0022-3565/07/3232-469–476$20.00
`THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS
`Copyright © 2007 by The American Society for Pharmacology and Experimental Therapeutics
`JPET 323:469–476, 2007
`
`Vol. 323, No. 2
`127183/3263802
`Printed in U S.A.
`
`Brain Penetration of the Oral Immunomodulatory Drug FTY720
`and Its Phosphorylation in the Central Nervous System during
`Experimental Autoimmune Encephalomyelitis: Consequences
`for Mode of Action in Multiple Sclerosis
`
`Carolyn A. Foster, Laurence M. Howard, Alain Schweitzer, Elke Persohn,
`Peter C. Hiestand, Bala´ zs Balatoni,1 Roland Reuschel, Christian Beerli, Manuela Schwartz,2
`and Andreas Billich
`Novartis Institutes for BioMedical Research, Vienna, Austria (C.A.F., L.M.H., B.B., R.R., M.S., A.B.); Novartis Institutes for
`BioMedical Research, Basel, Switzerland (P.C.H., C.B.); Novartis Pharma AG, Basel, Switzerland (A.S.); and Novartis Pharma
`AG, Muttenz, Switzerland (E.P.)
`Received June 24, 2007; accepted August 3, 2007
`
`ABSTRACT
`FTY720
`[2-amino-2-[2-(4-octylphenyl)ethyl]propane-1,3-diol
`hydrochloride]
`is an oral sphingosine-1-phosphate receptor
`modulator under development for the treatment of multiple
`sclerosis (MS). The drug is phosphorylated in vivo by sphin-
`gosine kinase 2 to its bioactive form, FTY720-P. Although treat-
`ment with FTY720 is accompanied by a reduction of the pe-
`ripheral lymphocyte count, its efficacy in MS and experimental
`autoimmune encephalomyelitis (EAE) may be due to additional,
`direct effects in the central nervous system (CNS). We now
`show that FTY720 localizes to the CNS white matter, preferen-
`tially along myelin sheaths. Brain trough levels of FTY720 and
`
`FTY720-P in rat EAE are of the same magnitude and dose
`dependently increase; they are in the range of 40 to 540 ng/g in
`the brain tissue at efficacious doses and exceed blood con-
`centrations severalfold. In a rat model of chronic EAE, pro-
`longed treatment with 0.03 mg/kg was efficacious, but limiting
`the dosing period failed to prevent EAE despite a significant
`decrease in blood lymphocytes. FTY720 effectiveness is likely
`due to a culmination of mechanisms involving reduction of
`autoreactive T cells, neuroprotective influence of FTY720-P
`in the CNS, and inhibition of inflammatory mediators in the
`brain.
`
`FTY720 is an oral sphingosine-1-phosphate (S1P) receptor
`modulator (Baumruker et al., 2007) under development for
`the treatment of multiple sclerosis (MS), representing the
`first of a new class of immunomodulatory agents. Promising
`results in phase II trials with relapsing MS patients (Kappos
`et al., 2006) mirror the striking efficacy of FTY720 in MS
`models of experimental autoimmune encephalomyelitis
`(EAE), shown by preventive and therapeutic treatment
`(Brinkmann et al., 2002; Fujino et al., 2003; Webb et al.,
`2004; Kataoka et al., 2005; Balatoni et al., 2007). FTY720 is
`converted in vivo to its biologically active phosphate ester
`
`This work was supported by Novartis Pharma AG.
`1 Current affiliation: Novartis Hungary Healthcare, Budapest, Hungary.
`2 Current affiliation: Evangelisches Krankenhaus, Vienna, Austria.
`Article, publication date, and citation information can be found at
`http://jpet.aspetjournals.org.
`doi:10.1124/jpet.107.127183.
`
`metabolite (FTY720-P), which acts as a high-affinity agonist
`for four of the five known G-protein-coupled S1P receptors,
`namely S1P1 and S1P3–5 (Brinkmann et al., 2002; Mandala
`et al., 2002). Sphingosine kinase (SPHK) 2 is the primary
`enzyme required for FTY720-P formation, as we and others
`subsequently confirmed in SPHK2 knockout mice (Kharel et
`al., 2005; Zemann et al., 2006). The fact that SPHK1 null
`mice become lymphopenic after FTY720 administration fur-
`ther supports the view that SPHK2 is sufficient for the func-
`tional activation of FTY720 (Allende et al., 2004).
