Article

Initiation of Antiviral B Cell Immunity Relies on Innate

Signals from Spatially Positioned NKT Cells
Graphical Abstract Authors
Mauro Gaya, Patricia Barral,
Marianne Burbage, ...,
Andreas Bruckbauer, Jessica Strid,
Facundo D. Batista

Correspondence
contact@maurogaya.com (M.G.),
fbatista1@mgh.harvard.edu (F.D.B.)

In Brief
NKT cells are required for the initial
formation of germinal centers and
production of effective neutralizing
antibody responses against viruses.

Highlights
d NKT cells promote B cell immunity upon viral infection

d NKT cells are primed by lymph-node-resident macrophages

d NKT cells produce early IL-4 wave at the follicular borders

d Early IL-4 wave is required for efficient seeding of germinal

centers

Gaya et al., 2018, Cell 172, 517–533

January 25, 2018 ª 2017 The Authors. Published by Elsevier Inc.
https://doi.org/10.1016/j.cell.2017.11.036
Article

Initiation of Antiviral B Cell Immunity Relies

on Innate Signals from Spatially
Positioned NKT Cells
Mauro Gaya,1,2,* Patricia Barral,2,3 Marianne Burbage,2 Shweta Aggarwal,2 Beatriz Montaner,2 Andrew Warren Navia,1,4,5
Malika Aid,6 Carlson Tsui,2 Paula Maldonado,2 Usha Nair,1 Khader Ghneim,7 Padraic G. Fallon,8 Rafick-Pierre Sekaly,7
Dan H. Barouch,1,6 Alex K. Shalek,1,4,5 Andreas Bruckbauer,2 Jessica Strid,9 and Facundo D. Batista1,2,10,11,*
1Ragon Institute of MGH, MIT and Harvard, Cambridge, MA 02139, USA
2The Francis Crick Institute, London NW1A 1AT, UK
3The Peter Gorer Department of Immunobiology, King’s College London, London SE1 9RT, UK
4Institute for Medical Engineering & Science, MIT, Cambridge, MA 02139, USA
5Broad Institute, Cambridge, MA 02142, USA
6Center for Virology and Vaccine Research, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02215, USA
7Case Western Reserve University, Cleveland, OH 44106, USA
8Institute of Molecular Medicine, Trinity College Dublin, Dublin, Ireland
9Division of Immunology and Inflammation, Department of Medicine, Imperial College London, London W12 0NN, UK
10Department of Microbiology and Immunobiology & HMS Center for Immune Imaging, Harvard Medical School, Boston, MA 02115, USA
11Lead Contact

*Correspondence: contact@maurogaya.com (M.G.), fbatista1@mgh.harvard.edu (F.D.B.)

https://doi.org/10.1016/j.cell.2017.11.036

SUMMARY through the production of highly specific antibodies. In the

B cells constitute an essential line of defense from
pathogenic infections through the generation of
class-switched antibody-secreting cells (ASCs) in
germinal centers. Although this process is known to
be regulated by follicular helper T (TfH) cells, the mech-
anism by which B cells initially seed germinal center
reactions remains elusive. We found that NKT cells, a
population of innate-like T lymphocytes, are critical
for the induction of B cell immunity upon viral infection.
The positioning of NKT cells at the interfollicular areas
of lymph nodes facilitates both their direct priming by
resident macrophages and the localized delivery of
innate signals to antigen-experienced B cells. Indeed,
NKT cells secrete an early wave of IL-4 and constitute
up to 70% of the total IL-4-producing cells during the
initial stages of infection. Importantly, the requirement
of this innate immunity arm appears to be evolution-
arily conserved because early NKT and IL-4 gene
signatures also positively correlate with the levels
of neutralizing antibodies in Zika-virus-infected ma-
caques. In conclusion, our data support a model
wherein a pre-TfH wave of IL-4 secreted by interfollic-
ular NKT cells triggers the seeding of germinal center
cells and serves as an innate link between viral infec-
tion and B cell immunity.
INTRODUCTION
B cells are key elements of the adaptive immune response
because they provide protection from pathogenic threats
past few years, there has been much attention on the generation
of broadly neutralizing antibodies against highly variable or
emerging pathogens, such as influenza, HIV, and Zika virus to
develop effective vaccines capable of conferring broad and
long-lasting immunity. Therefore, it is crucial to understand
how viruses elicit B cell immunity during an infection process.
Pathogen recognition occurs in secondary lymphoid organs
such as the spleen and lymph nodes (Batista and Harwood,
2009). In the lymph nodes, B cells recognize pathogens on the
surface of subcapsular sinus (SCS) macrophages, migratory
dendritic cells, and follicular dendritic cells (Carrasco and Ba-
tista, 2007; Gaya et al., 2015; Heesters et al., 2013; Junt et al.,
2007; Phan et al., 2007; Qi et al., 2006). Following pathogen
recognition, B cells process and present pathogen-derived pep-
tides via major histocompatibility complex (MHC) class II mole-
cules (Amigorena et al., 1994). This results in the recruitment of
specific CD4+ T cells, which help the B cells to differentiate
into either antibody-secreting plasma cells or germinal center
(GC) cells. Within germinal centers, B cells undergo class-switch
recombination and cycles of proliferation, somatic hypermuta-
tion, and affinity-based selection, finally resulting in the produc-
tion of high-affinity antibodies (Victora and Nussenzweig, 2012).
These processes are strictly dependent on co-stimulatory sig-
nals, mainly CD40-L, ICOS, and cytokines, such as interferon g
(IFN-g), interleukin-4 (IL-4), and IL-21, provided by follicular
helper T (TfH) cells (Weinstein et al., 2016). Because TfH cells
are known to appear later during the immune response, how
B cells initially seed germinal center reactions is still unclear.
In addition to pathogen-derived peptides, B cells are able to
present glycolipid antigens, such as those present on the surface
of Sphingomonas species (spp.), Borrelia burgdorferi, and Strep-
tococcus pneumoniae, on CD1d molecules (Barral et al.,
2008; Kinjo et al., 2011; Leadbetter et al., 2008). Populations of

Cell 172, 517–533, January 25, 2018 ª 2017 The Authors. Published by Elsevier Inc. 517
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
 A Uninfected CD1d-/- ****
11 Uninfected

Fas+ GL7+ B cells (%)

10 5

104 0.9% 9.5% 4.6%

CD1d-/-
5.5
10 3

0
Fas

0
0 103 104 105
GL7

B Uninfected CD1d-/-

CXCR5+PD-1+T cells (%)

10 5 24 Uninfected

104 4.3% 17.9% 18.6%

CD1d-/-
12
10 3
CXCR5

0
0
0 103 104 105
PD-1

C Uninfected CD1d-/-
IgD
GL7

D E
**
** * 18000
α-IAV IgG1 ASC/LN

Uninfected
GL7+ structures / section

10 0.4 Uninfected
GL7+ area /IgD+ area

CD1d-/- CD1d-/-
9000
5 0.2

0
0 0

F Uninfected Wild type/Vaccinia CD1d-/-/Vaccinia **

18 Uninfected
Fas+ GL7+ B cells (%)

105
Wild type/Vaccinia
0.3% 14.7% 7.8%
104 CD1d-/-/Vaccinia
9
103
Fas

0
0
0 103 104 105
GL7

G Uninfected Wild type/Vaccinia CD1d-/-/Vaccinia

CXCR5+PD-1+T cells (%)

24 Uninfected
105
Wild type/Vaccinia
104 4.7% 19.8% 17.0%
CD1d-/-/Vaccinia
12
103
CXCR5

0
0
0 103 104 105
PD-1

(legend on next page)

518 Cell 172, 517–533, January 25, 2018

CD1d-restricted, innate-like T cells, known as natural killer (NK)T
cells, recognize these lipids and provide cognate help to B cells.
This NKT cell-mediated help enhances the production of anti-
bodies by B cells (Bai et al., 2013; Barral et al., 2008; Leadbetter
et al., 2008). Because there are no reports of lipid antigens orig-
inating from viruses, it is currently unclear whether NKT cells
have a role in inducing antiviral B cell responses.
Here we show that, following respiratory viral infection, NKT
cell-deficient mice are impaired in their ability to form germinal
centers and immunoglobulin G (IgG) class-switched antibody-
secreting cells (ASCs). In a striking contrast to B cell activation
by glycolipid antigens, CD1d-mediated interactions between
B cells and NKT cells are dispensable during viral infection.
Instead, NKT cells promote B cell immunity through the early
and localized production of IL-4 at the follicular borders, a pro-
cess that is likely driven by CXCR3 and requires initial priming
by resident macrophages through CD1d and IL-18. Importantly,
this seems to be a conserved process because transcriptomic
analysis of Zika-virus-infected macaques shows that early NKT
and IL-4 gene signatures positively correlate with the levels of
Zika-specific neutralizing antibodies. Therefore, our results indi-
cate that the production of IL-4 by NKT cells immediately after
viral infection contributes to a pre-TfH IL-4 wave at the follicular
borders that facilitates the initial seeding of germinal center cells.
RESULTS
NKT Cells Promote B Cell Immunity during Viral
Infection
The role of NKT cells in inducing the humoral response to glyco-
lipid-bearing antigens has been established (Bai et al., 2013;
Barral et al., 2008; Leadbetter et al., 2008); however, because
virus-derived lipid antigens are unknown, we sought to investi-
gate the role of NKT cells in the induction of B cell responses dur-
ing viral infections. Accordingly, we intranasally infected
C57BL/6 wild-type and CD1d / mice, which lack CD1d and
type I and II NKT cells, with 200 plaque-forming units (PFUs) of
influenza A virus, Puerto Rico/8/34 (PR8) strain. After 7, 9, and
21 days of infection, lung-draining lymph nodes were harvested,
and adaptive immune response development was assessed
by flow cytometry. We observed a progressive expansion of
Fas+GL7+ germinal center cells and CXCR5+PD-1+ TfH cells dur-
ing the course of infection in wild-type animals compared with
uninfected ones (Figures 1A and 1B; Figures S1A–S1D). Interest-
ingly, the formation of germinal center cells was significantly
reduced in NKT cell-deficient mice at early time points, whereas
the expansion TfH cells remained similar throughout the infection
process (Figures 1A and 1B; Figures S1A–S1D). We further
compared the spatial organization of germinal centers in wild-
type and CD1d / mice by immunohistochemistry. In wild-type
animals, influenza infection induces both the enlargement of
IgD+ B cell follicles as well as the formation of follicular GL7+
structures corresponding to germinal centers (Figure 1C). In
contrast, in NKT cell-deficient animals, germinal center struc-
tures were reduced by day 9 of infection, a defect that
was accompanied by a strong decrease in the generation of
influenza-specific IgG1 ASCs (Figures 1C–1E). Interestingly,
CD1d / mice also displayed a significant reduction in germinal
center formation when infected with the vaccinia virus Western
Reserve strain, suggesting that the ability of NKT cells to pro-
mote B cell responses is a general feature of viral infections (Fig-
ures 1F and 1G).
To address whether NKT cells would also induce B cell immu-
nity after vaccination with viral antigens, we intranasally immu-
nized wild-type and CD1d / mice with hemagglutinin (HA)
trimmer from the PR8 influenza strain in the presence or absence
of two different adjuvants: alpha-galactosylceramide (a-GalCer)
or poly(I:C). We found that HA-specific germinal centers devel-
oped in lung-draining lymph nodes only when HA was adminis-
tered together with either of these adjuvants (Figures S1E
and S1F). Interestingly, the formation of HA-specific germinal
centers and ASCs is completely abrogated in NKT-cell-deficient
mice when a-GalCer was co-administered, demonstrating that
NKT cells are essential to induce B cell immunity when exoge-
nous glycolipids are used as adjuvants (Figures S1G and S1H).
In addition, when poly(I:C) was co-administered, we observed
a diminution in the number of IgM ASCs and a consistent,
although non-significant, reduction in HA-specific germinal cen-
ters in CD1d / animals. These results indicate that NKT cells
may also enhance B cell immunity during intranasal vaccination
in the absence of exogenous glycolipids (Figures S1I and S1J).
NKT Cells Deliver Non-cognate Help to B Cells to Induce
Antiviral Immunity
Having shown that CD1d / mice have diminished B cell re-
sponses during influenza and vaccinia virus infection, we sought
to determine the mechanism by which NKT cells help B cells to
mount an antiviral immune response. In the case of bacterium-
derived glycolipids, NKT cells differentiate into CXCR5+PD-
1+Bcl6+ follicular helper NKT (NKTfH) cells, which engage in
CD1d-mediated cognate interactions with B cells, produce
IL-21, and support germinal center responses (Chang et al.,
2011; King et al., 2011). To address whether NKT cells differen-
tiate into NKTfH cells after viral infection, wild-type mice were

Figure 1. NKT Cell-Deficient Mice Exhibit Impaired Antiviral B Cell Responses

(A and B) Flow cytometry analysis of mediastinal lymph nodes (LN) from control (black), wild-type (blue), or CD1d / (red) mice on day 9 of influenza infection.
Contour plots display the percentage of (A) germinal center and (B) TfH cells.
(C) Confocal microscopy images of mediastinal lymph node sections from mice treated as in (A) and (B), stained with antibodies to IgD (green) and GL7 (magenta).
Scale bar, 450 mm.
(D) Quantification of the number and follicular area occupied by germinal centers in confocal images shown in (C) by ImageJ.
(E) ELISPOT analysis of IgG1 influenza-specific ASCs in mediastinal lymph nodes from wild-type and CD1d / mice treated as described in (A) and (B).
(F and G) Flow cytometry analysis of mediastinal lymph nodes of from control (black), wild-type (blue), or CD1d / (red) mice on day 9 of vaccinia virus infection.
Contour plots display the percentage of (F) germinal center and (G) TfH cells.
Data show one representative result from three experiments, where each dot represents one mouse. Data represent mean ± SEM; two-tailed paired Student’s
t test, *p < 0.05, **p < 0.01, and ****p < 0.0001.

Cell 172, 517–533, January 25, 2018 519

A Gated on B220- LN cells B NKT cells CD4+ T cells

CXCR5+PD-1+ cells (%)

20
105 NKT cells CD4 T cells
+
105 NKT cells
CD4+ T cells
10 4
104 1.5% 20.0%
10
103 103

CXCR5
TCRβ

0 0
0
0 103 104 105 0 103 104 105 0 103 104 105 0 3 6 9
CD1d-tetramer CD4 PD-1 Days after infection

C *** ** D Targeted CD1d allele

12 16
CD1d-tet+ cell density
GL7
CD1d-tet+ cell number

CD1d-tet Exon 2 Lox P Exon 3 Lox P Exon 4

(10-5 μm-2)

6 8

Exon 2 Lox P Exon 4

0 0
Outside Inside Outside Inside
GC GC GC GC

E B cells F CD19-Cre **** G Mb1-Cre ****

105 100 100 100 100
Cre+ Cre-

CD1d+ B cells (%)

Cre+ Cre-
Events (% of max)

CD1d+ B cells (%)

Events (% of max)
104
50 50 50 50
103
CD19

0
0 0 0 0
0 103 104 105 0 103 104 105 0 103 104 105 Cre- Cre+
Cre- Cre+
IgD CD1d CD1d

H CD1d CD19-Cre- CD1d CD19-Cre+ I CD1d CD19-Cre- CD1d CD19-Cre+

CXCR5+PD-1+T cells (%)
15 105 20
10 5
Fas+ GL7+ B cells (%)

10.1% 10.2% 10 4 18.3 % 16.5 %

104
103 10
7.5
103
0
CXCR5

0
Fas

0 0
0 103 104 105 Cre- 0 103 104 105 Cre- Cre+
Cre+
GL7 PD-1

J CD1d Mb1-Cre- CD1d Mb1-Cre+ K CD1d Mb1-Cre- CD1d Mb1-Cre+

CXCR5+PD-1+T cells (%)

10 5 12 10 16
Fas+ GL7+ B cells (%)

8.9% 8.2% 10 4 15.9 % 14.9 %

104
103
6 8
103
0
CXCR5

0
Fas

0 0
0 103 104 105 Cre- Cre+ 0 103 104 105 Cre- Cre+
GL7 PD-1

Figure 2. NKT Cells Deliver Non-cognate Help to B Cells during Viral Infection
(A and B) Flow cytometry analysis of wild-type mice infected with influenza virus.
(A) Gating strategy for NKT and CD4+ T cells. B220+ cells were excluded from the analysis.
(legend continued on next page)

520 Cell 172, 517–533, January 25, 2018

intranasally infected with influenza virus, and mediastinal lymph
nodes were harvested after 0–9 days. Flow cytometry analysis
revealed that, although conventional CD4+ T cells gradually ex-
press CXCR5, PD-1, Bcl6, and SLAM and produce high levels
of IL-21 after infection, TCRb+CD1d-tetramer+ NKT cells do
not upregulate Bcl6, CXCR5, or PD-1 and express moderate
levels of SLAM and IL-21 (Figures 2A and 2B; Figures
S2A–S2D). These results indicate that NKT cells do not acquire
NKTfH cell phenotype after viral infection.
To further explore the localization of NKT cells within the
tissues of infected mice, we stained mediastinal lymph nodes
of wild-type mice with R-phycoerythrin-labeled, PBS-57-loaded
CD1d-tetramer as reported previously (Lee et al., 2015). Lymph
node sections were further stained for GL7 and analyzed by
confocal microscopy. We found that tetramer-labeled cells
localized in the periphery of GL7+ areas, whereas almost no
tetramer-labeled cells were found inside these GL7+ areas (Fig-
ure 2C). This observation indicates that NKT cells are excluded
from the germinal center after viral infection.
To directly address whether CD1d-mediated cognate interac-
tion between B cells and NKT cells is required for the develop-
ment of the antiviral B cell response, we generated CD1dflox/flox
mice to delete CD1d specifically on B cells. To this end, two
loxP sites were inserted, flanking exon 3 of CD1d, and
CD1dflox/flox mice were then crossed with mice expressing the
Cre recombinase under the promoter of CD19 or Mb1 genes
(Figure 2D; Figures S2E–S2H). Deletion of CD1d was observed
in approximately 80% of B cells in CD1dflox/floxCD19-Cre+ mice
and in 100% of B cells in CD1dflox/floxMb1-Cre+ mice, indicating
that the Mb1-Cre system is a more efficient tool to delete CD1d
on B cells (Figures 2E–2G). Then, groups of CD1dflox/floxCD19-
Cre+ and CD1dflox/floxMb1-Cre+ and their corresponding Cre
littermates were intranasally infected with influenza virus, and
mediastinal lymph nodes were harvested after 9 days. The
formation of germinal centers and TfH cells was similar
in CD1dflox/flox CD19-Cre and CD19-Cre+ mice and in
CD1dflox/flox Mb1-Cre and Mb1-Cre+ animals, indicating that
CD1d-mediated interactions between B cells and NKT cells are
not required for the development of germinal centers during viral
infection (Figures 2H–2K).
An Early NKT Cell Wave of IL-4 Precedes Late IL-4
Production by TfH Cells
One mechanism by which NKT cells might initiate antiviral B cell
immunity in a non-cognate manner is likely to be via the produc-
tion of B cell-modulatory cytokines. To examine the ability of
NKT cells to produce these molecules, we first assessed their
kinetics of activation after influenza infection by measuring the
levels of the surface activation marker CD69. We observed
that, although most NKT cells were negative for CD69 in unin-
fected mice (day 0), the number of CD69+ NKT cells peaked
3 days after infection, indicating that activated NKT cells accu-
mulated early after viral challenge (Figure 3A). Interestingly, influ-
enza infection not only results in an upregulation of activation
markers but also in a robust production of IFN-g and IL-4, with
a similar percentage of NKT cells producing either cytokine (Fig-
ure 3B; Figure S3A). Notably, even though NKT cells produce
IFN-g after infection, they represent only 1% of the total IFN-g+
cell population in lymph nodes (Figure 3C). In striking contrast,
NKT cells represent up to 50%–70% of the total IL-4+ cells after
influenza infection (Figure 3C); however, we did not detect NKT
cell production of other type 2 cytokines, such as IL-5 and
IL-13 (Figure S3B). These observations demonstrate that,
although the population of NKT cells is much smaller than those
of conventional T cells, they represent the main source of IL-4 in
lymph nodes at the early stages of influenza infection.
Previous studies have shown that TfH cells constitute essen-
tially all of the IL-4-producing T cells in draining lymph nodes
after infection (King and Mohrs, 2009; Reinhardt et al., 2009;
Yusuf et al., 2010). However, these cells usually accumulate later
during the infection process. To gain insight into the kinetics of
IL-4 production by NKT and TfH cells after influenza infection,
we took advantage of IL-4 GFP reporter mice, which concomi-
tantly express endogenous levels of IL-4 in addition to GFP
(Mohrs et al., 2001). Flow cytometry analysis revealed an accu-
mulation of NKT cells shortly after influenza infection, reaching
a maximum number on day 3, and a gradual but continuous
accumulation of TfH cells, reaching a maximum number on
day 9 (Figure 3D). In agreement with previous studies, we
observed GFP expression in approximately 20%–40% of NKT
cells even on day 0, suggesting that NKT cells already produce
IL-4 mRNAs at steady state (Figure 3E; Lee et al., 2013; Mohrs
et al., 2005). However, the percentage of GFP+ NKT cells peaks
up to 60% to 80% after 3 days of infection, indicating that influ-
enza infection rapidly induces IL-4 production by NKT cells (Fig-
ure 3E). Notably, both steady-state IL-4 producers and those
that acquire the de novo ability to produce IL-4 can contribute
to the early pool of IL-4 secretors (Figures S3C–S3E). In contrast,
the percentage of GFP+ TfH cells increases gradually after infec-
tion, with around 10% of GFP+ TfH cells on day 3 and 30% of
GFP-expressing TfH cells on day 9 of infection (Figure 3E). These
results indicate that, although NKT cells accumulate and
become IL-4 producers rapidly after infection, TfH cells differen-
tiate and produce IL-4 later throughout the infection process.
To assess the contribution of NKT and TfH cells to the pool of
IL-4-producing cells at different times of infection, we gated on

