Article
Reference
Macroautophagy in dendritic cells controls the homeostasis and stability of regulatory T cells
NIVEN, Jennifer, et al.
Abstract
Regulatory T cells (Tregs) play a crucial role in controlling autoimmune and inflammatory responses. Recent studies have demonstrated that dendritic cells (DCs) contribute to the homeostasis of peripheral Tregs. Autophagy, a critical pathway for cellular homeostasis, is active in DCs and is upregulated in different inflammatory conditions. We have shown that Tregs are expanded and have phenotypic alterations and impaired suppressive functions in mice with autophagy-deficient DCs. RNA profiling of Tregs revealed that autophagy in DCs is required to stabilize Treg expression signatures. This phenotype is linked to the downregulation of ICOS-Ligand expression in autophagy-deficient DCs, a consequence of the accumulation of ADAM10, the metalloproteinase responsible for its cleavage. Upon inflammation, in antigen-induced arthritis, mice with autophagy-deficient DCs exhibit increased synovial inflammation and cartilage and bone erosion correlating with Treg-to-Th17 conversion. Our data reveal a mechanism that couples autophagy deficiency in DCs to the function, homeostasis, and stability of Tregs.
NIVEN, Jennifer, et al . Macroautophagy in dendritic cells controls the homeostasis and stability of regulatory T cells. Cell Reports , 2019, vol. 28, no. 1, p. 21-29.e6
DOI : 10.1016/j.celrep.2019.05.110 PMID : 31269441
Available at:
http://archive-ouverte.unige.ch/unige:125385
Disclaimer: layout of this document may differ from the published version.
1 / 1
Cell Reports, Volume28
Supplemental Information
Macroautophagy in Dendritic Cells Controls the Homeostasis and Stability of Regulatory T Cells
Jennifer Niven, Natacha Madelon, Nicolas Page, Assunta Caruso, Guillaume Harlé, Sylvain
Lemeille, Christian A. Seemayer, Stéphanie Hugues, and Monique Gannagé
A B
CD4+ Foxp3+
Absolute count (x105)
Foxp3+ CD25 Neg Foxp3+ CD25 Low
*
Absolute count (x106)
C
Absolute count (x105) 2.0
4.0 6.0
1.0 2.0 3.0
2.0 4.0 6.0
*** *
Foxp3+ CD25 High
Absolute count (x106) 2.5 2.0 1.5 1.0 0.5
ns
Figure S1
2 4 6 8
10 ***
% of Live CD4+
0
Foxp3+ Neuropilin+
% of Live CD4+ LN
CD4+ Foxp3+
0 5 10 15 20
25 *
ns
E
ns ns
Memory T cells (CD62LLow CD44High) Naive T cells
(CD62LHigh CD44 Low)
CD25High CD25Low CD25High CD25Low Foxp3+ CD25 Neg Foxp3+ CD25 Low Foxp3+ CD25 High CD4+ Foxp3+
CD25Low
19.6% CD25Low
40.1%
CD25High 43.6%
CD25High 71.9%
Foxp3 CD25
Central T reg Effector T reg
Central T reg Effector T reg
% of CD4+ Foxp3+ % of thymic SP CD4+
H
% of CD4+ Foxp3+
Central T reg Effector T reg CD69+
% of thymic SP CD4+ % of thymic SP CD4+
% of thymic SP CD4+
Atg5+/+DC Atg5 -/- DC
F
I
% of CD4+ Foxp3Neg 0
2 4 6
0.0 0.5 1.0 1.5
0.0 0.5 1.0 1.5 2.0
0 1 2
ns ns 3 ns
20 40 60 80
0 10 20 30 40 50
0 20 40 60 80 100
0 5 10 15 20 25
% of CD4+ Foxp3Neg
10 20 30 40
0
Foxp3+ ki67+
% of thymic SP CD4+ ns
D
% of CD4+ Foxp3+
Foxp3+ CD25 Neg Foxp3+ CD25 Low Foxp3+ CD25 High
