Pathophysiological characteristics of ... - Wiley Online Library

2 downloads 0 Views 2MB Size Report
(Immudex, Copenhagen, Denmark) as previously described.17. These. DMRs were used for detection of PPI-specific CD8+ T cells. 2.3 | Immunophenotyping ...
Received: 9 November 2016

Revised: 2 March 2017

Accepted: 5 April 2017

DOI: 10.1111/pedi.12536

ORIGINAL ARTICLE

Pathophysiological characteristics of preproinsulin-specific CD8+ T cells in subjects with juvenile-onset and adult-onset type 1 diabetes: A 1-year follow-up study Mahinder Paul1 | Darshan Badal2 | Neenu Jacob2 | Devi Dayal2 | Rakesh Kumar2 | Anil Bhansali1 | Sanjay Kumar Bhadada1 | Naresh Sachdeva1 1 Department of Endocrinology, Post Graduate Institute of Medical Education and Research (PGIMER), Chandigarh, India 2

Department of Pediatrics, Post Graduate Institute of Medical Education and Research (PGIMER), Chandigarh, India Correspondence Naresh Sachdeva, Department of Endocrinology, Post Graduate Institute of Medical Education and Research (PGIMER), Chandigarh 160012, India. Email: [email protected] Funding Information Department of Biotechnology (DBT); Ministry of Science and Technology, Government of India, Grant/Award number: BT/PR13480/ Med/30/273/2010.

Aims/Hypothesis: Among the beta-cell associated antigens, preproinsulin (PPI) has been shown to play a key role in the pathogenesis of type 1 diabetes (T1D). PPI-specific autoreactive CD8+ T cells emerge early during beta-cell destruction and persist in peripheral circulation during diabetes progression. However, the influence of insulin therapy on phenotype of autoreactive CD8+ T cells in T1D including, juvenile-onset T1D (JOT1D), and adult-onset T1D (AOT1D) is not yet known. Methods: We followed the time course of PPI-specific CD8+ T cells in JOT1D and AOT1D subjects that achieved glycemic control after 1 year of insulin therapy, using major histocompatibility complex-I (MHC-I) dextramers by flow cytometry. Results and Discussion: At follow-up, PPI-specific CD8+ T cells could be detected consistently in peripheral blood of all T1D subjects. Proportion of PPI-specific effector memory (TEM) subsets decreased, while central memory T (TCM) cells remained unchanged in both groups. Expression of granzyme-B and perforin in PPI-specific CD8+ T cells also remained unchanged. Further, on analysis of B-chain and signal peptide (SP) specific CD8+ T cell responses separately, we again observed decrease in TEM subset in both the groups, while increase in naive (TN) subset was observed in B-chain specific CD8+ T cells only. Conclusion: Our study shows that PPI-specific CD8+ T cells can be detected in both JOT1D and AOT1D subjects over a period of time with reliable consistency in frequency but variable pathophysiological characteristics. Insulin therapy seems to reduce the PPI-specific TEM subsets; however, the PPI-specific TCM cells continue to persist as attractive targets for immunotherapy. KEYWORDS

CD8+ T cells, dextramers, insulin therapy, preproinsulin, type 1 diabetes

ABBREVIATIONS

FITC fluorescein isothiocyanate

7-AAD 7-Aminoactinomycin D

FMO fluorescence minus one

ADA American Diabetes Association

GAD glutamic acid decarboxylase

AOT1D adult-onset type 1 diabetes

GZM granzyme

APC allophycocyanin

HbA1c glycated hemoglobin

BIMAS bioinformatics and molecular analysis section

HLA human leukocyte antigen

CD cluster of differentiation

IA-2 islet antigen 2

DMR dextramer

IEDB immune epitope database

Pediatric Diabetes. 2017;1–12.

wileyonlinelibrary.com/journal/pedi

© 2017 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

1

2

PAUL ET AL.

JOT1D juvenile-onset type 1 diabetes

advances have been made to characterize antigen-specific T cells

MFI mean fluorescence intensity

using highly sensitive multimers such as dodecamers22 or using mass

MHC major histocompatibility complex

cytometry.23 Increase in PPI-specific CD8+ T cells has been shown to

PBMCs peripheral blood mononuclear cells

be associated with decline in fasting C-peptide levels.20 It has been

PE phycoerythrin

observed that PPI-specific CD8+ T cells isolated from peripheral

PECy7 phycoerythrin-cyanine7

blood of recent-onset and long-standing T1D subjects are more dif-

PerCP peridinin-chlorophyll

ferentiated24 and express markers of memory.20 Certain immune

PFN perforin

interventions have showed modulation in the functional phenotype

PHA phytohaemagglutinin

of these cells.25 Earlier, by using MHC-I multimers, Roep et al26

PPI preproinsulin

reported that clinical efficacy of proinsulin gene-therapy improved C-

SSP sequence-specific primer

peptide levels in the treatment group that was associated with

T1D type 1 diabetes

decline in proinsulin-specific CD8+ T cells. Therefore, determination

TCM central memory T cells

of frequency and phenotype of PPI-specific CD8+ T cells is becoming

TCR T cell receptor

increasingly beneficial in monitoring disease progression and efficacy

TEFF effector T cells

of therapeutic approaches in T1D advocating the utility of MHC mul-

TEM effector memory T cells

timers as useful biomarkers.26–29

TN naive T cells

Most T1D subjects require lifelong exogenous insulin therapy to control blood glucose levels. However, the long-term effects of insulin administration on autoreactive CD8+ T cells are largely unknown.

1 | I N T RO D UC T I O N

Also, there is scant data on the time course of PPI-specific CD8+ T cells during the clinical development of different forms of autoim-

Type 1 diabetes (T1D) is an autoimmune disease characterized by the

mune diabetes, especially in juvenile vs adult onset T1D. In this con-

destruction of pancreatic beta-cells by infiltrating immune cells lead-

text, we analyzed PPI-specific CD8+ T cells using MHC-I DMRs in

ing to insulin deficiency.1 T1D usually strikes in children (juvenile-

newly diagnosed JOT1D and AOT1D subjects and followed those

onset T1D [JOT1D]) though the onset can also occur in adults (adult-

subjects that achieved glycemic control after 1 year of insulin therapy

onset T1D [AOT1D]). Age of clinical onset of T1D is determined by

to investigate changes in the frequency and phenotype of PPI-

the intensity of the beta-cell destruction, a process modulated by

specific CD8+ T cells.

both genetic and environmental factors.2 It has been demonstrated that autoimmune infiltrate in insulitic lesions comprises of open and dynamic cell population constantly reseeded with both T and B cells.3–5 However, destruction of beta-cells has been shown to be mediated mainly by CD8+ T cells that recognize islet associated antigens6,7 and release granules containing granzymes (GZM) and perfo-

2 | RESEARCH DESIGN AND METHODS 2.1 | Subjects

rin (PFN).8 Among the beta-cell associated antigens, preproinsulin

The study was conducted at the departments of Endocrinology and

(PPI) is now known to play a key role in disease initiation and

Pediatrics at the Post Graduate Institute of Medical Education and

progression5,8–15 and PPI-specific CD8+ T cell clones isolated from

Research (PGIMER), Chandigarh, India. The study cohort comprised

T1D subjects have been shown to effectively destroy beta-cells

of newly diagnosed 45 JOT1D subjects (mean [SEM] age:

9,10,16

in vitro.

