40-dependent pathway (18), and killer cell Ig-like receptors are differentially expressed in T cells from PNH patients (19). In one recent study, there was expansion of autoreactive, CD1d-restricted, GPI-specific T cells in PNH patients, which appeared to be directed to the GPI-anchor itself (17). However, the molecular mechanisms responsible for the aberrant immune responses in PNH patients are not well understood.

RNA sequencing (RNA-seq) is a powerful method to quantitate gene expression, splice variants, and single-nucleotide variants, and for the discovery of novel transcripts and fusion genes (20, 21). To identify aberrant molecular mechanisms involved in immune targeting of HSCs in BM, we applied RNA-seq to examine the transcriptome of T cell subsets from PNH patients and healthy control subjects. We identified novel pathways in T cell subsets from PNH patients by RNA-seq. Our results demonstrate the value of RNA-seq in capturing the full dynamic range of the transcriptome of T cells in PNH, and in identifying key factors and pathways involved in regulation of T cells. We successfully validated dysregulated expression of several candidate genes. Increased expression of specific genes may underline BM failure in PNH patients.

Blood samples were collected after written informed consent in accordance with the Declaration of Helsinki and under protocols approved by the Institutional Review Board of the National Heart, Lung, and Blood Institute (www.clinicaltrials.gov, NCT00071045). Median age of 15 PNH patients was 41 y (range 20–59 y). Samples from 15 PNH patients and 15 age- and sex-matched healthy control subjects were used in this study. Twelve of 15 PNH patients had a prior history of IST for AA. Among three patients examined by RNA-seq, one had no prior history of IST. Median duration of prior IST from sampling was 6 y (3 mo to 24 y). ISTs were not being administered to any patient at the time of sampling. Demographic and clinical characteristics of PNH patients are summarized in Table I.

Table I.
Patient characteristics
Patient No.Age (y)SexPNH Granulocytes (%)Antecedent AAHistory of Prior ISTDuration of Prior IST from SamplingThrombotic EventsANCARCPlateletsExperiments
41 85 No None — No 2.02 156.0 73 RNA-seq, RT-qPCR 
49 29 Yes rATG+CsA, hATG, TPO-RA 5 y Yes 1.80 99.9 71 RNA-seq, RT-qPCR 
23 65 Yes rATG+CsA 8 y No 1.90 120.3 127 RNA-seq, RT-qPCR 
32 99 Yes rATG+CsA 7 y No 2.21 226.1 274 RT-qPCR 
33 95 Yes rATG+CsA 5 y No 2.79 165.2 191 RT-qPCR 
56 Yes hATG+CsA+TPO-RA 3 mo No 1.36 160.5 20 RT-qPCR 
59 61 Yes hATG+CsA+TPO-RA 3 y No 2.12 43.3 88 RT-qPCR 
27 98 Yes Cyclophosphamide+CsA 4 y No 3.11 178.8 138 Flow cytometry 
44 98 Yes hATG+CsA 24 y No 1.75 210.9 403 Flow cytometry 
10 56 78 No None — No 5.27 183.2 262 Flow cytometry 
11 43 95 Yes Alemtuzumab, rATG+CsA, hATG+CsA 9 y No 2.94 261.6 214 Flow cytometry 
12 46 81 Yes None — No 0.26 15.3 18 Flow cytometry 
13 20 46 Yes rATG+CsA, alemtuzumab 7 y No 1.75 107.6 236 Flow cytometry 
14 39 22 Yes hATG+CsA 13 y No 2.79 93.3 138 Flow cytometry 
15 24 73 Yes hATG+CsA, alemtuzumab 1 y No 0.70 5.8 Flow cytometry 
Patient No.Age (y)SexPNH Granulocytes (%)Antecedent AAHistory of Prior ISTDuration of Prior IST from SamplingThrombotic EventsANCARCPlateletsExperiments
41 85 No None — No 2.02 156.0 73 RNA-seq, RT-qPCR 
49 29 Yes rATG+CsA, hATG, TPO-RA 5 y Yes 1.80 99.9 71 RNA-seq, RT-qPCR 
23 65 Yes rATG+CsA 8 y No 1.90 120.3 127 RNA-seq, RT-qPCR 
32 99 Yes rATG+CsA 7 y No 2.21 226.1 274 RT-qPCR 
33 95 Yes rATG+CsA 5 y No 2.79 165.2 191 RT-qPCR 
56 Yes hATG+CsA+TPO-RA 3 mo No 1.36 160.5 20 RT-qPCR 
59 61 Yes hATG+CsA+TPO-RA 3 y No 2.12 43.3 88 RT-qPCR 
27 98 Yes Cyclophosphamide+CsA 4 y No 3.11 178.8 138 Flow cytometry 
44 98 Yes hATG+CsA 24 y No 1.75 210.9 403 Flow cytometry 
10 56 78 No None — No 5.27 183.2 262 Flow cytometry 
11 43 95 Yes Alemtuzumab, rATG+CsA, hATG+CsA 9 y No 2.94 261.6 214 Flow cytometry 
12 46 81 Yes None — No 0.26 15.3 18 Flow cytometry 
13 20 46 Yes rATG+CsA, alemtuzumab 7 y No 1.75 107.6 236 Flow cytometry 
14 39 22 Yes hATG+CsA 13 y No 2.79 93.3 138 Flow cytometry 
15 24 73 Yes hATG+CsA, alemtuzumab 1 y No 0.70 5.8 Flow cytometry 

ISTs were not being administered to any patient at the time of sampling.

ANC, absolute neutrophil count; ARC, absolute reticulocyte count; F, female; hATG, horse ATG; M, male; rATG, rabbit ATG; TPO-RA, thrombopoietin receptor agonist.

For cell sorting, PBMCs from PNH patients and healthy control subjects were separated from peripheral blood samples using Lymphocyte Separation Medium (MP Biomedicals, Santa Ana, CA). For flow cytometry, PBMCs were cryopreserved in RPMI-1640 (Life Technologies, Gaithersburg, MD) supplemented with 20% heat-inactivated FBS (Sigma-Aldrich, St. Louis, MO) and 10% DMSO, according to the standard protocol, until use.

The following fluorochrome-conjugated mAbs were purchased from commercial vendors and used for surface staining: anti–CD4-V500, anti–CD8-allophycocyanin-H7, anti–CD45RA-PE-Cy7, anti–CD45RO-allophycocyanin, anti–CCR7-AF700, anti–CD95-PE, anti–PD-1-FITC, and anti–CD69-FITC (BD Biosciences, San Jose, CA); anti–CD3-BV605 (BioLegend, San Diego, CA); anti–CD14-Pacific Blue and anti–CD19-Pacific Blue (Life Technologies, Carlsbad, CA); anti–CD30 ligand/TNFSF8-AF488 and anti–A20/TNFAIP3-AF488 (Novus Biologicals, Littleton, CO); and anti–CD27-PC5 (Beckman Coulter, Indianapolis, IN). The fixable violet amine reactive dye (ViViD; Invitrogen/Molecular Probes, Eugene, OR) was used to eliminate dead cells by flow cytometry.

For RNA extraction, freshly isolated PBMCs were sorted on the same day of blood draw to obtain four different T cell (CD3+ CD14 CD19 ViViD) populations (CD4+ naive [CD45RA+ CD45RO], CD4+ memory [CD45RA CD45RO+], CD8+ naive [CD45RA+ CD45RO], and CD8+ memory [CD45RA CD45RO+] T cells) by FACS using an Aria II instrument (Becton Dickinson, Franklin Lakes, NJ). Gating strategies for sorting T cell subsets are summarized in Fig. 1A.

FIGURE 1.

Isolation and molecular characterization of T cell subsets from PNH patients. (A) Gating strategy for T cell subsets for cell sorting. PBMCs were stained with ViViD, anti–CD14-Pacific Blue, anti–CD19-Pacific Blue, anti–CD3-BV605, anti–CD4-V500, anti–CD8-allophycocyanin-H7, anti–CD45RA-PE-Cy7, and anti–CD45RO-allophycocyanin. Lymphocytes were gated based on their scatter characteristics or forward scatter height versus forward scatter area. Live T cells were gated based on positive staining for CD3 and negative staining for ViViD, CD14, and CD19 to remove dead cells, monocytes, and B cells. CD4+ and CD8+ T cells were then gated based on the characteristic expression patterns of CD45RA and CD45RO. Four different T cell subsets [CD4+ naive (CD45RA+ CD45RO), CD4+ memory (CD45RA CD45RO+), CD8+ naive (CD45RA+ CD45RO), and CD8+ memory (CD45RA CD45RO+) T cells] were sorted for RNA-seq. (B) Transcriptional validation of the TN and memory T cell identity by expression levels of CD4, CD8, CCR5, and EOMES. FPKM, fragments per kilobase of transcript per million mapped reads. *p < 0.05.

FIGURE 1.

Isolation and molecular characterization of T cell subsets from PNH patients. (A) Gating strategy for T cell subsets for cell sorting. PBMCs were stained with ViViD, anti–CD14-Pacific Blue, anti–CD19-Pacific Blue, anti–CD3-BV605, anti–CD4-V500, anti–CD8-allophycocyanin-H7, anti–CD45RA-PE-Cy7, and anti–CD45RO-allophycocyanin. Lymphocytes were gated based on their scatter characteristics or forward scatter height versus forward scatter area. Live T cells were gated based on positive staining for CD3 and negative staining for ViViD, CD14, and CD19 to remove dead cells, monocytes, and B cells. CD4+ and CD8+ T cells were then gated based on the characteristic expression patterns of CD45RA and CD45RO. Four different T cell subsets [CD4+ naive (CD45RA+ CD45RO), CD4+ memory (CD45RA CD45RO+), CD8+ naive (CD45RA+ CD45RO), and CD8+ memory (CD45RA CD45RO+) T cells] were sorted for RNA-seq. (B) Transcriptional validation of the TN and memory T cell identity by expression levels of CD4, CD8, CCR5, and EOMES. FPKM, fragments per kilobase of transcript per million mapped reads. *p < 0.05.

