Depletion of abundant plasma proteins for extracellular vesicle proteome characterization: benefits and pitfalls

Blood extracellular vesicles (EVs) play essential roles in cell-cell communication and their molecular cargo is a promising source of disease biomarkers. However, proteomic characterization of plasma-derived EVs is challenged by the presence of highly abundant plasma proteins, which limits the detection of less abundant proteins, and by the low number of EVs in biological fluids. The aim of this study was to investigate if the removal of abundant plasma proteins prior to EV isolation could improve plasma-derived EV characterization by LC-MS/MS and expand the proteome coverage. Plasma depletion was performed using a single-use spin column and EVs were isolated from only 100 µL of non-depleted and depleted plasma by size exclusion chromatography. Afterwards, EVs were characterized by nanoparticle tracking analysis and mass spectrometry-based proteomics using a data-independent acquisition approach. Depleted plasma-derived EVs had higher particle concentrations and particle-to-protein ratios. Depletion did increase the protein coverage with a higher number of identifications in EVs from depleted plasma (474 proteins) than from non-depleted (386 proteins). However, EVs derived from non-depleted plasma carried a slightly higher number of common EV markers. Overall, our findings suggest that plasma depletion prior to EV isolation by size exclusion chromatography provides higher yield and protein coverage, but slightly lower identification of EV markers. This study also showed the possibility to characterize the proteome of EVs derived from small plasma volumes, encouraging the clinical feasibility of the discovery of EV biomarkers.

Prolongation of rat-to-mouse islets xenograft survival by co-transplantation of autologous IL-10 differentiated murine tolerogenic dendritic cells

BACKGROUND

Tolerogenic dendritic cells (DCs) represent a promising approach to promote transplantation tolerance. In this study, the potential of autologous bone marrow (BM)-derived murine DC to protect rat-to-mouse islets xenografts was analyzed.

METHODS

Tolerogenic DCs were generated by differentiating BM cells in the presence of granulocyte-macrophage colony-stimulating factor and interleukin 10 (IL-10, IL-10 DC). The phenotype of IL-10 DC was characterized in vitro by expression of costimulatory/inhibitory molecules (flow cytometry) and cytokines (Luminex and ELISA), their function by phagocytosis and T-cell stimulation assays. To study transplant tolerance in vivo, rat islets were transplanted alone or in combination with autologous murine IL-10 DC under the kidney capsule of streptozotocin-induced diabetic C57BL/6 mice. Xenograft survival was evaluated by monitoring glycemia, cellular infiltration of xenografts by microscopy and flow cytometry 10 days post-transplantation.

RESULTS

Compared with control DC, IL-10 DC exhibited lower levels of major histocompatibility complex class II, costimulatory molecules (CD40, CD86, CD205), lower production of pro-inflammatory cytokines (IL-12p70, TNF, IL-6), and higher production of IL-10. Phagocytosis of xenogeneic rat splenocytes was not impaired in IL-10 DC, whereas stimulation of T-cell proliferation was reduced in the presence of IL-10 DC. Xenograft survival of rat islets in diabetic mice co-transplanted with autologous murine IL-10 DC was significantly prolonged from 12 to 21 days, without additional immunosuppressive treatment. Overall, infiltration of xenografts by T cells and myeloid cells was not different in IL-10 DC recipient mice, but enriched for CD8⁺ T cells and myeloid cells with suppressor-associated phenotype.

CONCLUSIONS

Autologous IL-10-differentiated DC with tolerogenic properties prolong rat-to-mouse islets xenograft survival, potentially by locally inducing immune regulatory cells, indicating their potential for regulatory immune cell therapy in xenotransplantation.

Small-Molecule Immunosuppressive Drugs and Therapeutic Immunoglobulins Differentially Inhibit NK Cell Effector Functions in vitro

