Role of Sortilin in Models of Autoimmune Neuroinflammation Eva Reuter, Juliane Weber, Magdalena Paterka, Robert Ploen, Tilman Breiderhoff, Jack van Horssen, Thomas E. This information is current as Willnow, Volker Siffrin and Frauke Zipp of September 27, 2021. J Immunol 2015; 195:5762-5769; Prepublished online 13 November 2015; doi: 10.4049/jimmunol.1403156

http://www.jimmunol.org/content/195/12/5762 Downloaded from

References This article cites 26 articles, 10 of which you can access for free at: http://www.jimmunol.org/content/195/12/5762.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 27, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2015 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Role of Sortilin in Models of Autoimmune Neuroinflammation

Eva Reuter,* Juliane Weber,* Magdalena Paterka,* Robert Ploen,* Tilman Breiderhoff,† Jack van Horssen,‡ Thomas E. Willnow,† Volker Siffrin,* and Frauke Zipp*

The proneurotrophin receptor sortilin is a with dual functions, being involved in intracellular protein transport, as well as cellular signal transduction. The relevance of the receptor for various neuronal disorders, such as dementia, seizures, and injury, is well established. In contrast, little is known about the role of sortilin in immune cells and inflammatory diseases. The aim of our study was to elucidate the distribution of sortilin in different immune cell types in mice and humans and to analyze its function in autoimmune CNS inflammation. Sortilin was expressed most profoundly in murine and human macrophages and dendritic cells and to a much lesser extent in B and T cells. In dendritic cells, sortilin had an impact on Ag processing. Accordingly, sortilin was highly expressed by infiltrated perivascular myeloid cells, mainly in vessel cuffs, in the CNS of patients suffering from multiple sclerosis, the most common inflammatory autoimmune disease of the CNS. Yet, sortilin -targeted mice (Sort12/2) and chimeras deficient in sortilin in the immune system were as susceptible as wild-type littermates to T cell–dependent experimental autoimmune encephalomyelitis. Considering our results and Downloaded from recent data from other investigators, we conclude that the proneurotrophin receptor sortilin plays a role in innate, rather than in adaptive, immune processes and, thus, not in autoimmune neuroinflammation. The Journal of Immunology, 2015, 195: 5762–5769.

ortilin was discovered as an intracellular sorting protein (1) especially important during early development and ageing of the and a member of the vacuolar protein sorting 10 protein nervous systems, as well as under conditions of acute insult to the S (VPS10p) domain receptor family (2). Other family CNS (e.g., injury) (15). http://www.jimmunol.org/ members in mammals include SORLA, SORCS1, SORCS2, and In addition to its function in signal transduction, sortilin is in- SORCS3 (1, 3–5). VPS10p domain receptors are type I trans- volved in intracellular transport of from the trans-Golgi membrane proteins with a large N-terminal luminal domain and a network to late via a mannose-6–phosphate–independent short cytoplasmic C terminus (reviewed in Ref. 6). In sortilin, the pathway (16). Because intracellular protein sorting is important for VPS10p domain makes up the entire extracellular portion; other Ag processing in APCs, sortilin might be involved in inflammatory family members have additional extracellular domains. Sortilin is processes. In line with this assumption, recent reports indicate that expressed in various cell types, especially in the , sortilin is involved in survival and activation via brain-derived including in cortical and hippocampal neurons, glia, and sympa- neurotrophic factor (BDNF) (17) and in T cell effector functions via by guest on September 27, 2021 thetic neurons (7), as well as in nonneuronal tissue, such as that in alteration of cytokine production. In this context, Herda et al. (18) muscle, liver, lung, testis, and gland. In line with this broad showed that cytotoxic T cells lacking sortilin produce less IFN-g tissue distribution, sortilin has been implicated in various disorders, and more granzyme A. In addition, Wa¨he et al. (19) suggested that including Alzheimer’s disease (8), frontotemporal lobar dementia sortilin transports target proteins to phagosomes in macrophages. (9), and seizures (10), as examples of CNS diseases, as well as in Combining sortilin’s role as a proneurotrophin receptor in the CNS dyslipidemia and in the risk for (11, 12). and its more recently described influence on several immune cell In neurons, sortilin acts as part of a membrane-bound coreceptor subsets, we were interested in its impact on inflammatory processes complex with p75NTR that is responsible for signaling in the CNS. by proneurotrophins (13, 14). This neuronal cell death pathway is The most common autoimmune neuroinflammatory disease is multiple sclerosis (MS), in which autoreactive T cells enter the *Department of Neurology, University Medical Center of the Johannes Gutenberg CNS and attack the myelin sheath, leading to axonal damage and University Mainz, 55131 Mainz, Germany; †Molecular Cardiovascular Research, neurologic symptoms. Different experimental autoimmune en- Max Delbrueck Center for Molecular Medicine Berlin-Buch, 13125 Berlin, Ger- many; and ‡Department of Molecular Cell Biology and Immunology, VU University cephalomyelitis (EAE) models simulate some of the characteristic Medical Center Amsterdam, 1081 BT Amsterdam, the Netherlands features of MS, such as massive leukocyte infiltration and axonal ORCIDs: 0000-0002-1676-7498 (T.B.); 0000-0002-1231-1928 (F.Z.). injury. The aim of this study was to analyze the impact of sortilin on Received for publication December 19, 2014. Accepted for publication October 9, autoimmune neuroinflammation with a focus on adaptive immu- 2015. nity, which is present in EAE. This work was supported by German Research Foundation Grant ZI 448/10-1. This To this end, we analyzed the expression of sortilin in different work is part of the doctoral thesis of J.W. cells associated with innate and adaptive immunity, as well as in the Address correspondence and reprint requests to Prof. Frauke Zipp, Department of CNS tissue of MS patients. Furthermore, we investigated the Neurology, University Medical Center of the Johannes Gutenberg University Mainz, Langenbeckstrasse 1, 55131 Mainz, Germany. E-mail address: frauke.zipp@unimedizin- consequences of sortilin deficiency in models of EAE using 2 2 mainz.de sortilin-deficient mice (Sort1 / ). In addition to these models of Abbreviations used in this article: BDNF, brain-derived neurotrophic factor; BMDC, autoimmune neuroinflammation, we analyzed the influence of bone marrow–derived DC; DC, dendritic cell; EAE, experimental autoimmune en- sortilin deficiency in a model of acute neuronal damage, namely cephalomyelitis; MCAO, middle cerebral artery occlusion; MCS, mean clinical score; MOG, myelin oligodendrocyte ; mNSS, modified neurological cerebral ischemia, in which “secondary” inflammatory mecha- severity score; MS, multiple sclerosis; RPMI/H, RPMI 1640 + 1% HEPES; VPS10p, nisms play a role (20). We used the middle cerebral artery oc- vacuolar protein sorting 10 protein; WT, wild-type. clusion (MCAO) model to detect differences in damage size or Copyright Ó 2015 by The American Association of Immunologists, Inc. 0022-1767/15/$25.00 functional outcome resulting from sortilin knockout. www.jimmunol.org/cgi/doi/10.4049/jimmunol.1403156 The Journal of Immunology 5763