Close modal

To validate RNA-seq data by flow cytometry, we quantified CD69 and TNFSF8 expression in T cell subsets using several mAbs, according to the established protocol with a minor modification, as previously described (8). In brief, PBMCs were incubated with a viability marker at room temperature for 20 min, washed, and then incubated with anti-human CCR7 at 37°C for 20 min. After washing, cells were further stained with a mixture of Abs against CD3, CD4, CD8, CD45RA, CD45RO, CD27, CD95, CD14, and CD19 with or without anti-CD69/TNFSF8 Ab, respectively, for 30 min on ice. Cells were either stimulated with or without Dynabeads Human T-Activator CD3/CD28 (Life Technologies) for 6 h before staining. At least 150,000 events gated on CD3+ T cells were acquired with Fortessa flow cytometer (BD Biosciences). Each T cell subset was defined as follows: central memory T cells (TCM), ViViD CD3+ CD4 (CD8)+ CD45RO+ CCR7+; effector memory T cells (TEM), ViViD CD3+ CD4 (CD8)+ CD45RO+ CCR7; terminally differentiated effector T cells (TEs), ViViD CD3+ CD4 (CD8)+ CD45RO CD45RA+ CCR7 CD27; and naive T cells (TN), CD3+ CD4 (CD8)+ CD45RO CD45RA+ CCR7+ CD27+ CD95. Quantification of PD-1 expression in T cell subsets has been described (22). For intracellular staining of TNFAIP3, cells were incubated with the cell-surface staining Ab mixture, as described earlier, and were fixed/permeabilized using the Cytofix/Cytoperm Fixation and Permeabilization Solution (BD Biosciences), according to the manufacturer’s protocol. Intracellular staining was performed using anti–A20/TNFAIP3-AF488 at 4°C for 30 min. Data were analyzed using FlowJo software version 9.6 (Tree Star, Ashland, OR).

Total RNA was isolated using the RNeasy Mini kit (Qiagen, Valencia, CA), according to the manufacturer’s instructions. RNA concentration was measured using the NanoDrop device (Peqlab, Erlangen, Germany). RNA quality was further assessed using the Agilent 2100 Bioanalyzer to obtain an RNA Integrity Number score.

Quality of total RNA extracted from three PNH patients and three healthy control subjects (CD4+ naive, CD4+ memory, CD8+ naive, and CD8+ memory T cells for each sample) was assessed using the Agilent 2100 Bioanalyzer. RNA-seq and analysis were performed by the Beijing Genomics Institute (Hong Kong) using the Illumina TruSeq Stranded Total RNA Library Prep kit and the Illumina HiSeq 2000 platform, according to the institute’s protocols.

Genes were compared with demonstrated differences in fragments per kilobase of transcript per million mapped reads (FPKM) between PNH and healthy control groups. EBSeq was used to identify differentially expressed genes (23). A threshold of Abs [log2 (Y/X)] ≥ 1 and posterior probability of being equally expressed (PPEE) ≤ 0.05 were used to identify differentially expressed RNAs between PNH patients and healthy control groups. Cummerbund was used for visualization of differential expression results. These data are available under Gene Expression Omnibus series accession number GSE83808.

Ingenuity Pathway Analysis (IPA) was used to determine differentially regulated biological pathways by loading the lists of statistically significant differentially expressed genes into IPA software (www.ingenuity.com). Statistically significant (p ≤ 0.05) biological pathways were reported. Graphical representations of the networks were generated with Path Designer.

Gene set enrichment analysis (GSEA) was performed as described previously (24). The gene expression signatures were analyzed using the java GSEA package (http://software.broadinstitute.org/gsea/index.jsp). The most differentially expressed genes ranked by ratio for each comparison were used to generate a signature for GSEA analysis. We compared the gene expression levels from two different samples (PNH patients versus healthy control subjects) for each T cell subset. GSEA was performed by computing overlaps with c2: curated gene sets (all canonical pathways, gene symbols) obtained from the Broad Institute (http://software.broadinstitute.org/gsea/msigdb; 1330 gene sets). We used the GSEA’s default statistical threshold of false discovery rate (FDR) <0.25.

For validation of RNA-seq data, quantitative real-time RT-PCR (RT-qPCR) was performed using RT2 SYBR Green ROX qPCR Mastermix (Qiagen) with adequate primers (Supplemental Table I) and analyzed by the ABI Prism 7900HT Sequence Detection System (Applied Biosystems, Grand Island, NY). All PCRs were in triplicate on 384-well plates, and mRNA expression relative to control β-actin was calculated using the 2−ΔΔCt method.

All statistical analyses were performed using GraphPad PRISM version 6.0 (GraphPad Software, La Jolla, CA). Data were represented as means ± SEM. A Student t test was used to calculate statistical significance between two groups. A two-tailed p value <0.05 was considered statistically significant.

RNA-seq was performed to examine differentially expressed genes in four different T cell populations (CD4+ naive, CD4+ memory, CD8+ naive, and CD8+ memory T cells) from three PNH patients (patients 1–3; Table I) and three healthy control subjects. Representative gating strategies for sorting of T cell subsets are shown in Fig. 1A. First, to confirm molecularly the identity of individual T cell subsets from PNH patients and control subjects, we subjected RNA-seq data to analysis of defining lineage markers (CD4 and CD8), CCR5 (a chemokine receptor predominantly expressed in memory T cells), and EOMES (a transcriptional factor preferentially expressed in effector and memory CD8+ T cells) (Fig. 1B) (2527). CD4 or CD8 expression was detected exclusively on CD4+ or CD8+ T cells, respectively. Significantly higher CCR5 expression was observed in CD4+ and CD8+ memory T cell subsets compared with corresponding TN subsets, respectively. EOMES expression was significantly higher in CD8+ memory T cells than CD8+ TNs.

Differentially expressed gene analysis of four T cell subsets from PNH and healthy control groups showed distinct gene expression signatures in individual T cell subsets (Fig. 2A, 2B, Supplemental Table II). In CD4+ TNs, 11 gene expression levels were significantly different: 5 upregulated (including SRRM2 and TNFSF8) and 6 downregulated genes (including GIMAP6) (>2-fold change, FDR < 0.05; Fig. 2A, 2B). In CD4+ memory T cells, 25 gene expression levels were significantly different: 15 upregulated (including JUND and TOB1) and 10 downregulated genes (including GIMAP4) (>2-fold change, FDR < 0.05; Fig. 2A, 2B). In CD8+ TNs, only two gene expression levels were significantly different: upregulated CTSW and downregulated RPL9 (>2-fold change, FDR < 0.05; Fig. 2A, 2B). In CD8+ memory T cells, seven gene expression levels were significantly different: two upregulated (CTSW and DPP4) and five downregulated genes (including SLC12A7) (>2-fold change, FDR < 0.05; Fig. 2A, 2B). Venn diagrams summarizing the differentially expressed gene analysis in the four T cells subsets from PNH patients compared with healthy control subjects is shown in Fig. 2C.

FIGURE 2.

RNA-seq in T cell subsets from PNH patients. (A) Expression levels of mRNAs illustrated in scatterplots for individual T cell subsets (CD4+ naive, CD4+ memory, CD8+ naive, and CD8+ memory T cells). Each point corresponds to one NCBI Reference Sequence (RefSeq) transcript with log10 FPKM values for T cell subsets from PNH patients and healthy control subjects. Log2 (Y/X) ≥ 1 and posterior probability of being equally expressed ≤0.05 were used to identify differentially expressed RNAs in PNH. Yellow and blue points correspond to upregulated and downregulated genes, respectively. Some of the upregulated or downregulated genes are indicated by arrows. (B) Heat map analysis and hierarchical clustering of mRNAs dysregulated in PNH patients compared with healthy control subjects for each T cell subset (CD4+ naive, CD4+ memory, CD8+ naive, or CD8+ memory T cells). A red-blue color scale depicts RNA expression levels (red indicates high; blue indicates low). (C) Venn diagram of differentially expressed genes (PNH versus healthy controls) in each T cell subset. (D) Heat map analysis and hierarchical clustering of 55 mRNAs dysregulated in PNH compared with healthy controls. A red-blue color scale depicts RNA expression levels (red indicates high; blue indicates low).

FIGURE 2.

RNA-seq in T cell subsets from PNH patients. (A) Expression levels of mRNAs illustrated in scatterplots for individual T cell subsets (CD4+ naive, CD4+ memory, CD8+ naive, and CD8+ memory T cells). Each point corresponds to one NCBI Reference Sequence (RefSeq) transcript with log10 FPKM values for T cell subsets from PNH patients and healthy control subjects. Log2 (Y/X) ≥ 1 and posterior probability of being equally expressed ≤0.05 were used to identify differentially expressed RNAs in PNH. Yellow and blue points correspond to upregulated and downregulated genes, respectively. Some of the upregulated or downregulated genes are indicated by arrows. (B) Heat map analysis and hierarchical clustering of mRNAs dysregulated in PNH patients compared with healthy control subjects for each T cell subset (CD4+ naive, CD4+ memory, CD8+ naive, or CD8+ memory T cells). A red-blue color scale depicts RNA expression levels (red indicates high; blue indicates low). (C) Venn diagram of differentially expressed genes (PNH versus healthy controls) in each T cell subset. (D) Heat map analysis and hierarchical clustering of 55 mRNAs dysregulated in PNH compared with healthy controls. A red-blue color scale depicts RNA expression levels (red indicates high; blue indicates low).

Close modal

Further, differentially expressed gene analysis was performed by combining CD4+ naive, CD4+ memory, CD8+ naive, and CD8+ memory T cells from PNH patients or healthy control subjects, respectively, to examine whether there were any differential expression changes in T cells from PNH that overcame the difference of cell lineage and functions. Of 55 gene expression levels that were significantly different, 41 were upregulated (including TNFAIP3, JUN, JUND, TOB1, TNFSF8, and CD69) and 14 downregulated (including GIMAP4) (>2-fold change, FDR < 0.05; Fig. 2D; Supplemental Table II). A heat map was generated to visualize differences in the 55 gene expression levels for each T cell subset (Fig. 2D).

First, we used IPA software to analyze downstream effects of the genes on biological functions, molecular networks, and regulatory pathways. By canonical pathway analysis, putative gene network interactions of differentially expressed genes were significantly enriched for canonical pathways of TNFR1, TNFR2, IL-17A, and CD27 signaling (p < 0.05; Fig. 3A). By network analysis, the top network functions were involved in cellular development, cellular growth and proliferation, and hematological system development and function, and involved in upregulated and downregulated genes such as TNFAIP3, JUN, JUND, and TOB1 (17 genes; Fig. 3B). These genes are directly or indirectly downstream of TCR.