Small-molecule immunosuppressive drugs (ISD) prevent graft rejection mainly by inhibiting T lymphocytes. Therapeutic immunoglobulins (IVIg) are used for substitution, antibody-mediated rejection (AbMR) and HLA-sensitized recipients by targeting distinct cell types. Since the effect of ISD and IVIg on natural killer (NK) cells remains somewhat controversial in the current literature, the aim of this comparative study was to investigate healthy donor's human NK cell functions after exposure to ISD and IVIg, and to comprehensively review the current literature. NK cells were incubated overnight with IL2/IL12 and different doses and combinations of ISD and IVIg. Proliferation was evaluated by ³[H]-thymidine incorporation; phenotype, degranulation and interferon gamma (IFNγ) production by flow cytometry and ELISA; direct NK cytotoxicity by standard ⁵¹[Cr]-release and non-radioactive DELFIA assays using K562 as stimulator and target cells; porcine endothelial cells coated with human anti-pig antibodies were used as targets in antibody-dependent cellular cytotoxicity (ADCC) assays. We found that CD69, CD25, CD54, and NKG2D were downregulated by ISD. Proliferation was inhibited by methylprednisolone (MePRD), mycophenolic acid (MPA), and everolimus (EVE). MePRD and MPA reduced degranulation, MPA only of CD56^bright NK cells. MePRD and IVIg inhibited direct cytotoxicity and ADCC. Combinations of ISD demonstrated cumulative inhibitory effects. IFNγ production was inhibited by MePRD and ISD combinations, but not by IVIg. In conclusion, IVIg, ISD and combinations thereof differentially inhibit NK cell functions. The most potent drug with an effect on all NK functions was MePRD. The fact that MePRD and IVIg significantly block NK cytotoxicity, especially ADCC, has major implications for AbMR as well as therapeutic strategies targeting cancer and immune cells with monoclonal antibodies.

Chronic and acute inflammatory conditions determine MEK1 or MEK2 usage as regulators of IL-1β and sIL-1ra expression (117.18)

Deregulation of the production of IL-1β and its natural inhibitor, the secreted form of IL-1 receptor antagonist (sIL-1Ra), plays an important role in chronic inflammatory diseases such as multiple sclerosis. Relevant to this condition direct cellular contact with stimulated T cells potently triggers cytokine production in human monocytes. In this study we investigated the implication of MEK1 and MEK2 in the control of IL-1β and sIL-1Ra production by human monocytes upon two different stimuli: (i) soluble extracts of plasma membranes from stimulated T cells (CEsHUT), mimicking cellular contact with T cells that is relevant to chronic/sterile inflammation; and (ii) LPS that is relevant to acute/infectious inflammation. By using MEK1/2 (U0126) and MEK1 (PD98059) inhibitors and siRNA specific to MEK1 and MEK2 we demonstrate that MEK1 and MEK2 are differentially involved in the induction of IL-1β production upon chronic/sterile and acute/infectious inflammatory conditions, MEK1 being dispensable to IL-1β induction in monocytes activated by LPS but required in CEsHUT-activated monocytes. Together with the fact that immunomodulatory treatments of multiple sclerosis - IFNβ and glatiramer acetate - are using MEK2 to induce sIL-1Ra production in human monocytes, MEK1 represents a potential therapeutic target which could participate in the restoration of IL-1β/sIL-1Ra balance in chronic/sterile inflammation without affecting normal response to infectious agents.

HDL interfere with the binding of T cell microparticles to human monocytes to inhibit pro-inflammatory cytokine production

Background

Direct cellular contact with stimulated T cells is a potent mechanism that induces cytokine production in human monocytes in the absence of an infectious agent. This mechanism is likely to be relevant to T cell-mediated inflammatory diseases such as rheumatoid arthritis and multiple sclerosis. Microparticles (MP) generated by stimulated T cells (MP_T) display similar monocyte activating ability to whole T cells, isolated T cell membranes, or solubilized T cell membranes. We previously demonstrated that high-density lipoproteins (HDL) inhibited T cell contact- and MP_T-induced production of IL-1β but not of its natural inhibitor, the secreted form of IL-1 receptor antagonist (sIL-1Ra).

Methodology/Principal Findings

Labeled MP_T were used to assess their interaction with monocytes and T lymphocytes by flow cytometry. Similarly, interactions of labeled HDL with monocytes and MP_T were assessed by flow cytometry. In parallel, the MP_T-induction of IL-1β and sIL-1Ra production in human monocytes and the effect of HDL were assessed in cell cultures. The results show that MP_T, but not MP generated by activated endothelial cells, bond monocytes to trigger cytokine production. MP_T did not bind T cells. The inhibition of IL-1β production by HDL correlated with the inhibition of MP_T binding to monocytes. HDL interacted with MP_T rather than with monocytes suggesting that they bound the activating factor(s) of T cell surface. Furthermore, prototypical pro-inflammatory cytokines and chemokines such as TNF, IL-6, IL-8, CCL3 and CCL4 displayed a pattern of production induced by MP_T and inhibition by HDL similar to IL-1β, whereas the production of CCL2, like that of sIL-1Ra, was not inhibited by HDL.