% of CD4+ Foxp3+
% of CD4+ Foxp3+
0 2 4 6 8 10
0 5 10 15 20
25 * **
** Atg5+/+DC
Atg5 -/- DC
40 60 50 90 70 80
G
Atg5+/+DC Atg5 -/- DC
Atg5+/+DC Atg5 -/- DC
Figure S1 (Related to Figure 1): Phenotypic changes of Tregs in mice lacking autophagy in their DCs is specific to the periphery
(A) Flow cytometry analysis of the frequency of CD4+ Foxp3+ cells in peripheral lymph nodes
of Atg5+/+ DC and Atg5-/- DC mice. (B) Absolute counts of the total CD4+ Foxp3+ population in live splenocytes of Atg5+/+ DC and Atg5-/- DC mice. Bar graphs indicate the mean of frequency, each dot represents an individual mouse. Combination of 6 experiments with 3-7 mice/group. (C) Flow cytometry analysis of the % of Foxp3+ Neuropilin-1+ cells within live CD4+ splenocytes in Atg5+/+ DC and Atg5-/- DC mice. (D) Left Panel: % of the 3 different subsets of Tregs according to CD25 expression in peripheral lymph nodes. Right Panel: Absolute count of the 3 different subsets of splenic Tregs according to CD25 expression. Bar graphs indicate the mean of frequency from a combination of 3 individual experiments with 3 to 7 mice/group. (E) Flow cytometry analysis of the % of CD4+ Foxp3+ cells (left) and of the % of Foxp3+ CD25 subsets gated on single positive (SP) thymic CD4+ population. (F) Flow cytometry analysis of the % of Foxp3+ Ki67+ gated on single positive (SP) thymic CD4+ population. Data represents a combination of 3 or 4 individual experiments with 3 to 4 mice/group (G) Representative FACS plots of CD25 expression (low and high) within central (CD62LHigh CD44Low) and effector Tregs (CD62LLow CD44High). Cells are gated on live CD4+ Foxp3+ splenocytes. Bar graph indicates the mean of frequency of CD25 expression in each subset. Data are a combination of 3 individual experiments with 3 mice/group. (H) Flow cytometry analysis of the expression of CD69 in both central (CD62LHigh CD44Low) or effector (CD62LLow CD44High) Treg subsets. Cells are gated on live CD4+ Foxp3+ splenocytes. Data represent a combination of 2 experiments with 3 mice/group. (I) Flow cytometry analysis of the frequency of naïve (CD62LHigh CD44Low) and memory CD4+ population (CD62LLow CD44High) in live CD4+ Foxp3Neg splenocytes of Atg5+/+ DC and Atg5-/- DC mice. Data represent a combination of 3 individual experiments with 3 mice/group.
!
D Post-sort CD25 Low
Live CD4+ Post-sort CD25 High
CD4
CD25CD62L
CD25High 4.87%
CD25Low 2.08%
85.4%
Pre-sort CD4+ CD62L+ Live CD4+
98.2% 98.9%
CD25+
Foxp3+
Gene name FoldChange padj
Igkv8-19 14.93 7.23392E-06
H2-Ab1 0.63 0.039946717
H2-Eb1 0.60 0.018416397
Cd22 0.60 0.037519942
Cd74 0.59 1.66376E-05
S100a4 0.59 0.005741112
H2-Aa 0.58 0.00014325
Baiap3 0.57 0.046995435
Iglc2 0.53 0.012040082
Faim3 0.53 0.043484867
Spib 0.48 0.048571409
H2-DMb2 0.48 0.043633553
Hspa1b 0.47 0.03764566
Ctsh 0.45 0.027890364
Gm4841 0.42 0.012040082
Hspa1a 0.38 0.000212083
Gm4951 0.35 1.92391E-05
Igkv8-24 0.16 0.000212083
Siglech 0.13 0.028709365
Igkv16-104 0.10 0.00014325