FIGURE 3.

Pathway analysis of RNA-seq data. (A) Top canonical pathways identified by IPA comparing the differentially expressed genes between T cell subsets from PNH patients and healthy control subjects. (B) Putative gene network interactions with the dysregulated mRNAs by pathway analysis in PNH. Dysregulated gene networks in T cell subsets from PNH patients compared with healthy control subjects are shown. Color intensity indicates upregulation (red) and downregulation (green). Solid and dotted lines represent direct and indirect relationships between genes, respectively. (C) GSEA revealed the IGF1 pathway, Pre-NOTCH Expression and Processing, AP-1 pathway, and ATF2 pathway as the most extensively upregulated pathways in CD4+ naive, CD4+ memory, CD8+ naive, and CD8+ memory T cells from PNH patients, respectively. FDRq, FDR adjusted q value; NES, normalized enrichment score.

FIGURE 3.

Pathway analysis of RNA-seq data. (A) Top canonical pathways identified by IPA comparing the differentially expressed genes between T cell subsets from PNH patients and healthy control subjects. (B) Putative gene network interactions with the dysregulated mRNAs by pathway analysis in PNH. Dysregulated gene networks in T cell subsets from PNH patients compared with healthy control subjects are shown. Color intensity indicates upregulation (red) and downregulation (green). Solid and dotted lines represent direct and indirect relationships between genes, respectively. (C) GSEA revealed the IGF1 pathway, Pre-NOTCH Expression and Processing, AP-1 pathway, and ATF2 pathway as the most extensively upregulated pathways in CD4+ naive, CD4+ memory, CD8+ naive, and CD8+ memory T cells from PNH patients, respectively. FDRq, FDR adjusted q value; NES, normalized enrichment score.

Close modal

We then annotated functionality for our RNA-seq data by performing GSEA on curated gene sets in the Molecular Signatures Database to gain insights of the molecular pathways underlying the biologic processes in each T cell subset from PNH patients. The most significantly upregulated gene sets in CD4+ TNs from PNH patients displayed gene signatures associated with the “IGF1 pathway” (Fig. 3C). The most significantly upregulated gene sets in CD4+ memory, CD8+ naive, and CD8+ memory T cells from PNH patients displayed gene signatures related to the “Pre-NOTCH expression and processing,” “AP-1 pathway,” and “ATF2 pathway,” respectively (Fig. 3C). GSEA revealed significantly enriched gene sets in the C2 collections, such as P53 downstream and the HIF pathways (Table II). Together, our RNA-seq and pathway analyses indicated that T cells in PNH were distinct from their normal counterparts, and that each T cell subset had molecularly and functionally distinct characteristics.

Table II.
Summary of the GSEA-identified gene sets
Gene SetsNESFDR
CD4+ TNs   
 Up   
  BIOCARTA_IGF1_PATHWAY 2.18 0.009 
  BIOCARTA_TOLL_PATHWAY 2.15 0.014 
  PID_FCER1_PATHWAY 2.07 0.017 
  PID_CD40_PATHWAY 1.95 0.017 
  PID_PDGFRA_PATHWAY 2.02 0.018 
 Down   
  REACTOME_PEPTIDE_CHAIN_ELONGATION −3.28 
  KEGG_RIBOSOME −3.27 
  REACTOME_INFLUENZA_VIRAL_RNA_TRANSCRIPTION_AND_REPLICATION −3.21 
  REACTOME_SRP_DEPENDENT_COTRANSLATIONAL_PROTEIN_TARGETING_TO_MEMBRANE −3.16 
  REACTOME_3_UTR_MEDIATED_TRANSLATIONAL_REGULATION −3.16 
CD4+ memory T cells   
 Up   
  REACTOME_PRE_NOTCH_EXPRESSION_AND_PROCESSING 2.22 0.001 
  PID_ATF2_PATHWAY 2.05 0.013 
  PID_LYSOPHOSPHOLIPID_PATHWAY 2.10 0.014 
  BIOCARTA_IL1R_PATHWAY 2.06 0.017 
  BIOCARTA_TNFR2_PATHWAY 2.01 0.021 
 Down   
  REACTOME_PEPTIDE_CHAIN_ELONGATION −2.98 
  REACTOME_SRP_DEPENDENT_COTRANSLATIONAL_PROTEIN_TARGETING_TO_MEMBRANE −2.96 
  KEGG_RIBOSOME −2.94 
  REACTOME_3_UTR_MEDIATED_TRANSLATIONAL_REGULATION −2.86 
  REACTOME_TRANSLATION −2.79 
CD8+ TNs   
 Up   
  PID_AP1_PATHWAY 2.09 0.008 
  PID_FGF_PATHWAY 2.10 0.014 
  PID_ATF2_PATHWAY 1.99 0.031 
  PID_PDGFRA_PATHWAY 1.91 0.041 
  PID_NFAT_TFPATHWAY 1.89 0.041 
 Down   
  KEGG_RIBOSOME −3.27 
  REACTOME_PEPTIDE_CHAIN_ELONGATION −3.22 
  REACTOME_3_UTR_MEDIATED_TRANSLATIONAL_REGULATION −3.13 
  REACTOME_SRP_DEPENDENT_COTRANSLATIONAL_PROTEIN_TARGETING_TO_MEMBRANE −3.12 
  REACTOME_INFLUENZA_VIRAL_RNA_TRANSCRIPTION_AND_REPLICATION −3.10 
CD8+ memory T cells   
 Up   
  PID_ATF2_PATHWAY 2.22 0.001 
  PID_P53_DOWNSTREAM_PATHWAY 2.15 0.002 
  PID_AP1_PATHWAY 2.00 0.009 
  PID_HIF1_TFPATHWAY 2.00 0.012 
  KEGG_SYSTEMIC_LUPUS_ERYTHEMATOSUS 1.93 0.023 
 Down   
  REACTOME_GENERATION_OF_SECOND_MESSENGER_MOLECULES −2.03 0.017 
  KEGG_PEROXISOME −2.06 0.027 
  REACTOME_INTEGRIN_CELL_SURFACE_INTERACTIONS −1.77 0.134 
  KEGG_AMINOACYL_TRNA_BIOSYNTHESIS −1.83 0.136 
  REACTOME_IMMUNOREGULATORY_INTERACTIONS_BETWEEN_A_LYMPHOID_AND_A_ −1.73 0.145 
  NON_LYMPHOID_CELL   
Gene SetsNESFDR
CD4+ TNs   
 Up   
  BIOCARTA_IGF1_PATHWAY 2.18 0.009 
  BIOCARTA_TOLL_PATHWAY 2.15 0.014 
  PID_FCER1_PATHWAY 2.07 0.017 
  PID_CD40_PATHWAY 1.95 0.017 
  PID_PDGFRA_PATHWAY 2.02 0.018 
 Down   
  REACTOME_PEPTIDE_CHAIN_ELONGATION −3.28 
  KEGG_RIBOSOME −3.27 
  REACTOME_INFLUENZA_VIRAL_RNA_TRANSCRIPTION_AND_REPLICATION −3.21 
  REACTOME_SRP_DEPENDENT_COTRANSLATIONAL_PROTEIN_TARGETING_TO_MEMBRANE −3.16 
  REACTOME_3_UTR_MEDIATED_TRANSLATIONAL_REGULATION −3.16 
CD4+ memory T cells   
 Up   
  REACTOME_PRE_NOTCH_EXPRESSION_AND_PROCESSING 2.22 0.001 
  PID_ATF2_PATHWAY 2.05 0.013 
  PID_LYSOPHOSPHOLIPID_PATHWAY 2.10 0.014 
  BIOCARTA_IL1R_PATHWAY 2.06 0.017 
  BIOCARTA_TNFR2_PATHWAY 2.01 0.021 
 Down   
  REACTOME_PEPTIDE_CHAIN_ELONGATION −2.98 
  REACTOME_SRP_DEPENDENT_COTRANSLATIONAL_PROTEIN_TARGETING_TO_MEMBRANE −2.96 
  KEGG_RIBOSOME −2.94 
  REACTOME_3_UTR_MEDIATED_TRANSLATIONAL_REGULATION −2.86 
  REACTOME_TRANSLATION −2.79 
CD8+ TNs   
 Up   
  PID_AP1_PATHWAY 2.09 0.008 
  PID_FGF_PATHWAY 2.10 0.014 
  PID_ATF2_PATHWAY 1.99 0.031 
  PID_PDGFRA_PATHWAY 1.91 0.041 
  PID_NFAT_TFPATHWAY 1.89 0.041 
 Down   
  KEGG_RIBOSOME −3.27 
  REACTOME_PEPTIDE_CHAIN_ELONGATION −3.22 
  REACTOME_3_UTR_MEDIATED_TRANSLATIONAL_REGULATION −3.13 
  REACTOME_SRP_DEPENDENT_COTRANSLATIONAL_PROTEIN_TARGETING_TO_MEMBRANE −3.12 
  REACTOME_INFLUENZA_VIRAL_RNA_TRANSCRIPTION_AND_REPLICATION −3.10 
CD8+ memory T cells   
 Up   
  PID_ATF2_PATHWAY 2.22 0.001 
  PID_P53_DOWNSTREAM_PATHWAY 2.15 0.002 
  PID_AP1_PATHWAY 2.00 0.009 
  PID_HIF1_TFPATHWAY 2.00 0.012 
  KEGG_SYSTEMIC_LUPUS_ERYTHEMATOSUS 1.93 0.023 
 Down   
  REACTOME_GENERATION_OF_SECOND_MESSENGER_MOLECULES −2.03 0.017 
  KEGG_PEROXISOME −2.06 0.027 
  REACTOME_INTEGRIN_CELL_SURFACE_INTERACTIONS −1.77 0.134 
  KEGG_AMINOACYL_TRNA_BIOSYNTHESIS −1.83 0.136 
  REACTOME_IMMUNOREGULATORY_INTERACTIONS_BETWEEN_A_LYMPHOID_AND_A_ −1.73 0.145 
  NON_LYMPHOID_CELL   

C2 (canonical pathways) dataset was used for GSEA. Gene sets were sorted by FDR. Top five enriched gene sets are shown.

NES, normalized enrichment score.

