Circulating Monocytes Are Reduced by Sphingosine-1-Phosphate Receptor Modulators Independently of S1P3

This information is current as Nuruddeen D. Lewis, Sokol A. Haxhinasto, Shawn M. of September 26, 2021. Anderson, Dimitria E. Stefanopoulos, Steven E. Fogal, Prathima Adusumalli, Sudha N. Desai, Lori A. Patnaude, Susan M. Lukas, Kelli R. Ryan, Anthony J. Slavin, Maryanne L. Brown and Louise K. Modis

J Immunol 2013; 190:3533-3540; Prepublished online 22 Downloaded from February 2013; doi: 10.4049/jimmunol.1201810 http://www.jimmunol.org/content/190/7/3533 http://www.jimmunol.org/ References This article cites 36 articles, 12 of which you can access for free at: http://www.jimmunol.org/content/190/7/3533.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 26, 2021 • No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*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 © 2013 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Circulating Monocytes Are Reduced by Sphingosine-1-Phosphate Receptor Modulators Independently of S1P3

Nuruddeen D. Lewis, Sokol A. Haxhinasto, Shawn M. Anderson, Dimitria E. Stefanopoulos, Steven E. Fogal, Prathima Adusumalli, Sudha N. Desai, Lori A. Patnaude, Susan M. Lukas, Kelli R. Ryan, Anthony J. Slavin, Maryanne L. Brown, and Louise K. Modis

Sphingosine-1-phosphate (S1P) receptors are critical for lymphocyte egress from secondary lymphoid organs, and S1P receptor modulators suppress lymphocyte circulation. However, the role of S1P receptors on monocytes is less clear. To elucidate this, we systematically evaluated monocytes in rats and mice, both in naive and inflammatory conditions, with S1P receptor modulators FTY720 and BAF312. We demonstrate that S1P receptor modulators reduce circulating monocytes in a similar time course Downloaded from as lymphocytes. Furthermore, total monocyte numbers were increased in the spleen and bone marrow, suggesting that S1P receptor modulation restricts egress from hematopoietic organs. Monocytes treated ex vivo with FTY720 had reduced CD40 expression and TNF-a production, suggesting a direct effect on monocyte activation. Similar reductions in protein expression and cytokine production were also found in vivo. Suppression of experimental autoimmune encephalomyelitis in mice and rats by FTY720 correlated with reduced numbers of lymphocytes and monocytes. These effects on monocytes were independent of S1P3,as http://www.jimmunol.org/ treatment with BAF312, a S1P1,4,5 modulator, led to similar results. These data reveal a novel role for S1P receptors on monocytes and offer additional insights on the mechanism of action of S1P receptor modulators in disease. The Journal of Immunology, 2013, 190: 3533–3540.

phingosine-1-phosphate (S1P) is a bioactive mediator by modulation of S1P receptors based on reduced numbers of mac- that regulates multiple cellular processes, including immune rophages in tissue, such as the following: the peritoneal cavity fol- S cell trafficking (1). A gradient of S1P is recognized by S1P lowing thioglycolate-induced peritonitis (10), peripheral nerves in receptors on immune cells, leading to their entry into circulation experimental autoimmune neuritis (15), the CNS in experimental

(2). The biological relevance of this pathway was highlighted autoimmune encephalomyelitis (EAE) (16), renal glomeruli in by guest on September 26, 2021 by the development of FTY720, a potent S1P1,3,4,5 modulator, as experimental mesangioproliferative glomerulonephritis (17), ath- a therapy for multiple sclerosis (3). FTY720 leads to internalization erosclerotic plaques (18), and heart allografts (19). However, an and degradation of S1P receptors on lymphocytes; consequently, effect on circulating monocytes has not been reported. Moreover, lymphocyte egress from secondary lymphoid organs is inhibited, FTY720 has been shown to increase the number of macrophages resulting in a significant decrease in circulating lymphocyte levels in mesenteric lymph nodes, suggesting that S1P receptors may be (4) and reduced lymphocyte infiltration into the CNS (5). important for monocyte egress from lymph nodes in a manner S1P receptors may also be involved in innate immune cell similar to lymphocytes (20). However, it is not clear whether this trafficking (2). In vitro, mast cells (6), dendritic cells (7), eosin- is a direct or indirect effect on either the numbers or trafficking of ophils (8), NK cells (9), macrophages (10), and monocytes (11) circulating monocytes. have been shown to migrate toward S1P, albeit weakly. In vivo, In this work, we examined the effect of S1P receptor modulators NK cell trafficking is dependent upon S1P (12); and similar to on monocytes in vivo and observed reduced circulating monocyte lymphocytes, FTY720 treatment restricts NK cell egress from numbers in mice and rats, in both normal and inflammatory con- lymph nodes and bone marrow (13). Additionally, migration of ditions. We demonstrate that this effect is independent of S1P3 by 2/2 2/2 in vitro generated, mature S1P1 or S1P3 dendritic cells is using S1P receptor modulators with different selectivities. Fur- defective following injection into mice (14). thermore, we provide evidence that FTY720 restricts monocyte The role of S1P receptors on monocytes in vivo is less clear. egress from the spleen and bone marrow, as monocyte cell num- Several reports suggest that monocyte trafficking may be restricted bers are significantly increased at these sites. We also show that monocytes express S1P receptors and that FTY720 can act di-

Department of Immunology and Inflammation, Boehringer Ingelheim Pharmaceuti- rectly upon monocytes to alter their activation, as demonstrated by cals, Ridgefield, CT 06877 reduced surface protein expression and cytokine production. Received for publication June 29, 2012. Accepted for publication January 22, 2013. These findings demonstrate an important role for S1P receptors on Address correspondence and reprint requests to Dr. Louise K. Modis, Department of monocytes in rodents and suggest that additional studies on the Immunology and Inflammation, Research and Development 1461, Boehringer Ingel- functions of S1P receptor modulators on humans may be merited. heim Pharmaceuticals, 900 Ridgebury Road, P.O. Box 368, Ridgefield, CT 06877. E-mail address: [email protected] Abbreviations used in this article: AUC, area under the curve; EAE, experimental Materials and Methods autoimmune encephalomyelitis; EdU, 5-ethynyl-29-deoxyuridine; S1P, sphingosine- Animals 1-phosphate. Female C57BL/6 mice (6–10 wk of age) were purchased from The Jackson Copyright Ó 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00 Laboratory. Female Lewis rats (150–200 g) were purchased from Charles www.jimmunol.org/cgi/doi/10.4049/jimmunol.1201810 3534 FTY720 REDUCES MONOCYTE NUMBERS IN THE CIRCULATION

