This Accepted Manuscript has not been copyedited and formatted. The final version may differ from this version.
Research Articles: Cellular/Molecular
Regional distribution of CNS antigens differentially determines T-cell mediated neuroinflammation in a CX3CR1 dependent manner
Aditya Rayasam1, Julie A. Kijak2, McKenna Dallmann2, Martin Hsu1, Nicole Zindl2, Anders Lindstedt2, Leah Steinmetz2, Jeffrey S. Harding3, Melissa G. Harris1, Jozsef Karman4, Matyas Sandor2,3 and Zsuzsanna Fabry2,3
1Department of Pathology and Laboratory Medicine, University of Wisconsin-Madison, Madison, WI, Neuroscience Training Program, University of Wisconsin-Madison, Madison, WI 2Department of Pathology and Laboratory Medicine, University of Wisconsin-Madison, Madison, WI 3Department of Pathology and Laboratory Medicine, University of Wisconsin-Madison, Madison, WI, Cellular and Molecular Pathology Graduate Program, University of Wisconsin-Madison, Madison, WI 4Department of Pathology and Laboratory Medicine, University of Wisconsin-Madison, Madison, WI, Genzyme Corporation, Cambridge, MA
DOI: 10.1523/JNEUROSCI.0366-18.2018
Received: 8 February 2018
Revised: 15 May 2018
Accepted: 18 May 2018
Published: 29 June 2018
Author contributions: A.R., J.S.H., M.G.H., J. Karman, M.S., and Z.F. designed research; A.R., J. Kijak, M.D., M.H., N.Z., A.L., and L.S. performed research; A.R. and M.H. analyzed data; A.R. wrote the first draft of the paper; A.R., J. Kijak, M.H., J.S.H., M.G.H., and Z.F. edited the paper; A.R. wrote the paper.
Conflict of Interest: The authors declare no competing financial interests.
Corresponding Author: Zsuzsanna Fabry, Email: [email protected], Phone: 608-265-8716
Cite as: J. Neurosci ; 10.1523/JNEUROSCI.0366-18.2018
Alerts: Sign up at www.jneurosci.org/cgi/alerts to receive customized email alerts when the fully formatted version of this article is published.
Accepted manuscripts are peer-reviewed but have not been through the copyediting, formatting, or proofreading process.
Copyright © 2018 the authors
1 Regional distribution of CNS antigens differentially determines T-cell 2 mediated neuroinflammation in a CX3CR1 dependent manner. 3 4 Aditya Rayasam1, Julie A. Kijak2, McKenna Dallmann2, Martin Hsu1, Nicole 5 Zindl2, Anders Lindstedt2, Leah Steinmetz2, Jeffrey S. Harding3, Melissa G. 6 Harris1, Jozsef Karman4, Matyas Sandor2,3, Zsuzsanna Fabry2,3* 7 8 1[1] Department of Pathology and Laboratory Medicine, University of Wisconsin-Madison, 9 Madison, WI [2] Neuroscience Training Program, University of Wisconsin-Madison, Madison, WI. 10 2[1] Department of Pathology and Laboratory Medicine, University of Wisconsin-Madison, 11 Madison, WI. 12 3[1] Department of Pathology and Laboratory Medicine, University of Wisconsin-Madison, 13 Madison, WI [2] Cellular and Molecular Pathology Graduate Program, University of Wisconsin- 14 Madison, Madison, WI. 15 4[1] Department of Pathology and Laboratory Medicine, University of Wisconsin-Madison, 16 Madison, WI [2] Genzyme Corporation, Cambridge, MA. 17
18 Corresponding Author: 19 20 Zsuzsanna Fabry* 21 Email: [email protected] 22 Phone: 608-265-8716 23 24 Author Contributions
25 A.R. and J.A.K. performed adoptive transfers and induced EAE. A.R. analyzed
26 immune responses and histological data. J.A.K., M.D., M.H., N.Z., A.L., and L.S.
27 helped perform immunohistochemistry and analyses, assisted with murine organ
28 harvests, as well as scored and weighed EAE mice. J.K. generated the construct.
29 J.S.H., M.G.H., M.H., Z.F., and M.S. provided critical conceptual guidance,
30 comments and reviews. A.R., M.S., and Z.F. wrote the main manuscript text.
31 A.R. and M.H. prepared figures. All authors reviewed the manuscript.
32 Competing interests
33 The authors declare no competing financial interests.
34 35 36 37
1
38 Abstract 39 40 T cells continuously sample central nervous system (CNS) derived antigens in
41 the periphery, yet it is unknown how they sample and respond to CNS antigens
42 derived from distinct brain areas. We expressed ovalbumin (OVA) neoepitopes in
43 regionally distinct CNS areas (Cnp-OVA and Nes-OVA mice) to test peripheral
44 antigen sampling by OVA-specific T cells under homeostatic and
45 neuroinflammatory conditions. We show that antigen sampling in the periphery is
46 independent of regional origin of CNS antigens in both male and female mice.
47 However, experimental autoimmune encephalomyelitis (EAE) is differentially
48 influenced in Cnp-OVA and Nes-OVA female mice. While there is the same
49 frequency of CD45high CD11b+ CD11c+ CX3CL1+ myeloid cell - T cell clusters in
50 neoepitope expressing areas, EAE is inhibited in Nes-OVA female mice and
51 accelerated in CNP-OVA female mice. Accumulation of OVA-specific T cells and
52 their immunomodulatory effects on EAE are CX3CR1 dependent. These data
53 show that despite similar levels of peripheral antigen sampling, CNS antigen
54 specific T cells differentially influence neuroinflammatory disease depending on
55 the location of cognate antigens and the presence of CX3CL1/CX3CR1
56 signaling.
57 58 Significance Statement
59 Our data show that peripheral T cells similarly recognize neoepitopes
60 independent of their origin within the CNS under homeostatic conditions.
61 Contrastingly, during ongoing autoimmune neuroinflammation, neoepitope
62 specific T cells differentially influence clinical score and pathology based on the
2
63 CNS regional location of the neoepitopes in a CX3CR1 dependent manner.
64 Altogether, we propose a novel mechanism for how T cells respond to regionally
65 distinct CNS derived antigens and contribute to CNS autoimmune pathology.
66 Introduction
67 Understanding how the central nervous system (CNS) and the immune
68 system communicate during health and disease is critical for elucidating how
69 CNS diseases initiate, progress, and resolve. It is becoming increasingly clear
70 that peripheral immune cells can recognize and respond to CNS derived antigens
71 under homeostatic conditions and even more so during neurotrauma and
72 neurological disease 1,2. In spite of this, the routes and mechanisms for how
73 different CNS cell-derived antigenic information reaches the periphery and how
74 antigen specific T cells migrate to different regions in the CNS and influence
75 resident cells are still not understood.
76 Accumulating data suggests that Multiple sclerosis (MS) patients can display
77 unique immune cell auto-reactivity to single or multiple distinct CNS antigens 3,4.
78 It is unclear how and why the nature of the autoreactive antigen, the cells that
79 express them, and/or their regional location within the CNS, influence the
80 immune responses targeted towards them and how these processes ultimately
81 affect autoimmune outcome 5-9. Nonetheless, dendritic cells (DCs) have been
82 identified to be crucial in these processes as they are essential in initiating and
83 facilitating how peripheral immune responses interact with CNS antigens 10.
84 In order to understand how the peripheral immune system responds to
85 distinct CNS cell derived and regionally located antigens, we generated
3
86 transgenic mice using the oligodendrocyte specific (2’,3’-Cyclic-nucleotide 3'-
87 phosphodiesterase (CNPase)) promoter and the neural progenitor cell restricted
88 Nestin promoter to express an enhanced green fluorescent protein (EGFP)-
89 tagged fusion protein containing ovalbumin (OVA) antigenic peptides in
90 CNPase+ and Nestin+ cells respectively (Cnp-OVA and Nes-OVA mice) 11. Using
91 these models, we tested whether anti-OVA peptide-specific sentinel OT-I (CD8+)
92 and OT-II (CD4+) T cells would similarly sample OVA antigens from different
93 CNS regions in the periphery under homeostatic or neuroinflammatory
94 conditions.