(B) Analysis of the percentage of NKT (blue) and CD4+ T (red) cells that acquire CXCR5 and PD-1 markers after infection.
(C) Confocal microscopy analysis of lymph nodes on day 9 of influenza infection incubated with CD1d tetramer (magenta) and GL7 (green). Charts show the
number and density of NKT cells inside and outside of germinal centers. Each dot represents a single germinal center. Scale bar, 60 mm.
(D) Schematic showing the CD1d locus targeted for deletion with two loxP sites flanking exon 3 of CD1d.
(E) Representative contour plot shows the gating strategy for B cells.
(F and G) Flow cytometry analysis of CD1d expression in B cells from (F) CD1dflox/floxCD19-Cre or (G) CD1dflox/floxMb1-Cre mice: Cre (blue) and Cre+ (red).
(H–K) Flow cytometry analysis of germinal center and TfH cell formation in (H and I) CD1dflox/floxCD19-Cre or (J and K) CD1dflox/floxMb1-Cre mice on day 9 of
influenza infection: Cre (blue dots) and Cre+ (red dots).
All data are representative of 3 independent experiments. Unless specified, each dot represents one mouse. Horizontal bars, mean; error bars, SEM; statistical
analysis, two tailed Student’s t test, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

Cell 172, 517–533, January 25, 2018 521

 A Lymph node cells NKT cells B NKT cells

Cytokine+ NKT cells (%)

MFI 50
105 d6 105
1107
NKT cells d5 929 104 28.3 % 29.8 %
104
d4 883
25
% Maximum
103 d3 1182 103
d2 746 IFN-γ+ IL-4+
0 d1 450 0

TCRβ
TCRβ

d0 380
0
0 103 104 105 0 103 104 105 0 103 104 105 0 103 104 105 IFN-γ+ IL-4+
CD1d-tetramer CD69 IFN-γ IL-4

C Lymph node cells IFN-γ+ cells Lymph node cells IL-4+ cells ****
100

NKT cells in cytokine+

LN population (%)
105 105

104 104 10
20.4% 0.5 % 0.2 % 60.0 %
103 103
1
0 0
IFN-γ cells
+ NKTs IL-4 cells
+
NKTs
SSC
SSC

0.1
0 10 3
10 4
10 5
0 10 10 3 4
10 5
0 10 3
10 4
10 5
0 10 10 3 4
10 5 IFN-γ+ IL-4+
IFN-γ CD1d-tetramer IL-4 CD1d-tetramer

D E NKT cells TfH cells

8 NKT cells GFP +
GFP+ 80 NKT cells
Number of cells (x104)

TfH cells d9: 65% d9: 29% IL-4 GFP+cells (%) TfH cells

4 d6: 62% d6: 11%

% Maximum

40
d3: 76% d3: 10%

d0: 40% d0: 5%

0 0
0 3 6 9 0 10 3
10 4
10 5
0 9
3 6
Days after infection IL-4 GFP Days after infection

F Gated on IL-4 GFP+ LN cells

Day 0 Day 3 Day 6 Day 9 Early NKT IL-4 Late TfH IL-4
80
% of total GFP+ LN cells

10 5
NKT 19.8% NKT 68.3% NKT 32.8% NKT 13.9% NKT cells
TfH cells
10 4
CD1d Tetramer

TfH 12.1% TfH 1.5% TfH 42.5% TfH 67.3% 40

103 104
0 103
0
0 103 104 105 0 3 6 9
CXCR5 Days after infection

Figure 3. NKT Cells Generate a Pre-TfH Cell Wave of IL-4

(A) Flow cytometry analysis of CD69 levels in the TCRb+ CD1d tetramer+ population after 0–6 days of influenza infection. MFI, mean fluorescence intensity.
(B) Representative plots showing the percentage of NKT cells producing IFN-g (blue) and IL-4 (red) after 3 days of influenza infection.
(C) Flow cytometry analysis of IFN-g (blue) and IL-4 (red) production on day 3 of influenza infection. Gates were drawn on IFN-g+ and IL-4+ lymph node cells to
analyze the percentage of NKT cells inside each cytokine+ population.
(D–F) Flow cytometry analysis of NKT (blue) and TfH (red) cells from IL-4 GFP mice infected with influenza virus.
(D) Numbers of NKT and TfH cells at different stages of influenza infection.
(E) GFP expression in NKT and TfH cells at different times of infection.
(F) Percentage of CD1d-tetramer+ NKT cells and CXCR5+ TfH cells inside the IL-4 GFP+ lymph node cell population at different stages of influenza infection.
In all charts, each dot represents one mouse. In graphs showing statistical analysis, data are representative of three independent experiments. Horizontal bars,
mean; error bars, SEM. Student’s t test, ****p < 0.0001.

522 Cell 172, 517–533, January 25, 2018

A Follicular area B Follicular borders
Wild type CD1d-/- ** Wild type CD1d-/- *
B220 18 30

CD1d-tet+ cells/follicle

CD1d-tet+ cells/border
CD1d-tet

9 15

0 0
WT CD1d-/- WT CD1d-/-

C D
Lymph node Follicular area Follicular borders ** Lymph node cells IL-4 GFP+ cells

GFP+CD1d-tet+ cells (%)

IL-4 GFP 100
10 5
CD1d-tet
B220 IL-4 GFP+ CD1d-tet- CD1d-tet+
104
50
103
1.3% 28% 72%
0

SSC
0
Follicular Follicular 0 103 104 105 0 103 104 105
area borders IL-4 GFP CD1d-tetramer

E IL-4 GFP+ lymph node cells F CD3e Zbtb16 (PLZF) CXCR6 CXCR3
14 10 12 12
Tet , Cluster 1
+

Tet+, Cluster 2
Expression level

Tet-, Cluster 3
Tet-, Cluster 4
tSNE2

7 5 6 6

0 0 0 0
tSNE1 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
Clusters
G Gata3 Tbx21 (T-bet) Rorc (RORγt) H
12 12 10 NKT cells CXCR3 levels
100 80 **
105
% CXCR3+ NKT cells

IL-4 GFP- IL-4 GFP+

Expression level

GFP- GFP+
104
6 6 5 50 40
103
0
SSC

0 0
0 103 104 105 0 103 104 105
GFP- GFP+
0 0 0 IL-4 GFP CXCR3
1 2 3 4 1 2 3 4 1 2 3 4
Clusters

I NKT cells CXCR3- CXCR3+ J

40
* 200 **
105 105 CXCR3-
(per 2x104 NKTs)
% IL-4+ NKT cells

CXCR3- CXCR3+
IL-4 secretors

11.1% CXCR3+
104 104 35.2%
20 100
103 103
0 0
SSC

SSC

0 0
0 103 104 105 0 103 104 105 CXCR3- CXCR3+
CXCR3 IL-4

(legend on next page)

Cell 172, 517–533, January 25, 2018 523

TCRb+ IL-4-GFP+ lymph node cells and analyzed the proportion
of this population that was CD1d-tetamer+ (NKT cells) or
CXCR5+ (TfH cells). Interestingly, we observed that, 3 days after
influenza infection, almost 70% of GFP+ cells were NKT cells,
whereas less than 2% were TfH cells (Figure 3F). This trend is
reversed around 6 days after infection, and, by day 9, less than
15% of GFP+ cells were NKT cells, whereas almost 70% were
TfH cells (Figure 3F). These results indicate that, during influenza
infection, there is an early wave of IL-4, in which NKT cells consti-
tute the main source of this cytokine, and a late wave of IL-4,
where TfH cells overcome NKT cells as the main IL-4 producers.
An Early NKT Cell Wave of IL-4 Occurs at the Follicular
Borders
So far, we have defined the temporal frame of IL-4 production by
NKT cells during the early stages of influenza infection. To gain
insight into the spatial distribution of these IL-4-producing NKT
cells, we infected wild-type, CD1d / , and IL-4 GFP reporter
mice with influenza virus and harvested mediastinal lymph nodes
after infection. The lymph nodes were incubated with labeled
PBS-57-loaded CD1d-tetramer, and sections were further
stained against B220 and CD169, a macrophage marker, and
analyzed by confocal microscopy. Although CD1d-tetramer+
cells were nearly undetectable in uninfected animals, they were
observed inside B cell follicles and in direct contact with
CD169+ macrophages at the subcapsular sinus and interfollicu-
lar areas by day 3 of infection (Figures 4A and 4B; Figure S4A). In
contrast, CD1d-tetramer+ cells were almost absent in lymph no-
des from CD1d / animals, indicating that CD1d-tetramer+ cells
are most likely NKT cells (Figures 4A and 4B). Interestingly,
although the great majority of CD1d-tetramer+ cells located
inside the B cell follicles do not express GFP, most of the
GFP+ cells appear to be located in the areas surrounding B cell
follicles (Figure 4C). These results indicate that early IL-4 produc-
tion is restricted to the periphery of the B cell follicles, where
antigen-specific B cells relocate to recruit T cell help after
activation.
To get insight into the extent of cellular heterogeneity within
the immune cells contributing to the early IL-4 wave, we charac-
terized the IL-4 GFP+ population by flow cytometry 3 days post-
infection with influenza virus. This experiment shows that
approximately 70% of this population corresponds to NKT cells
(CD1d-tet+), whereas the remaining 30% (CD1d-tet ) are mostly
T cells (TCRb+), excluding a significant contribution of ILC2s to
the early wave of IL-4 (Figure 4D; Figure S4B). To further study
these two cell populations, IL-4 GFP+ CD1d-tet+ and IL-4
GFP+ CD1d-tet cells were single cell-sorted and analyzed by
RNA sequencing (RNA-seq). After processing (STAR Methods),
we retained 222 cells from the CDd1-tet+ (n = 113) and CDd1-
tet (n = 109) populations in our single-cell RNA-seq analysis.
T-distributed stochastic neighbor embedding (t-SNE) and
shared nearest neighbor (SNN) clustering revealed that single
cells from the two subsets (CD1d-tet+ and CD1d-tet ) each
bifurcate into two separate clusters, reflecting divergent tran-
scriptional states within these populations (Figures 4E; Fig-
ure S4C; Table S1). Independent of the cluster, most of these
cells express CD3e, confirming their T cell origin (Figure 4F).
A feature of the CD1d-tet subset clusters is the expression of
CCR7, L-selectin, and S1PR1, mRNAs encoding proteins that
control T cell migration (Figure S4D). Furthermore, one of the
CD1d-tet clusters is characterized by cells expressing high
levels of CD4 and the other one by cells expressing CD8a. Impor-
tantly, only a small proportion of the cells in these two clusters
express Bcl-6, ruling out a significant contribution of TfH cells
to this early IL-4 wave (Figures 3F; Figures S4D and S4E). On
the other hand, significant numbers of cells from the CD1d-tet+
subset express PLZF, a transcription factor that directs the
NKT cell effector program (Figure 4F). Interestingly, a large pro-
portion of cells from both CD1d-tet+ clusters express the chemo-
kine receptor CXCR6, whereas only one of them expresses
CXCR3 (Figure 4F). Both CXCR3+ and CXCR3 clusters express
GATA-3, whereas the CXCR3+ group specifically expresses
T-bet, and the CXCR3 one expresses RORgt, an observation
that was further confirmed by flow cytometry (Figures 4G; Fig-
ures S4F and S4G).
CXCR3 has been reported to recruit T cells to the interfollicular
zones, the areas where we find most of the IL-4 GFP+ NKT cells
after infection (Groom et al., 2012; Sung et al., 2012). Interest-
ingly, flow cytometry analysis of lymph nodes from influenza-
infected mice revealed that 80% of IL-4 GFP+ NKT cells express
CXCR3, whereas only 30% of IL-4 GFP NKT cells express
this marker (Figure 4H). Therefore, we wondered whether

Figure 4. The Early IL-4 Wave Is Localized at the Periphery of B Cell Follicles
(A–C) Confocal microscopy analysis of (A and B) wild-type and CD1d / and (C) IL4-GFP mice on day 3 of influenza infection. Lymph nodes were labeled with
CD1d tetramer (magenta) and anti-B220 antibody (green in A and B and white in C). The arrows in (C) indicate CD1d tetramer+ cells expressing IL-4 (IL-4
GFP, green). Scale bars, 300 mm (lymph node) and 60 mm (section).
(D) Flow cytometry analysis of IL-4 GFP+ cells in mediastinal lymph nodes on day 3 of influenza infection, showing CD1d-tet and CD1d-tet+ cells.
(E) t-SNE plots of CD1d-tet and CD1d-tet+ subsets. The CD1d-tet+ population separates into two clusters (1, dark purple; 2, light purple), as does the CD1d-tet
population (3, light green; 4, dark green).
(F and G) Expression distribution (violin plots) in each population (horizontal axes) for (F) CD3e, Zbtb16, CXCR6, and CXCR3 and (G) Gata3, Tbx21, and Rorc.
Expression levels are log2(tpm+1).
(H) Flow cytometry analysis of CXCR3 levels in IL-4-GFP (red) and GFP+ (blue) NKT cells on day 3 of influenza infection. Lines connect cells from the
same mouse.
(I) Flow cytometry analysis of IL-4 production by CXCR3 (red) and CXCR3+ (blue) NKT cells on day 3 of influenza infection. Lines connect cells from the
same mouse.
(J) ELISPOT analysis of IL-4-producing CXCR3 (red) and CXCR3+ (blue) NKT cells sorted on day 3 of influenza infection. Each dot represents pooled NKT cells
from 6 mice.
In graphs showing statistical analysis, data are representative of at least two independent experiments. Horizontal bars, mean; error bars, SEM. Student’s t test,
*p < 0.05, **p < 0.01.

