1Laboratory of Immune Regulation, Institute of Pharmaceutical Sciences, Seoul National University, Seoul, Republic of Korea. 2BK21 Plus program, College of Pharmacy, Seoul National University, Seoul, Republic of Korea. 3Laboratory of Toxicology, College of Pharmacy, Seoul National University, Seoul, Republic of Korea. 4Center for Integrative Rheumatoid Transcriptomics and Dynamics, The Catholic University of Korea, Seoul, Republic of Korea. 5Division of Cardiology, Department of Internal Medicine, Severance Hospital, College of Medicine, Yonsei University, Seoul, Republic of Korea. *e-mail: [email protected]
Atherosclerosis is a chronic inflammatory disease within the subendothelial area of arteries and is one of the leading causes of death worldwide1. In addition to the critical role of
hyperlipidemia in atherogenesis, it is well demonstrated that both innate immune cells and adaptive immune cells contribute to the development of atherosclerosis2. For example, macrophages and dendritic cells (DCs) recognize fatty materials such as long-chain saturated fatty acids and modified low-density lipoprotein (LDL) through scavenger receptors and Toll-like receptors (TLRs) and secrete inflammatory cytokines and chemokines, which aggravate vascular inflammation. In addition, the TH1 and TH17 subsets of helper T cells are found in the lesion; TH1 cells seem to be proath-erogenic, whereas the role of TH17 cells remains controversial3,4.
There is a higher incidence of atherosclerosis among patients with systemic autoimmune disorders; these atherosclerosis-related autoimmune diseases include psoriasis, rheumatoid arthritis and systemic lupus erythematosus (SLE), all of which are known to be mediated by autoreactive T cells5–7. Moreover, cholesterol-lowering treatments, such as statins and a low-fat diet, have been shown to ameliorate psoriasis and SLE, suggestive of a detrimental role for hyperlipidemia in the autoimmune diseases8,9. Among the athero-sclerosis-associated autoimmune diseases, psoriasis and SLE are thought to be mediated by TH17 cells10,11 and follicular helper T cells (TFH cells)12, respectively, whereas both helper T cell subsets seem to contribute to rheumatoid arthritis13. It has been demonstrated, through the use of animal models of experimental autoimmune encephalomyelitis, that atherogenic hyperlipidemia promotes auto-immune TH17 cell responses in vivo14. Whether atherogenic hyper-lipidemia also affects autoimmune TFH cell responses has remained poorly understood.
While insufficient TFH cell responses result in inadequate host defense against infection, aggressive TFH cell responses can result
in autoimmunity. The differentiation and maintenance of TFH cells, therefore, must be tightly orchestrated to ensure sufficient B cell responses without inducing autoimmunity. Patients with SLE exhibit an increased frequency of CXCR5+PD-1+ TFH cells15. The disease activity of SLE, including the SLE disease-activity index and the level of antibodies to double-stranded DNA (anti-dsDNA), is positively associated with the levels of circulating triglycerides and cholesterol16,17. Despite the evident association between athero-sclerosis and autoimmune SLE, little is known about whether and how the atherogenic environment affects the development of auto-immune TFH cell responses in vivo. In the present study, we found that atherogenic dyslipidemia promoted autoimmune and immu-nization-induced germinal-center (GC) reactions in a TLR4- and IL-27-dependent manner.
ResultsAtherogenic dyslipidemia augments autoimmune lupus. As a first step in investigating the role of atherogenic dyslipidemia in autoimmune lupus, we determined if the development of auto-antibodies and a lupus phenotype was affected by an atherogenic environment in vivo. We transferred bone marrow cells from lupus-prone BXD2 mice into wild-type (WT) mice or apolipo-protein E–deficient (Apoe–/–) mice in which the bone marrow had been ablated, to generate WTBXD2 mice and ApoEBXD2 mice, respec-tively. After a 6-week reconstitution period, the recipients were fed a high-fat diet (HFD) for another 6 weeks to induce hyperlipidemia (Fig. 1a). The two groups exhibited similar weights throughout the experiment (data not shown). The concentration of total immuno-globulin G (IgG) and IgG2c was significantly higher in the serum of ApoEBXD2 mice than in that of WTBXD2 mice, but the concentra-tion of total IgG1 was not (Fig. 1b). More notably, the level of IgG autoantibodies to dsDNA and rheumatoid factor IgG, particularly
Atherogenic dyslipidemia promotes autoimmune follicular helper T cell responses via IL-27Heeju Ryu1,2, Hoyong Lim1, Garam Choi1,2, Young-Jun Park1,2, Minkyoung Cho1, Hyeongjin Na1,2, Chul Won Ahn2,3, Young Chul Kim2,3, Wan-Uk Kim4, Sang-Hak Lee5 and Yeonseok Chung 1,2*
The incidence of atherosclerosis is higher among patients with systemic lupus erythematosus (SLE); however, the mechanism by which an atherogenic environment affects autoimmunity remains unclear. We found that reconstitution of atherosclerosis-prone Apoe–/– and Ldlr–/– mice with bone marrow from lupus-prone BXD2 mice resulted in increased autoantibody production and glomerulonephritis. This enhanced disease was associated with an increase in CXCR3+ follicular helper T cells (TFH cells). TFH cells isolated from Apoe–/– mice had higher expression of genes associated with inflammatory responses and SLE and were more potent in inducing production of the immunoglobulin IgG2c. Mechanistically, the atherogenic environment induced the cytokine IL-27 from dendritic cells in a Toll-like receptor 4 (TLR4)-dependent manner, which in turn triggered the differentia-tion of CXCR3+ TFH cells while inhibiting the differentiation of follicular regulatory T cells. Blockade of IL-27 signals diminished the increased TFH cell responses in atherogenic mice. Thus, atherogenic dyslipidemia augments autoimmune TFH cell responses and subsequent IgG2c production in a TLR4- and IL-27-dependent manner.
IgG2c, was significantly higher in the former group than in the lat-ter (Fig. 1c,d). Among mice maintained on a normal chow diet, the levels of autoantibodies in WTBXD2 mice were comparable to those in ApoEBXD2 mice (Supplementary Fig. 1a,b), indicative of a crucial role for atherogenic dyslipidemia in augmenting the production of autoantibodies. IgG2c autoantibodies are known to be more patho-genic than IgG1 in SLE due to their ability to activate the comple-ment pathway and immunoglobulin receptors of the Fcγ R family18. Histological assessment of the kidneys indicated that ApoEBXD2
mice fed an HFD exhibited more-severe glomerulonephritis, with expanded double layers around glomerulus, crescentic tubules and deposition of IgG immunocomplexes, than that of WTBXD2 mice fed an HFD (Fig. 1e,f).
To determine if the enhanced autoimmune lupus-like pheno-types could be induced in LDL receptor–deficient (Ldlr–/–) mice, another widely used animal model of atherosclerosis, we performed a similar bone marrow–transfer study (Fig. 1g) and found that Ldlr–/– recipients of BXD2 bone marrow (LDLRBXD2 mice) exhibited
BXD2 BM cells
a b c
fd
g
j k l
h i
e
800*
**
*** ****** *
**
**
600
400
WTBXD2
*
**
*
**
**
**
******
**
** ****
*
*
1.5
1.0
0.5
0.0
1.5
1.0
0.5
0.0
200
0
8
6
4
2
0
1.5
1.0
0.5
0.0
1.52.5 1.5
1.0
0.5
0.0
2.0
1.5
1.0
0.5
0.0
1/10 1/50 1/250 1/10 1/50 1/250
Serum dilution
Serum dilution
BXD2 BM cells1,500
1,000
500
0HDL LDL TG VLDL
BM-ablated recipients
Lipi
d (m
g/dl
)
IgG
(m
g/m
l)
IgG
1 (m
g/m
l)
IgG
2c (
mg/
ml)
WT
HFD
Ldlr–/–
WTBXD2 LDLRBXD2
WTBXD2 LDLRBXD2
WTBXD2 LDLRBXD2
Tub
ule
area
(%
)
0
1
2
3
***
Tub
ule
area
(%
)
0
1
2
3
4
**
IgG
MF
I (A
U)
0
1
2
3WTBXD2 LDLRBXD2
WTBXD2
LDLRBXD2
0.8
0.6
0.4
0.2
0.0
0.8
0.6
0.4
0.2
0.01/10 1/50 1/250 1/10 1/50 1/250
8
6
4
2
0
1.5 2.5
2.0
1.5
1.0
0.5
0.0
1.0
0.8
0.6
0.4
0.2
0.0
0.8
0.6
0.4
0.2
0.01/10 1/50 1/250 1/10
Serum dilution
Serum dilution
1/50 1/250
1.0
0.5
0.0
1/10 1/50 1/250 1/10 1/50 1/250
1.0
0.5
0.0HDL LDL TG VLDL
Lipi
d (m
g/dl
)
IgG
(m
g/m
l)
IgG
1 (m
g/m
l)
IgG
2c (
mg/
ml)
Ant
i-dsD
NA
IgG
O.D
. (A
U)
WTBXD2
BM-ablated recipients
WT
WTBXD2
HFD
Apoe–/–
ApoEBXD2
ApoEBXD2 WTBXD2 ApoEBXD2
ApoEBXD2WTBXD2
ApoEBXD2
RF
IgG
O.D
. (A
U)
Ant
i-dsD
NA
IgG
1 O
.D. (
AU
)
Ant
i-dsD
NA
IgG
2c O
.D. (
AU
)
Ant
i-dsD
NA
IgG
O.D
. (A
U)
RF
IgG
O.D
. (A
U)
Ant
i-dsD
NA
IgG
1 O
.D. (
AU
)
Ant
i-dsD
NA
IgG
2c O
.D. (
AU
)
*
IgG
MF
I (A
U)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Fig. 1 | Autoantibodies and autoimmune glomerulonephritis in Apoe–/– and Ldlr–/– recipients of BXD2 bone marrow. a,g, Experimental procedure (left): WT mice and Apoe–/– mice (a) or Ldlr–/– mice (g) in which the bone marrow was ablated (BM-ablated recipients) were given intravenous transfer of bone marrow cells from lupus-prone BXD2 mice (BXD2 BM cells) and were fed an HFD. Right, concentration of high-density lipoprotein (HDL), LDL, triglycerides (TG) and very low-density lipoprotein (VLDL) (horizontal axis) in plasma from mice as at left (key). b,h, ELISA of total IgG, IgG1 and IgG2c antibodies in the serum of mice as in a (b) or g (h). c,i, ELISA of autoantibodies to dsDNA and rheumatoid factor (RF) in the serum (dilution, horizontal axis) of mice as in a (c) or g (i), presented (in arbitrary units (AU)) as optical density (O.D.) measured at 450 nm. d,j, ELISA of autoreactive IgG-subclass autoantibodies to dsDNA in the serum (dilution, horizontal axis) of mice as in a (b) or g (h), presented as in c,i. e,k, H&E staining (left) of the kidneys of mice as in a (e) or g (k), and quantification of tubule area (right) in the kidneys of mice as at left. Scale bars (left), 100 μ m. f,l, Immunofluorescence staining (left) for IgG in the kidneys of mice as in a (f) or g (l), and mean fluorescent intensity (MFI) of IgG (right) in the kidneys of mice as at left. Scale bars (left), 100 μ m. Each symbol represents an individual mouse. *P < 0.05, **P < 0.01 and ***P < 0.001, compared with WTBXD2 mice (unpaired Student’s t-test). Data are from one experiment representative of four independent experiments with n = 5 mice per group (a–d,g–j; box plot in a,g (center line, median; box limits, minimum to maximum); mean + s.e.m. in b,h) or are representative of four independent experiments with two kidney sections per mouse (n = 5 per group) (e,k; mean + s.e.m. at right) or three kidney sections per mouse (n = 5 per group) (f,l).
significantly larger amounts of IgG2c autoantibodies and more-severe glomerulonephritis than that of their WT counterparts (WTBXD2 mice) (Fig. 1h–l). Together these results demonstrated that the atherogenic environment within Apoe–/– and Ldlr–/– mice aug-mented the production of pathogenic IgG2c autoantibodies, which led to severe autoimmune lupus in our experimental settings in vivo.