In this study, for validation of the RNA-seq data, we chose 11 genes (TNFAIP3, JUN, JUND, TOB1, CTSW, GIMAP4, GIMAP6, TRMT112, NR4A2, TNFSF8, and CD69) for the following reasons: 1) these are the important mediators involved in positive or negative regulation for T cells; 2) dysregulation of these genes is associated with autoimmune diseases, including multiple sclerosis (MS), systemic lupus erythematosus (SLE), and inflammatory bowel disease (2830); 3) dysregulation of these genes mainly contributed to the abnormalities of the signaling pathway; and 4) expression levels (FPKM) of the selected seven genes detected by RNA-seq were higher than those of other candidate genes, facilitating validation.

Differential expression levels of TNFAP3, JUN, JUND, TOB1, CTSW, GIMAP4, GIMAP6, TRMT112, and NR4A2 were confirmed by RT-qPCR. Higher TNFAIP3 expression was confirmed in all four T cell subsets from PNH compared with heathy controls, and TNFAIP3 expression levels were especially high in CD8+ memory T cells (p < 0.05; Fig. 4A). Higher JUN expression was validated in CD4+ memory T cells and CD8+ memory T cells and TNs from PNH patients compared with heathy control subjects, showing especially high JUN expression in CD8+ memory T cells (p < 0.05; Fig. 4B). TOB1 expression was higher in both CD4+ TN and memory T cell subsets from PNH patients compared with healthy control subjects (p < 0.05; Fig. 4C), whereas TOB1 expression tended to be higher in CD8+ TN and memory T cell subsets from PNH patients, but without statistical significance. All four T cell subsets from PNH patients displayed tendencies for higher JUND expression than those from healthy control subjects, but without statistically significant differences (Fig. 4D). CTSW expression was not statistically significantly different between PNH patients and healthy control subjects in both CD4+ and CD8+ T cell subsets (Fig. 4E). We chose three downregulated genes (GIMAP4, GIMAP6, and TRMT112) for further validation of RNA-seq results by RT-qPCR. Expression of GIMAP4, GIMAP6, and TRMT112 was significantly downregulated in CD4+ T cell subsets from PNH patients compared with healthy control subjects (p < 0.05; Fig. 4F–H). NR4A2 (a member of the Notch signaling pathway) contributed to the enrichment of the pathway in CD4+ memory T cells from PNH patients by GSEA. We therefore validated NR4A2 expression level, confirming increased expression in CD4+ and CD8+ memory T cell subsets from PNH patients compared with heathy control subjects (p < 0.05; Fig. 4I).

FIGURE 4.

Validation of RNA-seq data. RT-qPCR analysis of gene expression (A) TNFAIP3, (B) JUN, (C) TOB1, (D) JUND, (E) CTSW, (F) GIMAP4, (G) GIMAP6, (H) TRMT112, and (I) NR4A2 in CD4+ and CD8+ T cell subsets from PNH patients (n = 7) and healthy control subjects (n = 7). Relative RNA expression of mRNA was calculated by normalizing to β-actin expression. Expression levels were shown as fold change compared with those of naive CD4+ T cells from healthy control subjects as a value of 1. (J) Protein expression levels of TNFAIP3 in T cell subsets from PNH patients and healthy control subjects were assessed by flow cytometry; representative histograms are shown. (K) Graphs represent corresponding quantitative analysis of TNFAIP3 MFI (geometric mean of fluorescence intensity) on T cells from PNH patients (n = 5) and healthy control subjects (n = 5) as described in (J). *p < 0.05, two-tailed Student t test.

FIGURE 4.

Validation of RNA-seq data. RT-qPCR analysis of gene expression (A) TNFAIP3, (B) JUN, (C) TOB1, (D) JUND, (E) CTSW, (F) GIMAP4, (G) GIMAP6, (H) TRMT112, and (I) NR4A2 in CD4+ and CD8+ T cell subsets from PNH patients (n = 7) and healthy control subjects (n = 7). Relative RNA expression of mRNA was calculated by normalizing to β-actin expression. Expression levels were shown as fold change compared with those of naive CD4+ T cells from healthy control subjects as a value of 1. (J) Protein expression levels of TNFAIP3 in T cell subsets from PNH patients and healthy control subjects were assessed by flow cytometry; representative histograms are shown. (K) Graphs represent corresponding quantitative analysis of TNFAIP3 MFI (geometric mean of fluorescence intensity) on T cells from PNH patients (n = 5) and healthy control subjects (n = 5) as described in (J). *p < 0.05, two-tailed Student t test.

Close modal

TNFAIP3 protein expression in T cell subsets from PNH patients was further assessed by flow cytometry, because transcript levels do not always correlate with protein expression levels, and this molecule plays a central role in infection and autoimmunity (28, 31). Protein expression of TNFAIP3 was significantly increased in CD4+ (TN, TCM, TEM, and TE) and CD8+ (TN, TCM, TEM, and TE) T cell subsets from PNH patients compared to healthy control subjects (Fig. 4J, 4K).

We further validated differential expression levels of CD69 and TNFSF8 (CD153) by flow cytometry, as these two molecules are cell surface markers associated with T cell activation (32, 33) and directly or indirectly affect downstream signaling (Fig. 3B). By flow cytometry, higher expression of CD69 was confirmed in CD4+ and CD8+ T cells from PNH patients compared with healthy control subjects (p < 0.05; Fig. 5A). Protein expression of CD69 was significantly induced following CD3/CD28 stimulation in T cells from both PNH patients and healthy control subjects. Of interest, T cells from PNH patients showed lower CD69 expression after CD3/CD28 stimulation, compared with those of healthy control subjects (p < 0.05; Fig. 5B).

FIGURE 5.

Expression of CD69 and TNFSF8 in T cells. (A) Cell surface expression levels of CD69 in T cells from PNH patients and healthy control subjects were assessed by flow cytometry (n = 8). Representative histograms are shown. Graphs represent corresponding quantitative analysis of CD69 MFI (geometric mean of fluorescence intensity) on T cells from PNH patients and healthy control subjects. (B) Cell surface expression levels of CD69 in T cells from PNH patients and healthy control subjects after CD3/CD28 stimulation. (C) Cell surface expression levels of TNFSF8 (CD153) in T cells from PNH patients and healthy control subjects were assessed by flow cytometry (n = 8). Representative histograms are shown. Graphs display corresponding quantitative analysis of TNFSF8 (CD153) MFI on T cells from PNH patients and healthy control subjects. (D) Cell surface expression levels of TNFSF8 in T cells from PNH patients and healthy control subjects after CD3/CD28 stimulation. *p < 0.05, two-tailed Student t test.

FIGURE 5.

Expression of CD69 and TNFSF8 in T cells. (A) Cell surface expression levels of CD69 in T cells from PNH patients and healthy control subjects were assessed by flow cytometry (n = 8). Representative histograms are shown. Graphs represent corresponding quantitative analysis of CD69 MFI (geometric mean of fluorescence intensity) on T cells from PNH patients and healthy control subjects. (B) Cell surface expression levels of CD69 in T cells from PNH patients and healthy control subjects after CD3/CD28 stimulation. (C) Cell surface expression levels of TNFSF8 (CD153) in T cells from PNH patients and healthy control subjects were assessed by flow cytometry (n = 8). Representative histograms are shown. Graphs display corresponding quantitative analysis of TNFSF8 (CD153) MFI on T cells from PNH patients and healthy control subjects. (D) Cell surface expression levels of TNFSF8 in T cells from PNH patients and healthy control subjects after CD3/CD28 stimulation. *p < 0.05, two-tailed Student t test.

Close modal

Higher expression of TNFSF8 was confirmed in CD4+ and CD8+ T cells from PNH patients (p < 0.05; Fig. 5C). CD4+ T cells from PNH patients showed higher TNFSF8 expression after CD3/CD28 stimulation, compared with those of healthy control subjects (p < 0.05; Fig. 5D).

By RNA-seq and flow cytometry, we verified that CD69 expression in T cells was higher in PNH patients than in healthy control subjects, a result consistent with recent activation of T cells by Ag. Recent studies suggest that chronic exposure to persistent Ags results in upregulation of PD-1 expression in T cells (34, 35). We therefore examined PD-1 expression of various T cell subsets from PNH patients by flow cytometry. When PD-1 expression in individual CD4+ T cell subsets was compared between PNH patients and healthy control subjects, significantly higher levels were seen on TEs in PNH patients (p < 0.05; Fig. 6A, 6B). In CD8+ T cell subsets, significantly higher PD-1 expression levels were seen on CD8+ TEM in PNH patients compared with healthy control subjects (p < 0.05; Fig. 6A, 6C). These results are similar to reports of PD-1 expression correlation with differentiation of CD4+ and CD8+ T cells (8, 22, 36).

FIGURE 6.

PD-1 expression in T cell subsets. (A) Representative graphs of PD-1 expression of CD4+ and CD8+ T cells in PNH patients and healthy control subjects. (B) PD-1 MFI (geometric mean of fluorescence intensity) on CD4+ and CD8+ T cells from PNH patients (n = 8) and heathy control subjects (n = 8). (C) PD-1 expression levels on CD4+ and CD8+ T cell subsets were analyzed in PNH patients (n = 8) and healthy control subjects (n = 8). Each T cell subset was defined as follows: TCM, ViViD CD3+ CD4 (CD8)+ CD45RO+ CCR7+; TEM, ViViD CD3+ CD4 (CD8)+ CD45RO+ CCR7; terminally differentiated TE, ViViD CD3+ CD4 (CD8)+ CD45RO CD45RA+ CCR7 CD27; and TNs, CD3+ CD4 (CD8)+ CD45RO CD45RA+ CCR7+ CD27+ CD95. *p < 0.05, two-tailed Student t test.

FIGURE 6.

PD-1 expression in T cell subsets. (A) Representative graphs of PD-1 expression of CD4+ and CD8+ T cells in PNH patients and healthy control subjects. (B) PD-1 MFI (geometric mean of fluorescence intensity) on CD4+ and CD8+ T cells from PNH patients (n = 8) and heathy control subjects (n = 8). (C) PD-1 expression levels on CD4+ and CD8+ T cell subsets were analyzed in PNH patients (n = 8) and healthy control subjects (n = 8). Each T cell subset was defined as follows: TCM, ViViD CD3+ CD4 (CD8)+ CD45RO+ CCR7+; TEM, ViViD CD3+ CD4 (CD8)+ CD45RO+ CCR7; terminally differentiated TE, ViViD CD3+ CD4 (CD8)+ CD45RO CD45RA+ CCR7 CD27; and TNs, CD3+ CD4 (CD8)+ CD45RO CD45RA+ CCR7+ CD27+ CD95. *p < 0.05, two-tailed Student t test.

