(PDF) Cold Exposure Protects from Neuroinflammation Through Immunologic Reprogramming

1

Cold Exposure Protects from Neuroinflammation Through

Immunologic Reprogramming

Martina Spiljar1, Karin Steinbach2, Dorothée Rigo1, Nicolas Suárez-Zamorano1, Ingrid Wagner2,

Noushin Hadadi1, Ilena Vincenti2, Nicolas Page2, Bogna Klimek2, Mary-Aude Rochat1, Mario

Kreutzfeldt2, Claire Chevalier1, Ozren Stojanović1, Matthias Mack3, Dilay Cansever4, Melanie

Greter4, Doron Merkler2*, Mirko Trajkovski1*

1 Department of Cell Physiology and Metabolism, Faculty of Medicine, Centre Médical Universitaire

(CMU), University of Geneva, Switzerland.

2 Department of Pathology and Immunology, Faculty of Medicine, Centre Médical Universitaire (CMU),

University of Geneva, Switzerland.

3 Department of Internal Medicine II - Nephrology, University Hospital Regensburg, Germany.

4 Institute of Experimental Immunology, University of Zurich, Switzerland.

*Correspondence and equal contribution

SUMMARY

Autoimmunity is energetically costly, but the impact of a metabolically active state on immunity

and immune-mediated diseases is unclear. Ly6Chi monocytes are key effectors in CNS

autoimmunity with elusive role in priming naïve autoreactive T cells. Here we provide unbiased

analysis of the immune changes in various compartments during cold exposure, and show that

this energetically costly stimulus markedly ameliorates active experimental autoimmune

encephalomyelitis (EAE). Cold exposure decreases MHCII on monocytes at steady-state and in

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

2

various inflammatory mouse models, and suppresses T cell priming and pathogenicity through

the modulation of monocytes. Genetic, or antibody-mediated monocyte depletion, or adoptive

transfer of Th1- or Th17-polarized cells for EAE abolish the cold-induced effects on T cells or

EAE, respectively. These findings provide a mechanistic link between environmental temperature

and neuroinflammation, and suggest competition between cold-induced metabolic adaptations

and autoimmunity as energetic trade-off beneficial for the immune-mediated diseases.

INTRODUCTION

Maintaining immunity requires substantial metabolic resources (Buck et al., 2017; Demas and

Nelson, 2012; Ganeshan et al., 2019; Hotamisligil, 2017; McDade, 2005; O'Neill et al., 2016) for

steady-state systemic immune surveillance and during inflammation, leading to the generation of

millions of immune cells daily. The life-history theory (Stearns, 1992) proposes that prioritization

of resources between biological programs would depend on the environment. In hostile

environments, resources are shifted away from growth and reproduction programs into

maintenance programs (Okin and Medzhitov, 2012; Stearns, 1992; Wang et al., 2019). In addition

to immunity, these maintenance traits include adaptations to energy scarcity, but also adaptations

to changes in the environmental temperature. Exposure to cold provokes a high cost–high gain

physiologic response aimed at reducing heat dissipation (e.g. by vasoconstriction) and increasing

heat production, mainly by activating the brown adipose tissue (BAT) and browning of the

subcutaneous adipose tissue (SAT). BAT catabolizes energy to enable generation of heat, a

function conferred by the uncoupling protein 1 (UCP1). The cold-induced BAT activation is

triggered by sympathetic innervation (Cannon and Nedergaard, 2004; Stojanovic et al., 2018).

Norepinephrine released from the nerve endings stimulates the beta 3 adrenergic receptor, and

agonists of this receptor are frequently used to mimic cold exposure and experimentally induce

thermogenesis.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

3

In autoimmunity, the organism develops an energetically costly pro-inflammatory immune

response (Okin and Medzhitov, 2012). Such high energy demanding processes may compete

with each other, and prioritization of one task over the other may be a result of an energetic trade-

off. This concept can be of particular interest for autoimmunity, in which introducing an additional

energy-costly program may result in milder immune response and disease outcome. In contrast

to the well-described immune status during obesity (Grant and Dixit, 2015; Hotamisligil, 2017;

Kohlgruber et al., 2016; Li et al., 2020; Man et al., 2017), the repercussions of a metabolically

active phenotype on the autoimmunity, and the potential systemic effects on the immune-

mediated diseases are unknown.

Experimental Autoimmune Encephalomyelitis (EAE) is an animal model for multiple sclerosis

(MS), the most frequent autoimmune demyelinating disease of the central nervous system (CNS).

Similar to MS, EAE is exacerbated by high caloric diet and obesity (Hasan et al., 2016; Piccio et

al., 2008; Timmermans et al., 2014; Winer et al., 2009), notably characterized by a positive energy

balance. CD4+ T cells and monocytes are important cellular players in the pathogenesis of EAE.

Among CD4+ T cells, distinct encephalitogenic subsets are essentially involved in disease

precipitation and are characterized by certain cytokine signatures, including interleukin17 (IL-17),

Granulocyte-Macrophage Colony Stimulating Factor (GM-CSF) and Interferon gamma (IFNg).

Monocytes and monocyte-derived cells are crucial for the resulting tissue destruction of the

central nervous system (CNS), particularly during the effector phase of EAE (Ajami et al., 2011;

Fife et al., 2000; Izikson et al., 2000; King et al., 2009; Mildner et al., 2009; Serbina and Pamer,

2006). Thereby, these monocytes and monocyte-derived cells can express pro-inflammatory

INOS (m1-like) or anti-inflammatory ARG1 (m2-like) and in addition can exert antigen presenting

function via major histocompatibility complex class II (MHCII) upregulation. The role of monocytes

in priming of autoreactive T cells in CNS autoimmunity, and whether cold exposure could affect

this interaction remain unclear.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

4

Our study provides an unbiased characterization of the immune system in different compartments

in healthy conditions and during CNS autoimmunity following exposure to lower environmental

temperature. We demonstrate that cold exposure renders monocytes less activated in the bone

marrow, the circulation and the CNS. The change in the monocyte phenotype affects priming of

pathogenic T cells during neuroinflammation, resulting in reduced T cell cytokine expression and

consequently attenuated EAE. Our data suggest competition between cold exposure and

autoimmunity leading to a constrained immune response, which could be of therapeutic

importance in neuroinflammation, and potentially other autoimmune diseases.

RESULTS

Cold Exposure Modulates Myeloid Cells in Bone Marrow at Steady State

To explore the impact of cold exposure on the immune system, we first focused on the bone

marrow where immune cells originate from. We performed RNA sequencing of the bone marrow

from mice exposed to 10°C for two weeks following an acclimatization period, and compared them

to mice kept at room temperature (Figures 1A and S1A). We shortlisted the significantly different

genes (p<0.05) for pathway and gene ontology analyses and considered pathways and terms

significantly deregulated or enriched when p<0.05. Among the 5088 enriched terms identified by

Metacore Gene Ontology Processes, “Myeloid Leukocyte Activation” was among the top five most

enriched terms (Figure 1B). Metacore Metabolic Network analysis identified 41 enriched networks

(Figure S1B), whereas Metacore Pathway Maps revealed 885 pathways that were changed.

Among the top 10 most significantly regulated ones (Figure S1C), four pathways were consistently

changed (>80% of genes in same direction) (Figures 1C and S1D). These were Macrophage and

DC shift, G-CSF-induced myeloid differentiation, IFN-alpha/beta signaling and Transcription

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

5

regulation-granulocyte development, which mainly contain genes important for myeloid cell

activation and differentiation.

We then examined the impact of cold exposure at cellular and protein level within the bone marrow

by flow cytometry. While cold exposure did not affect the number of total bone marrow cells

(Figure S1E), there was a decrease in Ly6Chi but not Ly6Cint or Ly6Clo monocytes (Figure 1D).

Ly6Chi monocytes showed phenotypical changes characterized by decreased MHCII expression

(Figure 1E), a marker for antigen presentation and activation. Within the total MHCII+ cells, the

population of MHCII+Ly6Chi monocytes was decreased, while the percentage of total MHCII+ cells

within the bone marrow remained unchanged (Figure 1E, gating strategy S1F). Differences in the

MHCII expression on protein level were consistent with the observations from RNA sequencing,

where MHCII genes were within the top downregulated pathways (Figure 1C).

To address whether cold exposure decreases the monocytes by affecting hematopoiesis, we

analyzed hematopoietic progenitors in the bone marrow similarly as previously described

(Hettinger et al., 2013; Luo et al., 2015; Merad et al., 2013). In three out of five experiments, cold

exposure decreased the absolute cell numbers (data not shown) and the percentage of monocyte-

dendritic cell progenitors (MDPs), common dendritic cell progenitors (CDP) and tendentially

common monocyte progenitors (cMoP), while the common lymphoid progenitors (CLP) and most

hematopoietic cells (Lin- Sca1+ c-KIT+ or LSK) remained unchanged (Figure 1F, gating strategy

S1F).

Beta 3 adrenergic receptor signaling is one of the main pathways activated by cold exposure

(Cannon and Nedergaard, 2004; Stojanovic et al., 2018). The beta 3 adrenoreceptor agonist

CL316,243 (CL316) decreased Ly6Chi monocytes, common dendritic cell progenitors and

common monocyte progenitors (Figure 1G), suggesting that beta 3 adrenoreceptor stimulation is

sufficient to partly mimic the effects of cold on the myeloid bone marrow cells and their

progenitors. Together, these results suggest that cold exposure decreases myeloid cell

progenitors and monocytes in the bone marrow, and downregulates monocyte MHCII expression.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

6

Cold Exposure Impacts Monocytes in Blood at Steady State

We next examined the systemic relevance of the cold exposure-mediated changes observed in

the bone marrow by FACS sorting CD115+ blood monocytes (Figure S2A) and subsequent

profiling them via RNA sequencing (Figures S2B and 2A). Among the 58 Gene Ontology

Biological Processes that were enriched, the top 15 terms included leukocyte migration, T cell co-

stimulation and antigen presentation via MHCII (Figure 2B), all processes important for the

immune system in health and disease. Such pathways were similarly found within the top 15 of

653 total Metacore Pathway Maps (no p-value cutoff) (Figure S2C). Metacore Metabolic Networks

analysis revealed that 13 metabolic terms were enriched (Figure S2D). Among the variety of

general cell membrane restructuring elements (p<0.05, total 12), “MHC class II protein complex”

was among the most enriched Cellular Components in Gene Ontology analysis (Figure S2E). To

further profile the effect of cold exposure on circulatory immune cells and in particular myeloid

cells, we performed high parameter flow cytometry (BD FACSymphony), followed by an unbiased

analysis using FlowSOM algorithm (Van Gassen et al., 2015) and visualization via UMAP plots.

Overall, we neither observed any major shifts in myeloid cell (Figure 2C), nor in T cell populations

(data not shown) or numbers, suggesting that cold exposure induces several gene expression

changes on the monocytes without significantly altering the overall immune cell composition.

These observations were further supported by flow cytometry, which revealed that cold exposure

consistently decreased MHCII expression of Ly6Chi monocytes, without altering their overall

number (Figures 2D and 2E, gating strategy Figure S2F). Similarly, as in the bone marrow, beta

3 adrenoreceptor stimulation mimicked the effects of cold exposure and reduced the MHCII

expression of Ly6Chi monocytes (Figure 2F). These data indicate that cold exposure induces

functional changes on the monocytes in the blood by reducing their MHCII expression, without

provoking major changes in the main immune cell populations.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

7

Cold Decreases Monocyte MHCII Expression in Inflammatory Mouse Models

To study the functional relevance of the observed cold-induced monocyte modulations, we used

different immunologic models. First, we amplified circulating monocytes and their MHCII

expression using B16-GM-CSF tumor cells (Dranoff et al., 1993; Menezes et al., 2016) that were

subcutaneously injected into cold exposed-, or room temperature-kept mice. GM-CSF initiates an

inflammatory phenotype in monocytes and depletion of monocyte GM-CSF signaling renders

mice resistant to EAE (Croxford et al., 2015). Cold exposure decreased MHCII expression of

monocytes without changing the monocyte percentage after implantation of the tumor cells

(Figures 3A). To confirm that this effect is monocyte-specific we also analyzed total CD11c+ cells,

the majority of which are dendritic cells that remained unchanged (Figure 3B). Thus, cold

exposure counteracted GM-CSF-induced increase of MHCII on Ly6Chi monocytes.

Next, we triggered a thioglycolate elicited peritoneal inflammation, in which monocytes are highly

attracted into the peritoneum. Cold exposure decreased MHCII expression of recruited Ly6Chi

monocytes/monocyte-derived cells within the peritoneum (Figure 3C), without affecting the

proportion of several other immune cell populations (Figure S3A). These data show that cold

exposure leads to a preferential decrease in the monocyte MHCII expression in various mouse

models irrespective of the inflammatory trigger.

