Type 1 Interferons Potentiate Human CD8 + T Cell Cytotoxicity Through a STAT4 and Granzyme B Dependent Pathway Short Running Title: Interferon Induced T cell cytotoxicity Brittney N. Newby 1 , Todd M. Brusko 1 , Baiming Zou 2 , Mark A. Atkinson 1 , Michael Clare-Salzler 1 , and Clayton E. Mathews 1 * 1 -Department of Pathology, Immunology and Laboratory Medicine, University of Florida, Gainesville, FL, USA. 2 -Department of Biostatistics, College of Public Health and Health Professions & College of Medicine, University of Florida, Gainesville, FL, USA. *Corresponding Author: Clayton E. Mathews, PhD. Department of Pathology, Immunology and Laboratory Medicine University of Florida College of Medicine 1275 Center Drive, Biomedical Sciences Building J597 P.O Box 100275, Gainesville, FL 32610. Email Address: [email protected]TEL: (352) 273-9269 ORCID: 0000-0002-8817-6355 Word Count: 4297, Abstract: 144, Figs: 5, Tables: 1, References: 62 Page 1 of 43 Diabetes Diabetes Publish Ahead of Print, published online September 6, 2017
42
Embed
Type 1 Interferons Potentiate Human CD8 B …...Type 1 Interferons Potentiate Human CD8+ T Cell Cytotoxicity Through a STAT4 and Granzyme B Dependent Pathway Short Running Title: Interferon
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Type 1 Interferons Potentiate Human CD8+ T Cell Cytotoxicity Through a STAT4 and Granzyme
B Dependent Pathway
Short Running Title: Interferon Induced T cell cytotoxicity
Brittney N. Newby1, Todd M. Brusko
1, Baiming Zou
2, Mark A. Atkinson
1, Michael Clare-Salzler
1, and
Clayton E. Mathews1*
1-Department of Pathology, Immunology and Laboratory Medicine, University of Florida, Gainesville,
FL, USA.
2-Department of Biostatistics, College of Public Health and Health Professions & College of Medicine,
University of Florida, Gainesville, FL, USA.
*Corresponding Author:
Clayton E. Mathews, PhD.
Department of Pathology, Immunology and Laboratory Medicine
University of Florida College of Medicine
1275 Center Drive, Biomedical Sciences Building J597
autoimmune disease risk variant of STAT4 confers increased sensitivity to IFN-alpha in lupus patients
in vivo. J Immunol 2009;182:34-38
Page 22 of 43Diabetes
Fig. Legends
Fig. 1: Short-term exposure of autoreactive CTL to Type 1 Interferons enhances cytotoxicity toward
beta cells. IGRP-specific CTL-avatars (IGRP-CTL) were exposed to interferon α, β, or γ for 2 hours.
Cytokines were removed by washing and then IGRP-CTL were co-cultured with dispersed primary
human islets (A-C) or BetaLox5 cells (βL5) (D-F) for 16 hours in a standard chromium release assay
(CML). (A-C) Box and whisker plots represent the percentage of dispersed human islets lysed by IGRP-
CTLs. Data for the interferon-mediated change in effector function for each individual T cell donor is
provided using a scatter plot with connecting lines. (D-F) Lines represent the percentage of specific lysis
induced by IGRP-CTL primed with IFNα (1000U/mL: A and D), IFNβ (500U/mL: B and E), or IFNγ
(1000U/mL: C and F) over several effector to target ratios (E:T). (G) IGRP-CTL were primed with
various concentrations of T1-IFN for 2 hours and co-cultured with βL5 at a 10:1 E:T for 16 hours in
CML assays. Data are plotted as mean ± SEM where T cells from each donor (7 donors for A-C, 6
donors for D-F, and 5 donors for G) were weighted equally. There were at least three separate
experiments for each donor. Statistical significance was assessed using a non-parametric paired T-test
with Wilcoxon Post Test Analysis. * P<0.05; ** P<0.01; ***P<0.001
Fig. 2: Granzyme B Expression is increased upon T1-IFN priming of Autoreactive CTLs.