Moreover, increase in the frequency and pathogenicity

8.40  0.53 years) and 21 AOT1D subjects (age: 29.08  1.37 years)

of PPI-specific CD8+ T cells has been associated with the severity of

and 10 healthy control subjects (age: 28.0  1.15 years). Diabetes

the disease.17

was diagnosed as per American Diabetes Association (ADA) criteria.

The AOT1D is characterized by a longer symptomatic period

Inclusion criteria for autoimmune diabetes were; presence of autoan-

before diagnosis, lower frequencies of insulin autoantibodies, better

tibodies to glutamic acid decarboxylase (GAD65) or islet antigen-2

preservation of residual beta-cell function and lower HLA-DR3/DR4

(IA-2) or insulin. Exclusion criteria included, anemia (Hb < 8.0 g/dL),

heterozygosity than JOT1D.18 However, blood glucose levels, gly-

any acute illness, other autoimmune diseases (including celiac dis-

cated hemoglobin (HbA1c), degree of metabolic decompensation or

ease), lymphomas, psychiatric illness, pregnancy, and prior insulin

frequency of T1D in first-degree relatives do not differ among the

therapy. The study was approved by institute’s ethics committee.

2 groups.18 Thus, assessment of anti beta-cell immune responses at

After obtaining informed consent in writing, fasting peripheral blood

disease onset and at regular intervals during disease progression, pro-

samples were obtained from all subjects in heparinized vacutainers at

vide better information in various forms of T1D than single time

the time of recruitment and at regular intervals of 3 months till 1 year

point observations. Beta-cell associated CD8+ T cells have been iden-

of insulin therapy. Fasting plasma C-peptide was determined by

tified10,19 and characterized from peripheral blood as well as pancre-

Electro-chemiluminescence Immunoassay (Roche Diagnostics, Basel,

atic islets7 in T1D subjects using major histocompatibility complex-1

Switzerland), while HbA1c was estimated every 3 months by cation-

(MHC-I) multimers such as, tetramers20 or dextramers (DMRs).17

exchange chromatography (Bio-Rad, Hercules, CA) in all the subjects.

DMRs outperform tetramers for detection of antigen-specific T cells

Genomic DNA was extracted from the whole blood using kit (Real

particularly with TCRs of low MHC affinity.21 More recently,

Biotech Corporation, Taipei, Taiwan) before HLA-I typing of the

3

PAUL ET AL.

subjects by PCR-SSP method in 96-well typing plates (Inno-train,

with PPI for 72 hours in both the groups at baseline as well as follow

Kronberg, Germany). PCR products were resolved by electrophoresis

up. A total of 106 PBMCs/mL were plated/well in triplicates in flat

on a 2% agarose gel, followed by determination of human leukocyte

bottom 24-well tissue culture plates (BD-Falcon) in RPMI-1640

antigen (HLA) alleles.

medium supplemented with 0.1% penicillin, streptomycin, 10% fetal calf serum, and stimulated with optimized concentration of PPI

2.2 | DMR synthesis and validation Eight different allophycocyanin (APC) labeled MHC-I DMRs loaded with 9-mer PPI-derived epitopes, having highest affinity for respective MHC-I molecules as determined by MHC-I peptide binding prediction softwares, T cell Epitope Prediction Tool of Immune Epitope Database (IEDB)

(10 μg/mL) (GL Biochem, Shanghai,China) or 5 μg/mL phytohaemagglutinin (PHA) (positive control) (Sigma) or without any stimulant (negative control) for 72 hours at 37 C in 5% CO2. Following incubation, cells were harvested, stained, and reanalyzed for PPI-specific CD8+ T cell subsets as described above.

(La Jolla, CA), Bioinformatics and Molecular Analysis Section (BIMAS) (Centre for Information Technology, NIH, Bethesda, MD) and SYFPEITHI (Biomedical Informatics, Heidelberg, Germany) were custom synthesized (Immudex, Copenhagen, Denmark) as previously described.17 These DMRs were used for detection of PPI-specific CD8+ T cells.

2.5 | Statistical analysis The results were documented as mean  standard error of mean (SEM). Paired t test was used to compare the frequency and subsets of PPI-specific CD8+ T cells before and after insulin therapy and in vitro stimulation with PPI. Unpaired t test was used to compare

2.3 | Immunophenotyping and analysis of PPI-specific CD8+ T cells

the means between different groups. For all analyses, P < .05 was

On the basis of achievement of glycemic control post-insulin therapy,

performed using Graph pad Prism (v-4.0, La Jolla, California) and

13 JOT1D and 11 AOT1D subjects were followed up for immunophe-

Microsoft Office Excel (2013).

considered as statistically significant. All the statistical analyses were

notyping of PPI-specific CD8+ T cells. Peripheral blood mononuclear cells (PBMCs) were isolated by density gradient centrifugation with ficoll (Sigma-Aldrich, St Louis, MO). DMR staining was performed

3 | RE SU LT S

as previously described.17 Cell viability of samples was assessed using 7-aminoactinomycin D (7-AAD) in a separate tube. Briefly, 2 × 106 PBMCs were incubated with 10 μL DMR (corresponding to the HLA

3.1 | Clinical characteristics of the recruited subjects

type of the subject) at room temperature in dark for 10 minutes. Anti-

A total of 45 JOT1D (mean [SEM] age, 8.40  0.53 years) and

CD3 Peridinin-chlorophyll (PerCP), anti-CD8 Phycoerythrin-Cyanine7

21 AOT1D subjects (age 29.08  1.37 years) were recruited at the

(PECy7), anti-CD45RA allophycocyanin-H7 (APC-H7), anti-CD197

time of diagnosis and their HbA1c, fasting plasma C-peptide levels,

Phycoerythrin (PE) and cocktail of Alexa-Fluor 700 labeled anti-CD4,

ketoacidosis events, BMI, and HLA-I type were determined. Numbers

anti-CD14, anti-CD16, and anti-CD19 were then added for 15 minutes

of subjects presenting with ketoacidosis were significantly higher

followed by washing with FACS buffer (BD Biosciences, San Jose, CA).