River Laboratories. Animals were fed and watered ad libitum under 12-h 2mML-glutamine, and 10 mM HEPES. FTY720 was dissolved in this light/dark cycles. All studies were performed in accordance with Institu- medium and was used at 10 nM. LPS was used at 10 ng/ml. After 2 h, cells tional Animal Care and Use Committee guidelines. were centrifuged and supernatants were collected and frozen until needed. Cytokine levels of TNF-a were assessed using a sandwich immunoassay Chemical synthesis and dosing (Meso Scale Discovery). Cells were also stained for CD40 and analyzed by flow cytometry. FTY720 was purchased from Cayman Chemical. BAF312 (21) was pre- pared as the zwitterion, according to the synthetic procedures described In vivo LPS challenge and cytokine measurement (22), and crystallized, as described (23). Compounds were dosed in a ve- hicle containing 0.5% methylcellulose (Spectrum Chemical) and 0.2% C57BL/6 mice were dosed with either 0.3 mg/kg FTY720, BAF312, or Tween 80 (EMD Chemicals). Mice and rats were dosed once daily by oral vehicle (0.5% methylcellulose and 0.2% Tween 80). After 22 h, mice re- gavage. Prophylactic dosing of mice began 1 d before immunization, and ceived an i.p. injection of 200 ng LPS or vehicle (saline). Two hours later, . whole blood was collected. The blood was then centrifuged, and the serum therapeutic dosing of mice began once mice had a clinical score 1or 2 a score of 1 for at least 2 d. For in vitro studies, cell culture medium was was stored at 20˚C until assessed for cytokine levels. Serum levels of used as the vehicle. CCL2 (MCP-1), TNF-a, IL-6, and IL-10 were measured using a multiplex sandwich immunoassay (Meso Scale Discovery). Whole-blood differential analysis EAE induction Whole blood was collected from animals by retro-orbital bleeding and Mice were s.c. injected with 100 mg myelin oligodendrocyte glycopro- analyzed immediately using the Hemavet 1700FS (Drew Scientific). tein35–55 (AnaSpec) in IFA (Difco) containing 250 mg Mycobacterium tuberculosis (Difco). On the same day and again 48 h later, mice received Tissue preparations and flow cytometry an i.p. injection of 200 ng pertussis toxin (List Biological Laboratories). Mice were scored as follows: 0, no symptoms; 1, limp tail; 2, limp tail and Whole blood was isolated retro-orbitally, and 100 ml was used for staining. Downloaded from Spleens were smashed in 2 ml Cell Staining Buffer (BioLegend) over a 40- inability to right itself when placed on its back; 3, partial hind limb pa- mm strainer, and 30 ml was used for staining. Cell Staining Buffer was ralysis; 4, complete hind limb paralysis; 5, complete hind limb paralysis added to make a final volume of 100 ml. The right tibia was used to isolate with forelimb involvement or moribund. EAE was induced in rats by s.c. bone marrow. Bone marrow was resuspended in 800 ml Cell Staining injection of 100 mg guinea pig myelin basic protein68–86 (AnaSpec) in IFA Buffer, and 100 ml was used for staining. TruStain fcX (10 mg/ml; Bio- containing 800 mg M. tuberculosis. Rats were scored as follows: 0, no Legend) was added directly to the samples and incubated on ice for 10–15 symptoms; 1, limp tail; 2, ataxia and/or altered gait; 3, partial hind limb min to block FcRs. The following Abs in a total volume of 50 ml were then paralysis; 4, complete hind limb paralysis; 5, complete hind limb paralysis added at the concentration suggested by the manufacturer. Abs were with forelimb involvement. http://www.jimmunol.org/ purchased from BioLegend (Ly6C [Ag], HK1.4 [clone]; CD115, AFS98; Statistics CD86, GL-1; I-A/I-E [MHC II], M5/114.15.2; CD274 [PD-L1], 10F.9G2; CD40, 3/23; CD69, H1.2F3; CD80, 16-10A1; CD44, IM7; CD62L, MEL- For the EAE clinical score data, area under the curve (AUC) was calculated 14; B220, RA3-6B2; CD4, RM4-5; CD3, 145-2C11; CD8, 53-6.7; Gr-1, and the Kruskal-Wallis test was performed, followed by a Dunn’s posttest to RB6-8C5; NK1.1, PK136), Becton Dickinson Biosciences (CD45, 30-F11; determine significance between multiple groups. For other data, either an CD11b, M1/70), and eBioscience (CD335 [NKp46], 29A1.4). Isotype unpaired t test was used when comparing two groups or a one-way control Abs were used at the same protein concentrations as their corre- ANOVA followed by the Newman-Keuls posttest was used for compar- sponding markers. Cells were stained in the dark on ice for 20 min, and ison between multiple groups. These calculations were performed using then 1.6 ml One-Step Fix/Lyse Solution (eBioscience) was added, mixed, GraphPad Prism 5.04.