Materials and Methods plated in 10 ml medium with 10% FCS, using 4 3 106 cells/petri dish. T cell culture Cells were stored at 37˚C. To induce generation of the dendritic cell (DC) subset, GM-CSF produced in HEK293 cells was added regularly. On day CD4 T effector cell differentiation. T cells were isolated from the spleen of 3, cultures were supplied with 10 ml fresh medium; on day 6, nonadherent 2 2 Sort1 / and wild-type (WT) littermates by MACS-Sort (direct CD4 Sort; cells were removed by gently washing, and half of the medium was Miltenyi Biotec), according to the manufacturer’s protocol. CD4+ T cells replaced. The cultures were harvested on day 10. Activated DCs were were cultured with APCs (1:10), and cytokines (all from R&D Systems, generated by adding LPS from Salmonella typhimurium 24 h before har- Wiesbaden, Germany) were added for Th cell differentiation into Th1 vesting. The harvested cells were stained for the DC markers CD11c and (IL-12 [10 ng/ml]), Th17 (TGF-b [1.5 ng/ml], IL-6 [20 ng/ml], IL-23 CD11b and for the activation markers CD80, CD86, and MHC class II with [20 ng/ml]), and regulatory T cells (TGF-b [3 ng/ml]). After 72 h, cells fluorescent Abs and measured by flow cytometry (FACSCanto II; BD). were harvested, and cytokine production was measured by flow cytometry Ag-processing capacity. To compare the capability of bone marrow–derived (FACSCanto II; BD, Heidelberg, Germany). DCs (BMDCs) from Sort12/2 and Sort1+/+ mice to process an Ag, im- 2 2 Th17 culture. Naive CD4+ T cells were isolated from Sort1 / mice or WT mature BMDCs were harvested and incubated with DQ OVA (Invitrogen) littermates by magnetic sorting and cocultured for 18 d with APCs from at 37˚C. A control was stored at 4˚C. Cells were plated in a 48-well plate, C57BL/6 mice (ratio 1:3–1:10), 12.5 mg/ml myelin oligodendrocyte gly- using 5 3 105 DCs in 0.5 ml medium/well. Different concentrations of DQ coprotein (MOG)35–55 (Pepceuticals, Leicester, U.K.), recombinant mouse OVA were used. After 2, 4, and 6 h, cells were stained with anti-CD11c, IL-6 (20 ng/ml; R&D Systems), recombinant mouse IL-23 (20 ng/ml; and fluorescence of the degraded DQ OVA in CD11c+ cells was measured R&D Systems), and recombinant human TGF-b (3 ng/ml; R&D Sys- via FACS (FACSCanto II; BD). tems) at 3 million cells/ml in 24-well plates. Cells were kept in culture for DC–T cell coculture. To assess the capability of BMDCs from Sort1+/+ and 7 d, the medium was supplemented with recombinant human IL-2 (Chiron Sort12/2 mice to induce T cell proliferation, DCs were stimulated with Therapeutics) every 3–4 d, and cells were restimulated after 1 wk in OVA and incubated with OT2-transgenic CD4+ T cells. To monitor cell culture with fresh APCs. Three days after the second restimulation, cells division, CFSE was used to mark the CD4+ cells. Spleens and lymph nodes were transferred to induce EAE. Cytokine expression was checked by flow of OT2-transgenic mice were removed to isolate CD4+ cells by positive Downloaded from cytometry (FACSCanto II; BD) before transfer. Analysis of flow cytometry magnetic sorting, following the manufacturer’s protocol (CD4 [L3T4] data was done using FlowJo software (TreeStar, Ashland, OR). MicroBeads, Miltenyi Biotec). The isolated cell fraction was labeled with fluorescent CD4+ Abs, and purity was measured with a FACSCanto II. The Dendritic cells proportion of CD4+ cells was usually .90%. Human PBMCs. PBMCs were isolated from buffy coats of healthy controls CD4+ cells were incubated at 37˚C for 30 min and then washed twice by Ficoll/Hypaque density-gradient centrifugation. To obtain a specific cell with preheated RPMI 1640 + 1% HEPES (RPMI/H). Up to 50 3 106 cells/10 ml + + + subset, positive magnetic sorting was performed using CD14 ,CD8 ,CD4 , were incubated in CFSE solution for 10 min at 37˚C (see “CFSE protocol” http://www.jimmunol.org/ or CD19+ beads (Miltenyi Biotec; see manufacturer’s protocol). Purity was below). checked with FACS analysis after labeling (FACSCanto II; BD) with fluo- Immature DCs were harvested and incubated without Ag or with CD4+ rescent Abs and was typically .95%. OVA peptide (1.04 mg/ml; Pepceuticals) or OVA protein (20 mg/ml; Murine bone marrow–derived dendritic cells. Cells were generated from Sigma, Munich, Germany) in culture medium in a 48-well plate. Triplets bone marrow from the femur and tibia using a modified version of the were used for each approach. After an incubation time of 30 min at 37˚C, method of Inaba et al. (21). Cells were rinsed out of the bone marrow cavity CFSE-labeled CD4+ cells were added at a ratio of 4:1 (T cells/DCs). After with 5% FCS and lysed with ammonium chloride. The precursor cells were 72 h, cells were harvested and labeled with CD4 Abs. The number of by guest on September 27, 2021