Monocytes can acquire antigen presenting functions, including antigen uptake and migration to

the lymph nodes (Guilliams et al., 2018; Jakubzick et al., 2017). We therefore investigated if

antigen presenting functions of monocytes, as well as dendritic cells (Ganguly et al., 2013), might

be changed upon cold exposure. For this purpose, we performed a FITC painting assay, in which

FITC is applied epicutaneously of one flank of the mouse and allows measuring migratory immune

cell properties into the draining lymph node (dLN) (Allan et al., 2006). We monitored the FITC

uptake in the ipsi- and contralateral LN at two different time points, where 12h indicates possible

initial differences in lymphatic draining, while 18h is the peak and allows an estimate of the overall

changes (Platt et al., 2013). At none of the monitored time points have we observed changes in

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

8

cell count, FITC uptake or MHCII expression on dendritic cells, monocytes and various dendritic

cell subsets (Figure S3B and data not shown), thus excluding that cold exposure affects antigen

drainage to the LN. Altogether, these data show that exposure to lower environmental

temperature leads to a decrease in the monocyte MHCII expression in various inflammatory

mouse models without affecting the lymphatic drainage.

Cold Exposure Attenuates Experimental Autoimmune Encephalomyelitis (EAE)

Monocytes and monocyte derived cells are important pathogenic players in chronic inflammatory

diseases, such as MS. Similarly, monocyte trafficking to the CNS is critical for EAE progression

(Ajami et al., 2011; King et al., 2009) and Ccr2-deficient mice, which lack monocytes outside the

bone marrow (Serbina and Pamer, 2006), show mild (Mildner et al., 2009) or no EAE (Fife et al.,

2000; Izikson et al., 2000). We therefore explored whether cold exposure could affect the EAE

disease outcome. For this purpose, we exposed mice to 10 °C for 2 weeks with an additional

initial acclimatization period of 5 days at 18°C and 5 days at 14°C before immunization, and kept

them at 10°C during the whole EAE (Figure 4A). Strikingly, exposing animals to cold led to a

pronounced attenuation of the EAE disease severity compared to room temperature kept controls

(Figures 4B and S4A). The cold exposed mice also showed delayed onset of the disease, together

with a lower maximum and cumulative EAE score, disease incidence and body weight loss

(Figures 4C-4G) compared to the RT control mice. Histologic evaluation revealed a reduced area

of demyelination in the spinal cord (Figure 4H) and a lower number of lesions (Figure S4B) in

spinal cords from cold exposed mice at the peak of EAE. Mac3+, CD3+ and APP lesions from

LFB-Pas per white matter area showed a tendential decrease (Figures S4C-S4G) without

reaching statistical significance. As the extent of EAE amelioration slightly differed between

experiments, a pool is shown (Figures 4 B-G) of experiments performed under the same

conditions except the housing room, which had no consistent impact.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

9

To gain insights into the level of protection by cold exposure on the EAE-induced gene expression

alterations in the spinal cord, we performed RNA sequencing on mice kept at room or cold

temperature both at steady state, or at EAE peak. Principle component and multidimensional

scaling analysis revealed that when the room temperature-kept animals reached peak of the

disease after EAE induction (Figure 4I), the immunized cold-exposed mice clustered together with

the healthy, steady-state animals (Figures 4J and S4H). Interestingly, all genes that were

increased in the room temperature mice with EAE compared to healthy controls were left

unchanged, or deregulated to a lower extent in the cold exposed mice under EAE (Figure 4K).

These data show that the EAE-provoked spinal cord gene expression alterations are either

reduced, or prevented by cold exposure. Gene expression comparison between spinal cords from

cold exposed mice vs. room temperature controls during EAE revealed 61 up- and 1579

downregulated genes of 15467 total genes (p<0.05, logFC>1) (Figure 4L). Among the top 984

regulated pathways identified by the Metacore Pathway Maps, “Macrophage and Dendritic cell

phenotype shift…”, was the most significantly changed one. This pathway includes MHCII genes

and other activation markers (Figures 4M and S4I), further indicating that the myeloid cell and

antigen presentation are the most prominent changes induced by cold also during EAE. The

pathway further reveals downregulation of CXCL10, a recently identified marker of a pathogenic

monocyte subset (Giladi et al., 2020), which we found similarly decreased in the bone marrow

(Figure 1C).

To investigate whether the length of the cold exposure before the EAE induction is important for

the EAE onset and disease progression, we pre-exposed mice to cold two days before

immunization, and compared them to the animals that were kept 2 weeks at 10°C as in the

standard protocol. Two days cold pre-exposure was sufficient to attenuate EAE, although to a

slightly lower extent than the 2 weeks pre-exposed ones (Figures 4N and 4O), and to reduce

monocyte MHCII (Figure 4P), suggesting that the length of the cold pre-treatment plays a modest

role in the disease outcome.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

10

To address whether warm exposure has an opposing effect on EAE, we exposed mice to 34°C

for seven days before immunization, and kept them at warm during EAE (Figure 4Q). Warm

exposure led to decreased food intake and body weight (Figure S4M, Figures 4O and U). Although

more mice died under warm exposure, the survival rate was not significantly different, and neither

was the day of onset (Figures S4N and S4O). However, in contrast to the effects seen during

cold, the MHCII expression of Ly6Chi monocytes was increased in the warm-exposed mice,

coupled to deteriorated EAE, enhanced body weight loss and increased maximum disease score

(Figures 4R-T and S4J, K). Together, these results suggest that the temperature is a critical

determinant for EAE severity, and that cold exposure is a potent stimulus that attenuates

neuroinflammation.

Long-term cold exposure increases energy expenditure by enhancing the thermogenesis in the

brown and the beige fat, enabling homeothermy (Chevalier et al., 2015). However, when

additional biologic maintenance programs are active, as for instance in inflammation, we

hypothesized that these responses (maintenance of homeothermy and autoimmunity) may be in

competition for energy. We therefore analyzed a variety of parameters involved in thermogenesis

in cold the exposed groups of mice. There was no difference in the body temperature between

the RT-kept, or cold exposed healthy vs. EAE mice (Figure S5A). While under RT, EAE reduced

the blood glucose levels, 3 hour fasting glycemia during cold was unchanged between EAE and

healthy mice (Figure S5B). To further address the functional properties of the BAT and SAT

during cold, we studied the oxygen consumption rates, which were similar between the cold

exposed EAE and healthy mice in both SAT and BAT (Figures S5C, D). Consistent with the lack

of functional differences, the weights of the BAT and SAT depots of the EAE mice compared to

the healthy cold exposed controls were unchanged (Figures S5E). While the expression of Cidea

and Pparg in the BAT was reduced in the cold exposed EAE mice, the rest of the brown fat

markers were unaltered, and there was no difference in expression in any of the browning markers

in the inguinal SAT (Figures S5F, G). Moreover, we detected no changes in the lipid droplet size

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

11

distribution between cold exposed healthy or EAE mice, together suggesting no detectable EAE-

induced limitation of the BAT and SAT thermogenic response during cold. However, the slightly

milder brown adipose tissue marker activity under EAE, but without overall reduction in the

thermogenesis and oxygen consumption rates may suggest competition for energetic resources

between of the two energy costly programs. Intriguingly, the lack of functional limitation in the

thermogenic response to cold, coupled to decreased neuroinflammation indicates prioritization of

the homeothermic response over the autoimmunity, resulting in ameliorated EAE severity.

Cold exposure activates BAT thermogenesis primarily through UCP1, which uncouples the

mitochondrial respiratory chain activity from the ATP biosynthesis. However, during prolonged

acclimatization to cold, the Ucp1-KO mice are able to compensate for the lack of UCP1 through

be additional UCP1-independent thermogenic mechanisms that can satisfy the thermogenic

needs, but at a very high energetic cost (Ikeda et al., 2017). Surprisingly, Ucp1 deletion rendered

mice even more protected from EAE compared to their wildtype littermates (Figure S5I),

associated with increased BW of the Ucp1 -KO compared to their WT littermates during cold and

EAE, consistent with the overall improved health status. These data demonstrate that the UCP1-

dependent BAT activity is dispensable for these effects. They also show that cold exposure

ameliorates EAE, but without a cost on the overall homeothermic response.

Cold Exposure Decreases Monocyte and T Cell Pathogenicity during EAE

Monocytes play an essential role in EAE development, progression and remission (Ajami et al.,

2011; Fife et al., 2000; Izikson et al., 2000; King et al., 2009; Mildner et al., 2009; Serbina and

Pamer, 2006), and they reach the CNS in inflammatory conditions through the blood stream. To

examine whether cold exposure affects monocytes in the circulation during EAE, we performed

RNA sequencing on blood monocytes after FACS sorting (Figure S5A, S6A, B) at the EAE onset.

Fold change analysis (p<0.05, logFC>1) revealed 660 up- and 86 downregulated genes of 10794

total genes (Figures 5A) in cold exposed compared to room temperature mice. Metacore

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

12

Metabolic Network analysis (p<0.05) showed 23 enriched networks (Figure 5B). Among the 3366

enriched terms identified using Metacore GO, the top 10 terms were mainly metabolic changes

(Figure 5C). Interestingly, Gene Ontology Cellular Components identified “MHC class II protein

complex” among the top 5 of 10 total enriched terms (Figure 5B middle). Within the Reactome

pathway analysis, pathways important for antigen presentation were within the 4 identified hits

(Figure 5B bottom).

The changes in MHCII antigen presentation machinery in blood monocytes at EAE onset between

the cold exposed and the room temperature mice were confirmed by flow cytometry (Figure S6C).

While the percentage of Ly6Chi/int/lo monocytes remained the same (Figure 5D), cold exposure led

to decreased MHCII of Ly6Chi monocytes, coupled to a partial lowering (p=0.069) in the

percentage of MHCII+Ly6Chi M within the total MHCII+ cells but no change in the percentage of

total MHCII expression of CD45+ cells (Figure 5E). Blood and lymph node immune cells at

different days of EAE were further investigated using high dimensional flow cytometry and

unbiased computed analysis and detected no alterations thus excluding changes in the myeloid

cell populations (Figures S6D and S6E). These analyses demonstrate that cold exposure reduced

MHCII expression of circulating Ly6Chi monocytes at EAE onset without shifting myeloid cell

populations.

Circulating Ly6Chi monocytes contribute to EAE pathology by migration to the inflamed CNS

(Ajami et al., 2011; Fife et al., 2000; Izikson et al., 2000; King et al., 2009; Mildner et al., 2009;

Serbina and Pamer, 2006). We therefore investigated if the changes detected in the circulating

monocytes may be seen within the inflamed CNS at EAE onset. Cold exposure decreased the

percentage of Ly6Chi monocytes/monocyte derived cells and their MHCII expression within the

CNS (Figure 5F). Furthermore, their Arg1 expression was increased, indicating a predominant

anti-inflammatory M2-like phenotype (Figure 5F). Percentage and MHCII expression of microglia

were similarly decreased (Figure 5G), together suggesting that they are activated to a lower extent

at EAE onset under cold exposure.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

13

Apart from monocytes and monocyte-derived cells, T cells are the main drivers of EAE

pathogenicity. Cold exposure led to a decrease of pathogenic T cell cytokine signature and a

reduction of GM-CSF/IL-17 and IFNg/IL-17 double producers in the CNS (Figure 5H). We

detected similar changes in the dLN (Figure 5I), whereas in the mesenteric lymph nodes only GM-

CSF was decreased (Figure S6F upper), and no changes were seen in the spleen (Figure S6F

bottom). In the dLN, MHCII expression of Ly6Chi monocytes was lower on day two (Figure 5J) but

not at onset of EAE, or at steady state (data not shown). There were no consistent differences in

dendritic cell populations or their activation status in the dLN at any time (Figure S6G).

Collectively, these data show that cold exposure attenuates Ly6Chi monocyte and T cell

pathogenicity during active EAE.

T Cell Priming by Monocytes is Critical for the Cold-Induced Attenuation of EAE

The described changes in two of the main pathogenic cell types mediating EAE provide

mechanistic support for the observed amelioration of EAE by cold exposure. We therefore

investigated what is the connection between monocytes and T cells during cold and EAE. Since

cold exposure reduced T cell cytokine expression early during EAE and the monocyte changes

were already detected under steady-state conditions, we raised the questions whether cold

exposure affects the priming of T cells and what is the role of monocytes at regulating this stage.

To investigate this, we used an adoptive transfer model for passive EAE induction, in which in

vitro Th1 polarized, transgenic, MOG-specific 2D2 T cells were injected into room temperature or

cold exposed mice (Figure 6A). Similar to the active EAE, cold exposure reduced MHCII

expression of Ly6Chi blood monocytes during the passive EAE onset after transfer of differentiated

2D2 cells, while monocyte percentages remained unchanged (Figures 6B-D and 6E). Priming T

cells in vitro completely abolished the cold exposure-induced amelioration of EAE (Figures 6F-K

upper). Similar results were obtained after adoptive transfer of Th17 differentiated cells (Figures

6F-6K bottom). These data suggest that the effects of cold exposure during the T cell priming

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

14

phase are critical for the attenuation of EAE, and that cold exposure mediated MHCII

downregulation may be responsible for these effects. These conclusions were supported with

data from cold exposed mice that were shifted to room temperature on the day of immunization.

Transferring cold exposed mice to room temperature prior to the T cell priming stage (Figure S7A)

abolished the cold-induced effects on the active EAE disease (Figure S7B) and MHC expression

on monocytes (Figure S7C). This is in agreement with the observation that the MHCII

downregulation in monocytes under cold exposure was not sufficient to ameliorate EAE following

adoptive transfer of pre-activated encephalitogenic T cells, thus making it unlikely that the

observed monocyte changes are downstream mediators of effector T cells.