95 CX3C chemokine receptor 1 (CX3CR1) has been shown to be expressed on
96 CD4 T cells residing in the lung and skin 12,13 and more recently CX3CR1 was
97 shown to differentiate between subsets of memory CD8 T cells 14 and non-
98 lymphoid tissue surveying peripheral memory T cells during lymphocytic
99 choriomeningitis viral infection (LCMV) 15. While the role of CX3CR1 in effector
100 CD8 T cell localization has been described in different tissues including the lung,
101 skin and spleen 14, the role of this receptor in T cell homing to the CNS during
102 autoimmunity has not been tested.
103 Here, we show that peripheral OVA specific T cells similarly sample CNS
104 antigens independent of the cellular or regional origin of antigens during
105 homeostatic conditions. Yet during neuroinflammatory disease, OVA specific T
106 cells exacerbate outcome in Cnp-OVA mice but ameliorate EAE in Nes-OVA
107 mice in a CX3CR1 dependent manner. These data highlight the importance of
108 the cellular source and regional location of autoreactive CNS antigens and reveal
4
109 a novel mechanism for how antigen specific T cells affect the neuroinflammatory
110 outcome of CNS autoimmunity.
111 112 Materials and Methods
113 Mice. The pZ/EG plasmid (generated by Dr. Andras Nagy, Mount Sinai Hospital,
114 Toronto, ON) was designed to express EGFP under the control of the CMV/β-
115 actin promoter. Two loxP sites flank a β-galactosidase expression cassette
116 adjacent to a neomycin resistance gene in the plasmid, which also contains an
117 ampicillin resistance gene between the EGFP expression cassette and the
118 CMV/β-actin promoter complex. The DNA coding sequence for OVA257–264-
119 OVA323–339-PCC88–104 peptides was inserted at the 3′ end of the open reading
120 frame of EGFP before the STOP codon using the QuikChange method
121 (Stratagene, La Jolla, CA). Pigeon cytochrome C peptide (PCC) is presented by
122 I-Ek, that is not present on the MHC Allele Ab expressing C57Bl/6J mouse
123 strain16. This peptide was inserted together with the OVA peptides to be used for
124 antibody purification of the construct or future application with the B10.A mouse
125 strain.
126 The pZ/EG-OVA plasmid was linearized using ScaI restriction enzyme digestion
127 and subsequently purified by electroelution and ethanol precipitation. Transgenic
128 pZ/EG-OVA mice were generated by microinjection of linearized, purified DNA
129 into one-cell C57BL/6 embryos, which were implanted into pseudopregnant
130 C57BL/6 mice. A founder pZ/EG-OVA mouse line was established. It (#147) was
131 crossed with Cnp1Cre/Cre mice and Nestin-cre mice to obtain double transgenic
132 offspring having constitutive Cre-mediated myelinating glial cell-specific
5
133 transgene expression and neural progenitor cell-specific transgene expression
134 respectively. Genotyping of the double transgenic mouse strains was done by
135 PCR on genomic DNA using these primer sequences for GFP: GFP sense: 5′-
136 CAC ATG AAG CAG CAC GAC TT-3′; GFP anti-sense: 5′-TGC TCA GGT AGT
137 GGT TGT CG-3′, and the following Cre driver line-specific primer sequences: Cre
138 mice – Cre-E3 sense: 5′-GCC TTC AAA CTG TCC ATC TC-3′; Cre anti-sense:
139 5′-CCC AGC CCT TTT ATT ACC AC-3′; and puro3: 5′-CAT AGC CTG AAG AAC
140 GAG A-3′; The Thy1.1, OVA257-264-specific OT-I, and OVA323–339-specific OT-II
141 mice were also purchased from Jackson, and the OT-I and OT-II mice were bred
142 to the Thy1.1 background. The Cnp1Cre/Cre mice were a generous gift from Dr.
143 Brian Popko (University of Chicago, Chicago, IL). B6.Cg-Tg( Nes-Cre)1/Kln/J
144 mice were purchased from The Jackson Laboratory and B6.1290Cx3cr1tm1Zm
145 (CX3CR1-/-) mice were purchased from Taconic. Male and female mice (8-14
146 weeks old) were used for peripheral T cell sampling experiments while female
147 mice (8–14 weeks old) were exclusively used for the EAE experiments. All
148 experiments were conducted in accordance with guidelines from the National
149 Institutes of Health and the University of Wisconsin-Madison Institutional Animal
150 Care and Use Committee.
151 EAE induction. EAE was induced by subcutaneous (s.c.) immunization with
152 100 μg MOG35–55 peptide (Genemed Synthesis, Inc., San Antonio, TX) emulsified
153 in CFA (Difco) supplemented with 5 mg/mL H37Ra and Mycobacterium
154 tuberculosis11. Pertussis toxin, 200 ng, (List Biological Laboratories, Inc.,
155 Campbell, CA) was injected i.p. on the day of EAE induction and again two days
6
156 later. Clinical scores were assessed in a double-blinded manner daily according
157 to the following scale: 0, no clinical symptoms; 1, limp/flaccid tail; 2, partial hind
158 limb paralysis; 3, complete hind limb paralysis; 4, quadriplegia; 5, moribund or
159 dead. Intermediate scores were assigned for intermediate symptoms.
160 Adoptive transfer. 106 Lymphocytes from pooled lymph node preparation from
161 OT-I Thy1.1 and OT-II Thy1.1 transgenic mice were adoptively transferred
162 intravenously (i.v.) into the retro-orbital vein of recipient mice11.
163 Lymphocyte isolation, and FACS. Mice were deeply anesthetized with
164 isoflurane and then transcardially perfused with cold PBS. Single cell
165 suspensions were made from cervical lymph nodes (CLNs) and spleens by
166 grinding the tissues between the frosted ends of glass slides11. Red blood cells
167 were lysed using ACK lysis buffer, and cells were washed with HBSS. Brain and
168 spinal cord tissues were minced with razor blades and pushed through 70 μm
169 nylon cell strainers. Cells were washed, resuspended in 70% Percoll and overlaid
170 with 30% Percoll. The gradient was centrifuged at 2400 rpm for 30 min at 4°C
171 without brake. The interface was removed and washed before plating. All
172 collected organs were weighed, and live cells were counted using a
173 hemocytometer.
174 Data were acquired on a BD LSR II flow cytometer (BD Biosciences) and
175 analyzed using FlowJo software (Tree Star, Inc., Ashland, OR).
176 Fluorescent microscopy. For frozen sections, mice were first perfused with cold
177 PBS, followed by perfusion with 4% PFA/PBS. Harvested tissues were left in
178 25% sucrose/PBS overnight at 4°C. 10-40 μm-thick tissue cryosections were cut
7
179 and stored at −80°C until staining. Floating sections were incubated in PBS 2
180 times for 10 minutes min at room temperature before applying primary
181 conjugated antibodies in FACS buffer (PBS/ with 2% BSA/ and 0.1% Sodium
182 Azide) with 0.1% Triton X-100 (1:1000) overnight at 37°C. Sections were then
183 washed 2 times for 10 min each time with PBS and secondary antibodies were
184 applied in PBS (1:500) for 2 hours if necessary. Lastly, sections were washed 3
185 times for 10 min each time with PBS and mounted with ProLong Gold antifade
186 reagent containing DAPI (Invitrogen). All images were acquired with a camera
187 (Optronics Inc., Goleta, CA) mounted on a fluorescence microscope (Olympus
188 BX41, Leeds Precision Instruments). The brightness/contrast of the acquired
189 digital images was applied equally across the entire image and equally to control
190 images and analyzed using Adobe Photoshop CS4 software (Adobe Systems
191 Inc., San Jose, CA).
192 Stereology. Quantification of myelination by optical density and area in were
193 quantified by immunocytochemistry and Image J software based on
194 FluoroMyelin. Quantification of T cells was counted manually by non-biased
195 stereology in a double-blinded manner. Quantification of Immunostaining was
196 performed with Image J to measure area fraction by non-biased stereology in a
197 double-blinded manner. For each histological quantification, a minimum of 5
198 separate sagittal sections for each region of each mouse was analyzed to best
199 reconstruct the overall 3-dimensional structure of the tissue. The number of
200 immunopositve cells were quantified by analyzing digital images collected using
201 Olympus software. For each sections 5-10 fields of view were randomly sampled
8
202 at 10x magnification from each brain region of interest starting with the olfactory
203 bulbs then back towards the brainstem and spinal cord. We utilized a
204 computerized threshold to detect markers for area fraction analyses.