FIGURE 1. Expression of sortilin in human and murine immune cells. (A) Sortilin expression in murine CD4+ and CD8+ T cells, CD11+ monocytes/ macrophages, and CD45R+ B cells were analyzed by Western blot. Cells were isolated from the spleen by MACS, and the purity of the cell population was controlled by flow cytometry. (B) Sortilin expression was detected in mature and immature murine DCs. (C) Sortilin was also detected in human CD4+,CD14+, and CD19+ cells. Cells were isolated from human blood, and PBMC were sorted into different subpop- ulations by MACS. In (B)and(C), membrane prepa- rations from WT and Sort12/2 mice were loaded as positive and negative controls, respectively. 5764 ROLE OF SORTILIN IN MODELS OF AUTOIMMUNE NEUROINFLAMMATION proliferating OT2 CD4+ cells was determined by FACS analysis (FACS- scribed previously (23). The study was approved by the institutional ethics Canto II; BD). review board of VU University Medical Center, and all donors or their next of CFSE-proliferation assay. To measure CD4+ T cell proliferation, cells were kin provided written informed consent for brain autopsy and use of material labeled with the fluorescent dye CFSE. T cells were then preincubated for and clinical information for research purposes. After deparaffinization and 15–30 min at 37˚C in culture medium and washed twice in prewarmed citrate Ag retrieval, sections were blocked with PBS containing 0.1% RPMI/H. Cells were resuspended in 10 ml prewarmed RPMI/H containing Triton-X and 5% milk powder for 1 h at room temperature. Next, sections 2.5 mM CFSE (Molecular Probes, Darmstadt, Germany) and incubated for were incubated with anti-sortilin (1:250) in PBS supplemented with 0.1% 10 min at 37˚C in the dark. The labeled cells were washed twice with cold Triton-X for 1 h at room temperature and subsequently stained with culture medium, counted, and cultured in 48-well plates for 3 d with the EnVision HRP kit, followed by 3,39-diaminobenzidine tetrahydro- different peptides at different concentrations. Cells were harvested, washed chloridedihydrate. Images were taken on a Leica DM4000B microscope with FACS buffer, stained with anti–CD8-allophycocyanin fluorescent Ab, (Leica Microsystems Heidelberg, Mannheim, Germany). For double stain- and measured on a FACSCanto II (BD). ing, sections were stained with anti-sortilin, as described above, and subse- quently with anti–Iba-1 and alkaline phosphatase–conjugated secondary goat Animal experiments anti-rabbit. Alkaline phosphatase was visualized by Liquid Permanent Red. All sections were counterstained with hematoxylin. Mice. All mice were kept under specific pathogen–free conditions. All animal experiments were performed in accordance with German Animal Protection Laws. For the Western blot analysis, Sort12/2 mice on a C57BL/6J back- Results ground and age/gender-matched WT animals were used (15). For all other Expression of sortilin in human and murine immune cells experiments, Sort12/2 and Sort1+/+ littermates were used. B6.2d2 mice (C57BL/6-Tg[Tcra2D2,Tcrb2D2]1Kuch/J), in which all CD4+ T cells are We analyzed the expression of sortilin in different human and murine MOG35–55 specific, were originally generated by V.K. Kuchroo. immune cell types by Western blot. In the murine system, we de- 2/2 Experimental autoimmune encephalomyelitis. For active EAE, Sort1 tected sortilin in all investigated cell types (Fig. 1). The highest Downloaded from and WT littermate mice were immunized s.c. with 250 mg MOG35–55 or amounts were found in macrophages and DCs. In T cells, sortilin whole-protein MOG (Pepceuticals) emulsified in CFA (BD Difco, Hei- was expressed more highly in CD4+ T cells compared with CD8+ delberg, Germany) and received 200 ng pertussis toxin (List Biological Laboratories) i.p. at the time of immunization and 48 h later. For passive T cells, and expression in B cells was between these levels. This EAE, 5 million in vitro–differentiated CD4+ Th17 T cells were transferred finding was true for murine and for human cells. Because the 2 2 into Rag1 / mice. Clinical signs of EAE were translated into mean clini- highest sortilin levels were found in DCs, we focused on this cell cal score (MCS): 0 = no detectable signs of EAE, 1 = complete tail paralysis,