(A-I) IGRP-CTL were primed with IFNα, IFNβ, or IFNγ for 2 hours. Cytokines were removed by
washing and then IGRP-CTL were incubated for an additional 4 hours in media or stimulated by co-
culture with βL5 cells. Representative histograms and mean fluorescence intensities for (A-C) GZMB,
(D-F) IFNγ, and (G-I) Fas Ligand are displayed. (J-M) To the determine the contribution of pathways
important for CTL-mediated killing, IGRP-CTL were co-cultured with βL5 cells (10:1 – E:T) in the
presence of inhibitors known to block CTL cytotoxic function. Bars represent the percentage of βL5 cell
lysis by IGRP-CTLs (J) in the absence of inhibitors, (K) with anti-Fas antibody, (L) with
Concanamycin; and (M) with Pan Caspase Inhibitor, Z-VAD-FMK. For representative histograms (A,
Page 23 of 43 Diabetes
D, and G) all lines within a panel are from a single donor. Data plotted in B, C, E, F and H-M are mean
± SEM where T cells from each donor (7 donors for B, C, E, F, and H-I and 5 donors for J-M) were
weighted equally. There were at least three separate experiments for each donor. Statistical significance
was assessed by a non-parametric paired T-test with Wilcoxon Post Test Analysis. * P<0.05; **
P<0.01; ***P<0.001
Fig. 3: T1-IFN induces phosphorylation of STAT1 and STAT4 in IGRP-CTLs. (A-C) Western blot
analysis of IGRP-CTLs treated with IFNα, IFNβ, or IFNγ for 15 minutes. (A) Representative blots and
densitometry analysis of phosphorylated and total (B) STAT1 and (C) STAT4. (D-F) Time course
analysis of pSTAT1 and pSTAT4 activation is shown by western blot. Densitometry was performed and
are plotted against time for (E) STAT1 and (F) STAT4. Phospho flow cytometry was performed on
IGRP-CTLs treated with IFNα (125U/mL & 1000U/mL) at several time points (5, 15, 30, 60, and
120min). (G) Representative histograms for phosphorylated STAT1 and STAT4 are displayed. (H & I)
Mean fluorescence intensities (MFI) of pSTAT1 and pSTAT4 are plotted against time. (J-L) IGRP-CTL
were primed with T1-IFN for 2 hours and assessed for T-bet and Eomes expression by flow cytometry.
Representative histograms and MFI of (J) T-bet and (K) Eomes is displayed. (L) IGRP-CTL were
pretreated with rapamycin for 16 hours, primed with IFNα for 2 hours, and subsequently assessed for
granzyme B (GZMB) expression by flow cytometry (representative histograms and MFI are plotted).
For representative western blots and histograms (A, D, G, and J-L) all lanes or lines within a panel are
from a single donor. Data plotted in B, C, E, F, and H-L are mean ± SEM; T cells from each donor (at
least 3 donors for B, C, E, F, and L, 5 donors for H and I, and 8 donors for J and K) were weighted
equally. There were at least three separate experiments for each donor. Statistical significance was
assessed by a non-parametric paired T-test with Wilcoxon Post Test Analysis. * P<0.05; ** P<0.01;
***P<0.001
Page 24 of 43Diabetes
Fig. 4: Inhibition of STAT4 reverses T1-IFN induced cytotoxicity. IGRP-CTLs were transfected with
Accell siRNAs specific for STAT1, STAT4 or non-targeting controls for 72 hours. (A-B) RT-PCR
analysis to assess gene silencing was performed for (A) STAT1 and (B) STAT4. All values were
normalized relative to 18S mRNA expression. (C-E) IGRP-CTLs siRNA transfectants were primed with
T1-IFN and assessed for granzyme B (GZMB) expression by flow cytometry. (C) Representative
histograms, and (D) Mean Fluorescence Intensities (MFI) for GZMB are plotted. (E) Viability analysis