Conclusions/Significance

HDL inhibit both MP_T binding to monocytes and the MP_T-induced production of some but not all cytokines, shedding new light on the mechanism by which HDL display their anti-inflammatory functions.

Blockade of T cell contact-activation of human monocytes by high-density lipoproteins reveals a new pattern of cytokine and inflammatory genes

BACKGROUND

Cellular contact with stimulated T cells is a potent inducer of cytokine production in human monocytes and is likely to play a substantial part in chronic/sterile inflammatory diseases. High-density lipoproteins (HDL) specifically inhibit the production of pro-inflammatory cytokines induced by T cell contact.

METHODOLOGY/PRINCIPAL FINDINGS

To further elucidate the pro-inflammatory functions of cellular contact with stimulated T cells and its inhibition by HDL, we carried out multiplex and microarray analyses. Multiplex analysis of monocyte supernatant revealed that 12 out of 27 cytokines were induced upon contact with stimulated T cells, which cytokines included IL-1Ra, G-CSF, GM-CSF, IFNgamma, CCL2, CCL5, TNF, IL-1beta, IL-6, IL-8, CCL3, and CCL4, but only the latter six were inhibited by HDL. Microarray analysis showed that 437 out of 54,675 probe sets were enhanced in monocytes activated by contact with stimulated T cells, 164 probe sets (i.e., 38%) being inhibited by HDL. These results were validated by qPCR. Interestingly, the cytokines induced by T cell contact in monocytes comprised IL-1beta, IL-6 but not IL-12, suggesting that this mechanism might favor Th17 polarization, which emphasizes the relevance of this mechanism to chronic inflammatory diseases and highlights the contrast with acute inflammatory conditions that usually involve lipopolysaccharides (LPS). In addition, the expression of miR-155 and production of prostaglandin E(2)-both involved in inflammatory response-were triggered by T cell contact and inhibited in the presence of HDL.

CONCLUSIONS/SIGNIFICANCE

These results leave no doubt as to the pro-inflammatory nature of T cell contact-activation of human monocytes and the anti-inflammatory functions of HDL.

High-density lipoproteins downregulate CCL2 production in human fibroblast-like synoviocytes stimulated by urate crystals

Introduction

To investigate whether monosodium urate (MSU) crystals induce the production of CCL2 (monocyte chemoattractant protein-1; MCP-1) in human fibroblast-like synoviocytes (FLS) and whether this mechanism would be affected by high-density lipoproteins (HDL).

Methods

Human FLS isolated from synovial tissue explants were stimulated with MSU crystals (0.01 to 0.5 mg/ml) or interleukin (IL)-1β (10 pg/ml) in the presence or absence of HDL (50 and 100 μg/ml). The production and expression of CCL2 was evaluated with ELISA, confocal microscopy, immunofluorescence microscopy, chemotaxis assay, and real-time quantitative PCR.

Results

Exposure of FLS to MSU crystals induced CCL2 accumulation in culture medium in a dose- and time-dependent manner, reaching a plateau at 50 to 75 μg/ml MSU crystals and 20 to 24 hours. Although low, the induced CCL2 levels were sufficient to trigger mononuclear cell migration. In resting FLS, CCL2 was localized in small cytoplasmic vesicles whose number diminished with MSU crystal stimulation. Concomitantly, MSU crystals triggered the induction of CCL2 mRNA expression. All these processes were inhibited by HDL, which cause a 50% decrease in CCL2 mRNA levels and a dose-dependent inhibition of the release of CCL2. Similar results were obtained when FLS were pretreated with HDL and washed before activation by MSU crystals or IL-1β, suggesting a direct effect of HDL on the FLS activation state.

Conclusions

The present results demonstrate that MSU crystals induce FLS to release CCL2 that is stored in vesicles in resting conditions. This mechanism is inhibited by HDL, which may limit the inflammatory process by diminishing CCL2 production and, in turn, monocytes/macrophages recruitment in joints. This study confirms the antiinflammatory functions of HDL, which might play a part in the limitation of acute gout attack.

