Abstract
An unresolved issue in autoimmunity is the lack of surrogate biomarkers of immunological self-tolerance for disease monitoring. Here, we show that peripheral frequency of a regulatory T cell population, characterized by the coexpression of CD3 and CD56 molecules (TR3-56), is reduced in individuals with new-onset type 1 diabetes (T1D). In three independent T1D cohorts, we find that low frequency of circulating TR3-56 cells is associated with reduced beta-cell function and with the presence of diabetic ketoacidosis. Since autoreactive CD8+ T cells mediate disruption of insulin-producing beta cells1,2,3, we demonstrate that TR3-56 cells can suppress CD8+ T cell functions in vitro by reducing the levels of intracellular reactive oxygen species. The suppressive function, phenotype and transcriptional signature of TR3-56 cells are also altered in children with T1D. Together, our findings indicate that TR3-56 cells constitute a regulatory cell population that controls CD8+ effector functions, whose peripheral frequency may represent a traceable biomarker for monitoring immunological self-tolerance in T1D.
This is a preview of subscription content, access via your institution
Access options
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$29.99 / 30 days
cancel any time
Subscribe to this journal
Receive 12 digital issues and online access to articles
$119.00 per year
only $9.92 per issue
Buy this article
- Purchase on Springer Link
- Instant access to full article PDF
Prices may be subject to local taxes which are calculated during checkout
Similar content being viewed by others
Data availability
The data that support the findings of this study are available from the corresponding authors upon request. Transcriptional data of the TR3-56, NK, CD3+CD56− and CD8+ cells from adult healthy individuals can be found in the Gene Expression Omnibus database under accession code GSE106082. Transcriptional data the of TR3-56 cells from children with T1D and healthy children can also be found in the Gene Expression Omnibus database under accession code GSE134916. Source data for Figs. 1–4 and Extended Data Figs. 1, 2, 3 and 6 are included with this paper.
References
Liblau, R. S., Wong, F. S., Mars, L. T. & Santamaria, P. Autoreactive CD8 T cells in organ-specific autoimmunity: emerging targets for therapeutic intervention. Immunity 17, 1–6 (2002).
Coppieters, K. T. et al. Demonstration of islet-autoreactive CD8 T cells in insulitic lesions from recent onset and long-term type 1 diabetes patients. J. Exp. Med. 209, 51–60 (2012).
Culina, S. et al. CD8+ T cell frequencies in the pancreas, but not in blood, distinguish type 1 diabetic patients from healthy donors>. Sci. Immunol. 3, eaao4013 (2018).
Schmidt, R. E., Murray, C., Daley, J. F., Schlossman, S. F. & Ritz, J. A subset of natural killer cells in peripheral blood displays a mature T cell phenotype. J. Exp. Med. 1, 351–356 (1986).
Zhou, J. et al. High circulating CD3+CD56+CD16+ natural killer-like T cell levels predict a better IVF treatment outcome. J. Reprod. Immunol. 97, 197–203 (2013).
Diao, H. et al. A possible role for NKT-like cells in patients with chronic hepatitis B during telbivudine treatment. Immunol. Lett. 160, 65–71 (2014).
Galgani, M. et al. Meta-immunological profiling of children with type 1 diabetes identifies new biomarkers to monitor disease progression. Diabetes 62, 2481–2491 (2013).
Atkinson, M. A. et al. How does type 1 diabetes develop? The notion of homicide or β-cell suicide revisited. Diabetes 60, 1370–1379 (2011).
Beato-Víbora, P. I. & Tormo-García, M. Á. Glycemic control and insulin requirements in type 1 diabetic patients depending on the clinical characteristics at diabetes onset. Endocr. Res. 39, 86–90 (2014).
Pipkin, M. E. et al. Interleukin-2 and inflammation induce distinct transcriptional programs that promote the differentiation of effector cytolytic T cells. Immunity 32, 79–90 (2010).
Kalia, V. et al. Prolonged interleukin-2Rα expression on virus-specific CD8+ T cells favors terminal-effector differentiation in vivo. Immunity 32, 91–103 (2010).
Alter, G., Malenfant, J. M. & Altfeld, M. CD107a as a functional marker for the identification of natural killer cell activity. J. Immunol. Methods 294, 15–22 (2004).
Wagner, J. A. et al. CD56bright NK cells exhibit potent antitumor responses following IL-15 priming. J. Clin. Invest. 127, 4042–4058 (2017).