among JOT1D (20/45) subjects compared with AOT1D (3/21) group

Cells were then permeabilized with Cytoperm-CytoFix solution and

(P = .007) indicating that disease onset might have occurred a long

incubated with anti-Granzyme-B-V450 and Perforin-Fluorescein-

time before diagnosis in JOT1D subjects. The AOT1D subjects had

isothiocyanate (FITC) for 30 minutes at 4 C. After washing, cells were

more residual beta-cell mass as indicated by higher fasting plasma C-

acquired on a flowcytometer (FACS Aria II) and analyzed using FACS

peptide levels (0.99  0.19 ng/mL) compared to JOT1D group

Diva software (6.01) (BD Biosciences). At least 1 million total events

(0.44  0.07 ng/mL) (P = .002). The mean (SEM) HbA1c was high

were acquired and a minimum of 500 000 CD3+ T lymphocytes were

in both the JOT1D (12.56  0.56%) and AOT1D (11.03  0.83%)

analyzed. Gating was first performed on lymphocytes followed by gat-

group at the time of diagnosis. As expected, diverse range of HLA-I

ing of CD3+CD4−CD14−CD16−CD19− cells to exclude CD4+ T cells,

alleles was observed in both the groups with HLA-A*02, HLA-A*24,

monocytes, NK cells, and B cells. CD8+ T cells recognizing the DMRs

HLA-B*08, HLA-B*40 being the most common alleles in all subjects.

were then gated and analyzed for markers of differentiation stages includ-

HLA-A*24 (14/45), HLA-B*40 (15/45) alleles were most common in

ing, naïve (TN) (CD45RA+ CD197+), central-memory (TCM) (CD45RA

JOT1D subjects, whereas HLA-A*02 (9/21), HLA-B*08 (8/21) were

−CD197+), effector-memory (TEM) (CD45RA−CD197−), effector (TEFF)

most frequent in AOT1D subjects. Recruited subjects were then trea-

(CD45RA+ CD197−) along with intracellular expression of GZM-B and

ted with insulin and followed up for glycemic control for minimum

PFN (Figure 2A). Relevant isotype controls and fluorescence minus one

1 year. Average daily insulin requirement was 24.78  2.29 and

(FMO) tubes were used to set gates. DMRs loaded with Flu-specific pep-

31.20  2.55 U in JOT1D and AOT1D subjects, respectively

tides were used as positive control, while DMRs carrying HIV-1-specific

(P = .09). Glycemic control in both the groups was evaluated every

peptides were used as negative controls Figure S1.17

3 months and those subjects that achieved HbA1c below 7.5% postinsulin therapy were reanalyzed for immunophenotyping of PPI-

2.4 | In vitro stimulation with PPI

specific CD8+ T cells. Amongst the subjects that were followed up, mean (SEM) age

We also analyzed changes in the mean (SEM) frequency of various

at presentation of disease was 9.61 (1.27) years in JOT1D and

subsets of PPI-specific CD8+ T cells following in vitro stimulation

31.73 (1.92) years in AOT1D group. Previously reported susceptible

4

PAUL ET AL.

A

B

JOT1D

AOT1D 14 12 10

15

%HbA1C

% HbA1C

20

10

6 4

5

2 0

0

Baseline

C

Baseline

Follow-up

D

JOT1D 1.2

Fasting plasma C-peptide (ng/ml)

Fasting plasma C-peptide (ng/ml)

8

1 0.8 0.6 0.4 0.2 0

Baseline

Follow-up AOT1D

1.4 1.2 1 0.8 0.6 0.4 0.2 0

Follow-up

Baseline

E

Follow-up

HLA-A 7 6 5 4 3 2 1 0 A*01

A*02

A*03

A*11

A*24 JOT1D

F

A*26

A*29

A*31

A*33

A*68

AOT1D

HLA-B 8 7 6 5 4 3 2 1 0 B*07 B*08 B*13 B*15 B*35 B*40 B*44 B*47 B*50 B*51 B*52 B*58 B*78 JOT1D

AOT1D

FIGURE 1

A and B, Changes in percent glycated hemoglobin (HbA1c) from baseline to post-insulin therapy (follow up) in (A) juvenile-onset type 1 diabetes (JOT1D) and (B) adult-onset type 1 diabetes (AOT1D) subjects. Further, (C-D) represents changes in fasting plasma C-peptide levels from baseline to follow up in (C) JOT1D and (D) AOT1D subjects. E and F, Represents distribution of (E) human leukocyte antigen-A (HLA-A) and (F) HLA-B alleles among JOT1D and AOT1D subjects

MHC-I allele, HLA-A*02 was observed in 33% (8/24), HLA-A*24 in

7.04 (0.08) % (53  0.9 mmol/mol) at follow up compared with

38 % (9/24) of subjects. At follow up, mean (SEM) duration of insu-

12.89 (0.99) % (116  10.8 mmol/mol) at baseline in JOT1D

lin therapy was 12.23 (0.20) months in JOT1D and 12.05 (0.30)

(P = .0001) and 7.34 (0.13) % (57  1.4 mmol/mol) at follow up as

months in AOT1D with daily insulin requirement of 31.54 (3.70) U

compared with 10.00 (0.50) % (86  5.5 mmol/mol) at baseline in

and 27.45 (2.31) U, respectively (P = .38). Glycemic control in both

AOT1D group (P = .0001) (Figure 1A,B). However, during this period,

the groups was achieved, with a mean (SEM) HbA1c of

mean (SEM) fasting plasma C-peptide levels fell significantly in both

5

PAUL ET AL.

TABLE 1

Clinical characteristics of recruited subjects at the time of recruitment (baseline) and post-insulin therapy (follow up)

Characteristic

JOT1D

Patients (n)

AOT1D

13

Age (y)

11

9.61 (1.27)

31.73 (1.92)

Insulin therapy (mo)

12.23 (0.20)

12.05 (0.30) 27.45 (2.31)

Average daily insulin requirement (U)

31.54 (3.70)

BMI (kg/m2)

16.93 (0.68)

19.40 (2.63)

Anti GAD65/IA2 (% subjects)

84.6%

72.7%

Anti-insulin antibodies (% subjects)

46.1%

36.3%

HbA1c (%, mmol/mol)

a

Fasting plasma C-peptide (ng/mL)b

Baseline

Follow up

Baseline

Follow up

12.89 (0.99), 116 (10.8)

7.04 (0.08), 53 (0.9)

10.0 (0.50), 86 (5.5)

7.34 (0.13), 57 (1.4)

0.10 (0.03)

0.72 (0.18)

0.17 (0.05)

0.48 (0.15)

Abbreviations: AOT1D, adult-onset type 1 diabetes; HbA1c, glycated hemoglobin; JOT1D, juvenile-onset type 1 diabetes. Data are presented as mean (SEM). a

Significant decline in mean (SEM) HbA1c was observed following insulin therapy in both JOT1D (P = .0001) and AOT1D (P = .0001) group.

b

Mean (SEM) fasting plasma C-peptide levels decreased significantly in both JOT1D (P = .04) and AOT1D (P = .03) groups.

the JOT1D (0.48  0.15–0.10  0.03 ng/mL) (P = .04) and AOT1D

difference was observed in frequency of any other PPI-specific CD8+

(0.72  0.18–0.17  0.05 ng/mL) (P = .03) groups (Table 1) indica-

T cell subset (Figure 3A,B).

tive of persistent beta-cell damage.