To directly test whether monocytes may influence T cells during their priming phase in cold, we

used two complementary approaches to deplete peripheral monocytes: anti-CCR2 (MC-21)

antibody and Ccr2-KO mice, both of which restrict monocytes to the bone marrow (Serbina and

Pamer, 2006). While anti-CCR2 completely abolished peripheral blood monocytes 2 days before

EAE onset, Ccr2-KO mice showed significant, but incomplete loss of monocytes from the blood

(Figure 6L), likely because during inflammation there could be a compensatory, CCR2-

independent mechanisms for partial monocyte exit from the bone marrow. While cold exposure

decreased pathogenic T cell cytokine expression in the isotype control mice, both monocyte

depletion mouse models did not show any T cell cytokine expression in the dLN (Figure 6 M)

irrespective of the housing temperature. Together, these data show that monocytes play a critical

role in T cell priming with subsequent implications on the T cell encephalitogenic capacities.

DISCUSSION

Our study has several fundamental implications for understanding the immune system functioning

in relation to an energy-demanding metabolic stimulus, such as the change in the environmental

temperature. The work provides a systematic overview of the immune cell changes in bone

marrow and blood as a consequence of cold exposure during steady-state, and shows that lower

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

15

environmental temperature markedly ameliorates neuroinflammation. The cold-induced gene

expression alterations were accompanied by phenotypic and functional changes of monocytes.

Reduced MHC class II expression and related pathways in these cells was accompanied with

lowered antigen presenting capacity following various inflammatory stimuli. This cold-induced

monocyte modulation resulted in reduced priming of autoreactive T cells, leading to attenuated

autoimmune CNS disease. These data reveal a hitherto underestimated role of monocytes during

T cell priming in autoimmunity.

Circulating Ly6Chi monocytes can acquire antigen presenting functions and can exert pro- and

anti-inflammatory tasks (Jakubzick et al., 2017). They therefore play an important role during

infections but are also involved in autoimmune pathogenesis. The importance of monocytes and

monocyte-derived cells during the effector phase of EAE is well established (Fife et al., 2000;

Izikson et al., 2000; King et al., 2009; Serbina and Pamer, 2006), and in particular the relevance

of migration of Ccr2+ monocytes into the CNS (Fife et al., 2000). This is in line with our observation

that cold exposure reduces monocyte-derived cells during EAE in the CNS. However, our data

also point out that the noted ameliorated EAE disease under cold exposure was associated with

a reduced generation of encephalitogenic T cells, which was most prominent during the early

phase of the disease. This suggests a cold-induced effect on the priming of naïve autoreactive

CD4+ T cells, since adoptive transfer of in vitro pre-activated 2D2 T cells abolished the observed

protection on EAE. While dendritic cells are supposed to be the primary antigen presenting cells

involved in T cell priming (Ganguly et al., 2013; Steinman and Banchereau, 2007), we did not find

indications that cold-exposure impacted dendritic cell function per se. Instead, our data rather

suggest that altered monocytic function was implied in the reduced priming of autoreactive T cells

under cold. Although it is known that monocytes can present antigens to T cells, their role in the

context of CNS autoimmunity for T cell priming remains debated (Guilliams et al., 2018; Jakubzick

et al., 2017). One previous study (Ko et al., 2014) described decreased but not abrogated IL-17

expression in Ccr2-DTR mice at EAE onset, and did not report differences in IFNg or GM-CSF.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

16

However, the Ccr2+ cell depletion was induced transiently and after immunization, and therefore

likely did not address the initial priming phase. Another study found a minor decrease in IFNg/GM-

CSF double producing CD4+ T cells, but no abrogation and no effect on IL-17 in Ccr2-KO mice

before EAE onset (Ronchi et al., 2016). This mild effect on T cell pathogenicity could likely be

explained by the experimental conditions, or by an incomplete monocyte depletion from the

circulation in the Ccr2-KO mice under inflammatory conditions. Indeed, the monocyte depletion

in our study was milder in the Ccr2-KO compared to the anti-CCR2 treated mice. We observed

complete abrogation of pathogenic T cell cytokines after acute depletion of monocytes using the

antibody mediated approach – effects also evident in our genetic, Ccr2-KO mice.

In addition to the decrease in monocyte-derived cells, cold exposure led to reduction of monocyte

MHCII expression in bone marrow, blood and CNS at steady state, or during the early phase of

EAE, which provides a mechanistic link to the observed reduction in T cell priming and

consequently ameliorated disease course. The fact that MHCII in monocytes remained reduced

even after adoptive transfer of pre-activated 2D2 cells that abrogated the protective effect of cold

on EAE, strengthens the conclusions that the cold-induced monocyte changes interfere upstream

of T cell effector cells. We excluded other potential mechanisms by which monocytes could affect

T cell priming - neither the uptake and transport of antigen by monocytes in FITC painted mice,

nor monocyte cytokine expression on transcriptional level at steady-state and EAE onset were

modified under cold. Previous studies have suggested a positive feedback loop between T cells

and monocytes, where T cells secrete GM-CSF to induce monocyte IL-1b expression (Croxford

et al., 2015), which in return amplifies expansion of GM-CSF+ Th17 cells (Mufazalov et al., 2017).

Our study adds an additional layer to this feedback loop, stressing that as an initial step, monocyte

MHCII is critical for T cell priming.

Cold exposure induces beta adrenergic stimulation of the BAT to activate lipolysis, glucose uptake

and mitochondrial biogenesis, and is extensively studied in context of obesity and metabolic

disorders (Jastroch, 2017). Cold-induced BAT activation and white fat browning result in a

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

17

thermogenic response that is energy demanding. As such, this metabolic response may require

an energetic trade-off that competes with other energy costly programs, as for instance immune

responses (Wang and Medzhitov, 2019). Several studies addressed the energy trade-off concept

in the context of immune defense during infections and found that energy conserving

hypometabolic states may favor the utilization of tissue tolerance as a defense against bacterial

pathogens (Ganeshan et al., 2019; Weis et al., 2017). Similar to cold, caloric restriction also

induces negative energy balance, reduces and modulates circulating monocytes (Jordan et al.,

2019), and lowers CNS monocytes and their pathogenicity (Cignarella et al., 2018; Piccio et al.,

2008). In contrast to caloric restriction that requires an intervention to reduce food intake, cold

exposure ameliorates EAE despite the excessive caloric uptake. Certainly, a competition between

thermogenesis and a destructive immune response would be favorable with regard to

autoimmune diseases, as we see in our study. It was speculated that at low temperatures,

defense of remaining thermogenesis over immunity may take place (Wang and Medzhitov, 2019).

Our study is in line with this scenario and shows that cold exposure dampens autoimmunity, thus

providing evidence and a conceptual advance in our understanding of the energetic trade-off at

the cost of autoimmunity. In support of these conclusions, cold exposure ameliorates EAE to a

higher extend in absence of UCP1 and following acclimatization and activation of the UCP1-

independent thermogenic mechanisms, which function at a very high energetic cost (Ikeda et al.,

2017; Reynes et al., 2019; Roesler and Kazak, 2020). The cold-induced immune changes seem

to originate already in the bone marrow and can be partially provoked by beta adrenergic agonism,

which may suggest contribution of the nervous system to the energy trade-off. With regard to this

concept, it is important to stress that while evidently protective against autoimmunity, cold

exposure increases susceptibility to certain viral infections (Foxman et al., 2015; Jaakkola et al.,

2014). Thus, our work could be relevant not only for CNS autoimmunity, but also other immune-

mediated or infectious diseases, which warrants further investigations.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

18

Our study reveals that cold exposure orchestrates immunologic reprogramming that protects from

CNS autoimmunity, showing that monocytes and the monocyte MHCII expression are crucially

implicated during T cell priming in the course of an autoimmune response. These findings provide

critical insights into the pathogenic mechanisms of neuroinflammation, and could lead to future

preventive and therapeutic approaches for MS and other autoimmune diseases.

ACKNOWLEDGEMENTS

We thank the iGE3 Genomics Platform at the University of Geneva for mRNA sequencing and

mapping reads, the Flow Cytometry core facility at the CMU of the University of Geneva for help

with flow cytometry, Christelle Veyrat-Duberex, Florian Visentin, Melis Colakoglu, Giovanni di

Liberto, Salvatore Fabbiano and Karim Hammad for experimental assistance, Sergei Startchik for

the help on automatized lipid droplet quantification, and Stephanie Hughes and Pierre

Guermonprez for organizing and providing B16-GMCSF cells, respectively, Thomas Korn for the

Th17 EAE protocol and Claes Wollheim for discussions. This work was supported by the

CONFIRM grant of the Hôpitaux Universitaires de Genève (HUG) Foundation (no. RC2-09) and

the Swiss Multiple Sclerosis Society Grant to DM and MT; the European Research Council (ERC)

under the European Union’s Horizon 2020 research and innovation programme (ERC

Consolidator Grant agreement No. 815962) to MT; Swiss National Science Foundation (SNSF)

grant to DM (310030_173010); and SNSF Professorship grants (PP00P3_144886 and

PP00P3_172906) to MT.

AUTHOR CONTRIBUTIONS

M.S. set up and performed experiments, analyzed and interpreted data, and drafted figures and

manuscript. K.S. set up and performed experiments, and analyzed and interpreted data. N.H. and

M.S. performed RNA-seq analysis. D.R., N.S.Z., M.A.R., C.C., I.V., N.P., B.K. and O.S. helped

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

19

with experiments. I.W. assisted in histological processing. M.G. assisted in setup of and D.C.

performed and analyzed FACSymphony. M.K. developed and applied computer-assisted image

analysis algorithms. MM provided the MC-21 antibody. D.M. and M.T. conceptualized, designed

and supervised the project, interpreted data, and wrote the manuscript.

DECLARATION OF INTERESTS

The authors declare no competing interests.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

20

FIGURE LEGENDS

Figure 1. Cold Exposure Impacts Monocytes in the Bone Marrow

(A) Volcano plot showing the up- and downregulated transcripts by RNA sequencing in the bone

marrow of cold exposed mice for 2 weeks at 10°C compared to room temperature counterparts

(n=6 mice per group).

(B) Metacore Gene Ontology Processes analysis displaying the top 5 enriched processes of mice

as in (A).

(C) Shown are 3 of 4 heatmaps that were regulated in the same direction (> 80% of genes in the

same direction) from the top 10 differentially regulated Metacore Pathway Maps of mice as in (A).

Full pathway names are “Macrophage and dendritic cell phenotype shift in cancer”, “Immune

response_IFN-aplha/beta signaling via JAK/STAT” and “G-CSF induced myeloid differentiation”.

(D) Flow cytometry analysis of bone marrow cells of mice as in (A). Percentage of Ly6C high (hi),

intermediate (int) and low (lo) monocytes of total, single CD45+ alive cells.

(E) Percentage of MHCII+ cells of Ly6Chi monocytes (left), MHCII+ Ly6Chi monocytes of total

MHCII+ cells (middle) and total MHCII+ cells of CD45+ cells of mice as in (A), as determined by

flow cytometry.

(F) Bone marrow immune cell progenitors of mice as in (A) were analyzed by flow cytometry and

percentage of total alive single CD45+ cells is shown for common dendritic cell progenitors (CDP),

cMoP (common monocyte progenitors), monocyte-dendritic cell progenitors (MDP), common

lymphoid progenitors (CLP) and Lin- Sca1+ c-KIT+ cells (LSK), as determined by flow cytometry.

(G) Immune cell progenitor flow cytometry analysis of mice that were i.p. injected with beta 3

adrenoreceptor agonist CL316,243 (CL316) or vector (PBS) daily during one week. Percentage

of total alive CD45+ cells is shown of CDP, cMoP, MDP, CLP, LSK and Ly6Chi monocytes.

(B-C) The cutoff on differentially regulated genes considered for the pathway analysis is p<0.05

and pathways are considered deregulated with p<0.05. Shown are –log(p-value).

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

21

(D-G) Each dot represents one animal. Shown is mean ±SD. Significance was calculated using

student’s t test: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Pool of five experiments (D), two experiments

(E), one representative experiment of three out of five that were similar (F) and one representative

of two are shown.

Figure 2. Cold Exposure Modulates Monocytes in the Circulation

(A) Volcano plot showing the up- and downregulated transcripts by RNA sequencing of

monocytes that were MACS (anti-PE) and FACS (CD115-PE+) sorted from blood of 2 weeks cold

exposed (10°C) mice compared to room temperature mice.

(B) 15 most enriched Gene Ontology Biological Processes of monocytes as in (A). Genes were

considered differentially regulated with p<0.05. Shown are –log(p-value).

(C) Blood immune cells of mice as in (A) visualized using UMAP and clustered using FlowSOM

algorithm in R. Ly6Chi monocytes (green), Ly6Clo monocytes (yellow), neutrophils (brown),

eosinophils (orange) and others. Heatmap shows median relative expression of all panel markers.

(D) Flow cytometry analysis of blood cells from mice as in (A). Percentage of Ly6C high (hi),

intermediate (int) and low (lo) monocytes of total, single, CD45+ cells.