The atherogenic environment promotes enhanced TFH cell responses and GC reactions. Since we observed a significantly greater abundance of autoantibodies in ApoEBXD2 mice, we next assessed GC reactions spontaneously generated in the recipients of BXD2 bone marrow. We observed a greater abundance of pea-nut agglutinin–positive GCs in the spleen of ApoEBXD2 mice than in that of WTBXD2 mice (Fig. 2a). The frequency of GL7+CD95+ GC B cells, as well as that of B220–CD138+ plasma cells, was also higher in ApoEBXD2 mice than in WTBXD2 mice (Fig. 2b and Supplementary Fig. 1c–g). Moreover, the frequency of PD-1+CXCR5+Foxp3– TFH cells was significantly higher in ApoEBXD2 mice than in WTBXD2 mice (Fig. 2c). The higher frequency of GC B cells and TFH cells in the atherogenic mice resulted in an increased ratio of TFH cells or GC B cells to follicular regulatory T cells (TFR cells) in the ApoEBXD2 mice. TFH cells can be subcategorized into three subsets on the basis of their surface expression of the chemokine receptors CXCR3 and CCR619. The frequency and number of cells in the CXCR3+CCR6– TFH cell population was significantly higher in ApoEBXD2 mice than in WTBXD2 mice, while differences in the abundance of the CXCR3–CCR6– and CXCR3–CCR6+ subpopulations were relatively marginal (Fig. 2d). Linear-regression analysis revealed that the level of anti-dsDNA IgG exhibited a positive correlation with the num-ber of CXCR3+CCR6– TFH cells (Fig. 2e). Moreover, the number of CXCR3+CCR6– TFH cells was strongly correlated with the abun-dance of IgG2c autoantibodies (Fig. 2f).
To determine if the proatherogenic environment also affected immunization-induced GC reactions, we immunized HFD-fed WT and Apoe–/– mice with chicken type II collagen in complete Freund’s adjuvant (CFA). The level of antigen-specific IgG2c, as well as the frequency of GC B cells, plasma cells and TFH cells in the draining lymph nodes, was significantly higher in Apoe–/– mice than in WT mice (Supplementary Fig. 2a,b). We obtained similar results for mice immunized with 4-hydroxy-3-nitrophenyl (NP)-conjugated ovalbumin (NP-OVA) in CFA. The greater production of NP-specific IgG2c in Apoe–/– mice was more prominent for high-affinity types than for those of global affinity (Supplementary Fig. 2c–e). Collectively, these results demonstrated that a proathero-genic environment resulted in enhanced TFH cells, particularly of the CXCR3+CCR6– subset, and GC reactions in lupus-prone mice as well as in WT mice after immunization with exogenous antigens.
TFH cells in atherogenic mice exhibit distinct gene expression and function. We next determined if proatherogenic conditions also affected the B cell–stimulatory function of TFH cells. We sorted TFH cells by flow cytometry from HFD-fed WT or Apoe–/– mice after immunization with keyhole limpet hemocyanin (KLH) in CFA (Fig. 3a) and co-cultured them with naive B cells in the presence of antibody to the invariant signaling protein CD3 and anti-IgM. TFH cells from Apoe–/– mice were far more potent in inducing IgG pro-duction from naive B cells than were those from WT mice (Fig. 3b), indicative of a qualitative change in TFH cells in the former group, in addition to the observed quantitative increase in the TFH cell population. Analysis of IgG subclasses revealed that the observed increase in IgG production was due mainly to an increase in IgG2c, not IgG1. Among TFH cell subsets obtained from HFD-fed Apoe–/– mice, CXCR3+ TFH cells were more potent than CXCR3− TFH cells in inducing IgG2c production from naive B cells (Supplementary Fig. 3a–c), an effect that was significantly attenuated by antibody to interferon-γ (IFN-γ ) (Supplementary Fig. 3d–f). The number
of TH1 cells and TH17 cells was also greater in ApoEBXD2 mice than in WTBXD2 mice (Supplementary Fig. 3g); however, CD44+CD4+ non-TFH cells failed to induce IgG2c production from B cells (Supplementary Fig. 3h), which suggested that rather than non-TFH cells (such as TH1 cells), TFH cells were responsible for the greater production of IgG2c in the atherogenic mice.
To further characterize the pathogenic features of TFH cells gen-erated in atherogenic conditions, we performed global gene-expres-sion analysis by RNA-based next-generation sequencing (RNA-seq) of TFH cells. At a threshold of a change in expression of greater than twofold, a P value of < 0.05 and a false discovery rate of < 0.1, 211 genes were upregulated and 142 genes were downregulated in TFH cells from Apoe–/– mice relative to their expression in TFH cells from WT mice (Fig. 3c, Supplementary Fig. 3i and Supplementary Table 1). Gene-ontology-terms analysis revealed significant enrich-ment for genes encoding molecules associated with the inflam-matory response, SLE and the IFN-γ -related pathway in TFH cells from the atherogenic (Apoe–/–) mice, relative to their expression in WT mice, a result confirmed by quantitative RT-PCR (Fig. 3d,e). Moreover, TFH cell signature genes, including Bcl6, Cxcr5 and Slamf5, were also slightly but significantly upregulated in TFH cells from the atherogenic (Apoe–/–) mice relative to their expression in TFH cells from WT mice (Fig. 3f). In parallel with the increased abundance of CXCR3+CCR6– TFH cells (Fig. 2d), Ifng and Cxcr3 were also upregu-lated in TFH cells from Apoe–/– mice relative to their expression in TFH cells from WT mice (Fig. 3f). These results together indicated that TFH cells generated in an atherogenic environment in vivo exhibited a distinct gene-expression profile and provided more-potent ‘help’ to B cells for the production of IgG2c.
Atherogenic dyslipidemia promotes IL-27 production by CD11b+ DCs. To explore the underlying basis of the augmented TFH cell responses in atherogenic mice, we analyzed T cell–stimu-latory cytokines, including IL-1β , IL-6, IL-10, IL-12, IL-23, IL-27 and TNF, in serum from WTBXD2 and ApoEBXD2 mice. The concen-tration of IL-10, IL-23 and TNF was below the limit of detection, while the concentration of IL-12p40 was comparable in the two groups. In contrast, the concentration of IL-1β , IL-6 and IL-27 was significantly higher in ApoEBXD2 mice than in WTBXD2 mice (Fig. 4a). Feeding mice an HFD induced little increase in IL-27 in the serum of WTBXD2 mice relative to its abundance in that of normal chow–fed WTBXD2 mice (Supplementary Fig. 4a), which indicated that the increase in IL-27 in the HFD-fed ApoEBXD2 mice was due to atherogenic hyperlipidemia rather than to obesity. IL-27 showed a stronger positive correlation with the number of TFH cells than did IL-6 (Supplementary Fig. 4b,c). The increased abundance of IL-27 was from cells of hematopoietic origin, since both Apoe–/– mice and Ebi3–/–Apoe–/– mice (atherogenic mice that lack IL-27) that were given WT bone marrow showed a similar concentration of IL-27 in serum (Supplementary Fig. 4d–f). Consistent with the increased abundance of IL-27, the expression of Ebi3 and Il27 (which encode the EBI3 subunit and p28 subunit, respectively, of IL-27) and, to a lesser extent, that of Il12a was much higher in DCs (B220–F4/80–
MHCIIhiCD11c+) from ApoEBXD2 mice than those from the WTBXD2 mice (Fig. 4b). In contrast, the expression of Il12b, Il23 and Il6 in the two groups appeared to be similar (Fig. 4b). When stimulated with lipopolysaccharide (LPS), DCs from the ApoEBXD2 mice secreted more IL-27 and, to a lesser extent, IL-6 than did DCs from the WTBXD2 mice, while differences in the production of IL-12p40 were not observed (Fig. 4c).
Conventional DCs can be subcategorizied into two populations on the basis of their surface expression of CD11b (integrin α M) and the co-receptor CD8α 20. Notably, we observed a significantly greater frequency and number of CD11b+ DCs in the spleen of ApoEBXD2 mice than in that of WTBXD2 mice (Fig. 4d). This DC subset had higher levels of Ebi3 and Il27 transcripts than did the CD8α + subset,
particularly in ApoEBXD2 mice (Fig. 4e). Consistent with that, the CD11b+ DCs from Apoe–/– mice were far more potent producers of IL-27 than were CD8α + DCs, after stimulation with LPS (Fig. 4f). Among CD11b+ DCs, the CD4– subset was more abundant than the CD4+ subset and had higher expression of Ebi3 than that of the CD4+ subset (Fig. 4g,h). These results demonstrated that the environment within atherogenic mice not only promoted the accu-mulation of CD4–CD11b+ DCs but also significantly enhanced the capacity of these DCs to produce IL-27.