On comparing the 2 groups, the JOT1D subjects demonstrated higher frequency of PPI-specific CD8+ TN cells as compared with AOT1D subjects at baseline (JOT1D, 52.06  3.96%; AOT1D,

3.2 | PPI-specific CD8+ T cells persist in JOT1D and AOT1D subjects with similar frequencies but variable pathophysiological characteristics over a period of time In this study, we followed the subjects achieving glycemic control and performed immunophenotyping of PPI-specific (DMR+) CD8+ T cells after 1 year of insulin therapy to see its effect on the pathophysiological characteristics of PPI-specific CD8+ T cells (Figure 2). PPI-specific CD8+ T cells were consistently detectable in JOT1D and AOT1D subjects from the time of diagnosis to until 1 year post-insulin therapy. Moreover, no significant change was observed in mean (SEM) frequency of PPI-specific CD8+ T cells in both JOT1D (baseline, 0.04  0.007%, follow up, 0.06  0.007%) (P = .14) and AOT1D (baseline, 0.05  0.02%, follow up, 0.07  0.02% (P = .45) subjects (Figure 2B). Also, the mean (SEM) frequency of the PPI-specific CD8+ T cells was comparable in both the JOT1D (0.06  0.007%) and AOT1D (0.07  0.02%) (P = .48) groups at the time of follow up. When

32.11  3.93%)

(P = .02)

as

well

as

follow

up

(JOT1D,

68.20  3.86%; AOT1D, 43.25  4.66%) (P = .01) (Figure 4A,B). At follow up, JOT1D subjects were found to have lower proportion of DMR+ CD8+ TEFF cells (JOT1D, 17.90  3.26%) as compared with AOT1D group (AOT1D, 36.15  6.37%) (P = .02) (Figure 4B). In addition, we also analyzed PPI-specific CD8+ T cells in few subjects with uncontrolled diabetes (HbA1c > 7.5). However, no significant change in PPI-specific CD8+ T cell frequency or their subsets was observed in either of the JOT1D and AOT1D groups (Table S2A,B). The absolute counts of PPI-specific CD8+ T cell subsets were also determined at both time points (Table S3A,B). Each T cell subset responds differently to in vitro stimulation either by proliferation or by differentiation to other subset. When, we analyzed changes in the subsets of PPI-specific CD8+ T cells following in vitro stimulation with PPI in both the groups at both time points. Post stimulation, the percent change in the mean (SEM) frequency of all the subsets (TCM, TN, TEM, TEFF) was comparable in both the groups at baseline and at follow up (Figure 5A,B).

we followed the T1D subjects individually as well, 10 of 13 (76.9%) JOT1D and 9 of 11 AOT1D (81.8%) subjects showed no change in their PPI-specific CD8+ T cell frequencies. PPI-specific CD8+ T cells were undetectable in most of the healthy controls with only 2 of 10 subjects showing detectable frequency (Table S1, Supporting Information).

3.3 | Intracellular expression of GZM-B and PFN does not change following insulin therapy

Further, we followed the changes in subsets of PPI-specific CD8

Autoreactive CD8+ T cells destroy beta-cells mainly by producing

+ T cells after treatment with insulin by analysing the surface expres-

GZM-B and PFN. The frequency of PPI-specific CD8+ T cells expres-

sion of CD197 and CD45RA (Figure S2). A significant decrease in

sing GZM-B and PFN was assessed before and after insulin therapy.

mean ( SEM) relative frequency of PPI-specific CD8+ TEM cells in

We observed no significant change in the frequency of PPI-specific

both JOT1D (20.23  3.00%–10.29  2.80%) (P = .04) and AOT1D

CD8+ T cells co-expressing GZM-B and PFN after insulin therapy in

subjects (28.47  4.12%–16.05  3.09%) (P = .04) was observed. In

either JOT1D (21.09  4.07%–20.97  2.41%) (P = .97) or AOT1D

addition, a relative increase in PPI-specific CD8+ TN population

(29.66  6.37%–19.18  3.53%) (P = .11) subjects (Figure 6).

(52.06  3.96%–68.20  3.86%) (P = .01) in JOT1D and AOT1D

In case of PPI-specific CD8+ T cells expressing GZM-B that is DMR

(32.11  3.93%–43.25  4.66%) (P = .05) groups during follow up as

+ CD8+ GZM-B+ T cells, we did not observe any significant change in

compared to baseline was observed (Table 2). No significant

the mean (SEM) frequency in both JOT1D (baseline, 27.72  4.98%;

6

PAUL ET AL.

A

B

C

D

JOT1D

AOT1D 0.3

0.25

% DMR+ CD8+ T cells

% DMR+ CD8+ T cells

0.3

0.2 0.15 0.1 0.05 0

Baseline

Follow-up

0.25 0.2 0.15 0.1 0.05 0

Baseline

Follow-up

FIGURE 2

A, Representative images showing gating strategy for immunophenotyping of preproinsulin (PPI)-specific (dextramer [DMR]+) CD8+ T cells in peripheral blood. A, Lymphocytes were gated as per forward and side scatter profiles. B, After analysis of cell viability by 7aminoactinomycin D (7-AAD) staining, CD4+ T cells, monocytes, natural (NK) cells, and B cells were excluded by anti-CD4, CD14, CD16, CD19 staining and remaining CD3+ T cells were gated for further selection. C, CD8+ T cells recognizing DMRs carrying PPI-derived peptide were selected as PPI-specific CD8 + T cells. D, The PPI-specific CD8+ T cells were then further characterized on the basis of expression of CD45RA and CD197 as naïve (TN) (CD45RA + CD197+), central-memory (TCM) (CD45RA−CD197+), effector-memory (TEM) (CD45RA−CD197−) and effectors (TEFF) (CD45RA+ CD197−). E, The PPI-specific CD8+ T cells were also analyzed for the expression of granzyme-B (GZM-B) and perforin (PFN). B, Mean (SEM) frequency of PPI-specific (DMR+) CD8+ T cells in subjects with juvenile-onset type 1 diabetes (JOT1D) and adult-onset type 1 diabetes (AOT1D) at the time of recruitment and follow up. Line graphs show change in frequency of PPI-specific CD8+ T cells from baseline to follow-up individually in (C) JOT1D and (D) AOT1D subjects. No significant change in mean (SEM) frequency of PPIspecific CD8 + T cells following insulin therapy was observed in both the groups. Also, the frequency of DMR + CD8 + T cells was comparable in both the groups at both time points. Paired t test was used to compare the frequencies of PPI-specific CD8+ T cells before and after insulin therapy, while unpaired t test was used to compare frequencies between the 2 groups

follow up, 32.91  4.17%) (P = .39) and AOT1D subjects (baseline,

of GZM-B or PFN on PPI-specific CD8+ T cells following insulin ther-

34.76  7.11%; follow up, 37.92  6.20%) (P = .85) (Figure 7A). Simi-

apy (Figure 7C,D).

larly, we did not observe any significant change in PPI-specific CD8+ T cells expressing PFN in both the JOT1D (baseline, 26.53  4.15%; follow up, 24.38  2.56%) (P = .65) and AOT1D subjects (baseline, 34.31  7.95%; follow up, 21.26  3.61%) (P = .10) (Figure 7B).