524 Cell 172, 517–533, January 25, 2018

CXCR3 and CXCR3+ NKT cells would have a differential ability
to produce IL-4 after infection. To this end, we analyzed IL-4 pro-
duction by both CXCR3 and CXCR3+ NKT cell groups on day 3
of influenza infection by flow cytometry and enzyme-linked im-
munospot (ELISPOT). Strikingly, the number of cells secreting
IL-4 in the CXCR3+ NKT cell sub-population is five times higher
than in the CXCR3 group (Figures 4I and 4J). These results sug-
gest that CXCR3 preferentially drives the recruitment of IL-4-pro-
ducing NKT cells to interfollicular areas.
Taken together, our results indicate that influenza infection
triggers two spatiotemporally divergent waves of IL-4: an early
wave, mainly produced by NKT cells and restricted to the periph-
ery of B cell follicles, and a late one, produced by germinal cen-
ter-resident TfH cells.
Resident Macrophages Trigger an Early NKT Cell Wave
of IL-4 and B Cell Immunity
To define the immune machinery involved in the induction of
this early NKT cell-derived IL-4, we specifically abrogated
CD1d expression on different immune cell populations by
crossing our CD1dflox/flox mice with animals expressing the
Cre recombinase under the control of the Mb1, Lyz2, or
CD11c gene promoters. Flow cytometry analysis showed an
effective deletion of CD1d on Gr-1+CD11b+ neutrophils/
myeloid-derived suppressor cells (MDSCs) in CD1dflox/flox-
+
Lyz2-Cre mice and on MHCII+CD11c+ dendritic cells in
flox/ex
CD1d CD11c-Cre+ mice (Figures S5A–S5C). However,
no deletion of CD1d could be obtained in lymph node
CD169+ macrophages in these transgenic lines (Figure S5D).
Groups of CD1dflox/floxMb1-Cre+, CD1dflox/floxLyz2-Cre+, and
CD1dflox/exCD11c-Cre+ and their corresponding Cre litter-
mates were intranasally infected with influenza virus, and
mediastinal lymph nodes were harvested after 3 or 9 days.
Interestingly, similar percentages of IL-4+ NKT cells (day 3)
and germinal centers and TfH cells (day 9) were observed
among these transgenic lines (Figures 2H–2K and 5A–5C; Fig-
ures S5E–58H). These results indicate that CD1d-mediated
cognate interactions between NKT cells and B cells, neutro-
phils/MDSCs, or dendritic cells are not required for NKT cell-
mediated induction of IL-4 production or germinal center
formation.
Because we were unable to delete CD1d on CD169+ macro-
phages by using CD11c or Lyz2 Cre lines, we decided to take
a macrophage depletion approach to assess the importance of
these macrophages in the induction of IL-4 production by NKT
cells. Accordingly, mice expressing the diphtheria toxin receptor
(DTR) under the CD169 promoter, here named CD169-diphtheria
toxin receptor mice, were treated with PBS or diphtheria toxin.
Although PBS injection has no effect on macrophages, a single
injection of diphtheria toxin is sufficient to deplete both subcap-
sular sinus and medullar macrophages from lymph nodes
(Figures 5D and 5E). After 4 days of PBS or diphtheria toxin
administration, mice were intranasally infected with influenza
virus, and mediastinal lymph nodes were harvested 3 or 9 days
later. Interestingly, as revealed by flow cytometry, the production
of IL-4 by NKT cells (day 3) as well as the formation of germinal
centers (day 9) were severely impaired in macrophage-depleted
mice (Figures 5F and 5G). Notably however, the development of
TfH cells remained unaltered in the absence of macrophages
(Figure 5H). These results indicate that CD169+ macrophages
are essential for early triggering of NKT cell effector programs
and the initiation of B cell response following viral infection.
Induction of NKT Cell-Derived IL-4 by Macrophages
Requires CD1d and IL-18
Because the depletion of CD169+ macrophages abolished
the production of IL-4 by NKT cells, we investigated whether
CD1d-mediated interactions between CD169+ macrophages
and NKT cells are required for NKT cell activation. To address
this question, we generated mixed bone marrow chimeras lack-
ing CD1d expression specifically on CD169+ macrophages. For
this, CD169-DTR mice were sub-lethally irradiated and, the
following day, adoptively transferred with either 80% CD169-
DTR cells + 20% wild-type cells or 80% CD169-DTR cells +
20% CD1d / cells. After 1, 4, and 7 weeks, mice were treated
with diphtheria toxin, and, at week 8 of reconstitution, both
groups of chimeras were intranasally infected with influenza virus
(Figure 6A). Although we observed a robust production of IL-4 in
CD169-DTR/wild-type chimeras by day 3 of infection, NKT cells
from CD169-DTR/CD1d / chimeras were impaired regarding
this cytokine (Figure 6B). This result demonstrates that close
interactions between CD169+ macrophages and NKT cells
through CD1d are necessary to induce IL-4 production by NKT
cells after viral infection.
The depletion of CD169+ macrophages has a more profound
effect on the induction of IL-4 production by NKT cells than solely
the deletion of CD1d on these cells. Therefore, we wondered
whether CD169+ macrophages could enhance IL-4 synthesis
through the secretion of co-stimulatory cytokines. To address
this question, we stained lymph nodes on day 2 of influenza
infection with antibodies against CD169, IL-1b, IL-18, and
IL-33. These cytokines are secreted early after infection and
have been shown to modulate the differentiation and function
of innate and adaptive lymphoid cells (Kastenmüller et al.,
2012; Sagoo et al., 2016). We detected the accumulation of
IL-1b, IL-18, and IL-33 in both subcapsular sinus and medullar
macrophages, suggesting that these resident macrophages
are a source of inflammatory cytokines (Figures 6C and 6D; Fig-
ure S6A). Interestingly, their production is not the consequence
of direct infection of macrophages because no viral replication
was detected on them when a recombinant influenza virus
carrying a GFP reporter gene was used (Manicassamy et al.,
2010; Figure S6B).
Next we tested whether macrophage-derived IL-1b, IL-18, or
IL-33 are required to induce NKT cell production of IL-4. These
three cytokines are known to signal through the MyD88 adaptor
protein (Adachi et al., 1998; Schmitz et al., 2005); hence, as a
first approach, we compared the ability of wild-type versus
MyD88 / NKT cells to produce IL-4 after influenza infection.
Our flow cytometry results show that, although a substantial pro-
portion of lymph node NKT cells from wild-type mice produce
IL-4, NKT cells from MyD88 / animals are severely impaired
in the production of this cytokine (Figure 6E). Importantly, their
impairment is not due to lack of Toll-like receptor (TLR) signaling
because NKT cells from TLR7 / mice secrete IL-4 to the same
extent as NKT cells from wild-type animals (Figure 6F). We then
Cell 172, 517–533, January 25, 2018 525
A B
CD1d Mb1-Cre- CD1d Mb1-Cre+ CD1d Lyz2-Cre- CD1d Lyz2-Cre+
20 40

IL-4+ NKT cells (%)

10 5
105

IL-4+ NKT cells (%)

10.5 % 15.0 % 14.3 % 18.7 %
104 104
10 20
103 103

0
TCRβ

TCRβ
0
0 0
0 103 104 105 Cre- Cre+ 0 103 104 105 Cre- Cre+
IL-4 IL-4

C D CD169DTR/+PBS
CD1d CD11c-Cre -
CD1d CD11c-Cre +
Lymph Node Subcapsular sinus Medulla
16
IL-4+ NKT cells (%)

105
CD169
10.4 % 11.6 % B220
10 4

8
103

0
TCRβ

0
0 103 104 105 Cre- Cre+
IL-4

E CD169DTR/+/DT F
CD169DTR/+/DT
Lymph Node Subcapsular sinus Medulla
CD169DTR/+/PBS
40
**

IL-4+ NKT cells (%)

10 5
CD169
B220 104 19.1 % 2.6 %
20
103

0
TCRβ

0
0 102 103 104 105 PBS DT
IL-4

G H
CD169DTR/+/PBS CD169DTR/+/DT CD169DTR/+/PBS CD169DTR/+/DT
****
CXCR5+ PD-1+ T cells (%)

12 26
Fas+ GL7+ B cells (%)

10 5
10 5

9.0% 1.3% 19.9% 18.9%

104 104
6 13
103 103
CXCR5

0 0
Fas

0 0
0 103 104 105 PBS DT 0 103 104 105 PBS DT
GL7 PD-1

Figure 5. Resident Macrophages Promote Early IL-4 Production by NKT Cells and Antiviral B Cell Immunity
(A–C) Flow cytometry analysis of IL-4 production by NKT cells in (A) CD1dflox/floxMb1-Cre, (B) CD1dflox/floxLyz2-Cre, and (C) CD1dflox/exCD11c-Cre mice on day 3
of influenza infection: Cre (blue dots) and Cre+ (red dots).
(D and E) Confocal microscopy analysis of CD169DTR/+ mice after 4 days of (D) PBS or (E) diphtheria toxin (DT) administration. Sections were stained with
antibodies to B220 (white) and CD169 (green). Scale bars, 300 mm (left) and 60 mm (center and right).
(F) Flow cytometry analysis of IL-4 production by NKT cells from CD169DTR/+/PBS-treated or CD169DTR/+/DT-treated mice on day 3 of influenza infection: PBS
(blue dots) and DT (red dots).
(G and H) Flow cytometry analysis of germinal center (G) and TfH (H) cells in CD169DTR/+/PBS-treated and CD169DTR/+/DT-treated mice on day 9 of influenza
infection: PBS (blue dots) and DT (red dots).
In all panels, each dot represents one mouse. Data show one representative result from three experiments. Data represent mean ± SEM; two-tailed paired
Student’s t test, **p < 0.01, ****p < 0.0001.

526 Cell 172, 517–533, January 25, 2018

A B
CD169/Wild type CD169/CD1d-/-
25
**
Irradiation Bone marrow Diphteria toxin Harvest

IL-4+ NKT cells (%)

CD169-DTR transfer administration infection LNs 10 5

18.9 % 7.3 %
104
12.5
-1 0 7 28 49 56 59 103
Days
0

TCRβ
i) 80% CD169-DTR 0
20% Wild type 0 103 104 105 CD169 CD169
ii) 80% CD169-DTR IL-4 WT CD1d-/-
20% CD1d-/-

C IL-1β IL-18 IL-33 D IL-1β IL-18 IL-33

CD169 CD169 CD169 CD169 CD169 CD169
Subcapsular sinus

IL-1β IL-18 IL-33 IL-1β IL-18 IL-33

Medulla

E F
Wild type TLR7 -/-
Wild type MyD88-/- *** 42
40

IL-4+ NKT cells (%)

105
IL-4+ NKT cells (%)

105
20.4% 5.0% 38.0 % 31.2 %
104 104
20 21
103 103
0
TCRβ
TCRβ

0 0
0
0 10 3
10 4
10 5 WT MyD88-/- 0 103 104 105 WT TLR7-/-
IL-4 IL-4

G H
Wild type IL-1R -/- Wild type IL-18R -/- **
40 IL-4+ NKT cells (%) 20
105
IL-4+ NKT cells (%)

10 5

24.2 % 24.8 % 13.9 % 7.0 %

104 104
20 10
103 103

0 0
TCRβ
TCRβ

0 0
WT IL-1R-/- 0 103 104 105 WT IL-18R-/-
0 103 104 105
IL-4 IL-4

I J
Wild type-CD45.1 IL-33R-/--CD45.2 NKT cells IL-4 secretors
140
50
IL-4+ NKT cells (%)

105
per 2x104 cells
IL-4 secretors

19.2% 20.1% 0 10
104
70 0.1 100
25
103

0 1 1000
TCRβ

0
0 0 0.1 1 10 100 1000
0 103 104 105 WT IL-33R-/-
IL-18 (ng/ml) IL-18 (ng/ml)
IL-4

Figure 6. Resident Macrophages Promote Early NKT Cell IL-4 Production through CD1d and IL-18
(A) Experimental scheme showing the strategy for the generation of chimeras lacking CD1d specifically on CD169+ macrophages.
(B) Flow cytometry analysis of IL-4 production by NKT cells from mice subjected to the experimental scheme show in (A): wild-type chimeras (blue dots) and
CD1d / chimeras (red dots).
(legend continued on next page)

Cell 172, 517–533, January 25, 2018 527

explored whether signaling through IL-1R, IL-18R, or IL-33R is
involved in this process. We performed influenza infection in
wild-type, IL-1R / , or IL-18R / mice and mixed bone marrow
chimeras transferred with 50% CD45.1 wild-type and 50%
CD45.2 IL-33R / cells. We found that IL-1R and IL-33R were
not required for IL-4 production because NKT cells from
IL-1R / mice and IL-33R / experimental chimeras did not
show an impairment in IL-4 production compared with their
wild-type counterparts (Figures 6G–6I). In contrast, the percent-
age of NKT cells producing IL-4 was significantly reduced in
IL-18R / mice, suggesting that IL-18 enhances IL-4 secretion
by NKT cells (Figure 6H). Remarkably, MyD88 and IL-18R were
intrinsically required by NKT cells because MyD88 / and
IL-18R / NKT cells displayed a reduced ability to produce
IL-4 compared with wild-type cells in a mixed bone marrow
chimera setup (Figures S6C–S6E). Furthermore, IL-18 could
induce IL-4 production on itself when added ex vivo to primary
NKT cells (Figure 6J). Altogether, our results indicate that the
IL-18R-MyD88 axis plays an important role in the production of
IL-4 by NKT cells.
An Early IL-4 Wave Initiates B Cell Immunity to Viral
Infection
Because of the central role of NKT cells in producing IL-4 early
after infection, we wondered whether CD1d / mice would
have reduced numbers of IL-4-secreting cells. To address this
question, wild-type and CD1d / mice were infected with influ-
enza virus, and the amount of IL-4+ cells was measured on
day 3. The number of cells secreting IL-4 was significantly
reduced in CD1d / compared with wild-type animals, indi-
cating that other immune cells cannot compensate for the
absence of NKT cells (Figure 7A). Next, to test whether the
impaired B cell responses observed in CD1d / animals could
be due to reduced IL-4 production, FVB wild-type and IL-4 /
mice were infected with influenza virus, and, 8 days later, medi-
astinal lymph nodes were harvested. A substantial expansion of
germinal center and TfH cells was observed in wild-type mice
compared with uninfected animals (Figures 7B and 7C). Interest-
ingly, in IL-4 / mice, there was a significant reduction in the pro-
portion of germinal centers, whereas the differentiation of TfH
cells remained similar to that of wild-type animals (Figures
7B and 7C). Furthermore, germinal center cells from IL-4 /
mice have a reduced capacity to class-switch to IgG1 (Fig-
ure 7D). These results show that IL-4 is necessary for the induc-
tion of the B cell response after influenza infection.
Then we asked whether the early wave of IL-4 was necessary
for the induction of B cell responses after influenza infection. To
address this question, wild-type mice were infected with influ-
enza virus, and a sub-group of them was treated with a blocking
IL-4 antibody during the first 3 days of infection. By day 8, we
(C and D) Confocal microscopy analysis on day 2 of influenza infection showing (C) SCS and (D) medullar areas. Sections were stained with antibodies to CD169
(green) and IL-1b/IL-18/IL-33 (magenta). Scale bars, 60 mm.
(E–I) Flow cytometry analysis of wild-type and (E) MyD88 / , (F) TLR7 / , (G) IL-1R / , and (H) IL-18R / mice or (I) IL-33R / chimeric mice on day 3 of influenza
infection. Charts show IL-4+ NKT cells in wild-type (blue dots) or mutant (red dots) populations.
(J) ELISPOT analysis of IL-4 production by sorted NKT cells incubated ex vivo with IL-18.
Each dot represents one mouse. Data show mean ± SEM and are representative of 3 independent experiments. Statistical analysis: two-tailed Student’s t test,
**p < 0.01, ***p < 0.001.
528 Cell 172, 517–533, January 25, 2018
observed a significant reduction in the formation of germinal
centers and in their ability to class-switch to IgG1 in mice treated
with anti-IL-4 (Figures 7E and 7F; Figure S7A). Importantly, when
we injected anti-IL-4 3 days prior to infection, this reduction was
not significant, suggesting that the blocking effect of this anti-
body only lasts for a few days (Figures S7B). Altogether, these
results suggest that the early IL-4 wave is necessary to promote
B cell immunity.
Subsequently, to check whether the recovery of the initial
IL-4 wave could rescue the germinal center defect observed
in NKT cell-deficient mice, we administered PBS or IL-4 com-
plexed with an anti-IL4 antibody (IL-4c) to groups of CD1d /
mice during the initial 2 days of infection. Remarkably, admin-
istration of IL-4c for the initial days of infection is sufficient to
restore a robust germinal center response and enhance IgG1
class-switching in the absence of NKT cells (Figures 7G
and 7H). In contrast, administration of more antigen-specific
TfH cell precursors to CD1d / animals did not rescue the
germinal center defect, indicating that the TfH cells might not
compensate for the absence of NKT cells (Figures S7C–S7F).
Altogether, our results indicate that the early wave of IL-4 trig-
gered after influenza infection is critical for the induction of
B cell responses.
Previous studies have shown that IL-4 supports the energy
and biosynthetic needs necessary for growth, proliferation, and
effector functions of immune cells (Dufort et al., 2007; Huang
et al., 2016; Ricardo-Gonzalez et al., 2010). Thus, it is feasible
that the early IL-4 wave would induce metabolic reprograming
on antigen-experienced B cells to fulfil the high metabolic re-
quirements associated with the generation of germinal centers.
To test this hypothesis, we initially cultured murine naive B cells
or B cells stimulated through their B cell receptor (BCR)
(anti-IgM) in the presence or absence of IL-4. Interestingly, we
observed that the addition of extracellular IL-4 to our ex vivo
B cell cultures increases the oxygen consumption rate (OCR) in
both naive and antigen-stimulated lymphocytes (Figure S7G).
To address whether early IL-4 secretion could induce metabolic
changes on B cells in vivo, we infected groups of mice with influ-
enza virus and treated them with either PBS or anti-IL4 blocking
antibody. We isolated lymph node cells on day 3 of infection,
cultured them with anti-IgM and anti-CD3ε for 16 hr, and
measured OCR levels in purified B cells after the incubation.
We found that B cells from lung-draining lymph nodes have a
higher OCR than B cells from non-draining lymph nodes, indi-
cating that there is a metabolic reprograming of B cells early after
influenza infection (Figure S7H). Interestingly, the ability of B cells
to increase their respiratory capacity after viral infection requires
early IL-4 production because the upregulation of the OCR is
diminished when the early IL-4 wave is inhibited (Figure S7H).
These results suggest that the early IL-4 wave triggered after
 A Wild type CD1d-/- ** B Wild type IL-4-/- **
5 15

Fas+ GL7+ B cells (%)

10 5
10 5

IL-4+ cells (x103)

13.3% 7.8%
10 4 0.25 % 0.08 % 104
2.5 7.5
103 103
0 0
SSC

Fas
0 0
WT CD1d-/- WT IL-4-/-
0 103 104 105 0 103 104 105
IL-4 GL7

C Wild type IL-4-/- D Wild type IL-4-/- *

CXCR5+PD-1+ T cells (%)

24 5.4
10 10

IgG1+ GC cells (%)

5 5

18.7% 18.0% 4.1% 1.8%

104 104
12 2.7
103 103
CXCR5

SSC
0 0
0 0
0 103 104 105 WT IL-4 -/-
0 103 104 105 WT IL-4-/-
PD-1 IgG1

* Control α-IL4 (d0-d3)

E Control α-IL4 (d0-d3) F
12
*
7.0
Fas+ GL7+ B cells (%)

10

IgG1+ GC cells (%)

5
10 5

2.3% 9.7% 4.1%

5.7% 104
104
6
3.5
103
103
SSC

0
0
Fas

0 0
0 103 104 105 Control α-IL4
0 10 3
10 4
10 5 Control α-IL4
IgG1 (d0-d3)
GL7 (d0-d3)

G CD1d-/- CD1d-/- + IL-4c (d0-d3) *** H CD1d-/- CD1d-/- + IL-4c (d0-d3) **

14 44
Fas+ GL7+ B cells (%)

105 105

IgG1+ GC cells (%)

2.9% 11.8% 14% 38%
104 104
7 22
103 103
SSC

0 0
Fas

0 0
0 10 3
10 4
10 5 CD1d-/- CD1d-/- 0 103 104 105 CD1d-/- CD1d-/-
GL7 +IL-4c IgG1 +IL-4c

Zika virus Harvest LNs Collect Serum IL-4

I K
infection Gene expression NAbs GRB2
UBE2V2
AMICA1
STK17B NKT
SETX
TMEM49
ETS1 ST6GAL1 CCR6 VAX2
STK4 ID 2
0 3 7 ITK BCL6 RBMS1
CCL26
GPR114
ST18 LRRC17 CENPA
AKT2 MYCN
AURKB
SOS1 ICAM 2 GIMAP4
MMP25
IL2RB KDM1 A RAMP3 JAG1
J Lymph node day 3 vs. Neutralizing antibodies day 7
PIK3CA
BAD
SDC1 GPX1
TK1
CCNB1
RASA2 CHAD SOX13 KCNK1
GATA3 AVPI1
NKT.POPULATIONS FAM69 A RNF208
NAA1 6 F10 TACC3
IL1R1 BLK ACTN 2
PLEKHF1
ITM2 A
TOP50. MOST.DEG.NKT1 CYFIP2
BATF NPL IL17RE PMF1 CDK1
EOMES IL4
TOP50. MOST.DEG.NKT2 IL6R CDCA8
CD8B STAT4 RRM1
RORA TESC GMNN
CD6 WDR76 ASF1B
TOP50. MOST.DEG.NKT17 STAT6 PDE9A
ZMYND1 1
IL18R1 TKT
IL24 CD2
KLF12 BCL11B
IL18 SIGNALING
HIVEP2 LEF1
PTPN22
SENP7 MAPK8
IL4.SUPERPATHWAY.GENECARD FASLG IL7R
GLB1 SYTL3
PRDM1 ISY1
CD28
LRRFIP1 CD247
STAT6 CHIP_SEQ TARGET GENES TMEM43 SKI ITPKB IL2RA
RAB35
TFH SIGNALING
CA6
ITGB2
RAB8B BSCL2 EIF2C2
FAM188 A
SAMM50 IL18
IKBKE
TBC1D2 3 PGM1
SYNGAP1
Genesets normalized enrichment score RGS12
TFPT RNF168

RFX3
-2
KIF13B
+2 MED15
Shared genes

(legend on next page)