Differential regulation of cytokine production by PI3Kdelta in human monocytes upon acute and chronic inflammatory conditions

Deregulated production of cytokines, including IL-1beta, IL-6 and TNF plays an important role in chronic inflammation. Relevant to this condition, direct cellular contact with stimulated T cells is a potent inducer of cytokine production in human monocytes/macrophages. We previously demonstrated that PI3Ks regulate differential production of IL-1beta and its specific inhibitor secreted IL-1 receptor antagonist (sIL-1Ra) by human monocytes. Here we show that in contrast with PI3Kalpha, beta and gamma, PI3Kdelta accounts for most of the PI3K-dependent signaling ruling the production of IL-1beta, IL-6, TNF and sIL-1Ra in monocytes activated by cellular contact with stimulated T cells (mimicked by CHAPS-solubilized membranes of stimulated T cells, CE sHUT) and lipopolysaccharides (LPS); the latter stimuli being relevant to chronic/sterile and acute/infectious inflammation, respectively. Interestingly, PI3Kdelta activity dampened the production of pro-inflammatory cytokines in LPS-activated monocytes, but induced it in CE sHUT-activated cells. In both CE sHUT- and LPS-activated monocytes PI3Kdelta regulated cytokine transcript expression through the phosphorylation/inactivation of glycogen synthase kinase-3beta (GSK3beta). The blockade of GSK3beta displayed inverse effects to those of PI3Kdelta blockade. Thus, by displaying opposite functions in conditions mimicking chronic/sterile and acute/infectious inflammation, i.e., by repressing pro-inflammatory cytokine expression in LPS-activated monocytes but inducing such mediators in T cell contact-activated monocytes, PI3Kdelta represents a potential therapeutic target specific to chronic/sterile inflammatory conditions.

Stimulated T cells generate microparticles, which mimic cellular contact activation of human monocytes: differential regulation of pro- and anti-inflammatory cytokine production by high-density lipoproteins

Imbalance in cytokine homeostasis plays an important part in the pathogenesis of chronic inflammatory diseases such as multiple sclerosis and rheumatoid arthritis. We demonstrated that T cells might exert a pathological effect through direct cellular contact with human monocytes/macrophages, inducing a massive up-regulation of the prototypical proinflammatory cytokines IL-1beta and TNF. This mechanism that might be implicated in chronic inflammation is specifically inhibited by high-density lipoproteins (HDL). Like many other stimuli, besides proinflammatory cytokines, the contact-mediated activation of monocytes induces the production of cytokine inhibitors such as the secreted form of the IL-1 receptor antagonist (sIL-1Ra). The present study demonstrates that stimulated T cells generate microparticles (MP) that induce the production of TNF, IL-1beta, and sIL-1Ra in human monocytes; the production of TNF and IL-1beta but not that of sIL-1Ra is inhibited in the presence of HDL. The results were similar when monocytes were stimulated by whole membranes of T cells or soluble extracts of the latter. This suggests that MP carry similar monocyte-activating factors to cells from which they originate. Thus, by releasing MP, T cells might convey surface molecules similar to those involved in the activation of monocytes by cellular contact. By extension, MP might affect the activity of cells, which are usually not in direct contact with T cells at the inflammatory site. Furthermore, this study demonstrates that HDL exert an anti-inflammatory effect in nonseptic activation of human monocytes, not only by inhibiting the production of IL-1beta and TNF but also, by leaving sIL-1Ra production unchanged.

Opposite Regulation of IL-1β and Secreted IL-1 Receptor Antagonist Production by Phosphatidylinositide-3 Kinases in Human Monocytes Activated by Lipopolysaccharides or Contact with T Cells