205 Antibodies The following fluorophore-conjugated antibodies were purchased
206 from BD Biosciences (Franklin Lakes, NJ): anti-CD4 (RM4-5), anti-CD8 (53–6.7),
207 anti-CD11c (557400) anti-CD90.1 (Thy1.1) (OX-7). All isotype controls were
208 purchased from BD Biosciences. Inc. GFP booster was purchased from Bulldog
209 Biosciences (Rochester, NY). Anti-Fcγ-R (2.4G2) was produced from a
210 hybridoma. IBA-1(019-19741) was purchased from WAKO (Richmond, VA).
211 FluoroMyelin (F34652) and Click-IT Plus TUNEL assay (C10617) was purchased
212 from Thermo Fisher Scientific (San Francisco, CA). Anti-CX3CR1 (SA011F11)
213 was purchased from Biolegend (San Diego, CA) and anti-CX3CL1 (126315) was
214 purchased from R and D systems (Minneapolis, MN). Anti-GFAP (AB5541), Anti-
215 Neun (MAB377B) and Anti-CNPase (AB9342) was purchased from EDM
216 Millipore (Hayward, CA).
217 Experimental Design and Statistical Analysis
218 Data for figure 1 was performed on a combination of naïve male and female mice
219 (n = 5). Sampling for tissue sections for figure 1 a, b, and c is detailed in
220 stereology section above. Mann Whitney U-test was performed for figure 1d and
221 included 2 independent experiments. P values for hippocampus, cortex,
222 brainstem, and cerebellum was 0.0079. Experimental design for figure 2 is
223 shown in figure 2a and was performed on a combination of male and female
224 mice (n = 6). Mann Whitney U-test was performed for figure 2c and d and
9
225 included 3 independent experiments (P = 0.0411). Experimental design for figure
226 3 is shown in figure 3a and was performed in female mice (n = 17, 11). Linear
227 regression was performed for figure 3b left and included 6 independent
228 experiments. P values were <0.0001. Mann Whitney U-test was performed for
229 figure 3b right and included 6 independent experiments. Sampling for tissue
230 sections for figure 3c is detailed in stereology section above. Mann Whitney U-
231 test was performed for figure 3d and included 3 independent experiments (n = 6).
232 P value was 0.0022. Experimental design for figure 4 is shown in figure 3a and
233 was performed in female mice (n = 5). Mann Whitney U-test was performed for
234 hippocampus/cortex (P = 0.0079), brainstem/cerebellum (P = 0.0079), and spinal
235 cord (P = 0.0159). Sampling for tissue sections for figure 4c is detailed in
236 stereology section above. Experimental design for figure 5 is shown in figure 3a
237 and was performed in female mice (n = 6). Mann Whitney U-test was performed
238 in figure 5b in 2 independent experiments (P = 0.0022). Mann Whitney U-test
239 was performed for figure 5d in 2 independent experiments (P = 0.0159). Mann
240 Whitney U-test was performed for figure 5e in 2 independent experiments (P =
241 0.0159 for diencephalon. P = 0.0079 for hippocampus and cortex). Experimental
242 design for figure 6a and b is shown in figure 3a and was performed in female
243 mice (n = 6). Mann Whitney U-test was performed for figure 6b in 2 independent
244 experiments (P = 0.0411). Experimental design for figure 6 d, e, and f is shown in
245 figure 6 c. Mann Whitney U-test was performed in figure 6e in 2 independent
246 experiments (P = < 0.0001). Experimental design for figure 7 is shown in figure
247 3a and was performed in female mice (n = 5). Sampling for tissue sections for
10
248 figure 7a is detailed in stereology section above. Mann Whitney U-test was
249 performed in figure 7b and c in 2 independent experiments (P = 0.0159).
250 Experimental design for figure 8 is shown in figure 3a and was performed in
251 female mice (n = 5). Sampling for tissue sections for figure 8a is detailed in
252 stereology section above. Mann Whitney U-test was performed in figure 8b (P =
253 0.0022). Mann Whitney U-test was applied to compare measures between two
254 groups and were computed using InStat software (GraphPad Software, La Jolla,
255 CA) to make statistical comparisons between groups (Fig. 1.-Fig. 8). Each group
256 of transgenic mice was compared to non-transgenic littermate controls. Multiple
257 comparisons were made using one-way ANOVA or two-way ANOVA where
258 appropriate. Linear regression was applied to access differences in EAE clinical
259 score (Fig 3 and Fig 6.). Data represent mean ± s.e.m. *P < 0.05, **P < 0.01,
260 ***P < 0.001, ****P < 0.0001. All quantifications were made in 5-10 sagittal
261 sections per mouse using 5-10 animals per transgenic mice. Exact n numbers,
262 number of independent experiments, p-values, and statistical tests are also listed
263 within the figure legends.
264 Ethics statement
265 C57BL/6 WT mice were obtained from The Jackson Laboratory (Bar Harbor,
266 ME). Transgenic pZ/EG-OVA mice were generated at the University of Wisconsin
267 Biotechnology Center Transgenic Facility by microinjection of linearized, purified
268 DNA into one-cell C57BL/6 embryos. The Cnp1Cre/Cre mice were a generous
269 gift from Dr. Brian Popko (University of Chicago, Chicago, IL. Experimental mice
270 underwent adoptive transfer and/or EAE induction. All animal procedures used in
11
271 this study were conducted in strict compliance with the National Institutes of
272 Health Guide for the Care and Use of Laboratory Animals and approved by the
273 University of Wisconsin Center for Health Sciences Research Animal Care
274 Committee. All mice (~25 g) were anesthetized with isoflurane for procedures,
275 and all efforts were made to minimize suffering.
276 Data availability
277 The data that support the findings of this study are available from the
278 corresponding author upon request.
279 Results
280 Peripheral antigen sampling by T cells of CNPase and Nestin derived
281 neoepitopes is similar under homeostatic conditions.
282 In order to test how distinct CNS cell derived neoepitopes are recognized
283 by peripheral T cells, we utilized the pZ/EG plasmid to create the transgenic
284 loxP-carrying pZ/EG-OVA mice having the DNA coding sequence for EGFP,
285 OVA peptides OVA257-264 (MHC class I-restricted, recognized by OT-I T cells) and
286 OVA323-339 (MHC class II-restricted, recognized by OT-II T cells), as well as
11 287 control peptide PCC88-104 . The pZ/EG-OVA mouse strain was crossed with the
288 Nestin-Cre or CNPase-Cre mice to obtain CnpCre +/- OVA fl/fl (Cnp-OVA) and
289 NesCre +/- OVA fl/fl mice (Nes-OVA) as well as respective littermates (CnpCre +/-
290 OVA -/- and NesCre +/- OVA -/- mice).
291 Immunofluorescence staining with GFP booster antibodies in combination
292 with CNPase confirmed that Cnp-OVA mice successfully incorporated EGFP into
293 CNP+ cells while the littermates did not (Fig. 1a). We then evaluated the regional
12
294 distribution of GFP expression in the brains of the Cnp-OVA and littermate
295 strains. This revealed that the GFP expression was more localized to the
296 brainstem and cerebellum in Cnp-OVA mice relative to other brain regions (Fig.
297 1d). We also confirmed that Nes-OVA mice successfully incorporated EGFP into
298 NeuN+ cells (Fig. 1b). Nes-OVA mice also minimally expressed GFP in some
299 GFAP+ cells (Fig. 1c). We did not visualize any noticeable GFP staining of
300 endothelial cells or microglia in the Nes-OVA mice (data not shown).
301 Furthermore, we evaluated the regional distribution of GFP in the brains of the
302 Nes-OVA and littermates, which revealed that the GFP expression was more
303 localized to the cortex and hippocampus in Nes-OVA mice relative to other brain
304 regions (Fig. 1d).
305 To test antigen sampling in the periphery from distinct CNS cell derived
306 neoepitopes under steady state conditions; we adoptively transferred naïve
307 CellTrace Violet-labeled CD8+Thy1.1+ OT-I and CD4+Thy1.1+OT-II T cells into
308 Cnp-OVA and Nes-OVA mice as well as their littermate strains (Fig. 2a). Five
309 days post-adoptive transfer we observed significant CellTrace Violet dilution of
310 OT-I cells in the cervical lymph nodes (CLNs) of Cnp-OVA and Nes-OVA mice
311 but not in littermate strains (Fig. 2 b,c). We also observed significant CellTrace
312 Violet dilution of OT-II cells in the cervical lymph nodes (CLNs) of Cnp-OVA and
313 Nes-OVA mice compared to littermate controls but not to the same extent (Fig. 2
314 b,d). These data show that despite the distinct difference in regional distribution
315 of neoepitopes in the Cnp-OVA and Nes-OVA mice, both strains display similar
316 levels of peripheral T cells surveillance.