on siRNA transfected CTLs is plotted. (F-I) IGRP-CTLs were pre-treated with Lisofylline (LSF) and
analyzed for pSTAT4 activation by (F) western blot and (G) phospho flow cytometry. LSF treated CTLs
were primed with IFNα and analyzed for GZMB expression by flow cytometry. (H) Representative
histograms, and (I) MFI for GZMB are shown. For representative western blots and histograms (C, F,
and H) all lanes or lines within a panel are from a single donor. Data plotted in A, B, D, E, G, and I are
mean ± SEM; T cells from each donor (3 donors for A and B, at least 5 donors for D, E, and G, and at
least 7 donors for I) were weighted equally. There were at least three separate experiments for each
donor. Statistical significance was determined with a non-parametric paired T-test with Wilcoxon Post
Test Analysis. * P<0.05; ** P<0.01; ***P<0.001
Fig. 5: T1-IFN induces pSTAT4 binding to the GZMB promoter to induce transcription. (A-B) qRT-
PCR analysis of (A) GZMB and (B) IFIH1 expression in IGRP-CTLs following 2-hour treatment with
interferons. (C-D) Chromatin immunoprecipitation (CHIP)-qPCR with SYBR Green was performed to
assess the binding of pSTAT1 and pSTAT4 to the GZMB promoter in IGRP-CTLs after treatment with
IFNα for 1 hour. CHIP-qPCR representing the binding of STAT1 and STAT4 to the (C) putative
proximal promoter- and (D) distal promoter-binding sites of GZMB is displayed. Data plotted in are
mean ± SEM; T cells from 3 donors were weighted equally. There were at least three separate
experiments for each donor. Statistical significance was determined by a non-parametric paired T-test
with Wilcoxon Post Test Analysis. * P<0.05; ** P<0.01; ***P<0.001
Page 25 of 43 Diabetes
Tables
Table 1. Peripheral blood donor sex (F, female; M, male) and age as well as CD8+ T cell transduction
efficiency for the experiments reported herein.
CD8+
T Lymphocytes - Donor Information
Sex Quantity Mean Age (years) [Range] Transduction Efficiency (%) [Range]
F 7 27.64 [14.17-38.00] 75.47 [43.6- 93.3]
M 12 29.07 [12.50-46.67] 62.71 [37.5 - 84.2]
Page 26 of 43Diabetes
IFNα
% S
peci
fic L
ysis
Cntl IFNα
IFNβ
% S
peci
fic L
ysis
Cntl IFNβ
IFNγ
% S
peci
fic L
ysis
Cntl IFNγ
* * NS
A. B. C.Figure 1
NS
*
*
**
% S
peci
fic L
ysis
25:1
10:1 5:1 1:1
CntlIFNβ
% S
peci
fic L
ysis
25:1
10:1 5:1 1:1
CntlIFNα
**
*****
*
% S
peci
fic L
ysis
25:1
10:1 5:1 1:1
CntlIFNγ
IFNγIFNβIFNαD. E. F.
% S
peci
fic L
ysis
IFNγ (U/mL)IFNα (U/mL)IFNβ (U/mL)
---
1000--
-32-
-63-
-125
-
-250
-
-500
-
-1000
-
--
16
--
32
--
63
--
125
--
250
--
500
********* ** ** **********
G.
Page 28 of 43Diabetes
Figure 2
CntlIFNβ IFNα
IFNγCntl
IFNβ IFNα
IFNγA.
Ant
igen
St
imul
ated
No
Ant
igen
Total GFP- GFP+
Total CD8+
GZM
B M
FI
B.
No Ant Ant.Stim
GFP+ CD8+
GZM
B M
FI
C.
Granzyme BD.
Total GFP- GFP+
Total CD8+
IFN
γ M
FI
E.
GFP+ CD8+
IFN
γ M
FI
F.
IFNγG. Total GFP- GFP+
Total CD8+
FasL
MFI
H.
GFP+ CD8+Fa
sL M
FII.
Fas Ligand
% S
peci
fic L
ysis
J.
% S
peci
fic L
ysis
K.
% S
peci
fic L
ysis
L.
% S
peci
fic L
ysis
M.
CntlIFNα (125U/mL)IFNα (1000U/mL)IFNβ (500U/mL)
Control anti-FAS CMA Z-VAD-FMK
** **** **
********
* ** **
Isotype ControlA
ntig
en
Stim
ulat
edN
o A
ntig
enA
ntig
en
Stim
ulat
edN
o A
ntig
en
No Ant Ant.Stim
No Ant Ant.Stim
No Ant Ant.Stim
No Ant Ant.Stim
No Ant Ant.Stim
Page 29 of 43 Diabetes
Figure 3IFNγIFNα IFNβ
---
+--
-32-
-125
-
-1000
-
--
500
--
125
--
32pSTAT1
tSTAT1
pSTAT4
tSTAT4
GAPDH
A.