Cell 172, 517–533, January 25, 2018 529

infection could be regulating seeding of germinal center cells
through the induction of metabolic changes in antigen-specific
B cells.
Early NKT Cell and IL-4 Signatures in Lymph Nodes from
Zika-Infected Macaques
We were interested to find out whether the early role for NKT cells
and IL-4 in the initiation of antiviral B cell immunity might extend
to other species. To address this question, we analyzed tran-
scriptomic data from lymph nodes of rhesus macaques that
were subcutaneously infected with Zika virus (Aid et al., 2017).
We assessed the transcriptomic profiles of lymph nodes ob-
tained on day 3 or 14 post-infection to identify pathways that
correlated with Zika-specific neutralizing antibody titers on day
7 and day 14, respectively (Figure 7I; Figure S7I). Pathway
enrichment analysis revealed a significant and positive correla-
tion of NKT cells (normalized enrichment score [NES] = 1.48,
false discovery rate [FDR] = 0.06), IL-4 (NES = 1.43, FDR =
0.02), IL-18 (NES = 1.42, FDR = 0.05), and STAT6 signaling
(NES = 1.83, FDR = 0.02) signatures as early as day 3 of infection
with neutralizing antibody titers at day 7 (Figure 7J). Furthermore,
gene-interacting network inference showed the high intercon-
nectivity of all these pathways, highlighted by co-expression of
their leading genes with a central role of IL-4 in connecting all
of these genes (Figure 7K). Importantly, no significant correlation
between TfH cell signatures (FDR > 0.20) and neutralizing anti-
bodies was found at this early time point (Figure 7J). In contrast,
a positive correlation between TfH cell signatures and neutral-
izing antibodies was observed on day 14 of infection (NES =
1.23, FDR = 0.13) (Figures S7J and S7K). These correlates are
in line with our previous findings, pointing out to a possible role
for NKT cells at the early stages of B cell immunity, with TfH cells
only appearing later on the infection process. Furthermore, these
analyses indicate that an early pre-TfH wave of NKT cells and
IL-4 might be conserved upon influenza and Zika virus infection
in mice and primates, respectively.
DISCUSSION
Several studies have focused on the use of NKT cell agonists
as vaccine adjuvants to boost NKT cell activity and immune
responses. In particular, immunization with inactivated influenza
virus or virus-derived proteins in combination with a-GalCer, a
synthetic glycosphingolipid, results in strong protection against
subsequent challenges with various strains of influenza virus
(Artiaga et al., 2016; Galli et al., 2007). However, whether NKT
cells could play a role in the induction of adaptive immune re-
sponses in the absence of exogenous glycolipids was not clear.
Here we showed that NKT cells are required for the early seeding
of germinal center B cells and the production of class-switched
antibody-secreting cells in response to influenza infection.
Furthermore, we found that NKT cells can enhance B cell immu-
nity to influenza-derived proteins when administered together
with TLR adjuvants. Our findings are in line with previous studies
showing that NKT cells enhance the activity of CD8+ T cells after
influenza infection even in the absence of exogenous glycolipids
(De Santo et al., 2008; Ishikawa et al., 2010). Therefore, the
importance of NKT cells in the induction of adaptive immunity
goes beyond the presence of glycolipid antigens.
How do NKT cells promote B cell responses during infection?
Previous work showed that, in response to bacterium-derived
glycolipids, NKT cells differentiate into NKTfH cells that engage
in prolonged cognate interactions with B cells. These CD1d-
mediated interactions rapidly induce germinal center and extra-
follicular plasma cell formation, leading to the production of high
titers of IgM and early class-switched antibodies (Bai et al., 2013;
Barral et al., 2008; Chang et al., 2011; King et al., 2011; Leadbet-
ter et al., 2008). However, we found that, upon influenza virus
infection, NKT cells do not differentiate into NKTfH cells, nor
they localize in germinal centers. Furthermore, CD1d-mediated
cognate interactions between B cells and NKT cells are dispens-
able for the induction of antiviral B cell responses. Therefore,
even though NKT cells promote B cell immunity in response to
both bacterium-derived glycolipids and viral infection, the mech-
anism behind this regulation varies depending on the absence or
presence of pathogen-derived glycolipids on the surface of the
infectious agent.
Our results show that, during the initial stages of a viral infec-
tion, NKT cells become the main source of IL-4 in draining lymph
nodes. Flow cytometry and single-cell RNA-seq analysis re-
vealed that around 70% of the IL-4-producing cells are NKT
cells, whereas the remaining ones are mainly T cells, excluding
a significant contribution by ILC2s and TfH cells to the IL-4 pro-
duction at early stages of infection. This finding was unexpected
because previous studies reported that TfH cells constitute
essentially all of the IL-4 producing cells in lymph nodes in

Figure 7. A Conserved Early IL-4 Wave Is Critical for Induction of B Cell Immunity
(A) Flow cytometry analysis on day 3 of influenza infection. Graphs show the number of IL-4+ lymph node cells in wild-type (blue dots) and CD1d /
(red dots) animals.
(B–D) Flow cytometry analysis on day 8 of influenza infection. Graphs show the percentage of (B) germinal center, (C) TfH, and (D) IgG1+ cells in wild-type
(blue dots) and IL-4 / mice (red dots).
(E and F) Flow cytometry analysis on day 8 of influenza infection of wild-type mice treated with PBS or blocking a-IL-4 antibody during the initial 3 days of infection.
Graphs show (E) germinal center and (F) IgG1+ cells in wild-type (blue dots) and a-IL-4 treated mice (red dots).
(G and H) Flow cytometry analysis on day 8 of influenza infection of CD1d / mice treated with PBS or IL-4 complex (IL-4c) during the initial 3 days of infection.
Graphs show (G) germinal center cells and (H) IgG1+ cells in wild-type (blue dots) and IL-4c-treated mice (red dots).
(I) Experimental scheme showing the Zika virus infection strategy.
(J) Linear regression model of gene signatures for NKT cells, TfH cells, IL-4, STAT6, and IL-18 from harvested lymph nodes on day 3 with neutralizing antibodies
measured in plasma on day 7. Red squares, positive correlation; blue squares, negative correlation; white squares, no significant correlation.
(K) Gene-interacting network inference representing interconnectivity of NKT cells, IL-4, STAT6, and IL-18 pathways.
In (A)–(H), each dot represents a single mouse. Data are representative of three independent experiments. Statistical analysis: two-tailed Student’s t test,
*p < 0.05, **p < 0.01, ***p < 0.001.

530 Cell 172, 517–533, January 25, 2018

response to infection with different pathogens (King and Mohrs,
2009; Reinhardt et al., 2009; Yusuf et al., 2010). However, this
previously unreported NKT cell-mediated wave of IL-4 produc-
tion occurs at the early stages of infection, even before the
appearance of TfH cells or germinal centers. Furthermore, this
phenomenon is restricted to the follicular borders, where B cells
migrate immediately after pathogen encounter. The spatiotem-
porally restricted production of IL-4 by NKT cells appears to
instruct antigen-activated B cells at the periphery of B cell folli-
cles early after infection for the subsequent seeding of germinal
centers. Interestingly, a recent report showed that TfH cells only
switch to IL-4 production at the late stages of the infection pro-
cess (Weinstein et al., 2016). Therefore, it will be important to
discern the contribution of the early NKT cell IL-4 wave at the
follicular borders versus the late TfH cell IL-4 wave at germinal
centers to the different processes involved in the development
of B cell immunity.
The positioning of NKT cells at the follicular borders during the
early stages of infection appears to be at the core of the NKT cell
ability to initiate B cell immunity. Single-cell RNA-seq and flow
cytometry analysis revealed that IL-4-producing NKT cells pref-
erentially express CXCR3, a chemokine receptor used by T cell
subsets to relocate to the interfollicular areas of the lymph
node (Groom et al., 2012; Sung et al., 2012). However, whether
NKT cells acquire CXCR3 after activation or whether a popula-
tion of NKT cells already expressing CXCR3 expands after infec-
tion remains to be elucidated. Interestingly, Th2 cells have also
been shown to locate and exert their functions at the periphery
of B cell follicles (Turqueti-Neves et al., 2014). Therefore, we
believe our study reveals a new layer of regulation of B cell immu-
nity at the initial stages of the infection process, when B cells
relocate to follicular borders and recruit T cell help.
The early production of IL-4 by NKT cells is mainly triggered by
CD169+ macrophages and is mediated by CD1d interactions
and IL-18 secretion. Interestingly, a great number of NKT-
macrophage contacts can be observed at the subcapsular sinus
and interfollicular areas as early as 2 days post-influenza infec-
tion. Our findings are in line with the notion that NKT cell activa-
tion after infection is the result of a combination of both CD1d
and cytokine signaling, with CD1d helping to position the cells
and cytokines inducing the effector functions of NKT cells
(Brennan et al., 2013). Furthermore, previous studies by others
and us have shown that these macrophages can present bacte-
rium-derived glycolipids on CD1d to mediate early NKT cell
expansion and that IL-18 can enhance NKT cell activity in
response to lipopolysaccharide (LPS), when co-administered
with a-GalCer or in autoimmune models (Barral et al., 2010; Häg-
glöf et al., 2016; Leite-De-Moraes et al., 2001; Nagarajan and
Kronenberg, 2007). Therefore, it seems that the induction of
diverse NKT cell effector functions by CD169+ macrophages
and IL-18 is a conserved mechanism observed upon infection
with various pathogens and in autoimmune disorders.
In this study, we have unveiled a new layer of regulation of
B cell immunity at the early stages of viral infection that involves
a tightly regulated NKT cell-IL-4-B cell axis. Interestingly, this
phenomenon seems to be conserved upon viral infection in other
species because our transcriptomic analysis of Zika-virus-in-
fected macaques revealed a positive correlation between early
NKT cell-IL-4 gene signatures and the serum levels of Zika
virus-neutralizing antibodies. We believe our findings raise the
prospect of harnessing the interplay between macrophages
and NKT cells and the early production of IL-4 at follicular
borders in future approaches for the design of new vaccination
strategies.
STAR+METHODS
Detailed methods are provided in the online version of this paper
and include the following:
d KEY RESOURCES TABLE
d CONTACT FOR REAGENT AND RESOURCE SHARING
d EXPERIMENTAL MODEL AND SUBJECT DETAILS
B Mice
B Infections and injections
B Flow Cytometry
B Immunohistochemistry
B qRT-PCR
B ELISPOT
B Extracellular flux assay
B scRNA-Seq Library Preparation
B Single-cell RNA-seq Data Analysis
B Transcriptomic analysis of samples from Zika infected
macaques
d QUANTIFICATION AND STATISTICAL ANALYSIS
d DATA AND SOFTWARE AVAILABILITY
SUPPLEMENTAL INFORMATION
Supplemental Information includes seven figures and one table and can be
found with this article online at https://doi.org/10.1016/j.cell.2017.11.036.
AUTHOR CONTRIBUTIONS
M.G. designed and performed experiments, generated CD1d-floxed mice,
analyzed the data, and wrote the manuscript. P.B. performed some initial
experiments, assist with the setup of in vivo infections, and contributed to sci-
entific discussions. A.W.N. and A.K.S. performed and analyzed single-cell
RNA-seq data. M.B., C.T., S.A., and P.M. performed experiments and contrib-
uted to scientific discussions. M.A. and D.H.B. analyzed transcriptomic data
from Zika-virus-infected macaques and contributed to scientific discussions.
K.G. and R.-P.S. contributed to scientific discussions regarding transcriptomic
analyses. P.G.F. provided IL-33R / bone marrow. B.M. contributed to the
generation of CD1d-floxed animals and provided technical support. A.B.
analyzed data and provided technical support for microscopy. U.N. edited
the manuscript. J.S. provided IL-4 / mice and contributed to scientific dis-
cussions. F.D.B. conceived the project, supervised the work, and wrote the
manuscript.
ACKNOWLEDGMENTS
We thank the flow cytometry facility for technical support and the biological
resource unit for the breeding of animals and assistance with the generation
of CD1d-floxed animals (at the Francis Crick and Ragon Institutes). We thank
Kathrin Kirsch for the maintenance of mouse colonies. We thank Neil Rogers
and Caetano Reis e Sousa for the provision of MyD88 / , TLR7 / , CD11c-
Cre, and Lyz2-Cre mice and for scientific discussions. We thank Michael
Reth for the Mb1-Cre strain. We thank Mark Wilson for the IL-4-GFP reporter
mice. We thank Andrew McKenzie for the IL-33R / bone marrow. We thank
Samuel W. Kazer and Sanjay M. Prakadan for assistance with RNA-seq
Cell 172, 517–533, January 25, 2018 531
analysis. We thank Adolfo Garcı́a-Sastre for the NS1-GFP virus. We thank Paul
Thomas for the PR8-OTII virus. We thank Daniel Lingwood for the HA trimer.
We thank the NIH Tetramer Core Facility for the provision of labeled CD1d tet-
ramers. We thank all members of the Lymphocyte Interaction Laboratory for
critical reading of the manuscript and scientific discussions. This work was
supported by the Francis Crick Institute, which receives its core funding
from Cancer Research UK (FC001035), the United Kingdom Medical Research
Council (FC001035), the Wellcome Trust (FC001035), the Center for HIV/AIDS
Vaccine Immunology and Immunogen Discovery of the NIH (UM1AI100663),
the Phillip T. and Susan M. Ragon Institute Foundation, and Bill and Melinda
Gates Foundation Innovation Award 228966 (to F.D.B.).
DECLARATION OF INTERESTS
The authors declare no competing interests.
Received: December 5, 2016
Revised: September 11, 2017
Accepted: November 20, 2017
Published: December 14, 2017
REFERENCES
Adachi, O., Kawai, T., Takeda, K., Matsumoto, M., Tsutsui, H., Sakagami, M.,
Nakanishi, K., and Akira, S. (1998). Targeted disruption of the MyD88 gene re-
sults in loss of IL-1- and IL-18-mediated function. Immunity 9, 143–150.
Aid, M., Abbink, P., Larocca, R.A., Boyd, M., Nityanandam, R., Nanayakkara,
O., Martinot, A.J., Moseley, E.T., Blass, E., Borducchi, E.N., et al. (2017). Zika
Virus Persistence in the Central Nervous System and Lymph Nodes of Rhesus
Monkeys. Cell 169, 610–620.e14.
Amigorena, S., Drake, J.R., Webster, P., and Mellman, I. (1994). Transient
accumulation of new class II MHC molecules in a novel endocytic compart-
ment in B lymphocytes. Nature 369, 113–120.
Artiaga, B.L., Yang, G., Hackmann, T.J., Liu, Q., Richt, J.A., Salek-Ardakani,
S., Castleman, W.L., Lednicky, J.A., and Driver, J.P. (2016). a-Galactosylcer-
amide protects swine against influenza infection when administered as a vac-
cine adjuvant. Sci. Rep. 6, 23593.
Bai, L., Deng, S., Reboulet, R., Mathew, R., Teyton, L., Savage, P.B., and
Bendelac, A. (2013). Natural killer T (NKT)-B-cell interactions promote pro-
longed antibody responses and long-term memory to pneumococcal capsular
polysaccharides. Proc. Natl. Acad. Sci. USA 110, 16097–16102.
Barral, P., Eckl-Dorna, J., Harwood, N.E., De Santo, C., Salio, M., Illarionov, P.,
Besra, G.S., Cerundolo, V., and Batista, F.D. (2008). B cell receptor-mediated
uptake of CD1d-restricted antigen augments antibody responses by recruiting
invariant NKT cell help in vivo. Proc. Natl. Acad. Sci. USA 105, 8345–8350.
Barral, P., Polzella, P., Bruckbauer, A., van Rooijen, N., Besra, G.S., Cerun-
dolo, V., and Batista, F.D. (2010). CD169(+) macrophages present lipid anti-
gens to mediate early activation of iNKT cells in lymph nodes. Nat. Immunol.
11, 303–312.
Batista, F.D., and Harwood, N.E. (2009). The who, how and where of antigen
presentation to B cells. Nat. Rev. Immunol. 9, 15–27.
Brennan, P.J., Brigl, M., and Brenner, M.B. (2013). Invariant natural killer
T cells: an innate activation scheme linked to diverse effector functions. Nat.
Rev. Immunol. 13, 101–117.
Carrasco, Y.R., and Batista, F.D. (2007). B cells acquire particulate antigen in a
macrophage-rich area at the boundary between the follicle and the subcapsu-
lar sinus of the lymph node. Immunity 27, 160–171.
Chang, P.-P., Barral, P., Fitch, J., Pratama, A., Ma, C.S., Kallies, A., Hogan,
J.J., Cerundolo, V., Tangye, S.G., Bittman, R., et al. (2011). Identification of
Bcl-6-dependent follicular helper NKT cells that provide cognate help for
B cell responses. Nat. Immunol. 13, 35–43.
De Santo, C., Salio, M., Masri, S.H., Lee, L.Y.-H., Dong, T., Speak, A.O., Por-
ubsky, S., Booth, S., Veerapen, N., Besra, G.S., et al. (2008). Invariant NKT
cells reduce the immunosuppressive activity of influenza A virus-induced
532 Cell 172, 517–533, January 25, 2018
myeloid-derived suppressor cells in mice and humans. J. Clin. Invest. 118,
4036–4048.
Dufort, F.J., Bleiman, B.F., Gumina, M.R., Blair, D., Wagner, D.J., Roberts,
M.F., Abu-Amer, Y., and Chiles, T.C. (2007). Cutting edge: IL-4-mediated pro-
tection of primary B lymphocytes from apoptosis via Stat6-dependent regula-
tion of glycolytic metabolism. J. Immunol. 179, 4953–4957.
Galli, G., Pittoni, P., Tonti, E., Malzone, C., Uematsu, Y., Tortoli, M., Maione, D.,
Volpini, G., Finco, O., Nuti, S., et al. (2007). Invariant NKT cells sustain specific
B cell responses and memory. Proc. Natl. Acad. Sci. USA 104, 3984–3989.
Gaya, M., Castello, A., Montaner, B., Rogers, N., Reis e Sousa, C., Bruckba-
uer, A., and Batista, F.D. (2015). Host response. Inflammation-induced disrup-
tion of SCS macrophages impairs B cell responses to secondary infection.
Science 347, 667–672.
Groom, J.R., Richmond, J., Murooka, T.T., Sorensen, E.W., Sung, J.H., Bank-
ert, K., von Andrian, U.H., Moon, J.J., Mempel, T.R., and Luster, A.D. (2012).
CXCR3 chemokine receptor-ligand interactions in the lymph node optimize
CD4+ T helper 1 cell differentiation. Immunity 37, 1091–1103.
Hägglöf, T., Sedimbi, S.K., Yates, J.L., Parsa, R., Salas, B.H., Harris, R.A.,
Leadbetter, E.A., and Karlsson, M.C.I. (2016). Neutrophils license iNKT cells
to regulate self-reactive mouse B cell responses. Nat. Immunol. 17,
1407–1414.
Heesters, B.A., Chatterjee, P., Kim, Y.-A., Gonzalez, S.F., Kuligowski, M.P.,
Kirchhausen, T., and Carroll, M.C. (2013). Endocytosis and recycling of
immune complexes by follicular dendritic cells enhances B cell antigen binding
and activation. Immunity 38, 1164–1175.
Huang, S.C.-C., Smith, A.M., Everts, B., Colonna, M., Pearce, E.L., Schilling,
J.D., and Pearce, E.J. (2016). Metabolic Reprogramming Mediated by the
mTORC2-IRF4 Signaling Axis Is Essential for Macrophage Alternative Activa-
tion. Immunity 45, 817–830.
Ishikawa, H., Tanaka, K., Kutsukake, E., Fukui, T., Sasaki, H., Hata, A., Noda,
S., and Matsumoto, T. (2010). IFN-g production downstream of NKT cell acti-
vation in mice infected with influenza virus enhances the cytolytic activities of
both NK cells and viral antigen-specific CD8+ T cells. Virology 407, 325–332.
Junt, T., Moseman, E.A., Iannacone, M., Massberg, S., Lang, P.A., Boes, M.,
Fink, K., Henrickson, S.E., Shayakhmetov, D.M., Di Paolo, N.C., et al. (2007).
Subcapsular sinus macrophages in lymph nodes clear lymph-borne viruses
and present them to antiviral B cells. Nature 450, 110–114.
Kastenmüller, W., Torabi-Parizi, P., Subramanian, N., Lämmermann, T., and
Germain, R.N. (2012). A spatially-organized multicellular innate immune
response in lymph nodes limits systemic pathogen spread. Cell 150,
1235–1248.
King, I.L., and Mohrs, M. (2009). IL-4-producing CD4+ T cells in reactive lymph
nodes during helminth infection are T follicular helper cells. J. Exp. Med. 206,
1001–1007.
King, I.L., Fortier, A., Tighe, M., Dibble, J., Watts, G.F.M., Veerapen, N., Haber-
man, A.M., Besra, G.S., Mohrs, M., Brenner, M.B., and Leadbetter, E.A. (2011).
Invariant natural killer T cells direct B cell responses to cognate lipid antigen in
an IL-21-dependent manner. Nat. Immunol. 13, 44–50.
Kinjo, Y., Illarionov, P., Vela, J.L., Pei, B., Girardi, E., Li, X., Li, Y., Imamura, M.,
Kaneko, Y., Okawara, A., et al. (2011). Invariant natural killer T cells recognize
glycolipids from pathogenic Gram-positive bacteria. Nat. Immunol. 12,
966–974.
Leadbetter, E.A., Brigl, M., Illarionov, P., Cohen, N., Luteran, M.C., Pillai, S.,
Besra, G.S., and Brenner, M.B. (2008). NK T cells provide lipid antigen-specific
cognate help for B cells. Proc. Natl. Acad. Sci. USA 105, 8339–8344.
Lee, Y.J., Holzapfel, K.L., Zhu, J., Jameson, S.C., and Hogquist, K.A. (2013).
Steady-state production of IL-4 modulates immunity in mouse strains and is
determined by lineage diversity of iNKT cells. Nat. Immunol. 14, 1146–1154.
Lee, Y.J., Wang, H., Starrett, G.J., Phuong, V., Jameson, S.C., and Hogquist,
K.A. (2015). Tissue-Specific Distribution of iNKT Cells Impacts Their Cytokine
Response. Immunity 43, 566–578.
Leite-De-Moraes, M.C., Hameg, A., Pacilio, M., Koezuka, Y., Taniguchi, M.,
Van Kaer, L., Schneider, E., Dy, M., and Herbelin, A. (2001). IL-18 enhances
IL-4 production by ligand-activated NKT lymphocytes: a pro-Th2 effect of IL-
18 exerted through NKT cells. J. Immunol. 166, 945–951.
Manicassamy, B., Manicassamy, S., Belicha-Villanueva, A., Pisanelli, G., Pu-
lendran, B., and Garcı́a-Sastre, A. (2010). Analysis of in vivo dynamics of influ-
enza virus infection in mice using a GFP reporter virus. Proc. Natl. Acad. Sci.
USA 107, 11531–11536.
Mohrs, M., Shinkai, K., Mohrs, K., and Locksley, R.M. (2001). Analysis of type 2
immunity in vivo with a bicistronic IL-4 reporter. Immunity 15, 303–311.
Mohrs, K., Wakil, A.E., Killeen, N., Locksley, R.M., and Mohrs, M. (2005).
A two-step process for cytokine production revealed by IL-4 dual-reporter
mice. Immunity 23, 419–429.
Nagarajan, N.A., and Kronenberg, M. (2007). Invariant NKT cells amplify the
innate immune response to lipopolysaccharide. J. Immunol. 178, 2706–2713.
Phan, T.G., Grigorova, I., Okada, T., and Cyster, J.G. (2007). Subcapsular
encounter and complement-dependent transport of immune complexes by
lymph node B cells. Nat. Immunol. 8, 992–1000.
Qi, H., Egen, J.G., Huang, A.Y.C., and Germain, R.N. (2006). Extrafollicular
activation of lymph node B cells by antigen-bearing dendritic cells. Science
312, 1672–1676.
Reinhardt, R.L., Liang, H.-E., and Locksley, R.M. (2009). Cytokine-secreting
follicular T cells shape the antibody repertoire. Nat. Immunol. 10, 385–393.
Ricardo-Gonzalez, R.R., Red Eagle, A., Odegaard, J.I., Jouihan, H., Morel,
C.R., Heredia, J.E., Mukundan, L., Wu, D., Locksley, R.M., and Chawla, A.
(2010). IL-4/STAT6 immune axis regulates peripheral nutrient metabolism
and insulin sensitivity. Proc. Natl. Acad. Sci. USA 107, 22617–22622.
Sagoo, P., Garcia, Z., Breart, B., Lemaı̂tre, F., Michonneau, D., Albert, M.L.,
Levy, Y., and Bousso, P. (2016). In vivo imaging of inflammasome activation
reveals a subcapsular macrophage burst response that mobilizes innate and
adaptive immunity. Nat. Med. 22, 64–71.
Satija, R., Farrell, J.A., Gennert, D., Schier, A.F., and Regev, A. (2015). Spatial
reconstruction of single-cell gene expression data. Nat. Biotechnol. 33,
495–502.
Schmitz, J., Owyang, A., Oldham, E., Song, Y., Murphy, E., McClanahan, T.K.,
Zurawski, G., Moshrefi, M., Qin, J., Li, X., et al. (2005). IL-33, an interleukin-1-
like cytokine that signals via the IL-1 receptor-related protein ST2 and induces
T helper type 2-associated cytokines. Immunity 23, 479–490.
Sung, J.H., Zhang, H., Moseman, E.A., Alvarez, D., Iannacone, M., Henrickson,
S.E., de la Torre, J.C., Groom, J.R., Luster, A.D., and von Andrian, U.H. (2012).
Chemokine guidance of central memory T cells is critical for antiviral recall
responses in lymph nodes. Cell 150, 1249–1263.
Thomas, P.G., Brown, S.A., Yue, W., So, J., Webby, R.J., and Doherty, P.C.
(2006). An unexpected antibody response to an engineered influenza virus
modifies CD8+ T cell responses. Proc. Natl. Acad. Sci. USA 103, 2764–2769.
Tirosh, I., Izar, B., Prakadan, S.M., Wadsworth, M.H., 2nd, Treacy, D., Trom-
betta, J.J., Rotem, A., Rodman, C., Lian, C., Murphy, G., et al. (2016). Dissect-
ing the multicellular ecosystem of metastatic melanoma by single-cell
RNA-seq. Science 352, 189–196.
Townsend, M.J., Fallon, P.G., Matthews, D.J., Jolin, H.E., and McKenzie, A.N.
(2000). T1/ST2-deficient mice demonstrate the importance of T1/ST2 in devel-
oping primary T helper cell type 2 responses. J. Exp. Med. 191, 1069–1076.
Trombetta, J.J., Gennert, D., Lu, D., Satija, R., Shalek, A.K., and Regev, A.
(2014). Preparation of Single-Cell RNA-Seq Libraries for Next Generation
Sequencing. Curr. Protoc. Mol. Biol. 107, 4.22.1–4.22.17.
Turqueti-Neves, A., Otte, M., Prazeres da Costa, O., Höpken, U.E., Lipp, M.,
Buch, T., and Voehringer, D. (2014). B-cell-intrinsic STAT6 signaling controls
germinal center formation. Eur. J. Immunol. 44, 2130–2138.
Victora, G.D., and Nussenzweig, M.C. (2012). Germinal centers. Annu. Rev.
Immunol. 30, 429–457.
Weinstein, J.S., Herman, E.I., Lainez, B., Licona-Limón, P., Esplugues, E.,
Flavell, R., and Craft, J. (2016). TFH cells progressively differentiate to regulate
the germinal center response. Nat. Immunol. 17, 1197–1205.
Yusuf, I., Kageyama, R., Monticelli, L., Johnston, R.J., Ditoro, D., Hansen, K.,
Barnett, B., and Crotty, S. (2010). Germinal center T follicular helper cell IL-4
production is dependent on signaling lymphocytic activation molecule recep-
tor (CD150). J. Immunol. 185, 190–202.
Cell 172, 517–533, January 25, 2018 533
STAR+METHODS