13
317 Antigen specific T cells differentially affect EAE in CNPase and Nestin
318 derived neoepitope-expressing transgenic mice.
319 We next tested how distinct CNS cell derived neoepitope specific T cells
320 affect ongoing neuroinflammation modeled in EAE. In order to do this, we
321 induced EAE in Cnp-OVA and Nes-OVA mice as well as their respective
322 littermate strains. Once EAE symptoms were detectable (day 7-10), we
323 adoptively transferred OVA specific T cells and monitored the EAE clinical scores
324 and pathology (Fig. 3a). Our data revealed that transfer of OT-I and OT-II cells
325 into Cnp-OVA mice with ongoing EAE significantly accelerated and augmented
326 EAE clinical score compared to littermate controls. Contrastingly, transfer of OVA
327 specific T cells into Nes-OVA mice with ongoing EAE significantly reduced EAE
328 clinical scores (Fig. 3b). Transfer of only OT-I and/or OT-II T cells at day 7-10 did
329 not elicit significant changes in EAE scores as well (data not shown).
330 Histopathological analysis using FluoroMyelin and staining verified EAE clinical
331 scores showing that Cnp-OVA mice had significantly more demyelination in the
332 spinal cord compared to littermate strains and Nes-OVA mice had significantly
333 less demyelination compared to littermate mice (Fig. 3 c,d). These data indicate
334 that distinct CNS cell derived neoepitope specific T cells differentially affect
335 ongoing MOG35-55-induced EAE clinical score and pathology in mice.
336 Antigen specific T cells differentially steer the localization of immune
337 responses in the CNS of Nestin and CNPase neoepitope-expressing mice
338 during EAE.
14
339 In order to understand the contrasting effect of the adaptive transfer of
340 OVA specific T cells on EAE clinical scores in Nes-OVA or Cnp-OVA mice, we
341 analyzed T cell distribution in the CNS tissues of Cnp-OVA and Nes-OVA mice.
342 Following quantification of CD4 T cells via immunohistochemistry, we found that
343 Cnp-OVA mice had a significantly higher number of CD4 T cells in the spinal
344 cord, brainstem and cerebellum relative to littermate strains at EAE day 30. On
345 the other hand, we observed that the Nes-OVA mice had significantly more CD4
346 T cells in the hippocampus and cortex relative to littermate strains at EAE day 30
347 (Fig. 4a left panel). Analyses of the CD8 T cell numbers revealed similar results,
348 showing that that Cnp-OVA mice had significantly more CD8 T cells in the
349 brainstem and cerebellum relative to littermates and the Nes-OVA mice had
350 significantly more CD8 T cells in the hippocampus and cortex relative to
351 littermates (Fig. 4a right panel). Analysis of Cnp-OVA mice at EAE day 30
352 revealed that although there were some CD4 T cells near the hippocampus in the
353 CA3 region, there were very little if any CD8 T cells in the hippocampus (Fig 4b).
354 Contrastingly, in the brainstem of Cnp-OVA mice at EAE day 30, there is a dense
355 population of both CD4 and CD8 T cells. Representative images of Nes-OVA
356 mice at EAE day 30 reveal a highly dense population of CD4 T cells throughout
357 the hippocampus and also sporadically distributed CD8 T cells in the same area
358 with limited amount of CD4 and CD8 T cells in the brainstem relative to Cnp-OVA
359 mice (Fig. 4b). These data suggest that distinct CNS cell derived neoepitope
360 specific T cells steer the localization of immune responses in mice with ongoing
361 EAE within the CNS towards the location of their cognate antigen.
15
362 Antigen specific T cells form clusters with CD11b+ CD11c+ myeloid cells in
363 distinct CNS regions during EAE.
364 Given previous data showing the importance of myeloid cells in facilitating
365 interactions between T cells of different antigen specificities 10,17,18, we evaluated
366 the frequency of CD45high CD11b+ CD11c+ cells in the brain during EAE in each
367 of the four mouse strains. In order to do this, we isolated whole brains from Cnp-
368 OVA and Nes-OVA mice as well as littermate strains at EAE day 30 that received
369 the transfer of OVA specific T cells at the onset of disease. Using flow cytometry,
370 we gated on the CD45high CD11b+ cells to distinguish infiltrating myeloid cells
371 from local microglia, and analyzed the CD11c+ expressing cells. These data
372 revealed that both the Cnp-OVA and Nes-OVA mice had higher numbers of
373 CD45high CD11b+ CD11c+ cells relative to their respective littermate mice from
374 the brain (Fig 5 a, b). Representative images from the brainstem of Cnp-OVA
375 mice depict myeloid cell-T cell clusters stained with CD11c and CD4 (Fig. 5c).
376 Furthermore, we quantified the number of CD11c+/CD3+ clusters in different
377 regions of the CNS using immunohistochemistry. We show the Cnp-OVA mice
378 had significantly more myeloid cells-T cell clusters in the brainstem, cerebellum,
379 and spinal cord relative to littermates (Fig. 5d). These data also revealed that the
380 Nes-OVA mice had significantly more myeloid cell-T cell clusters in the
381 diencephalon, hippocampus and cortex compared to their littermates (Fig. 5e).
382 Altogether, this suggests that transfer of antigen specific T cells into mice with
383 ongoing EAE leads to enrichment of myeloid-T cell clusters in distinct CNS
384 regions.
16
385 CX3CR1 expression on antigen specific T cells drives their localization to
386 specific regions in the CNS and influences EAE outcome.
387 To understand the mechanism of how neoepitope specific T cells migrate
388 towards the brainstem and cerebellum in Cnp-OVA mice and towards the
389 hippocampus and cortex in Nes-OVA mice during EAE, we performed
390 immunohistochemistry on an array of cytokines known to be involved in inducing
391 T cell migration during autoimmune neuroinflammation. These results identified
392 CX3CL1 as the most abundant cytokine with regional expression levels highly
393 correlated to the EGFP location in Cnp-OVA and Nes-OVA mice shown in figure
394 1. Although CX3CL1 has been shown to be prominently expressed on neurons
395 during EAE 19-23, several groups have identified CX3CL1 expression on CD11c+
396 cells which is important for the recruitment of encephalitogenic T cells during
397 various diseases 24-26. Given these data, we performed immunohistochemistry
398 and performed flow cytometry on brains from Cnp-OVA and Nes-OVA mice as
399 well as their respective littermate strains at EAE day 30 following the transfer of
400 OVA specific T cells at the onset of disease. Representative images indicate that
401 myeloid cell-T cell clusters in Cnp-OVA mice contain CD11c+ CX3CL1+
402 expressing cells mainly in the brainstem whereas in myeloid cell-T cell clusters in
403 Nes-OVA mice, CD11c+ CX3CL1+ expressing cells localize mainly in the cortex
404 (Fig. 6a). Flow cytometry of the whole brains revealed that Cnp-OVA mice and
405 Nes-OVA mice both had a higher percentage of CD45high CD11c+ CX3CL1+
406 expressing cells compared to their respective littermate strains (Fig. 6b).
17
407 To test whether CX3CR1 expression on antigen specific T cells drives
408 their migration to distinct brain regions, we induced EAE in Cnp-OVA and Nes-
409 OVA mice as well as their littermate strains. Once EAE symptoms began (day 7-
410 10), we adoptively transferred CX3CR1 KO and WT OVA specific T cells into the
411 host mice and monitored the EAE clinical scores and pathology (Fig. 6c). We
412 then isolated the brainstem and cerebellum as well as the hippocampus and
413 cortex from these mice to find out where the CX3CR1+ T cells were located in
414 each of the mice. Our results also show that CX3CR1+ T cells primarily localized
415 to the brainstem and cerebellum of Cnp-OVA mice and to the hippocampus and
416 cortex of Nes-OVA mice (Fig. 6 d, e). More importantly, we observed that
417 adoptive transfer of CX3CR1 KO OVA specific T cells into Cnp-OVA mice and
418 Nes-OVA mice did not affect EAE scores compared to the transfer of WT OVA
419 specific T cells into the same host mice (Fig. 6f). These results indicate that the
420 influence of distinct CNS cell derived antigen specific T cells on mice with EAE is
421 dependent on CX3CR1 driven migration to distinct areas.