Arb
itrar
y U
nits
IFNγIFNα IFNβ
---
+--
-32 -
-125 -
-1000
-
--
500
--
125
--
32
pSTAT1/STAT1B.
IFNγIFNα IFNβ
---
+--
-32 -
-125 -
-1000
-
--
500
--
125
--
32
Arb
itrar
y U
nits pSTAT4/STAT4
C.
D.
pSTAT1
tSTAT1
pSTAT4
tSTAT4GAPDH
IFNγ (1000 U/mL)IFNα (1000U/mL)
IFNβ(500U/mL)
---
+--
-+-
--+
---
+--
-+-
--+
---
+--
-+-
--+
E.
Arb
itrar
y U
nits
pSTAT1/STAT1
0 20 40 60Time (min)
F.
Arb
itrar
y U
nits
pSTAT4/STAT4
0 20 40 60Time (min)
Cntl IFNα (125U/mL) IFNα (1000U/mL)
pSTAT1 pSTAT4
5’
15’
30’
60’
120’
pSTAT1
pSTA
T1 M
FI
G. H.
0 30 9060Time (min)
120
pSTAT4
pSTA
T4 M
FI
0 30 9060Time (min)
120
I.
J. No Antigen
AntigenStimulated
% o
f Max
T-Bet
K. NoAntigen
AntigenStimulated
% o
f Max
EOMES
T-B
et M
FINo
AntigenAntigen
StimulatedEO
MES
MFI
NoAntigen
AntigenStimulatedL.
Control 1nM Rapamycin 1µM Rapamycin
Granzyme B
% o
f Max
Rapamycin Cntl 1nM 1µM
GZM
B n
MFI NS
NS
CntlIFNα
*****
**
*** ** **
****
*
CntlIFNγIFNαIFNβ
*
*******
***
**** **
*
*
***
** ** *
15 min 30 min 60 min
CntlIFNγ IFNα IFNβ
Page 30 of 43Diabetes
GZM
B M
FI
D.
NT ControlSTAT1 siRNASTAT4 siRNA
---
---
+--
+--
-+-
-+-
--+
--+
IFNαCntl
***
NS
Figure 4
% o
f Max
C.
Granzyme B
Media NT Control STAT1 STAT4IFNαCntl IFNα + siRNA
+--
% V
iabi
lity
E.
NT ControlSTAT1 siRNASTAT4 siRNA
---
---
+--
-+-
-+-
--+
--+
CntlIFNα
Fold
Cha
nge
A.
Cntl
STAT1STA
T4 NT
**
H.
Granzyme B
% o
f Max
0 U/mL 125 U/mL 1000 U/mLIFNαCntl IFNα + LSF
Fold
Cha
nge
B.
Cntl
STAT1STA
T4 NT
**
pSTAT4-Tyr693
F.
tSTAT4
GAPDH
Lisofylline
IFNα (U/mL) 1250 1000
- + - + - +
GZM
B M
FI
I.
IFNα (U/mL) 0 125 1000
†***
*
†
†CntlLSF
pSTA
T4 M
FI
G.
IFNα (U/mL) 0 32 125 1000
CntlLSF
††
†† ††
* * *
Page 31 of 43 Diabetes
Fold
Cha
nge
A.
CntlIFNγIFNαIFNβ
GZMB
***
Fold
Cha
nge
B.
CntlIFNγIFNαIFNβ
IFIH1
**
% o
f Inp
ut
C.