The receptors TLR4 and LXR regulate IL-27 production from DCs under atherogenic conditions. We next sought to determine the molecular mechanism by which the atherogenic environment triggered the production of IL-27 from DCs. Atherogenic lip-ids such as oxidized LDL and long-chain saturated fatty acids can be recognized by the receptors TLR4, LOX-1 and CD3614,21. The expression of Tlr2, Tlr4 and Olr1 (which encodes LOX-1) was sig-nificantly higher in CD11b+ DCs from Apoe–/– mice than in those from WT mice (Supplementary Fig. 4g). To determine if TLR4 had
a role in the enhanced production of IL-27 and GC reactions in atherogenic mice, we generated bone marrow chimeras by transfer-ring bone marrow cells from WT or TLR4-deficient mice into WT or Apoe–/– mice in which the bone marrow was ablated, to generate WTWT, ApoEWT and ApoETLR4 mice, and fed those host mice an HFD after a reconstitution period (Fig. 5a). Consistent with the observa-tions noted for ApoEBXD2 mice (Fig. 4a), we observed a significantly higher concentration of IL-6 and IL-27 in serum from ApoEWT mice than in that of WTWT mice, while the concentration IL-12p40 remained similar in these mice (Fig. 5b). In contrast, the higher concentration of IL-6 and IL-27 was not observed in ApoETLR4 mice, indicative of an essential role for TLR4 on hematopoietic cells in the increased production of these cytokines in the atherogenic mice. Oxidized LDL substantially induced the production of IL-6 and IL-27, but not that of IL-12p40, by DCs from the ApoEWT mice, in a dose-dependent manner, relative to such production by DCs from the WTWT mice, and this was significantly attenuated in the DCs from ApoETLR4 mice, relative to that in DCs from ApoEWT mice (*P < 0.05, **P < 0.01 and ***P < 0.001 (unpaired Student’s t-test);
WTBXD23
***** *** ***
** *
2
1
0
4
2
3
1
0
4 15 5
4
3
2
1
0
2.0
1.5
1.0
0.5
0.0
10
5
0
2
3
1
0
ApoEBXD2
WTBXD2a
d
e f
b c
ApoEBXD2
WTBXD2
2.90
43.7
0 102 103 104
80
*
**
**
*
60
40
20
0
50
30
10
Cel
ls (
× 1
04 )
40
20
0CXCR3+
CCR6–CXCR3–
CCR6–CXCR3–
CCR6+CXCR3+
CCR6–CXCR3–
CCR6–CXCR3–
CCR6+
0
103
104
105
CC
R6-
PE
-Cy7
CXCR3-PE
2.0
1.5
1.0
0.5
0.00 10 20 30 40 50
CXCR3+CCR6– TFH cells (× 104)
Ant
i-dsD
NA
IgG
50.2
1.15
59.2
37.1
ApoEBXD2
GC
B c
ell f
ollic
les/
mm
2
GL7
+C
D95
+ c
ells
(%
)
B22
0– CD
138+
cel
ls (
%)
PD
-1+C
XC
R5+
Fox
p3– c
ells
(%
)
TF
H/T
FR
Cel
ls (
%)
GC
B /T
FR
2.0
1.5
1.0
0.5R2 = 0.6985P = 0.0026
R2 = 0.3964P = 0.0511
R2 = 0.7431P = 0.0013
0.00 10 20 30
CXCR3–CCR6– TFH cells (× 104)
Ant
i-dsD
NA
IgG
0.8
0.6
0.4
0.2
0.00 10 20 30 40 50
CXCR3+CCR6– cells (× 104)
Ant
i-dsD
NA
IgG
2cR2 = 0.3044P = 0.0983
0.8
0.6
0.4
0.2
0.00 10 20 30
CXCR3–CCR6– cells (× 104)
Ant
i-dsD
NA
IgG
2cFig. 2 | Analysis of GC reactions and TFH cell responses in Apoe–/– recipients of BXD2 bone marrow. a, GC staining (left) in the spleen of bone marrow chimeras generated and treated as described in Fig. 1 (left margin): green, peanut agglutinin; red, IgD. Scale bars, 200 μ m. Right, quantification of GC B cell follicles in the spleen of mice as at left (key). b,c, Frequency of GL7+CD95+ cells (GC B cells) and B220–CD138+ cells (plasma cells) (b) and PD-1+CXCR5+Foxp3– cells (TFH cells), as well as the ratio of TFH cells to TFR cells (TFH/TFR) and of GC B cells to TFR cells (GC B/TFR) (c), among splenocytes from mice as in a (key in a), assessed by flow cytometry. d, Flow cytometry (left half) of cells from the spleen mice as in a (above plots), and frequency (middle) and number (right) of each TFH cell subset (horizontal axis) in the spleen of mice as at left (key in a). e,f, Linear-regression analysis of the abundance of each TFH cell subset (horizontal axis) and the level of anti-dsDNA IgG (e) or IgG2c (f) in WTBXD2 and ApoEBXD2 mice. Each symbol represents an individual mouse; small horizontal lines (a–d) indicate the mean (+ s.e.m.)). *P < 0.05, **P < 0.01 and ***P < 0.001, compared with WTBXD2 mice (unpaired Students t-test). Data are from one experiment representative of four independent experiments with three spleen sections per mouse (n = 5 per group) (a) or n = 5 mice per group (b–f).
Supplementary Fig. 5a). Moreover, while the concentration of total IgG and IgG2c, but not that of IgG1, was significantly higher in serum from ApoEWT mice than in that of WTWT mice, their con-centrations in ApoETLR4 mice were indistinguishable from those in WTWT mice (Fig. 5c). When these mice were immunized with NP-OVA in CFA, the frequency of GC B and TFH cells in the drain-ing lymph nodes was significantly higher in ApoEWT mice than in WTWT mice, and the frequency of these cells in ApoETLR4 mice was comparable to that of WTWT mice (Fig. 5d). Similarly, although the amount of NP-specific total IgG and IgG2c in ApoETLR4 mice was higher than that of WTWT mice, it was significantly lower than that of ApoEWT mice (Fig. 5e).
On the other hand, analysis of energy metabolism (by measure-ment of the oxygen-consumption rate and extracellular acidifica-tion rate) revealed that DCs from HFD-fed Apoe–/– mice exhibited a significantly higher glycolysis, glycolytic capacity and respira-tory reserve capacity than that of DCs from HFD-fed WT mice
(Supplementary Fig. 5b–f), indicative of robust upregulation of energy-generating pathways in DCs exposed to an atherogenic environment. Moreover, BODIPY-staining analysis showed that the intracellular lipid content was greater in DCs from ApoEBXD2 mice than in those from WT mice, a result that was not observed for T cells and B cells (Fig. 5f and Supplementary Fig. 5g); this indi-cated abnormal lipid metabolism in the DCs. Receptors of the LXR (‘liver X receptor’) family have a central role in the regulation of intracellular lipid metabolism22,23. Notably, the abundance of Nr1h2 transcripts (which encode LXRβ ) and LXRβ protein, but not that of Nr1h3 transcripts (which encode LXRα ), was significantly lower in DCs from Apoe–/– mice than in those from WT mice (Fig. 5g and Supplementary Fig. 5h–j). To determine if the reduced expression of LXRβ contributed to the increased production of IL-27 in athero-genic mice, we measured the production of IL-27 by DCs stimulated in the presence of the LXR agonist GW396524. GW3965 signifi-cantly reduced the expression of Ebi3 and Il27 while increasing
2.0
1.5
1.0
0.5
0.0
20
15
10
5
0
Hist1h
3g
S100a
8Ifit
m1
Ifitm
2Bcl6
Cxcr5
Slamf5 ll2
1 ll4 lfng
Cxcr3
–5–10 0
EASE score (–log10)
5 10
Systemic lupus erythematosusInflammatory response
Toll-like receptor 4 binding
Cholesterol biosynthesis processCholesterol metabolic process
Cellular response to interferon-gammaCellular response to lipopolysaccharide
Response to interferon-gammaPositive regulation of inflammatory response
Positive regulation of interferon-gamma production
20
25
15
10
5
0
WT
**
** ***
* *
**
* **
***20
25
4
3
mR
NA
exp
ress
ion
(rel
ativ
e)
mR
NA
exp
ress
ion
(rel
ativ
e)
2
1
0
4
3
2
1
0
4 Apoe
Apoe 1
Ldlr
Ifng Cxcr3
Ifitm2
Ifitm1
Bcl6
S100A8
–4 –2
3
2
2 4
P v
alue
(–l
og10
)
1
00
Expression (log2 fold)
15
10
5
0
HFDKLH-CFA immunization
Flow-sortedCD4+PD-1+CXCR5+ T cell
Apoe–/–WT Apoe–/–
IgG
(ng
/ml)
IgG
1 (n
g/m
l)
IgG
2c (
ng/m
l)
IgG
2c/Ig
G1
a
c
f
d
e
b WT
Hist1h3g
Fig. 3 | Transcriptomic and functional analyses of TFH cells isolated from atherogenic mice. a, Experimental procedure (left): WT or Apoe–/– mice were immunized with KLH-CFA and fed an HFD, then TFH cells from those mice were sorted by flow cytometry (as CD4+PD-1+CXCR5+) and analyzed further. b, ELISA of total IgG, IgG1 and IgG2c, as well as the ratio of IgG2c to IgG1 (IgG2c/IgG1), in supernatants of TFH cells from mice as in a (key) and naive B cells co-cultured for 7 d in the presence of anti-CD3 and anti-IgM. c, Gene expression plotted against P values, presented as a volcano plot showing the expression (log2 values) of genes in TFH cells from Apoe–/– mice as in a relative to their expression in in TFH cells from WT mice as in a (and the significance of that difference): red, P < 0.05; orange, change in expression of over onefold; green, P < 0.05 and change in expression of over onefold; black, neither P < 0.05 nor a change in expression of over onefold. d, Gene-ontology analysis of mice as in a, showing the EASE score (Expression Analysis Systematic Explorer software application) of genes encoding molecules in upregulated (red) or downregulated (green) pathways (left margin). e, Quantitative RT-PCR analysis of mRNA from selected genes (horizontal axis) in TFH cells as in a; results were normalized to those of the control gene Actb. f, Quantitative RT-PCR analysis of mRNA from TFH cell–related genes (horizontal axis) in TFH cells as in a; results normalized as in e. Each symbol (b,e,f) represents an individual mouse. *P < 0.05 and **P < 0.01, compared with WT (unpaired Student’s t-test). Data are from one experiment representative of three independent experiments with n = 5 mice per group (b; mean + s.e.m.) or are representative of two independent experiments with n = 4 mice per group (c–f; box plots (e,f) as in Fig. 1a).
the expression of Srebf1 (which encodes the transcription factor SREBP-1c), a well-known target of LXRs (Fig. 5h). Consequently, the production of IL-27 by the DCs from Apoe–/– mice was sig-nificantly diminished by treatment with GW3965 (Fig. 5i). These results indicated that TLR4 was required for the increased produc-tion of IL-27, and that restoration of LXRβ activity diminished that increase in the production of IL-27 by DCs from atherogenic mice.
IL-27 is essential for the enhanced TFH cell and GC reactions in ath-erogenic mice. To determine the role of IL-27 in the increase in TFH cells and GC reactions in atherogeic mice, we comparatively analyzed HFD-fed WT, Apoe–/– and Ebi3–/–Apoe–/– mice. As expected, IL-27 was absent from the serum of Ebi3–/–Apoe–/– mice, whereas IL-1β , IL-6 and IL-12p40 were marginally affected (Fig. 6a). Although total IgG and IgG2c were consistently higher in the serum of Apoe–/– mice than in that of WT mice, in Ebi3–/–Apoe–/– mice, their concentration was the same as that seen in the WT mice (Supplementary Fig. 6a).