type and investigated sortilin expression in different maturation http://www.jimmunol.org/ 2 = partial hind limb paralysis, 3 = complete bilateral hind limb paralysis, 4 = total paralysis of forelimbs and hind limbs, and 5 = death. Bone marrow chimeras Recipient animals were sublethally irradiated with 2 3 5 Gy. Bone marrow was isolated from donor mice by flushing the femur and tibia. The cell suspension was depleted of CD90+ T cells by MACS. Eight hours after radiation, 15 3 106 bone marrow cells were transferred i.v. into recipients. Reconstitution was allowed to take place for 8 wk and was checked by flow cytometry analysis of lymphocyte numbers in peripheral blood. For the first 4 wk, mice received prophylactic antibiotics with enrofloxacin in by guest on September 27, 2021 drinking water. Middle artery occlusion MCAO was performed on Sort12/2 mice and WT controls (11 male animals/group; procedure performed on groups unblinded) by intraluminal insertion of a silicone-coated filament (70 fine MCAO suture, tip diameter 0.19 mm; Doccol, Redlands, CA). Briefly, anesthesia was induced with 3% isoflurane and maintained with 1–2% isoflurane in 70% N2O and 30% O2 via facial mask. The filament was introduced via the external carotid artery and advanced. Occlusion of the middle cerebral artery was verified by laser Doppler flowmetry. The filament was withdrawn 45 min after MCAO to allow reperfusion. As an indication that the MCAO procedure was successful, neurologic deficit was measured using the modified neurological severity score (mNSS) (22) directly prior to the MCAO procedure and 2 and 24 h afterward. Twenty-four hours after MCAO, animals were anesthetized and trans- cardially perfused with saline and paraformaldehyde. For infarct size measurement, 20-mm-thick coronal cryosections were cut every 400 mm and stained with cresyl violet (Sigma-Aldrich, St. Louis, MO). Infarct volume was measured using ImageJ (version 1.37 V; National Institutes of Health, Bethesda, MD) and corrected for brain edema by subtracting the volume difference between the ischemic and nonischemic hemispheres from the infarct volume. Western blot Membrane proteins of the respective cells were separated by SDS-PAGE and transferred to a polyvinylidene difluoride membrane. Western blot 2/2 A analysis was performed using mouse anti-sortilin (BD Transduction Lab- FIGURE 2. Th cell development and function in Sort1 mice. ( ) + 2/2 oratories, Franklin Lakes, NJ) and rabbit anti-actin (Sigma-Aldrich) to- Cytokine production in in vitro–differentiated CD4 T cells from Sort1 gether with peroxidase-conjugated secondary antisera. and WT littermates. No significant difference in cytokine profile could be detected between the two groups with regard to IL-17 or IFN-g production. Immunohistochemistry (B) Proliferation of CD4+ T cells from Sort12/2 and WT littermates after Five well-characterized active MS lesions were collected to determine the CD3/28 stimulation was analyzed after 72 h by CFSE-proliferation assay. expression of sortilin (in collaboration with the Netherlands Brain Bank, No significant difference in CFSE fluorescence was detected (i.e., no Amsterdam, the Netherlands). Lesion classification was analyzed as de- differences in proliferation rate). The Journal of Immunology 5765

FIGURE 3. Analysis of maturation and acti- vation status and costimulation capacity of in vitro–generated DCs from Sort12/2 mice and WT littermates. (A) Analysis of Ag-processing capacity by detection of DQ OVA fluorescence intensity. Increase in mean fluorescence of DQ OVA (compared with baseline) is significantly lower in Sort12/2 DCs compared with WT DCs after 4 and 6 h (Student t test; *p , 0.05). (B) CFSE-proliferation assay of DCs from Sort12/2 and WT DCs cocultured with WT CD4+ T cells showed no difference in T cell proliferation. Downloaded from

states. The results showed no significant differences between Sort12/2 CD4 T cells produced IFN-g. In Th1 and Th17 T cells, sortilin expression in mature versus immature DCs (Fig. 1B). we did not detect a relevant disturbance in T effector cell differ-

entiation caused by loss of sortilin. http://www.jimmunol.org/ Characterization of Th cell development and function in Sort12/2 mice T cell proliferation. We further checked the T cell–proliferation capacity of naive CD4+ T cells from Sort12/2 and WT littermates T cell differentiation. In the next step, we investigated the dif- after stimulation with CD3/CD28. A CFSE-proliferation assay + 2/2 ferentiation capacity of naive CD4 T cells isolated from Sort1 showed comparable proliferation in both groups (Fig. 2B). mice, as well as their WT littermates. We differentiated them in vitro and analyzed cytokine production by flow cytometry. Role of sortilin in DCs After differentiation into Th17 cells for 72 h, 10% of the Because Western blot data showed sortilin to be highly expressed in 2 2 WT CD4+ cells, as well as the Sort1 / T cells, produced IL-17 DCs, we analyzed the relevance of sortilin in DC functions such as