Close modal

In this study, RNA-seq was used to examine the transcriptome in T cell subsets sorted by FACS-based strategy, resulting in identification of a novel molecular mechanism that underlays the aberrant T cell immune status in PNH. Our data demonstrate the feasibility of isolation and molecular characterization of purified primary T cell subsets in PNH patients and healthy control subjects by massive parallel transcriptome sequencing. To date, no study has used RNA-seq to analyze the transcriptome of immune cells in PNH. In other autoimmune diseases, this novel method has yielded insights (3739). In SLE T cells, transcripts for hundreds of genes are consistently altered, for which pathway analysis highlights induction of pathways related to mitochondria, nucleotide metabolism, and DNA replication (38). Transcriptome analysis of CD4+ T cells in celiac disease shows evidence that BACH2 plays an important role in pathogenesis (39). A study using transcriptional profiling of purified CD8+ T cells from patients with anti-neutrophil cytoplasmic Ab–associated vasculitis and SLE suggests that the subset of genes defining the poor prognostic group is enriched for genes involved in the IL-7R pathway and TCR signaling, and those expressed by CD8+ memory T cells (40).

In our experiments, canonical pathway analysis revealed TNFR1 and TNFR2 signaling to be most significantly affected in PNH. Association of TNFR signaling in PNH was further supported by GSEA. TNFR signaling mediates activation of the transcription factor AP-1 (30). ATF2 is one of the members of the ATF/CREB family of bZIP transcription factors that contribute to cellular responses to stresses, such as DNA damage (41). Like JUN and FOS, ATF2 is also characterized by a basic structural region and a leucine zipper domain that are crucial for AP-1 homodimerization and heterodimerization (42). TNFR signaling is dysregulated in various autoimmune diseases. Indeed, elevated TNF levels have been present in rheumatoid arthritis (RA), psoriasis, inflammatory bowel disease, and MS (43, 44). Inhibitors of TNF are clinically efficacious at reducing inflammation associated with autoimmune diseases (45). TNFR signaling is also dysregulated in other BM failure diseases, such as AA and myelodysplastic syndrome (46, 47). T cells overexpressing IFN-γ and TNF-α are concentrated in the circulating or BM T cells of AA (7, 48), which would in turn suppress hematopoiesis (49). TNF-α and IFN-γ are also overexpressed in the BM mononuclear cells of Fanconi anemia (congenital BM failure) patients, and TNF-α suppresses erythropoiesis in vitro (50).

By GSEA, each T cell subset showed dysregulated signaling pathways. The IGF1/IGF1R pathway regulates cell cycle progression, apoptosis, and the translation of proteins, and abnormalities in this pathway have been reported with autoimmunity such as in Graves’ disease, Crohn’s disease, and RA (51).Within this pathway, JUN, FOS, IGF1R, PIK3CA, PIK3R1, and SOS1 contributed to the enrichment. IGF-1 inhibits apoptosis of immature CD45RA+ and mature CD45RO+ T cells. IGF-1 also promotes the transition of cord blood CD45RA+ T cells to a CD45RO phenotype by increasing RA to RO conversion after Ag stimulation (52). Increased IGF1 pathway signaling in CD4+ TNs might reflect promotion of naive cells to memory phenotype after Ag stimulation in PNH patients. The Notch signaling pathway was one of the most significantly enriched gene sets in CD4+ memory T cells from PNH patients. A recent study identified NOTCH signaling as a primary driver of Th1-mediated pathogenesis in AA, and NOTCH may represent a target for therapeutic intervention in BMF mouse models (53). Several genes (including FURIN, KAT2B, CREBBP, and NR4A2) contributed to the enrichment of gene sets identified by GSEA: especially, upregulation of NR4A2 was validated in CD4+ memory T cells by RT-qPCR. NR4A2 (an orphan nuclear receptor with an unknown ligand) is associated with T cell subset communication (54) and induced by the macrophage inflammatory response (55). Increased NR4A2 expression is seen in psoriasis and decreases after anti-TNF treatment (56). Upregulation of NR4A2 expression is observed in MS (57, 58), and enforced expression augments promoter activities of IL-17 and IFN-γ genes, leading to excessive production of these cytokines (58). NR4A2 is upregulated in both chronic infection and tumor-induced T cell dysfunction (5961), but it remains to be determined whether its role is central. Increased expression of NR4A2 in PNH patients might reflect abnormal T cell responses to Ags. In CD8+ naive or CD8+ memory T cells, the AP-1 pathway or the ATF2 pathway was one of the most significantly upregulated pathways in PNH patients, respectively. Of the AP-1 pathway, FOSL2, JUN, DUSP1, FOS, JUND, JUNB, BCL2L11, TGFB1, CDKN1B, ELF1, EP300, NFATC2, BAG1, and MT2A were significant contributors to enrichment. In the ATF2 pathway, DUSP5, DUSP1, FOS, DUSP10, JUN, SOCS3, JUNB, JUND, and DDIT3 were identified as mainly contributing to enrichment. Upregulation of JUN was validated in CD8+ TNs and memory T cells by qPCR. Therefore, specific dysregulation of T cell intracellular signaling may contribute to BM failure and the inflammatory environment in PNH (62).

To validate our RNA-seq data, we selected 11 candidate genes that are important mediators in regulation for T cells and are dysregulated in other autoimmune diseases. Higher or lower expressions of nine genes and proteins were validated in T cells from PNH by RT-qPCR (JUN, TNFAIP3, TOB1, GIMAP4, GIMAP6, TRMT112, and NR4A2) and flow cytometry (CD69 and TNFSF8). Discordance between RNA-seq results and qPCR data in JUND, CTSW, and TOB1 in CD8+ T cells may be caused by variation in expression levels in individual patients and the limited number of samples examined. JUN belongs to the AP-1 transcription family, which regulates nuclear gene expression associated with T cell activation or effector functions (30). High expression levels of JUN were evident, especially in memory T cells, by RT-qPCR, and these results were comparable with RNA-seq results (Fig. 2D). JUN is an “immediate early gene” and is responsive to mitogenic stimuli, as well as DNA damage and stress (42). JUN expression levels are tightly controlled by a combination of protein stability and a short mRNA half-life. JUN transcription is induced by SP1, NF-κB, ternary complex factors, MEF2, or CCAAT-binding transcription factors (63). CD69 is an early activation marker of lymphocytes at sites of chronic inflammation (64), and it also plays an essential role in regulation of inflammation (33). Expression of CD69 can be induced in vitro by a wide variety of agents, such as anti-CD3/TCR, anti-CD2 mAb, activators of protein kinase C, and PHA (65). Soon after stimulation of T cells through the TCR, CD69 mRNA levels are transiently increased. Inducible CD69 expression by mitogenic signals is regulated by the transcription factor AP-1 (66). In this work, CD69 expression in T cells was higher in PNH patients than in healthy control subjects, which may be explained by higher AP-1 activity in T cells from PNH patients. The higher expression levels of CD69 in T cells from PNH patients have been reported by other groups using flow cytometry, consistent with the aberrant immunity in PNH associated with BM failure (67). The inability to upregulate CD69 after TCR activation may lead to T cell dysfunction in PNH patients (68). The PD-1 pathway exerts critical inhibitory functions in the setting of persistent antigenic stimulation, as well as during encounter of self-antigens, in chronic viral infections, and in response to tumor Ags (34, 35). Overexpressed genes in CD8+ T cells from lymphocytic choriomeningitis virus–infected mice (59) and tumor-infiltrating CD8+ T cells of an autochthonous mouse melanoma (60) show a characteristic signature associated with T cell dysfunction, including increased expression of PD-1 and/or CD69 (61). Upregulation of both CD69 and PD-1 in T cells therefore might reflect acute or chronic antigenic stimulation in PNH. TNFSF8 (CD153), the ligand for CD30 (CD30L), a membrane-associated glycoprotein related to TNF, is expressed mainly by activated T cells (69). TNFSF8 belongs to the TNF family as a costimulatory signaling molecule for T cells, and the TNFSF8–CD30 interaction costimulates T cell activation and proliferation (32). Collectively, our results and the work of others support a molecular mechanism of antigenic-driven activation of T cells in PNH.

We could successfully validate the increased gene and protein expression levels of TNFAIP3 in T cell subsets from PNH patients. TNFAIP3 is highly expressed in CD8+ T cells during chronic infection and in tumors (5961), indicating some commonalities with infection and cancer. TNFAIP3 encodes a protein (A20) that attenuates TNF-induced NF-κB activation (70), and cell type–specific TNFAIP3 deficiency is associated with multiple autoimmune diseases (28, 7173). Heterozygous germline mutations in TNFAIP3 leading to A20 haploinsufficiency have been reported in families with early-onset systemic autoinflammatory disease (71). In myeloid-specific A20-deficient mice, specific ablation of Tnfaip3 in myeloid cells results in spontaneous development of a severe destructive polyarthritis with features of RA (72). TOB1 in the Tob/BTG antiproliferative (APRO) family plays a critical role in adaptive T cell immune responses in the development of experimental autoimmune encephalomyelitis (29). Expression of TNFAIP3 is increased in PBMCs from RA patients. TNFAIP3 is rapidly induced by TNF, and the expression of TNF is known to be markedly elevated in active RA (74). These reports and our data suggest that negative regulators, TNFAIP3 and TOB1, may attenuate an aberrant T cell activation in response to Ag.