When the mice were immunized with NP-OVA in CFA, NP-specific IgG and IgG2c antibodies were higher in Apoe–/– mice than in WT mice, but Ebi3–/–Apoe–/– mice showed no such elevation in anti-body titers (Fig. 6b). Moreover, the frequency of GC B cells and TFH cells in Ebi3–/–Apoe–/– mice was lower than that in Apoe–/– mice, while it was comparable to the frequency of such cells in WT mice (Fig. 6c). Similar results were obtained for mice immunized with chicken type II collagen (Supplementary Fig. 6b,c). Unlike those from the Apoe–/– mice, TFH cells from the Ebi3–/–Apoe–/– mice did not trigger the production of IgG2c from naive B cells in vitro (Fig. 6d) or exhibit increased expression of genes encoding molecules asso-ciated with inflammatory responses and SLE, including Hist1h3g, S100a8 and Ifitm1, or TFH cell–related genes (Supplementary Fig. 6d,e). The number of CXCR3+CCR6– TFH cells was significantly lower in Ebi3–/–Apoe–/– mice than in Apoe–/– mice (Fig. 6e).
To demonstrate the role of IL-27 more definitively, we studied mice deficient in the cytokine receptor IL-27R (Il27ra–/– mice).
a
c d e
f g h
bWTBXD2 ApoEBXD2
100
80
60 *
IL-6
(pg
/ml)
40
20
0
250
200
150
*
IL-6
(pg
/ml)
100
**
IL-2
7 (p
g/m
l)
WT
N.D.
150
100
50
0
CD8α+ CD11b+
Apoe–/–
WT
Apoe–/
–W
T
Apoe–/
–
**
IL-2
7 (p
g/m
l)
250
200
150
100
50
0
***
CD
4+ D
Cs
(%)
80
60
40
20
0
**
CD4–
CD11b+CD4+
CD11b+
DC
s (×
103 )
40
20
30
10
0E
bi3
expr
essi
on (
rela
tive)
II27
expr
essi
on (
rela
tive)
** **
**
CD4–
CD11b+CD4+
CD11b+
2
4
6
0
** **
CD4–
CD11b+CD4+
CD11b+
3
2
1
0
IL-1
2p40
(pg
/ml)
100
80
60
40
20
0
**
**
CD8α+ CD11b+
DC
s (%
)
80
60
40
20
0
**
CD8α+ CD11b+
DC
s (×
103 )
80
100
60
40
20
0
**
Ebi
3 ex
pres
sion
(re
lativ
e)
II27
expr
essi
on (
rela
tive)
CD8α+ CD11b+
30
20
10
0
*
**
CD8α+ CD11b+
4
3
2
1
0
**
IL-2
7 (p
g/m
l)
100
80
60
40
20
0
IL-1
2p40
(pg
/ml)
100
80
60
40
20
0
*
IL-1
β (p
g/m
l)
100
80
60
40
20
0
*
****
mR
NA
exp
ress
ion
(rel
ativ
e)
6
4
2
0
II12a
II12b
II13a
Ebi3 II27 II6
Fig. 4 | Atherogenic dyslipidemia induces IL-27 production by DCs. a, Concentration of various cytokines in the serum of WTBXD2 and ApoEBXD2mice (key). Data are from one experiment representative of three independent experiments with n = 5 mice per group (mean + s.e.m.). b, Quantitative RT-PCR analysis of mRNA (horizontal axis) in the CD11c+ DCs of mice as in a (key in a); results normalized to those of Actb. Data are representative of two independent experiments with n = 4 mice per group (box plots as in Fig. 1a). c, Concentration of IL-27 secreted, in the presence of LPS (100 ng/ml), by CD11c+ DCs of mice as in a (key in a). Data are from one experiment representative of two independent experiments with n = 5 cultures per group (mean + s.e.m.). d, Frequency (assessed by flow cytometry) and absolute number of DC subsets (horizontal axis) in mice as in a (key in a). Data are from one experiment representative of three independent experiments with n = 5 mice per group. e, Quantitative RT-PCR analysis of Ebi3 and Il27 mRNA in splenocytes obtained from mice as in a (key in a), then sorted by flow cytometry into subsets (horizontal axis); results normalized as in b. f, Concentration of IL-27 secreted, in the presence of LPS (100 ng/ml), by various subsets of cells (above plots) from HFD-fed WT and Apoe–/– mice (key). Data are representative of two independent experiments with n = 4 mice per group (e,f; box plots (e) as in Fig. 1a; mean + s.e.m. in f). g, Frequency of CD4+ cells among CD11b+ DCs (assessed by flow cytometry) (left), and number of CD4–CD11b+ or CD4+CD11b+ DCs (right), in mice as in f (key in f). Data are from one experiment representative of two independent experiments with n = 5 mice per group. h, Quantitative RT-PCR analysis of Ebi3 and Il27 mRNA in CD4–CD11b+ and CD4+CD11b+ DCs (horizontal axis) sorted by flow cytometry from HFD-fed WT or Apoe–/– mice (key as in f); results normalized as in b. Data are representative of two independent experiments with n = 4 mice per group (box plots as in Fig. 1a). Each symbol represents an individual mouse; small horizontal lines (d,g) indicate the mean (+ s.e.m.). *P < 0.05 and **P < 0.01, compared with WTBXD2 mice (a–d) or WT mice (e–h) (unpaired Student’s t-test).
Apoe–/– mice in which the bone marrow was ablated were given adoptive transfer of WT or Il27ra–/– bone marrow cells before being fed an HFD, and the recipient mice were immunized with KLH in CFA. We found a significantly lower abundance of anti-KLH IgG, especially IgG2c, and TFH cells, particularly the CXCR3+ subset, in the recipients of Il27ra–/– bone marrow than in those given WT bone marrow (Fig. 6f–h). When we performed a simi-lar experiment with mixed–bone marrow chimeras generated by the transfer a 1:1 mixture of WT (CD45.1+) bone marrow cells and Il27ra–/– (CD45.2+) bone marrow cells into Apoe–/– mice, we observed a similarly lower frequency of CXCR3+ TFH cells among Il27ra–/– CD4+ T cells (Supplementary Fig. 6f), indicative of a T cell–intrinsic role for IL-27 signal in promoting the differentia-tion of CXCR3+ TFH cells in atherogenic environment. In contrast, the frequency of TFR cells was higher in the Il27ra–/– CD4+ T cell population than in the WT CD4+ T cell population (Fig. 6i and Supplementary Fig. 6g). Injection of anti-IL-6 also significantly diminished the frequency of TFH cells, GC B cells and plasma cells in Apoe–/– mice immunized with KLH in CFA (Supplementary Fig. 6h–k). Anti-IL-6 also decreased the frequency of TH17 cells, but not that of TH1 cells. However, unlike the blockade of IL-27 sig-naling, treatment with anti-IL-6 decreased not only the production
of IgG2c and the number of CXCR3+ TFH cells but also the produc-tion of IgG1 and the number of CXCR3– TFH cells.
Since IL-27 induces signaling via the transducers STAT1 and STAT325, we next investigated the role of these STAT proteins in the enhanced TFH cell responses in atherogenic mice. We generated mixed–bone marrow chimeras by transferring a 1:1 mixture of WT (CD45.1+) bone marrow cells and either Stat1–/– bone marrow cells or bone marrow cells from mice with conditional deletion of Stat3 in CD4+ T cells (CD4Stat3 mice) (all CD45.2+) into Apoe–/ –mice, then fed the recipients an HFD and subsequently immunized them with KLH in CFA. The frequency of TFH cells was lower among STAT1- or STAT3-deficient CD4+ T cells than among their respective WT counterparts (Supplementary Fig. 6l). When we analyzed subsets among the TFH cell population, we observed a significant reduction in the CXCR3+ subset among STAT1-deficient TFH cells but not among STAT3-deficient TFH cells (Fig. 6j). When CXCR3–PD-1+CXCR5+ TFH cells sorted by flow cytometry were stimulated with IL-27, STAT1-deficient T cells failed to induce the expression of Cxcr3 or Ifng, while STAT3-defient T cells showed no defect in inducing those genes but failed to induce Il21 (Fig. 6k). These results demonstrated an essential role for IL-27 in the differentiation of CXCR3+ TFH cells and subse-quent GC reactions in atherogenic mice (Supplementary Fig. 7).
a b c
d f
g h i
WTWT ApoEWT ApoETLR4
WT or TIr4–/– BM cells
WT
WTWT ApoEWT ApoETLR4
HFDApoe–/–
WT Apoe–/–
ApoEBXD2
WTBXD2
100
50
0
Eve
nts
(% o
f max
)
BODIPY
***** ***
0
1
2
3
4
GL7
+C
D95
+ c
ells
(%
)
1.2
0.0
0.4
0.8
Nr1
h3 e
xpre
ssio
n (r
elat
ive)
**1.5
1.0
0.5
0.0Nr1
h2 e
xpre
ssio
n (r
elat
ive)
***
**
WTApo
E 0.1
0.2 1
3
2
1
0
GW3965 (μM)
Sre
bf1
expr
essi
on (
rela
tive) ***
**6
4
2
0
GW3965 (μM)
Ebi
3 ex
pres
sion
(re
lativ
e) ****** **
3
4
5
2
1
0
GW3965 (μM)
II27
expr
essi
on (
rela
tive)
** ***300
200
100
0
GW3965 (μM)
Il-27
(pg
/ml)
e*** ***
0
1
2
3
4
5
PD
-1+C
XC
R5+
Fox
p3– c
ells
(%
)
*
* *
2.0##
####
1.5
1.0
0.5
0.01/10 1/30 1/90
IgG
O.D
. (A
U)
1.5
1.0
0.5
0.01/10 1/30Serum dilution
1/90
IgG
1 O
.D. (
AU
)
#
***
*
######
#
##
1/10 1/30 1/90
2.0
1.5
1.0
0.5
0.0
IgG
2c O
.D. (
AU
)
**2,500
2,000
1,500
1,000
500
0
BO
DIP
Y M
FI (
AU
)
100
** **
120
80
60
40
20
0IL
-6 (
pg/m
l)
* *100
120
80
60
40
20
0
IL-2
7 (p
g/m
l)
100
120
80
60
40
20
0
IL-1
2p40
(pg
/ml)
***3
2
1
0
IgG
2c (
mg/
ml)
* **0.5
0.4
0.3
0.2
0.1
0.0
IgG
1 (m
g/m
l)
* *3
2
1
0
IgG
(m
g/m
l)
WTApo
E 0.1
0.2 1
WTApo
E 0.1
0.2 1
WT
ApoE 0.
10.