(Fig. 2A). Among the Th1-differentiated cells, ∼40% of WT and Ag processing or presentation. Accordingly, we analyzed several by guest on September 27, 2021

FIGURE 4. Histological analysis of sortilin distribution in human brain tissue from MS patients. (A) Sortilin is highly expressed in active MS lesions (left panel). Magnification (original magnification 320) of the marked area shows sortilin+ cells accumulating in the periventricular cuffs. Bold arrows point to sortilin+ macrophages (iden- tified by their morphology). Thin ar- rows denote sortilin2 lymphocytes. (B) Confocal microscopy showed no merge between CD3+ cells and sortilin. (C) Double staining with sortilin and Iba-1 showed that sortilin+ cells were micro- glia/macrophages. Original magnifica- tion 3200. 5766 ROLE OF SORTILIN IN MODELS OF AUTOIMMUNE NEUROINFLAMMATION

DC surface markers for cell maturation, including CD86 and MHC apoptosis. Considering both findings, sortilin may also influence class II, but failed to detect any significant difference between inflammatory processes in the CNS. Therefore, we analyzed Sort12/2 and WT DCs (data not shown). Because the loss of sortilin sortilin distribution in CNS tissue from patients suffering from did not influence DC maturation under these cell culture conditions, MS, the most common autoimmune inflammatory disease of the we next compared the Ag-processing capacities of Sort12/2 and CNS. Staining of MS lesions showed sortilin to be highly WT DCs. DQ OVA allows the measurement of the Ag-processing expressed, especially in the perivascular cuffs (Fig. 4A). Double status because its enzymatic cleavage leads to fluorescent products staining with CD3 ruled out T cells as the predominant sortilin- that can be detected by flow cytometry. We measured the mean expressing cells (Fig. 4B). Instead, sortilin-expressing cells + fluorescence intensity of DQ OVA at different time points and found turned out to be Iba-1 microglia/macrophages, as shown by a significant reduction in Sort12/2 DCs compared with WT DCs double staining (Fig. 4C). PCR analysis was also used to dem- after 4 and 6 h of incubation (Fig. 3A). These data indicate that the onstrate the presence of sortilin mRNA in human microglia loss of sortilin impairs Ag processing in DCs. However, further cultures. coculture experiments with OT2 CD4+ T cells and WT or Sort12/2 Sortilin in EAE DCs revealed that this reduction in Ag processing did not reduce T cell proliferation in a CFSE-proliferation assay (Fig. 3B). Active EAE with MOG peptide. To investigate the functional Taken together, these data suggest that sortilin plays a role in Ag consequence of our findings, we used EAE, an animal model of 2/2 processing in DCs but is not essential for effective T cell activation. MS, and compared the disease course in Sort1 animals (15) with that in their WT littermates. After active immunization with Detection of sortilin in the MS brain Downloaded from MOG35–55 peptide, we detected the first disease symptoms at day As mentioned earlier, sortilin seems to be involved in multiple 14; the peak of disease, as identified by the highest MCSs, oc- inflammatory processes. However, there are also data concerning curred at ∼day 22 (Fig. 5A). There was no significant difference 2 2 its neuroprotective role in neuronal cells in different in vitro and between the disease course in Sort1 / mice and their WT lit- in vivo models, wherein loss of sortilin leads to reduced neuronal termates, with respect to onset or severity. http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 5. Effect of sortilin on disease course and cytokine profiles in active EAE with MOG peptide. (A) EAE was induced by immuni- zation with MOG peptide. The MCS showed no significant differences between the groups (Mann–Whitney U Test). Pooled data from three ex- periments are shown. (B) Flow cytom- etry analysis of lymphocytes isolated from the CNS of EAE mice showed comparable cell numbers and cytokine production in Sort12/2 and WT litter- mates. (C) Flow cytometry analysis of DCs isolated from the brain of EAE mice. Comparable amounts of CD45+ CD11c+ and Ly6c+ cells are seen in the groups. The Journal of Immunology 5767

We analyzed CNS immune cells by flow cytometry at the peak of disease (Fig. 5B). Cytokine production of CD4+ T cells revealed comparable amounts of IL-17–, IFN-g–, and TNF-a–producing cells after stimulation. The amount of cytokines produced by CD4+ T cells from the spleen also showed no significant differ- ences (data not shown), nor did Sort12/2 versus WT CD4+ T cells with respect to the common surface markers CD11c, MHC class II, and Ly 6 (Fig. 5C). Active EAE with whole MOG protein. In contrast to immunization with the MOG35–55 peptide, which needs no further processing, immunization with whole MOG protein allows for testing of the Ag-processing and Ag-presenting capacities of APCs and DCs. If loss of sortilin leads to a dysfunction of protein processing in DCs, one would expect an effect on EAE initiation or disease course in Sort12/2 mice compared with WT mice. Fig. 6A shows disease onset in both groups around day 10 and the peak of the disease around day 14, with an MCS of 3 in Sort12/2 and WT animals. Additionally, no differences were detected between the genotype groups were regard to the remission phase. Downloaded from Passive EAE by transfer of Th17 CD4+ T cells. Because alterations in T cell function of Sort12/2 T cells were reported (18), these may be exacerbated in passive EAE, which is dependent on the encephalitogenicity of the transferred T cells. Therefore, we backcrossed our Sort12/2 strain with MOG-transgenic 2d2 mice. We cultured naive CD4+ T cells from 2d2-Sort12/2 (2d2 Sort) and http://www.jimmunol.org/ 2d2-Sort1+/+ (2d2 WT) mice, according to the previously de- scribed protocols, and transferred fully activated, IL-17–producing T cells into Rag2/2 animals. For both groups, onset of disease occurred on day 20. The MCS was slightly higher in mice that received 2d2 WT Th17, reflecting a more severe disease course, but this difference was not statistically significant (Fig. 6B). EAE in bone marrow chimera. Sortilin exhibits functions in neu-