Yi, J. S., Holbrook, B. C., Michalek, R. D., Laniewski, N. G. & Grayson, J. M. Electron transport complex I is required for CD8+ T cell function. J. Immunol. 177, 852–862 (2006).
Bai, A. et al. NADH oxidase-dependent CD39 expression by CD8+ T cells modulates interferon gamma responses via generation of adenosine. Nat. Commun. 9, 8819 (2015).
Nazarewicz, R. R., Bikineyeva, A. & Dikalov, S. I. Rapid and specific measurements of superoxide using fluorescence spectroscopy. J. Biomol. Screen. 18, 498–503 (2013).
Criddle, D. N. et al. Menadione-induced reactive oxygen species generation via redox cycling promotes apoptosis of murine pancreatic acinar cells. J. Biol. Chem. 281, 40485–40492 (2006).
Terrazzano, G. et al. T cell activation induces CuZn superoxide dismutase (SOD)-1 intracellular re-localization, production and secretion. Biochim. Biophys. Acta 1843, 265–274 (2014).
Wicker, L. S. et al. Type 1 diabetes genes and pathways shared by humans and NOD mice. J. Autoimmun. 25, 29–33 (2005).
Roep, B. O. The role of T-cells in the pathogenesis of Type 1 diabetes: from cause to cure. Diabetologia 46, 305–321 (2003).
Dirice, E. et al. Increased β-cell proliferation before immune cell invasion prevents progression of type 1 diabetes. Nat. Metab. 1, 509–518 (2019).
Pugliese, A. Autoreactive T cells in type 1 diabetes. J. Clin. Invest. 127, 2881–2891 (2017).
Pinkse, G. G. M. et al. Autoreactive CD8 T cells associated with β cell destruction in type 1 diabetes. Proc. Natl Acad. Sci. USA 102, 18425–18430 (2005).
Sivori, S. et al. Human NK cells: surface receptors, inhibitory checkpoints, and translational applications. Cell. Mol. Immunol. 16, 430–441 (2019).
Tan, T. G., Mathis, D. & Benoist, C. Singular role for T-BET+CXCR3+ regulatory T cells in protection from autoimmune diabetes. Proc. Natl Acad. Sci. USA 113, 14103–14108 (2016).
van Halteren, A. G., Kardol, M. J., Mulder, A. & Roep, B. O. Homing of human autoreactive T cells into pancreatic tissue of NOD-scid mice. Diabetologia 48, 75–82 (2005).
Shan, Z., Xu, B., Mikulowska-Mennis, A. & Michie, S. A. CCR7 directs the recruitment of T cells into inflamed pancreatic islets of nonobese diabetic (NOD) mice. Immunol. Res. 58, 351–357 (2014).
Rudensky, A. Y. Regulatory T cells and Foxp3. Immunol. Rev. 241, 260–268 (2011).
Godfrey, D. I. & Kronenberg, M. Going both ways: immune regulation via CD1d-dependent NKT cells. J. Clin. Invest. 114, 1379–1388 (2004).
Kuylenstierna, C. et al. NKG2D performs two functions in invariant NKT cells: direct TCR-independent activation of NK-like cytolysis and co-stimulation of activation by CD1d. Eur. J. Immunol. 41, 1913–1923 (2011).
Wirasinha, R. C. et al. GPR65 inhibits experimental autoimmune encephalomyelitis through CD4+ T cell independent mechanisms that include effects on iNKT cells. Immunol. Cell Biol. 96, 128–136 (2018).
Lin, Y.-L. & Lin, S.-C. Analysis of the CD161-expressing cell quantities and CD161 expression levels in peripheral blood natural killer and T cells of systemic lupus erythematosus patients. Clin. Exp. Med. 17, 101–109 (2017).
Mingari, M. C., Pietra, G. & Moretta, L. Immune checkpoint inhibitors: anti-NKG2A antibodies on board. Trends Immunol. 40, 83–85 (2019).
De Simone, M. et al. Transcriptional landscape of human tissue lymphocytes unveils uniqueness of tumor-infiltrating T regulatory cells. Immunity 45, 1135–1147 (2016).
Schmidleithner, L. et al. Enzymatic activity of HPGD in Treg cells suppresses Tconv cells to maintain adipose tissue homeostasis and prevent metabolic dysfunction. Immunity 50, 1232–1248.e14 (2019).
Craig, M. E., Hattersley, A. & Donaghue, K. C. Definition, epidemiology and classification of diabetes in children and adolescents. Pediatr. Diabetes 10, 3–12 (2009).