422 Antigen specific T cells differentially affect myeloid cell morphology and
423 distribution in CNPase and Nestin derived neoepitope-expressing
424 transgenic mice during EAE.
425 In order to evaluate whether or not the differential distribution of OVA-
426 specific T cells influence myeloid cells during EAE, we performed
427 immunohistochemistry on sections from Nes-OVA and Cnp-OVA mice without
428 EAE and at EAE day 30. These experiments revealed that Nes-OVA mice had
429 higher levels of IBA-1 in the cortex compared to Cnp-OVA mice at EAE day 30 or
18
430 compared to No EAE mice (Fig. 7a, b). On the other hand, Cnp-OVA mice had
431 higher levels of IBA-1 staining in the brainstem relative to Nes-OVA mice at EAE
432 day 30 and No EAE mice (Fig 7 a, c). These results suggest that the differential
433 distribution of T cells in the Nes-OVA and Cnp-OVA mice during EAE correlates
434 with augmented myeloid cell activation by measure of IBA-1 expression.
435 Antigen specific T cells differentially influence TUNEL expression on
436 neurons in CNPase and Nestin derived neoepitope-expressing transgenic
437 mice during EAE.
438 To address whether the increased immune cell load in different brain areas
439 influenced neuronal death regionally in Cnp-OVA and Nes-OVA mice, we
440 performed a TUNEL assay in conjunction with NeuN staining on brain sections
441 from Cnp-OVA and Nes-OVA mice at EAE day 30. These results revealed that
442 Nes-OVA mice had a greater expression of double expressing TUNEL+ NeuN+
443 cells in the cortex at EAE day 30 compared to Cnp-OVA mice. On the other
444 hand, Cnp-OVA mice had a greater higher expression of double expressing
445 TUNEL+ NeuN+ cells in the brainstem at EAE day 30 compared to Nes-OVA
446 mice (Fig 8 a,b). These results indicate that while Nes-OVA mice display
447 reduced EAE scores compared to Cnp-OVA mice, they may have dysfunction
448 related to higher order brain areas such as memory and cognition due to the
449 neuronal death observed in the cortex.
450 Discussion
451 Our data highlight the importance of the cellular source and regional location
452 of autoreactive CNS antigens while also reveal a novel mechanism for how
19
453 antigen specific T cells migrate and influence outcome during CNS autoimmunity.
454 We have previously shown that antigen sampling and immunosurveillance in the
455 CNS are similar to non-immune privileged tissues but antigen specific T cells
456 infiltrate the CNS only when neuroinflammation is present 11. Here, we studied
457 how the same T cells respond to their cognate antigens when the antigens are
458 expressed distinctly in different CNS cell types and regions as well as how they
459 influence the outcome of neuroinflammatory disease. We show that T cell
460 peripheral sampling from the CNS is independent of where and what cell type the
461 neoepitopes are derived from. We also show that CX3CR1 dependent T cell
462 migration to distinct brain regions significantly influences the outcome of ongoing
463 neuroinflammatory disease.
464 Using different antigen expressing CNS cell specific T cell transgenic models,
465 other labs have shown that multiple antigen specific T cells can influence the
466 course of neuroinflammatory disease 27,28. Several models have been utilized to
467 induce expression of foreign antigens into various CNS cells utilizing
468 oligodendrocyte 27-32, astrocyte 33 and neuronal promoters 34-36, yet, there are no
469 studies to test whether the same antigen specific T cells can influence ongoing
470 myelin initiated autoimmune disease differentially based on the location and
471 cellular source of the epitopes. Following MOG35-55 induced neuroinflammation,
472 we show that antigen specific T cells influence clinical outcome in a distinct
473 manner depending on the source and location of the neoepitopes.
474 The influence of multiple antigen specific T cells on EAE also depends on
475 the regional location of myeloid cell-T cell clusters within the CNS. This suggests
20
476 that CD11c expressing infiltrating cells could provide a platform for interaction
477 between T cells of different antigen specificities. Our flow cytometry data suggest
478 that most of the CD11c+ cells are peripherally derived (CD45high CD11b+
479 CD11c+). Several studies, including this one, have shown that myeloid cells can
480 contribute to the inflammatory microclusters seen in EAE and MS lesions. These
481 clusters have been described before and can strongly express chemokines such
482 as CCL5, CX3CL1, CXCL19, and CXCL10 which are critical for the recruitment
483 of encephalitogenic T cells 41. Future studies to evaluate whether inhibiting
484 myeloid cell function and activity can influence T cell migration and phenotype in
485 our system is imperative. Moreover, it will be key to identify the myeloid cell
486 specific factors responsible for augmenting pro-inflammatory cascades involving
487 T cells of multiple antigenic specificities.
488 Finally, our data indicate that the influence of the transfer of antigen
489 specific T cells into ongoing MOG induced EAE is dependent on their expression
490 of CX3CR1. Although several studies identify CX3CL1 to be prominently
491 expressed on neurons during EAE19-22,42, the role of the ligand on neural cells
492 during autoimmune diseases of the CNS is unclear. Interestingly CX3CR1 KO
493 mice have more severe EAE compared to WT, yet this has been suggested to be
494 mainly due to the role of NK cells and microglia. NK cells have been suggested
495 to limit autoimmune responses by killing DCs and producing regulatory
496 cytokines43,44, while microglial CX3CR1 is suggested to quell their activation and
497 limit their pro-inflammatory influence on EAE45. These data highlight the
498 importance of targeting chemokine receptors in a cell specific manner when
21
499 treating disease. Our data suggests that the CX3CL1 on dendritic cells plays a
500 major role in facilitating the recruitment CX3CR1 dependent T cell responses,
501 which is in line with other work assessing CX3CL1 dependent migration and
502 immune surveillance 14,15, 19,24-26,45. These data strongly suggest that CX3CR1+ T
503 cells are responding to DC expression of CX3CL1+ providing novel information
504 about how CX3CR1 contributes to the localization of antigen specific T cells in
505 the CNS.
506 Our data show that transgenic mice that express OVA antigens in Nestin+
507 cells have lower EAE clinical scores following OVA specific T cell transfer given
508 that the immune response is steered from motor to non-motor associated brain
509 areas such as the hippocampus and cortex. While this effect is beneficial for EAE
510 clinical score, a key question is how this influences the brain areas where the T
511 cells end up and their impact on hippocampal and cortical dependent brain
512 functions. Recent studies have suggested that neuronal produced factors can
513 direct T cell phenotype 46,47 to protect mice from the detrimental effects of
514 autoimmune disease 48 so it is possible that these brain areas are better
515 equipped to handle pro-inflammatory environments. Nonetheless, our data reveal
516 that there is in fact more neuronal death in the cortex of Nes-OVA mice
517 compared to Cnp-OVA mice at EAE day 30 suggesting that these mice could
518 have higher order brain deficits in memory or cognition. While other murine
519 models have been developed to understand how autoimmune T cells affect
520 neuroinflammation in higher order brain areas 52,53, our models uniquely allow us
521 to study novel questions about the influence of multiple antigen specific T cells
22
522 on neuroinflammatory outcome without damaging the brain via cerebral injections
523 of foreign antigens.
524 In patients affected by CNS autoimmune diseases such as MS, recent
525 work has identified that inflammatory lesions affect different brain regions
526 distinctly54 and that unique clinical outcomes can develop as a response to where
527 neuroinflammatory responses are located within the CNS55. Future studies entail
528 identifying whether the distribution of myeloid cell-T cell clusters and T cell
529 expression of CX3CR1 within the human brain correlates with MS outcome.
530 Moreover, since the effect of antigen specific T cells on ongoing disease was
531 dependent on CX3CR1 signaling, modifying CX3CR1 dependent T cell
532 responses might be beneficial for patients inflicted with autoimmune disease.
533 Altogether, these data reveal new information for how antigen specific peripheral
534 immune cells respond to regionally distinct CNS derived antigens under naïve
535 conditions while also unveil a novel mechanism for how antigen specific T cells
536 modulate neuroinflammatory disease outcome.
537 Acknowledgements
538 We would like to thank Khen Macvilay, Laura Schmitt Brunold, and Satoshi
539 Kinoshita for excellent technical assistance. We would also like to thank Eric
540 Silignavong and Christian Gerhart for assistance. This research was supported
541 by NIH/NIGMS grant T32 GM007507 (Neuroscience Training Program), NIH
542 grants R01-NS37570 (Z.Fabry) and R01-AI048087 (M. Sandor), and AHA pre-
543 doctoral grant 1525500022 (A. Rayasam).