ronal and immune cells. Yet, a clear distinction between these by guest on September 27, 2021 functions is not possible with the obligate knockout mouse line, which lacks sortilin in all tissues. To analyze the effect resulting from the loss of sortilin only in the immunological compartment, we created bone marrow chimeric mice by transfer of sortilin bone marrow into WT mice (Sort1+/+) after sublethal irradiation. As a control, we transferred WT bone marrow into WT mice. After a reconstitution phase of 8 wk, we induced active EAE with MOG35–55, as previously described. Fig. 6C shows a comparable disease course in both groups, indicating that sortilin is not rele- vant for EAE, a chronic neuroinflammation model with modest involvement of the adaptive immune system.

Cerebral ischemia (MCAO) FIGURE 6. Effect of sortilin on disease course in additional EAE Infarct size after 45 min of transient MCAO did not significantly models. (A) EAE was induced by immunization with MOG protein 2/2 6 6 1–225 . differ between Sort1 and WT mice (42.9 18.3 versus 38.3 The MCS score showed no significant differences between the two groups 3 14.2 mm , respectively; n = 11; p = 0.52, t test; Fig. 7). Also, (Mann–Whitney U test). Results from one experiment. (B) Passive EAE clinical deficit, as measured by mNSS to confirm the success of was induced by transfer of in vitro–differentiated CD4+ Th17 Sort12/2 or the MCAO procedure, did not differ significantly between groups CD4+ Th17 Sort1+/+ cells into Rag2/2 mice. The MCS showed no sig- (5.1 6 1.3 versus 4.2 6 1.4 in knockout and WT animals, re- nificant difference between the two groups (Mann–Whitney U test). Re- spectively; p = 0.19, Mann–Whitney U test). However, it should sults are representative of two experiments. (C) Active EAE in bone be noted that longer survival times would be necessary to detect marrow chimeras. MCS in WT-WT and WT-Sort mice showed a slightly clear behavioral changes in MCAO mice. milder disease course in the WT-Sort chimera, but it was not statistically significant (Mann–Whitney U test). Discussion Experimental evidence increasingly implicates the proneurotrophin regulated in stress situations and secreted in an autocrine manner. receptor sortilin in the regulation of neuronal cell death and survival, Suppression of sortilin expression by small interfering RNA in their as well as in cellular processes in the immune system. However, there model led to a reduced of BDNF and increased apoptosis are few reports concerning the distribution of sortilin in different of cultured B cells. Rogers et al. (24) found that sortilin in complex immune cells. Fauchais et al. (17) showed that sortilin plays a role in with p75NTR mediated apoptosis of NK cells. Blocking of sortilin B cell survival by regulating the transport of BDNF, which is up- with led to a reduced death of NK cells. In contrast to 5768 ROLE OF SORTILIN IN MODELS OF AUTOIMMUNE NEUROINFLAMMATION

FIGURE 7. Sortilin knockout and ce- rebral ischemia. (A) Infarct size (45 min transient MCAO was performed in male Sort12/2 mice and WT controls) did not differ significantly between groups (n = 11/group). (B) Neurological deficit, as measured by mNSS, did not differ sig- nificantly between Sort12/2 and WT mice (n = 11/group). (C) Representative coronal cresyl violet–stained cryosections. Downloaded from http://www.jimmunol.org/