Ighv5-12 0.03 7.23392E-06
Ighv1-43 0.02 0.000418818
G
Gene name Fold Change padj
Fancf 2.87 0.010188
Hs6st1 2.33 8.74E-05
Spred1 2.06 0.006503
Akt1 1.92 1.27E-06
Rab21 1.83 1.61E-05
Marveld1 1.79 0.005941
Sppl3 1.76 0.005988
Sirt1 1.70 0.006061
Fam102b 1.70 0.005988
Reps1 1.68 5.04E-06
Mprip 1.67 0.000156
Furin 1.56 0.045001
Fbrsl1 1.55 0.045851
Smad4 1.53 0.000492
Rcor1 1.52 0.006208
Phlda1 0.62 0.018443
Igkc 0.61 0.00088
Cd74 0.58 2.01E-07
Rgs2 0.58 0.007929
Gem 0.55 0.049834
Cd19 0.54 0.00635
H2-Aa 0.53 0.000562
Cr2 0.51 0.008225
Il9r 0.49 0.020313
H2-Ab1 0.45 3.65E-09
Gm4951 0.45 0.006061
Cd22 0.42 2.76E-05
Iglc2 0.40 0.00111
Fosb 0.37 0.031496
Hspa1b 0.27 6.97E-14
Hspa1a 0.22 3.11E-11
Iglv3 0.19 0.000413
Igkv8-21 0.13 0.013723
Igkv8-30 0.07 2.01E-07
Igkv4-86 0.00 1.81E-05
F E
Expression (fold) 100
10-5 10-10
10-15
10-2 10-1 100 101 102 Higher in Atg5-/- DC Higher in Atg5+/+ DC
P (log10)
CD25Low subset Hspa1a
Hspa1b
Figure S2
5 10 15 20
0 10 20 30
5 10 15 20 25 30
20 40 60 80
Foxp3+ CD25 Neg Foxp3+ CD25 Low Foxp3+ CD25 High
% of Live CD4+
CD4+ Foxp3+
% of CD4+ Foxp3+ % of CD4+ Foxp3+ % of CD4+ Foxp3+
p=0.05 p=0.05
** *
Atg14 +/+ DC Atg14 - /- DC
A B
0 20 40 60
20 40 60
% of CD4+ Foxp3+
% of CD4+ Foxp3+
C entral T reg
**
Effector T reg CD62LLow CD44 High CD62LHigh CD44 Low
*
C
Figure S2 (Related to Figure 2): Differential gene expression in CD25High and CD25Low Tregs from Atg5-/- DC mice
(A) Bar graphs showing the mean frequency of CD4+ Foxp3+ cells in live CD4 cells from the spleen of Atg14+/+
DC and Atg14-/- DC mice. Data represent a combination of 2 individual experiments with 3 to 4 mice/group.
(B) Flow cytometry analysis of the 3 different subsets of Tregs according to CD25 expression in splenocytes of Atg14+/+ DC and Atg14-/- DC mice. Data represent a combination of 2 individual experiments with 3 to 4 mice/group. (C) Bar graphs showing the mean frequency of effector Tregs (CD44High CD62LLow) and central Tregs (CD62LHigh and CD44Low) in CD4+ Foxp3+ cells from Atg14+/+ DC and Atg14-/- DC mice.
(D) Representative flow cytometric analysis showing the sorting strategy of Treg subsets. Left panel pre-sort gating strategy: the FACS plot (on the left) is showing the gating strategy after CD4+ T cell isolation by magnetic beads, cells are gated on CD4+CD62L+ subsets and are sorted according to CD25 expression (on the right). Right panel: post sort analysis of Foxp3+ expression in CD25High Tregs (CD4+ CD62L+ CD25High)and CD25Low Tregs (CD4+ CD62L+ CD25Low)populations. (E) Volcano plot comparing differential gene expression in CD25Low Tregs isolated from secondary lymphoid organs of Atg5+/+ DC and Atg5-/- DC mice. Combination of 3 individual repeats with 3/4 mice per group. (F) and (G) List of genes that are differentially expressed in CD25High Tregs (F) and CD25Low Tregs (G)from Atg5+/+ DC compare to Atg5-/- DC mice. Fold changes are labelled in blue (up-regulated in Atg5+/+ DC) and red (down-regulated in Atg5+/+DC). Padj (p adjusted values).