2 1
Fig. 5 | TLR4 and LXR regulate IL-27 production by DCs from atherogenic mice. a, Experimental procedure: WT or Apoe–/– mice in which the bone marrow was ablated were given intravenous transfer of bone marrow cells from WT or Tlr4–/– mice, then were fed an HFD and immunized with OVA-NP in CFA. b, Concentration of cytokines in the serum of mice as in a (key). c, ELISA of total IgG, IgG1 and IgG2c antibodies in the serum of mice as in a (key in b). d, Frequency of GC B cells (left) and TFH cells (right) in the draining lymph nodes of mice as in a (key in b), assessed by flow cytometry. e, ELISA of NP-specific IgG, IgG1 and IgG2c antibodies the serum (dilution, horizontal axis) of mice as in a (key in b). f, Flow cytometry of CD11c+ DCs obtained from WTBXD2 and ApoEBXD2 mice (key) and stained with the fluorescent dye BODIPY (left), and mean fluorescent intensity of BODIPY in those cells (right). g, Quantitative RT-PCR analysis of Nr1h3 and Nr1h2 mRNA in CD11c+ DCs isolated from the spleen of WT or Apoe–/– mice (key); results normalized to those of Actb. h, Quantitative RT-PCR analysis of Srebf1, Ebi3 and Il27 mRNA in CD11c+ DCs as in g (key in g), stimulated with LPS (100 ng/ml) in the presence of increasing concentrations (wedge; horizontal axis) of GW3965; results normalized as in g. i, Concentration of IL-27 produced by CD11c+ DCs as in g (key in g) after stimulation as in h (horizontal axis). Each symbol (b–i) represents an individual mouse; small horizontal lines (d) indicate the mean (+ s.e.m.). * or #P < 0.05, ** or ##P < 0.01 and ***P < 0.001, between the two groups (all panels except e) or compared with ApoEWT (* in e) or WTWT (# in e) (unpaired Student’s t-test). Data are from one experiment representative of two (b–e,g–i) or four (f) independent experiments with n = 4 mice per group (b–e) or n = 5 mice per group (f–i).
Patients with hyperlipidemia have elevated serum IL-27 and anti-bodies. To demonstrate the clinical relevance of our findings, we measured the serum concentration of IL-6 and IL-27 in a cohort of patients with hypercholesterolemia and healthy donors. Although total cholesterol and LDL cholesterol were significantly higher in the patients, both groups exhibited comparable body mass index, medi-cal comorbidities and use of medication (Fig. 7a and Supplementary Table 2). The plasma concentration of IL-6 was similar in the two groups; however, the plasma concentration of IL-27 was significantly higher in the patients than in the healthy donors (Fig. 7b). Moreover, the plasma from the patients exhibited a slightly but significantly
higher concentration of IgG autoantibodies to dsDNA (Fig. 7c). The concentration of total IgG was also significantly higher in these patients, presumably due to an increase in IgG1 and IgG3 rather than in IgG4 (Fig. 7d). The IgG1 and IgG3 subclasses in humans are homologs of IgG2 subclasses in mice26. These results demonstrated that patients with hypercholesterolemia exhibited elevated IL-27 associated with an increase in plasma IgG1 and IgG3.
DiscussionBeyond the key role of hyperlipidemia and fatty materials in athero-sclerosis, a growing body of evidence indicates a pathogenic role for
Fig. 6 | IL-27 is necessary for the increased TFH cell responses and GC reactions in atherogenic mice. a, Concentration of cytokines in the serum of WT, Apoe–/– and Ebi3–/–Apoe–/– mice (key) fed an HFD for 4 weeks and then immunized with OVA-NP in CFA. b, ELISA of NP-specific IgG, IgG1 and IgG2c antibodies in the serum (dilution, horizontal axis) of mice as in a (key in a). c, Frequency of GC B cells (left) and TFH cells (right) in mice as in a (key in a), assessed by flow cytometry. d, ELISA of IgG, IgG1 and IgG2c, and the ratio of IgG2c to IgG1, in TFH cells sorted by flow cytometry from mice as in a (key in a) and then co-cultured with naive B cells. N.D.,. not detected. e, Absolute number of cells in various TFH cell subsets (horizontal axis) in mice as in a (key in a). f, ELISA of KLH-specific IgG, IgG1 and IgG2c antibodies in the serum (dilution, horizontal axis) of Apoe–/– mice in which the bone marrow was ablated, that were reconstituted by intravenous transfer of bone marrow cells from WT mice (ApoEWT) or Il27ra–/– mice (ApoEIl27ra) (key) and then fed an HFD before being immunized with KLH in CFA. g–i, Frequency of GC B cells and TFH cells (g), subsets of TFH cells (horizontal axis) (h) and TFR cells (i) in mice as in f (key in f), assessed by flow cytometry. j, Frequency of various subsets of TFH cells (horizontal axis) in Apoe–/– mice in which the bone marrow was ablated, that were reconstituted by intravenous transfer of mixture of bone marrow cells from congenic WT mice or Stat1–/– mice (left) or CD4Stat3 mice (right) (donor in key) and then fed an HFD; cells were assessed by flow cytometry. k, Quantitative RT-PCR analysis of Cxcr3, Ifng and Il21 mRNA in CXCR3– TFH cells isolated from WT, Stat1–/– or CD4Stat3 mice (above plots) and stimulated with IL-27; results normalized to those of Actb. Each symbol represents an individual mouse; small horizontal lines (c,e,g–j) indicate the mean (+ s.e.m.).*P < 0.05, **P < 0.01 and ***P < 0.001, between groups (unpaired Student’s t-test). Data are from one experiment representative of three independent experiments with n = 5 mice per group (a–e and f–i) or one representative experiment of two independent experiments with n = 4 mice per group (j) or are representative of two independent experiments with n = 4 mice per group (k).
these conditions in autoimmune diseases, including lupus. Given the fact that cardiovascular disease is a leading cause of death and disabil-ity27, it is important to study cross-regulatory mechanisms that govern the pathogenesis of atherosclerosis and related autoimmune diseases. Using in vivo animal models of atherosclerosis and autoimmune lupus, we demonstrated that atherogenic dyslipidemia significantly increased the production of autoantibodies, particularly IgG2c, and the severity of SLE by enhancing TFH cell responses. Our findings have identified a previously unknown mechanism by which atherogenic dyslipidemia promotes TFH cell responses and GC reactions to lupus-associated self antigens as well as exogenous antigens via IL-27 in vivo.
IL-6, IL-12 and IL-27 are known to induce TFH cell differentia-tion by inducing IL-21 production from CD4+ T cells28,29. We found that the abundance of IL-27 in serum was significantly higher in patients with hypercholesterolemia as well as in atherogenic Apoe–/– or Ldlr–/– mice. IL-27 was originally reported to promote TH1 immu-nity30; however, it is also known to augment TFH cell responses and GC reactions by activating STAT3 in T cells28,31. More notably, we observed that blockade of IL-27 signal in atherogenic mice almost completely abolished the increased production of IgG2c and the gen-eration of CXCR3+ TFH cells to the levels similar to those of WT mice. Studies of T cells deficient in STAT1 or STAT3 revealed that STAT1 was required for the expression of Cxcr3 and Ifng by TFH cells in atherogenic mice in a cell-intrinsic manner, in agreement with pub-lished studies32,33. IL-6 production was also increased in atherogenic mice; however, we propose that the increased IL-27, rather than the increased IL-6, could have accounted for the enhancement of TFH cell and GC reactions, on the basis of our findings that (i) blockade of IL-27 signal specifically reduced the features of GC reactions in atherogenic mice; (ii) neutralization of IL-6 non-specifically reduced overall GC reactions, including CXCR3– TFH cells and IgG1 produc-tion; and (iii) IL-27, not IL-6, was increased in patients with hyper-cholesterolemia. Our study also proposes a role for IL-27 signaling in regulating TFR cells responsible for controlling GC reactions34,35.
Consistent with the findings from a sanroque lupus model36, IFN-γ production by CXCR3+ TFH cells appeared to be critical for
the induction of IgG2c from B cells. TH1 cells have been shown to mediate the generation of low-affinity neutralizing antibodies in the absence of TFH cells in a viral infection model37. Although we can-not rule out the possibility of a contribution from TH1 cells to the increased antibody production in atherogenic mice, our observation that TFH cells isolated from the atherogenic mice were far superior to non-TFH cells in stimulating IgG2c production from B cells strongly suggested that TFH cells had a major role in the enhanced antibody production in atherogenic mice. The substantial increase in the pro-duction of high-affinity IgG2c and formation of GCs further sup-ported the proposal of a crucial role for TFH cells in the enhanced antibody production in atherogenic mice. IgG2 subclasses are more pathogenic than IgG1 or IgG3 in murine models of autoimmunity, due to their ability to activate the complement pathway and to bind to activating Fc receptors18. In this context, it is noteworthy that patients with hypercholesterolemia exhibited higher levels of IgG1 and IgG3, which suggested that atherogenic dyslipidemia might also promote the production of IgG1 and IgG3 via IL-27 in humans.
How does atherogenic dyslipidemia increase IL-27 in circula-tion? We propose that there are multiple mechanisms by which an atherogenic environment induces IL-27 production from DCs. First, we observed a significant increase in the CD8α −CD11b+ DC subset in the atherogenic mice. CD11b+ DCs are known to be more potent than CD11b– DCs in the secretion of IL-6 and IL-2738,39. Notably, DCs from the atherogenic mice had lower expression of Nr1h2 (which encodes LXRβ ), and activation of LXRs significantly reduced the production of IL-27 by the DCs, which would sug-gest that reduced expression of LXRβ contributed to the increased production of IL-27 by DCs in atherogenic mice. That hypothesis is supported by a published study showing that LXR-deficient DCs upregulate their production of pro-inflammatory cytokines40. Hence, we propose that an atherogenic environment ‘preferen-tially’ increases CD8α −CD11b+ DCs and triggers IL-27 produc-tion from them by downregulating LXRβ expression. Second, the fatty materials with increased abundance in atherogenic mice stimulate IL-27 production by DCs. Oxidized LDL and saturated
Controla
d
b cPatient
400 ***
300
200LD
L-C
(m
g/dl
)
100
0
60
40
20
0
***
IgG
(m
g/m
l)
25
20
15
10
5
0
**
IgG
1 (m
g/m
l)
800
600
400
200
0
***
IgG
3 (μ
g/m
l)
80
60
40
20
0
IgG
4 (μ
g/m
l)
30
10
0
20
IL-6
(pg
/ml)
***1.0
0.8
0.6
0.4
0.2
0.0
Ant
i-dsD
NA
IgG
*300
200
100
0
IL-2
7 (p
g/m
l)Fig. 7 | Patients with hypercholesterolemia exhibit increased autoantibodies and IL-27. a, Concentration of LDL cholesterol (LDL-c) in the plasma of healthy donors (Control) or patients with hypercholesterolemia (key). b, Concentration of cytokines in the plasma of subjects as in a (key in a). c, Level of autoreactive IgG antibodies to dsDNA in subjects as in a (key in a). d, Concentration of total IgG and IgG-subclass antibodies in subjects as in a (key in a). Each symbol represents an individual subject. *P < 0.05, **P < 0.01 and ***P < 0.001, compared with healthy donors (unpaired Student’s t-test). Data are from one experiment representative of two technical replicates with n = 24 subjects per group (box plots as in Fig. 1a).
long-chain fatty acids are known to induce IL-27 production by DCs41,42. Our study here identified a critical role for TLR4 in the increase in IL-27 in atherogenic mice, which would suggest that fatty materials trigger IL-27 production from DCs directly or syn-ergistically in concert with other pathogen-associated molecular patterns via TLR4. Third, DCs from atherogenic mice exhibited increased expression of pattern-recognition receptors, including TLR2, TLR4 and LOX-1. DCs with higher expression of pattern-recognition receptors are probably more sensitive to pathogen-associated molecular patterns for the production of cytokines, including IL-2743,44. Moreover, given that TLR4 signaling sup-presses LXR activity23, the increased expression of TLR4 might have a role in downregulating LXRβ expression in DCs. Finally, the upregulated energy-generating metabolic pathways observed in the DCs from atherogenic mice might increase IL-27 produc-tion, since it has been proposed that energy metabolism signifi-cantly affects cytokine production by DCs45,46. Further studies will be needed to determine details of the molecular mechanisms by which TLR4, LXRβ and energy metabolism regulate IL-27 produc-tion in atherogenic environment.