(E) Percentage of blood MHCII+ cells of Ly6Chi monocytes (left), MHCII+ Ly6Chi monocytes of total

MHCII+ cells (middle) and total MHCII+ cells of total, single CD45+ cells of mice as in (A).

(F) Blood cell analysis after 1 week of daily i.p. injected beta 3 adrenoreceptor agonist CL316,243

or vector (PBS). Percentage of MHCII+ cells of Ly6Chi monocytes (left), MHCII+ Ly6Chi monocytes

of total MHCII+ cells (middle) and total MHCII+ cells of CD45+ cells.

(D-F) Each dot represents one animal. Shown is mean ±SD. Significance was calculated using

student’s t test: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Pool of three experiments (D, E).

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

22

Figure 3. Cold Exposure Decreases Monocyte MHCII in Inflammatory Mouse Models

(A) Blood Ly6Chi monocytes and their MHCII expression were analyzed by flow cytometry on

different days after B16-GMCSF s.c. tumor implantation in mice that were cold (10°C) or room

temperature exposed for 2 weeks before and during tumor growth.

(B) Blood CD11c+MHCII+ cells and MHCII mean fluorescence intensity (MFI) of MHCII+CD11c+

cells were determined by flow cytometry from mice as in (A).

(C) Flow cytometry analysis of peritoneal fluid cells 24h after i.p. injection with thioglycollate into

2 weeks cold exposed (10°C) or room temperature mice.

(A-C) Each dot represents one animal. Shown is mean ±SD. Significance was calculated using

multiple t-test with Holm-Sidak correction (A, B) or student’s t test: (C), ∗p < 0.05, ∗∗p < 0.01.

Figure 4. Cold Exposure Attenuates Neuroinflammation

(A-B) Scheme showing the experimental setup for EAE and cold exposure (A). EAE was induced

by s.c. immunization with MOG35-55 peptide in complete Freund’s adjuvant (day 0) and pertussis

toxin (day 0 and 2). Mice were exposed to cold or room temperature (10°C) for two weeks before

and during EAE. Clinical symptoms of EAE were monitored according to a standardized scoring

system (B).

(C-G) Day of EAE onset, i.e. when first symptoms occur (C), maximum disease score, i.e. the

highest score an individual mouse reached during the experiment (D), cumulative disease score,

i.e. the sum of all scores each individual mouse reached during the experiment (E), the

percentage of disease free mice (F) and body weight (G) of mice as in (A).

(H) Quantification of demyelinated area expressed as percentage of white matter (WM) detected

by Luxol Fast Blue and Periodic acid–Schiff (LFB-PAS) of spinal cords from mice as in (A).

(I-M) Spinal cords of healthy and peak EAE mice that were kept at room temperature or 2 weeks

cold exposed were used for RNA sequencing. Scores of the respective mice at sacrifice (I),

principal component analysis (J), relativeness analysis of logFC of room temperature healthy vs

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

23

EAE spinal cords (red) and cold exposure healthy vs EAE spinal cords (blue) with p<0.05, FC>2

(K), volcano plot (gene Gpx3 was excluded for better visualization) (L) and genes of top1

deregulated pathway “Macrophage and DC phenotype shift in cancer” from Metacore Pathway

Maps with p<0.05 (M).

(N-O) Scheme showing the experimental setup. Mice were either exposed to cold temperature

for two weeks or two days before and continued during EAE and compared to room temperature

controls (N). EAE was induced as in (A) and EAE scores are shown (O).

(P) Percentage of MHCII expression on Ly6Chi blood monocytes was determined using flow

cytometry on day zero and day four after EAE induction.

(Q-R) Scheme showing the experimental setup (Q). Mice were exposed to warm temperature of

34°C for one week before and during EAE (as in (A)) and compared to room temperature controls.

Each curve represents one individual mouse (R).

(S) Percentage of MHCII expression of Ly6Chi blood monocytes from mice as in (Q) was

determined using flow cytometry on day 0 of EAE.

(T-U) Maximum disease score (T) or delta of body weight gain from day zero to day six of warm

exposure (U).

(B-S) Data (B-G) represents pool of 3 experiments (n = 8-12 mice per group per experiment).

Shown is mean ±SEM, two-way ANOVA (B; G, O, R); student’s t test (C-E, I, S-U); Mantel-Cox

(F); multiple t-test with Holm-Sidak correction (P), mean ±SD, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure 5. Cold Exposure Reduces Monocyte and T Cell Pathogenicity during EAE

(A) Volcano plot identifying the up- and downregulated transcripts after RNA sequencing of

monocytes that were MACS (anti-PE) and FACS (CD115-PE+) sorted from blood of mice at onset

of EAE (as in Figure 4A) and exposed to cold (10°C) or room temperature for 2 weeks before and

continued during EAE.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

24

(B-C) Top deregulated and enriched Metacore Metabolic Networks (B top), Gene Ontology

cellular components (B middle), Reactome pathways (B bottom) and Gene Ontology Processes

(C). Genes were considered when p<0.05 and for Reactome pathways in addition FC>2. Data of

sequencing as in (A).

(D-E) Flow cytometry analysis of blood cells of mice as in (A) at EAE onset. Percentage of Ly6C

high (left), intermediate (middle) and low (right) monocytes of total, single, CD45+ alive cells (D).

Percentage of MHCII+ cells of Ly6Chi monocytes (left), MHCII+ Ly6C+ monocytes of total MHCII+

cells (middle) and total MHCII+ cells of CD45+ cells (E). Pool of 3 experiments.

(F-H) Flow cytometry analysis of CNS cells from mice as in (A) at EAE onset. Percentage of

Ly6Chi monocytes/monocyte-derived cells of total, single, CD45+ alive cells (F, left). Percentage

of MHCII, arginase 1 (Arg1) and iNOS expression of Ly6Chi monocytes/monocyte-derived cells

(F, right). Percentage of microglia of total, single, CD45+ alive cells and percentage of MHCII of

microglia (G). Percentage of corresponding cytokine expression as indicated in CD4+ T cells (H).

(I) Flow cytometry analysis of dLN cells from mice as in (A) at EAE onset. Percentage of cytokine

expression in CD4+ T cells. Observed in 2 out of 3 experiments.

(J) Flow cytometry analysis of dLN cells from mice as in (A) on day two after EAE induction.

Percentage of Ly6Chi monocytes of total, single, CD45+ alive cells (upper panel). Percentage of

MHCII on Ly6Chi monocytes (lower panel).

(D-J) Shown is mean ±SD, significance was calculated using student’s t test, * p<0.05, ** p<0.01,

***p<0.001. Shown is 1 representative example of 3 (F), a pool of 2 experiments (H) or one

representative experiment of two out of three that were similar (I).

Figure 6. Monocyte Regulation of the T Cell Priming is Critical for Cold-Induced EAE

Attenuation

(A) Scheme showing experimental setup for cold exposure and adoptive transfer EAE. T cells

from transgenic 2D2 mice or actively immunized wildtype donor mice were in vitro differentiated

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

25

towards Th1 or Th17 respectively and transferred into room temperature or cold exposed (2

weeks, 10°C) mice.

(B-E) Flow cytometry analysis of blood cells of mice as in (A) at EAE onset. Percentage of Ly6Chi

(B), intermediate (C) and low monocytes of total, single, CD45+ cells (D). Percentage of MHCII+

cells of Ly6Chi monocytes (E).

(F-K) EAE symptoms of mice as in (A) were scored (F). Onset day of EAE disease (G), maximum

score (H), cumulative score (I), percentage of disease free animals (J) and body weight (K). All

panels are shown following adoptive transfer of Th1 (upper) and Th17 (lower) cells.

(L-M) Isotype control antibody, CCR2 antibody injected or Ccr2-knockout (KO) mice were housed

at room temperature or cold (10°C) for 2 weeks before and during EAE and were s.c. immunized

with MOG35-55 peptide in complete Freund’s adjuvant. Two days before disease onset,

monocyte depletion efficiency was analyzed in the blood (L) and at onset T cell cytokine

expression in the draining lymph nodes (M) via flow cytometry. Data represents mean ± SD,

significance was calculated using multiple t-test with Holm-Sidak correction, *p<0.05, **p<0.01,

***p<0.001.

(B-E, F-K, L-M) Pool of 3 experiments (B-E, F-K upper panel), student’s t test with mean ± SD (B-

E, G-I), two-way ANOVA with mean ± SEM (F,K) or Mantel-Cox (J). Multiple t-tests with Holm-

Sidak correction with mean ±SD, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001 (L-M).

Supplemental Figure Legends

Figure S1. (Related to Figure 1)

(A) Multidimensional scaling and principal component analysis of the bone marrow from mice

after 2 weeks of cold exposure at 10°C compared to room temperature mice.

(B) 10 most enriched Metacore Metabolic Networks of mice as in (A).

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

26

(C) Metacore Pathway Maps analysis identifying the 10 most differentially regulated pathways of

mice as in (A).

(D) 1 of 4 heatmaps that were consistently changed (> 80% of genes in the same direction) from

the top 10 deregulated Metacore Pathway Maps of mice as in (A).

(E) Absolute lymphocyte numbers within the bone marrow of mice as in (A) as determined by flow

cytometry. Shown is a pool of 5 experiments, mean ±SD.

(F) Bone marrow monocytes were gated as live, single, CD45+, Ly6G-, CD11b+, CD115+, Ly6C+

cells and MHCII+ population determined. Bone marrow progenitors were all gated as single, live,

CD45+, Lin- (NK1.1, CD3, CD19, Ter119, Ly6G) and CD11b-, Sca1+, CD117hi, CD115+, Ly6Chi

(cMoP), CD11b-, Sca1+, CD117hi, CD115+, Ly6C- (MDP), CD11b-, Sca1+, CD117-, CD115+,

CD135+ (MDP), CD117+, CD127+, Sca-1+ (CLP), Sca-1+, CD117+ (LSK).

(C-D) The cutoff on differentially regulated genes considered for the pathway analysis is p<0.05

and pathways are considered deregulated with p<0.05. Shown are –log(p-value).

Figure S2. (Related to Figure 2)

(A) Gating strategy for the FACS (CD115-PE+) sort after MACS (anti-PE) sort before RNA

sequencing from blood of 2 weeks cold exposed (10°C) compared to room temperature mice.

Sorted were DAPI-, single, CD115+ cells.

(B) Multidimensional scaling (left) and principal component analysis (right) after RNA sequencing

of monocytes as in (A).

(C-E) 10 most differentially regulated Metacore Pathway Maps without p-value cutoff on

considered genes (C), Metacore Metabolic Networks (D) and Gene Ontology Cellular

Components (E). Genes were considered as differentially regulated genes with p<0.05 and

pathways are considered deregulated or enriched with p<0.05. Shown are –log(p-value).

(F) For flow cytometry analysis, blood monocytes were gated as Ly6G-, CD11b+, CD115+,

Ly6Chi/int/lo and MHCII+ positive population was gated.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

27

Figure S3. (Related to Figure 3)

(A) Peritoneal fluid cells were analyzed by flow cytometry 24h after i.p. injection with thioglycollate

into two weeks cold exposed (10°C) or room temperature mice. Shown is mean ±SD.

(B) Flow cytometry analysis of lymph node draining cells 12 and 18 hours after FITC skin painting

on one flank of 2 weeks cold exposed (10°C) or room temperature mice. Percentage of panDCs

(upper) and Ly6Chi monocytes (bottom), their FITC uptake and MHCII expression. Shown is mean

±SD, two-way ANOVA. Every dot represents one animal and the experiment is representative of

two.

Figure S4. (Related to Figure 4)

(A) Individual EAE (as in Figure 4A) curve of each room temperature and cold exposed mouse.

(B-G) Histology on sagittal spinal cord sections of cold exposed and room temperature peak EAE

mice. Number of lesions that were found in both LFB-PAS and APP staining were quantified in

APP staining (B). Immunohistochemistry using anti-mac3 (C) and anti-CD3 antibody (D). APP

staining was quantified within areas identified as demyelinated in LFB-Pas staining and presented

as percentage per white matter area (E). EAE disease score at sacrifice of mice in which histology

was performed (F). LFB-Pas stain shows demyelinated areas within the white matter in purple

(first panel), anti-mac3 stains mac3+ macrophages (second panel), anti-CD3 stains T cells in

brown (third panel) and amyloid precursor protein (APP) accumulates at axons with impaired

axonal transport, a surrogate for acute axonal damage (fourth panel). Arrows identify punctuated

APP-positive axonal spheroids. (G). Anti-mac3, anti-CD3 and anti-APP stainings were developed

with DAB and nuclei counterstained with hemalum.

(H, I) Multidimensional scaling analysis of spinal cord from mice as in Figure 4J (H). Top 35

Metacore Metabolic Network pathways of mice with EAE, genes with p<0.05 were included for

analysis (I).

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

28

(J-O) EAE disease curve of warm exposed (34°C, started one week before immunization) or room

temperature mice. Arrows indicate individual mice that died or had to be sacrificed, which received

a score of five on the day of death (J). Body weight during the course of EAE (K). Food intake of

two mice per cage within 24 h on day 6 of warm exposure (L) and body weight on day six of warm

exposure (M), before immunization. Survival of mice during EAE (N) and day of EAE onset (O).