3.4 | B chain and signal peptide-(SP) specific CD8+ T cell responses

In order to assess expression of GZM-B and PFN on per cell

In order to find out whether the observed changes in PPI-specific

basis by PPI-specific CD8+ T cells, we also compared the mean fluo-

CD8+ T cells post insulin therapy are restricted to mature insulin or

rescence intensity (MFI) of GZM-B and PFN on DMR+ CD8+ T cells.

evenly distributed across all regions of PPI, including the signal pep-

Similar to frequency, no significant change was observed in the MFI

tide (SP), we assessed the B chain and SP specific CD8+ T cell

7

PAUL ET AL.

TABLE 2

(A) Frequencies of PPI-specific (DMR+) CD8+ T cells and their subsets in JOT1D group at baseline and follow-up. (B) Frequencies of PPI-specific (DMR+) CD8+ T cells and their subsets in AOT1D group at baseline and follow up Baseline

P value

Follow up

(A) JOT1D DMR + CD8+ T cells

0.04 (0.007)

0.06 (0.007)

TCM

5.07 (1.78)

3.62 (1.12)

.14 .42

TN

52.06 (3.96)

68.20 (3.86)

.01

TEM

20.23 (3.00)

10.29 (2.80)

.04

TEFF

22.63 (3.76)

17.90 (3.26)

.35

GZM-B + PFN+

21.09 (4.07)

20.97 (2.41)

.97

DMR + CD8+ T cells

0.05 (0.02)

0.07 (0.02)

.45

TCM

5.60 (1.66)

4.56 (1.12)

.56

TN

32.11 (3.93)

43.25 (4.66)

.05

TEM

28.47 (4.12)

16.05 (3.09)

.04

(B) AOT1D

TEFF

33.72 (8.59)

36.15 (6.37)

.82

GZM-B + PFN+

29.66 (6.37)

19.18 (3.53)

.11

Abbreviations: AOT1D, adult-onset type 1 diabetes, GZM, granzyme; JOT1D, juvenile-onset type 1 diabetes, TCM, central memory T cells, TN, naïve T cells, TEM, effector memory T cells; TEFF, effector T cells, PFN, perforin. Data are presented as mean ( SEM)%.

B

JOT1D

105 p=0.01

Relative percentage

90

Follow-up

75 60 45 30

p=0.04

p=0.35

p=0.42

15

AOT1D

80

Baseline

70 Relative percentage

A

Baseline Follow-up

p=0.05

p=0.82

60 p=0.04

50 40 30 p=0.56

20 10

0

0 TCM

TN

T

EM

T EFF

TCM

TN

T

EM

T EFF

FIGURE 3

Relative proportion of preproinsulin (PPI)-specific dextramer+ (DMR+) CD8+ T cell subsets in (A) juvenile-onset type 1 diabetes (JOT1D) and (B) adult-onset type 1 diabetes (AOT1D) at baseline and follow up. After 1 year of insulin therapy, significant decrease in PPI-specific CD8+ TEM cells in both JOT1D (P = .04) and AOT1D (P = .04) groups was observed. PPI-specific CD8 + TN cells increased significantly in JOT1D (P = .01) and in AOT1D (P = .05) group. Paired t test was used to compare the frequencies of DMR+ CD8+ T cell subsets before and after insulin therapy

JOT1D

80

p=0.02

60

AOT1D p=0.28

p=0.26

40 p=0.83

20

100

JOT1D

p=0.01

AOT1D

80 p=0.02

60 40 p=0.35

p=0.55

20 0

0 TCM FIGURE 4

B Relative percentage

Relative percentage

A

TN

TEM

TEFF

TCM

TN

TEM

TEFF

Mean (SEM) frequencies of preproinsulin (PPI)-specific dextramer+ (DMR+) CD8+ T cell subsets in juvenile-onset type 1 diabetes (JOT1D) and adult-onset type 1 diabetes (AOT1D) subjects at (A) baseline and (B) follow up. After 1 year of insulin therapy, AOT1D subjects were found to have more PPI-specific CD8+ T cells with effector phenotype (P = .02). However, relative proportion of naïve cells was higher in JOT1D subjects at both time points. Frequencies of PPI-specific CD8+ T cell subsets between the groups were compared using unpaired t test

8

PAUL ET AL.

FIGURE 5

Percent change in the mean ( SEM) frequency of preproinsulin (PPI)-specific dextramer+ (DMR+) CD8+ T cell subsets following in vitro stimulation with PPI in (A) juvenile-onset type 1 diabetes (JOT1D) and (B) adult-onset type 1 diabetes (AOT1D) groups. Delta changes in the frequency of all PPI-specific CD8+ T cell subsets following stimulation with PPI were comparable at both the time points. Paired t test was used to compare the delta changes in the frequency of PPI-specific CD8+ T cell subsets before and after stimulation with PPI (24.95  3.52% vs 13.70  3.30%) (P = .01) in B chain carrier group. Further, relative frequency of PPI-specific CD8+ TN cells increased at the time of follow up in these subjects (32.30  7.24% vs 56.50  7.65%) (P = .01). However, frequencies of TCM (P = .95) and TEFF cells (P = .23) remained unchanged. Among the SP-specific CD8 + T cell responses, we observed significant decrease in PPI-specific CD8+ TEM (24.30  3.72% vs 13.07  2.81%) (P = .05) population whereas TCM (P = .30), TN (P = .37) and TEFF (P = .38) cells were relatively unchanged (Table 3A,B).

4 | DI SCU SSION It is proposed that PPI acts as a key autoantigen mediating the destruction of beta-cells. At the same time, long-term intensive

FIGURE 6

Mean (SEM) frequency of preproinsulin (PPI)-specific dextramer+ (DMR+) CD8+ T cells expressing granzyme-B (GZM-B) and perforin (PFN) in juvenile-onset type 1 diabetes (JOT1D) and adult-onset type 1 diabetes (AOT1D) subjects at baseline and follow up. We observed no significant change in frequency of PPI-specific CD8+ GZM-B+ PFN+ T cells following insulin therapy in both the groups. Paired t test was used to compare the frequency of PPIspecific CD8+ GZM-B+ PFN+ T cells before and after insulin therapy, while unpaired t test was used to compare the 2 groups

exogenous insulin therapy is believed to exert an immunomodulatory effect on beta-cell specific immune responses.30–32 CD8+ T cells expressing TCRs-specific for PPI epitopes can recognize the antigen presented by APCs and differentiate into effector and memory cells causing damage to beta-cells. In this context, we measured the CD8+ T cell responses to PPI in JOT1D and AOT1D groups at the time of diagnosis and following glycaemic control post-insulin therapy, an approach better than measuring antibodies to insulin or beta-cell autoantigens. CD8+ T cells against PPI-derived epitopes