Our study had some limitations, such as the relatively small number of patient samples and limited number of validated genes. However, validating our RNA-seq data for several upregulated and downregulated genes supported the reliability of our RNA-seq results. The roles of GIMAPs have been studied in apoptotic functions (75) and in Th cell differentiation (76, 77). Downregulation of GIMAP4 and GIMAP6 in CD4+ T cells might have an antiapoptotic role in T cells from PNH patients. TRMT112 is tRNA methyltransferase, but its role in T cells is unknown. Another limitation in this study is that the majority of the patients had a prior history of IST, for antecedent AA. However, the duration of prior IST from sampling date was quite long, and ISTs were not being administered to any patient at the time of sampling. GPI-AP T cells were not examined because of the low frequency of this population (78). Because our T cells were unstimulated, it is possible that we failed to detect difference in gene expression apparent only on activation. Nonetheless, we have demonstrated molecular signatures associated with positive and negative regulators in T cells, suggesting novel pathophysiologies in PNH. Although directly addressing mechanisms was difficult in our study, molecular signatures in T cells from PNH patients were associated with T cell dysfunction (such as increased expression of PD-1, CD69, NR4A2, and TNFAIP3) (61), as in chronic infection and cancer. We expect that elucidation of the transcriptome of highly purified human T cell subsets in PNH will be a valuable resource, informing the validation and discovery of novel pathophysiologic factors and therapeutic targets in other autoimmune diseases.

In conclusion, using RNA-seq to examine the transcriptome of T cell subsets from PNH, we identified novel pathways associated with disease. Our results implicate some key molecules and pathways involved in regulation of T cells in PNH. Understanding these pathways may provide new therapeutic strategies to modulate T cell immune responses in BM failure.

We thank Olga Rios, Barbara Weinstein, and Kinneret Broder for assistance in obtaining patient and healthy volunteer samples, and Marie J. Desierto and Maria del Pilar Fernandez Ibanez for technical assistance. We thank Honglan Gou and Qiongzhi He from BGI for help with conducting the experiments and analysis for RNA-seq.

This work was supported by the Intramural Research Program of the National Institutes of Health, National Heart, Lung, and Blood Institute and The Aplastic Anemia and MDS International Foundation (to K.H.).

The RNA sequencing data presented in this article have been submitted to the National Center for Biotechnology Information’s Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/) under accession number GSE83808.

The online version of this article contains supplemental material.

Abbreviations used in this article:

AA

aplastic anemia

ATG

anti-thymocyte globulin

BM

bone marrow

CsA

cyclosporine

FDR

false discovery rate

FPKM

fragment per kilobase of transcript per million mapped reads

GPI-AP

GPI-anchored protein

GSEA

gene set enrichment analysis

HSC

hematopoietic stem cell

IPA

Ingenuity Pathway Analysis

IST

immunosuppressive therapy

MS

multiple sclerosis

PNH

paroxysmal nocturnal hemoglobinuria

qPCR

quantitative PCR

RA

rheumatoid arthritis

RNA-seq

RNA sequencing

RT-qPCR

quantitative real-time RT-PCR

SLE

systemic lupus erythematosus

TCM

central memory T cell

TE

effector T cell

TEM

effector memory T cell

TN

naive T cell.