544 Figure Legends
23
545 Figure 1 | GFP expression of CNPase and Nestin derived antigens in the
546 CNS. (a) Representative image of brain section from brainstem stained with
547 CNPase (left panel), GFP (center panel) and merged (right panel) in CnpCre
548 +/- OVA fl/fl mice. Scale bar = 10 μm. White arrows depict co-localization of cell
549 specific markers with GFP tagged neoepitope expression. (b) Representative
550 image of brain section from cortex stained with NeuN (left panel), GFP (center
551 panel) and merged (right panel) in NesCre +/- OVA fl/fl mice. Scale bar = 10 μm.
552 White arrows depict co-localization of cell specific markers with GFP tagged
553 neoepitope expression. (c) Representative image of brain section stained with
554 GFAP (left panel), GFP (center panel) and merged (right panel) in NesCre +/-
555 OVA fl/fl mice. Scale bar = 10 μm. White arrows depict co-localization of cell
556 specific markers with GFP tagged neoepitope expression. (d) Quantification of
557 regional GFP expression by measure of area fraction in different CNS areas of
558 CnpCre +/- OVA fl/fl and NesCre +/- OVA fl/fl mice (n=5; 2 independent
559 experiments). (Data represent mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001,
560 ****P < 0.0001. Mann-Whitney U test (d).
561 Figure 2 | Systemic antigen sampling by T cells of CNPase and Nestin
562 derived antigens is similar under homeostatic conditions. (a) Experimental
563 design to test systemic T cell proliferation in response to CNS derived OVA
564 antigens. Mice received 106 epitope-specific T cells I.V. FACS staining was
565 performed on CLN derived lymphocytes 5 days post-transfer. (b) Histograms
566 show representative CellTrace Violet dilution of CD8+Thy1.1+ cells (top) and
567 CD4+Thy1.1+ cells (bottom) from CLNs of each genotype of mice. (c)
24
568 Quantification of mean number of CellTrace Violet low/int CD8+/Thy1.1+ cells
569 from CLNs of each mouse. (d) Quantification of mean number of CellTrace Violet
570 low/int CD4+/Thy1.1+ cells from CLNs of each mouse. (n=6; 3 independent
571 experiments). Data represent mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001,
572 ****P < 0.0001. Mann-Whitney U test (c, d).
573 Figure 3 | Antigen specific T cells differentially affect EAE in CNPase and
574 Nestin derived neoepitope-expressing transgenic mice. (a) Experimental
575 design to test the influence of secondary antigen specific T cells on ongoing
576 EAE. EAE was induced at day 0 in CnpCre +/- OVA fl/fl, CnpCre +/- OVA -/-,
577 NesCre +/- OVA fl/fl and NesCre +/- OVA -/- mice. OVA peptide specific OT-I
578 Thy1.1 and OT-II Thy1.1 T cells were transferred I.V. at onset of disease (EAE
579 day 7-10). EAE scores were monitored and mice were sacrificed at EAE day 30.
580 (b) Clinical score data in mice from day 0 to day 30 EAE (left panel) and
581 average clinical scores for each mouse at terminal time point (right panel). (n =
582 17, 11 mice, 6 independent experiments). (c) Representative images of spinal
583 cord in mice with No EAE and at EAE day 30 stained with FluoroMyelin-PE.
584 White line depicts area of demyelination. Scale bar = 50 μm. The green box
585 indicates the regions where image was magnified. (d) Quantification of
586 normalized demyelination in thoracic spinal cord of mice at EAE day 30. (n = 6
587 mice, 3 independent experiments). Data represent mean ± s.e.m. *P < 0.05, **P
588 < 0.01, ***P < 0.001, ****P < 0.0001. (linear regression (b left panel (EAE plot)),
589 Mann-Whitney U test (b right panel, d).
25
590 Figure 4 | Antigen specific T cells differentially steer the localization of
591 immune responses in the CNS of Nestin and CNPase neoepitope-
592 expressing mice during EAE. (a) Quantification of CD4 T cells (left panel) per
593 10 mm2 different brain regions and spinal cord at EAE day 30. Quantification of
594 CD8 T cells (right panel) per 10 mm2 different brain regions and spinal cord at
595 EAE day 30. (n = 5, 2 independent experiments) (b) Representative images of
596 hippocampus (top panel) and brainstem (bottom panel) sections stained with
597 CD4 (red), CD8 (green), and DAPI (blue) in CnpCre +/- OVA fl/fl and NesCre +/-
598 OVA fl/fl mice at EAE day 30. Scale bar = 80 μm. White “*” depict CD8 T cells and
599 white “#” depict CD4 T cells. (Data represent mean ± s.e.m. *P < 0.05, **P <
600 0.01, ***P < 0.001, ****P < 0.0001. Mann-Whitney U test (a).
601
602 Figure 5 | Antigen specific T cells form clusters with CD11b+ CD11c+
603 myeloid cells in distinct CNS regions during EAE. (a) Representative gating
604 of CD45high CD11b+ CD11c+ myeloid cells from mice. Infiltrating myeloid cells
high 605 are gated as CD45 CD11b+ CD11c+. Gating of CD11c+ cells based on IgG1
606 isotype control as stated in the materials and methods. (b) Quantification of
607 CD45high CD11b+CD11c+ cells per gram tissue in whole brain at EAE day 30. (n
608 = 6, 2 independent experiments). (c) Representative images of myeloid cell-T cell
609 clusters in brainstem of CnpCre +/- OVA fl/fl mice at EAE day 30. Scale bar = 20
610 μm (CD4 = red, CD11c = green, DAPI = blue) White arrows depict the
611 localization of the clusters. (d) Quantification of CD3/CD11c+ clusters in different
612 brain areas and spinal cord at EAE day 30 in CnpCre +/- OVA fl/fl and CnpCre +/-
26
613 OVA -/- mice (n= 5, 2 independent experiments). (e) Quantification of
614 CD3/CD11c+ clusters in different brain areas and spinal cord at EAE day 30 in
615 NesCre +/- OVA fl/fl and NesCre +/- OVA -/- mice (n= 5, 2 independent
616 experiments). Data represent mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001,
617 ****P < 0.0001. Mann-Whitney U test (b, d, e).
618 Figure 6 | CX3CR1 expression on antigen specific T cells drives their
619 localization to specific regions in the CNS and influences EAE outcome. (a)
620 High magnification representative image stained with CX3CL1 (first panel),
621 CD11c (second panel), IBA-1 (third panel) and merged (fourth panel) of
622 brainstem in CnpCre +/- OVA fl/fl mice and cortex in NesCre +/- OVA fl/fl mice at
623 EAE day 30. Scale bar = 5 μm. White arrows depict CD11c+ CX3CL1 expressing
624 cells. (b) Quantification of percentage of CD45high CD11b+ CD11c+ that are
625 CX3CL1+ in the whole brain of mice at EAE day 30. (n = 6, 2 independent
626 experiments) (c) Experimental design to test the influence of secondary antigen
627 specific T cells deficient in CX3CR1 on ongoing EAE. EAE was induced at day 0
628 in CnpCre +/- OVA fl/fl , CnpCre +/- OVA -/- , NesCre +/- OVA fl/fl and NesCre +/-
629 OVA -/- . CX3CR1 deficient OVA peptide specific OT-I Thy1.1 and OT-II Thy1.1 T
630 cells were transferred I.V. at onset of disease (EAE day 7-10). EAE scores were
631 monitored and mice were sacrificed at EAE day 30. (d) Representative gating of
632 CX3CR1+ Thy1.1+ single positive, double positive, and double negative cells
633 from cells gated on combined single positive CD4 and CD8 T cells from
634 brainstem and cerebellum (top row) and hippocampus and cortex (bottom row).
635 (e) Quantification of absolute number of CX3CR1+ Thy1.1+ cells per gram tissue
27
636 in brainstem and cerebellum (left panel) and quantification of absolute number of
637 CX3CR1+ Thy1.1+ cells per gram tissue in hippocampus and cortex (right
638 panel). (n = 6, 2 independent experiments) (f) EAE clinical score data in mice
639 from day 0 to day 30 Data are pooled from 2 independent experiments (n= 8)
640 EAE was induced at day 0 in CnpCre +/- OVA fl/fl , CnpCre +/- OVA -/- , NesCre
641 +/- OVA fl/fl and NesCre +/- OVA -/- . CX3CR1 deficient OVA peptide specific OT-I
642 Thy1.1 and OT-II Thy1.1 T cells were transferred I.V. at onset of disease (EAE
643 day 7-10). EAE scores were monitored and mice were sacrificed at EAE day 30.