The unbalanced production of IL-1beta and its natural, specific inhibitor, the secreted IL-1R antagonist (sIL-1Ra), plays an important role in chronic/sterile inflammation. Relevant to this condition is direct cellular contact with stimulated T cells which is a potent inducer of cytokine production in human monocytes/macrophages. We previously demonstrated that activation of PI3Ks is a prerequisite of the transcription of the sIL-1Ra gene in human monocytes activated by IFN-beta. In this study, we addressed the question of PI3K involvement in the production of IL-1beta and sIL-1Ra in monocytes activated by cellular contact with stimulated T cells (mimicked by CHAPS-solubilized membranes of stimulated T cells (CE(sHUT))), and a crude preparation of LPS, to compare stimuli relevant to chronic/sterile and acute/infectious inflammation, respectively. In monocytes activated by either CE(sHUT) or LPS, the inhibition of PI3Ks abrogated sIL-1Ra transcript expression and sIL-1Ra production, demonstrating that PI3Ks control the induction of sIL-1Ra gene transcription. In contrast, PI3K inhibition increased the production of IL-1beta protein in both CE(sHUT)- and LPS-activated monocytes, the enhancement being drastically higher in the former. This was not due to changes in IL-1beta mRNA steady-state levels or transcript stability, but to the involvement of PI3Ks in the repression of IL-1beta secretion. The downstream PI3K effector, Akt, was implicated in this process. The present results demonstrate that PI3Ks are involved in the inhibition of IL-1beta secretion and in the induction of sIL-1Ra production in human blood monocytes by controlling different mechanisms in conditions mimicking chronic/sterile (CE(sHUT)) and acute/infectious (LPS) inflammation.

The production of IL-1 receptor antagonist in IFN-beta-stimulated human monocytes depends on the activation of phosphatidylinositol 3-kinase but not of STAT1

IFN-beta induces the production of secreted IL-1R antagonist (sIL-1Ra) without triggering synthesis of the agonist IL-1beta in human monocytes. This might account for its anti-inflammatory properties. Canonically, IFN-beta signals through activation of JAK/STAT pathway, although PI3K and MAPK have also been involved. In this study, the role of PI3K, MEK1, and STAT1 in IFN-beta-induced sIL-1Ra production is investigated in freshly isolated human blood monocytes. PI3K, but not MEK1 activation is essential for sIL-1Ra production in monocytes treated with IFN-beta, as demonstrated by using the respective inhibitors of PI3K and MEK1, Ly294002 and PD98059. The use of cycloheximide and actinomycin D shows that sIL-1Ra was an immediate early gene induced by IFN-beta and that PI3K was controlling sIL-1Ra gene transcription. Although both inhibitors of PI3K and MEK1 diminished the Ser(727) phosphorylation of STAT1 induced by IFN-beta, only Ly294002 inhibited sIL-1Ra production. Furthermore, the inhibition of STAT1-Ser(727) phosphorylation by Ly294002 did not affect STAT1 translocation, suggesting that STAT1 was not involved in sIL-1Ra gene induction. This was confirmed in monocytes that were transfected with small interfering RNA specifically targeting STAT1. Indeed, monocytes in which effective STAT1 gene knockdown was achieved were fully responsive to IFN-beta in terms of sIL-1Ra production. Taken together, the present data demonstrate that the induction of sIL-1Ra transcription and production by IFN-beta in human monocytes involved PI3K, but not STAT1 activation.

Differential Induction of IL-1β and TNF by CD40 Ligand or Cellular Contact with Stimulated T Cells Depends on the Maturation Stage of Human Monocytes

Cellular contact with stimulated T cells potently induces cytokine production in monocytes, a mechanism that is likely to be relevant to chronic inflammation. Although the identity of surface molecules involved in this process remains elusive, CD40 and its ligand, CD40L, are thought to be implicated, considering that they are expressed at the inflammatory site. To ascertain the involvement of CD40L, we compared the activation of three different types of human monocytic cells, i.e., freshly isolated monocytes, monocytes primed with IFN-gamma (IFN-gamma-macrophages), and THP-1 cells. These cells were activated by either membranes isolated from stimulated T cells (HUT-78 or T lymphocytes) to mimic cellular contact, soluble extracts from isolated membranes, or CD40L trimer (CD40LT). The production of TNF and IL-1beta was induced by membranes of stimulated T cells in the three types of target cells, whereas CD40LT induced TNF production in IFN-gamma-macrophages only. Similar results were obtained with soluble extracts of T cell membranes, demonstrating that the difference between membranes and CD40LT was not due to the particulate form of membranes. CD40LT induced neither transcript nor protein of cytokines in monocytes, whereas in IFN-gamma-macrophages, IL-1beta and TNF mRNA were observed, but only TNF was measured in cell supernatants. Finally, anti-CD40L Abs failed to inhibit TNF and IL-1beta production induced in IFN-gamma-macrophages by solubilized membranes, whereas TNF production induced by CD40LT was inhibited. These results demonstrate that CD40L is not required in monocyte activation by direct cellular contact with stimulated T cells, although soluble CD40LT induces the production of TNF in IFN-gamma-macrophages.