(B-F, J-O) Shown is mean ±SEM, two-way ANOVA (J, K); Mantel-Cox (N), student’s t test (B-F,

L-M, O) with mean ±SD, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure S5. (Related to Figure 5)

(A) Body temperature shown is the mean temperature of the two eyes as determined via infrared

images of mice on day 7 of EAE and steady state. Mice were cold exposed for 2 weeks before

and during EAE or steady state. Control mice were at room temperature for the same duration.

(B) Glycemia of mice as in (A) on day 8 of EAE after 3 hours of fasting.

(C-D) Oxygen consumption rate of interscapular brown (BAT, C) and inguinal subcutaneous

adipose tissue (SAT, D) on day 8 of EAE or in steady state after 3 weeks of cold exposure (CE)

and 3 hours of fasting. Technical replicates were used for BAT.

(E) Adipose tissue weights of interscapular brown (BAT, left), inguinal subcutaneous adipose

tissue (SAT, middle) and perigonadal visceral adipose tissue (VAT, right) of mice as in (C).

(F-G) mRNA expression of browning markers relative to B2m and 36b4 in interscapular brown

(BAT, F) and inguinal subcutaneous adipose tissue (SAT, G) of mice as in (B) as determined by

qPCR.

(H) Quantification of lipid droplet size distribution in ingSAT. Slides were analyzed in technical

duplicates from different layers of the tissue and averaged.

(I) Symptoms of EAE (left) and body weight (right) in Ucp1-knockout (KO) and their respective

wildtype (WT) littermate controls housed at room temperature or cold for 1 week before and during

EAE. Shown is 1 representative (n=10-15 per group) of 2 experiments.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

29

(A-G, I) Multiple t-test with Holm-Sidak correction (A-B, F-G), student t-test (C, D), mean ±SD,

two-way ANOVA with mean ±SEM (I), ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Figure S6. (Related to Figure 6)

(A) Gating strategy for the FACS sort after MACS sorting before RNA sequencing (as in Figure

6A). Sorted were DAPI-, CD115+, single cells.

(B) Multidimensional scaling and principal component analysis of FACS sorted blood monocytes

at EAE onset as in Figure 6A.

(C) Gating strategy for MHCII expression on Ly6Chi blood monocytes of mice as in Figure 6A.

Cells were gated as single, Ly6G-, CD115+CD11b+, Ly6Chi, MHCII+.

(D-E) Blood (D) and lymph node (E) immune cells of mice as in Figure 6A on different days after

immunization were visualized using UMAP and clustered using the FlowSOM algorithm in R.

Heatmap shows median relative expression of all panel markers.

(F) Flow cytometry analysis of mesenteric LN (upper panel) or spleen cells (lower panel) of mice

as in Figure 6A at EAE onset. Percentage of cytokine expression in CD4+ T cells. Shown is mean

±SD, significance was calculated using student’s t test, * p<0.05.

(G) Flow cytometry analysis of lymph node dendritic cells and their MFI of MHCII on different days

after immunization.

Figure S7. (Related to Figure 6)

(A-B) Scheme showing experimental setup for cold exposure and EAE (A). Cold exposure either

started two weeks before and continued during EAE (blue, CE) or switched to room temperature

from the day of immunization (violet, CE-RT) and compared to room temperature mice (red, B).

EAE disease curve shown as mean ±SEM, two-way ANOVA, ∗p < 0.05.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

30

(C) Percentage of MHCII expression of Ly6Chi blood monocytes via flow cytometry on day zero,

day three and day seven after EAE induction. Multiple t-test with Holm-Sidak correction with mean

±SD, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

31

REFERENCES

Ajami, B., Bennett, J.L., Krieger, C., McNagny, K.M., and Rossi, F.M. (2011). Infiltrating

monocytes trigger EAE progression, but do not contribute to the resident microglia pool. Nat

Neurosci 14, 1142-1149.

Allan, R.S., Waithman, J., Bedoui, S., Jones, C.M., Villadangos, J.A., Zhan, Y., Lew, A.M.,

Shortman, K., Heath, W.R., and Carbone, F.R. (2006). Migratory dendritic cells transfer antigen

to a lymph node-resident dendritic cell population for efficient CTL priming. Immunity 25, 153-

162.

Buck, M.D., Sowell, R.T., Kaech, S.M., and Pearce, E.L. (2017). Metabolic Instruction of

Immunity. Cell 169, 570-586.

Cannon, B., and Nedergaard, J. (2004). Brown adipose tissue: function and physiological

significance. Physiol Rev 84, 277-359.

Cignarella, F., Cantoni, C., Ghezzi, L., Salter, A., Dorsett, Y., Chen, L., Phillips, D., Weinstock,

G.M., Fontana, L., Cross, A.H., et al. (2018). Intermittent Fasting Confers Protection in CNS

Autoimmunity by Altering the Gut Microbiota. Cell Metab 27, 1222-1235 e1226.

Croxford, A.L., Lanzinger, M., Hartmann, F.J., Schreiner, B., Mair, F., Pelczar, P., Clausen,

B.E., Jung, S., Greter, M., and Becher, B. (2015). The Cytokine GM-CSF Drives the

Inflammatory Signature of CCR2+ Monocytes and Licenses Autoimmunity. Immunity 43, 502-

514.

Davies, L.C., Rosas, M., Jenkins, S.J., Liao, C.T., Scurr, M.J., Brombacher, F., Fraser, D.J.,

Allen, J.E., Jones, S.A., and Taylor, P.R. (2013). Distinct bone marrow-derived and tissue-

resident macrophage lineages proliferate at key stages during inflammation. Nat Commun 4,

1886.

Demas, G., and Nelson, R. (2012). Ecoimmunology (Oxford University Press).

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

32

Dobin, A., Davis, C.A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., Batut, P., Chaisson,

M., and Gingeras, T.R. (2013). STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29,

15-21.

Dranoff, G., Jaffee, E., Lazenby, A., Golumbek, P., Levitsky, H., Brose, K., Jackson, V.,

Hamada, H., Pardoll, D., and Mulligan, R.C. (1993). Vaccination with irradiated tumor cells

engineered to secrete murine granulocyte-macrophage colony-stimulating factor stimulates

potent, specific, and long-lasting anti-tumor immunity. Proc Natl Acad Sci U S A 90, 3539-3543.

Fife, B.T., Huffnagle, G.B., Kuziel, W.A., and Karpus, W.J. (2000). CC chemokine receptor 2 is

critical for induction of experimental autoimmune encephalomyelitis. J Exp Med 192, 899-905.

Foxman, E.F., Storer, J.A., Fitzgerald, M.E., Wasik, B.R., Hou, L., Zhao, H., Turner, P.E., Pyle,

A.M., and Iwasaki, A. (2015). Temperature-dependent innate defense against the common cold

virus limits viral replication at warm temperature in mouse airway cells. Proc Natl Acad Sci U S

A 112, 827-832.

Ganeshan, K., Nikkanen, J., Man, K., Leong, Y.A., Sogawa, Y., Maschek, J.A., Van Ry, T.,

Chagwedera, D.N., Cox, J.E., and Chawla, A. (2019). Energetic Trade-Offs and Hypometabolic

States Promote Disease Tolerance. Cell 177, 399-413 e312.

Ganguly, D., Haak, S., Sisirak, V., and Reizis, B. (2013). The role of dendritic cells in

autoimmunity. Nat Rev Immunol 13, 566-577.

Giladi, A., Wagner, L.K., Li, H., Dorr, D., Medaglia, C., Paul, F., Shemer, A., Jung, S., Yona, S.,

Mack, M., et al. (2020). Cxcl10(+) monocytes define a pathogenic subset in the central nervous

system during autoimmune neuroinflammation. Nat Immunol 21, 525-534.

Grant, R.W., and Dixit, V.D. (2015). Adipose tissue as an immunological organ. Obesity (Silver

Spring) 23, 512-518.

Guilliams, M., Mildner, A., and Yona, S. (2018). Developmental and Functional Heterogeneity of

Monocytes. Immunity 49, 595-613.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

33

Hasan, M., Seo, J.E., Rahaman, K.A., Kang, M.J., Jung, B.H., and Kwon, O.S. (2016).

Increased levels of brain serotonin correlated with MMP-9 activity and IL-4 levels resulted in

severe experimental autoimmune encephalomyelitis (EAE) in obese mice. Neuroscience 319,

168-182.

Hettinger, J., Richards, D.M., Hansson, J., Barra, M.M., Joschko, A.C., Krijgsveld, J., and

Feuerer, M. (2013). Origin of monocytes and macrophages in a committed progenitor. Nat

Immunol 14, 821-830.

Hotamisligil, G.S. (2017). Foundations of Immunometabolism and Implications for Metabolic

Health and Disease. Immunity 47, 406-420.

Ikeda, K., Kang, Q., Yoneshiro, T., Camporez, J.P., Maki, H., Homma, M., Shinoda, K., Chen,

Y., Lu, X., Maretich, P., et al. (2017). UCP1-independent signaling involving SERCA2b-

mediated calcium cycling regulates beige fat thermogenesis and systemic glucose homeostasis.

Nat Med 23, 1454-1465.

Izikson, L., Klein, R.S., Charo, I.F., Weiner, H.L., and Luster, A.D. (2000). Resistance to

experimental autoimmune encephalomyelitis in mice lacking the CC chemokine receptor

(CCR)2. J Exp Med 192, 1075-1080.

Jaakkola, K., Saukkoriipi, A., Jokelainen, J., Juvonen, R., Kauppila, J., Vainio, O., Ziegler, T.,

Ronkko, E., Jaakkola, J.J., Ikaheimo, T.M., et al. (2014). Decline in temperature and humidity

increases the occurrence of influenza in cold climate. Environ Health 13, 22.

Jakubzick, C.V., Randolph, G.J., and Henson, P.M. (2017). Monocyte differentiation and

antigen-presenting functions. Nat Rev Immunol 17, 349-362.

Jastroch, M. (2017). Uncoupling protein 1 controls reactive oxygen species in brown adipose

tissue. Proc Natl Acad Sci U S A 114, 7744-7746.

Jordan, S., Tung, N., Casanova-Acebes, M., Chang, C., Cantoni, C., Zhang, D., Wirtz, T.H.,

Naik, S., Rose, S.A., Brocker, C.N., et al. (2019). Dietary Intake Regulates the Circulating

Inflammatory Monocyte Pool. Cell 178, 1102-1114 e1117.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

34

King, I.L., Dickendesher, T.L., and Segal, B.M. (2009). Circulating Ly-6C+ myeloid precursors

migrate to the CNS and play a pathogenic role during autoimmune demyelinating disease.

Blood 113, 3190-3197.

Ko, H.J., Brady, J.L., Ryg-Cornejo, V., Hansen, D.S., Vremec, D., Shortman, K., Zhan, Y., and

Lew, A.M. (2014). GM-CSF-responsive monocyte-derived dendritic cells are pivotal in Th17

pathogenesis. J Immunol 192, 2202-2209.

Kohlgruber, A.C., LaMarche, N.M., and Lynch, L. (2016). Adipose tissue at the nexus of

systemic and cellular immunometabolism. Semin Immunol 28, 431-440.

Li, C., Spallanzani, R.G., and Mathis, D. (2020). Visceral adipose tissue Tregs and the cells that

nurture them. Immunol Rev 295, 114-125.

Luo, Y., Chen, G.L., Hannemann, N., Ipseiz, N., Kronke, G., Bauerle, T., Munos, L., Wirtz, S.,

Schett, G., and Bozec, A. (2015). Microbiota from Obese Mice Regulate Hematopoietic Stem

Cell Differentiation by Altering the Bone Niche. Cell Metab 22, 886-894.

Man, K., Kutyavin, V.I., and Chawla, A. (2017). Tissue Immunometabolism: Development,

Physiology, and Pathobiology. Cell Metab 25, 11-26.

McCarthy, D.J., Chen, Y., and Smyth, G.K. (2012). Differential expression analysis of multifactor

RNA-Seq experiments with respect to biological variation. Nucleic Acids Res 40, 4288-4297.

McDade, T.W. (2005). Life history, maintenance, and the early origins of immune function. Am J

Hum Biol 17, 81-94.

McInnes, L., Healy, J., and Melville, J. (2018). Umap: Uniform manifold approximation and

projection for dimension reduction. arXiv preprint arXiv:180203426.

Menezes, S., Melandri, D., Anselmi, G., Perchet, T., Loschko, J., Dubrot, J., Patel, R., Gautier,

E.L., Hugues, S., Longhi, M.P., et al. (2016). The Heterogeneity of Ly6C(hi) Monocytes Controls

Their Differentiation into iNOS(+) Macrophages or Monocyte-Derived Dendritic Cells. Immunity

45, 1205-1218.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

35

Merad, M., Sathe, P., Helft, J., Miller, J., and Mortha, A. (2013). The dendritic cell lineage:

ontogeny and function of dendritic cells and their subsets in the steady state and the inflamed

setting. Annu Rev Immunol 31, 563-604.