`Emerging evidence suggests that the effectiveness of
`FTY720 in the central nervous system (CNS) extends beyond
`immunomodulation to encompass other aspects of MS patho-
`physiology, including an influence on the blood-brain barrier
`and glial repair mechanisms that could ultimately contribute
`to restoration of nerve function (Baumruker et al., 2007;
`
`ABBREVIATIONS: FTY720, fingolimod, 2-amino-2-[2-(4-octylphenyl)ethyl]propane-1,3-diol hydrochloride; MS, multiple sclerosis; EAE, experi-
`mental autoimmune encephalomyelitis; SPHK, sphingosine kinase; CNS, central nervous system; DA, Dark Agouti; CFA, complete Freund’s
`adjuvant; CsA, cyclosporine A; QWBA, quantitative whole-body autoradiography; CSF, cerebrospinal fluid; ANOVA, analysis of variance; AUC,
`area under the curve; FTY720-P, 2-amino-2-[2-(4-octylphenyl)ethyl)propane-1,3-diol-1-(dihydrogen phosphate); FK506, tacrolimus.
`
`469
`
`Apotex v. Novartis
`IPR2017-00854
`NOVARTIS 2038
`
`

`

`470
`
`Foster et al.
`
`Jung et al., 2007; Osinde et al., 2007). A key consideration
`behind this concept is the finding that FTY720 distributes to
`the brain (Meno-Tetang et al., 2006), which contains endog-
`enous SPHK2 for the phosphorylation of FTY720 (Billich et
`al., 2003). Moreover, neurons and glial cells (astrocytes, mi-
`croglia, oligodendrocytes) in the brain differentially express
`S1P receptors (Fig. 1), thus raising the possibility for recep-
`tor activation in situ by FTY720-P. So far, there is no infor-
`mation on the presence of FTY720-P in the brain or potential
`concentrations therein. Our primary aim was to investigate
`the distribution of FTY720 and its phosphorylated form in
`the CNS after clinically relevant doses in two different EAE
`models. We provide preclinical evidence that the bioactive
`metabolite FTY720-P distributes to the CNS white matter,
`suggesting the potential for functional interaction with glial
`cells bearing S1P receptors in the brain and spinal cord.
`
`Materials and Methods
`Animals. For EAE, female RT11 Lewis rats from Charles River
`(Sulzfeld, Germany) and Dark Agouti (DA) rats from Harlan Winkel-
`mann (Borchen, Germany) were kept under standardized light- and
`climate-controlled conditions with free access to food and water.
`Age-matched rats were acclimatized for at least 1 week before dis-
`tribution into the experimental groups. For autoradiography, male
`pigmented LE/CR WIGA rats (Charles River; 198–238 g) were
`housed individually in metabolism cages. All experiments conformed
`to Novartis animal care regulations and were approved by the Aus-
`trian and Swiss health authorities in compliance with international
`animal welfare standards according to the European Communities
`Council Directive and the guidelines set forth in the National Insti-
`tutes of Health Guide for the Care and Use of Laboratory Animals
`(Institute of Laboratory Animal Resources, 1996).
`EAE Induction and Clinical Scoring. Animals were lightly
`anesthetized by isoflurane inhalation (0.5% Forane; Abbott Labora-
`tories, Vienna, Austria) and given a single intradermal 200-␮l inoc-
`ulation in the dorsal base of the tail root. The immunization mixture
`for DA rats consisted of syngeneic CNS antigen (4 parts brain to 6
`parts of spinal cord) in phosphate-buffered saline emulsified 1:1 in
`incomplete Freund’s adjuvant supplemented with 200 ␮g of heat-
`inactivated Mycobacterium tuberculosis (strain H37 RA; DIFCO, BD
`Diagnostics, Oxford, UK); the adjuvant is henceforth referred to as
`complete Freund’s adjuvant (CFA). For Lewis rats, the inoculum
`contained guinea pig spinal cord in phosphate-buffered saline emul-
`sified 1:1 in CFA. As an adjuvant control for all EAE studies, animals
`were injected with CFA alone and vehicle-treated. The rats were
`weighed every other day and scored daily for neurological signs as
`
`follows: 0, no clinical deficit; 1, complete loss of tail tonus; 2, limb
`weakness or ataxia; 3, full paralysis of hind or forelimbs; or 4,
`tetraparalysis or moribund. Animals with a score of 4 were sacrificed
`if weight loss indicated little chance of recovery, in accordance with
`animal welfare standards. Mortality due to sacrifice or spontaneous
`EAE-related death was indicated (†) and recorded as a 4 on thegiven
`day; this death score continued to be included in the clinical assess-
`ment, but body weight measurements were not carried forward.