responses in all T1D subjects separately. For this purpose, we

can be detected even in T1D subjects negative for insulin autoanti-

divided all the subjects, independent of their age at onset, into

bodies and unlike insulin-autoantibodies, these assays do not show

2 groups; A01, 03, and A011 carriers (B chain carrier group) (n = 9)

positivity following administration of exogenous insulin. Use of

as HLA-I DMRs used for these alleles were loaded with B chain-

such highly specific yet sensitive assay like ours, are encouraged

derived peptides and the SP-carrier group (n = 15) where DMRs car-

now-a-days to identify and characterize immune cells associated

rying SP-derived peptides (A*02, A*24, B*08, B*35, B*51)

with pathology of T1D. Previous studies comparing the clinical characteristics of JOT1D

were used. Firstly, we observed no significant change in overall frequency of

and AOT1D subjects reported better preservation of beta-cell mass

DMR+ CD8+ T cells following insulin therapy in both B chain carrier

and lower frequency of autoantibodies in AOT1D subjects.2,18 Fur-

group (baseline; 0.04  0.005% vs follow up; 0.07  0.02%) and the

ther, Pruul et al33 reported variation in expression of genes involved

SP-carrier

group

(baseline;

0.05  0.01%

vs

follow

up;

0.06  0.007%).

in co-stimulatory pathways in the peripheral blood of newly diagnosed JOT1D and AOT1D subjects. The JOT1D group exhibit upre-

Among the B chain-specific responses, we observed significant

gulated CD80-CTLA4 pathway, whereas, in AOT1D group, CD86-

decrease in relative frequency of PPI-specific CD8+ TEM cells

CD28 pathway, and TGF-β production were activated.33 However,

9

PAUL ET AL.

FIGURE 7 Mean (SEM) frequency of preproinsulin (PPI)-specific dextramer+ (DMR+) CD8+ T cell expressing (A) granzyme (GZM-B) and (B) perforin (PFN) in juvenile-onset type 1 diabetes (JOT1D) and adult-onset type 1 diabetes (AOT1D) subjects at baseline and follow up. No significant change in frequency of PPI-specific CD8+ T cells expressing GZM-B or PFN was observed following the insulin therapy. C, Expression of GZM-B and (D) PFN per PPI-specific CD8+ T cell was also assessed in terms of mean fluorescence intensity (MFI) before and after insulin therapy in both the groups. Again, no significant change in MFI of GZM-B or PFN was observed in PPI-specific CD8+ T cells in both the groups. Paired t test was used to compare the percentages and MFI of PPI-specific CD8+ T cells expressing GZM-B or PFN before and after insulin therapy

there is scant data on differentiation status and phenotype of autoreactive CD8+ T cells in these groups. In our previous study, we demonstrated that PPI-specific CD8+ T cells from JOT1D subjects exhibit frequency and pathophysiological characteristics similar to

TABLE 3

(A) Frequencies of B-chain-specific (DMR+) CD8+ T cells and their subsets at baseline and follow-up. (B) Frequencies of SPspecific (DMR+) CD8+ T cells and their subsets at baseline and follow-up

those from AOT1D subjects.17 In this study, we now demonstrate

Baseline

Follow up

P value

that PPI-specific CD8+ T cells can be detected consistently in periph-

(A)

eral blood of subjects with autoimmune diabetes over a period of

DMR + CD8+ T cells

0.04 (0.005)

0.07(0.019)

time using MHC DMRs. There was no change in the frequency of

TCM

3.45 (1.27)

3.41 (1.05)

0.95

B chain .18

PPI-specific CD8+ T cells in the peripheral blood of both the groups

TN

32.30 (7.24)

56.50 (7.65)

0.01

from the time of diagnosis to until follow up. Consistent with previ-

TEM

24.95 (3.52)

13.70 (3.30)

.01

ous studies, we found higher levels of fasting C-peptide levels in

TEFF

39.18 (6.38)

26.41 (3.41)

.23

AOT1D compared with JOT1D.34,35 At the same time, we found sig-

(B)

nificant decline in the fasting C-peptide levels in both the groups.

DMR + CD8+ T cells

0.05 (0.01)

0.06 (0.007)

.57

Thus persistence of the PPI-specific CD8+ T cells appears to lead to

TCM

6.52 (1.50)

4.54 (0.86)

.30

unremitting decline in beta-cell mass.

SP

TN

47.34 (5.42)

54.28 (7.07)

.37

When, we followed the changes in phenotype of PPI-specific

TEM

24.30 (3.72)

13.07 (2.81)

.05

CD8+ T cells in the subjects achieving well-controlled HbA1c postin-

TEFF

21.83 (4.84)

28.11 (6.08)

.38

sulin therapy, JOT1D subjects demonstrated higher proportion of PPI-specific naïve and low proportion of TEFF cells compared with

Abbreviations: DMR, dextramers, SP, signal peptide, TCM, central memory T cells, TN, naïve T cells, TEM, effector memory T cells, TEFF, effector T cells.

AOT1D group. Higher proportion of TEFF cells in AOT1D could be

Data are presented as mean (SEM)%.

10

PAUL ET AL.

expected due to persistent residual beta-cell mass leading to acti-

molecules might be promoting further decline in beta-cell mass and

vated antigenic presentation whereas JOT1D subjects at this point

here again, antigenic memory could be an important factor in their

were found to have almost diminished beta-cell mass. Persistence of

unaltered expression. Therefore, modulation or depletion of memory

higher proportion of PPI-specific CD8+ TN cells following higher rates

T cells appears to be a valid strategy in controlling autoimmunity and

of ketoacidosis and higher HbA1c at the time of diagnosis indicates

rescuing beta-cell damage in T1D individuals.41,42

that substantial autoimmune damage has already occurred in the

Collectively, our results highlight that PPI-specific CD8+ T cells

JOT1D group leading to very low beta-cell mass. In addition, further

can be detected in peripheral blood of subjects with autoimmune dia-

increase in the relative percentage of PPI-specific CD8+ TN cell popu-

betes consistently over a period of time, and insulin therapy seems to

lation in the peripheral blood of T1D subjects, especially in the B

influence the phenotype but not the frequency of such autoreactive

chain group, may be attributed to intensive insulin therapy coupled

T cells. The PPI-specific T cells particularly those with memory phe-

with decreased beta-cell mass. Furthermore, we also found significant

notype can be targeted using antigen-specific approaches and moni-

decrease in PPI-specific CD8+ TEM cells in both the groups at the

toring epitope specific T cell repertoires would be highly useful in

time of follow up. Reduction in the frequency of TEM cells in periph-

assessing clinical efficacy of such therapies.43

eral blood may occur due to their selective migration to pancreas or could be attributed to tolerance induced by insulin therapy.30,31 Previous studies have reported that CD8+ T cells recognizing leader sequence and B chain-derived peptides are predominant in T1D subjects.15,20 Administration of exogenous insulin presumably leads to increased presentation of A and B chain-derived peptides, which could have total different immunological effects on the subsets of B-chain and SP specific CD8 T cells. Theoretically, B chain-specific CD8 T cells should be affected more; however, SP specific CD8+ T cell response also changed, probably due to other regulatory mechanisms induced by insulin in a non-antigen specific manner. It seems that, presentation of SP-derived peptides from endogenous PPI production was still able to sustain the TCM and TEFF population. Animal studies have also shown that insulin-based immunotherapy or peripheral proinsulin expression reduces CD4+ TEM subsets by indu-