and incubated with the cells for 30 min at room temperature. The cell by guest on September 26, 2021 solution was then centrifuged, and the supernatant was discarded. The cells Results were then washed twice with Cell Staining Buffer, resuspended in 200 ml Cell Staining Buffer, run on the FACS Canto II (Becton Dickinson Bio- FTY720 reduces circulating levels of monocytes and NK cells sciences), and analyzed using Flow Jo 7.6.3. without affecting neutrophil levels The total cell numbers shown in Fig. 1B were obtained by multiplying To determine the effects of FTY720 on innate immune cells, we the percentage positive of each cell type (as determined by flow cytometry) by the whole blood cell count obtained from the Hemavet, then dividing dosed mice with FTY720 at 0.03, 0.3, and 3 mg/kg and performed that number by 100. For Fig. 2B, the total cell counts were obtained by a differential analysis of peripheral blood collected at 4 and 24 h multiplying the percentage of monocytes by the total cell concentrations. after dosing. In agreement with previous reports, using size-based The cell concentrations were derived by taking the total number of cells analysis (Hemavet), we found a significant, dose-dependent de- run through the flow cytometer for each sample, divided by the run time of the sample to get cells/time. This number was then divided by the speed of crease in circulating lymphocytes at both time points when com- the cytometer to get the cell concentration. The cell counts were then pared with vehicle alone (up to an 87% reduction). We also observed multiplied by their respective dilutions and the total volume of the cell a significant reduction in circulating monocytes. Monocyte levels suspension. Mice were assumed to have a total blood volume of 58.5 ml/kg were decreased up to 78% in a dose-dependent manner with body weight. FTY720 treatment (Fig. 1A). We also monitored circulating S1P receptor expression analysis neutrophil levels and found that there was no significant effect in response to FTY720 treatment (Fig. 1A). S1P receptor expression data were obtained from publicly available micro- Using flow cytometry, we confirmed the lymphocyte and mono- array datasets using NextBio [www.nextbio.com (24)]. The data were pro- cessed and normalized by NextBio to compare between different datasets, as cyte lowering in mice and extended our analysis to additional cell described. Expression levels of S1P receptors were obtained for CD8+ Tcells types. Mice were dosed with 0.03, 0.1, 0.3, and 3 mg/kg FTY720, (GSE11446), B220+ B cells (GSE6980), NK1.1+ NK cells (GSE6506), and whole blood immune cell counts were assessed by flow + + granulocytes (Mac-1 Gr-1 ; GSE10246), Ly6C high and low monocytes cytometry. We confirmed that CD4+ (97% reduction) and CD8+ (GSE17256), and corticotectal neurons (GSE17784). All cell types were + obtained from C57BL/6 mice. Data are presented as the mean 6 SD. T cells (91% reduction) along with B220 B cells (60% reduction) were decreased in peripheral blood 24 h after dosing with all tested Monocyte enrichment and stimulation doses of FTY720 (Fig. 1B). Significant decreases in monocytes + + + + Blood was isolated from C57BL/6 mice and lysed using RBC lysis buffer (CD11b CD115 ; 56% reduction) and NK cells (NK1.1 NKp46 ; (BioLegend), according to the manufacturer’s instructions. The EasySep 43% reduction) were also detected; however, no effect on neu- Mouse Monocyte Enrichment kit (Stemcell Technologies) was used to trophils (CD11b+ Gr-1+) was observed (Fig. 1B). Thus, the number enrich for monocytes. This was performed according to the manufacturer’s of circulating monocytes and NK cells is reduced following instructions. Cell purity was assessed by CD11b+ Ly6G2 according to the manufacturer’s instructions and was on average 90.2% pure. FTY720 with a similar time course as lymphocytes. Monocytes were cultured at 37˚C in RPMI 1640 medium containing We dosed rats with FTY720 at 0.03, 0.1, and 0.3 mg/kg and 10% FBS, 10% L929-conditioned medium, 13 penicillin/streptomycin, performed a differential analysis of peripheral blood collected at 4 The Journal of Immunology 3535 Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 1. FTY720 reduces circulating lymphocyte, monocyte, and NK cell levels, but not neutrophils. (A) C57BL/6 mice were dosed with vehicle, 0.03 mg/kg, 0.3 mg/kg, or 3 mg/kg FTY720 (n = 7 mice per group). Whole blood was taken at 0, 4, and 24 h after the administration of the dose. Lymphocyte, neutrophil, and monocyte cell counts were assessed using the Hemavet. (B) C57BL/6 mice were dosed with vehicle, 0.03, 0.1, 0.3, or 3 mg/kg FTY720. After 24 h, whole blood was collected and CD4+ T cells, CD8+ T cells, B220+ B cells, CD11b+ CD115+ monocytes, NK1.1+ NKp46+ NK cells, and CD11b+ Gr-1+ neutrophil cell counts were assessed by flow cytometry (n = 29–30 mice in the untreated and vehicle groups; n = 10 in the 0.03, 0.1, and 0.3 mg/kg groups; and n = 20 in the 3 mg/kg group). (C) Lewis rats were dosed with vehicle, 0.03 mg/kg, 0.1 mg/kg, or 0.3 mg/kg FTY720. Whole blood was taken at 0, 4, and 24 h after the administration of the dose. Lymphocyte, neutrophil, and monocyte cell counts were assessed using the Hemavet (n = 12 rats for the 0-h time point, and n = 5 rats per group for all other time points). *p , 0.05, **p , 0.01, ***p , 0.001 versus vehicle. Mean 6 SEM is shown. and 24 h after dosing. Similar to our results in mice, we found monocyte levels were also altered in lymphoid tissues. We dosed a significant, dose-dependent decrease in circulating lymphocytes at mice with FTY720 at 0.3 mg/kg and 2 mg 5-ethynyl-29-deoxy- both time points when compared with vehicle alone (up to an 88% uridine (EdU) (a 5-bromo-29-deoxyuridine alternative) and iso- reduction; Fig. 1C). Monocyte levels were also decreased in a dose- lated the blood, spleen, and bone marrow 4 h later and performed dependent manner with FTY720 treatment up to an 80% reduction flow cytometry to assess monocyte levels in these tissues. In (Fig. 1C). We also monitored circulating neutrophil levels by addition to our observation that FTY720 reduced circulating Hemavet, and, although a decrease in response to FTY720 treat- monocyte levels, we found that FTY720 causes a significant ac- ment was observed, it was not a significant effect (Fig. 1C). cumulation of monocytes in both the spleen and bone marrow (Fig. 2B). This effect was observed for both Ly6Chigh and Ly6Clow FTY720 restricts monocyte egress from the spleen and bone monocytes. Additionally, we calculated the sum of the total marrow monocytes in these tissues and found that there were significantly Due to the decreased levels of circulating monocytes observed more Ly6Chigh monocytes, but not Ly6Clow monocytes (Fig. 2C). upon administration of FTY720, we wanted to determine whether These data suggest that for the Ly6Clow monocytes, the changes 3536 FTY720 REDUCES MONOCYTE NUMBERS IN THE CIRCULATION we observed were due to cell relocation. However, the observed demonstrate that FTY720 acts directly on monocytes to alter changes for Ly6Chigh monocytes could be due to increased pro- surface protein expression and cytokine production. liferation or reduced cell turnover. We next analyzed the number S1P receptor modulation in vivo alters monocyte surface of EdU+ bone marrow monocytes and found that FTY720 in- protein expression and cytokine production creased their numbers, suggesting an increase in proliferation (Fig. 2D). This effect, however, was restricted to the Ly6Chigh subset, To assess whether S1P receptor modulation alters monocyte ac- thus explaining our results found in Fig. 2C. These results dem- tivation in vivo, mice were dosed with 0.3 mg/kg FTY720. Twenty- onstrate that FTY720 causes monocytes to accumulate in the four hours later, we collected peripheral blood and assessed the spleen and bone marrow while also causing increased proliferation cell surface expression of multiple proteins on monocyte subsets. in the bone marrow. We found that on Ly6Chigh monocytes, PD-L1, CD40, and CD69 were decreased by FTY720, whereas CD80 was increased (Fig. FTY720 directly restricts monocyte activation 4A). On Ly6Clow monocytes, CD40 and CD69 were also decreased Our previous results demonstrate that FTY720 changes monocyte by FTY720 compared with control. We also monitored cell apo- levels by reducing their numbers in the circulation and increasing ptosis and necrosis using annexin V and propidium iodide and their numbers in the spleen and bone marrow. These data, however, found no significant difference between the groups (data not do not address whether this effect is direct or indirect. The first step shown). These results suggest that FTY720 alters monocyte sur- we took to address this was to analyze S1P receptor expression on face protein expression in vivo and may affect monocyte function. monocyte subsets and other immune cells by accessing and ana- Because we demonstrated that FTY720 reduced circulating lyzing publicly available microarray datasets using NextBio (Fig. levels of monocytes and altered their surface protein expression and Downloaded from 3A). These data demonstrate that monocytes express S1P recep- cytokine production, we wanted to determine whether S1P receptor tors with the highest levels observed for S1pr2 and S1pr5. These modulation would have an impact on serum cytokine levels after an data also confirm previous publications that lymphocytes express inflammatory stimulus in vivo. To test this, we dosed mice with 0.3 high levels of S1pr1 and NK cells have high levels of S1pr5. mg/kg FTY720 and BAF312, a S1P1,4,5 modulator, for 22 h and Expression of S1pr4 was not included in these datasets; never- challenged them with 200 ng LPS. After 2 h, we collected whole

