Ab0029 circulating cd39+ treg cells in early rheumatoid arthritis facilitate the antiinflammatory action of methotrexate and serve as early biomarkers of clinical response

BackgroundFoxP3+ regulatory CD4+ T cells (Tregs) are key to the homeostasis of the immune system 1-3. Stressed cells at inflammatory foci release adenine nucleotides to the extracellular space4 that can act as enhancers of inflammation4 but are rapidly hydrolysed by the sequential action of ectonucleotidases CD39 and CD73 rendering the antiinflammatory agent adenosine4. CD39, the rate-limiting enzyme in this cascade, is highly expressed by a subset of human FoxP3+Tregs (Treg39+)1-3. Methotrexate (MTX), the first line of treatment in Rheumatoid Arthritis (RA)4, inhibits AICAR transformylase which results in an enhanced release of extracellular adenine nucleotides4 to be metabolized by Treg39+ cells; hence MTX may cooperate with Treg39+ cells in the control of inflammation.ObjectivesTo examine the relation of CD39 expression on Treg cells of early RA (ERA) patients with the ex vivo and in vivo effect of MTX.MethodsPeripheral blood was obtained from 63 DMARD- and steroid- naïve ERA patients with a disease duration < 24 weeks and 63 age and gender-matched healthy controls (HC). 33 ERA patients donated blood again 12 months after initiating MTX (ERA-R). The frequency of Treg and Treg cell subsets was assessed by flow cytometry. CD4+CD25+CD127- (total T reg), CD4+CD27+CD127-CD39+ Treg (Treg39+) and CD4+CD25-CD39- responder T (Tresp39-) cells were isolated by sorting. The suppressor potency of Tregs was determined in cocultures of isolated Tregs with Tresp. Proliferation was determined by CFSE dilution; cytokine secretion was measured by ELISA of culture supernatants. Disease activity was assessed using the DAS28-ESR score. Low disease activity (LDA) was defined as a DAS28 ≤ 3.2.ResultsThe proportion of Tregs that expressed CD39 (Treg39+) was significantly increased in ERA. Total ERA Tregs were more potent suppressors than HC Tregs, and the difference was reduced by adenosine A2AR antagonists. In vitro, MTX further heightened the total Treg cell potency, with greater amplification in ERA vs HC and this was also reversed by A2AR antagonists. The potency of isolated Treg39+ and its enhancement by MTX were comparable for ERA and HC; the potency of isolated Treg39- was inferior to that of Treg39+ cells in both groups of subjects and was not modified by A2AR antagonists; this further suggests that the differences observed in assays using total Tregs are due to the increased ERA Treg39+ frequency. Patients who achieved LDA had significantly higher basal frequencies of Treg39+ cells and multiple logistic regression (OR 1.93, 95% CI 1.33-4.51) showed that this was independent of basal disease activity, RF or ACPA titres. The Receiver Operating Characteristic (ROC) analysis area under the curve (AUC) was 0.97 (0.94-0.99) for Treg39+ cells. The relative risk (RR) of achieving LDA for patients with a cTreg39+ frequency above the p75 value observed in HC was 13.4 (2.9-75.6) (Fisher’s exact test). The proportion of Treg39+ significantly decreased in all patients whether they had or not achieved LDA, and the frequency and function of ERA-R Treg cells were not different from HC. In fact, 12 months after initiating treatment with MTX, the circulating Treg39+ cell frequency was no longer elevated in ERA; however, its association with the clinical response remained: the cTreg39+ cell frequency observed at 12 months was still significantly higher in those patients who had achieved LDA. This suggests that Treg cell expression of CD39 is upregulated during the initial stages of ERA as a negative feedback mechanism of inflammation.ConclusionMTX cooperates with Treg39+ cells to control inflammation and the pretreatment Treg39+ frequency in ERA is associated with the clinical response to MTX at 12 months.References[1]Peres RS, et al. Proc Natl Acad Sci U S A. 2015.[2]Deaglio S, et al. J Exp Med. 2007.[3]Borsellino G, et al. Blood. 2007.[4]Montesinos MC, et al. Arthritis Rheum. 2007.Disclosure of InterestsAlejandro Villalba: None declared, Laura Nuño: None declared, Marta Benito-Miguel: None declared, Irene Monjo: None declared, Marta Novella-Navarro: None declared, Diana Peiteado: None declared, Sara García-Carazo: None declared, Alejandro Balsa: None declared, Maria-Eugenia Miranda-Carus Grant/research support from: Gebro Pharma - Unrestricted research grant