ACKNOWLEDGEMENTS We thank Pinaki Dutta, Rama Walia, and Ashu Rastogi, Department of Endocrinology, PGIMER, Chandigarh for help in recruitment of study subjects. We also thank Ravi Sharma, Department of Ophthalmology, PGIMER, Chandigarh, Raj Davinder, Vivek Sharma, and Harmanpreet Kaur Department of Endocrinology, PGIMER, Chandigarh for help in performing routine diagnostic investigations of the study subjects. We also thank participants in giving consent and providing blood samples whenever required. This study was supported by Department of Biotechnology (DBT), Ministry of Science and Technology, Government of India (Grant No. BT/PR13480/Med/30/ 273/2010).

cing Tregs.36,37 In a similar study, Martinuzzi et al38 using ELISpot assay, reported a significant reduction in the PPI-specific IFN-γ pro-

Conflict of interest

ducing CD8+ T cells, illustrating decline in T cell responses following insulin therapy in T1D subjects. Today ELISpot-like assays have been

The authors declare that there are no conflicts of interest associated

replaced by those examining the complete phenotype of antigen-

with this manuscript.

specific T cells. Using our DMR-based assay we could thus identify naïve precursors in addition to the antigen experienced T cells. Another subset of autoreactive CD8+ T cells, particularly those with TCM phenotype have been reported to distinguish T1D from 24

20

Author contribution N.S. conceptualized the study and its design. M.P., D.B., and N.-

reported that TCM

J. performed the experiments and acquired the data. N.S., D.D., R.K.,

cells persist at similar frequency in both recent onset and long stand-

A.B., and S.K.B. recruited the subjects. N.S. and D.D. supervised the

ing T1D subjects with median diabetes duration of 4 years. Persist-

study. N.S. and M.P. interpreted the results critically, reviewed, and

ence of PPI-specific CD8+ T cells especially with memory phenotype

edited the manuscript. All authors revised the article and approved

is considered as a biggest hurdle for any immunotherapeutic

the final version of the manuscript. N.S. is the guarantor of this work

approach. In our study subjects we observed that frequency of PPI-

and, as such, had complete access to all of the data in the study and

specific TCM cells does not change even after insulin therapy sugges-

takes responsibility for the integrity of the data and the accuracy of

tive of their role in maintaining autoimmunity. In fact maintenance of

the data analysis.

healthy subjects in a recent study.

Luce et al

TCM is less dependent on antigenic presentation by antigen presenting cells once the T cells have been sensitized and intensive insulin therapy or declining beta cell mass had minimal impact on this subset. The long-term persistence of such TCM cells can act as a reservoir to replenish the effector subsets to propagate the autoimmune process of beta-cell destruction; thereby posing a major hurdle in immunotherapeutic approaches to prevent disease progression. As a surrogate marker of cytolytic activity,39,40 we observed that there was no decrease in expression of GZM-B and PFN even 1 year post insulin therapy and it appears that the unabated expression of such effector

RE FE RE NC ES 1. Mallone R, Martinuzzi E, Blancou P, et al. CD8+ T-cell responses identify β-cell autoimmunity in human type 1 diabetes. Diabetes. 2007;56 (3):613-621. 2. Sabbah E, Savola K, Ebeling T, et al. Genetic, autoimmune, and clinical characteristics of childhood-and adult-onset type 1 diabetes. Diabetes Care. 2000;23(9):1326-1332. 3. Magnuson AM, Thurber GM, Kohler RH, Weissleder R, Mathis D, Benoist C. Population dynamics of islet-infiltrating cells in autoimmune diabetes. Proc Natl Acad Sci USA. 2015;112(5):1511-1516.

PAUL ET AL.

4. Willcox A, Richardson S, Bone A, Foulis A, Morgan N. Analysis of islet inflammation in human type 1 diabetes. Clin Exp Immunol. 2009;155 (2):173-181. 5. Pathiraja V, Kuehlich JP, Campbell PD, et al. Proinsulin specific, HLADQ8 and HLA-DQ8 transdimer restricted, CD4+ T cells infiltrate the islets in type 1 diabetes. Diabetes. 2015;64(1):172-182. 6. Unger WW, Pearson T, Abreu JR, et al. Islet-specific CTL cloned from a type 1 diabetes patient cause beta-cell destruction after engraftment into HLA-A2 transgenic NOD/scid/IL2RG null mice. PLoS One. 2012;7(11):e49213. 7. Coppieters KT, Dotta F, Amirian N, et al. Demonstration of islet-autoreactive CD8 T cells in insulitic lesions from recent onset and longterm type 1 diabetes patients. J Exp Med. 2012;209(1):51-60. 8. Knight RR, Kronenberg D, Zhao M, et al. Human β-cell killing by autoreactive preproinsulin-specific CD8 T cells is predominantly granulemediated with the potency dependent upon T-cell receptor avidity. Diabetes. 2013;62(1):205-213. 9. Skowera A, Ellis RJ, Varela-Calviño R, et al. CTLs are targeted to kill β cells in patients with type 1 diabetes through recognition of a glucose-regulated preproinsulin epitope. J Clin Invest. 2008;118 (10):3390. 10. Kronenberg D, Knight RR, Estorninho M, et al. Circulating preproinsulin signal peptide–specific CD8 T cells restricted by the susceptibility molecule HLA-A24 are expanded at onset of type 1 diabetes and kill β-cells. Diabetes. 2012;61(7):1752-1759. 11. Marron MP, Graser RT, Chapman HD, Serreze DV. Functional evidence for the mediation of diabetogenic T cell responses by HLA-A2. 1 MHC class I molecules through transgenic expression in NOD mice. Proc Natl Acad Sci USA. 2002;99(21):13753-13758. 12. Trudeau JD, Kelly-Smith C, Verchere CB, et al. Prediction of spontaneous autoimmune diabetes in NOD mice by quantification of autoreactive T cells in peripheral blood. J Clin Investig. 2003;111(2):217. 13. Takaki T, Marron MP, Mathews CE, et al. HLA-A* 0201-restricted T cells from humanized NOD mice recognize autoantigens of potential clinical relevance to type 1 diabetes. J Immunol. 2006;176(5):32573265. 14. Kent SC, Chen Y, Bregoli L, et al. Expanded T cells from pancreatic lymph nodes of type 1 diabetic subjects recognize an insulin epitope. Nature. 2005;435(7039):224-228. 15. Toma A, Laïka T, Haddouk S, et al. Recognition of human proinsulin leader sequence by class I-restricted T-cells in HLA-A* 0201 transgenic mice and in human type 1 diabetes. Diabetes. 2009;58 (2):394-402. 16. Bulek AM, Cole DK, Skowera A, et al. Structural basis for the killing of human beta cells by CD8+ T cells in type 1 diabetes. Nat Immunol. 2012;13(3):283-289. 17. Sachdeva N, Paul M, Badal D, et al. Preproinsulin specific CD8+ T cells in subjects with latent autoimmune diabetes show lower frequency and different pathophysiological characteristics than those with type 1 diabetes. Clin Immunol. 2015;157(1):78-90. 18. Karjalainen J, Salmela P, Ilonen J, Surcel H-M, Knip M. A comparison of childhood and adult type I diabetes mellitus. N Engl J Med. 1989;320(14):881-886. 19. Velthuis JH, Unger WW, Abreu JR, et al. Simultaneous detection of circulating autoreactive CD8+ T-cells specific for different islet cellassociated epitopes using combinatorial MHC multimers. Diabetes. 2010;59(7):1721-1730. 20. Luce S, Lemonnier F, Briand J-P, et al. Single insulin-specific CD8+ T cells show characteristic gene expression profiles in human type 1 diabetes. Diabetes. 2011;60(12):3289-3299. 21. Dolton G, Lissina A, Skowera A, et al. Comparison of peptide–major histocompatibility complex tetramers and dextramers for the identification of antigen-specific T cells. Clin Exp Immunol. 2014;177 (1):47-63. 22. Huang J, Zeng X, Sigal N, et al. Detection, phenotyping, and quantification of antigen-specific T cells using a peptide-MHC dodecamer. Proc Natl Acad Sci USA. 2016;113(13):E1890-E1897.