Mildner, A., Mack, M., Schmidt, H., Bruck, W., Djukic, M., Zabel, M.D., Hille, A., Priller, J., and

Prinz, M. (2009). CCR2+Ly-6Chi monocytes are crucial for the effector phase of autoimmunity

in the central nervous system. Brain 132, 2487-2500.

Mufazalov, I.A., Schelmbauer, C., Regen, T., Kuschmann, J., Wanke, F., Gabriel, L.A.,

Hauptmann, J., Muller, W., Pinteaux, E., Kurschus, F.C., et al. (2017). IL-1 signaling is critical

for expansion but not generation of autoreactive GM-CSF+ Th17 cells. EMBO J 36, 102-115.

O'Neill, L.A., Kishton, R.J., and Rathmell, J. (2016). A guide to immunometabolism for

immunologists. Nat Rev Immunol 16, 553-565.

Okin, D., and Medzhitov, R. (2012). Evolution of inflammatory diseases. Curr Biol 22, R733-740.

Piccio, L., Stark, J.L., and Cross, A.H. (2008). Chronic calorie restriction attenuates

experimental autoimmune encephalomyelitis. J Leukoc Biol 84, 940-948.

Platt, A.M., Rutkowski, J.M., Martel, C., Kuan, E.L., Ivanov, S., Swartz, M.A., and Randolph,

G.J. (2013). Normal dendritic cell mobilization to lymph nodes under conditions of severe

lymphatic hypoplasia. J Immunol 190, 4608-4620.

Reynes, B., van Schothorst, E.M., Keijer, J., Palou, A., and Oliver, P. (2019). Effects of cold

exposure revealed by global transcriptomic analysis in ferret peripheral blood mononuclear

cells. Sci Rep 9, 19985.

Roesler, A., and Kazak, L. (2020). UCP1-independent thermogenesis. Biochem J 477, 709-725.

Ronchi, F., Basso, C., Preite, S., Reboldi, A., Baumjohann, D., Perlini, L., Lanzavecchia, A., and

Sallusto, F. (2016). Experimental priming of encephalitogenic Th1/Th17 cells requires pertussis

toxin-driven IL-1beta production by myeloid cells. Nat Commun 7, 11541.

Serbina, N.V., and Pamer, E.G. (2006). Monocyte emigration from bone marrow during bacterial

infection requires signals mediated by chemokine receptor CCR2. Nat Immunol 7, 311-317.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

36

Stearns, S.C. (1992). The evolution of life histories.

Steinman, R.M., and Banchereau, J. (2007). Taking dendritic cells into medicine. Nature 449,

419-426.

Stojanovic, O., Kieser, S., and Trajkovski, M. (2018). Common traits between the beige fat-

inducing stimuli. Curr Opin Cell Biol 55, 67-73.

Timmermans, S., Bogie, J.F., Vanmierlo, T., Lutjohann, D., Stinissen, P., Hellings, N., and

Hendriks, J.J. (2014). High fat diet exacerbates neuroinflammation in an animal model of

multiple sclerosis by activation of the Renin Angiotensin system. J Neuroimmune Pharmacol 9,

209-217.

Van Gassen, S., Callebaut, B., Van Helden, M.J., Lambrecht, B.N., Demeester, P., Dhaene, T.,

and Saeys, Y. (2015). FlowSOM: Using self-organizing maps for visualization and interpretation

of cytometry data. Cytometry A 87, 636-645.

Wang, A., Luan, H.H., and Medzhitov, R. (2019). An evolutionary perspective on

immunometabolism. Science 363.

Wang, A., and Medzhitov, R. (2019). Counting Calories: The Cost of Inflammation. Cell 177,

223-224.

Weis, S., Carlos, A.R., Moita, M.R., Singh, S., Blankenhaus, B., Cardoso, S., Larsen, R.,

Rebelo, S., Schauble, S., Del Barrio, L., et al. (2017). Metabolic Adaptation Establishes Disease

Tolerance to Sepsis. Cell 169, 1263-1275 e1214.

Winer, S., Paltser, G., Chan, Y., Tsui, H., Engleman, E., Winer, D., and Dosch, H.M. (2009).

Obesity predisposes to Th17 bias. Eur J Immunol 39, 2629-2635.

Yu, G., and He, Q.Y. (2016). ReactomePA: an R/Bioconductor package for reactome pathway

analysis and visualization. Mol Biosyst 12, 477-479.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

37

METHODS

RESOURCES TABLE

REAGENT or RESOURCE

SOURCE

IDENTIFIER

Chemicals, Peptides, Recombinant Proteins

2-Mercaptoethanol

Biorad

1610710

Acetone

Sigma

179124

Brefeldin A

BioLegend

420601

CL316,243

Sigma

C5976

Collagenase A

Roche

11088882001

Collagenase D

Sigma

11088866001

DAB

DAKO

K5001

DAPI

ThermoFisher

D1306

Dibutyl phthalate,99%

Sigma

524980

DMEM

ThermoFisher

41965-039

DNAse

Roche

10104159001

FA Folic Acid solution

Sigma

F8758

Fluorescein isothiocyanate isomer I

Sigma

F7250

Hemalun

Merck

1092490.500

Heparin

Sigma

H3393

hTGF-beta

Peprotech

100-21C

IL-12

Peprotech

210-12

IL-18

MBL (Biozol)

B004-5

IL-2

ebioscience

34-8021-82

IL-23

Miltenyi

130-096-676

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

38

IL-6

Peprotech

216-16

IL-7

Peprotech

217-17

Ionomycin

Sigma

IO634

L-Arginine

Sigma

11009

L-Asparagine

Sigma

A0884

L-Glutamin

Bioconcept

5-10K00H

Lithium Carbonate

Merck

5671.0250

Luxol Fast Blue

BDH laboratories

340443P

Monensin

BioLegend

420701

NEAA 100x

Gibco

11140-035

Penicilin-Streptomycin solution

AMIMED

4-01F-00H

Percoll

GE Healthcare

17-0891-01

Periodoc Acid

Merck

1005240100

PMA

Sigma

P1585

Schiff's reagent

Merck

1.09033.0500

Sodium Pyruvate

Sigma

S8636

Thioglycollate medium

Sigma

T9032

VIT Vitamin mix solution

Sigma

M6985

Critical Commercial Assays

AccuCheck Counting Beads

Molecular Probes

PCB100

BD FACS Lysing Solution

BD

349202

CD4(l3t4) microbeads mouse

Miltenyi

130-117-043

cDNA Reverse Transcription Kit

Applied Biosystems

4368814

Fixation buffer

BioLegend

420801

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

39

FoxP3/Transcription Factor Staining Buffer

Set

eBioscience

00-5523-00

heat inactivated mycobacterium tuberculosis

H3/RA

BD Difco

231141

Incomplete Freund`s adjuvant

BD Difco

263910

Intracellular Staining Perm Wash Buffer

BioLegend

421002

MOG peptide 35-55

Peptides&elephants

N/A

Nextera kit

Illumina

N/A

Pertussis toxin

Sigma/ Calbiochem

P7208/ 516560

RBC Lysis buffer

BioLegend

420301

RNAlater stabilization solution

ThermoFisher

AM7020

RNeasy plus Micro Kit

QIAGEN

74034

RNeasy plus Mini Kit

QIAGEN

74134

PowerUp SYBR Green Master Mix

Applied Biosystems

A25741

SMART-Seq v4 ultra-low input kit

Clontech

N/A

TRIzol reagent

ThermoFisher

15596026

TruSeq stranded mRNA kit

Illumina

N/A

Versacomp beads

Beckman Coulter

B22804

Zombie NIR Fixable Viability Kit

BioLegend

423105

Antibodies

Anti-mouse HRP

DAKO

K4001

Anti-rabbit HRP

DAKO

K4003

Anti-rat HRP

Vectorlab

MP-7404-50

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

40

Armenian hamster anti-CD11c (clone N418)

BioLegend

117328, 117308,

117343, 17309,

117312

Armenian hamster anti-CD11c (clone N418)

eBioscience

103618

Armenian hamster anti-CD3 (clone 145-2C11)

ThermoFisher,

BioLegend

11-0031-82,

100328

Armenian Hamster anti-CD3 (clone 145-

2C11)

BioLegend

100314

Armenian hamster anti-CD49a (clone Ha31/8)

BD

564667

Mouse anti-APP A4 a.a. 66-81 (clone 22C11)

Sigma (formerly Millipore)

MAB348

Mouse anti-CD64 (clone X54-5/7.1)

BioLegend

139308, 139309

Mouse anti-CX3CR1 (clone SA011F11)

BioLegend

149013, 123135

Mouse anti-NK1.1 (clone PK136)

BioLegend

108727

Mouse anti-XCR1 (clone ZET)

BioLegend

562290

Rabbit anti-CD3e (polyclonal)

DAKO

A0452

Rat anti-B220 (clone RA3-6B2)

BioLegend

740262

Rat anti-CCR2 (MC-21)

Provided by Matthias

Mack

N/A

Rat anti-CD103 (clone M290)

BD

564322

Rat anti-CD107b (Mac3) (clone M3/84)

BioLegend

108502

Rat anti-CD115 (clone AFS98)

BioLegend

135505, 135515,

135523

Rat anti-CD117 (clone ACK2, 2B8)

BioLegend

135111, 105827

Rat anti-CD11b (clone M1/70)

BioLegend

101242, 101228,

101222, 101263

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

41

Rat anti-CD11b (clone M1/70)

BD

35011482

Rat anti-CD127 (clone A7R34)

BioLegend

135009

Rat anti-CD135 (clone A2F10)

BioLegend

135310

Rat anti-CD16/32 TruStain FcX (clone 93)

BioLegend

101320

Rat anti-CD19 (clone 6D5)

BioLegend

115534

Rat anti-CD4 (clone GK1.5)

BioLegend

100422

Rat anti-CD45 (clone 30-F11)

BioLegend

103114, 103108

Rat anti-CD8 (clone 53-6.7)

BD

563786

Rat anti-F4/80 (clone BM8)

BioLegend

123114

Rat anti-F4/80 (clone BM8)

BioLegend

107622

Rat anti-GM-CSF (clone MP1-22E9)

ThermoFisher

11733181

Rat anti-IFNg (clone XMG1.2)

BioLegend

508510, 505830

Rat anti-IFNg (XMG1.2)

Bio X Cell

BE0055

Rat anti-IL-17 (clone TC11-18H10.1)

BioLegend

506904

Rat anti-Ly6C (clone HK1.4)

BioLegend

128036, 128037

Rat anti-Ly6G (clone 1A8)

BioLegend

127622, 126615

, 565707

Rat anti-MHCII (clone M5/114.15.2)

BioLegend

107620, 107606,

107631, 107622

Rat anti-mouse CD49d (clone R1-2)

BioLegend

149027

Rat anti-NOS2 (clone CXNFT)

ThermoFisher

53-5920-82

Rat anti-PDCA-1 (clone 927)

BioLegend

127012

Rat anti-Sca1 (clone D7)

BioLegend

108116

Rat anti-Siglec-F (clone E50-2440)

BD

108749

Rat anti-Ter119 (clone Ter119)

BioLegend

116228

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

42

Rat IgG2b isotype

Bio X Cell

BE0090

Sheep anti-Arg1

R&D

IC5868P

Syrian Hamster anti-CD28 (clone 37.51)

BioLegend

102112

Other

Attune NxT

ThermoFisher

N/A

AutoMACS Pro

Miltenyi

N/A

AxioScan Z1

Zeiss

N/A

Bioanalyzer Agilent 2100

Agilent

N/A

BD FACSAria II

BD Biosciences

N/A

BD FACSymphony

BD Biosciences

N/A

BD LSRFortessa

BD Biosciences

N/A

TProfessional Basic Thermocycler

Biometra

N/A

Bio-Rad S3

Bio-Rad

N/A

FLIR E60 thermal camera

FLIR systems

N/A

HiSeq 4000

Illumina

N/A

Light Cycler 480-II

Roche

N/A

MoFlo Astrios

Beckman Coulter

N/A

NanoDrop

ThermoFisher

N/A

Pannoramic 250 FLASH II

3Dhistech

N/A

Seahorse XF24 Analyzer

Agilent

N/A

Tissue Lyser II

QIAGEN

N/A

Experimental Models: Organisms/ Strains

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

43

Mouse: 2D2

The Jackson Laboratory

006912

Mouse: C57BL/6

Charles River, Janvier

Mouse: Ccr2-KO

The Jackson Laboratory

004999

Mouse: Ucp1-KO

The Jackson Laboratory

003124

Experimental Models: Cell Lines

B16-GMCSF cells

Dranoff et al., 1993

N/A

Oligonucleotides

Name

Forward Primer

Reverse Primer

M. musculus

Ucp1

5′-GGCCTCTACGACTCAGTCCA-3′

5′-TAAGCCGGCTGAGATCTTGT-3′

M. musculus

Cidea

5′-TGCTCTTCTGTATCGCCCAGT-3′

5′-GCCGTGTTAAGGAATCTGCTG-3′

M. musculus

Ppargc1a

5′-

AAGTGTGGAACTCTCTGGAACTG-

3′

5′-

GGGTTATCTTGGTTGGCTTTATG-3′

M. musculus

Ppara

5′-AGAGCCCCATCTGTCCTCTC-3′

5′-ACTGGTAGTCTGCAAAACCAAA-

3′

M. musculus

Pparg

5′-AGGCGAGGGCGATCTTGACAG-

3′

5′-AATTCGGATGGCCACCTCTTTG-

3′

M. musculus

Prdm16

5′-CAGCACGGTGAAGCCATT-3′

5′-GCGTGCATCCGCTTGTG-3′

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

44

M. musculus

36B4

5′-

GTACCCATTGATGATGGAGTGT-3′

5′-TGAAGTGCTCGACATCACAG-3′

M. musculus

B2m

5′-TTGTCTCACTGACCGGCCT-3′

5′-TATGTTCGGCTTCCCATTCTCC-

3′

Software and Algorithms

FastQC v.0.11.5

https://www.bioinformatics.babraham.a

c.uk/projects/fastqc/

N/A

FLIR Tools+ software

FLIR technologies

FlowJo 10.2

TreeStar

N/A

GraphPad 8

Prism

N/A

HTSeq v0.9.1

https://www-

huber.embl.de/users/anders/HTSeq/

N/A

MetaCore

Clarivate Analytics

N/A

Pannoramic Viewer software

3DHistech

N/A

Picard tools v.1.141

https://picard.sourceforge.net

N/A

QuPath v0.2.2

QuPath

N/A

R language and environment for statistical

computing and graphics

https://www.r-project.org

N/A

R/Bioconductor package EdgeR v. 3.4.2

(McCarthy et al., 2012)