Revised criteria for diagnosis of coeliac disease. Report of Working Group of European Society of Paediatric Gastroenterology and Nutrition. Arch. Dis. Child. 65, 909–911 (1990).
Hanley, P., Lord, K. & Bauer, A. J. Thyroid disorders in children and adolescents: a review. JAMA Pediatr. 170, 1008–1019 (2016).
Bruzzaniti, S. et al. An immunometabolic pathomechanism for chronic obstructive pulmonary disease. Proc. Natl Acad. Sci. USA 116, 15625–15634 (2019).
McGinnes, K., Chapman, G., Marks, R. & Penny, R. A fluorescence NK assay using flow cytometry. J. Immunol. Methods 86, 7–15 (1986).
Tallerico, R. et al. Human NK cells selective targeting of colon cancer-initiating cells: a role for natural cytotoxicity receptors and MHC class I molecules. J. Immunol. 190, 2381–2390 (2013).
Terrazzano, G. et al. Differential involvement of CD40, CD80, and major histocompatibility complex class I molecules in cytotoxicity induction and interferon-gamma production by human natural killer effectors. J. Leukoc. Biol. 72, 305–311 (2002).
Sabbatini, M. et al. Oscillatory mTOR inhibition and Treg increase in kidney transplantation. Clin. Exp. Immunol. 182, 230–240 (2015).
Del Fabbro, C., Scalabrin, S., Morgante, M. & Giorgi, F. M. An extensive evaluation of read trimming effects on Illumina NGS data analysis. PLoS ONE 8, e85024 (2013).
Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 17, 10–12 (2011).
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Pertea, M. et al. StringTie enables improved reconstruction of a transcriptome from RNA-seq reads. Nat. Biotechnol. 33, 290–295 (2015).
Anders, S., Pyl, P. T. & Huber, W. HTSeq: a Python framework to work with high-throughput sequencing data. Bioinformatics 31, 166–169 (2015).
Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).
Anders, S. & Huber, W. Differential expression analysis for sequence count data. Genome Biol. 11, R106 (2010).
Acknowledgements
We thank M. Montagna and all members of the IEOS-CNR for their technical support, and M. Carrara and F. Marabita for assistance in the transcriptomic analysis. This paper was supported by grants from the Juvenile Diabetes Research Foundation (grant no. 2-SRA-2018-479-S-B to M.G.; grant no. 1-SRA-2018-477-S-B to P.D.C); the European Foundation for the Study of Diabetes (EFSD/JDRF/Lilly Programme 2016 to M.G.); the National Multiple Sclerosis Society (grant no PP-1804-30725 to M.G.); Fondazione Italiana Sclerosi Multipla (grant no. 2016/R/18 to G.M.; grant no. 2018/R/4 to V.D.R.); Ministero della Salute (grant no. GR-2016-02363725 to V.D.R.); Università degli Studi di Napoli Federico II (STAR Program Linea 1–2018 to V.D.R.); European Research Council menTORingTregs grant no. 310496 to G.M.); Telethon (grant no. GGP 17086 to G.M.); Grant Fondazione Italiana Sclerosi Multipla (grant no. 2018/S/5 to G.M.); and the Italian Ministry of Health Giovani Ricercatori (grant no. GR-2016-02363749 to C. Procaccini). This work has also been supported by Italian Ministry of Health Ricerca Corrente - IRCCS MultiMedica.
Author information
Authors and Affiliations
Contributions
S.B., V.R., M.S., A.T.P., A.G., C.L.R. and S.D.S. performed the experiments and data analyses. F.P. and S.B. performed the flow cytometry experiments and data analyses. S.B., V.R., M.S., G.T., G.R., A. Porcellini, V.D.R., C. Procaccini and M.G. analysed the data and interpreted the results. A. Puca and P.D.C. analysed the transcriptional data. V.D.R., A. Porcellini and C. Procaccini were involved in the discussion about the data. J.L., C. Porcellini, V.F., E.M., R.T. and A.F. provided the patient samples and were involved in the discussion about the data. P.D.C., G.T., G.R., J.L., G.M. and M.G. designed the study and wrote the manuscript.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Peer review information Primary Handling Editor: Christoph Schmitt.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data
Extended Data Fig. 1 TR3-56 cell enumeration predicts residual β-cell function and DKA in pre-puberty T1D subjects at disease onset.