644 Data represent mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001, ****P <
645 0.0001. (linear regression (f (EAE plot)), Mann-Whitney U test (b, e).
646 Figure 7 | Antigen specific T cells differentially affect myeloid cell
647 morphology and distribution in CNPase and Nestin derived neoepitope-
648 expressing transgenic mice during EAE. (a) Representative images of cortex
649 (top panel) and brainstem (bottom panel) sections stained with IBA-1 of naïve
650 CnpCre +/- OVA fl/fl and NesCre +/- OVA fl/fl mice as well as CnpCre +/- OVA fl/fl
651 and NesCre +/- OVA fl/fl mice at EAE day 30. Scale bar = 10 μm. (b)
652 Quantification of IBA-1+ cells by measure of area fraction in the cortex of naïve
653 CnpCre +/- OVA fl/fl and NesCre +/- OVA fl/fl mice as well as CnpCre +/- OVA fl/fl
654 and NesCre +/- OVA fl/fl mice at EAE day 30. (n = 6, 2 independent experiments)
655 (c) Quantification of IBA-1+ cells by measure of area fraction in the brainstem of
656 naïve CnpCre +/- OVA fl/fl and NesCre +/- OVA fl/fl mice as well as CnpCre +/-
657 OVA fl/fl and NesCre +/- OVA fl/fl mice at EAE day 30. (n = 6, 2 independent
28
658 experiments). Data represent mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001,
659 ****P < 0.0001. Mann-Whitney U test (b, c).
660 Figure 8 | Antigen specific T cells differentially influence TUNEL expression
661 on neurons in CNPase and Nestin derived neoepitope-expressing
662 transgenic mice during EAE. (a) Representative images of cortex and
663 brainstem sections stained with NeuN (red), TUNEL (greeen) and DAPI (blue) of
664 CnpCre +/- OVA fl/fl and NesCre +/- OVA fl/fl mice at EAE day 30. Scale bar = 10
665 μm. (b) Quantification of the percentage NeuN+ cells also positive for TUNEL in
666 the cortex and brainstem of CnpCre +/- OVA fl/fl and NesCre +/- OVA fl/fl mice at
667 EAE day 30 (n =5).(Data represent mean ± s.e.m. *P < 0.05, **P < 0.01, ***P <
668 0.001, ****P < 0.0001. Mann-Whitney U test (b).
669
670
671 References
672 673 1 Engelhardt, B., Vajkoczy, P. & Weller, R. O. The movers and shapers in 674 immune privilege of the CNS. Nat Immunol 18, 123-131, doi:10.1038/ni.3666 675 (2017). 676 2 Ransohoff, R. M. & Engelhardt, B. The anatomical and cellular basis of 677 immune surveillance in the central nervous system. Nat Rev Immunol 12, 678 623-635, doi:10.1038/nri3265 (2012). 679 3 Kerlero de Rosbo, N. et al. Reactivity to myelin antigens in multiple sclerosis. 680 Peripheral blood lymphocytes respond predominantly to myelin 681 oligodendrocyte glycoprotein. J Clin Invest 92, 2602-2608, 682 doi:10.1172/JCI116875 (1993). 683 4 Malyavantham, K. et al. Humoral Responses to Diverse Autoimmune Disease- 684 Associated Antigens in Multiple Sclerosis. PLoS One 10, e0129503, 685 doi:10.1371/journal.pone.0129503 (2015). 686 5 Krakowski, M. L. & Owens, T. Naive T lymphocytes traffic to inflamed central 687 nervous system, but require antigen recognition for activation. European 688 journal of immunology 30, 1002-1009, doi:10.1002/(SICI)1521- 689 4141(200004)30:4<1002::AID-IMMU1002>3.0.CO;2-2 (2000).
29
690 6 Qing, Z., Sewell, D., Sandor, M. & Fabry, Z. Antigen-specific T cell trafficking 691 into the central nervous system. J Neuroimmunol 105, 169-178 (2000). 692 7 Ji, Q., Castelli, L. & Goverman, J. M. MHC class I-restricted myelin epitopes are 693 cross-presented by Tip-DCs that promote determinant spreading to CD8(+) T 694 cells. Nat Immunol 14, 254-261, doi:10.1038/ni.2513 (2013). 695 8 Kadowaki, A. et al. Gut environment-induced intraepithelial autoreactive 696 CD4(+) T cells suppress central nervous system autoimmunity via LAG-3. 697 Nature communications 7, 11639, doi:10.1038/ncomms11639 (2016). 698 9 Flach, A. C. et al. Autoantibody-boosted T-cell reactivation in the target organ 699 triggers manifestation of autoimmune CNS disease. Proc Natl Acad Sci U S A 700 113, 3323-3328, doi:10.1073/pnas.1519608113 (2016). 701 10 Karman, J., Ling, C., Sandor, M. & Fabry, Z. Dendritic cells in the initiation of 702 immune responses against central nervous system-derived antigens. 703 Immunol Lett 92, 107-115, doi:10.1016/j.imlet.2003.10.017 (2004). 704 11 Harris, M. G. et al. Immune privilege of the CNS is not the consequence of 705 limited antigen sampling. Sci Rep 4, 4422, doi:10.1038/srep04422 (2014). 706 12 Staumont-Sallé, D. et al. CX₃CL1 (fractalkine) and its receptor CX₃CR1 707 regulate atopic dermatitis by controlling effector T cell retention in inflamed 708 skin. J Exp Med 211, 1185-1196, doi:10.1084/jem.20121350 (2014). 709 13 Mionnet, C. et al. CX3CR1 is required for airway inflammation by promoting T 710 helper cell survival and maintenance in inflamed lung. Nat Med 16, 1305- 711 1312, doi:10.1038/nm.2253 (2010). 712 14 Böttcher, J. P. et al. Functional classification of memory CD8(+) T cells by 713 CX3CR1 expression. Nat Commun 6, 8306, doi:10.1038/ncomms9306 (2015). 714 15 Gerlach, C. et al. The Chemokine Receptor CX3CR1 Defines Three Antigen- 715 Experienced CD8 T Cell Subsets with Distinct Roles in Immune Surveillance 716 and Homeostasis. Immunity 45, 1270-1284, 717 doi:10.1016/j.immuni.2016.10.018 (2016). 718 16 Kaye, J. et al. Selective development of CD4+ T cells in transgenic mice 719 expressing a class II MHC-restricted antigen receptor. Nature 341, 746-749, 720 doi:10.1038/341746a0 (1989). 721 17 Anandasabapathy, N. et al. Flt3L controls the development of radiosensitive 722 dendritic cells in the meninges and choroid plexus of the steady-state mouse 723 brain. The Journal of experimental medicine 208, 1695-1705, 724 doi:10.1084/jem.20102657 (2011). 725 18 Karman, J., Ling, C., Sandor, M. & Fabry, Z. Initiation of immune responses in 726 brain is promoted by local dendritic cells. J Immunol 173, 2353-2361 (2004). 727 19 Sunnemark, D. et al. CX3CL1 (fractalkine) and CX3CR1 expression in myelin 728 oligodendrocyte glycoprotein-induced experimental autoimmune 729 encephalomyelitis: kinetics and cellular origin. J Neuroinflammation 2, 17, 730 doi:10.1186/1742-2094-2-17 (2005). 731 20 Lou, Z. Y., Chen, C., He, Q., Zhao, C. B. & Xiao, B. G. Targeting CB(2) receptor as 732 a neuroinflammatory modulator in experimental autoimmune 733 encephalomyelitis. Mol Immunol 49, 453-461, 734 doi:10.1016/j.molimm.2011.09.016 (2011).