N/A

R/Bioconductor package topGO

https://bioconductor.org/packages/rele

ase/bioc/vignettes/topGO/inst/doc/topG

O.pdf

N/A

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

45

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Mice

Male 8 weeks old C57BL/6J mice were obtained from Charles River, or Janvier (France). Ucp1-

KO (The Jackson Laboratory), their littermate controls, Ccr2-KO (The Jackson Laboratory) and

2D2 (The Jackson Laboratory) mice were bred in house. Mice were housed in a specific pathogen

free (SPF) facility in 12 h day/night cycles, fed irradiated standard chow diet and given water from

autoclaved bottles. Cold exposure was applied in a light- and humidity-controlled climatic chamber

(TSE, Germany) under SPF conditions, at 10 °C for 2 weeks before immunization with an

additional initial acclimatization period of 5 days at 18°C and 5 days at 14°C, as well as during

EAE. Warm exposure of 34°C was applied for 7 days before immunization, and during EAE. All

climate chamber experiments were performed with 2 mice per cage without bedding and nesting

material for enrichment to ensure controlled temperature conditions. CL316,243 (Sigma) was i.p.

injected daily 7 days before immunization and until sacrifice. Body temperature was determined

with infrared camera FLIR E60 (FLIR, UK) taking the mean temperature of both eyes using the

FLIR Tools+ software. All animal experiments were performed at the University of Geneva with

authorization by the responsible Cantonal authority and in accordance with the Swiss law for

animal protection.

ReactomePA R-package

(Yu and He, 2016)

N/A

Rgraphviz

https://www.bioconductor.org/package

s/release/bioc/html/Rgraphviz.html

N/A

STAR aligner v.2.6.0c

(Dobin et al., 2013)

N/A

Seahorse Wave Desktop Software

Agilent

N/A

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

46

METHODS DETAILS

EAE induction and scoring

For active EAE immunization, complete Freund`s adjuvant was prepared by mixing 10mg/ml heat

inactivated mycobacterium tuberculosis H3/RA (Difco) with incomplete Freund`s adjuvant (Difco).

To create an emulsion, this was mixed 1:1 with MOG peptide 35-55 (4 mg/ml, peptides &

elephants). Mice were anesthetized and 50 ul of the emulsion was injected into the left and right

flank each (200 ug peptide/mouse), followed by i.v. injection of 300 or 400 ng pertussis toxin

(Sigma or Calbiochem). After two days, mice received additional 300 or 400 ng pertussis toxin

i.p.

For Th1 adoptive transfer EAE, splenocytes of 2D2 T cell receptor transgenic mice were cultured

for 2 days with MOG35-55 peptide (20 µg/ml), IL-2 (5 ng/ml, ebioscience) and IL-7 (5 ng/ml,

Peprotech), 4 days with IL-2 (5 ng/ml) and IL-7 (5 ng/ml) and 1 day with anti-CD3/28 (0.5 ng/ul,

Biolegend), IL-12 (20 ng/ml, Peprotech) and IL-18 (25 ng/ml, MBL (Biozol)) at 37°C. Per mouse,

1-2 million cells were transferred (i.p.) and animals were injected with 67 ng pertussis toxin on

day 0 (i.v.) and day 2 (i.p.) of EAE. Congenic animals (CD45.1 and CD45.2) were used to later

distinguish endogenous and transferred T cells.

For Th17 adoptive transfer EAE, wildtype donor mice were actively immunized as described

above and after 7 days their splenocytes lysed, cultured together with cells from dLNs for 3 days

with IL-6 (5 ng/ml, Peprotech), hTGF-beta (0.25 ng/ml, Peprotech), anti-IFNg (14 ug/ml, Bio X

Cell), IL-23 (6.5 ng/ml, Miltenyi) and MOG peptide (35 ug/ml) and sorted via positive selection of

CD4+ T cells (MACS, Miltenyi). Per recipient, 3 million CD4+ T cells were transferred by iv. injection

and animals were injected i.p. with pertussis toxin on day 0 and day 2 of EAE (300 ng).

Starting at the disease onset, approx. 8-10 days after induction of EAE, disease severity was

monitored using an established scale from 0 to 4 to describe clinical symptoms: 0, no detectable

signs of classical EAE; 0.5, paresis of distal limb tail; 1.0, tail paralysis; 1.5, hind-limb weakness;

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

47

2, hind-limb paresis; 2.5, partial hind-limb paralysis; 3, complete bilateral hind-limb paralysis; 3.5,

ascending paralysis to torso; 4, paralysis of forelimbs and hind limbs. Mice with scores >3.5 were

sacrificed. Dead animals were excluded or in warm exposure experiment given a score of 5 on

death day due to increased deaths and high scores. Animals that reach a score >1.5 were

provided with their chow diet, watered, inside the cage.

B16-GMCSF tumor and thioglycolate elicited peritoneal inflammation

GMCSF-expressing B16 melanoma tumor cells (Dranoff et al., 1993) were cultured and s.c.

engrafted as described (Menezes et al., 2016) into cold exposed and room temperature mice.

Blood was drawn on several days after tumor implantation in heparin (Sigma) to analyze

circulating monocyte expansion and phenotype using flow cytometry. For peritoneal inflammation,

cold-exposed and room temperature mice were i.p. injected with 1ml of 4% thioglycolate (Sigma),

as described before (Davies et al., 2013). Animals were sacrificed after 24h for analysis of

peritoneal monocytes and monocyte derived cells using flow cytometry.

FITC migration of DCs using skin painting of the flank

Mice were anesthetized with isoflourane, a small area of one flank (about 1x1cm) was shaved

and 20 ul of FITC solution was applied, consisting of 1% FITC (Sigma-Aldrich: Fluorescein

isothiocyanate isomer I) in Aceton/Dibutylphthalate (1:1, Sigma). After 12h and 18h, the 3 draining

lymph nodes (axillary, brachial and inguinal) of ipsi- and contralateral side were harvested and

cells harvested as described above.

Antibody treatment for CCR2+ monocyte depletion

Monocytes were depleted by daily i.p. injections of anti-CCR2 mAb (clone: MC-21; 40 μg), starting

one day before immunization on 8 consecutive days until sacrifice. Control mice received isotype

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

48

rat IgG2b (Bio X Cell) at the same dose. Depletion efficiency was tested 2 days before EAE onset

and sacrifice.

Histology

For histologic analyses of spinal cord, mice were kept until 1-2 days after peak, perfused with

PBS and subsequently 4% paraformaldehyde (PFA) and additionally immersion fixed in 4% PFA.

Sectioning, embedding and staining were performed as previously described (Steinbach et al.,

2016). Mice shown for histologic evaluation were control mice of an experiment and therefore

injected with diphtheria toxin (Sigma). Stainings were developed with DAB (DAKO) and nuclei

counterstained with hemalum (Merck). Sections were scanned using Pannoramic Digital Slide

Scanner 250 FLASH II (3DHISTECH). Quantifications were performed blinded to the

experimental condition. Anti-CD3 (polyclonal, DAKO) and anti-mac3 (M3/84, Biolegend) stainings

were automatically quantified using Definiens Developer software with a custom-made script.

Quantifications of Luxol Fast Blue (BDH laboratories)-Perdiodic Acid (Merck)-Schiff’s (Merck)

(LFB-Pas) lesion areas, anti-APP (22C11, Millipore) and lesion number were performed manually

using Pannoramic Viewer software (3DHISTECH). Number of lesions were counted in APP

staining and only lesions were taken into account that were found in LFB-Pas as well. APP was

quantified within lesions marked in LFB-Pas staining and reported per total white matter area. For

representative LFB-Pas images, contrast was equally adjusted in the compared conditions in

Adobe Photoshop.

For histologic analyses of adipose tissues, tissues were fixed in 4% paraformaldehyde (Sigma-

Aldrich), paraffin embedded, cut in sections (5 μm) and stained with hematoxylin-eosin (H&E).

Images were acquired using AxioScan Z1 (Zeiss) and the cell/lipid droplet size quantified

with Qupath software (V0.2.2) using machine-learning classifiers.

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

49

Immune cell isolation from tissues

Peripheral blood samples (10-30 ul) were obtained by facial vein puncture in 1 ul heparin (10U/ul,

Sigma). Blood erythrocytes were lysed and cells fixed using BD FACS Lysing Solution. For other

tissues that required red blood cell lysis, RBC Lysis buffer (Biolegend) was used. For the isolation

of immune cells from the CNS, mice were anesthetized and transcardially perfused with PBS.

Brain and spinal cord were minced, digested in DMEM (ThermoFisher) with Collagenase A

(1mg/ml, Sigma) and DNaseI (0.1 mg/ml, Roche) for 1h at 37°C and homogenized using 70-μm

cell strainers (BD). Leukocytes were separated using a discontinuous Percoll (GE healthcare)

gradient (30% / 70%, 2500 RPM, 30 min, 4°C). Bone marrow cells were flushed from 1 femur per

mouse. For analysis of DC and monocyte populations, axillary and brachial lymph nodes or spleen

were cut into small pieces and digested in DMEM with collagenase D (1mg/ml, Roche) and

DNaseI (10ug/ml, Roche) for 30-35 min at 37°C under constant pipetting. For studies of all other

immune cells, lymph nodes or spleens were mashed through a 70 um cell strainer. For

splenocytes, red blood cell lysis was performed (BioLegend).

Flow cytometry

Dead cells were excluded from the analysis using Zombie NIR or Aqua Fixable Viability kit

(BioLegend) for all tissues except blood. Surface staining was performed with directly labeled

antibodies in FACS buffer (DPBS, 2.5% FCS, 10mM EDTA). Compensation controls were set up

using Versacomp beads (Beckman Coulter). Antibodies were from BioLegend if not stated

otherwise. For staining of bone marrow progenitors and mature monocytes, the following

antibodies were used: Lin (CD3 (145-2C11), CD19 (6D5), NK1.1 (PK136), Ly6G (1A8), Ter-119),

CD127 (A7R34), CD117 (ACK2, 2B8), CD115 (AFS98), CD135 (A2F10), Ly6C (HK1.4), CD11b

(M1/70), Sca1 (D7), CD11c (N418), MHCII (M5/114.15.2). In figures, monocytes and monocyte-

derived cells were abbreviated as “M” for simplicity reasons. Both Ly6Chi monocytes and

monocyte derived cells were defined mainly by their Ly6C expression (in addition to and exclusion

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

50

of other markers), but irrespective of a co-expression of CD11c or pro/anti-inflammatory

cytokines. “M” therefore comprises both, undifferentiated Ly6Chi monocytes and monocytes that

already gained APC or macrophage like functions but still express Ly6C. For characterization of

blood and bone marrow monocytes, the following antibodies were used: CD45 (30-F11), Ly6G,

CD11b, CD115, Ly6C, MHCII. For characterization of T cells in various tissues, the following

antibodies were used: CD45, CD11b, Ly6G, CD4 (GK1.5), CD3, CD8 (53-6.7, BD), GM-CSF

(MP1-22E9, ThermoFisher), IL-17 (TC11-18H10.1), IFNg (XMG1.2). Unspecific binding to FC

receptors was blocked using anti-CD16/32 (BioLegend). Intracellular stainings were performed

according to manufacturer’s instructions: Fixation/Intracellular Staining Permeabilization Wash

Buffer (BioLegend) and Fixation buffer (BioLegend) were used for Arg1 (IC5868P, R&D), iNOS

(CXNFT), IL-17, IFNg and GM-CSF. To assess for intracellular cytokine production of T cells,

leukocytes were cultured for 4 h in the presence of PMA (end 50ng/ml, Sigma) and Ionomycin

(end 1ug/ml) (Sigma), monensin (1:1000, BioLegend), and brefeldin A (1:1000, BioLegend).