1
Young
,
N. S.
,
G.
Meyers
,
H.
Schrezenmeier
,
P.
Hillmen
,
A.
Hill
.
2009
.
The management of paroxysmal nocturnal hemoglobinuria: recent advances in diagnosis and treatment and new hope for patients.
Semin. Hematol.
46
(
1
Suppl. 1
):
S1
S16
.
2
Rosse
,
W. F.
1996
.
Epidemiology of PNH.
Lancet
348
:
560
.
3
Takeda
,
J.
,
T.
Miyata
,
K.
Kawagoe
,
Y.
Iida
,
Y.
Endo
,
T.
Fujita
,
M.
Takahashi
,
T.
Kitani
,
T.
Kinoshita
.
1993
.
Deficiency of the GPI anchor caused by a somatic mutation of the PIG-A gene in paroxysmal nocturnal hemoglobinuria.
Cell
73
:
703
711
.
4
Young
,
N. S.
,
R. T.
Calado
,
P.
Scheinberg
.
2006
.
Current concepts in the pathophysiology and treatment of aplastic anemia.
Blood
108
:
2509
2519
.
5
Risitano
,
A. M.
,
H.
Kook
,
W.
Zeng
,
G.
Chen
,
N. S.
Young
,
J. P.
Maciejewski
.
2002
.
Oligoclonal and polyclonal CD4 and CD8 lymphocytes in aplastic anemia and paroxysmal nocturnal hemoglobinuria measured by V beta CDR3 spectratyping and flow cytometry.
Blood
100
:
178
183
.
6
Nakao
,
S.
,
A.
Takami
,
H.
Takamatsu
,
W.
Zeng
,
N.
Sugimori
,
H.
Yamazaki
,
Y.
Miura
,
M.
Ueda
,
S.
Shiobara
,
T.
Yoshioka
, et al
.
1997
.
Isolation of a T-cell clone showing HLA-DRB1*0405-restricted cytotoxicity for hematopoietic cells in a patient with aplastic anemia.
Blood
89
:
3691
3699
.
7
Sloand
,
E.
,
S.
Kim
,
J. P.
Maciejewski
,
J.
Tisdale
,
D.
Follmann
,
N. S.
Young
.
2002
.
Intracellular interferon-gamma in circulating and marrow T cells detected by flow cytometry and the response to immunosuppressive therapy in patients with aplastic anemia.
Blood
100
:
1185
1191
.
8
Hosokawa
,
K.
,
P.
Muranski
,
X.
Feng
,
D. M.
Townsley
,
B.
Liu
,
J.
Knickelbein
,
K.
Keyvanfar
,
B.
Dumitriu
,
S.
Ito
,
S.
Kajigaya
, et al
.
2016
.
Memory stem T cells in autoimmune disease: high frequency of circulating CD8+ memory stem cells in acquired aplastic anemia.
J. Immunol.
196
:
1568
1578
.
9
Scheinberg
,
P.
,
O.
Nunez
,
B.
Weinstein
,
P.
Scheinberg
,
A.
Biancotto
,
C. O.
Wu
,
N. S.
Young
.
2011
.
Horse versus rabbit antithymocyte globulin in acquired aplastic anemia.
N. Engl. J. Med.
365
:
430
438
.
10
Scheinberg
,
P.
,
N. S.
Young
.
2012
.
How I treat acquired aplastic anemia.
Blood
120
:
1185
1196
.
11
Risitano
,
A. M.
,
J. P.
Maciejewski
,
S.
Green
,
M.
Plasilova
,
W.
Zeng
,
N. S.
Young
.
2004
.
In-vivo dominant immune responses in aplastic anaemia: molecular tracking of putatively pathogenetic T-cell clones by TCR beta-CDR3 sequencing.
Lancet
364
:
355
364
.
12
Young
,
N. S.
,
J. P.
Maciejewski
,
E.
Sloand
,
G.
Chen
,
W.
Zeng
,
A.
Risitano
,
A.
Miyazato
.
2002
.
The relationship of aplastic anemia and PNH.
Int. J. Hematol.
76
(
Suppl. 2
):
168
172
.
13
Yoshizato
,
T.
,
B.
Dumitriu
,
K.
Hosokawa
,
H.
Makishima
,
K.
Yoshida
,
D.
Townsley
,
A.
Sato-Otsubo
,
Y.
Sato
,
D.
Liu
,
H.
Suzuki
, et al
.
2015
.
Somatic mutations and clonal hematopoiesis in aplastic anemia.
N. Engl. J. Med.
373
:
35
47
.
14
Dunn
,
D. E.
,
P.
Tanawattanacharoen
,
P.
Boccuni
,
S.
Nagakura
,
S. W.
Green
,
M. R.
Kirby
,
M. S.
Kumar
,
S.
Rosenfeld
,
N. S.
Young
.
1999
.
Paroxysmal nocturnal hemoglobinuria cells in patients with bone marrow failure syndromes.
Ann. Intern. Med.
131
:
401
408
.
15
Kulagin
,
A.
,
I.
Lisukov
,
M.
Ivanova
,
I.
Golubovskaya
,
I.
Kruchkova
,
S.
Bondarenko
,
V.
Vavilov
,
N.
Stancheva
,
E.
Babenko
,
A.
Sipol
, et al
.
2014
.
Prognostic value of paroxysmal nocturnal haemoglobinuria clone presence in aplastic anaemia patients treated with combined immunosuppression: results of two-centre prospective study.
Br. J. Haematol.
164
:
546
554
.
16
Karadimitris
,
A.
,
J. S.
Manavalan
,
H. T.
Thaler
,
R.
Notaro
,
D. J.
Araten
,
K.
Nafa
,
I. A.
Roberts
,
M. E.
Weksler
,
L.
Luzzatto
.
2000
.
Abnormal T-cell repertoire is consistent with immune process underlying the pathogenesis of paroxysmal nocturnal hemoglobinuria.
Blood
96
:
2613
2620
.
17
Gargiulo
,
L.
,
M.
Papaioannou
,
M.
Sica
,
G.
Talini
,
A.
Chaidos
,
B.
Richichi
,
A. V.
Nikolaev
,
C.
Nativi
,
M.
Layton
,
J.
de la Fuente
, et al
.
2013
.
Glycosylphosphatidylinositol-specific, CD1d-restricted T cells in paroxysmal nocturnal hemoglobinuria.
Blood
121
:
2753
2761
.
18
Terrazzano
,
G.
,
M.
Sica
,
C.
Becchimanzi
,
S.
Costantini
,
B.
Rotoli
,
S.
Zappacosta
,
F.
Alfinito
,
G.
Ruggiero
.
2005
.
T cells from paroxysmal nocturnal haemoglobinuria (PNH) patients show an altered CD40-dependent pathway.
J. Leukoc. Biol.
78
:
27
36
.
19
van Bijnen
,
S. T.
,
M.
Withaar
,
F.
Preijers
,
A.
van der Meer
,
T.
de Witte
,
P.
Muus
,
H.
Dolstra
.
2011
.
T cells expressing the activating NK-cell receptors KIR2DS4, NKG2C and NKG2D are elevated in paroxysmal nocturnal hemoglobinuria and cytotoxic toward hematopoietic progenitor cell lines.
Exp. Hematol.
39
:
751
762.e3
.
20
Wang
,
Z.
,
M.
Gerstein
,
M.
Snyder
.
2009
.
RNA-Seq: a revolutionary tool for transcriptomics.
Nat. Rev. Genet.
10
:
57
63
.
21
Garber
,
M.
,
M. G.
Grabherr
,
M.
Guttman
,
C.
Trapnell
.
2011
.
Computational methods for transcriptome annotation and quantification using RNA-seq.
Nat. Methods
8
:
469
477
.
22
Legat
,
A.
,
D. E.
Speiser
,
H.
Pircher
,
D.
Zehn
,
S. A.
Fuertes Marraco
.
2013
.
Inhibitory receptor expression depends more dominantly on differentiation and activation than “Exhaustion” of human CD8 T cells.
Front. Immunol.
4
:
455
.
23
Leng
,
N.
,
J. A.
Dawson
,
J. A.
Thomson
,
V.
Ruotti
,
A. I.
Rissman
,
B. M.
Smits
,
J. D.
Haag
,
M. N.
Gould
,
R. M.
Stewart
,
C.
Kendziorski
.
2013
.
EBSeq: an empirical Bayes hierarchical model for inference in RNA-seq experiments.
Bioinformatics
29
:
1035
1043
.
24
Subramanian
,
A.
,
P.
Tamayo
,
V. K.
Mootha
,
S.
Mukherjee
,
B. L.
Ebert
,
M. A.
Gillette
,
A.
Paulovich
,
S. L.
Pomeroy
,
T. R.
Golub
,
E. S.
Lander
,
J. P.
Mesirov
.
2005
.
Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles.
Proc. Natl. Acad. Sci. USA
102
:
15545
15550
.
25
Bleul
,
C. C.
,
L.
Wu
,
J. A.
Hoxie
,
T. A.
Springer
,
C. R.
Mackay
.
1997
.
The HIV coreceptors CXCR4 and CCR5 are differentially expressed and regulated on human T lymphocytes.
Proc. Natl. Acad. Sci. USA
94
:
1925
1930
.
26
Banerjee
,
A.
,
S. M.
Gordon
,
A. M.
Intlekofer
,
M. A.
Paley
,
E. C.
Mooney
,
T.
Lindsten
,
E. J.
Wherry
,
S. L.
Reiner
.
2010
.
Cutting edge: the transcription factor eomesodermin enables CD8+ T cells to compete for the memory cell niche.
J. Immunol.
185
:
4988
4992
.
27
Intlekofer
,
A. M.
,
N.
Takemoto
,
E. J.
Wherry
,
S. A.
Longworth
,
J. T.
Northrup
,
V. R.
Palanivel
,
A. C.
Mullen
,
C. R.
Gasink
,
S. M.
Kaech
,
J. D.
Miller
, et al
.
2005
.
Effector and memory CD8+ T cell fate coupled by T-bet and eomesodermin.
Nat. Immunol.
6
:
1236
1244
.
28
Catrysse
,
L.
,
L.
Vereecke
,
R.
Beyaert
,
G.
van Loo
.
2014
.
A20 in inflammation and autoimmunity.
Trends Immunol.
35
:
22
31
.
29
Schulze-Topphoff
,
U.
,
S.
Casazza
,
M.
Varrin-Doyer
,
K.
Pekarek
,
R. A.
Sobel
,
S. L.
Hauser
,
J. R.
Oksenberg
,
S. S.
Zamvil
,
S. E.
Baranzini
.
2013
.
Tob1 plays a critical role in the activation of encephalitogenic T cells in CNS autoimmunity.
J. Exp. Med.
210
:
1301
1309
.
30
Baud
,
V.
,
M.
Karin
.
2001
.
Signal transduction by tumor necrosis factor and its relatives.
Trends Cell Biol.
11
:
372
377
.
31
Coornaert
,
B.
,
I.
Carpentier
,
R.
Beyaert
.
2009
.
A20: central gatekeeper in inflammation and immunity.
J. Biol. Chem.
284
:
8217
8221
.
32
Blazar
,
B. R.
,
R. B.
Levy
,
T. W.
Mak
,
A.
Panoskaltsis-Mortari
,
H.
Muta
,
M.
Jones
,
M.
Roskos
,
J. S.
Serody
,
H.
Yagita
,
E. R.
Podack
,
P. A.
Taylor
.
2004
.
CD30/CD30 ligand (CD153) interaction regulates CD4+ T cell-mediated graft-versus-host disease.
J. Immunol.
173
:
2933
2941
.
33
Sancho
,
D.
,
M.
Gómez
,
F.
Sánchez-Madrid
.
2005
.
CD69 is an immunoregulatory molecule induced following activation.
Trends Immunol.
26
:
136
140
.
34
Francisco
,
L. M.
,
P. T.
Sage
,
A. H.
Sharpe
.
2010
.
The PD-1 pathway in tolerance and autoimmunity.
Immunol. Rev.
236
:
219
242
.
35
Sharpe
,
A. H.
,
E. J.
Wherry
,
R.
Ahmed
,
G. J.
Freeman
.
2007
.
The function of programmed cell death 1 and its ligands in regulating autoimmunity and infection.
Nat. Immunol.
8
:
239
245
.
36
Jaafoura
,
S.
,
M. G.
de Goër de Herve
,
E. A.
Hernandez-Vargas
,
H.
Hendel-Chavez
,
M.
Abdoh
,
M. C.
Mateo
,
R.
Krzysiek
,
M.
Merad
,
R.
Seng
,
M.
Tardieu
, et al
.
2014
.
Progressive contraction of the latent HIV reservoir around a core of less-differentiated CD4+ memory T Cells.
Nat. Commun.
5
:
5407
.
37
Hrdlickova
,
B.
,
V.
Kumar
,
K.
Kanduri
,
D. V.
Zhernakova
,
S.
Tripathi
,
J.
Karjalainen
,
R. J.
Lund
,
Y.
Li
,
U.
Ullah
,
R.
Modderman
, et al
.
2014
.
Expression profiles of long non-coding RNAs located in autoimmune disease-associated regions reveal immune cell-type specificity.
Genome Med.
6
:
88
.
38
Bradley
,
S. J.
,
A.
Suarez-Fueyo
,
D. R.
Moss
,
V. C.
Kyttaris
,
G. C.
Tsokos
.
2015
.
T cell transcriptomes describe patient subtypes in systemic lupus erythematosus.
PLoS One
10
:
e0141171
.
39
Quinn
,
E. M.
,
C.
Coleman
,
B.
Molloy
,
P.
Dominguez Castro
,
P.
Cormican
,
V.
Trimble
,
N.
Mahmud
,
R.
McManus
.
2015
.
Transcriptome analysis of CD4+ T cells in coeliac disease reveals imprint of BACH2 and IFNγ regulation.
PLoS One
10
:
e0140049
.
40
McKinney, E. F., P. A. Lyons, E. J. Carr, J. L. Hollis, D. R. Jayne, L. C. Willcocks, M. Koukoulaki, A. Brazma, V. Jovanovic, D. M. Kemeny, et al. 2010. A CD8+ T cell transcription signature predicts prognosis in autoimmune disease. Nat. Med. 16: 586–591, 1p following 591.