`Test Compounds for in Vivo Evaluation. Unlabeled and 14C-
`labeled FTY720, as well as cyclosporine A (CsA), were supplied by
`Novartis Pharma AG (Basel, Switzerland). The radiochemical purity
`of [14C]FTY720, which was labeled in position 2, was shown by
`high-pressure liquid chromatography to be 98% with a specific ac-
`tivity of 35 ␮Ci/mg. FTY720 was dissolved in water, and CsA was
`dosed in the Neoral vehicle. Both drugs were freshly prepared and
`given p.o. once daily by gavage at a dosing volume of 5 ml/kg body
`weight. For prophylactic and therapeutic treatment, oral dosing
`started on day 0 at immunization and at the peak of disease in fully
`established EAE, respectively.
`Peripheral Leukocyte Counts. Rats were lightly anesthetized
`by isoflurane inhalation and 100 ␮l of blood from the retro-orbital
`venous plexus was collected in EDTA-coated tubes (Sarstedt AG,
`Nu¨ mbrecht, Germany). Automated differential leukocyte analysis
`was performed on the HESKA Vet ABC-Diff Hematology Analyzer
`(Heska Corp., Fort Collins, CO).
`Autoradiography of 14C-Labeled FTY720. Quantitative whole-
`body autoradiography (QWBA) and light microscopic autoradiogra-
`phy were performed to assess the uptake and tissue distribution of
`[14C]FTY720 radioactivity in male pigmented rats (n ⫽ 6) following
`seven oral doses at 7.5 mg/kg/d. At 8, 24, and 168 h after the last
`dose, the animals were deeply anesthetized with isoflurane and
`submerged in a dry ice-hexane bath at ⫺70°C for at least 20 min. The
`frozen carcasses were rapidly shaven and stored below ⫺20°C until
`embedment in an ice-cold aqueous solution of 2% carboxymethylcel-
`lulose. They were then frozen for approximately 30 min in a dry
`ice-hexane mixture at ⫺70°C, followed by an overnight stabilization
`at ⫺20°C. Lengthwise sections (40 ␮m thick) were obtained in a
`cryomicrotome (Leica Microsystems, Nussloch, Germany) at ⫺20°C.
`Whole-body autoradiograms were obtained by autoradioluminogra-
`phy. Briefly, sections with a paper backing were placed on Fuji
`BASIII imaging plates (Fuji Photo Film, Tokyo, Japan) for 1 day at
`room temperature in a lead shielding box. After exposure (detection
`of approximately 1.5 dpm/mg), the imaging plates were first kept in
`the dark for 3 to 5 min and then transferred to a Fuji BAS 2000 TR
`phosphorimaging device (Fuji Photo Film) for scanning at a 100-␮m
`step with a 1024 gradation. Images were prepared by re-exposing the
`sections
`onto Super Resolution storage
`phosphor
`screens
`(PerkinElmer, Shelton, CT) for 1 day at room temperature and
`
`Fig. 1. Cartoon depicting hierarchy of S1P receptor
`expression on rat glial subpopulations (Rao et al., 2003;
`Tham et al., 2003; Toman et al., 2004; Yu et al., 2004):
`implications for FTY720-P-mediated repair in the CNS.
`bFGF, basic fibroblast growth factor; CSF, colony-stim-
`ulating factor; GDNF, glial-derived neurotrophic factor;
`IFN␥, interferon ␥; IL1, interleukin-1; IL12, interleu-
`kin-12; MA, macrophage/microglia; NGF, nerve growth
`factor; NO, nitric oxide; TNF␣, tumor necrosis factor ␣.
`
`

`

`scanning at a 42-␮m step (Cyclone PhosphorImager; Packard Instru-
`ment, Meriden, CT). The image files were processed using Photoshop
`Elements 2.0 software (Adobe Systems, San Jose, CA). Levels of
`radioactivity in the tissues were determined by comparative densi-
`tometry, as described previously (Schweitzer et al., 1987).
`For light microscopy, brain and spinal cord samples from all
`animals were fixed in 3% glutaraldehyde in 0.1 M cacodylate buffer,
`pH 7.4, for 2 days at 4°C. Postfixation was performed with 1%
`osmium tetroxide in 0.1 M cacodylate buffer, pH 7.4, for 2 h at4°C.