!
0 100 200 300 400 500500 1000 1500
CD4+ FOXP3POS CD4+ FOXP3NEG Isotype ICOSL Ab Isotype ICOSL Ab
*** ns
C on ta ct d ur at ion (s )
2 3 4 5 6 7
ICOS-L Splenic DCs
**
MFI
0.6 0.8 1.0
Relative MFI
ICOS-L BM DCs
*
A
B
Atg14+/+DCAtg14 -/- DC
Figure S3
Figure S3 (Related to Figure 3): ICOS-L expression controls the stability of DC/Treg interactions (A) Total interactions observed between CD4+ Foxp3NEG T cells or CD4+ Foxp3POS Tregs with immature wild type BMDCs. BMDCs were incubated pior to the co-culture with ICOS-L antibody or an isotype control, for 30 minutes. All T cells in the field of view were counted. Combination of 2 individual experiments. ***P<0.0001 (Mann Whitney test). (B) Graphs showing the MFI of ICOS-Ligand expression on splenic DCs (left) or BMDCs (right) from Atg14+/+ DC and Atg14-/- DC mice. Data are from 2 experiments with 2/3 mice per group.
!
Ratio99mTc(mBSA/PBS)
A
B
C
D
Ratio of 99mTc uptake
Day 1 Day 3
ns Relative fold change (2^-∆∆CT)
Relative fold change (2^-∆∆CT) Relative fold change (2^-∆∆CT)
PBS mBSA
PBS mBSA PBS mBSA
10ug mBSA Control
Anti-mBSA Titer Anti-mBSA Titer Anti-mBSA Titer
IL-6 IL-1β IL-23
IFNᵞ
Atg5 +/+ DC Atg5 - /- DC
0 1 2 3
0 1 2 3 4 1000 2000 3000 4000
0 2 4 6 1000 2000 3000 4000
0 2 4 6 8 10 12 14
0 200 400 600 800 1000 20003000
104 105 106 107 108
104 105 106 107 108
102 103 104 105 106 107 108
Total IgG IgG1 IgG2A
IFNᵞ pg/ml
ns ns
Atg5+/+ DC Atg5-/- DC
Figure S4
E
CFSELow% of live CD4+
Atg5 +/+ DC Atg5 - /- DC
0.0 0.5 1.0 1.5 Atg5 +/+ DC
Atg5 - /- DC
Figure S4 (Related to Figure 4): Autophagy deletion in DCs does not affect the innate immune response during antigen-induced-arthritis
(A) Early joint inflammation assessment by external gamma counting of 99mTc up-take in the inflamed knee on d1 and d3 post initiation of arthritis. Results are expressed as the ratio of 99mTc up-take in the left knee (mBSA injected) over the right control knee (PBS injected). Data shown are a combination of 3 experiments with 8 to 10 mice/group. (B) QPCR analysis of pro-inflammatory cytokines expression at d4 post arthritis: mRNA transcripts levels of IL-6, IL-1! and IL-23 are shown. Data shown are a combination of 2 experiments with 4 to 5
mice/group (C) IFN" production from inguinal LN cultures at d7 post arthritis. Cells were cultured in the presence of mBSA (antigen) or PBS (control). Cytokine concentration was determined by ELISA at d3.
Combination of 2 experiments, 6 to 8 mice/group
(D) Antibodies against mBSA under AIA conditions. From the left to the right: total IgG, IgG1 and IgG2A titers of anti-mBSA antibodies in the sera of AIA mice. Data shown are a combination of 3 experiments with 8 to 12 mice/group.
(E) CFSE labelled CD4+ T cell from inguinal LN at d7 post arthritis were cultured in the presence of mBSA (antigen) or PBS (control). Bar graphs represents the % of CD4+ CFSElow cells at 72h post proliferation in the presence of mBSA. One representative experiment with 7-9 mice/group is shown.