B cells, sortilin in NK cells did not affect intracellular trafficking did not detect an influence of sortilin, because Sort12/2 mice and autocrine production of . These data point to showed a disease course indistinguishable from that in WT mice. opposing effects of sortilin in immune cell function. Our data now This observation was true for active immunization with MOG, as show that the largest amounts of sortilin are found on myeloid cells well as for transfer of completely differentiated Th17 cells in pas- and DCs, both in mice and humans. In line with this cellular dis- sive EAE. Therefore, one explanation might be the greater impact by guest on September 27, 2021 tribution, the expression of sortilin in active lesions of MS patients of IFN-g in inflammatory bowel disease compared with EAE. is also highest in myeloid cells. Interestingly, loss of sortilin ex- Nevertheless, isolation of CD4+ T cells from the CNS of diseased pression reduced the Ag-processing capacity in Sort12/2 DCs EAE mice revealed no significant difference in the number, acti- compared with WT DCs, the most effective APCs. In line with this vation state, or cytokine production of CD4+ T cells. Based on our observation, Wa¨he et al. (19) described that sortilin is involved in observation of reduced Ag-processing capacity, we further analyzed the transport of lysosomal proteins in a Golgi-dependent route to DCs from the CNS of mice with EAE, mimicking, in part, the phagosomes in macrophages. Upregulation of sortilin in macro- activation and maturation marker aspect of MS pathology; however, phages was also detected during the process of mycobacterial we found no significant influence of sortilin loss. killing (25), emphasizing its role in correct cell function during As in our models of autoimmune neuroinflammation, we found response to Ags. Despite the above evidence implicating sortilin in no significant difference in the extent of the damaged area between cellular protein degradation, the reduction in Ag processing in our Sort12/2 mice and B6-WT mice in the ischemic stroke model study did not impact the activation of sortilin-deficient T effector MCAO. Although there is certainly an inflammatory component in cells. T cells cocultured with DCs from Sort12/2 mice showed stroke, it is not as prevalent as in classical neuroinflammatory comparable proliferation in a CFSE-proliferation assay, pointing to diseases, and the innate and adaptive immune systems play dif- alternative pathways in intracellular protein transport compensating ferent roles at different time points following acute injury (20, 26) for the loss of sortilin. Concerning T cell function, the ability of than do those in classic neuroinflammation. naive CD4+ T cells to differentiate into various T effector cells was In summary, we showed that sortilin is expressed in different not affected by the loss of sortilin, because comparable amounts of immune cell populations, including DCs, where it is involved in Ag cytokine producers were detected in our T cell cultures. In contrast, processing. The receptor is also strongly expressed in MS lesions, Herda et al. (18) described a reduced production of IFN-g in CTLs, where it is mainly found in infiltrated macrophages and activated as well as in CD4+ Th1 cells. In the same study, production of microglial cells. However, in contrast to previous data in other granzyme A was increased in both cell types. This altered cytokine inflammatory models, sortilin does not significantly impact cell secretion in Sort12/2 mice led to alterations in two infection function in autoimmune T cell–dependent neuroinflammation in models, including a bacterial infection model with Listeria mono- EAE, because no relevant effect of sortilin loss on disease course cytogenes,inwhichSort12/2 mice showed reduced IFN-g serum was observed. levels and a reduced clearance of liver enzymes as an indicator of reduced physiologic clearance capacity. Furthermore, in a model of Acknowledgments 2/2 inflammatory bowel disease, mice receiving Sort1 CD4 T cells We thank H. Ehrengard, C. Liefla¨nder, B. Hohmann C. Oswald, and K. showed milder colitis pathology and reduced weight loss compared Kampf for expert technical support. We also thank Dr. Darragh O’Neill for with transfer of WT CD4 T cells. However, in our EAE models, we proofreading the manuscript. The Journal of Immunology 5769

Disclosures 14. Teng, H. K., K. K. Teng, R. Lee, S. Wright, S. Tevar, R. D. Almeida, P. Kermani, R. Torkin, Z. Y. Chen, F. S. Lee, et al. 2005. ProBDNF induces neuronal apo- The authors have no financial conflicts of interest. ptosis via activation of a receptor complex of p75NTR and sortilin. J. Neurosci. 25: 5455–5463. 15. Jansen, P., K. Giehl, J. R. Nyengaard, K. Teng, O. Lioubinski, S. S. Sjoegaard, References T. Breiderhoff, M. Gotthardt, F. Lin, A. Eilers, et al. 2007. Roles for the pro- 1. Petersen, C. M., M. S. Nielsen, A. Nykjaer, L. Jacobsen, N. Tommerup, receptor sortilin in neuronal development, aging and brain injury. H. H. Rasmussen, H. Roigaard, J. Gliemann, P. Madsen, and S. K. Moestrup. Nat. Neurosci. 10: 1449–1457. 1997. Molecular identification of a novel candidate sorting receptor purified 16. Nielsen, M. S., P. Madsen, E. I. Christensen, A. Nykjaer, J. Gliemann, D. Kasper, from human brain by receptor-associated protein affinity chromatography. J. R. Pohlmann, and C. M. Petersen. 2001. The sortilin cytoplasmic tail conveys Biol. Chem. 272: 3599–3605. Golgi- transport and binds the VHS domain of the GGA2 sorting 2. Marcusson, E. G., B. F. Horazdovsky, J. L. Cereghino, E. Gharakhanian, and protein. EMBO J. 20: 2180–2190. S. D. Emr. 1994. The sorting receptor for yeast vacuolar carboxypeptidase Y is 17. Fauchais, A. L., F. Lalloue´, M. C. Lise, A. Boumediene, J. L. Preud’homme, encoded by the VPS10 gene. Cell 77: 579–586. E. Vidal, and M. O. Jauberteau. 2008. Role of endogenous brain-derived neu- 3. Hermey, G., I. B. Riedel, W. Hampe, H. C. Schaller, and I. Hermans-Borgmeyer. rotrophic factor and sortilin in B cell survival. J. Immunol. 181: 3027–3038. 1999. Identification and characterization of SorCS, a third member of a novel 18. Herda, S., F. Raczkowski, H. W. Mittrucker,€ G. Willimsky, K. Gerlach, receptor family. Biochem. Biophys. Res. Commun. 266: 347–351. A. A. Kuhl,€ T. Breiderhoff, T. E. Willnow, B. Do¨rken, U. E. Ho¨pken, and 4. Jacobsen, L., P. Madsen, S. K. Moestrup, A. H. Lund, N. Tommerup, A. Nykjaer, A. Rehm. 2012. The sorting receptor Sortilin exhibits a dual function in exocytic L. Sottrup-Jensen, J. Gliemann, and C. M. Petersen. 1996. Molecular charac- trafficking of interferon-g and granzyme A in T cells. Immunity 37: 854–866. terization of a novel human hybrid-type receptor that binds the alpha2- 19. Wa¨he, A., B. Kasmapour, C. Schmaderer, D. Liebl, K. Sandhoff, A. Nykjaer, macroglobulin receptor-associated protein. J. Biol. Chem. 271: 31379–31383. G. Griffiths, and M. G. Gutierrez. 2010. Golgi-to-phagosome transport of acid 5. Rezgaoui, M., G. Hermey, I. B. Riedel, W. Hampe, H. C. Schaller, and sphingomyelinase and prosaposin is mediated by sortilin. J. Cell Sci. 123: 2502– I. Hermans-Borgmeyer. 2001. Identification of SorCS2, a novel member of the 2511. VPS10 domain containing receptor family, prominently expressed in the de- 20. Zipp, F., and O. Aktas. 2006. The brain as a target of inflammation: common