theless, other studies have shown low-level expression of S1pr4 on blood and assessed the levels of proinflammatory cytokines. We http://www.jimmunol.org/ monocytes as compared with other immune cells (25). found that induction of IL-6, TNF-a, CCL2, and IL-10 by LPS To further determine whether FTY720 acts directly upon was significantly decreased by FTY720 and BAF312 (Fig. 4B). monocytes, we enriched monocytes from the blood (∼90.2% pure) These results demonstrate that S1P receptor modulation decreases and cultured them in the presence of FTY720 (10 nM) and LPS cytokine production after an LPS challenge. (10 ng/ml). After 2 h, we found FTY720 reduced CD40 protein levels on monocytes (Fig. 3B). In the presence of LPS, there was FTY720 reduces monocytes during EAE an induction of CD40 levels that was significantly reduced with To determine whether the impact of FTY720 on monocytes is FTY720. Furthermore, we analyzed cytokine production and transient or sustained in a long-term model of inflammation, we found LPS induced the production of TNF-a, which was signifi- induced EAE in both mice and rats and measured the effects of 0.03 by guest on September 26, 2021 cantly reduced in the presence of FTY720 (Fig. 3C). These data and 0.3 mg/kg FTY720 on clinical disease and circulating levels

FIGURE 2. FTY720 restricts monocyte egress from the spleen and bone marrow while enhancing proliferation. C57BL/6 mice were orally dosed with vehicle or 0.3 mg/kg FTY720 and simulta- neously given an i.p. injection of 2 mg EdU. After 4 h, the blood, spleen, and bone marrow were isolated. Single-cell suspensions were made and analyzed by flow cytometry. (A) Monocyte gating is shown for each tissue. (B) Total cell counts are shown for total monocytes, Ly6Chigh monocytes, and Ly6Clow monocytes in the blood, spleen, and bone marrow. (C) The sum of the cell counts is shown for Ly6Chigh and Ly6Clow monocytes. (D) Gating of bone marrow EdU+ monocytes and total cell counts are shown. n = 8–10 mice per group. *p , 0.05, **p , 0.01, ***p , 0.001 versus vehicle. The Journal of Immunology 3537

FIGURE 3. FTY720 acts directly on monocytes to alter cell surface expression and cytokine pro- duction. (A) S1P receptor mRNA expression is shown for the indicated cell types. Data were ob- tained from publicly available microarray datasets using NextBio, as described in Materials and Methods.(B and C) Monocytes were isolated from peripheral blood and cultured for 2 h in the presence of 10 nM FTY720 and 10 ng/ml LPS. (B) CD40 levels measured by flow cytometry (n = 8). (C) TNF- a cytokine levels measured by sandwich immuno- assay (n = 14). *p , 0.05 versus vehicle and ###p , 0.001 versus LPS. Downloaded from