**
Proximal PromoterCntlIFNα
IgG
pSTAT1
pSTAT4
tSTA
T1
tSTA
T4
% o
f Inp
ut
D.Distal Promoter
CntlIFNα
IgG
pSTAT1
pSTAT4
tSTA
T1
tSTA
T4
Figure 5Page 32 of 43Diabetes
Supplemental Table 1
Antibody Information
Page 33 of 43 Diabetes
Supplemental Table 2
Taq Man Probes
Probe Probe ID
STAT1 Hs01013996_m1
STAT4 Hs01028017_m1
Granzyme B Hs00188051_m1
GAPDH Hs02758991_g1
18S Hs03003631_g1
Page 34 of 43Diabetes
Supplemental Table 3
CHIP Primer Sequences
Antibody Clone
GZMB Promoter Primer Pair 1 Forward 5’-tcacttcataggcttgggttcct -3’
GZMB Promoter Primer Pair 1 Reverse 5’- ctctgggtgcttgtgtgagaatc -3’
GZMB Promoter Primer Pair 2 Forward 5’-ctgtgagcctgttatgtgctgag -3’
GZMB Promoter Primer Pair 2 Reverse 5’- ggacgtttgtggtgctaaattgc -3’
Page 35 of 43 Diabetes
Supplemental Table 4. IFNα and IFNβ pretreatment of human IGRP-CTL for 2 hours increases the lytic function of these cells. These data accompany Figures 1D-1F. Data reported in this table are the averages of the technical replicates for the individual donor. These averages were used to generate the Mean + SEM reported in Figures 1D-1F. E:TRatio=25:1Treatment Donor1 Donor2 Donor3 Donor4 Donor5 Donor6Control 44.50 24.80 68.40 64.60 18.10 49.78IFNα 76.30 63.50 88.40 74.20 50.50 72.21IFNβ 60.35 69.70 88.80 74.40 52.00 69.89IFNγ 50.30 24.20 65.20 75.30 18.77 54.76 E:TRatio=10:1Treatment Donor1 Donor2 Donor3 Donor4 Donor5 Donor6Control 19.30 12.02 33.80 47.20 7.20 21.58IFNα 37.90 34.30 62.50 65.80 20.30 28.88IFNβ 26.37 30.30 66.20 65.10 21.30 25.21IFNγ 22.40 11.12 31.50 51.70 7.40 21.66 E:TRatio=5:1Treatment Donor1 Donor2 Donor3 Donor4 Donor5 Donor6Control 9.20 7.70 19.11 24.00 3.80 9.02IFNα 20.50 15.60 35.60 35.00 12.00 17.43IFNβ 13.71 15.90 36.70 36.00 9.30 17.60IFNγ 11.80 6.80 18.70 28.10 5.20 10.14 E:TRatio=1:1Treatment Donor1 Donor2 Donor3 Donor4 Donor5 Donor6Control 2.70 2.80 6.10 1.05 0.63 1.38IFNα 4.30 5.20 9.40 1.80 1.25 1.95IFNβ 1.98 4.80 7.70 1.60 0.82 1.12IFNγ 4.10 2.30 4.20 6.80 0.01 0.79
Page 36 of 43Diabetes
Supplemental Figure 1
Protocol for Generating Human CTL Avatars
Supplemental Figure 1. Schematic representation of protocol for generating human CTL avatars. Peripheral blood samples are obtained from consenting human subjects. Samples undergo negative selection and FACs cell sorting for isolation of naïve CD8+ T lymphocytes. These cells are subsequently activated in vitro using CD3/CD28 conjugated beads, which act to mimic TCR and co- stimulatory signals that trigger T cell proliferation and differentiation. After 48 hours, T cells are transduced with lentiviral vectors which permanently integrate sequences for antigen specific TCRs into the genome of the lymphocyte. Newly transduced avatars are expanded over an additional 7 days to yield a large pool of antigen-specific CTLs for study.
2 days
Page 37 of 43 Diabetes
Supplemental Figure 2IGRP Specific CD8+ T cells
Unstained IGRP265-273 Dextramer-PE
eGFP
MART122-35 Dextramer-PE
PE PE PE
Donor 1
Unstained IGRP265-273 Dextramer-PE
eGFP
MART122-35 Dextramer-PE
PE PE PE
Donor 2
Unstained IGRP265-273 Dextramer-PE
eGFP
MART122-35 Dextramer-PE
PE PE PE
Donor 3
Unstained IGRP265-273 Dextramer-PE
eGFP
MART122-35 Dextramer-PE
PE PE PE
B. MART-1 Specific T cells
A.