11

23. Wiedeman AE, James EA, Greenbaum C, Long SA. Extensive characterization of islet-reactive CD8 T cells in type 1 diabetes by mass cytometry (CyTOF). J Immunol. 2016;196(suppl 1):54.27. 24. Skowera A, Ladell K, McLaren JE, et al. β-Cell–specific CD8 T cell phenotype in type 1 diabetes reflects chronic autoantigen exposure. Diabetes. 2015;64(3):916-925. 25. Cernea S, Herold KC. Monitoring of antigen-specific CD8 T cells in patients with type 1 diabetes treated with antiCD3 monoclonal antibodies. Clin Immunol. 2010;134(2):121-129. 26. Roep BO, Solvason N, Gottlieb PA, et al. Plasmid-encoded proinsulin preserves C-peptide while specifically reducing proinsulin-specific CD8+ T cells in type 1 diabetes. Science translational medicine. 2013;5 (191):191ra82. 27. Boitard C. T-lymphocyte recognition of beta cells in type 1 diabetes: clinical perspectives. Diabetes Metab. 2013;39(6):459-466. 28. Mallone R, Roep BO. Biomarkers for immune intervention trials in type 1 diabetes. Clin Immunol. 2013;149(3):286-296. 29. Abreu J, Roep BO. Targeting proinsulin-reactive CD8+ T cells: a new direction for type 1 diabetes treatment. Expert Rev Clin Immunol. 2013;9:1001-1003. 30. Orban T, Farkas K, Jalahej H, et al. Autoantigen-specific regulatory T cells induced in patients with type 1 diabetes mellitus by insulin Bchain immunotherapy. J Autoimmun. 2010;34(4):408-415. 31. Bonifacio E, Ziegler A-G, Klingensmith G, et al. Effects of high-dose oral insulin on immune responses in children at high risk for type 1 diabetes: the Pre-POINT randomized clinical trial. JAMA. 2015;313 (15):1541-1549. 32. Tiittanen M, Huupponen JT, Knip M, Vaarala O. Insulin treatment in patients with type 1 diabetes induces upregulation of regulatory Tcell markers in peripheral blood mononuclear cells stimulated with insulin in vitro. Diabetes. 2006;55(12):3446-3454. 33. Pruul K, Kisand K, Alnek K, et al. Differences in B7 and CD28 family gene expression in the peripheral blood between newly diagnosed young-onset and adult-onset type 1 diabetes patients. Mol Cell Endocrinol. 2015;412:265-271. 34. Bollyky JB, Xu P, Butte AJ, Wilson DM, Beam CA, Greenbaum CJ. Heterogeneity in recent-onset type 1 diabetes—a clinical trial perspective. Diabetes Metab Res Rev. 2015;31(6):588-594. 35. Davis AK, DuBose SN, Haller MJ, et al. Prevalence of detectable Cpeptide according to age at diagnosis and duration of type 1 diabetes. Diabetes Care. 2015;38(3):476-481. 36. Zhang J, Gao W, Yang X, et al. Tolerogenic vaccination reduced effector memory CD4 T cells and induced effector memory Treg cells for type I diabetes treatment. PLoS One. 2013;8(7):e70056. 37. Thayer TC, Pearson JA, De Leenheer E, et al. Peripheral proinsulin expression controls low-avidity proinsulin-reactive CD8 T cells in type 1 diabetes. Diabetes. 2016;65(11):3429-3439. 38. Martinuzzi E, Novelli G, Scotto M, et al. The frequency and immunodominance of islet-specific CD8+ T-cell responses change after type 1 diabetes diagnosis and treatment. Diabetes. 2008;57(5):13121320. 39. Takata H, Takiguchi M. Three memory subsets of human CD8+ T cells differently expressing three cytolytic effector molecules. J Immunol. 2006;177(7):4330-4340. 40. Chattopadhyay PK, Betts MR, Price DA, et al. The cytolytic enzymes granyzme A, granzyme B, and perforin: expression patterns, cell distribution, and their relationship to cell maturity and bright CD57 expression. J Leukoc Biol. 2009;85(1):88-97. 41. Ehlers MR, Rigby MR. Targeting memory T cells in type 1 diabetes. Curr Diab Rep. 2015;15(11):1-10. 42. Rigby MR, DiMeglio LA, Rendell MS, et al. Targeting of memory T cells with alefacept in new-onset type 1 diabetes (T1DAL study): 12 month results of a randomised, double-blind, placebo-controlled phase 2 trial. Lancet Diabetes Endocrinol. 2013;1(4):284-294. 43. Paul M, Jacob N, Sachdeva N. Regulatory T cells in treatment of type-1 diabetes: types and approaches. Diabetes Res Open J. 2015;1 (3):54-66.

12

PAUL ET AL.

SUPPORTING INFORMATION Additional Supporting Information may be found online in the supporting information tab for this article.

How to cite this article: Paul M, Badal D, Jacob N, Dayal D, Kumar R, Bhansali A, Bhadada SK, Sachdeva N. Pathophysiological characteristics of preproinsulin-specific CD8+ T cells in subjects with juvenile-onset and adult-onset type 1 diabetes: A 1-year follow-up study. Pediatr Diabetes. 2017;0:1–12. https://doi.org/10.1111/pedi.12536