KEY RESOURCES TABLE

REAGENT or RESOURCE SOURCE IDENTIFIER

Antibodies
PE-Cy7 Rat Anti-Mouse CD19, Clone 1D3 ThermoFisher Cat# 25-0193-82, RRID:AB_657663
Scientific
Alexa Fluor 488 Mouse Anti-Mouse CD95 (APO-1/ Fas), ThermoFisher Cat# 53-0951-82, RRID:AB_10671269
Clone 15A7 Scientific
Alexa Fluor 647 Rat Anti-Mouse/Human GL-7 Biolegend Cat# 144606, RRID:AB_2562185
(T and B cell activation Marker), Clone GL7
PE Rat Anti-Mouse CD4, Clone GK1.5 Biolegend Cat# 100408, RRID:AB_312693
APC Rat Anti-Mouse CD4, Clone GK1.5 Biolegend Cat# 100412, RRID:AB_312697
Biotin Rat Anti-Mouse CD185 (CXCR5), Clone SPRCL5 ThermoFisher Cat# 13-7185-82, RRID:AB_2572800
Scientific
PerCP eFluor 710 Rat Anti-Mouse CD185 (CXCR5), ThermoFisher Cat# 46-7185-82, RRID:AB_2573837
Clone SPRCL5 Scientific
FITC Hamster Anti-Mouse CD279 (PD-1), Clone J43 ThermoFisher Cat# 11-9985-82, RRID:AB_465472
Scientific
Pacific Blue Rat Anti-Mouse/Human CD45R (B220), Biolegend Cat# 103227, RRID:AB_492876
Clone RA3-6B2
Brilliant Violet 650 Rat Anti-Mouse CD45R (B220), Biolegend Cat# 103241, RRID:AB_11204069
Clone RA3-6B2
APC Rat Anti-Mouse MHC Class II (I-A/I-E), Clone ThermoFisher Cat# 17-5321-82, RRID:AB_469455
M5/114.15.2 Scientific
PerCP/Cy5.5 Armenian Hamster Anti-Mouse CD11c, Biolegend Cat# 117328, RRID:AB_2129641
Clone N418
eFluor 450 Armenian Hamster Anti-Mouse CD11c, ThermoFisher Cat# 48-0114-82, RRID:AB_1548654
Clone N418 Scientific
APC Rat Anti-Mouse CD11b, Clone M1/70 BD Biosciences Cat# 553312, RRID:AB_398535
eFluor 450 Rat Anti-Mouse CD11b, Clone M1/70 ThermoFisher Cat# 48-0112-82, RRID:AB_1582236
Scientific
eFluor 450 Rat Anti-Mouse CD3, Clone 17A2 ThermoFisher Cat# 48-0032-82, RRID:AB_1272193
Scientific
eFluor 450 Rat Anti-Mouse CD5, Clone 53-7.3 ThermoFisher Cat# 48-0051-82, RRID:AB_1603250
Scientific
PE Rat Anti-Mouse CD127 (IL-7Ra), Clone A7R34 Biolegend Cat# 135010, RRID:AB_1937251
FITC Armenian Hamster Anti-Mouse TCRb, Clone ThermoFisher Cat# 11-5961-82 RRID:AB_465323
H57-597 Scientific
APC Armenian Hamster Anti-Mouse TCRb, Clone ThermoFisher Cat# 17-5961-82 RRID:AB_469481
H57-597 Scientific
PE PBS57-loaded Mouse CD1d tetramer NIH Tetramer N/A
Core Facility
PE-Cy7 Armenian Hamster Anti-Mouse CD69, ThermoFisher Cat# 25-0691-82, RRID:AB_469637
Clone H1.2F3 Scientific
PE Rat anti-Mouse CD1d, Clone 1B1 BD Biosciences Cat# 553846, RRID:AB_2073521
Pacific Blue Rat Anti-Mouse Ly6G-Ly6C (Gr-1), Biolegend Cat# 108430, RRID:AB_893556
Clone RB6-8C5
V450 Rat Anti-Mouse IgG1, Clone A85-1 BD Biosciences Cat# 562107, RRID:AB_10894002
PerCP/Cy5.5 Mouse anti-Mouse CD45.1, Clone A20 ThermoFisher Cat# 45-0453-82, RRID:AB_1107003
Scientific
Pacific Blue Mouse anti-Mouse CD45.2, Clone 104 Biolegend Cat# 109820, RRID:AB_492872
(Continued on next page)

e1 Cell 172, 517–533.e1–e7, January 25, 2018

Continued
REAGENT or RESOURCE SOURCE IDENTIFIER
PerCP/Cy5.5 Rat anti-Mouse CD45, Clone 30-F11 Biolegend Cat# 103132, RRID:AB_893340
PE-Cy7 Armenian Hamster Anti-Mouse CD183 Biolegend Cat# 126516, RRID:AB_2245493
(CXCR3), Clone CXCR3-173
APC Armenian Hamster Anti-Mouse CD183 Biolegend Cat# 126512, RRID:AB_1088993
(CXCR3), Clone CXCR3-173
Alexa Fluor 488 Rat Anti-Mouse CD150 (SLAM), Biolegend Cat# 115916, RRID:AB_528744
Clone TC15-12F12.2
APC Rat Anti-Mouse CD8a, Clone 53-6.7 Biolegend Cat# 100712, RRID:AB_312751
Purified Rat Anti-Mouse CD16/32, Clone 93 Biolegend Cat# 101302, RRID:AB_312801
Alexa Fluor 647 Mouse Anti-Mouse GATA-3, Biolegend Cat# 653810, RRID:AB_2563217
Clone 16E10A23
APC Rat Anti-Mouse RORgt, Clone B2D ThermoFisher Cat# 17-6981-82, RRID:AB_2573254
Scientific
PE-Cy7 Mouse Anti-Mouse T-bet, Clone 4B10 ThermoFisher Cat# 25-5825-82, RRID:AB_11042699
Scientific
Alexa Fluor 647 Mouse Anti-Mouse Bcl-6, BD Biosciences Cat# 561525, RRID:AB_10898007
Clone K112-91
PE-Cy7 Rat Anti-Mouse IFN-g, Clone XMG1.2 ThermoFisher Cat# 25-7311-41, RRID:AB_1257211
Scientific
APC Rat Anti-Mouse IL-4, Clone 11B11 ThermoFisher Cat# 17-7041-82, RRID:AB_469494
Scientific
InVivoMab Rat Anti-Mouse IL-4, Clone 11B11 BioXCell Cat# BE0045, RRID:AB_1107707
Alexa Fluor 488 Rat Anti-Mouse CD169 (Siglec-1), Produced in the lab N/A
Clone SER-4
FITC Rat Anti-Mouse IgD, Clone 11-26c.2a BD Biosciences Cat# 553439, RRID:AB_394859
Alexa Fluor 647 Rat Anti-Mouse CD207 (Langerin), Dendritics Cat# DDX0362, RRID:AB_1148742
Clone 929F3.01
Purified Goat Anti-Mouse IL-1b, Polyclonal R&D Systems Cat# AF-401-NA, RRID:AB_416684
Purified Rabbit Anti-Mouse IL-18, Polyclonal Abcam Cat# ab71495, RRID:AB_1209302
Purified Goat Anti-Mouse IL-33, Polyclonal R&D Systems Cat# AF3626, RRID:AB_884269
Alexa Fluor 555 Goat anti-Rabbit IgG (H + L) ThermoFisher Cat# A32732 RRID:AB_2633281
Scientific
Alexa Fluor 555 Donkey anti-Goat IgG (H + L) ThermoFisher Cat# A-21432, RRID:AB_141788
Scientific
Purified Goat anti-R-Phycoerythrin, Polyclonal Novus Biologicals Cat# NB100-78573, RRID:AB_1085348
Biotin Goat Anti-Mouse IgM, Human adsorbed Southern Biotech Cat# 102008, RRID:AB_616726
Biotin Goat Anti-Mouse IgG1, Human adsorbed Southern Biotech Cat# 107008, RRID:AB_609714
Purified Armenian Hamster anti-Mouse CD3ε, Biolegend Cat# 100302, RRID:AB_312667
Clone 145-2C11
Bacterial and Virus Strains
Influenza virus A/Puerto Rico/8/1934 (PR8) H1N1 Caetano Reis e N/A
Sousa’s lab
PR8-OTII Thomas et al., 2006 N/A
Influenza NS1-GFP Manicassamy N/A
et al., 2010
Vaccinia virus Western Reserve Caetano Reis e N/A
Sousa’s lab
Chemicals, Peptides, and Recombinant Proteins
PE Streptavidin Biolegend Cat# 405204
eFluor 450 Streptavidin ThermoFisher Cat# 48-4317
Scientific
(Continued on next page)