30
735 21 Lou, Z. Y., Zhao, C. B. & Xiao, B. G. Immunoregulation of experimental 736 autoimmune encephalomyelitis by the selective CB1 receptor antagonist. 737 Journal of neuroscience research 90, 84-95, doi:10.1002/jnr.22721 (2012). 738 22 Limatola, C. & Ransohoff, R. M. Modulating neurotoxicity through 739 CX3CL1/CX3CR1 signaling. Front Cell Neurosci 8, 229, 740 doi:10.3389/fncel.2014.00229 (2014). 741 23 Panek, C. A. et al. Differential expression of the fractalkine chemokine 742 receptor (CX3CR1) in human monocytes during differentiation. Cell Mol 743 Immunol 12, 669-680, doi:10.1038/cmi.2014.116 (2015). 744 24 Papadopoulos, E. J. et al. Fractalkine, a CX3C chemokine, is expressed by 745 dendritic cells and is up-regulated upon dendritic cell maturation. European 746 journal of immunology 29, 2551-2559, doi:10.1002/(SICI)1521- 747 4141(199908)29:08<2551::AID-IMMU2551>3.0.CO;2-T (1999). 748 25 Kanazawa, N. et al. Fractalkine and macrophage-derived chemokine: T cell- 749 attracting chemokines expressed in T cell area dendritic cells. European 750 journal of immunology 29, 1925-1932, doi:10.1002/(SICI)1521- 751 4141(199906)29:06<1925::AID-IMMU1925>3.0.CO;2-U (1999). 752 26 Nukiwa, M. et al. Dendritic cells modified to express fractalkine/CX3CL1 in 753 the treatment of preexisting tumors. European journal of immunology 36, 754 1019-1027, doi:10.1002/eji.200535549 (2006). 755 27 Saxena, A. et al. Cutting edge: Multiple sclerosis-like lesions induced by 756 effector CD8 T cells recognizing a sequestered antigen on oligodendrocytes. J 757 Immunol 181, 1617-1621 (2008). 758 28 Na, S. Y. et al. Naive CD8 T-cells initiate spontaneous autoimmunity to a 759 sequestered model antigen of the central nervous system. Brain 131, 2353- 760 2365, doi:10.1093/brain/awn148 (2008). 761 29 Cao, Y. et al. Induction of experimental autoimmune encephalomyelitis in 762 transgenic mice expressing ovalbumin in oligodendrocytes. European journal 763 of immunology 36, 207-215, doi:10.1002/eji.200535211 (2006). 764 30 Schildknecht, A. et al. Antigens expressed by myelinating glia cells induce 765 peripheral cross-tolerance of endogenous CD8+ T cells. European journal of 766 immunology 39, 1505-1515, doi:10.1002/eji.200839019 (2009). 767 31 Hövelmeyer, N. et al. Apoptosis of oligodendrocytes via Fas and TNF-R1 is a 768 key event in the induction of experimental autoimmune encephalomyelitis. J 769 Immunol 175, 5875-5884 (2005). 770 32 Leone, D. P. et al. Tamoxifen-inducible glia-specific Cre mice for somatic 771 mutagenesis in oligodendrocytes and Schwann cells. Mol Cell Neurosci 22, 772 430-440 (2003). 773 33 Cornet, A. et al. Enterocolitis induced by autoimmune targeting of enteric 774 glial cells: a possible mechanism in Crohn's disease? Proc Natl Acad Sci U S A 775 98, 13306-13311, doi:10.1073/pnas.231474098 (2001). 776 34 Sanchez-Ruiz, M. et al. Molecular mimicry between neurons and an 777 intracerebral pathogen induces a CD8 T cell-mediated autoimmune disease. J 778 Immunol 180, 8421-8433 (2008).
31
779 35 Scheikl, T. et al. Cutting edge: neuronal recognition by CD8 T cells elicits 780 central diabetes insipidus. J Immunol 188, 4731-4735, 781 doi:10.4049/jimmunol.1102998 (2012). 782 36 Casanova, E. et al. A CamKIIalpha iCre BAC allows brain-specific gene 783 inactivation. Genesis 31, 37-42 (2001). 784 37 Grifka-Walk, H. M., Lalor, S. J. & Segal, B. M. Highly polarized Th17 cells 785 induce EAE via a T-bet independent mechanism. European journal of 786 immunology 43, 2824-2831, doi:10.1002/eji.201343723 (2013). 787 38 Jadidi-Niaragh, F. & Mirshafiey, A. Th17 cell, the new player of 788 neuroinflammatory process in multiple sclerosis. Scandinavian journal of 789 immunology 74, 1-13, doi:10.1111/j.1365-3083.2011.02536.x (2011). 790 39 Kalinke, U. & Prinz, M. Endogenous, or therapeutically induced, type I 791 interferon responses differentially modulate Th1/Th17-mediated 792 autoimmunity in the CNS. Immunology and cell biology 90, 505-509, 793 doi:10.1038/icb.2012.8 (2012). 794 40 Rostami, A. & Ciric, B. Role of Th17 cells in the pathogenesis of CNS 795 inflammatory demyelination. J Neurol Sci 333, 76-87, 796 doi:10.1016/j.jns.2013.03.002 (2013). 797 41 Paterka, M. et al. Gatekeeper role of brain antigen-presenting CD11c+ cells in 798 neuroinflammation. EMBO J 35, 89-101, doi:10.15252/embj.201591488 799 (2016). 800 42 Harrison, J. K. et al. Role for neuronally derived fractalkine in mediating 801 interactions between neurons and CX3CR1-expressing microglia. Proc Natl 802 Acad Sci U S A 95, 10896-10901 (1998). 803 43 Hertwig, L. et al. CX3CR1-dependent recruitment of mature NK cells into the 804 central nervous system contributes to control autoimmune 805 neuroinflammation. Eur J Immunol 46, 1984-1996, 806 doi:10.1002/eji.201546194 (2016). 807 44 Huang, D. et al. The neuronal chemokine CX3CL1/fractalkine selectively 808 recruits NK cells that modify experimental autoimmune encephalomyelitis 809 within the central nervous system. FASEB J 20, 896-905, doi:10.1096/fj.05- 810 5465com (2006). 811 45 Garcia, J. A. et al. Regulation of adaptive immunity by the fractalkine receptor 812 during autoimmune inflammation. J Immunol 191, 1063-1072, 813 doi:10.4049/jimmunol.1300040 (2013). 814 46 Liu, Y. et al. FoxA1 directs the lineage and immunosuppressive properties of 815 a novel regulatory T cell population in EAE and MS. Nat Med 20, 272-282, 816 doi:10.1038/nm.3485 (2014). 817 47 Liu, Y. et al. Neuronal IFN-beta-induced PI3K/Akt-FoxA1 signalling is 818 essential for generation of FoxA1(+)Treg cells. Nat Commun 8, 14709, 819 doi:10.1038/ncomms14709 (2017). 820 48 Flügel, A. et al. Neuronal FasL induces cell death of encephalitogenic T 821 lymphocytes. Brain Pathol 10, 353-364 (2000). 822 49 Balschun, D. et al. Vesicular glutamate transporter VGLUT1 has a role in 823 hippocampal long-term potentiation and spatial reversal learning. Cereb 824 Cortex 20, 684-693, doi:10.1093/cercor/bhp133 (2010).
32
825 50 Cheng, X. R., Yang, Y., Zhou, W. X. & Zhang, Y. X. Expression of VGLUTs 826 contributes to degeneration and acquisition of learning and memory. 827 Neurobiol Learn Mem 95, 361-375, doi:10.1016/j.nlm.2011.01.010 (2011). 828 51 Herman, M. A., Ackermann, F., Trimbuch, T. & Rosenmund, C. Vesicular 829 glutamate transporter expression level affects synaptic vesicle release 830 probability at hippocampal synapses in culture. J Neurosci 34, 11781-11791, 831 doi:10.1523/JNEUROSCI.1444-14.2014 (2014). 832 52 Bonfiglio, T. et al. Prophylactic versus Therapeutic Fingolimod: Restoration of 833 Presynaptic Defects in Mice Suffering from Experimental Autoimmune 834 Encephalomyelitis. PLoS One 12, e0170825, 835 doi:10.1371/journal.pone.0170825 (2017). 836 53 Howe, C. L., Lafrance-Corey, R. G., Sundsbak, R. S. & Lafrance, S. J. 837 Inflammatory monocytes damage the hippocampus during acute 838 picornavirus infection of the brain. J Neuroinflammation 9, 50, 839 doi:10.1186/1742-2094-9-50 (2012). 840 54 Haider, L. et al. The topograpy of demyelination and neurodegeneration in 841 the multiple sclerosis brain. Brain 139, 807-815, doi:10.1093/brain/awv398 842 (2016). 843 55 Parisi, L. et al. Patterns of regional gray matter and white matter atrophy in 844 cortical multiple sclerosis. J Neurol 261, 1715-1725, doi:10.1007/s00415- 845 014-7409-5 (2014). 846
33