Absolute cell numbers were quantified using AccuCheck Counting Beads (Invitrogen). Flow

cytometric samples were acquired on an Attune NxT (ThermoFisher) or LSRFortessa (BD) flow

cytometer using compensation controls. FlowJo software (Treestar, Version 10) was used for

analysis. Further antibodies used in DC analysis or high dimensional analysis, that were not yet

mentioned before are CD49a (Ha31/8, BD), CD64 (X54-5/7.1), CX3CR1 (SA011F11), XCR1

(ZET), B220 (RA3-6B2), CD103 (M290, BD), F4/80 (BM8), CD49d (R1-2), PDCA-1 (927), Siglec-

F (E50-2440, BD). High-dimensional flow cytometry analysis was performed using BD

FACSymphony. Raw data was pre-processed using FlowJo followed by transformation, percentile

normalization in R, dimensionality reduction and visualization by Uniform Manifold Approximation

and Projection (UMAP) (McInnes et al., 2018). The FlowSOM algorithm was used for automated

clustering (Van Gassen et al., 2015) using UMAP with overlaid mean marker expression values

and a heatmap of median expression values (Hartmann et al., 2016; Mrdjen et al., 2018). Data

shown in Figure 2C is the same as in Figure 5E but was visualized together with data from other

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

51

time points to highlight shifts of populations over the time in Figure 5E.

Oxygen Consumption Rate (OCR). OCR was determined as before (Dunham-Snary et al.,

2014), in 3.2-8mg punches (2.5 mm) of adipose tissue. After washing in wash buffer (25mM

HEPES, 1% fatty acid-free BSA, 2.5 mM glucose, in salt solution: 4mM KCl, 115 mM NaCl, 1 mM

MgCl2, 1.5 mM KH2PO4, 2 mM MgSO4, 1.5 mM CaCl2 at pH 7.4), tissue punches were placed

in islet capture plates (model XF24, Seahorse Bioscience) and washed with wash and

subsequently with respiration buffer (0.1% fatty acid-free BSA, 10 mM glucose, 2 mM L-glutamine,

0.5 mM L-carnitine, 1 mM sodium pyruvate in salt solution at pH 7.4), in which they were further

equilibrated for 50 or 60 min at 37°C. Samples were analyzed using a Seahorse XF24 Analyzer

(Seahorse Bioscience) and Wave software with 3 minutes mix; 2 minutes rest; 3 minutes

measurements in basal condition.

RNA Extraction for RNA sequencing and qPCR

After preparation from mice, spinal cord and adipose tissues were stored at -80°C until further

processed. For RNA isolation, frozen tissues were mechanically homogenized in 1 ml Trizol

(Thermo Fisher Scientific) with 1 stainless steel bead (5 mm) per sample using the TissueLyserII

(Qiagen) by shaking for 50 s at 30 Hz. 200 μL chloroform was added to homogenized samples in

trizol, followed by shaking for 15 s and centrifugation (15 min, 12000 RCF, 4°C). The chloroform

phase was collected, shaken for 15 s with 500 μL isopropanol and centrifuged (15 min, 12000

RCF, 4°C). The pellet obtained was washed twice with 70% ethanol (10 min, 8000 RCF, 4°C) and

resuspended in 50 μL PCR-grade water. Bone marrows were flushed right after sectioning from

mouse, cells centrifuged, loaded onto shredder columns (Qiagen). Shredded bone marrow cells

or FACS sorted cells were frozen (-80°C) in RLT buffer with 1% beta-mercaptoethanol until RNA

extraction via RNA mini kit or micro kit (Qiagen), respectively. For Sequencing, RNA integrity

number (RIN) was determined (Bioanalyzer 2100, Agilent). For qPCR, RNA concentration and

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

52

purity were determined (NanoDrop, Thermo Fisher Scientific) and 1 μg RNA was used per sample

for retrotranscription using the High-Capacity cDNA Reverse Transcription Kit (Applied

Biosystems). Expression of mRNA was measured by qPCR on a LightCycler 480 II (Roche) with

PowerUp SYBR Green Master Mix (Applied Biosystems). Relative mRNA expression was

normalized to at least two house keeping genes and to room temperature control mice.

Blood monocyte purification for RNA sequencing

Blood was taken retro-orbital under terminal anesthesia in polypropylen facs tubes containing 7ul

heparin (10U/ul) and coated with 100 IU/mL heparin. Anti-CD115-PE was directly added and

MACS separation was performed using Anti-PE MicroBeads (Miltenyi) and AutoMACS Pro

Seperator (Miltenyi). Immediately after, monocytes were sorted as CD115+, DAPI-

(ThermoFisher) cells (FACSAriaII, BD, Bio-Rad S3 or MoFlo Astrios, Beckman Coulter).

RNAseq sequencing and analysis

mRNA sequencing was performed at the iGE3 Genomics Platform at the CMU of the University

of Geneva. Libraries for sequencing were prepared with the TruSeq stranded mRNA kit (Illumina)

for bone marrow and spinal cord. The SMART-Seq v4 ultra-low input kit (Clontech) combined with

the Nextera kit (Illumina) were used respectively for cDNA amplification and library preparation

for monocytes (steady state and EAE). Bone marrow, spinal cord, and monocytes (EAE) were

sequenced with read length SR50 and steady state monocytes with read length SR100 (Illumina

HiSeq 4000). The sequencing quality control was done with FastQC v.0.11.5

(http://www.bioinformatics.babraham.ac.uk/projects/fastqc/). The reads were mapped

with STAR aligner (Dobin et al., 2013) to the UCSC Mus musculus mm10 reference. The

transcriptome metrics were evaluated with the Picard tools v.1.141

(http://picard.sourceforge.net/). The table of counts with the number of reads mapping to each

gene feature of the Mus musculus mm10 reference was prepared with HTSeq v0.9.1 (htseq-

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

53

count, http://www-huber.embl.de/users/anders/HTSeq/). Raw counts were then further processed

and analyzed by R/Bioconductor package EdgeR v. 3.4.2 (McCarthy et al., 2012a), for differential

expression analysis. The counts were normalized according to the library size and filtered. Only

genes having log count per million reads (cpm)>0 were kept for further analyses. After

normalization of the counts, transcript abundances were compared in pairwise conditions in a

modified Fischer exact test (as implemented in edgeR). The differentially expressed genes p-

values were corrected for multiple testing error with a 5% FDR (false discovery rate). The

correction used is Benjamini-Hochberg (BH). Transcripts with log(cpm)>0 and p≤0.05 were

subsequently subjected to pathway analysis. For the relativeness and reactome analysis, genes

were called significantly differentially expressed between any given two conditions when p<0.05

and FC>2. Reactome pathways database was used with ReactomePA R-package (Yu and He,

2016), reporting the enrichment ratio (# DEGs/Total Genes in the dataset) and FDR-adjusted

Pvalue computed by Fisher exact test. For gene ontology analysis, R/Bioconductor package

topGO (https://bioconductor.org/packages/release/bioc/vignettes/topGO/inst/doc/topGO.pdf),

together with Rgraphviz was used. MetaCore (Clarivate Analytics) Pathway Maps, Gene Ontology

and Metabolic Networks were used.

Statistical analysis

To assess significant differences, GraphPad 8.0 (Prism) was used. For two groups, two-tailed

Student’s t test was used. For more than two groups, multiple t-tests were applied with Holm-

Sidak correction. Two-waxy ANOVA was used to compare two variables between groups. Mantel-

Cox was applied to test the incidence or survival distributions of two groups. P < 0.05 was

considered significant (*), p < 0.01 (**), p < 0.001 (***).

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

Figure 1

A

D

RT CE

0.0

0.1

0.2

0.3

% Ly6Cint M of alive

RT CE

0

2

4

6

% MHCII of Ly6Chi M

***

RT CE

0

1

2

3

4

%MHCII+Ly6Chi of MHCII+

**

RT CE

0

10

20

30

40

50

%MHCII+ of CD45+

RT CE

0.00

0.05

0.10

0.15

0.20

% CDP

*

RT CE

0.00

0.02

0.04

0.06

0.08

0.10

% MDP

*

RT CE

0.00

0.01

0.02

0.03

0.04

0.05

% CLP

RT CE

0.00

0.05

0.10

0.15

0.20

0.25

% LSK

E

F

B

C

G

RT CE

0.0

0.2

0.4

0.6

0.8

1.0

% Ly6Clo M of alive

RT CE

0

2

4

6

8

% Ly6Chi M of alive

*

RT CE

0.10

0.15

0.20

0.25

% cMoP

p=0.095

PBS CL316

0.00

0.05

0.10

0.15

% cMoP

*

PBS CL316

0.00

0.02

0.04

0.06

0.08

% CLP

PBS CL316

0.0

0.2

0.4

0.6

% LSK

PBS CL316

0.00

0.05

0.10

0.15

0.20

% CDP

*

PBS CL316

0

1

2

3

4

%Ly6Chi M of alive

*

0 50 100 150

Myeloid leucocyte activation

Export from cell

Secretion

Cell activation

Immune system process

-log(p-value)

GO Processes

PBS CL316

0.00

0.02

0.04

0.06

0.08

% MDP

p=0.1

RT1

RT2

RT3

RT4

RT5

RT6

CE1

CE2

CE3

CE4

CE5

CE6

B2m

CD80

Csf1

Csf1r

Csf2ra

Csf2rb

Cxcl10

Ern1

Gas6

H2-D1

H2-K1

H2-Q10

Hgf

Hif1a

Ifngr1

Il12a

Il12rb1

Il12rb2

Il13ra1

Il1b

Il1r1

Il4

Il4ra

Inhba

Irf4

Irf5

Itgam

Itgb2

Jag1

Kdm6b

Lilrb4

Mapk13

Mapk14

Mertk

Msr1

Notch1

Pik3r3

Postn

Ptger2

Ptpn6

Rbpj

Ripk1

Socs3

Stat1

Stat3

Stat6

Tgfbr1

Thbs1

Tlr2

Tlr4

Tlr7

Tnf

Tnfrsf1a

Tnfsf10

Xbp1

Macrophage and DC shift

-2

0

2

RT1

RT2

RT3

RT4

RT5

RT6

CE1

CE2

CE3

CE4

CE5

CE6

Adar

B2m

Bcl2

Casp-7

Cbl

Cxcl10

Cxcl9

Eif2ak1

Eif2ak2

Gbp3

Gbp4

H2-D1

H2-K1

H2-Q10

HIF1A

Ifit3

Irf1

Irf7

Irf9

Isg20

Mx1

Ncoa7

Oas1a

Oas1b

Oas1g

Socs3

Stat1

Stat2

Stat3

Stat4

Stat6

Tap1

Tnfsf10

Usp18

Xaf1

IFN-alpha/beta signaling

-2

0

2

RT1

RT2

RT3

RT4

RT5

RT6

CE1

CE2

CE3

CE4

CE5

CE6

Ccr2

Cebpa

Cebpe

Csf3r

Gfi1

Itgam

Itgb2

Lcn2

Ltf

Map2k1

Mapk3

Mpo

Myc

Prtn3

Ptpn11

Ptpn6

Rasgrp1

Rb1

Sfpi1

Socs3

Stat3

G-CSF-induced myeloid differentiation

-2

-1

0

1

2

8430408G22Rik

Apol11a

C130026I21Rik Cd209f

Col1a1

Ctsk

Oasg1

Cyr61

Hspa1a

Ifitm1

Marco

Upp1

0

5

10

Log2

fold change

-Log10 P

Hspa1b

Col1a1

Vsig4

Rspo2

Mmp13

Cxcr1

Irg1

Oas2

NS

p-value

2

p-value and log FC

-1 1

02

The copyright holder for this preprintthis version posted August 30, 2020. ; https://doi.org/10.1101/2020.08.28.269563doi: bioRxiv preprint

Figure 2

A

RT CE

0

1

2

3

4

% Ly6Clo M of CD45+

0

5

10

15

20

% MHCII of Ly6Chi M

***

RT CE

0.0

0.5

1.0

1.5

2.0

% Ly6Cint M of CD45+

RT CE

0

20

40

60

80

% MHCII+ of CD45+

B

D

E

RT CE

0

5

10

% Ly6Chi M of CD45+

PBS CL316

0

5

10

15

% MHCII+ of Ly6Chi M

*

PBS CL316

0.0

0.5

1.0

1.5

2.0

2.5

%MHCII+Ly6Chi of MHCII+

*

0 2 4 6 8

Positive regulation of macrophage chemotaxis

Ureteric bud development

T cell activation

Regulation of fever generation

Positive regulation of heat generation

Antigen processing, presentation of peptide or polysaccharide antigen via MHCII

T cell costimulation

Regulation of cytosolic Ca2+ in PLC-activating G protein-coupled signaling

Negative regulation of interleukin-6 secretion

Response to bacterium

Regulation of B cell proliferation

Neutrophil chemotaxis

Leukocyte migration involved in inflammatory response

Adaptive immune response

Positive regulation of cytosolic calcium ion concentration

-log(p-adjust.)

GO BP

PBS CL316

0

20

40

60

80

% MHCII+ of CD45+

RT CE

0

1

2

Cold Exposure Protects from Neuroinflammation Through Immunologic Reprogramming

Abstract