`The tissues were dehydrated in graded acetone solutions and em-
`bedded in Epon. Semithin sections were dipped in Ilford L4 emulsion
`(Ilford, Mobberley, Cheshire, UK) in the dark at 4°C. Dipped sections
`were developed in Kodak D19 (Eastman Kodak, Rochester, NY) after
`31 weeks exposure, stopped in distilled water, fixed in Ilford Hypam
`rapid fixer (Ilford), and counterstained with toluidine blue. Light
`microscopic examination of sections was performed independently by
`three pathologists. Labeling was identified by more silver grains over
`the cells than the background without tissue.
`Quantification of FTY720 and FTY720-P in the Blood,
`Brain, and Cerebrospinal Fluid. Plasma, whole blood, brain, and
`cerebrospinal fluid (CSF) were collected from EAE rats at 24 h after
`the last FTY720 dose to obtain trough levels. Concentrations of
`FTY720 and FTY720-P were determined by high-pressure liquid
`chromatography (Agilent 1100; Agilent, Waldbronn, Germany) with
`mass spectrometric detection as described previously for serum and
`other tissues (Zemann et al., 2006). For measurements in plasma,
`heparinized blood, or CSF, 20- to 100-␮l aliquots were spiked with
`internal standards (final concentration, 0.5 ␮g/ml) and extracted
`with chloroform/methanol at acidic pH; extracts were dried and
`reconstituted in methanol/0.2% formic acid. Samples were chromato-
`graphed on a Luna C8 column (3 ␮, 2⫻ 50 mm; Phenomenex,
`Torrence, CA) equipped with a C4 wide-bore precolumn. The ana-
`lytes were eluted with a gradient (eluent A, 10 mM ammonium
`acetate containing 0.08% HCOOH in water; eluent B, 10 mM ammo-
`nium acetate containing 0.08% HCOOH in MeOH; 50–98% B in 14
`min) at a flow of 0.4 ml/min at 40°C. Analytes were detected by
`electrospray-ionization liquid chromatography with tandem mass
`spectroscopy using an API 4000 QTrap instrument (MDS Sciex,
`Concord, ON, Canada). The optimal collision energies for FTY720
`and FTY720-P were 23 and 25 V, respectively. The multiple reaction
`monitoring transitions monitored for FTY720 and FTY720-P were
`m/z 308/255 and 388/255, respectively.
`For CNS determination, half of a rat brain (approximately 750 mg)
`was emulsified 1:1 in 2 ml of MeOH/H2O (1:1) with a glass homog-
`enizer (Potter S; Braun, Melsungen, Germany). The homogenate was
`divided for separate determination of FTY720 and FTY720-P, then
`aliquots were spiked with their respective internal standards (35 ␮l)
`to give a final concentration of 0.5 ␮g/ml. Methanol (400 ␮l) was
`added to 0.5 ml of homogenate, followed by 20 min of mixing on a
`
`TABLE 1
`FTY720 therapeutic treatment effect in DA rat model of EAEa
`
`FTY720 Exposure in CNS
`
`471
`
`rotary shaker and 5 min of sonication at room temperature. Super-
`natants were harvested after centrifugation at 12,000g; the pellet
`was extracted once with 200 ␮l of methanol, then the supernatant
`was harvested after centrifugation. Combined supernatants were
`subjected to solid-phase extraction using STRATA-X-C 33 ␮m Cation
`Mixed-Mode Polymer Phase columns (Phenomenex). The columns
`were eluted with 3 ml of MeOH/1% NH4OH. The eluate was dried, its
`residue was dissolved in a 35:65 mixture of eluents A and B (see
`above), and then samples were chromatographed as described above.
`An alternative method for blood or homogenized brain comprised
`two extractions with acetonitrile/methanol/ethyl acetate (5:3:2), dis-
`solution of the residue of the evaporated organic phase in methanol/
`water/ammonia (50:48.75:1.25), and chromatography on a Zorbax
`SB-C8 0.5- ⫻ 75-mm column (3.5 ␮m; Agilent Technologies) at 50°C
`with a linear gradient from 0 to 95% B within 10 min at 20 ␮l/min
`(solvent A, 0.2% HCOOH in water; B, 0.2% HCOOH in acetonitrile).
`In this case, a Micromass Micro triple quadrupole mass spectrometer
`with electrospray source (Waters AG, Rupperswil, Switzerland) was
`used, and multiple reaction monitoring transitions were monitored
`for FTY720 and FTY720-P at m/z 308/105 and 388/255, respectively.
`Both analytical methods yielded identical results.