41
Bhoumik
,
A.
,
P.
Lopez-Bergami
,
Z.
Ronai
.
2007
.
ATF2 on the double - activating transcription factor and DNA damage response protein.
Pigment Cell Res.
20
:
498
506
.
42
Lopez-Bergami
,
P.
,
E.
Lau
,
Z.
Ronai
.
2010
.
Emerging roles of ATF2 and the dynamic AP1 network in cancer.
Nat. Rev. Cancer
10
:
65
76
.
43
Wong
,
M.
,
D.
Ziring
,
Y.
Korin
,
S.
Desai
,
S.
Kim
,
J.
Lin
,
D.
Gjertson
,
J.
Braun
,
E.
Reed
,
R. R.
Singh
.
2008
.
TNFalpha blockade in human diseases: mechanisms and future directions.
Clin. Immunol.
126
:
121
136
.
44
Kontermann
,
R. E.
,
P.
Scheurich
,
K.
Pfizenmaier
.
2009
.
Antagonists of TNF action: clinical experience and new developments.
Expert Opin. Drug Discov.
4
:
279
292
.
45
Croft
,
M.
,
C. A.
Benedict
,
C. F.
Ware
.
2013
.
Clinical targeting of the TNF and TNFR superfamilies.
Nat. Rev. Drug Discov.
12
:
147
168
.
46
Zeng
,
W.
,
S.
Kajigaya
,
G.
Chen
,
A. M.
Risitano
,
O.
Nunez
,
N. S.
Young
.
2004
.
Transcript profile of CD4+ and CD8+ T cells from the bone marrow of acquired aplastic anemia patients.
Exp. Hematol.
32
:
806
814
.
47
Gañán-Gómez
,
I.
,
Y.
Wei
,
D. T.
Starczynowski
,
S.
Colla
,
H.
Yang
,
M.
Cabrero-Calvo
,
Z. S.
Bohannan
,
A.
Verma
,
U.
Steidl
,
G.
Garcia-Manero
.
2015
.
Deregulation of innate immune and inflammatory signaling in myelodysplastic syndromes.
Leukemia
29
:
1458
1469
.
48
Dufour
,
C.
,
A.
Corcione
,
J.
Svahn
,
R.
Haupt
,
N.
Battilana
,
V.
Pistoia
.
2001
.
Interferon gamma and tumour necrosis factor alpha are overexpressed in bone marrow T lymphocytes from paediatric patients with aplastic anaemia.
Br. J. Haematol.
115
:
1023
1031
.
49
Selleri
,
C.
,
T.
Sato
,
S.
Anderson
,
N. S.
Young
,
J. P.
Maciejewski
.
1995
.
Interferon-gamma and tumor necrosis factor-alpha suppress both early and late stages of hematopoiesis and induce programmed cell death.
J. Cell. Physiol.
165
:
538
546
.
50
Dufour
,
C.
,
A.
Corcione
,
J.
Svahn
,
R.
Haupt
,
V.
Poggi
,
A. N.
Béka’ssy
,
R.
Scimè
,
A.
Pistorio
,
V.
Pistoia
.
2003
.
TNF-alpha and IFN-gamma are overexpressed in the bone marrow of Fanconi anemia patients and TNF-alpha suppresses erythropoiesis in vitro.
Blood
102
:
2053
2059
.
51
Smith
,
T. J.
2010
.
Insulin-like growth factor-I regulation of immune function: a potential therapeutic target in autoimmune diseases?
Pharmacol. Rev.
62
:
199
236
.
52
Tu
,
W.
,
P. T.
Cheung
,
Y. L.
Lau
.
2000
.
Insulin-like growth factor 1 promotes cord blood T cell maturation and inhibits its spontaneous and phytohemagglutinin-induced apoptosis through different mechanisms.
J. Immunol.
165
:
1331
1336
.
53
Roderick
,
J. E.
,
G.
Gonzalez-Perez
,
C. A.
Kuksin
,
A.
Dongre
,
E. R.
Roberts
,
J.
Srinivasan
,
C.
Andrzejewski
Jr.
,
A. H.
Fauq
,
T. E.
Golde
,
L.
Miele
,
L. M.
Minter
.
2013
.
Therapeutic targeting of NOTCH signaling ameliorates immune-mediated bone marrow failure of aplastic anemia.
J. Exp. Med.
210
:
1311
1329
.
54
Wang
,
M.
,
D.
Windgassen
,
E. T.
Papoutsakis
.
2008
.
Comparative analysis of transcriptional profiling of CD3+, CD4+ and CD8+ T cells identifies novel immune response players in T-cell activation.
BMC Genomics
9
:
225
.
55
Pei
,
L.
,
A.
Castrillo
,
M.
Chen
,
A.
Hoffmann
,
P.
Tontonoz
.
2005
.
Induction of NR4A orphan nuclear receptor expression in macrophages in response to inflammatory stimuli.
J. Biol. Chem.
280
:
29256
29262
.
56
O’Kane
,
M.
,
T.
Markham
,
A. N.
McEvoy
,
U.
Fearon
,
D. J.
Veale
,
O.
FitzGerald
,
B.
Kirby
,
E. P.
Murphy
.
2008
.
Increased expression of the orphan nuclear receptor NURR1 in psoriasis and modulation following TNF-alpha inhibition.
J. Invest. Dermatol.
128
:
300
310
.
57
Satoh
,
J.
,
M.
Nakanishi
,
F.
Koike
,
S.
Miyake
,
T.
Yamamoto
,
M.
Kawai
,
S.
Kikuchi
,
K.
Nomura
,
K.
Yokoyama
,
K.
Ota
, et al
.
2005
.
Microarray analysis identifies an aberrant expression of apoptosis and DNA damage-regulatory genes in multiple sclerosis.
Neurobiol. Dis.
18
:
537
550
.
58
Doi
,
Y.
,
S.
Oki
,
T.
Ozawa
,
H.
Hohjoh
,
S.
Miyake
,
T.
Yamamura
.
2008
.
Orphan nuclear receptor NR4A2 expressed in T cells from multiple sclerosis mediates production of inflammatory cytokines.
Proc. Natl. Acad. Sci. USA
105
:
8381
8386
.
59
Doering
,
T. A.
,
A.
Crawford
,
J. M.
Angelosanto
,
M. A.
Paley
,
C. G.
Ziegler
,
E. J.
Wherry
.
2012
.
Network analysis reveals centrally connected genes and pathways involved in CD8+ T cell exhaustion versus memory.
Immunity
37
:
1130
1144
.
60
Giordano
,
M.
,
C.
Henin
,
J.
Maurizio
,
C.
Imbratta
,
P.
Bourdely
,
M.
Buferne
,
L.
Baitsch
,
L.
Vanhille
,
M. H.
Sieweke
,
D. E.
Speiser
, et al
.
2015
.
Molecular profiling of CD8 T cells in autochthonous melanoma identifies Maf as driver of exhaustion.
EMBO J.
34
:
2042
2058
.
61
Speiser
,
D. E.
,
P. C.
Ho
,
G.
Verdeil
.
2016
.
Regulatory circuits of T cell function in cancer.
Nat. Rev. Immunol.
16
:
599
611
.
62
van Bijnen
,
S. T.
,
D.
Wouters
,
G. J.
van Mierlo
,
P.
Muus
,
S.
Zeerleder
.
2015
.
Neutrophil activation and nucleosomes as markers of systemic inflammation in paroxysmal nocturnal hemoglobinuria: effects of eculizumab.
J. Thromb. Haemost.
13
:
2004
2011
.
63
Sng
,
J. C.
,
H.
Taniura
,
Y.
Yoneda
.
2004
.
A tale of early response genes.
Biol. Pharm. Bull.
27
:
606
612
.
64
Moretta
,
A.
,
A.
Poggi
,
D.
Pende
,
G.
Tripodi
,
A. M.
Orengo
,
N.
Pella
,
R.
Augugliaro
,
C.
Bottino
,
E.
Ciccone
,
L.
Moretta
.
1991
.
CD69-mediated pathway of lymphocyte activation: anti-CD69 monoclonal antibodies trigger the cytolytic activity of different lymphoid effector cells with the exception of cytolytic T lymphocytes expressing T cell receptor alpha/beta.
J. Exp. Med.
174
:
1393
1398
.
65
López-Cabrera
,
M.
,
A. G.
Santis
,
E.
Fernández-Ruiz
,
R.
Blacher
,
F.
Esch
,
P.
Sánchez-Mateos
,
F.
Sánchez-Madrid
.
1993
.
Molecular cloning, expression, and chromosomal localization of the human earliest lymphocyte activation antigen AIM/CD69, a new member of the C-type animal lectin superfamily of signal-transmitting receptors.
J. Exp. Med.
178
:
537
547
.
66
Castellanos
,
M. C.
,
C.
Muñoz
,
M. C.
Montoya
,
E.
Lara-Pezzi
,
M.
López-Cabrera
,
M. O.
de Landázuri
.
1997
.
Expression of the leukocyte early activation antigen CD69 is regulated by the transcription factor AP-1.
J. Immunol.
159
:
5463
5473
.
67
Zhang
,
T.
,
Y.
Liang
,
R.
Fu
,
L. J.
Li
,
J.
Wang
,
H.
Liu
,
H. L.
Wang
,
E. B.
Ruan
,
W.
Qu
,
G. J.
Wang
, et al
2010
.
[Quantity and function of T cell subsets in patients with paroxysmal nocturnal hemoglobinuria].
Zhongguo Shi Yan Xue Ye Xue Za Zhi
18
:
721
725
.
68
Critchley-Thorne
,
R. J.
,
N.
Yan
,
S.
Nacu
,
J.
Weber
,
S. P.
Holmes
,
P. P.
Lee
.
2007
.
Down-regulation of the interferon signaling pathway in T lymphocytes from patients with metastatic melanoma.
PLoS Med.
4
:
e176
.
69
Smith
,
C. A.
,
H. J.
Gruss
,
T.
Davis
,
D.
Anderson
,
T.
Farrah
,
E.
Baker
,
G. R.
Sutherland
,
C. I.
Brannan
,
N. G.
Copeland
,
N. A.
Jenkins
, et al
.
1993
.
CD30 antigen, a marker for Hodgkin’s lymphoma, is a receptor whose ligand defines an emerging family of cytokines with homology to TNF.
Cell
73
:
1349
1360
.
70
Werner
,
S. L.
,
J. D.
Kearns
,
V.
Zadorozhnaya
,
C.
Lynch
,
E.
O’Dea
,
M. P.
Boldin
,
A.
Ma
,
D.
Baltimore
,
A.
Hoffmann
.
2008
.
Encoding NF-kappaB temporal control in response to TNF: distinct roles for the negative regulators IkappaBalpha and A20.
Genes Dev.
22
:
2093
2101
.
71
Zhou
,
Q.
,
H.
Wang
,
D. M.
Schwartz
,
M.
Stoffels
,
Y. H.
Park
,
Y.
Zhang
,
D.
Yang
,
E.
Demirkaya
,
M.
Takeuchi
,
W. L.
Tsai
, et al
.
2016
.
Loss-of-function mutations in TNFAIP3 leading to A20 haploinsufficiency cause an early-onset autoinflammatory disease.
Nat. Genet.
48
:
67
73
.
72
Matmati
,
M.
,
P.
Jacques
,
J.
Maelfait
,
E.
Verheugen
,
M.
Kool
,
M.
Sze
,
L.
Geboes
,
E.
Louagie
,
C.
Mc Guire
,
L.
Vereecke
, et al
.
2011
.
A20 (TNFAIP3) deficiency in myeloid cells triggers erosive polyarthritis resembling rheumatoid arthritis.
Nat. Genet.
43
:
908
912
.
73
Sisto
,
M.
,
S.
Lisi
,
D. D.
Lofrumento
,
G.
Ingravallo
,
E.
Maiorano
,
M.
D’Amore
.
2011
.
A failure of TNFAIP3 negative regulation maintains sustained NF-κB activation in Sjögren’s syndrome.
Histochem. Cell Biol.
135
:
615
625
.
74
Maxwell
,
J. R.
,
I. R.
Gowers
,
K. P.
Kuet
,
A.
Barton
,
J.
Worthington
,
A. G.
Wilson
.
2012
.
Expression of the autoimmunity associated TNFAIP3 is increased in rheumatoid arthritis but does not differ according to genotype at 6q23.
Rheumatology (Oxford)
51
:
1514
1515
.
75
Schnell
,
S.
,
C.
Démollière
,
P.
van den Berk
,
H.
Jacobs
.
2006
.
Gimap4 accelerates T-cell death.
Blood
108
:
591
599
.
76
Filén
,
S.
,
R.
Lahesmaa
.
2010
.
GIMAP proteins in T-lymphocytes.
J. Signal Transduct.
2010
:
268589
.
77
Filén
,
J. J.
,
S.
Filén
,
R.
Moulder
,
S.
Tuomela
,
H.
Ahlfors
,
A.
West
,
P.
Kouvonen
,
S.
Kantola
,
M.
Björkman
,
M.
Katajamaa
, et al
.
2009
.
Quantitative proteomics reveals GIMAP family proteins 1 and 4 to be differentially regulated during human T helper cell differentiation.
Mol. Cell. Proteomics
8
:
32
44
.
78
Katagiri
,
T.
,
H.
Kawamoto
,
T.
Nakakuki
,
K.
Ishiyama
,
M.
Okada-Hatakeyama
,
S.
Ohtake
,
Y.
Seiki
,
K.
Hosokawa
,
S.
Nakao
.
2013
.
Individual hematopoietic stem cells in human bone marrow of patients with aplastic anemia or myelodysplastic syndrome stably give rise to limited cell lineages.
Stem Cells
31
:
536
546
.

The authors have no financial conflicts of interest.

Supplementary data