Cell 172, 517–533.e1–e7, January 25, 2018 e2

Continued
REAGENT or RESOURCE SOURCE IDENTIFIER
PR8 Hemagglutinin trimer From Daniel N/A
Lingwood
Poly I:C HMW Invivogen Cat# 31852-29-6
Alpha Gal-Cer KRN7000 Enzo Life Sciences Cat# BML-SL232
Diphtheria Toxin from Corynebacterium diphtheriae Sigma Cat# D0564
Recombinant Murine Interleukin 4 Prepotech Cat# 214-14
LIVE/DEAD Fixable Blue ThermoFisher Cat# L23105
Scientific
Cell Stimulation Cocktail plus protein transport inhibitor ThermoFisher Cat# 00-4975-03
Scientific
Influenza A H1N1 (A/Puerto Rico/8/34) Hemagglutinin Sino Biological Cat# 11684-V08H
Streptavidin-Alkaline Phosphatase Sigma Cat# S2890
Recombinant Mouse Interleukin 18 R&D Systems Cat# 9139-IL
AffiniPure F(ab’)2 Fragment Goat Anti-Mouse IgM, m Jackson Cat# 115-006-020
Chain Specific ImmunoResearch
3-Amino-9-ethylcarbazole Sigma Cat# A5754
Critical Commercial Assays
Foxp3 Transcription Factor Staining Buffer Set ThermoFisher Cat# 00-5523-00
Scientific
Cytofix/Cytoperm BD Biosciences Cat# 554722
RNeasy Micro Kit QIAGEN Cat# 74004
SuperScript III first-strand system ThermoFisher Cat# 18080051
Scientific
SYBR green PCR master mix ThermoFisher Cat# 4309155
Scientific
Mouse IL-4 ELISPOT kit BD Biosciences Cat# 551017
Pan B cell Isolation Kit, mouse Miltenyi Cat# 130-095-813
Seahorse XF Cell Mito Stress Test Kit Agilent Cat# 103015-100
Maxima Reverse Transcriptase ThermoFisher Cat# EP0741
Scientific
Deposited Data
Single-Cell RNA-Seq data This study GEO: GSE103753
Experimental Models: Cell Lines
Dog (female), MDCK cells Caetano Reis e N/A
Sousa’s lab
Chicken, DF-1 cells Caetano Reis e N/A
Sousa’s lab
Experimental Models: Organisms/Strains
Mouse: Wild type (C57BL/6J) The Jackson Strain Code: JAX 000664
Laboratory
Mouse: Wild type (FVB/NCrl) Charles Rivers Strain Code: 207
Laboratory
Mouse: B6.SJL-Ptprca Pepcb/BoyJ (CD45.1) The Jackson Strain Code: JAX 002014
Laboratory
Mouse: B6;129-Siglec1 < tm1(HBEGF)Mtka > (CD169-DTR) Riken BioResource Strain Code: RBRC04395
Center
Mouse, C.Cg-Cd19tm1(cre)Cgn Ighb/J (CD19-Cre) The Jackson Strain Code: JAX 004126
Laboratory
Mouse, B6.129S6-Del(3Cd1d2-Cd1d1)1Sbp/J The Jackson Strain Code: JAX 008881
(CD1d KO) Laboratory
(Continued on next page)

e3 Cell 172, 517–533.e1–e7, January 25, 2018

Continued
REAGENT or RESOURCE SOURCE IDENTIFIER
tm1Imx
Mouse, B6.129S7-Il1r1 /J (IL-1R KO) The Jackson Strain Code: JAX 003245
Laboratory
Mouse, B6.129P2-Il18r1tm1Aki/J (IL-18R KO) The Jackson Strain Code: JAX 004131
Laboratory
Mouse, B6.129P2-Lyz2tm1(cre)Ifo/J (Lyz2-Cre) The Jackson Strain Code: JAX 004781
Laboratory
Mouse, B6.Cg-Tg(Itgax-cre)1-1Reiz/J (CD11c-Cre) The Jackson Strain Code: JAX 008068
Laboratory
Mouse, B6.129S1-Tlr7tm1Flv/J (TLR7 KO) The Jackson Strain Code: JAX 008380
Laboratory
Mouse, C57BL/6-Tg(TcraTcrb)425Cbn/Crl (OT-II) The Jackson Strain Code: JAX 004194
Laboratory
Mouse, B6.129P2(SJL)-Myd88tm1.1Defr/J (MyD88 KO) The Jackson Strain Code: JAX 009088
Laboratory
Mouse, C.129-Il4tm1Lky/J (IL-4/GFP-enhanced transcript, 4Get). The Jackson Strain Code: JAX 004190
Backcrossed to B6 background for 10 generations. Laboratory
Mouse, B6.C(Cg)-Cd79atm1(cre)Reth/EhobJ (Mb1-Cre) The Jackson Strain Code: JAX 020505
Laboratory
Mouse, B6.129P2-Il4tm1Cgn/J (IL-4 KO). Backcrossed The Jackson Strain Code: JAX 002253
to FVB/N background for 10 generations. Laboratory
Mouse, Il1rl1tm1Anjm (IL-33R KO) Townsend MGI: 2386675
et al., 2000
Mouse, CD1d flox/flox CD19-Cre This study N/A
Mouse, CD1d flox/flox Mb1-Cre This study N/A
Mouse, CD1d flox/flox Lyz2-Cre This study N/A
Mouse, CD1d flox/flox CD11c-Cre This study N/A
Mouse, B6.Cg-Tg(Pgk1-flpo)10Sykr/J (FLPo) The Jackson Strain Code: JAX 011065
Laboratory
Oligonucleotides
Primer: HPRT Forward GCCCTTGACTATAATGAGTACTTCAGG This study N/A
Primer: HPRT Reverse TTCAACTTGCGCTCATCTTAGG This study N/A
Primer: IL-4 Forward ACGAGGTCACAGGAGAAGGGA This study N/A
Primer: IL-4 Reverse AGCCCTACAGACGAGCTCACTC This study N/A
Primer: IL-5 Forward TGACAAGCAATGAGACGATGAGG This study N/A
Primer: IL-5 Reverse ACCCCCACGGACAGTTTGATTC This study N/A
Primer: IL-13 Forward CCTCTGACCCTTAAGGAGCTTAT This study N/A
Primer: IL-13 Reverse CGTTGCACAGGGGAGTCTT This study N/A
Primer: IL-21 Forward TCAGCTCCACAAGATGTAAAGGG This study N/A
Primer: IL-21 Reverse GGGCCACGAGGTCAATGAT This study N/A
Software and Algorithms
Diva BD Biosciences http://www.bdbiosciences.com/us/
instruments/clinical/software/flow-
cytometry-acquisition/bd-facsdiva-
software/m/333333/overview
FlowJo (version 10.1) FlowJo, LLC https://www.flowjo.com
Zen Zeizz https://www.zeiss.com/microscopy/
int/downloads/zen.html
ImageJ (version 1.50c) NIH Image https://imagej.nih.gov/ij/
Imaris (version 8.2.0) Bitplane http://www.bitplane.com/
GraphPad Prism (version 7.0) GraphPad https://www.graphpad.com/
Software Inc scientific-software/prism/
(Continued on next page)

Cell 172, 517–533.e1–e7, January 25, 2018 e4

Continued
REAGENT or RESOURCE SOURCE IDENTIFIER
ELISPOT reader with immunospot software Cellular Technology Ltd N/A
InDesign CS5.5 Adobe http://www.adobe.com/products/
indesign.html
R (Version 3.3.2), Seurat package (version 1.4.0.16) R Core Team https://www.r-project.org/
GSEA Broad Institute http://software.broadinstitute.org/
gsea/index.jsp
Gene functional annotation and network inference http://cytoscape.org/ http://genemania.org/

CONTACT FOR REAGENT AND RESOURCE SHARING

Further information and requests for resources and reagents should be directed to, and will be fulfilled by, the Lead Contact, Facundo
D. Batista (FBATISTA1@mgh.harvard.edu).

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Mice
8-10 week old wild-type C57BL/6 and FVB/N mice were obtained from Jackson and Charles River Laboratories. Congenic CD45.1
mice were from the Francis Crick Institute, UK. Siglec1DTR/+ (CD169-DTR) mice were obtained from the Riken BioResource Center,
Japan. CD19-Cre, CD1d / , IL-1R / and IL-18R / mice were obtained from Jackson Laboratories, USA. Lyz2-Cre, Cd11c-Cre,
TLR7 / , OT-II and Myd88 / mice were obtained from Dr. Caetano Reis e Sousa, Francis Crick Institute, UK. IL-4/GFP-enhanced
transcript (4Get) mice were obtained from Dr. Mark Wilson, Francis Crick Institute, UK. FVB/N IL-4 / mice were obtained from Dr.
Jessica Strid, Imperial College London, UK. Mb1-Cre mice were obtained from Dr. Michael Reth, Max Planck Institute, Germany.
IL-33R / bone marrows (Townsend et al., 2000) were obtained from Andrew McKenzie, MRC laboratory of Molecular Biology,
UK and Padraic Fallon, Trinity College Dublin, Ireland.
For the generation of CD1d conditional mice, we used homologous recombination to insert loxP sites flanking exon 3 of cd1d1
gene in B6 embryonic stem (ES) cells. Identification of ES cell clones with a single correct insertion of the targeted locus in the genome
was performed through long-range PCR and Southern blot. Transgenic mice derived from these ES cells were crossed with B6 mice
expressing the FLP recombinase to remove the frt-flanked PGK-neomycin sequence used for selection of ES cells that have inserted
the targeted vector into their genome. Mice carrying the flox allele were crossed with mice expressing the Cre recombinase under the
promoters of CD19, Mb1, Lyz2 and CD11c genes. Finally, mice were crossed again with the same floxed mice to make them homo-
zygous for the floxed allele.
For the generation of mixed bone marrow chimeras, CD45.1 C57BL/6 or CD169-DTR of 6-8 weeks of age were irradiated with 2
doses of 500 rad, 4 hours apart. One day later, bone marrow cells were injected i.v. in recipient animals (2x106 cells in total). Exper-
imental animals were kept on acidified water for one week prior and 3 weeks post irradiation treatment. Chimeras were used after
8 weeks of reconstitution.
Mice were bred and maintained at the animal facilities of the Francis Crick and Ragon Institutes. All mice were maintained in indi-
vidually ventilated cages and were used at the age of 8 to 12 weeks. Up to five mice per cage were housed in individually ventilated
cages in a 12 hr light/dark cycle, with room temperature at 22 C (19 C-23 C change). They were fed with autoclaved standard pellet
chow and reverse osmosis water. All cages contained 5 mm of aspen chip and tissue wiles for bedding and a mouse house for envi-
ronmental enrichment. Littermates (males or females) were randomly assigned to experimental groups. Generally between 4 to 8
mice were used per experimental group. All experiments were approved by the Animal Ethics Committee of Cancer Research UK
and the United Kingdom Home Office, and by the Institutional Animal Care and Use Committee of the United States.

Infections and injections

Mice were anesthetized i.p. with Ketamine/Xylazine (100 mg/kg body) and intranasally infected with 200 PFU of Influenza virus
A/Puerto Rico/8/1934 (PR8) H1N1 strain or 100 PFU of PR8-OTII virus (Thomas et al., 2006) or 103 PFU of Influenza NS1-GFP virus
(Manicassamy et al., 2010) or 200 PFU of Vaccinia virus Western Reserve strain in 20 ml of PBS. Influenza virus was amplified on
MDCK cells and VACV on DF-1 cells. Purification of viral particles was performed in a sucrose 30% cushion at 25, 000 RPM for 2 hours
in an SW32Ti rotor. For protein immunizations, 5 mg of PR8 Hemagglutinin (HA) trimmer, kindly provided by Daniel Lingwood, was
intranasally administered together with 10 mg of Poly I:C (Invivogen) or 2 mg of a-GalCer (Enzo) in 20 ml of PBS. For depletion of
CD169+ macrophages, mice were i.p. injected with 500 ng of Diphtheria toxin (Sigma) in 200 ml of PBS. For in vivo IL-4 blocking,
mice were i.p. injected with 1.5 mg of anti-IL4 (11B11, BioXCell) in 200 ml of PBS at different days of infection. For the administration
of IL-4 complex (IL-4c), 5 mg of recombinant mouse IL-4 (PrepoTech) was complexed to 25 mg of anti-mouse IL-4 (BioXCell, 11B11),

e5 Cell 172, 517–533.e1–e7, January 25, 2018

diluted in 200 ml of PBS and administered i.p. on days 0, 1 and 2 of infection. For adoptive transfers, sorted NKT cells (8.105), OT-II
CD4+ isolated CD4+ T cells (2.104) or bone marrow cells (2.106) were resuspended in 200 ml of PBS and injected in the tail vein.

Flow Cytometry
For labeling of surface markers, single cell suspensions from lymph nodes or lungs were incubated with anti-mouse CD16/32
(BD Biosciences) and LIVE/DEAD Fixable blue (Life Technologies) in PBS 2% FCS for 20 minutes on ice to block nonspecific antibody
binding and exclude dead cells. Cells were then washed and stained for 20 minutes on ice with the indicated anti-mouse antibodies,
PBS57-loaded CD1d tetramer (kindly provided by NIH tetramer core facility) and/or biotinylated HA (Sino biological). When using
biotinylated antibodies or proteins, cells suspensions were washed following antibody labeling and incubated for 20 minutes on
ice with labeled streptavidin. If labeling of transcription factors was performed, cells were fixed and permeabilized with Foxp3 tran-
scription factor staining buffer set (eBioscience) according to manufacturer instructions and incubated for 30 min with anti-mouse
antibodies. Finally, cells were either resuspended in 400ml of FACS buffer and analyzed on a Fortessa cytometer (BD Biosciences)
or in RPMI media and used for cell sorting on a FACSAria II (BD, Bioscience).
For analysis of cytokine production, lymph node single cell suspensions were incubated 4 hours at 37 C in RPMI complete medium
supplemented with 1X Cell stimulation cocktail plus protein transport inhibitor (eBioscience). After surface staining, cells were fixed
and permeabilized with Cytofix/Cytoperm (BD Biosciences) according to manufacturer instructions and incubated for 30 min with
anti-mouse antibodies. Finally, cells were resuspended in 400ml of FACS buffer and data were collected on a Fortessa cytometer
(BD Biosciences) and analyzed with FlowJo (TreeStar).

Immunohistochemistry
For surface staining, cryostat sections (10mm thick) of lymph nodes were dried, fixed in 4% PFA for 10 minutes, blocked with PBS 5%
BSA and then incubated with the following anti-mouse antibodies in 1% BSA for 1 hour: B220 Pacific Blue (RA3-6 B2, Biolegend),
CD169 AF488 (Ser-4, ATCC), IgD FITC (11-26c.2a, BD Biosciences), Langerin AF647 (929F3.01, Dendritics) and GL-7 AF647 (GL7,
Biolegend). For intracellular staining, sections were permeabilized with PBS Triton 0.3% for 5 minutes before blocking and then incu-
bated with anti-mouse IL-1b (polyclonal, R&D), IL-18 (polyclonal, Abcam) or IL-33 (polyclonal, R&D), followed by incubation with
AF555 anti-goat or anti-rabbit IgG (Invitrogen). CD1d tetramer staining was performed as previously described (Lee et al., 2015).
Briefly, lymph nodes were incubated overnight with PE-labeled PBS-57 loaded CD1d tetramer in 2% FCS at 4 C. Next day, lymph
nodes were washed and fixed with 4% PFA for 1 hour and snap frozen. 10mm sections were blocked with 5% BSA and stained with
anti-PE antibody (polyclonal, Novus Biologicals) followed by goat anti-rabbit AF555 (Life Technology). Imaging was carried out on a
LSM 780 (Zeiss) inverted confocal microscope using a Plan-Apochromat 40x NA 1.3 oil immersion objective.

qRT-PCR
NKT cells, TfH cells and conventional CD4+ T cells from mediastinal lymph nodes were sorted at day 3 and 9 of infection using the
gating strategy shown in Figure S2A. RNA was extracted from sorted cells using the RNeasy micro Kit (QIAGEN). Between 103
and 104 cells were used in each condition. mRNA was reversely transcribed into cDNA using random hexamers from the SuperScript
III first-strand system (Thermofisher). cDNA was amplified using custom primer sets (Sigma) and SYBR green PCR master mix
(Thermofisher) on a 7500 real-time PCR system (Applied Biosystems). The relative mRNA abundance of target genes was determined
by means of the DDCT method, using HPRT as house keeping gene.

ELISPOT
To measure influenza-specific ASCs, Enzyme-linked immunosorbent spot (ELISPOT) multiscreen filtration plates (Millipore) were
activated with absolute ethanol, washed with PBS and coated overnight at 4 C with 1 mg/ml of PR8 virus or 2 mg/ml of PR8 HA
(Sino Biological) diluted in PBS. Plates were subsequently blocked for 1 hour with complete medium and incubated for 24 hours
at 37 C with serial dilutions of lymph node single cell suspensions. Plates were washed with PBS 0.01% Tween-20 and incubated
for 1 hour with 1mg/ml of biotinylated anti-IgM or IgG1 (Southern Biotech) diluted in PBS 1% BSA. Then, plates were washed and
incubated for 1 hour with 1mg/ml Streptavidin-Alkaline Phosphatase (Sigma). Finally plates were washed and developed with
3-amino-9-ethyl-carbazole (Sigma).
To measure IL-4 production after influenza infection, NKT cells were sorted from mediastinal lymph nodes (TCRb+ CD1d-tetramer+
CXCR3+/ ) and incubated in RPMI complete medium supplemented with 1X Cell stimulation cocktail (eBioscience) for 32 hours at a
density of 2.104 cells/well. To measure IL-4 production in the presence of IL-18, NKT cells were sorted from spleen (TCRb+ CD1d-
tetramer+) and incubated with increasing concentrations of mouse recombinant IL-18 (R&D systems) for 32 hours at a density of 2.104
cells/well. IL-4 secreting cells were detected using the mouse IL-4 ELISPOT kit from BD Biosciences following manufacturer’s
instructions.

Extracellular flux assay

Splenic naive B cells were purified using Pan B cell Isolation Kit (Miltenyi) and stimulated with 10 ng/ml IL-4 (PrepoTech) and/or
5 mg/ml F(ab’)2 anti-IgM (Jackson ImmunoResearch) for 16 hours. Lymph node cells were obtained after 3 days of influenza infection,
cultured with 5 mg/ml F(ab’)2 anti-IgM and 3 mg/ml anti-CD3ε (Biolegend, 145-2C11) and B cells were purified after 16 hours. B cells

Cell 172, 517–533.e1–e7, January 25, 2018 e6

were resuspended in Seahorse medium supplemented with 11mM glucose (Sigma Aldrich), 2 mM glutamax (Life Technologies) and
1 mM Sodium pyruvate (Life Technologies). Cells were settled on 96-well assay plate (Seahorse Bioscience) coated with poly-lysine
(Sigma Aldrich) for 30 minutes. Data was recorded with the XF96 Extracellular Flux analyzer. Oligomycin A (3 mM), FCCP (5 mM) and
Rotenone (5 mM) (all from Agilent) were added sequentially at indicated time points to generate the OCR profile.

scRNA-Seq Library Preparation

Plate-based scRNA-Seq libraries were prepared as previously described (Trombetta et al., 2014). Briefly, cells were FACS sorted into
96-well plates containing 10 mL of RLT lysis buffer (QIAGEN) supplemented with 1% b-mercaptoethanol (Aldrich) in each well. Two
plates of GFP+CD1d-tetramer+ and two plates of the GFP+CD1d-tetramer- cells were sorted. Smart-Seq2 whole transcriptome
amplification (WTA) was performed as per (Trombetta et al., 2014) with Maxima Reverse Transcriptase used in place of SuperScript
II (Life Technologies). Libraries were sequenced on an Illumina NextSeq500, with 30 base-pair paired-end reads, to an average depth
of 670k ± 30k reads per cell.

Single-cell RNA-seq Data Analysis

Sequenced reads were aligned to the UCSC mm10 mouse transcriptome and gene expression was calculated using RSEM as pre-
viously described (Tirosh et al., 2016). Cells with greater than 1,000 genes mapped and with more than 10,000 transcripts detected
were used for downstream phenotypic analysis. Genes expressed in fewer than 15 cells were filtered out from further analysis, leav-
ing 222 cells and 9,271 genes. Expression data was log transformed (Log2(tpm+1)) before further analysis. Clustering and differential
expression were performed using the Seurat package (version 1.4.0.16) in R (version 3.3.2) (Satija et al., 2015). Briefly, we identified
708 variable genes with log-mean expression values greater than 0.05 and dispersion (variance/mean) greater than 0.8. A principal
components analysis was then run over variable genes, and the first five principal components were selected for downstream ana-
lyses based on a combination of Q-Q plots and the standard deviation of PCs as determined by an elbow plot (calculated using the
Jackstraw algorithm in Seurat). tSNE and SNN clustering (FindClusters function in Seurat with k.param = 20 & resolution = 0.6) were
used to depict and cluster the cells in each analysis, respectively. Genes differentially expressed between identified clusters were
calculated using the bimodal distribution option. Differentially expressed markers were removed if identified in less than 20 percent
of cells in a given cluster or if the average log fold change between clusters was less than 0.5. Heatmaps depict the 25 most signif-
icantly differentially expressed genes (p values < 0.002) in each cluster (full data available in Table S1). Violin plots were generated
over curated NKT genes.