An elevated level of IL-27 is found in the atherosclerotic plaques of patients with coronary artery disease. It upregulates expression of the integrin ligands ICAM-1 and VCAM-1, IL-6, and the che-mokines CCL5 and CXCL10 in arterial endothelial cells, which leads to the infiltration of inflammatory immune cells into the lesions47,48. The role of IL-27 in lupus remains controversial. IL-27 is positively correlated with the renal SLE disease activity index49, and lack of the IL-27 receptor ameliorates disease severity in the sanroque mouse model of lupus31. In contrast, administration of IL-27 or enhancing IL-27 signaling ameliorates the severity of lupus in an animal model50. Our findings support the hypoth-esis that IL-27 is pathogenic in lupus because it increases TFH cell responses and promotes the production of pathogenic subclasses of IgG in atherogenic environments in vivo. In summary, our study has identified the ‘hyperlipidemia–TLR4–IL-27–CXCR3+ TFH cell’ axis as a previously unknown mechanism by which the atherogenic environment promotes GC reactions, which might explain the tight association between atherosclerosis and SLE in humans. Targeting IL-27, therefore, might be effective for the treatment of antibody-mediated autoimmune diseases, including SLE, in patients with hypercholesterolemia.
MethodsMethods, including statements of data availability and any asso-ciated accession codes and references, are available at https://doi.org/10.1038/s41590-018-0102-6.
Received: 15 December 2016; Accepted: 15 March 2018; Published: xx xx xxxx
References 1. Weber, C. & Noels, H. Atherosclerosis: current pathogenesis and therapeutic
options. Nat. Med. 17, 1410–1422 (2011). 2. Hansson, G. K. & Libby, P. The immune response in atherosclerosis: a
double-edged sword. Nat. Rev. Immunol. 6, 508–519 (2006). 3. Laurat, E. et al. In vivo downregulation of T helper cell 1 immune responses
reduces atherogenesis in apolipoprotein E-knockout mice. Circulation 104, 197 (2001).
4. Danzaki, K. et al. Interleukin-17A deficiency accelerates unstable atherosclerotic plaque formation in apolipoprotein E-deficient mice. Arterioscler. Thromb. Vasc. Biol. 32, 273 (2012).
5. Goodson, N., Marks, J., Lunt, M. & Symmons, D. Cardiovascular admissions and mortality in an inception cohort of patients with rheumatoid arthritis with onset in the 1980s and 1990s. Ann. Rheum. Dis. 64, 1595–1601 (2005).
6. Kimball, A. B. et al. Cardiovascular disease and risk factors among psoriasis patients in two US healthcare databases, 2001–2002. Dermatology 217, 27–37 (2008).
7. Roman, M. J. et al. Prevalence and correlates of accelerated atherosclerosis in systemic lupus erythematosus. N. Engl. J. Med. 349, 2399–2406 (2003).
8. Yu, H.-H. et al. Statin reduces mortality and morbidity in systemic lupus erythematosus patients with hyperlipidemia: a nationwide population-based cohort study. Atherosclerosis 243, 11–18 (2015).
9. Ghazizadeh, R., Tosa, M. & Ghazizadeh, M. Clinical improvement in psoriasis with treatment of associated hyperlipidemia. Am. J. Med. Sci. 341, 394–398 (2011).
10. Cai, Y., Fleming, C. & Yan, J. New insights of T cells in the pathogenesis of psoriasis. Cell. Mol. Immunol. 9, 302–309 (2012).
11. Ryu, H. & Chung, Y. Regulation of IL-17 in atherosclerosis and related autoimmunity. Cytokine 74, 219–227 (2015).
12. Choi, J.-Y. et al. Circulating follicular helper–like T cells in systemic lupus erythematosus: association with disease activity. Arthritis Rheumatol. 67, 988–999 (2015).
13. McGeachy, M. J., Singh, D., Henkel, M. & Moreland, L. Th17/TfH cells in rheumatoid arthritis: correlations with disease activity and therapy response. J. Immunol. 196, 51.23 (2016).
14. Lim, H. et al. Proatherogenic conditions promote autoimmune T helper 17 cell responses in vivo. Immunity 40, 153–165 (2014).
15. Park, H.-J. et al. Insights into the role of follicular helper T cells in autoimmunity. Immune Netw. 14, 21–29 (2014).
16. Yuan, J., Li, L. I., Wang, Z., Song, W. & Zhang, Z. Dyslipidemia in patients with systemic lupus erythematosus: association with disease activity and B-type natriuretic peptide levels. Biomed. Rep. 4, 68–72 (2016).
17. Chung, C. P. et al. Inflammatory mechanisms affecting the lipid profile in patients with systemic lupus erythematosus. J. Rheumatol. 34, 1849 (2007).
18. Baudino, L., Azeredo da Silveira, S., Nakata, M. & Izui, S. Molecular and cellular basis for pathogenicity of autoantibodies: lessons from murine monoclonal autoantibodies. Springer Semin. Immunopathol. 28, 175–184 (2006).
19. Tangye, S. G., Ma, C. S., Brink, R. & Deenick, E. K. The good, the bad and the ugly—TFH cells in human health and disease. Nat. Rev. Immunol. 13, 412–426 (2013).
20. Shortman, K. & Liu, Y.-J. Mouse and human dendritic cell subtypes. Nat. Rev. Immunol. 2, 151–161 (2002).
21. Reynolds, C. M. et al. Dietary saturated fatty acids prime the NLRP3 inflammasome via TLR4 in dendritic cells—implications for diet-induced insulin resistance. Mol. Nutr. Food Res. 56, 1212–1222 (2012).
22. Hong, C. & Tontonoz, P. Liver X receptors in lipid metabolism: opportunities for drug discovery. Nat. Rev. Drug Discov. 13, 433 (2014).
23. Castrillo, A. et al. Crosstalk between LXR and Toll-like receptor signaling mediates bacterial and viral antagonism of cholesterol metabolism. Mol. Cell 12, 805–816 (2003).
24. Collins, J. L. et al. Identification of a nonsteroidal liver X receptor agonist through parallel array synthesis of tertiary amines. J. Med. Chem. 45, 1963–1966 (2002).
25. Stumhofer, J. S. et al. Interleukins 27 and 6 induce STAT3-mediated T cell production of interleukin 10. Nat. Immunol. 12, 1363–1371 (2007).
26. Snapper, C. M. et al. Induction of IgG3 secretion by interferon gamma: a model for T cell-independent class switching in response to T cell-independent type 2 antigens. J. Exp. Med. 175, 1367 (1992).
27. Mozaffarian, D. F. et al. Executive summary: heart disease and stroke statistics—2016 update: a report from the American Heart Association. Circulation 133, 447–454 (2016).
28. Batten, M. et al. IL-27 supports germinal center function by enhancing IL-21 production and the function of T follicular helper cells. J. Exp. Med. 207, 2895 (2010).
29. Nurieva, R. I. et al. Generation of T follicular helper cells is mediated by interleukin-21 but independent of T helper 1, 2, or 17 cell lineages. Immunity 29, 138–149 (2008).
30. Owaki, T. et al. A role for IL-27 in early regulation of Th1 differentiation. J. Immunol. 175, 2191 (2005).
31. Vijayan, D. et al. IL-27 directly enhances germinal center B cell activity and potentiates lupus in Sanroque mice. J. Immunol. 197, 3008 (2016).
32. Koch, M. et al. The transcription factor T-bet controls regulatory T cell homeostasis and function during type 1 inflammation. Nat. Immunol. 10, 595–602 (2009).
33. Barbi, J., Oghumu, S., Lezama-Davila, C. M. & Satoskar, A. R. IFN-γ and STAT1 are required for efficient induction of CXC chemokine receptor 3 (CXCR3) on CD4+ but not CD8+ T cells. Blood 110, 2215–2216 (2007).
34. Chung, Y. et al. Follicular regulatory T (Tfr) cells with dual Foxp3 and Bcl6 expression suppress germinal center reactions. Nat. Med. 17, 983–988 (2011).
35. Chang, J.-H. & Chung, Y. Regulatory T cells in B cell follicles. Immune Netw. 14, 227–236 (2014).
36. Lee, S. K. et al. Interferon-γ excess leads to pathogenic accumulation of follicular helper T cells and germinal centers. Immunity 37, 880–892 (2012).
37. Miyauchi, K. et al. Protective neutralizing influenza antibody response in the absence of T follicular helper cells. Nat. Immunol. 17, 1447 (2016).
38. Wang, L. et al. Selective depletion of CD11c+CD11b+ dendritic cells partially abrogates tolerogenic effects of intravenous MOG in murine EAE. Eur. J. Immunol. 46, 2454–2466 (2016).
39. Shin, C. et al. CD8α – dendritic cells induce antigen-specific T follicular helper cells generating efficient humoral immune responses. Cell Reports 11, 1929–1940 (2015).
40. Joseph, S. B., Castrillo, A., Laffitte, B. A., Mangelsdorf, D. J. & Tontonoz, P. Reciprocal regulation of inflammation and lipid metabolism by liver X receptors. Nat. Med. 9, 213 (2003).
41. Jin, W. et al. Elevated circulating interleukin-27 in patients with coronary artery disease is associated with dendritic cells, oxidized low-density lipoprotein, and severity of coronary artery stenosis. Mediators Inflamm. 2012, 10 (2012).
42. Leng, C.-H. et al. A recombinant lipoprotein containing an unsaturated fatty acid activates NF-κ B through the TLR2 signaling pathway and induces a differential gene profile from a synthetic lipopeptide. Mol. Immunol. 47, 2015–2021 (2010).
43. Kim, H. S., Go, H., Akira, S. & Chung, D. H. TLR2-mediated production of IL-27 and chemokines by respiratory epithelial cells promotes bleomycin-induced pulmonary fibrosis in mice. J. Immunol. 187, 4007 (2011).
44. Gringhuis, S. I. et al. Fucose-based PAMPs prime dendritic cells for follicular T helper cell polarization via DC-SIGN-dependent IL-27 production. Nat. Commun. 5, 5074 (2014).
45. Mills, E. L., Kelly, B. & O’Neill, L. A. J. Mitochondria are the powerhouses of immunity. Nat. Immunol. 18, 488 (2017).