of lymphocytes, monocytes, and neutrophils. When treated pro- (reductions of 89, 66, and 47%, respectively) and 0.3 mg/kg (88, 70, phylactically, FTY720 suppressed clinical disease in mice at 0.3 and 64%, respectively; Fig. 5C). mg/kg (Fig. 5A). Peripheral blood was analyzed at day 14 post- Monocyte trafficking is not dependent on S1P3 immunization, and the levels of lymphocytes, monocytes, and http://www.jimmunol.org/ neutrophils were significantly reduced in mice at both tested S1P3 has been implicated in controlling monocyte trafficking (10); doses, as follows: 0.03 mg/kg (reductions of 56, 61, and 56%, therefore, we wanted to assess whether decreased circulating respectively) and 0.3 mg/kg (reductions of 86, 82, 74, respec- monocytes could be achieved with a S1P that tively; Fig. 5A). does not agonize S1P3. We synthesized BAF312, a S1P1,4,5 agonist, We also treated EAE mice with a therapeutic regimen of FTY720 confirmed its potency (EC50) in vitro with GTPgS assays for S1P (i.e., daily dosing began after mice developed disease; Fig. 5B). receptors, and demonstrated potencies of 0.2 nM for S1P1, .10 mM Although disease was not completely abolished at 0.3 mg/kg, the for S1P3, and 0.5 nM for S1P5. FTY720 was assayed in parallel clinical score was significantly decreased, which correlated with and demonstrated potencies of 2 nM for S1P1, 175 nM for S1P3, by guest on September 26, 2021 decreased levels of circulating monocytes and lymphocytes. These and 2.5 nM for S1P5 (L. Patnaude, S. Lukas, S. Haxhinasto, and results indicate that FTY720 also restricts circulating levels of L. Modis, manuscript in preparation). We tested BAF312 for po- monocytes under sustained inflammatory conditions. tency in naive (0.03, 0.3, and 3 mg/kg) and EAE (0.03 and 0.3 In rats, prophylactic treatment with 0.03 and 0.3 mg/kg FTY720 mg/kg) conditions. In naive mice, similar to FTY720, we observed significantly reduced clinical disease. Peripheral blood was ana- significant decreases in lymphocytes and monocytes at 4 h that lyzed at the end of the study, and lymphocytes, monocytes, and continued through 24 h (reductions of 83 and 86%, respectively; neutrophils were also significantly reduced at doses of 0.03 mg/kg Fig. 6A), and we did not observe any effects on neutrophils. In

FIGURE 4. S1P receptor modulation al- tered monocyte protein expression and re- duced serum cytokine levels after a LPS challenge. (A) C57BL/6 mice were dosed with 0.3 mg/kg FTY720 or vehicle, and, after 24 h, peripheral blood was isolated. Monocyte subsets (Ly6Chigh and Ly6Clow) were assessed by flow cytometry for cell surface expression of several proteins. Mean fluorescence intensity data are shown (n =5– 6 mice per group). *p , 0.05, **p , 0.1 versus vehicle. (B)C57BL/6micewere dosed with 0.3 mg/ml FTY720, BAF312, or vehicle. After 22 h, mice received an i.p. injection of 200 ng LPS or vehicle. Two hours later, whole blood was collected, and IL-6, TNF-a, IL-10, and CCL2 were as- sessed by a sandwich immunoassay. *p , 0.05, **p , 0.01, ***p , 0.001. 3538 FTY720 REDUCES MONOCYTE NUMBERS IN THE CIRCULATION Downloaded from

FIGURE 5. FTY720 suppresses clinical signs of disease in mouse and rat EAE while decreasing lymphocytes, monocytes, and neutrophils. EAE was induced in both C57BL/6 mice and Lewis rats, as described in Materials and Methods. Animals with EAE were either untreated, or treated daily with a vehicle, 0.03 mg/kg, or 0.3 mg/kg FTY720. (A) Prophylactic dosing of mice began 1 d before immunization, and clinical scores were evaluated daily beginning on day 7 postimmunization. Whole blood was assessed for circulating lymphocytes, monocytes, and neutrophils at day 14 postimmunization. For http://www.jimmunol.org/ clinical score data, n = 9–10 mice per group. For cell counts, n = 4–5 mice per group. (B) Therapeutic dosing of mice began once mice had a clinical score .1 or a score of 1 for at least 2 d. Whole-blood assessment of circulating lymphocytes, monocytes, and neutrophils was performed at day 16 post- immunization. For clinical score data, n = 16–21 mice per group. For cell counts, n = 4–5 mice per group. (C) Prophylactic dosing of rats began 1 d before immunization, and clinical scores were evaluated daily beginning on day 7 postimmunization. Whole-blood assessment of circulating lymphocytes, monocytes, and neutrophils was performed at day 20 postimmunization. For clinical score data, n = 10 rats per group. For cell counts, n = 5 rats per group. Mean 6 SEM is shown. Significant differences in the clinical score data were assessed using the AUC data from individual mice. #p , 0.05 for 0.03 mg/kg versus vehicle, **p , 0.01, and ***p , 0.001 for 0.3 mg/kg versus vehicle.

EAE, we demonstrated that BAF312 significantly reduced clinical atherosclerotic plaques in ApoE-deficient mice (18), peritoneal by guest on September 26, 2021 scores when dosed prophylactically or therapeutically in mice at cavity with thioglycolate-induced peritonitis (10), peripheral 0.3 mg/kg (Fig. 6B, 6C). As expected, lymphocyte levels were nerves in experimental autoimmune neuritis (15), CNS in EAE significantly reduced when assessed near the peak of disease (16), renal glomeruli in experimental mesangioproliferative glo- (i.e., days 14 and 16 postimmunization for the prophylactic [73% merulonephritis (17), and heart allografts (19). However, no reduction] and therapeutic [37% reduction] regimens, respec- mechanism has been demonstrated. Given that tissue macrophages tively). We also observed significant decreases in monocyte lev- typically derive from circulating monocytes, and under conditions els, up to a 61% reduction with the prophylactic regimen, and up of tissue injury there is increased monocyte trafficking into dis- to a 57% reduction with the therapeutic regimen. BAF312 also eased tissues (26, 27), our results demonstrating decreased cir- inhibited clinical disease in rats in a dose-dependent manner, culating monocyte levels upon FTY720 administration provide an which correlated with reduced levels of circulating lymphocytes explanation for these findings. and monocytes (92 and 66% reductions, respectively; Fig. 6D). Our observation that FTY720 altered several cell surface These results indicate that direct S1P3 agonism is not necessary markers on monocytes suggests that FTY720 may alter monocyte for an impact on monocyte numbers in circulation. function. In both the Ly6Chigh and Ly6Clow monocyte subset, we observed significant decreases in CD69 expression. CD69 has Discussion been demonstrated to interact with S1P1 in T cells and inhibit In this study, we demonstrate in mice and rats using two different T cell responsiveness to S1P (28). Because FTY720 causes in- S1P receptor modulators (FTY720 and BAF312) that S1P receptors ternalization and degradation of S1P1, CD69 may be internalized are involved in maintaining monocyte numbers in the circulation also due to their interaction. Additionally, ligation of CD69 on under normal and inflammatory conditions. We further demonstrate monocytes triggers monocyte activation, which may be reduced that FTY720 causes monocytes to accumulate in the spleen and by FTY720 (29). Similarly, CD40 was also reduced on both bone marrow. These observed effects on monocytes are similar to monocyte subsets, further suggesting that FTY720 may decrease the effects observed on lymphocytes, that S1P receptor modulators monocyte activation. These data are also substantiated by our restrict egress from hematopoietic organs. We also demonstrate that observation that FTY720 reduced monocyte cytokine production FTY720 acts directly upon isolated monocytes to alter their ac- and serum cytokine levels after an in vivo challenge with LPS. We tivation, as demonstrated by reduced surface protein expression and hypothesize that this reduced serum response can be mediated by cytokine production. Similar results were also observed in vivo. both reduced monocyte numbers and activation state. Thus, our findings reveal that S1P receptors play a significant role Pathogenesis in EAE models can be blocked by either reducing in monocyte trafficking and may also alter monocyte function. T cells (30) or monocytes, as illustrated by treating mice and rats A reduction in the number of macrophages has been observed in with clodronate liposomes (31). Similarly, GM-CSF2/2 mice, several tissues of FTY720-treated mice, including the following: which have a monocytic defect, are protected from EAE (30–32). The Journal of Immunology 3539 Downloaded from http://www.jimmunol.org/