a, Box plots indicate the percentage (left) and absolute number (right) of circulating TR3-56 cells in pre-puberty T1D subjects at disease onset from Italian cohort compared with healthy subjects, after adjustment for age, sex and BMI. Data are presented as box plots (min, max, median, and 25th and 75th percentiles), each dot represents a individual subjects (n=86 healthy subjects; n=128 T1D for percentage of TR3-56 cells and n=126 T1D for absolute number of TR3-56 cells). *p<0.0001 by two-tailed Mann-Whitney U-test. b, Box plots indicate the percentage of necrotic (left) and apoptotic (right) rate of circulating TR3-56 cells in healthy subjects (n=47) and T1D children at disease onset (n=82) from Italian cohort. Data are presented as box plots (min, max, median, and 25th and 75th percentiles), each dot represents a individual subjects. *p<0.0001 by two-tailed Mann-Whitney U-test. c, Left, logistic regression modeling shows that percentage of TR3-56 cells predicts the presence or absence of DKA in pre-puberty T1D subjects at diagnosis (n=128) from Italian cohort. T1D subjects were dichotomized on the basis of the presence (Yes) or absence (No) of DKA at disease diagnosis. Low numbers of TR3-56 cells at diagnosis associated with presence of DKA. Right, ROC curve of the model-based prognostic scores of TR3-56 cells for the presence of DKA. AUC=0.72. d, Left, logistic regression modeling shows that absolute number of TR3-56 cell counts predicts the presence or absence of DKA in pre-puberty T1D subjects at diagnosis (n=126) from Italian cohort. Right, ROC curve of the model-based prognostic scores of TR3-56 cells for the presence of DKA. AUC=0.67. e, Left, logistic regression modeling shows that percentage of circulating TR3-56 cells predicts the presence or absence of DKA in post-puberty young adults T1D (n=19) from Italian cohort. Right, ROC curve of the model-based prognostic scores of TR3-56 cells for the presence of DKA. AUC=0.88. f, Left, logistic regression modeling shows that absolute number of TR3-56 cells predicts presence of DKA in post-puberty young adults T1D (n=18) from Italian cohort. Right, ROC curve of the model-based prognostic scores of TR3-56 cells for the presence of DKA. AUC=0.81.
Extended Data Fig. 2 TR3-56 cells in T1D subjects with other autoimmune diseases.
a, Left, logistic regression modeling shows that percentage of TR3-56 cells predicts the presence or absence of DKA in children (n=23) that developed after diagnosis of T1D another autoimmune conditions (CD or AIT). T1D subjects were dichotomized on the basis of the presence (Yes) or absence (No) of DKA at disease diagnosis. Right, ROC curve of the model-based prognostic scores of TR3-56 cells for the presence of DKA. AUC=0.87. b, Left, logistic regression modeling shows that peripheral frequency of TR3-56 cells associated with presence of DKA in children (n=21) that at T1D diagnosis are already affected by other autoimmune conditions. Right, ROC curve of the model-based prognostic scores of TR3-56 cells for the presence of DKA. AUC=0.67.
Extended Data Fig. 3 Correlation between TR3-56 cells and fasting C-peptide in the absence of outliers.
a, Scatter plot showing statistical correlation between frequency of TR3-56 cells and fasting C-peptide in the absence of TR3-56 cell outliers (n=5) in pre-puberty T1D subjects (n=123) at disease onset from Italian cohort. Red line indicates regression line and shading indicates confidence interval. r=0.52, p<0.0001 by two-tailed Pearson’s correlation. b, Scatter plot showing statistical correlation between absolute numbers of TR3-56 cells and C-peptide in the absence of TR3-56 cell outliers (n=7) in pre-puberty T1D subjects (n=119) at disease onset from Italian cohort. Red line indicates regression line and shading indicates confidence interval. r=0.31, p=0.0007 by two-tailed Pearson’s correlation. c, Scatter plot showing positive correlation between the frequency of circulating TR3-56 cells and serum levels of fasting C-peptide in absence of TR3-56 outliers (n=4) in Swedish cohort of T1D children (n=32) at disease onset; Red line indicates regression line and shading indicates confidence interval. r=0.72, p<0.0001 by two-tailed Pearson’s correlation. To identify outliers ROUT (Q=0.1%) method has been applied.