Supplemental Figure 2. Human CTL Avatars are Antigen Specific. Following expansion, human CTL avatars were stained with PE-conjugated dextramers and subjected to flow cytometric analysis. (A) Dot plots of IGRP-specific CTL avatars from three donors are displayed. Dextramer binding as displayed by PE fluorescence is compared to eGFP expression, which serves as a fluorescent tag for positively transduced IGRP CTLs. IGRP-specific CTLs selectively bind to the IGRP265-273 dextramer, but show no binding to the MART-122-35 dextramer. (B) Additonally, MART-1-specific CTL avatars were generated,stained, and analyzed for dextramer binding by flow cytometry. While the lentiviral construct used to generate these human CTL avatars does not contain a fluorescent tag, these cells selectively bind the MART-122-35 dextramer serving as a positive control for the MART-122-35 dextramer.
Page 38 of 43Diabetes
Supplemental Figure 3
Supplemental Figure 3. CTL avatars display primarily an effector/ effector memory phenotype with marked levels of CD27 expression. (Α & B). The gating strategy for the surface characterization of PBMCs, which were used as a gating control, as well as the CTL avatars. (C) Compares the percentage of live CD8+ T cells that show differential profiles for CD45RA and CD45RO expression. (D) Displays the percentage of live CD8+ T cells that are positive or negative for CCR7. (E) Compares the percentage of live CD8+ T cells that show differential profiles for CD27 and CD28 expression.
Live CD8s
Live CD8s
Page 39 of 43 Diabetes
Supplemental Figure 4
Supplemental Figure 3. CTL Avatars are characterized by high levels of granzyme B and robust production of IFNγ and TNFα following PMA/Ionomycin stimulation. Α). The gating strategy used for gating of live CD8+ T cells. B-D) Compares the cytotoxin and cytokine profiles for control PBMCs, unstimulated CTL avatars, and CTL avatars stimu-lated with PMA and Ionomycin. E) Compares the percentage of live stimulated CTL avatars that are positive for Granzyme B (GZMB), IFNγ, TNFα, IL-17A, IL-5, and IL10.
Page 40 of 43Diabetes
Supplemental Figure 5
A. Gating Strategy for Live Cells
B. C. ControlIFNγ IFNα IFNβ
4 hours 16 hours
Supplemental Figure 5. Viability Analysis of IGRP CTLs following IFN-primingViability analysis was performed on IGRP-CTLs using flow cytometry following 2-hour cytokine priming. Representative Gating Strategy shown (A). Viability determined by Live/Dead Viability dye staining after 2 hour IFN-priming and 4 hour incubation (B) or 16 hour incubation (C) in media.
Page 41 of 43 Diabetes
Isotype ControlMedia ControlIFNα (1000U/mL)
pSTAT3 pSTAT5 pSTAT6
0’
5’
15’
30’
60’
120’
APC PE PE
A.
B.
PE
Isotype ControlIL-2 (1000U/mL)
pSTAT5
Supplemental Figure 6
Supplemental Figure 6. Phospho Flow Cytometry for pSTAT3, pSTAT5, and pSTAT6 following T1-IFN Treatment. (A) Phospho Flow Cytometry was performed on IGRP-CTLs treated with IFNα (1000U/mL) over several time points (5, 15, 30, 60, and 120min) to characterize T1-IFN STAT signaling. Representative histograms for phosphorylated STAT3, STAT5, and STAT6 are displayed. (B) IGRP-CTLs were treated with IL-2 (1000U/mL) for 30 minutes and assessed for phosphorylated STAT5 by phospho flow cytometry. A representative histogram is displayed. (C) IGRP-CTLs were treated with IL-4 (1000U/mL) for 30 minutes and assessed for phosphorylated STAT6 by phospho flow cytometry. A representa-tive histogram is displayed.
C.
PE
Isotype ControlIL-4 (100ng/mL)
pSTAT6
Page 42 of 43Diabetes
Supplemental Figure 7
Supplemental Figure 7. Information regarding Primer Pairs used for Chromatin ImmunoprecipitationSchematics detailing the sequences, expected size product and position of the primer pairs within the promoter of granzyme B are displayed (A & B).