`Statistical Analysis. A one-way analysis of variance (ANOVA)
`was used to compare all data sets using SigmaStat for Windows,
`version 3.11 (Systat Software Inc., Richmond, CA). Differences be-
`tween groups were analyzed using the post hoc Tukey test for pair-
`wise multiple comparison. For EAE, area under the curve (AUC)
`values for body weight loss and clinical grade scores were evaluated
`during the entire prophylactic treatment period or after the initia-
`tion of therapeutic dosing. Probabilities (p) ⱕ 0.05 were considered to
`be statistically significant.
`
`Results
`FTY720 Provides Sustained Protection in EAE. Two-
`week therapeutic treatment with 0.03 to 0.9 mg/kg FTY720
`dose dependently inhibited progression of established dis-
`ease in the DA rat model of chronic EAE compared with
`vehicle controls, which exhibited sustained neurological def-
`icits throughout the 2-month observation (Table 1; Fig. 2).
`Evidence for rapid and full disease suppression was consis-
`tently observed after administration of 0.3 mg/kg FTY720,
`whereas the minimum effective dose providing almost com-
`plete protection even 1 month after discontinuation was 0.1
`mg/kg (p ⫽ 0.00002). A plateau in efficacy appeared to be
`reached by 0.3 mg/kg since there was no difference in the
`cumulative disease score between this dose and the 3-fold
`higher one of 0.9 mg/kg (Table 1). The very low dose of 0.03
`
`Treatment Groups: Oral Dose:
`from Days 12 to 25
`
`mg/kg
`Vehicle (n ⫽ 17)
`FTY720, 0.03 mg (n ⫽ 8)
`FTY720, 0.1 mg (n ⫽ 19)
`FTY720, 0.3 mg (n ⫽ 19)
`FTY720, 0.9 mg (n ⫽ 9)
`CsA, 25 mg (n ⫽ 18)
`
`EAE
`Onsetb
`
`Day
`
`8.4 ⫾ 0.4
`8.8 ⫾ 0.3
`8.4 ⫾ 0.3
`9.0 ⫾ 0.3
`8.3 ⫾ 0.2
`9.0 ⫾ 0.3
`
`Maximum Weight Lossc
`
`Maximum Diseased
`
`Cumulative Disease Score from
`Days 13 to 56e
`
`%
`
`Day
`
`Score
`
`Day
`
`AUC
`
`p vs. Vehicle
`
`23.6 ⫾ 0.02
`16.4 ⫾ 0.03
`12.0 ⫾ 0.02
`9.6 ⫾ 0.01
`15.4 ⫾ 0.02
`16.9 ⫾ 0.02
`
`20.7 ⫾ 0.7
`21.9 ⫾ 1.0
`20.4 ⫾ 0.2
`20.2 ⫾ 0.2
`20.2 ⫾ 0.1
`28.4 ⫾ 2.4
`
`2.9 ⫾ 0.2
`2.3 ⫾ 0.4
`1.2 ⫾ 0.3
`0.3 ⫾ 0.1
`0.6 ⫾ 0.4
`3.3 ⫾ 0.1
`
`22.6 ⫾ 1.5
`28.5 ⫾ 2.6
`24.0 ⫾ 1.9
`24.2 ⫾ 2.0
`25.7 ⫾ 3.7
`33.9 ⫾ 0.6
`
`79.0 ⫾ 12.6
`64.0 ⫾ 14.7
`17.0 ⫾ 3.7
`7.8 ⫾ 2.5
`7.7 ⫾ 3.3
`52.8 ⫾ 6.7
`
`N.S.
`0.00002
`⬍0.00001
`0.0005
`N.S.
`
`n, number of rats per group; N.S., not significant.
`a Syngeneic antigen-induced EAE (Fig. 2); data shown as mean ⫾ S.E.M.
`b Based on disease score ⱖ 1.
`c Highest weight loss starting 1 week after dosing (days 20–56) compared with adjuvant control.
`d Initial peak of disease severity starting 1 week after dosing (days 20–56).
`e Level of significance (p) determined by ANOVA of AUC values for clinical scores from days 13 to 56 compared with the positive control.
`
`

`

`472
`
`Foster et al.
`
`Fig. 2. Sustained efficacy of FTY720 after cessation in DA rat model of
`EAE. Representative disease course in syngeneic antigen-induced EAE in
`DA rats, depicted by the vehicle control (f; n ⫽ 17) and shown as mean
`score ⫾ S.E. Oral therapeutic dosing started at the peak of established
`disease on day 12 and continued for 2 weeks. CsA at 25 mg/kg (䡺; n ⫽ 18)
`suppressed EAE signs during treatment, but animals became severely
`paralyzed, and 22% died (†) upon drug discontinuation. In sharp contrast,
`FTY720 at 0.1 (E; n ⫽ 19) and 0.3 mg/kg (F; n ⫽ 19) significantly
`prevented wasting and recurrence of neurological deficits, as detailed in
`Table 1. Although 0.03 mg/kg FTY720 (ⴱ; n ⫽ 8) tended to reduce the
`disease burden, it was not statistically different to the vehicle (Table 1).