FIGURE 6. BAF312 restricts lymphocyte and monocyte circulation and suppresses clinical signs of disease in mouse and rat EAE. (A) C57BL/6 naive mice were dosed with vehicle, 0.03 mg/kg, 0.3 mg/kg, or 3 mg/kg BAF312. Whole blood was taken at 0, 4, and 24 h after the administration of the dose. by guest on September 26, 2021 Lymphocyte, neutrophil, and monocyte cell counts were assessed using the Hemavet (n = 7 mice per group). *p , 0.05, **p , 0.01, ***p , 0.001 versus vehicle. (B–D) EAE was induced in both C57BL/6 mice and Lewis rats, as described in Materials and Methods. Animals with EAE were either untreated, or treated daily with a vehicle, 0.03 mg/kg, or 0.3 mg/kg BAF312. (B) Prophylactic dosing of mice began 1 d before immunization, and clinical scores were evaluated daily beginning on day 7 postimmunization. Whole-blood assessment of circulating lymphocytes, monocytes, and neutrophils was performedat day 14 postimmunization. For clinical score data, n = 9–10 mice per group. For cell counts, n = 4–5 mice per group. (C) Therapeutic dosing of mice began once mice had a clinical score .1 or a score of 1 for at least 2 d. Whole-blood assessment of circulating lymphocytes, monocytes, and neutrophils was performed at day 16 postimmunization. For clinical score data, n = 16–21 mice per group. For cell counts, n = 4–5 mice per group. (D) Prophylactic dosing of rats began 1 d before immunization, and clinical scores were evaluated daily beginning on day 7 postimmunization. Whole-blood assessment of cir- culating lymphocytes, monocytes, and neutrophils was performed at day 20 postimmunization. For clinical score data, n = 10 rats per group. For cell counts, n = 5 rats per group. The untreated and vehicle-treated groups shown in (A–C) are the same data as shown in Figs. 1A, 5A, and 5B because the compounds FTY720 and BAF312 were given during the same experiment. Mean 6 SEM is shown. Significant differences in the clinical score data were assessed using the AUC data from individual mice. *p , 0.05, **p , 0.01, ***p , 0.001 for 0.3 mg/kg versus vehicle.