veloping mouse brain. Mech. Dev. 100: 335–338. pathways link inflammatory and neurodegenerative diseases. Trends Neurosci. Downloaded from 6. Willnow, T. E., C. M. Petersen, and A. Nykjaer. 2008. VPS10P-domain receptors 29: 518–527. - regulators of neuronal viability and function. Nat. Rev. Neurosci. 9: 899–909. 21. Inaba, K., M. Inaba, N. Romani, H. Aya, M. Deguchi, S. Ikehara, S. Muramatsu, 7. Sarret, P., P. Krzywkowski, L. Segal, M. S. Nielsen, C. M. Petersen, J. Mazella, and R. M. Steinman. 1992. Generation of large numbers of dendritic cells from T. Stroh, and A. Beaudet. 2003. Distribution of NTS3 receptor/sortilin mRNA mouse bone marrow cultures supplemented with granulocyte/macrophage J. Comp. Neurol. and protein in the rat . 461: 483–505. colony-stimulating factor. J. Exp. Med. 176: 1693–1702. 8. Carlo, A. S., C. Gustafsen, G. Mastrobuoni, M. S. Nielsen, T. Burgert, D. Hartl, 22. Pfeilschifter, W., F. Bohmann, P. Baumgarten, M. Mittelbronn, J. Pfeilschifter, M. Rohe, A. Nykjaer, J. Herz, J. Heeren, et al. 2013. The pro-neurotrophin re- E. Lindhoff-Last, H. Steinmetz, and C. Foerch. 2012. Thrombolysis with ceptor sortilin is a major neuronal receptor for catabolism of

recombinant tissue plasminogen activator under dabigatran anticoagulation in http://www.jimmunol.org/ amyloid-b peptide in the brain. J. Neurosci. 33: 358–370. experimental stroke. Ann. Neurol. 71: 624–633. 9. Lewis, J., and T. E. Golde. 2010. Sorting out frontotemporal dementia? Neuron 23.vanHorssen,J.,C.M.Vos,L.Admiraal, E. S. van Haastert, L. Montagne, 68: 601–603. 10. Le, A. P., and W. J. Friedman. 2012. Matrix metalloproteinase-7 regulates P. van der Valk, and H. E. de Vries. 2006. Matrix metalloproteinase-19 is cleavage of pro- and is neuroprotective following kainic acid- highly expressed in active multiple sclerosis lesions. Neuropathol. Appl. induced seizures. J. Neurosci. 32: 703–712. Neurobiol. 32: 585–593. 11. Kjolby, M., O. M. Andersen, T. Breiderhoff, A. W. Fjorback, K. M. Pedersen, 24. Rogers, M. L., S. Bailey, D. Matusica, I. Nicholson, H. Muyderman, P. Madsen, P. Jansen, J. Heeren, T. E. Willnow, and A. Nykjaer. 2010. Sort1, P. C. Pagadala, K. E. Neet, H. Zola, P. Macardle, and R. A. Rush. 2010. ProNGF encoded by the cardiovascular risk locus 1p13.3, is a regulator of hepatic lipo- mediates death of Natural Killer cells through activation of the p75NTR-sortilin protein export. Cell Metab. 12: 213–223. complex. J. Neuroimmunol. 226: 93–103. 12. Musunuru, K., A. Strong, M. Frank-Kamenetsky, N. E. Lee, T. Ahfeldt, 25. Gutierrez, M. G., B. B. Mishra, L. Jordao, E. Elliott, E. Anes, and G. Griffiths. K. V. Sachs, X. Li, H. Li, N. Kuperwasser, V. M. Ruda, et al. 2010. From 2008. NF-kappa B activation controls phagolysosome fusion-mediated killing of noncoding variant to phenotype via SORT1 at the 1p13 locus. Nature mycobacteria by macrophages. J. Immunol. 181: 2651–2663. by guest on September 27, 2021 466: 714–719. 26. Jolivel, V., F. Bicker, F. Biname´, R. Ploen, S. Keller, R. Gollan, B. Jurek, 13. Nykjaer, A., R. Lee, K. K. Teng, P. Jansen, P. Madsen, M. S. Nielsen, J. Birkenstock, L. Poisa-Beiro, J. Bruttger, et al. 2015. Perivascular microglia C. Jacobsen, M. Kliemannel, E. Schwarz, T. E. Willnow, et al. 2004. Sortilin is promote blood vessel disintegration in the ischemic penumbra. Acta Neuro- essential for proNGF-induced neuronal cell death. Nature 427: 843–848. pathol. 129: 279–295.