Opposite effects of IFN beta on cytokine homeostasis in LPS- and T cell contact-activated human monocytes

Multiple sclerosis (MS) is an immune-mediated disease improved by interferon-beta (IFNbeta) therapy. IFNbeta may owe its anti-inflammatory property to its ability to induce interleukin-1 receptor antagonist (IL-1Ra) without triggering IL-1beta synthesis in human monocytes. Furthermore, we recently demonstrated that IFNbeta inhibits the production of IL-1beta and tumor necrosis factor-alpha (TNF) in human monocytes activated by cellular contact with stimulated T cells, a mechanism which we suspected of playing an important part in the pathogenesis of chronic inflammatory diseases including MS. Here we compare modulatory effects of IFNbeta on the production of proinflammatory cytokines (IL-1beta, IL-1alpha, TNF, and IL-6) and IL-1Ra in human monocytes stimulated by lipopolysaccharides (LPS) and isolated plasma membranes of stimulated T cells (msHUT), which are likely to reflect monocyte activation in acute and chronic inflammation, respectively. In monocytes activated by either LPS or msHUT, IFNbeta did not modulate the secretion of IL-1alpha and IL-6, but it enhanced the production of IL-1Ra in a dose-dependent manner. However, in monocytes activated by msHUT, the expression of cell-associated and intracellular IL-1alpha was inhibited by IFNbeta, correlating with the inhibition of IL-1alpha transcript. IFNbeta inhibited the expression (mRNA) and production (protein) of IL-1beta and TNF, while enhancing those of IL-1Ra in monocytes activated by msHUT. In contrast, in monocytes activated by LPS, IFNbeta enhanced the expression and production of IL-1beta, TNF, and IL-1Ra, suggesting that it did not display anti-inflammatory properties in these conditions. This study demonstrates that IFNbeta displays opposite effects depending on the type of activation of human monocytes, suggesting that it may affect different pathogenic mechanisms in opposite ways.

The active metabolite of leflunomide, A77 1726, inhibits the production of prostaglandin E(2), matrix metalloproteinase 1 and interleukin 6 in human fibroblast-like synoviocytes

OBJECTIVES

To investigate the effects of the active metabolite of leflunomide, A77 1726, on fibroblast-like synoviocytes. In rheumatoid arthritis (RA) synoviocytes participate in tissue destruction by producing metalloproteinases (MMP), prostaglandin E(2) (PGE(2)) and interleukin (IL) 6, which are involved in extracellular matrix degradation, resorption of the mineral phase and osteoclast-mediated bone resorption.

METHODS

Human synoviocytes were stimulated with IL-1alpha or tumour necrosis factor alpha (TNF-alpha) in the presence of A77 1726. Culture supernatants were analysed for production of interstitial collagenase (MMP-1), tissue-inhibitor of metalloproteinases 1 (TIMP-1), PGE(2) and IL-6. Total RNA was isolated and analysed for steady-state levels of MMP-1, cyclooxygenase-2 (COX-2) and IL-6 mRNA.

RESULTS

A77 1726 inhibited the production of PGE(2) in synoviocytes activated by TNF-alpha and IL-1alpha with median inhibitory concentrations (IC(50)) of 7 and 3 microM respectively. In contrast, MMP-1 and IL-6 production was inhibited at high A77 1726 concentrations (> 10 microM), whereas TIMP-1 was not affected. The inhibition of MMP-1 and IL-6 production was due to the known inhibitory effect of A77 1726 on pyrimidine synthesis, as it was reversed by the addition of uridine. This did not apply to PGE(2) production, which was inhibited via direct action of A77 1726 on COX-2, as shown by the increasing amount of substrate (arachidonic acid) in the culture medium.

CONCLUSION

This study shows that some of the beneficial effect of leflunomide in RA patients may be due to the inhibition of PGE(2), IL-6 and MMP-1 production in synoviocytes. This effect, coupled with its multiple inhibitory effects on T lymphocyte functions, might account for the significant reduction in the rate of disease progression in RA patients treated with leflunomide.