46. O’Neill, L. A. J. & Pearce, E. J. Immunometabolism governs dendritic cell and macrophage function. J. Exp. Med. 213, 15–23 (2016).
47. Dorosz, S. A. et al. Role of calprotectin as a modulator of the IL27-mediated proinflammatory effect on endothelial cells. Mediators Inflamm. 2015, 16 (2015).
48. Qiu, H.-N., Liu, B., Liu, W. & Liu, S. Interleukin-27 enhances TNF-α -mediated activation of human coronary artery endothelial cells. Mol. Cell. Biochem. 411, 1–10 (2016).
49. Xia, L. P., Li, B. F., Shen, H. & Lu, J. Interleukin-27 and interleukin-23 in patients with systemic lupus erythematosus: possible role in lupus nephritis. Scand. J. Rheumatol. 44, 200–205 (2015).
50. Pan, H.-F., Tao, J.-H. & Ye, D.-Q. Therapeutic potential of IL-27 in systemic lupus erythematosus. Expert Opin. Ther. Targets 14, 479–484 (2010).
AcknowledgementsWe thank K.-W. Kim and S.-J. Bae for support in flow cytometry; S. Akira (Osaka University) and S-Y. Sung (Seoul National University) for Stat3fl/fl and Tlr4–/– mice; H. S. Kim (Ulsan University) for Stat1–/– mice; J.-H. Choi (Hanyang University) for Apoe–
/– mice; S-A. Yoo (The Catholic University of Korea) for animal models; the entire Chung lab for suggestion and discussion; and J. Reynolds, G. Martinez and D.-S. Kuen for proofreading the manuscript. Supported by the National Research Foundation of Korea (research grants 2014R1A2A1A11054364, 2017R1A2B3007392 and 2017K1A1A2004511 to Y.C., and 2015R1D1A1A01059719 to M.C.).
Author contributionsH.R., H.L., G.C. and H.N. performed the in vitro and in vivo experiments; Y.-J.P. generated Ebi3–/–Apoe–/– mice; M.C. generated the Tlr4–/– bone marrow chimeras; C.W.A. and Y.C.K. provided Ldlr–/– mice; H.R., W.-U.K. and Y.C. analyzed the data; S.-H.L. provided the human plasma samples and patient information; H.R. and Y.C. wrote the manuscript; and all authors complied in submission of the manuscript for publication.
Competing interestsThe authors declare no competing interests.
Additional informationSupplementary information is available for this paper at https://doi.org/10.1038/s41590-018-0102-6.
Reprints and permissions information is available at www.nature.com/reprints.
Correspondence and requests for materials should be addressed to Y.C.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
MethodsEthics statement. All experiments were performed according to protocols approved by the Institutional Animal Care and Use Committees of Seoul National University (SNU-160422-3-6). Applicable human subject protocols were approved by the Institutional Review Boards at Seoul National University (IRB No. E1707/001-006) and Yonsei University (4-2016-1010). All subjects provided informed consent. The characteristics of each subject are provided in Supplementary Table 2.
Mice. Apoe–/–, BXD2, Ebi3–/–, Cd4-cre, Ldlr–/– and Il27ra–/– mice were purchased from Jackson Laboratory. Tlr4–/– and Stat3fl/fl mice were provided by S.-Y. Sung (Seoul National University) and S. Akira (Osaka University). Stat1–/– mice were provided by H.S. Kim (Ulsan University). C57BL/6 mice were purchased from Orient Bio. Female mice were used throughout all experiments. Ebi3–/–Apoe–/– mice were generated by the interbreeding of Ebi3–/– and Apoe–/–mice. In some experiments, we established bone marrow (BM) chimeric mice. In brief, 7- to 8-week-old mice were given intraperitoneal injection of busulfan in DMSO (50 mg per kg body weight) to ablate BM in the recipient mice, and BM cells from the indicated mice were adoptively transferred. The recipients were maintained on normal chow diet for 6 weeks of reconstitution before additional treatment. All mice were bred and maintained in a specific pathogen-free facility at the vivarium of the Seoul National University, with free access to gamma-irradiated standard cereal-based diets (Ziegler) or gamma-irradiated high fat diets (40% energy from fat) (Research Diets).
Immunization. WT, Apoe–/– or Ebi3–/–Apoe–/– mice or the recipient mice in the bone marrow–chimerism studies were subcutaneously immunized with 100 μ g of KLH (Sigma) or NP-OVA (Biosearch Technologies) emulsified in CFA. 7 d later, lymphoid cells from the spleen and the draining lymph nodes were analyzed. In some experiments, mice were immunized with 200 μ g of chicken type II collagen (Chondrex) emulsified in CFA, as described previously51.
Immunohistochemical analysis. The spleen or kidneys of mice were embedded in Tissue-Tek O.C.T compound (Sakura Finetek) and were cut into 7-μ m sections. The slides were incubated with 3% BSA in PBS for 30 min and then were stained with FITC peanut agglutinin (Sigma), PE anti-IgD (11–26 c.2a, Biolegend), FITC anti-C3 (MP), or FITC anti-IgG (Sigma). The slides were mounted with ProLong Gold Antifade Reagent (Molecular Probes). Kidneys were incubated in 4% formalin for 12 h at 4 °C for fixation and were sent to the Pathology Center at Seoul National University College of Medicine for hematoxylin-and-eosin staining. Images were acquired with the Vectra 3.0 Automated Quantitative Pathology Imaging System (PerkinElmer) and were analyzed with inForm (PerkinElmer).
Flow cytometry. Lymphoid cells were obtained and then were stained with PerCp-Cy5.5 or APC-Cy7 anti-CD4 (GK1.5, ThermoFisher), FITC anti-PD-1 (J43, ThermoFisher), biotin anti-CXCR5 (L138D7, Biolegend), APC streptavidin (Biolegend), PE anti-CXCR3 (CXCR3-173, ThermoFisher), PE-Cy7 anti-CCR6 (29-2L17, Biolegend), FITC anti-GL-7 (GL7, Biolegend), PE anti-CD95 (15A7, ThermoFisher), PerCp-Cy5.5 anti-CD138 (281-2, Biolegend), Alexa488 anti-CD62L (MEL-14, Biolegend), PE-Cy7 anti-CD44 (IM7, Biolegend), PE anti-CD25 (PC61, Biolegend), PE anti-IgD (11-26 c.2a, Biolegend), APC anti-B220 (RA3-6B2, Biolegend), APC anti-F4/80 (BM8, Biolegend), APC-Cy7 anti-I-A/I-E (M5/114.15.2, Biolegend), PerCp-Cy5.5 anti-CD11c (N418, Biolegend), FITC anti-CD11b (M1/70, Biolegend) or PE-Cy7 anti-CD8a (53-6.7, Biolegend). Foxp3 staining was performed by using a Foxp3/Transcription factor staining buffer set (ThermoFisher) and eFluor450 anti-Foxp3 (FJK-16S, ThermoFisher). For BODIPY staining, lymphoid cells were incubated with 1 μ M BODIPY in the dark for 30 min before washing. For STAT staining, cells were permeabilized with methanol for 30 min after fixation, then were stained with Alexa647 anti-STAT1 (4a, BD Bioscience) or Pacific Blue anti-STAT3 (4/P-STAT3, BD Bioscience). The cells were analyzed with a FACSVerse, LSRFortessa, FACSAria III or FACSCalibur (BD Bioscienc), and data were analyzed using FlowJo software (TreeStar).
Cell isolation and differentiation. For preparation of primary DCs, spleens were surgically removed and were teased into small pieces by using a gentleMACS dissociator (Miltenyi Biotec), then were digested with RPMI 1640 medium containing 10% of FBS, 0.5 mg/ml of collagenase (ThermoFisher) and 30 μ g/ml of DNase I (Bio Basic) for 30 min at 37 °C. Then, CD11c+ cells were isolated by MACS (Miltenyi Biotec). Isolated DCs were stimulated with LPS (100 ng/ml,
Sigma) for 4 h (for RT-PCR) or 24 h (for ELISA). CD4+ T cells were isolated by MACS (Miltenyi Biotec) and the cells were further sorted as TFH cells (CD4+PD-1+CXCR5+) by a FACSAria III flow cytometer (BD Bioscience). Naive B cells (B220+GL-7–IgD+; 1 × 105 cells per well) were isolated by FACSAria III and were co-cultured with TFH cells (4 × 104 cells per well) for 7 d in in RPMI-1640 supplemented with 10% FBS in the presence of anti-CD3 (145.2C11, Bio X cell), anti-IFN-γ (XMG1.2, Bio X Cell) and anti-IgM (AffiniPure F(ab’)2 Fragment Goat anti-IgM, μ chain specific, 115-006-020, Jackson ImmunoResearch).
ELISA. To measure the levels of antibodies to dsDNA and rheumatoid factors, serum was collected from mice and were measured by ELISA as described previously52. In brief, ELISA plates (Greiner Bio-one) were coated with 0.01% of poly-L-lysine (Sigma-Aldrich) for 1 h before coating with 5 μ g/ml of calf thymus dsDNA (Sigma-Aldrich) and rabbit IgG (Sigma-Aldrich) overnight at 4 °C. The plates were blocked for 2 h with 5% skim milk in PBS. Serum was diluted in the same solution, transferred to the plates and incubated for 2 h at room temperature. Then, the assay was performed with HRP-conjugated detection antibody for total IgG (1010-05), IgG1 (1070-05) or IgG2c (1079-05) (all from SouthernBiotech). To measure the levels of IgG in culture supernatants or serum, IgG was quantified with a total IgG capture antibody (donkey anti-mouse IgG (H+ L), 715-005-151, Jackson ImmunoResearch) and HRP-conjugated detection antibody for IgG, IgG1, IgG2b, or IgG2c (identified above). Serum from OVA-NP immunized mice was collected, and NP-specific IgG, IgG1, IgG2b and IgG2c antibodies were measured by NP-BSA (Biosearch Technologies) ELISA. Other ELISA kits, including IL-27, and human IgG, IgG1, IgG3 and IgG4, were purchased from ThermoFisher, and assays were conducted as described in the manufacturer’s instruction. The concentration of TG, HDL, and LDL was determined by TRIGL, HDL-Cholesterol plus 3rd generation, or LDL-cholesterol Gen.3 (all from Roche), respectively. The concentration of VLDL was determined by Freidewald’s formula.
RT-PCR. Total RNA was prepared by using Trizol (Invitrogen), and cDNA was synthesized with a RevertAid First Strand cDNA Synthesis Kit (ThermoFisher) according to the manufacturer’s protocol. Gene expression was measured by quantitative 7500 Fast RT-PCR (Applied Biosystems). The levels of gene expression were normalized to Actb. The primer pairs used in quantitative RT-PCR are described in Supplementary Table 3.