It has also been shown that inhibition of the M-CSF receptor trafficking is dependent upon S1P3 (10). However, when the authors 2/2 (CD115) reduces clinical disease during EAE (33). Additionally, compared wild-type mice containing S1P3 bone marrow with 2/2 2/2 monocytes from CCR2 mice are reduced in number and un- S1P3 mice containing wild-type bone marrow, no difference was able to traffic into the CNS (34); consequently, these mice are also observed in macrophage infiltration into the peritoneum. This ob- protected from EAE (35). Our data are consistent with these servation, along with our data utilizing BAF312, suggests that high observations, as we see reductions in the proinflammatory Ly6C monocyte expression of S1P3 is not solely responsible for monocyte monocyte subset, which would be expected to directly affect the trafficking. levels of macrophages at sites of inflammation. Furthermore, the Although other groups have mentioned that they did not observe reduction in monocyte numbers parallels that of lymphocytes, both reductions in circulating monocytes with FTY720 treatment (15, in timing and with respect to the PK/PD effect (both are maximally 18, 20, 36, 37), monocytes were not the focus of these studies, reduced at 0.3 mg/kg in healthy mice). Therefore, the reduction in monocyte levels were assessed in a disease model in which the number of circulating monocytes may be partially responsible monocytes are already affected (18), or surface markers were used for the observed efficacy of S1P receptor modulators in EAE. that are not specific for monocytes (36). Our studies, however, We demonstrated that BAF312, which does not modulate S1P3, provide an in-depth analysis of the effects of FTY720 and BAF312 reduced circulating monocyte levels in a dose-dependent manner on monocytes at several doses under multiple conditions and similar to FTY720, suggesting that monocytes traffic indepen- were consistent in both mice and rats. To confirm that the role of dently of S1P3. This is in contrast to a report that monocyte FTY720 on monocytes trafficking is direct, a future step using 3540 FTY720 REDUCES MONOCYTE NUMBERS IN THE CIRCULATION mice in which S1P receptors are specifically deleted from myeloid 17.Martini,S.,S.Kra¨mer, T. Loof, Y. Wang-Rosenke, U. Daig, K. Budde, H.-H. Neumayer, and H. Peters. 2007. S1P modulator FTY720 limits matrix cells in adult mice in an inducible manner is necessary. expansion in acute anti-thy1 mesangioproliferative glomerulonephritis. Am. J. In summary, our data indicate that S1P receptor modulators Physiol. Renal Physiol. 292: F1761–F1770. lead to a significant reduction in circulating monocyte levels in 18. Keul, P., M. To¨lle, S. Lucke, K. von Wnuck Lipinski, G. Heusch, M. Schuchardt, M. van der Giet, and B. Levkau. 2007. The sphingosine-1-phosphate analogue mice and rats. Although an impact of FTY720 on monocytes in FTY720 reduces atherosclerosis in apolipoprotein E-deficient mice. Arterioscler. humans at therapeutic doses (0.5 mg in multiple sclerosis and up Thromb. Vasc. Biol. 27: 607–613. to 5 mg in trials) has not been reported, cynomologous monkeys 19. Ogawa, R., M. Takahashi, S. Hirose, H. Morimoto, H. Ise, T. Murakami, T. Yasue, K. Kuriyama, M. Hongo, E. Kobayashi, and U. Ikeda. 2007. A novel treated with FTY720 at 0.5 mg/kg (∼2 mg per animal) showed – sphingosine-1-phosphate receptor agonist KRP-203 attenuates rat autoimmune a reduction in circulating monocyte levels (38). Given the expres- myocarditis. Biochem. Biophys. Res. Commun. 361: 621–628. 20.Singer,I.I.,M.Tian,L.A.Wickham,J.Lin,S.S.Matheravidathu, sion of S1P receptors on human monocytes and macrophages (11, M. J. Forrest, S. Mandala, and E. J. Quackenbush. 2005. Sphingosine-1- 39), additional studies on monocyte levels and functional markers in phosphate agonists increase macrophage homing, lymphocyte contacts, and humans may be warranted. In conclusion, we have performed a endothelial junctional complex formation in murine lymph nodes. J. Immunol. 175: 7151–7161. systematic, carefully controlled analysis and uncovered novel 21. Gergely, P., B. Nuesslein-Hildesheim, D. Guerini, V. Brinkmann, M. Traebert, roles of S1P receptor modulation on monocytes in rodents. C. Bruns, S. Pan, N. S. Gray, K. Hinterding, N. G. Cooke, et al. 2012. The se- lective sphingosine 1-phosphate receptor modulator BAF312 redirects lymphocyte distribution and has species-specific effects on heart rate. Br.J.Pharmacol.167: 1035–1047. Disclosures 22. Pan, S., W. Gao, N. Gray, Y. Mi, and Y. Fan, and The Genomics Institute of the The authors are employed by Boehringer Ingelheim Pharmaceuticals. Novartis Research Foundation. 2004. Immunosuppressant compounds and com- positions. World Intellectual Property Organization, Geneva, Switzerland. 23. Liu, Y., D. Papoutsakis, and E. Roddy, and Novartis Pharma AG. 2010 Downloaded from New polymorphic form of 1-(4-{ L-[ (E)-4-cyclohexyl–3-trifluoromethyl-ben- References zyloxyimino]-ethyl)-2-ethyl-bezy L)-azetidine-3-carboxylic. World Intellectual 1. Cyster, J. G., and S. R. Schwab. 2012. Sphingosine-1-phosphate and lymphocyte Property Organization, Geneva, Switzerland. egress from lymphoid organs. Annu. Rev. Immunol. 30: 69–94. 24. Kupershmidt, I., Q. J. Su, A. Grewal, S. Sundaresh, I. Halperin, J. Flynn, 2. Rivera, J., R. L. Proia, and A. Olivera. 2008. The alliance of sphingosine-1- M. Shekar, H. Wang, J. Park, W. Cui, et al. 2010. Ontology-based meta-analysis phosphate and its receptors in immunity. Nat. Rev. Immunol. 8: 753–763. of global collections of high-throughput public data. PLoS One 5: e13066. 3. Brinkmann, V., A. Billich, T. Baumruker, P. Heining, R. Schmouder, G. Francis, 25. Gautier, E. L., T. Shay, J. Miller, M. Greter, C. Jakubzick, S. Ivanov, J. Helft,

S. Aradhye, and P. Burtin. 2010. (FTY720): discovery and devel- A. Chow, K. G. Elpek, S. Gordonov, et al; Immunological Genome Consortium. http://www.jimmunol.org/ opment of an oral drug to treat multiple sclerosis. Nat. Rev. Drug Discov. 9: 883– 2012. Gene-expression profiles and transcriptional regulatory pathways that 897. underlie the identity and diversity of mouse tissue macrophages. Nat. Immunol. 4. Cyster, J. G. 2005. Chemokines, sphingosine-1-phosphate, and cell migration in 13: 1118–1128. secondary lymphoid organs. Annu. Rev. Immunol. 23: 127–159. 26. Iijima, N., L. M. Mattei, and A. Iwasaki. 2011. Recruited inflammatory mono- 5. Kappos, L., J. Antel, G. Comi, X. Montalban, P. O’Connor, C. H. Polman, cytes stimulate antiviral Th1 immunity in infected tissue. Proc. Natl. Acad. Sci. T. Haas, A. A. Korn, G. Karlsson, and E. W. Radue; FTY720 D2201 Study USA 108: 284–289. Group. 2006. Oral fingolimod (FTY720) for relapsing multiple sclerosis. 27. Shi, C., and E. G. Pamer. 2011. Monocyte recruitment during infection and N. Engl. J. Med. 355: 1124–1140. inflammation. Nat. Rev. Immunol. 11: 762–774. 6. Jolly, P. S., M. Bektas, A. Olivera, C. Gonzalez-Espinosa, R. L. Proia, J. Rivera, 28. Shiow, L. R., D. B. Rosen, N. Brdickova´, Y. Xu, J. An, L. L. Lanier, J. G. Cyster, S. Milstien, and S. Spiegel. 2004. Transactivation of sphingosine-1-phosphate and M. Matloubian. 2006. CD69 acts downstream of interferon-a/b to inhibit receptors by FcepsilonRI triggering is required for normal mast cell degranu- S1P1 and lymphocyte egress from lymphoid organs. Nature 440: 540–544.