`Nevertheless, this very low dose as well as the other FTY720 treatments
`completely protected against EAE-related deaths, compared with three in
`the vehicle (days 15, 30, and 42) and four in CsA-treated animals (days
`34, 35, 40, and 41).
`
`mg/kg also tended to diminish the overall disease burden and
`prevent a marked rebound. In contrast, severe paralysis re-
`occurs following cessation of classic immunosuppressive
`agents, such as CsA at 25 mg/kg (Fig. 2) or FK506 at 4 mg/kg
`(data not shown), that were used as reference compounds. All
`doses of FTY720 completely prevented moribund incidence
`compared with 18 and 22% with vehicle and CsA, respec-
`tively (Fig. 2).
`Prolongation of Low-Dose FTY720 Steadily Reduces
`EAE Signs after Therapeutic Treatment. To further ex-
`plore the long-term efficacy of low-dose FTY720 in a more
`clinically relevant setting, 0.03 to 0.3 mg/kg was adminis-
`tered for 3 weeks in DA rats with established EAE (Fig. 3).
`By the last week of therapeutic treatment, three different
`studies consistently demonstrated a significant decrease in
`disease signs with 0.03 mg/kg versus the vehicle. Higher
`doses of 0.1 mg/kg (data not shown) and 0.3 mg/kg were even
`more efficacious. Moreover, the mortality rate was markedly
`reduced with FTY720 at 0.03 mg/kg (8.9% death), with com-
`plete protection at 0.1 and 0.3 mg/kg compared with 29.2%
`deaths in the positive control.
`In contrast, 2-week prophylactic treatment with 0.03
`mg/kg FTY720 in the Lewis rat model of acute EAE merely
`delayed the onset of paralysis by approximately 1 day but
`had no protective effect on disease development (Fig. 4).
`Increasing the 2-week dose to 0.3 mg/kg markedly prevented
`neurological deficits during the 1-month study (p ⫽ 0.0003);
`however, restricting the treatment to days 0 to 6 failed to
`stop EAE induction, with clinical signs appearing 1 week
`after drug cessation.
`Influence of FTY720 on Circulating Lymphocyte
`Counts during EAE. Given that the above disparities in
`
`Fig. 3. Prolonged therapeutic treatment with low-dose FTY720 progres-
`sively reduces EAE burden in DA rats. Change in clinical scores, along
`with EAE-related deaths (†), during the disease course in DA rats immu-
`nized with syngeneic CNS antigens. Data were pooled from three EAE
`studies and expressed as mean ⫾ S.E. Animals received daily treatment
`p.o. with vehicle (f; n ⫽ 48) or FTY720 for 3 weeks, starting from day 11.
`Prolongation of 0.03 mg/kg FTY720 dosing (O, n ⫽ 45) gradually reduced
`the EAE burden and weight loss such that animals no longer exhibited
`ataxia or limb paralysis (p ⫽ ⬍0.001), in contrast with vehicle controls,
`which displayed sustained disease throughout the 5-week observation
`period. As expected, the 0.3 mg/kg dose (F; n ⫽ 33) rapidly suppressed the
`EAE grade to a level near baseline, similar to the adjuvant controls that
`were injected with CFA alone and vehicle-treated (data not shown). Both
`doses of FTY720 profoundly prevented mortalities, except for 4 in the 0.03
`mg/kg group (days 15 and 21), compared with 14 in the vehicle (days 13,
`14, 17, 18, 19, 20, 22, 23, 25, and 26). Level of significance was determined
`by ANOVA of AUC values from days 12 to 33 versus the positive control.
`Blood was collected on days 14 and 33 (Œ) for differential leukocyte
`analysis, as shown in Fig. 5B. ⴱⴱⴱ, p ⬍ 0.001.