Transcriptomic analysis of samples from Zika infected macaques

Genes signatures for NKT, NKT subpopulations, IL4, IL18, TfH, STAT6 and B cell signaling were compiled from
different public repository (http://software.broadinstitute.org/gsea/msigdb/, http://pathcards.genecards.org/, PMID:16698943,
PMCID:PMC3528072). We correlated the expression of these signatures in lymph nodes from ZIKA virus infected Rhesus monkeys
to neutralizing antibodies measured in plasma from the same animals on day 7 and day 14 (Aid et al., 2017). We used a linear regres-
sion analysis (R language) to identify genes that correlated significantly to neutralizing antibodies or viral loads using a p value cut-off
of 0.05. Significant genes were then used to perform a genesets enrichment analysis using the above signatures (http://software.
broadinstitute.org/gsea/index.jsp). Significant genes were used to infer gene interacting networks using GeneMania
(http://genemania.org/) and Cytoscape (http://cytoscape.org/).

QUANTIFICATION AND STATISTICAL ANALYSIS

Statistical parameters including the exact value of n with the description of what n represents, the mean, the SEM and the p value are
reported in the Figures and the Figure Legends. Statistical analyses were performed using Prism (GraphPad Software), two-tailed
t test, p values < 0.05 were considered significant. In figures, asterisks stand for: *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001).

DATA AND SOFTWARE AVAILABILITY

The single-cell RNA-seq data reported in this paper have been deposited in the NCBI Gene Expression Omnibus (GEO) database
under accession number GEO: GSE103753.

e7 Cell 172, 517–533.e1–e7, January 25, 2018

Supplemental Figures

A Day 7 B Day 7
CD1d-/- * CD1d-/-

CXCR5+PD-1+ T cells (%)

5 14

Fas+ GL7+ B cells (%)

10 5
10 5

3.6% 1.7% 5.1% 4.9%

104 104
2.5 7
103 103

CXCR5
0
0
Fas

0 0
0 103 104 105 WT CD1d-/- 0 103 104 105 WT CD1d-/-
GL7 PD-1

C Day 21 D Day 21
CD1d-/- CD1d-/-

CXCR5+PD-1+ T cells (%)

10 5
Fas+ GL7+ B cells (%) 13 10 5 40
10.8% 10.6%
104 104 20.8% 20.3%
6.5 20
103 10 3

CXCR5
0 0
Fas

0 0
0 103 104 105 WT CD1d-/- 0 103 104 105 WT CD1d-/-
GL7 PD-1

E HA trimer HA trimer + α-Galcer HA trimer + Poly I:C *

105
*
16.0% 34 HA trimer
Fas GL7 B cells (%)

0.5% 8.6%
HA trimer + α-Galcer
10 4

HA trimer + Poly I:C

103 17
+

0
Fas

0 103 104 105 0

GL7

F Non-immunized α-Galcer Poly I:C

34 Non-immunized
Fas GL7 B cells (%)

10 5
0.8% 0.0% 0.0% α-Galcer
104 Poly I:C
17
+

103
0
+
Fas

0
0 103 104 105
GL7

H Wild type/ HA + α-Galcer CD1d-/-/ HA + α-Galcer

G * 104 * 46
*
104
Fas+ GL7+ B cells (%)

Wild type 105 Wild type

α-HA IgG1 ASC/LN
α-HA IgM ASC/LN

25.4% 0.5%
CD1d-/- CD1d-/-
10 4

102 10 2
23
103

0
Fas

100 100 0
0 103 104 105
GL7

I * J Wild type/ HA + Poly I:C CD1d-/-/ HA + Poly I:C

α-HA IgG1 ASC/LN (x102)

120 25 10
Fas+ GL7+ B cells (%)

Wild type 105 Wild type

α-HA IgM ASC/LN

6.4% 4.4%
CD1d -/- CD1d-/-
10 4

60 12 5
103

0
Fas

0 0 0
0 103 104 105
GL7

(legend on next page)

Figure S1. NKT Cell-Deficient Mice Display Early Impairment in B Cell Immunity, Related to Figure 1
(A–D) Flow cytometry analysis of mediastinal lymph node cells from wild type or CD1d
/
mice that were intranasally infected with 200 PFU of influenza virus for
(A-B) 7 or (C-D) 21 days. Representative contour plots display the percentage of (A-C) B220+ cells bearing biomarkers of germinal cell activity, Fas+GL7+, and
(B–D) CD4+ T cells bearing biomarkers of TfH cells, CXCR5+PD-1+. Dot plots show quantification of (A-C) germinal center and (B-D) TfH cells.
(E and F) Flow cytometry analysis of mediastinal lymph nodes harvested at day 10 from wild type mice that were intranasally challenged with (E) 10 mg of HA
trimmer or (F) PBS in the presence or absence of poly I:C or a-GalCer. Representative contour plots display the percentage of HA-specific B220+ cells differ-
entiated into Fas+GL7+ germinal center cells.
(G and I) ELISPOT analysis of IgG1 HA-specific ASCs in mediastinal lymph nodes from wild type and CD1d
/
mice treated as in figure S1E.
(H and J) Flow cytometry analysis of HA-specific germinal center cells in mediastinal lymph nodes from wild type and CD1d
/
mice treated as in figure S1E.
In all quantification charts, each dot represents one mouse. Data are representative from two experiments. Statistical analysis two-tailed Student’s t test;
*p < 0.05.
 A Gating strategy B
Day 3 Day 9

NKT-TfH / CD4+ T cells

105 105 105 1024
IL-21 IL-21
NKT cells

Fold Increase
104 104 104 TfH cells

103 103 103 32

CXCR5
0
TCRβ

CD4
0 0 CD4+ T cells

0 103 104 105 0

0 10 10 3 4
10 5
0 10 3
10 4
10 5
NKT TfH NKT TfH
CD1d-Tetramer FSC PD-1

C ****
240 CD4+ T cells 240 **** CD4+ T cells
105 105
NKT cells NKT cells
104 104
Bcl-6 MFI

Bcl-6 MFI
TfH cells TfH cells
TfH 120 TfH 120
10 3
103
Bcl-6

Bcl-6
0 0 CD4+ NKT
CD4+ NKT 0 0
0 103 104 105 0 103 104 105
CD1d-Tetramer CD1d-Tetramer

D *** *** ** ***

7000 CD4+ T cells 2800 CD4+ T cells
105 105
TfH NKT cells TfH NKT cells
104

SLAM MFI
104
SLAM MFI

NKT TfH cells TfH cells

3500 NKT 1400
103 103

0 CD4+ 0 CD4+
SLAM

SLAM

0 0
0 103 104 105 0 103 104 105
CD1d-Tetramer CD1d-Tetramer

E H Southern blot strategy

Neomycine probe
Wild type locus
Clones 1 2 3
Sac I 10.6 kb Sac I

Southern probe CD1d (exon 3)

Targeted locus

Sac I 7.3 kb Sac I Sac I

Southern probe lox P site CD1d (exon 3) Neor FRT site

Neomycine probe

7.1 kb 6.6 kb
Primer 1 Primer 2 Primer 3 Primer 4

F Long range PCR strategy G Southern blot strategy

Primers 1-2 Primers 3-4 Southern probe
Kb Kb Kb

8 8 10
6 6 8
5 5 6

Clones 1 2 3 1 2 3 Clones 1 2 3

(legend on next page)

Figure S2. Expression Analysis of TfH Markers by NKT Cells and Strategy for the Generation of CD1d-Floxed ESCs, Related to Figure 2
(A) Flow cytometry gating strategy for the analysis of NKT cells (TCRb+ CD1d tetramer+), CD4+ T cells (TCRb+ CD1d tetramer- CD4+) and TfH cells (TCRb+ CD1d
tetramer- CD4+ CXCR5+ PD-1+) in mediastinal lymph nodes at day 9 of influenza infection.
(B) RT-qPCR analysis of IL-21 expression at day 3 and 9 of influenza infection by sorted NKT and TfH cells compared to CD4+ cells.
(C and D) Representative contour plots show the expression of (C) Bcl-6 and (D) SLAM by NKT, TfH and CD4+ T cells at day 3 and 9 of infection. Quantification on
the right charts shows the mean fluorescence intensity for (C) Bcl-6 and (D) SLAM.
(E) Scheme showing the wild type and targeted CD1d locus containing two loxP sites flanking the exon 3 of CD1d as well as two FRT sites for the removal of the
Neomycin resistance cassette.
(F) Specific fragments amplified by long-range PCR for the 50 end (7.1kb) and 30 end (6.6kb) of the clones that correctly performed homologous recombination.
(G) DNA from three positive clones was digested with SacI restriction enzyme and Southern blot performed using a labeled Southern probe. The upper band
corresponds to the wild type locus (10.6kb) and the lower band to the targeted locus (7.3kb).
(H) Southern blot of the three positive clones performed with a labeled Southern probe against the Neomycin cassette. Orange arrow indicates the single 5.2kb
band recognized by this probe, indicating single insertion of the construct. The clone 2, in red, was chosen for injections.
In all quantification charts, each dot represents one mouse. Data are representative from three experiments. Statistical analysis two-tailed Student’s t test;
**p < 0.01, ***p < 0.001 and ****p < 0.0001.
 A B C
**** 80
Groups Day -1: Adoptive transfer

Cytokine+ NKT cells / LN

106 2400

NKT/CD4+ T cells
Cytokine+ cells / LN

1 IL4-GFP+ CD45.2 No adoptive transfer

Fold Increase
10 4
2 IL4-GFP- CD45.1
40 No adoptive transfer
1200
102 3 IL4-GFP- CD45.1 + IL4-GFP+ CD45.2 NKTs
N.D.
0 4 IL4-GFP- CD45.1 + IL4-GFP- CD45.2 NKTs
100 0

-4

3
IFN-γ IL-4 IFN-γ+ IL-4+

-5
+ +

-1
IL

IL

IL
D Lymph Nodes - Day 3 Group 1: IL-4 GFP+ CD45.2 mice Group 2: IL-4 GFP- CD45.1 mice

105 105 58.1% 105 0% 0%

99.0%
NKT cells
104 104 104

103 103 103

CD45.2
CD45.2
TCRβ

0 0
0

0 103 104 105 0 103 104 105 0 103 104 105 0 103 104 105 0 103 104 105
CD1d-Tetramer CD45.1 IL-4 GFP CD45.1 IL-4 GFP

Group 3: CD45.1 mice + IL-4 GFP+ CD45.2 NKT cells Group 4: CD45.1 mice + IL-4 GFP- CD45.2 NKT cells

105 105
0.12% 75.0% 0.11% 33.3%

10 4 10 4

103 103
CD45.2
CD45.2

0 0

0 103 104 105 0 103 104 105 0 103 104 105 0 103 104 105
CD45.1 IL-4 GFP CD45.1 IL-4 GFP

E Lungs - Day 3 Group 1: IL-4 GFP+ CD45.2 mice Group 2: IL-4 GFP- CD45.1 mice

105 105 56.7% 105 0% 0%

99.5%
NKT cells
104 104 10 4

103 103 103

CD45.2
CD45.2
TCRβ

0 0 0

0 103 104 105 0 103 104 105 0 103 104 105 0 103 104 105 0 103 104 105
CD1d-Tetramer CD45.1 IL-4 GFP CD45.1 IL-4 GFP

Group 3: CD45.1 mice + IL-4 GFP+ CD45.2 NKT cells Group 4: CD45.1 mice + IL-4 GFP- CD45.2 NKT cells

105 105
0.26% 87.5% 0.27% 22.2%

10 4 104

103 103
CD45.2
CD45.2

0 0

0 103 104 105 0 103 104 105 0 103 104 105 0 103 104 105
CD45.1 IL-4 GFP CD45.1 IL-4 GFP

(legend on next page)

Figure S3. Analysis of the NKT Cell Source of IL-4, Related to Figure 3
(A) Quantification of IFN-g+ (blue) and IL-4+ (red) producing lymph node (left chart) or NKT cells (right chart) in mediastinal lymph nodes after 3 days of influenza
infection.
(B) RT-qPCR analysis of IL-4, IL-5 and IL-13 expression at day 3 of influenza infection by sorted NKT cells compared to CD4+ T cells. Each dot represents
one mouse.
(C) Adoptive transfer strategy for the analysis of the source of IL-4+ NKT cells after influenza infection.
(D and E) Flow cytometry analysis at day 3 of influenza infection of (D) mediastinal lymph node or (E) lung NKT cells from mice that were treated as in (C) to detect
IL4-GFP expression.
Data are representative from two experiments.
(legend on next page)
Figure S4. Single-Cell RNA-Seq Analysis of IL-4+ Lymph Node Cells, Related to Figure 4
(A) Confocal microscopy of mediastinal lymph nodes sections at day 3 of influenza infection. Lymph nodes were labeled with PBS-57 loaded CD1d tetramer
(magenta) and antibodies to B220 (white) and CD169 (green). IF, interfollicular. Scale bars, 300 mm (left); 60 mm (right). Chart on the right shows the quantification
of NKT cells-macrophages contacts observed in the different areas of the lymph node at day 0, 1, 2 and 3 of infection in mediastinal lymph nodes. Data are
representative from three experiments.
(B) Flow cytometry analysis of IL4-GFP+ CD1d-tet- cells in mediastinal lymph nodes at day 3 of influenza infection showing TCRb- and TCRb + cells.
(C) Single-cell RNA-seq heatmap showing the expression of 25 significant marker genes (Table S1) for the CD1d-tet+ clusters (columns 1 and 2) and CD1d-tet-
clusters (columns 3 and 4). Single cells and genes were determined on the basis of unsupervised hierarchical clustering (Table S1).
(D) Expression distribution (violin plots) in each population (horizontal axes) for Ccr7, Sell, S1pr1, CD4, CD8a and Bcl6. Expression levels are Log2(tpm+1).
(E) Flow cytometry analysis of IL4-GFP+ TCRb+ CD1d-tet- cells in mediastinal lymph nodes at day 3 of influenza infection showing CD4+ and CD8a + cells.
(F) Representative contour plot show the expression of GATA-3 by NKT cells, ILC2s (Live/Dead- B220- CD3- CD5- CD11b- CD11c- CD127+ CD45+ GATA-3+) and
CD4+ T cells at day 3 of infection. Quantification on the right chart shows the mean fluorescence intensity for GATA-3.
(G) Flow cytometry analysis of RORgt+ and T-bet+ NKT cells in mediastinal lymph nodes at day 3 of influenza infection. Histogram shows the levels of CXCR3 in
RORgt+ and T-bet+ populations. Dot graphs show the quantification of CXCR3+ NKT cells in the RORgt+ (red dots) and T-bet+ populations (blue dots), with lines
connecting NKT cells from the same mouse.
In graphs showing statistical analysis, data are representative of at least two independent experiments. Statistical analysis two-tailed Student’s t test; **p < 0.01,
***p < 0.001 and ****p < 0.0001.
 A Neutrophils/MDSCs Lyz2-Cre **** B

CD1d+ NFs/MDSCs (%)

100 100 100 CD1d Lyz2-Cre-
10 5

Events (% of max)
Events (% of max)
CD1d Lyz2-Cre+
104 Cre+ Cre-
CD1d-/-
50 50 50
103

0
Gr-1

0 0 0
0 103 104 105 0 103 104 105 Cre- Cre+ 0 103 104 105
CD11b CD1d CD1d

C Dendritic cells CD11c-Cre ****

10 5 100 100
Cre+ Cre -
Events (% of max)

CD1d+ DCs (%)

104

50 50
103
MHCII

0
0 0
0 103 104 105 0 103 104 105 Cre- Cre+
CD11c CD1d

D Wild type CD1d-/- CD1d Lyz2-Cre+ CD1d CD11c-Cre+

CD169

CD1d

E CD1d Lyz2-Cre- CD1d Lyz2-Cre+ F CD1d Lyz2-Cre- CD1d Lyz2-Cre+

CXCR5+PD-1+T cells (%)

12 50
10 5
10 5
Fas+ GL7+ B cells (%)

7.6% 7.9%
.26.8 25.2 %
104 104
6 25
103 103
CXCR5

0
0
Fas

0 0
0 103 104 105 0 103 104 105 Cre-
Cre- Cre+ Cre+
GL7 PD-1

G CD1d CD11c-Cre- CD1d CD11c-Cre+ * H CD1d CD11c-Cre- CD1d CD11c-Cre+

CXCR5+PD-1+T cells (%)

12 20
105 105
Fas+ GL7+ B cells (%)

6.6% 8.2% 13.6% 11.4 %

104 104

6 10
103 103
CXCR5

0
0
Fas

0 0
0 103 104 105 0 103 104 105 Cre-
Cre- Cre+ Cre+
GL7 PD-1

(legend on next page)

Figure S5. Conditional Targeting of CD1d in the Myeloid Lineage, Related to Figure 5
(A–C) Flow cytometry plots showing the gating strategy for (A) Gr-1+CD11b+ neutrophils (NFs) and myeloid-derived suppressor cells (MDSCs) and (C)
MHCII+CD11c+ dendritic cells (DCs). Red and blue histograms depict the levels of CD1d in the gated populations in (A) CD1dflox/floxLyz2-Cre- versus Cre+ and (B)
CD1d
/
mice or (C) CD1dflox/exCD11c-Cre- versus Cre+ mice. These quantitative differences are depicted via dot plots; blue and red dots show CD1d+ cells in
the Cre- and Cre+ animals, respectively.
(D) Confocal microscopy of mediastinal lymph nodes sections from wild type, CD1d
/
, CD1dflox/floxLyz2-Cre+ and CD1dflox/exCD11c-Cre mice. Sections were
stained with antibodies to CD169 (green) and CD1d (magenta). Scale bars, 60 mm.
(E–H) Flow cytometry analysis of mediastinal lymph node cells at day 9 of influenza infection. Representative contour plots show the percentage of B cells and
T cells from (E-F) CD1dflox/floxLyz2-Cre- and Cre+ or (G-H) CD1dflox/exCD11c-Cre- and Cre+ mice that acquire (E and G) germinal center markers, Fas and GL7, or
(F and H) TfH cell markers, CXCR5 and PD-1. Quantifications are shown in dot plots, Cre- mice (blue dots) and Cre+ mice (red dots).
In all panels, each dot represents a single mouse. Horizontal bars – mean; error bars – SEM. Data are representative of three independent experiments. Statistical
analysis, two tailed Student’s t test, *p < 0.05, ****p < 0.0001.
 SCS Medulla
A 30 80
IL-1β IL-1β
Cytokine+ CD169+

Cytokine+ CD169+
IL-18 IL-18
cells / section

cells / section
IL-33 IL-33
15 40

0 0

B
20
**
CD169 GFP Langerin GFP CD169+ cells

GFP+ cells/section
B220 B220 B220 B220 Langerin+ cells

10

C Mixed bone marrow chimeras D Mixed bone marrow chimeras

Wild type-CD45.1 Wild type-CD45.2 Wild type-CD45.1 MyD88-/--CD45.2

14
**
14

IL-4+ NKT cells (%)

IL-4+ NKT cells (%)

105 105
14.0% 13.1% 10.8% 4.9%
104 104
7 7
103 103

0 0
TCRβ
TCRβ

0 0
0 103 104 105 WT WT 0 103 104 105 WT MyD88/-
IL-4 IL-4

E Mixed bone marrow chimeras

Wild type-CD45.1 IL-18R-/--CD45.2

22
*
IL-4+ NKT cells (%)