lation and chemotaxis. J. Exp. Med. 199: 959–970. 29. De Maria, R., M. G. Cifone, R. Trotta, M. R. Rippo, C. Festuccia, A. Santoni, by guest on September 26, 2021 7. Maeda, Y., H. Matsuyuki, K. Shimano, H. Kataoka, K. Sugahara, and K. Chiba. and R. Testi. 1994. Triggering of human monocyte activation through CD69, 2007. Migration of CD4 T cells and dendritic cells toward sphingosine 1- a member of the natural killer cell gene complex family of signal transducing phosphate (S1P) is mediated by different receptor subtypes: S1P regulates the receptors. J. Exp. Med. 180: 1999–2004. functions of murine mature dendritic cells via S1P receptor type 3. J. Immunol. 30. Codarri, L., G. Gyu¨lve´szi, V. Tosevski, L. Hesske, A. Fontana, L. Magnenat, 178: 3437–3446. T. Suter, and B. Becher. 2011. RORgt drives production of the cytokine GM-CSF 8. Roviezzo, F., F. Del Galdo, G. Abbate, M. Bucci, B. D’Agostino, E. Antunes, in helper T cells, which is essential for the effector phase of autoimmune neu- G. De Dominicis, L. Parente, F. Rossi, G. Cirino, and R. De Palma. 2004. Human roinflammation. Nat. Immunol. 12: 560–567. eosinophil chemotaxis and selective in vivo recruitment by sphingosine 1- 31. King, I. L., T. L. Dickendesher, and B. M. Segal. 2009. Circulating Ly-6C+ phosphate. Proc. Natl. Acad. Sci. USA 101: 11170–11175. myeloid precursors migrate to the CNS and play a pathogenic role during au- 9. Kveberg, L., Y. Bryceson, M. Inngjerdingen, B. Rolstad, and A. A. Maghazachi. toimmune demyelinating disease. Blood 113: 3190–3197. 2002. Sphingosine 1 phosphate induces the chemotaxis of human natural killer 32. El-Behi, M., B. Ciric, H. Dai, Y. Yan, M. Cullimore, F. Safavi, G.-X. Zhang, cells: role for heterotrimeric G proteins and phosphoinositide 3 . Eur. J. B. N. Dittel, and A. Rostami. 2011. The encephalitogenicity of T(H)17 cells is Immunol. 32: 1856–1864. dependent on IL-1- and IL-23-induced production of the cytokine GM-CSF. Nat. 10. Keul, P., S. Lucke, K. von Wnuck Lipinski, C. Bode, M. Gra¨ler, G. Heusch, and Immunol. 12: 568–575. B. Levkau. 2011. Sphingosine-1-phosphate receptor 3 promotes recruitment of 33. Uemura, Y., H. Ohno, Y. Ohzeki, H. Takanashi, H. Murooka, K. Kubo, and monocyte/macrophages in inflammation and atherosclerosis. Circ. Res. 108: I. Serizawa. 2008. The selective M-CSF receptor tyrosine inhibitor 314–323. Ki20227 suppresses experimental autoimmune encephalomyelitis. J. Neuro- 11. Ulrych, T., A. Bo¨hm, A. Polzin, G. Daum, R. M. Nu¨sing, G. Geisslinger, immunol. 195: 73–80. T. Hohlfeld, K. Schro¨r, and B. H. Rauch. 2011. Release of sphingosine-1- 34. Mildner, A., H. Schmidt, M. Nitsche, D. Merkler, U.-K. Hanisch, M. Mack, phosphate from human platelets is dependent on thromboxane formation. J. M. Heikenwalder, W. Bru¨ck, J. Priller, and M. Prinz. 2007. Microglia in the adult Thromb. Haemost. 9: 790–798. brain arise from Ly-6ChiCCR2+ monocytes only under defined host conditions. 12. Walzer, T., L. Chiossone, J. Chaix, A. Calver, C. Carozzo, L. Garrigue-Antar, Nat. Neurosci. 10: 1544–1553. Y. Jacques, M. Baratin, E. Tomasello, and E. Vivier. 2007. Natural killer cell 35. Fife, B. T., G. B. Huffnagle, W. A. Kuziel, and W. J. Karpus. 2000. CC che- trafficking in vivo requires a dedicated sphingosine 1-phosphate receptor. Nat. mokine receptor 2 is critical for induction of experimental autoimmune en- Immunol. 8: 1337–1344. cephalomyelitis. J. Exp. Med. 192: 899–905. 13. Jenne, C. N., A. Enders, R. Rivera, S. R. Watson, A. J. Bankovich, J. P. Pereira, 36. Ishii, M., J. G. Egen, F. Klauschen, M. Meier-Schellersheim, Y. Saeki, J. Vacher, Y. Xu, C. M. Roots, J. N. Beilke, A. Banerjee, et al. 2009. T-bet-dependent S1P5 R. L. Proia, and R. N. Germain. 2009. Sphingosine-1-phosphate mobilizes os- expression in NK cells promotes egress from lymph nodes and bone marrow. teoclast precursors and regulates bone homeostasis. Nature 458: 524–528. J. Exp. Med. 206: 2469–2481. 37. Liesz, A., L. Sun, W. Zhou, S. Schwarting, E. Mracsko, M. Zorn, H. Bauer, 14. Rathinasamy, A., N. Czeloth, O. Pabst, R. Fo¨rster, and G. Bernhardt. 2010. The C. Sommer, and R. Veltkamp. 2011. FTY720 reduces post-ischemic brain lymphocyte origin and maturity of dendritic cells determine the pattern of sphingosine 1- influx but does not improve outcome in permanent murine cerebral ischemia. PLoS phosphate receptors expressed and required for efficient migration. J. Immunol. One 6: e21312. 185: 4072–4081. 38. Food and Drug Administration Peripheral and Central Nervous System 15. Zhang, Z., Z.-Y. Zhang, U. Fauser, and H. J. Schluesener. 2008. FTY720 ame- Drugs Advisory Committee. Fingolimod (NDA 22-527) Background Package. liorates experimental autoimmune neuritis by inhibition of lymphocyte and June 10, 2010. Available at: http://www.fda.gov/downloads/advisorycommittees/ monocyte infiltration into peripheral nerves. Exp. Neurol. 210: 681–690. committeesmeetingmaterials/drudr/peripheralandcentralnervoussystemdrugs 16. Rausch, M., P. Hiestand, C. A. Foster, D. R. Baumann, C. Cannet, and M. Rudin. advisorycommittee/ucm214670.pdf. Accessed: March 5, 2012. 2004. Predictability of FTY720 efficacy in experimental autoimmune encepha- 39. Duong, C. Q., S. M. Bared, A. Abu-Khader, C. Buechler, A. Schmitz, and lomyelitis by in vivo macrophage tracking: clinical implications for ultrasmall G. Schmitz. 2004. Expression of the lysophospholipid receptor family and in- superparamagnetic iron oxide-enhanced magnetic resonance imaging. J. Magn. vestigation of lysophospholipid-mediated responses in human macrophages. Reson. Imaging 20: 16–24. Biochim. Biophys. Acta 1682: 112–119.