`
`Fig. 4. Efficacy of FTY720 prophylaxis in EAE is more dependent on dose
`and duration compared with therapeutic treatment. Clinical scores ⫾
`S.E. in Lewis rat model of EAE (nine animals per group) showing typical
`monophasic disease in the vehicle control (f). Preventive therapy with
`FTY720 was effective when 0.3 mg/kg was given for 2 weeks (F) but not
`when dosing was restricted to the first 7 days (E). A 10-fold lower dose of
`0.03 mg/kg (⽧) failed to prevent disease development. As expected, 25
`mg/kg CsA (ⴱ) was fully protective during treatment, yet paralysis oc-
`curred almost 2 weeks after drug cessation, resulting in a death (†). Blood
`was collected at 0, 6, and 24 h and on 3 subsequent days (Œ) for differ-
`ential leukocyte analysis, as shown in Fig. 5A.
`
`EAE-efficacy appeared to be less related to the FTY720 dose
`but more to its duration, we sought to investigate the tem-
`poral relationship between peripheral lymphocyte counts
`and EAE treatment regimens. Earlier studies showed that
`
`

`

`0.03 and 0.3 mg/kg FTY720 can decrease the peripheral
`lymphocyte count by approximately 20 and 70%, respec-
`tively, within 6 h after a single oral dose in naive Lewis rats
`(Brinkmann et al., 2002), resulting in an ED50 of 0.09 ⫾ 0.01
`mg/kg by 48 h. Likewise during EAE, 0.3 mg/kg FTY720
`already reached maximum reduction (approximately 90%) of
`lymphocytes by 6 h in theLewis rat (Fig. 5A). Furthermore,
`it was already recognized that daily doses of 0.03 mg/kg
`FTY720 for 1 week can reduce lymphocyte counts by up to
`80% compared with placebo-treated Lewis rats in a heart
`allograft model (Nikolova et al., 2000). We have extended
`these findings to EAE and demonstrate that 0.03 mg/kg
`therapeutic dosing for 3 weeks in DA rats leads to a highly
`significant reduction in circulating lymphocytes by days 14
`and 33 versus vehicle (Fig. 5B), i.e., 52 and 67%, respectively.
`At the early time point, 0.3 mg/kg FTY720 led to over twice
`the reduction in lymphocyte numbers compared with 0.03
`mg/kg (p ⬍ 0.001), but by day 33, there was no difference
`between these two doses (Fig. 5B).
`Taking into account that 7-day preventive dosing with 0.3
`mg/kg FTY720 failed to suppress EAE (Fig. 4), yet lympho-
`cytes were reduced by 75% versus vehicle (Fig. 5A), we agree
`with previous suggestions (Webb et al., 2004) that FTY720 or
`its phosphate are apt to exert additional effects beyond the
`induction of peripheral lymphopenia. For example, although
`0.03 mg/kg FTY720 decreased the circulating lymphocytes by
`at least 50% shortly after drug initiation (Fig. 5B), this low
`dose failed to significantly prevent EAE signs when treat-
`ment is limited to 2 weeks in a prophylactic setting (Fig. 4).
`To explore whether additional effects of FTY720 and
`FTY720-P in the brain are possible at all, based on available
`drug concentrations in that tissue, we examined the distri-
`bution of FTY720 in the rat and determined levels of
`FTY720/FTY720-P in the brain.
`Quantitative Whole-Body Autoradiography and My-
`elin Sheath Distribution of [14C]FTY720. First, QWBA
`was used to investigate the distribution of [14C]FTY720-
`related radioactivity in vivo and, in particular, its uptake
`into the CNS. Pigmented rats received [14C]FTY720 for 1
`week at a high oral dose of 7.5 mg/kg/d. By 24 h after the
`seventh and last dose (Fig. 6A), elevated amounts of ex-
`travascular radioactivity were detected in the adrenal cortex,
`kidney (cortex-medullary junction), nasal turbinates, pitu-
`
`FTY720 Exposure in CNS
`
`473
`
`itary gland, preputial gland, and stomach (glandular mu-
`cosa); maximal levels of radioactivity occurred in the brain,
`epididymis, eye (ocular membranes, vitreous body), and tes-
`tis. At 168 h, residual radioactivity was still observed in most
`of these tissues, equivalent to approximately 1.4% of the
`administered dose, but the highest concentrations were
`found in the brain (reticular nucleus, corpus callosum, cere-
`bellar white matter), preputial gland, and spinal cord (Fig.
`6B). The distribution pattern after multiple doses was simi-
`lar to that after a single dose (data not shown), especially the
`preferential localization to brain and spinal cord at 168 h.
`The accumulation factor (i.e., ratio between the tissue con-
`centration after multiple versus single dosing) was 3.2 for
`brain and 3.7 for spinal cord at 24 h after the seventh dose
`compared with the single dose.
`Light microscopic evaluation showed that the 14C labeling
`in brain and spinal cord was confined to the myelin sheets
`(Fig. 7). Neurons were free of grains, except for some back-
`ground labeling.