105
15.8% 10.4%
104
11
103

0
TCRβ

0
0 103 104 105 WT IL-18R-/-
IL-4

Figure S6. Intrinsic Role of MyD88 and IL-18R in NKT Cell Production of IL-4, Related to Figure 6
(A) Quantification of IL-1b+, IL-18+ and IL-33+ cells in the SCS and medullar areas from the images shown in Figures 6C and 6D.
(B) Confocal microscopy of mediastinal lymph nodes from wild type mice that were infected with 200 PFU of PR8 NS1-GFP and sacrificed after 3 days. Sections
were stained with antibodies to CD169 (red), B220 (gray) and Langerin (magenta). Scale bars, 60 mm. Quantification of GFP+ cells is shown in the right chart.
(C–E) Flow cytometry analysis of chimeric mice that were generated by transferring a 1:1 bone marrow mixture of wild type (CD45.1+) and (C) wild type (CD45.2+),
(D) MyD88
/
(CD45.2+) or (E) IL-18R
/
(CD45.2+) to sub-lethally g-irradiated CD45.1 recipient mice. Eight weeks after bone marrow transplant, mice were
infected with 200 PFU of influenza virus and mediastinal lymph nodes were harvested after further 3 days. Flow cytometry panels show the production of IL-4 by
CD45.1+ and CD45.2+ NKT cells in the different sets of chimeras. Dot graphs show the quantification of IL-4+ NKT cells in the CD45.1+ population (blue dots) and
the CD45.2+ population (red dots), with lines connecting the two NKT cell subpopulation coming from the same chimera.
In all panels, each dot represents a single mouse. Horizontal bars – mean; error bars – SEM. Data are representative of two independent experiments. Statistical
analysis, two tailed Student’s t test, *p < 0.05, **p < 0.01.
 A Uninfected α *
0.4 Uninfected

GL7+ area /IgD+ area

IgD
GL7 Control
α-IL4 (d0-d3)
0.2

B Control α-IL4 (day -3) C D Lymph node

8
105 Endogenous OT-II

Fas+ GL7+ B cells (%)

10 5

7.8% 6.4% i.v OT-II T cells Harvest mLNs

104 104
4
103 -1 0 7 103
Days
0
0

CD4
Fas

0
0 10 3
10 4
10 5
Control α-IL4 0 103 104 105
GL7 (day -3) CD45.1

E ** F
Endogenous T cells OT-II T cells Wild type + OT-II CD1d-/- + OT-II *
CXCR5+ PD-1+ T cells (%)

100 2.4

GL7+ CXCR5+ cells (%)

105 105
2.1% 0.9%
104 104
13% 84%
50 1.2
103 103
CXCR5

CXCR5

0 0
0 0
0 103 104 105 Endogenous OT-II 0 103 104 105 WT CD1d-/-
PD-1 GL7

G Oligomycin FCCP Rotenone H Oligomycin FCCP Rotenone **** ****

**** *

Max OCR (pmol/min)

120 110 60 60
Max OCR (pmol/min)
OCR (pmol/min)

OCR (pmol/min)

60 55 30 30

0 0 0 0
0 30 60 90 120 No stimulus 0 30 60 90 120 Non-draining LN
Minutes + IL-4 Minutes Mediastinal LN
α-IgM
Mediastinal LN / α-IL-4
α-IgM + IL-4

I Zika virus Harvest LNs and collect serum K

infection Gene expression and NAbs IL1R1

NKT cells
0.015429415
SMOX
Shared genes
ABI3BP JAG1
0.011072036
0.010043709
0.011657506 0.010829483 EMB
0.01519685
S100A4 SERINC2
0.016368505
HSD11B1
0.01318893 DAPK2
PRELID2 0.016291324
0.010493873
0.014082469
0.011003553
SDC1 0.013271378
0.027093891
IL12RB2
0 14 TNFRSF21
0.011596566
0.016046662
LAX1 0.015700813
0.010854919 0.011594475
0.021436013 0.024041943IL2RB
0.025380895
0.02187106
GLT25D2
0.015635138

IL-18 0.015477868
0.015140184 IGF1R
0.012223331
0.0118996
0.01298504
RGS1 0.012155684
0.021018473
MMP25
ITGB5 LY6E 0.010310485
0.010615125 0.011974632
PPM1J ITGB4
0.010576856
0.012295994 SEPP1
J Lymph node day 14 vs. Neutralizing antibodies day 14
0.012392153 0.014356517 0.012749335
0.015547134 0.015786204
0.012666095 0.014845156
IL17RE STOM0.010053013KLRD1
0.017748201
0.014697292 0.013883438 0.021536583
0.012121715
0.015470931 0.0193451290.014257311
NSA2 0.011824749 DOPEY2
JUN DLG5 CXCR3 0.01838226
CHAD
NOP58
0.019416662
0.011988523 TNS4
0.013870903 0.010203674 0.012169804

NKT.POPULATIONS
0.010741385 0.011468454 0.024453746
0.01119987 0.011534583 0.012416903
0.018724404
0.012484936 MYCN 0.026429152
LEF1
REXO4 NR1H2
0.010380626
0.018384848
0.018884826 0.018018289 0.010821869 NPL 0.024766233 0.012952557 0.013281825
0.012372096 NKG7 FGL2
0.016697038
0.010803388 0.019846603 0.018380737
0.014397028 0.01114185 0.014366682
0.024408918 0.014546911 0.0168531310.012966084
WDR76 PIAS2 0.012680381
0.016242001
0.015358985 0.014478071 RAMP3
0.015070294 LZTFL1
TOP50. MOST.DEG.NKT1
0.015073859 0.021915143
0.013336151 0.012216817
0.015909238 0.012319515 0.022630535
0.021525236 0.015622756
0.011797441
0.017880553 0.0127865390.015282566 0.02527271
0.012307171 0.017929636 0.0214925370.01246645 0.0108482
0.016033484
0.028943405 0.012016931
0.011852172 KCNK1
0.018027455 0.011699469 0.01223024
0.012954631 0.013973099
0.012140085 0.012918495
SAMM50 0.0188610130.019491881 0.015089986
0.016516179
0.022146184
0.011858172 0.021689225IL23R 0.015702274
0.013282981 0.012160562
B cells
0.014758236 0.011489145 0.013006836 0.013410385
0.011200354
0.014188903 0.023279125
0.01496506
HAUS3 0.014385412 0.014758769 0.012152531 0.01182013

TOP50. MOST.DEG.NKT2
0.015487644 0.010114142 0.01426356 0.012382553 0.010942082 0.012266212
0.016195277 0.012067639 0.025139615 0.014442964 0.010597174
0.012470361 0.022872396 0.012702811
0.0109117040.01923993 0.013117665 0.010548177
ACD CREBBP 0.015949557 0.015236114 0.025105948
0.011752614 0.019264285 BLK
0.017569305 0.015386345
0.011562778 0.0130232360.015751053
0.017598122 0.01193442 0.022727065 0.019486673 0.011779791 0.017603973
0.017353406 0.010221882 0.017691877
0.013915279 C8A 0.010677717
IL-4
0.012835389
0.013202892 0.011997514 0.012796179
0.015359912
RASA2
TOP50. MOST.DEG.NKT17
0.01148795 0.013382702
MAPK9 0.016460529
0.013167338
0.010346037 0.017676152 0.010473339 0.019568302 0.019858915 0.010072585
0.015253469
0.014969677 0.013622384
0.018407391
0.017531956 0.019008962
0.012216643 0.015445159
0.020095417 0.017799322 0.017631747STAT4
0.018375628
0.010765362
0.015953310.013619585 0.011542165 SYK 0.014893764 0.012215213
0.010998044
0.01975779
0.010576462
0.016110955 0.010426386 FGF18
0.011319747
0.011845951 IKBKB CDH17
0.010140981 0.012525561
0.011709638
0.0129039530.015109989 0.022537433 FCER1G 0.02436001
0.01932949 CD8B
0.010239518 0.01071016
0.018386347
0.019401101
0.014756741
0.012266147
0.016478315 0.019934522 0.016367253 0.011317985 0.010143301
0.011454908
0.012060402 BCL11B 0.012196074
HIPK2 0.015686754
0.010649092
0.01794713
0.012017687 0.01586429 0.013558399
0.016010523
0.017516598
0.018996222 0.016142515
0.012560469 0.014118372 0.017370477 0.018122507 0.018085927
0.011026372
0.015472411 0.014927189
0.020435147
0.011014057 0.014162485FGFR3
0.011067537

IL18 SIGNALING
0.016973529
0.012123006 0.0121084630.010864384
0.016515503 0.011351636
IL7R
0.014697029
0.013609584 0.016408242
0.017461097 0.018260669
0.016756127
0.012162902
0.015341596 0.014413077 IL10
0.011005486 0.010123569 0.015307621 0.013506545 0.0114115510.019811174 0.014642898
0.0201090160.013141713 0.01288066
0.013935988
0.010016678 ITGB1
0.022851929 0.01405701 0.017454084
0.028544348 0.013778259IL6R
0.013826034 0.013029292 0.01438726
SOS1 PTPN6 DOK2 0.013504313 0.014178259 0.01924158 0.017502153
0.010022604
0.018670155 0.01429501
ID2 0.011133187
0.01189978
0.01457029 0.015154514
0.011157078
0.017771358
0.012071574
0.011179866
0.018555671
0.015419094
0.026962144 IL18R1 0.017983206
0.017094193 0.01050905 0.010053773
0.015855901 0.012062198 0.014532668
0.018380629
IL4 0.020613749 0.016881403
0.012463135 0.010650929
FOS 0.022283358 0.020470127
0.013430852 0.021557692 0.015242511
0.012265869 0.01909311
0.01457225 CD46
IL4.SUPERPATHWAY.GENECARD
0.02075929 0.014598424 0.023875691
0.014262081 0.012234556 0.014802155 0.010140702
0.01796581 0.020615323 0.021126982 0.016847864
0.015919045
0.011807513 0.015290955
0.018186785
0.016529085
0.011052549 0.010804154
0.022675749
0.01646053 BLNK CR1
0.019055948 0.0222582 CD28
0.011615388
0.020834110.01456436
0.016252514
0.017538473 0.010160125
CD247
0.012511780.02000524
0.020541523
0.013444564 0.010308822
0.01561761
0.015290309 0.016188892
0.021475537
0.012495874
0.010817757
0.02271407
0.027155261
0.015857125
0.016241718IKBKG
0.013313063
0.017606745
0.012674276
0.012128019 0.010078654
0.012785444
0.0118290930.012375739 0.011458408 MALT1 0.012516972
0.011248583 0.023192078 0.011918926
RBPJ
0.012631252 0.019966897
NAA15 HTT0.011128846 0.010231567 0.012519578 0.024636596 0.011609847
GTF3A
0.010252288 0.019081626
0.02714273 0.010215194 0.01680042
0.01639617 0.013472381

STAT6 CHIP_SEQ TARGET GENES

0.010242475
0.015535207
0.014385676 0.018094767
0.0225418580.013337469
0.011972174 0.015153943 0.012768193 0.010223722
0.011460828
0.018284597
AKT2 ARHGAP25 0.017572574 0.016137786 0.023352617
0.011038058 0.010160251 0.021056995 CBLB
MFNG PLCL2
0.013588131 0.018218767
0.016040852 0.011226529
0.010315209 0.010070749 0.010225759
0.019026786
0.019043218 0.013478593
0.01326289
0.016599866
0.016752426 0.023438638
0.019190107 0.010816079 0.011549573 0.020272927
0.011943626
0.017233074
0.02115039
0.012959125 0.025361568 0.010391866
0.010786807 0.012120673
0.018868322
SOCS5 0.013662432AKT3
0.013456334
0.012118213
0.02288302
0.01974293
0.025012394 0.01786066
0.01596571
0.016494822 ICOS 0.019275704
0.018102817
0.011486704
0.014396441
0.012217976
ST6GAL1
0.010164227
0.013485089 0.011639956
CD3G PIAS4 0.013408039
0.011012814 0.013829924 0.020411162 0.011256514 0.019865047 0.021956917
IL2RG 0.01194442

TFH SIGNALING
0.016143924 0.017440299 0.012575164
0.010278405
0.013020146
0.013563842 0.012880244 0.011282202
0.015273024 0.010432524
LTA SELP 0.023925828 0.016336827
0.019667512 0.020390745
0.014642892
0.015034821
0.011850583
0.018879987
0.011716519
0.018463304 0.024837721
0.019056708 0.010674533
0.015197326 0.011901393
0.012124317 0.012727001
0.015235992 0.016518379
0.010752411 EPAS1
0.012762003 RBMS1
0.011148687
0.010971621
0.011232376 0.015155939 0.014245229 0.011014584 IL16 0.023318753
0.017014984 0.01215963
0.011802149
0.014201781 0.012606612 ABL1 C7
0.01192374LY96
0.020312231
0.010711143
0.017110558
0.013376961
NSMCE1
0.019880466 DPP4 0.014719022
0.011036912
0.012134356
0.016481398
0.015629105
0.021984056
0.011245846 0.012453869 0.0106632680.012528501
0.014649993 0.017748993 C2
JAK2 SETX
0.020539515
0.020390470.016721515
0.015989935
0.021171153 0.0105913 0.015308663 0.011733402 0.012605351
0.018221427
0.013636362
0.010162055 0.024552554 SOCS3
0.014329486 0.010500984 0.012606218 ILK
MTOR 0.011163474 0.023238668
0.018055838
0.017778596
0.011650304
0.016963454 0.020786095
0.01341601
0.012649393
0.020701809
0.012395194
0.016361717 0.011438731 0.0123226480.010138639
0.014466414
0.012005043
0.0118528830.010127493 0.013972893 0.011313702 0.015323334 0.013766988 0.0112163470.011730175 0.013150808
0.010904187 0.019118035
0.019449813 0.014392233
0.017696970.014355966 F13A1 0.014572669 0.01141884
0.012852606 0.011243172
BCL2
0.010847453 0.02787192
0.013832759
ITGAL
0.025844308 DARS2 0.014820142
0.010508782

SIGNALING BY BCR
0.013721234 0.014805953 0.013660352
0.022239044 0.013736916 0.01077254
0.023584234 0.017119745 0.016666163
0.020039102 STK4
0.013508984
0.013023692
0.0139914960.014817731 0.011125219 0.022896403
0.011446012
0.010590819 0.017499158 0.016422013
0.013709232
0.013598777
0.010283172
CD37 CTLA4 0.01040281
0.018833889
0.0168711
0.017022973 0.016328994 0.020163452
0.013631523
0.012944256 0.023602813 0.012904036 0.013183873
0.021667331 0.012352059
0.012071137
0.013975258
0.010728808
0.010055012
0.017268775
0.012403385
0.014885793 0.012441435
0.01051914 0.019712845 0.0169122 0.014223583
0.016470721
0.012672219 0.011987714
0.016996382 0.02045104 0.012638848
0.011017904 0.012936182
0.010994026 ITK 0.01305955
B_CELL_DIFFERENTIATION
0.013370895 0.01257035 0.010141686
0.010110741 0.018770345 0.0111325 0.010647717 0.017339904
0.013265601 0.010056509 0.016414134
0.010534098 0.0141636080.011335699 PRKCH
0.011991788
0.010221333
0.015244135
0.01100086
0.019989913PPP3CA
0.01361251 PDZD2 0.018452639
0.012497327
0.01257054
0.022676373 0.019746821
0.016250093 0.012417934 0.014703322 0.020646857
0.015395921 0.02250453 0.012559289 0.012650777
0.015612876
0.022369953
0.0194265060.013611216 0.017876709 0.014895173
0.011509113
0.011009477 0.01735941
0.015406711
0.013177319 0.013092094
LAT2
0.012618517
0.01068719

GO_B_CELL_PROLIFERATION
0.011079663
SENP7 0.018641843
0.028772091
0.011997484 0.01326291 0.0108434 0.012785671
0.016628686
0.012945394
0.015565435 0.015906308 PLCG1
0.014832935
0.016253738 0.012140101 NFAT5
0.021482557 0.011422238PIK3R1
0.013796593
0.012712552
PIK3R5 SDK1
0.019513523
0.010092102 0.018580282
0.024038987
0.022100780.015137965
RAC1 0.013783992 ELMO1 THADA0.0137217830.01341504
KIAA0748 0.014050107 0.020789623

GO_B_CELL_ACTIVATION
0.013748328 0.012073907 0.011343884
0.010132535 0.021527624 0.017174827
0.01962966 0.02021398 0.010287361 0.022720525
0.010576494 0.02306794
0.011925466 0.010089644 CD84
ISY1 NUP210 0.013633448 PRKCD PTPRC
0.013000086
0.010749958
0.014653462 ATP2A1
0.017157072 CARD11
0.022594852
0.012786892 0.018419405
0.017203087 MYH10
IL24
0.015307196 0.0119200260.014008874 0.010614986

STAT6
EOMES POLR1A
0.013279185
PRKD3
0.01600918
0.024862885
NFATC2 TfH cells
Genesets normalized enrichment score
0.0119275390.01688171 0.013244064
CA6 0.012110713 0.01189715 LCP2 0.01206414
0.011056357 NFKBIB 0.012809452
0.012805958 0.013091532 NFATC3
LRRC16A RORA HIPK1 0.013081491
0.010461881

CUBN 0.019292485 0.012091244 TMEM68

TXK
DDX26B TLR1
0.017498158
VAV1
0.010950093
RAB7A ZCCHC6
-2 0 +2 STK17B

(legend on next page)

Figure S7. Importance of Early IL-4 Production on B Cell Immunity, Related to Figure 7
(A) Representative confocal microscopy images of mediastinal lymph node sections from mice that were treated as in Figure 7E. Sections were stained with
antibodies to the B lymphocyte marker, IgD (green) and the germinal center marker, GL7 (magenta). Scale bar, 450 mm. Quantification of the germinal center area
in confocal images by ImageJ is shown in the right chart.
(B) Flow cytometry analysis at day 8 of mediastinal lymph nodes from wild type mice that were treated with PBS or anti-IL4 blocking antibody on day
3, and
infected with 200 PFU of influenza virus at day 0. Contour plots show Fas+GL7+ germinal center cells. The responses are quantified in right charts.
(C) Experimental scheme showing the strategy for the generation of antigen-specific TfH cell precursors.
(D–F) Flow cytometry analysis of mediastinal lymph node cells from mice treated as in (C). Representative contour plots display (D) the gating strategy used to
differentiate endogenous CD4+ T cells from adoptively transferred OT-II cells, (E) the percentage of endogenous and exogenous cells bearing the TfH cell markers
PD-1 and CXCR5 and (F) the percentage of B cells that acquire GL7 marker. Quantifications are shown on the right charts.
(G and H) OCR at baseline and after sequential treatment with oligomycin, FCCP and Rotenone of B cells that were (G) stimulated ex vivo with anti-IgM and/or IL-4
or (H) isolated from lymph node cells of influenza-infected mice (day 3). Bar charts show the respiratory capacity obtained as (OCR after FCCP)- (basal OCR).
Each dot represents an individual measurement.
(I) Experimental scheme showing Zika virus infection strategy and time points at which lymph nodes and serum were collected from macaques.
(J) Linear regression model of gene signatures for NKT cells, TfH cells, B cells, IL-4, STAT6 and IL-18 from harvested lymph nodes at day 14 with neutralizing
antibodies measured in plasma at day 17. Red squares represent positive correlation, blue squares represent negative correlation and white squares represent no
significant correlation.
(K) Gene interacting network inference using GeneMANIA software representing interconnectivity of NKT, TfH cells, B cells, IL-4, STAT6 and IL-18 pathways and
the co-expression of their leading genes.
Unless stated, each dot represents a single mouse. Horizontal bars – mean; error bars – SEM. Data are representative of two independent experiments. Statistical
analysis, two tailed Student’s t test, *p < 0.05, **p < 0.01 and ****p < 0.0001.