Immunoblot analysis. CD11c+ DCs were isolated from the indicated mice by MACS (Miltenyi Biotec). Cells were washed with cold PBS and lysed with RIPA lysis buffer containing protease inhibitor cocktail (GenDepot), and cell lysates (40 μ g of proteins) were separated by SDS-PAGE and transferred to Immobilon-P PVDF membranes (Millipore). The blots were blocked with 5% nonfat dry milk in TBST (0.05%) and were probed with antibodies at 4 °C overnight. The following antibodies were used for immunoblot analysis: anti-LXRβ (14278-1-AP, 1:2,000 dilution, Proteintech), anti-β -actin (sc-47778, 1:1,000 dilution, Santa Cruz Biotechnology), anti-rabbit IgG-HRP (sc-2004, 1:5,000 dilution, Santa Cruz Biotechnology) and anti-mouse IgG-HRP (sc-2005, 1:5,000 dilution, Santa Cruz Biotechnology).
Statistical analysis. Data were analyzed with GraphPad Prism 7 (GraphPad Software). Statistics were calculated with the two-tailed Student’s t-test. P values are presented within each figure or figure legend.
Reporting Summary. Further information on experimental design is available in the Nature Research Reporting Summary.
Data availability statement. The data that support the findings of this study are available from the corresponding author upon request. RNA-seq data have been deposited in the Gene Expression Omnibus under accession code GSE111779.
(anti-VEGF) hexapeptide inhibits collagen-induced arthritis and VEGF-stimulated productions of TNF-α and IL-6 by human monocytes. J. Immunol. 174, 5846 (2005).
52. Kim, Y. U., Lim, H., Jung, H. E., Wetsel, R. A. & Chung, Y. Kim, Y. U., Lim, H., Jung, H. E., Wetsel, R. A. & Chung, Y. Regulation of autoimmune germinal center reactions in lupus-prone BXD2 mice byfollicular helper T cells. PLoS One 10, e0120294 (2015).
Initial submission Revised version Final submission
Life Sciences Reporting SummaryNature Research wishes to improve the reproducibility of the work that we publish. This form is intended for publication with all accepted life science papers and provides structure for consistency and transparency in reporting. Every life science submission will use this form; some list items might not apply to an individual manuscript, but all fields must be completed for clarity.
For further information on the points included in this form, see Reporting Life Sciences Research. For further information on Nature Research policies, including our data availability policy, see Authors & Referees and the Editorial Policy Checklist.
Experimental design1. Sample size
Describe how sample size was determined. To get a significance of 0.05 with 2-sided test, power of 80, and effect size of 3.1 or 2.0, we need 3 or 5 mice per group. All experiments were conducted based on above calculation.
2. Data exclusions
Describe any data exclusions. No data were excluded from the analysis
3. Replication
Describe whether the experimental findings were reliably reproduced.
All experiments were conducted at least 2 times and the numbers of attempts were stated in the figure legends. All attempts at replication was successful.
4. Randomization
Describe how samples/organisms/participants were allocated into experimental groups.
Mice were allocated into the experimental groups based on their genotype or the way they were treated. Human samples were allocated based on modified UK criteria classified as "definite, probable, or possible".
5. Blinding
Describe whether the investigators were blinded to group allocation during data collection and/or analysis.
Mice and human samples were labeled as numeric numbers and analyzed. The data were analyzed by at least two different individuals.
Note: all studies involving animals and/or human research participants must disclose whether blinding and randomization were used.
6. Statistical parameters For all figures and tables that use statistical methods, confirm that the following items are present in relevant figure legends (or in the Methods section if additional space is needed).
n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement (animals, litters, cultures, etc.)
A description of how samples were collected, noting whether measurements were taken from distinct samples or whether the same sample was measured repeatedly
A statement indicating how many times each experiment was replicated
The statistical test(s) used and whether they are one- or two-sided (note: only common tests should be described solely by name; more complex techniques should be described in the Methods section)
A description of any assumptions or corrections, such as an adjustment for multiple comparisons
The test results (e.g. P values) given as exact values whenever possible and with confidence intervals noted
A clear description of statistics including central tendency (e.g. median, mean) and variation (e.g. standard deviation, interquartile range)
Clearly defined error bars
See the web collection on statistics for biologists for further resources and guidance.
2
nature research | life sciences reporting summ
aryJune 2017
SoftwarePolicy information about availability of computer code
7. Software
Describe the software used to analyze the data in this study.
We used FlowJo v.10 for FACS analysis, inForm for histological analysis, and Prism 7 for statistics, .
For manuscripts utilizing custom algorithms or software that are central to the paper but not yet described in the published literature, software must be made available to editors and reviewers upon request. We strongly encourage code deposition in a community repository (e.g. GitHub). Nature Methods guidance for providing algorithms and software for publication provides further information on this topic.
Materials and reagentsPolicy information about availability of materials
8. Materials availability
Indicate whether there are restrictions on availability of unique materials or if these materials are only available for distribution by a for-profit company.
All materials are commercially available.
9. Antibodies
Describe the antibodies used and how they were validated for use in the system under study (i.e. assay and species).
For flow cytometry analysis, cells were stained with PerCp-Cy5.5 or APC-Cy7 anti-CD4 (GK1.5, ThermoFisher), FITC anti-PD-1 (J43, ThermoFisher), biotin anti-CXCR5 (L138D7, Biolegend), APC streptavidin (Biolegend), PE anti-CXCR3 (CXCR3-173, ThermoFisher), PE-Cy7 anti-CCR6 (29-2L17, Biolegend), FITC anti-GL-7 (GL7, Biolegend), PE anti-CD95 (15A7, ThermoFisher), PerCp-Cy5.5 anti-CD138 (281-2, Biolegend), Alexa488 anti-CD62L (MEL-14, Biolegend), PE-Cy7 anti-CD44 (IM7, Biolegend), PE anti-CD25 (PC61, Biolegend), PE anti-IgD (11-26c.2a, Biolegend), APC anti-B220 (RA3-6B2, Biolegend), APC anti-F4/80 (BM8, Biolegend), APC-Cy7 anti-I-A/I-E (M5/114.15.2, Biolegend), PerCp-Cy5.5 anti-CD11c (N418, Biolegend), FITC anti-CD11b (M1/70, Biolegend), or PE-Cy7 anti-CD8a (53-6.7, Biolegend). Foxp3 staining was performed by using Foxp3/Transcription factor staining buffer set (ThermoFisher) and eFluor450 anti-Foxp3 (FJK-16S, ThermoFisher). For BODIPY staining, lymphoid cells were incubated with 1 μM BODIPY in dark for 30 minutes before washing. For STAT staining, cells were permeabilized with methanol for 30 minutes after fixation, and were stained with Alexa647 anti-STAT1 (4a, BD Bioscience) or Pacific Blue anti-STAT3 (4/P-STAT3, BD Bioscience). For Western blot, anti-LXRβ (14278-1-AP, 1:2000 dilution, Proteintech), anti-β-actin (sc-47778, 1:1000 dilution, Santa Cruz Biotechnology), anti-rabbit IgG-HRP (sc-2004, 1:5000 dilution, Santa Cruz Biotechnology), and anti-mouse IgG-HRP (sc-2005, 1:5000 dilution, Santa Cruz Biotechnology) were used. For ELISA, total IgG, IgG1, IgG2b, or IgG2c were purchased from Southern Biotech, and other ELISA kits, including IL-27, human IgG, IgG1, IgG3, and IgG4 were purchased from ThermoFisher, and were conducted as described in manufacturer’s instruction
10. Eukaryotic cell linesa. State the source of each eukaryotic cell line used. No eukaryotic cell lines were used.
b. Describe the method of cell line authentication used. Describe the authentication procedures for each cell line used OR declare that none of the cell lines used have been authenticated OR state that no eukaryotic cell lines were used.
c. Report whether the cell lines were tested for mycoplasma contamination.
Confirm that all cell lines tested negative for mycoplasma contamination OR describe the results of the testing for mycoplasma contamination OR declare that the cell lines were not tested for mycoplasma contamination OR state that no eukaryotic cell lines were used.
d. If any of the cell lines used are listed in the database of commonly misidentified cell lines maintained by ICLAC, provide a scientific rationale for their use.
Provide a rationale for the use of commonly misidentified cell lines OR state that no commonly misidentified cell lines were used.
3
nature research | life sciences reporting summ
aryJune 2017
Animals and human research participantsPolicy information about studies involving animals; when reporting animal research, follow the ARRIVE guidelines
11. Description of research animalsProvide details on animals and/or animal-derived materials used in the study.
Age of 7-to-8-week-old female mice were used in our study. Apoe–/–, BXD2, Ebi3–/–, Ldlr–/– and Il27ra–/– mice were purchased from Jackson Laboratory (Maine, USA). Tlr4–/– and Stat3fl/fl were kindly provided by Drs. Seung-Yong Sung and Shuzuo Akira. Stat1–/– mice were kindly provided by Dr. Hun Sik Kim (Ulsan University). C57BL/6 mice were purchased from Orient Bio (Gyeonggi, South Korea).
Policy information about studies involving human research participants
12. Description of human research participantsDescribe the covariate-relevant population characteristics of the human research participants.
Patients with hypercholesterolemia were defined based on modified UK criteria ("definite","probable" or "possible". Healthy controls were defined as people who were not under cholesterol-lowering treatment, and the levels of cholesterol in sera < 200 mg/dL. Total 24 plasma samples from the patients and healthy controls were collected, respectively. Both groups were comparable in terms of ratio of male (37.5%), age, body mass index, comorbidities, and use of medication.
nature research | flow cytom
etry reporting summ
aryJune 2017
1
Corresponding author(s): Yeonseok Chung
Initial submission Revised version Final submission
Flow Cytometry Reporting Summary Form fields will expand as needed. Please do not leave fields blank.
Data presentationFor all flow cytometry data, confirm that:
1. The axis labels state the marker and fluorochrome used (e.g. CD4-FITC).
2. The axis scales are clearly visible. Include numbers along axes only for bottom left plot of group (a 'group' is an analysis of identical markers).
3. All plots are contour plots with outliers or pseudocolor plots.
4. A numerical value for number of cells or percentage (with statistics) is provided.
Methodological details5. Describe the sample preparation. All the lymph nodes or spleens were prepared individually. Single cell
suspensions were stained according to manufacturer's protocols.
6. Identify the instrument used for data collection. FACS Aria III (for sorting), Verse, Calibur, Fortessa, Canto (for analyzing) from BD were used for data collection.
7. Describe the software used to collect and analyze the flow cytometry data.
Diva, Cellquest Pro., or Suite were used for data collection, and FlowJo v.10 were used for analysis.
8. Describe the abundance of the relevant cell populations within post-sort fractions.
Purity confirmed by FACS was >95% after sorting.
9. Describe the gating strategy used. Gating strategies are provided as Supplementary Figure 1c. Briefly, DCs as F4/80-B220-CD11c+MHCIIhi, naive B cells as B220+IgD+GL7-, Tfh cells as CD4+PD-1+CXCR5+Foxp3-, GC B as B220+GL7+CD95+, plasma cell as B220-CD138+, and Tfr as Cd4+PD-1+CXCR5+Foxp3+.
Tick this box to confirm that a figure exemplifying the gating strategy is provided in the Supplementary Information.