Extended Data Fig. 4 TR3-56 cells suppress CD107a/LAMP-1 and IFN-γ in both autologous and allogeneic conditions, require cell-to-cell contact and is independent from CD56 molecules.
a, Representative flow cytometry histograms showing CD107a/LAMP-1 and IFN-γ staining of CTLs after 4 hours of culture with anti-CD3 plus anti-CD28 microbeads alone (grey), in the presence of autologous or allogeneic TR3-56 cells (blue) as indicated. Dotted lines indicate unstimulated CTLs. Numbers indicate percentage of positive cells. Data are from one representative experiment out of four. b, Representative flow cytometry histograms showing CD107a/LAMP-1 and IFN-γ staining of CTLs cultured for 4 hours with anti-CD3 plus anti-CD28 microbeads alone (grey), in the presence of TR3-56 cells or when TR3-56 cells were separated by transwell (TW) plate system (as indicated). Dotted lines indicate unstimulated CTLs. Numbers indicate percentage of positive cells. Data are from one representative experiment out of six. c, Representative flow cytometry histograms showing CD107a/LAMP-1 and IFN-γ staining of CTLs after 4 hours of culture with anti-CD3 plus anti-CD28 microbeads alone (grey), or in the presence of TR3-56 cells (blue), either in the presence of the control 345.134 IgG2a or the anti-CD56 neutralizing mAb, as indicated. Dotted lines indicate unstimulated CTLs. Numbers indicate percentage of positive cells. Data are from one representative experiment out of three.
Extended Data Fig. 5 Menadione pre-treated CTLs are resistant to TR3-56 cell suppressive activity.
CD107a/LAMP-1 and IFN-γ staining of CTLs cultured for 4 hours in the presence or absence of anti-CD3 plus anti-CD28 microbeads alone or in the presence of TR3-56 cells; light blue lines indicate CTLs pre-treated for 15 minutes with 0.05 mM menadione. Dotted lines indicate unstimulated cells. Numbers indicate percentage of positive cells. Data are from one representative experiment out of six.
Extended Data Fig. 6 Phenotype of peripheral TR3-56 cells in healthy and T1D subjects.
a, Representative flow-cytometry plots showing the gating strategy used to evaluate the expression of CD4 and CD8 on TR3-56 cells (upper panels) and the frequency of invariant (i)NKT cells, evaluated by Vα24 expression and CD1d tetramers loaded with a-Galactosyl ceramide (CD1d-aGal) binding on TR3-56 lymphocytes (lower panels) on both healthy and T1D at-onset subjects, as indicated. Numbers in plots indicate percent of positive cells. b, Column bar showing the TCR Vβ family expression in TR3-56 cells from healthy subjects (yellow) and T1D children (turquoise) at diagnosis, as indicated. Data are from n=5 healthy subjects and n=3 T1D subjects. Data are expressed as mean ± SEM. No statistical significance differences are identified by two-way ANOVA-corrected for multiple comparison using Bonferroni test (p >0.9999).
Extended Data Fig. 7 Hypothetic model showing the regulatory function of TR3-56 cells and β-cell integrity in healthy and autoimmune conditions.
In healthy subjects, normal number and suppressive function of TR3-56 cells control self-reactive CD8+ T cells (green), possible contributing to maintenance of immune self-tolerance and insulin production by live β-islet cells (red). Right, in autoimmune T1D, a lower frequency and a reduced functional capacity of TR3-56 cells correlated with reduced β-cell mass, reduced serum levels of C-peptide and progressive lost of immunological self-tolerance. The schematic model was prepared using the Motifolio Scientific Illustration Toolkit (Motifolio).
Supplementary information
Supplementary Information
Supplementary Figs. 1–9 and Tables 1–4
Rights and permissions
About this article
Cite this article
Terrazzano, G., Bruzzaniti, S., Rubino, V. et al. Type 1 diabetes progression is associated with loss of CD3+CD56+ regulatory T cells that control CD8+ T-cell effector functions. Nat Metab 2, 142–152 (2020). https://doi.org/10.1038/s42255-020-0173-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1038/s42255-020-0173-1
This article is cited by
-
Progression of type 1 diabetes is associated with high levels of soluble PD-1 in islet autoantibody-positive children
Diabetologia (2024)
-
The Increased TIGIT-Expressing CD3+CD56+ Cells Are Associated with Coronary Artery Disease and Its Inflammatory Environment
Inflammation (2023)
-
High levels of blood circulating immune checkpoint molecules in children with new-onset type 1 diabetes are associated with the risk of developing an additional autoimmune disease
Diabetologia (2022)
-
Immunological Monitoring in Beta Cell Replacement: Towards a Pathophysiology-Guided Implementation of Biomarkers
Current Diabetes Reports (2021)
-
Plasma circulating miR-23~27~24 clusters correlate with the immunometabolic derangement and predict C-peptide loss in children with type 1 diabetes
Diabetologia (2020)