`Levels of FTY720 and FTY720-P in Blood and Brain.
`Given that exposure of the CNS after oral dosage of FTY720
`to normal rats was shown in the above QWBA study and in a
`previous pharmacokinetic analysis (Meno-Tetang et al.,
`2006), we next sought to determine concentrations of FTY720
`and its phosphorylated form in blood and brain of rats dis-
`eased with EAE. In fact, levels of FTY720-P in brain have so
`far not been reported at all. Samples from both acute and
`chronic models of EAE (in Lewis and DA rats, respectively)
`were obtained 24 h after the last dose to determine trough
`levels. FTY720 and FTY720-P were found both in the blood
`and brain, with ratios between approximately 0.8 and 3 (Ta-
`ble 2). Brain levels of both forms were considerably higher
`(by factors of 10 to 27 after 23 doses) than those in blood, as
`seen before for FTY720 in normal rats (Meno-Tetang et al.,
`2006). Brain concentrations of the two forms were of compa-
`rable magnitude. The observed amounts of both FTY720 and
`its phosphorylated form in the DA rat brain increased with
`the oral dose of FTY720, as did the blood levels; thus, the
`brain/blood ratio was relatively constant in DA rats (21–27
`for FTY720; 14–17 for FTY720-P). We also followed the time
`course of FTY720 and FTY720-P concentrations in blood and
`brain (Table 2). Although blood levels were constant over 21
`
`Fig. 5. FTY720 rapidly reduces the peripheral
`lymphocyte count, with sustained activity at a low
`subtherapeutic dose during EAE. A, time course of
`lymphocyte count in the Lewis rat model of EAE
`following prophylactic treatment from days 0 to 13
`with vehicle (f), 0.3 mg/kg FTY720 (E), or 25
`mg/kg CsA (ⴱ); n ⫽ 6 per group, represented as
`mean ⫾ S.D. At 6 h postimmunization, FTY720
`had already reduced the lymphocyte number to
`the maximal extent. CsA, despite full efficacy dur-
`ing treatment (Fig. 4), did not significantly alter
`the lymphocyte count. Level of significance was
`based on ANOVA comparison with naive animals
`at day 0. B, reduction of peripheral lymphocytes
`during early and late EAE in DA rats therapeuti-
`cally treated from days 11 to 33 with vehicle (u);
`n ⫽ 21) or FTY720 at 0.03 (o; n ⫽ 26) and 0.3 (f;
`n ⫽ 13) mg/kg. Data were pooled from three EAE
`studies (Fig. 3) and expressed as mean percent-
`age ⫾ S.E. of lymphocytes in the adjuvant control
`(䡺; n ⫽ 13). ANOVA was performed against the
`vehicle of antigen-immunized animals. ⴱ, p ⱕ 0.05;
`ⴱⴱ, p ⱕ 0.01; ⴱⴱⴱ, p ⱕ 0.001.
`
`

`

`474
`
`Foster et al.
`
`Fig. 6. High-dose [14C]FTY720 autoradiography in rats.
`Representative whole-body, midsagittal autoradiolumino-
`grams taken 24 (A) and 168 (B) h after administration of
`seven daily oral doses of 7.5 mg/kg [14C]FTY720 to pig-
`mented rats. White lines (B) point to increased label in the
`brain and spinal cord by 168 h after the last dose.
`
`FTY720 within the CNS. Interestingly, the brain/blood ratio
`remained relatively constant (21–27 and 14–17 for FTY720
`and FTY720-P, respectively) after 3 weeks of treatment de-
`spite the dose range (0.03–0.3 mg/kg; Table 2).
`Regardless of how the FTY720-P levels are attained in the
`brain, we speculate that this phosphorylated metabolite may
`promote endogenous repair mechanisms in the CNS via S1P
`receptors on glial and/or neuronal cells. It is noteworthy that
`the nervous system is a major locus for constitutive S1P
`receptor expression in glial cells and neurons (Ishii et al.,
`2004). Four of the five known S1P receptor subtypes display
`a distinct distribution pattern within specific brain regions
`and cell lineages, as illustrated in Fig. 1. CNS expression of
`S1P5, for example, is restricted to oligodendrocytes (Terai et
`al., 2003) and expressed throughout development to the ma-
`ture myelin-forming cell. Subsequent to the discovery that
`S1P acts as an important regulator of cell growth, it has
`become increasingly clear that this sphingolipid mediator
`may induce the survival of such cells in the CNS (Ishii et al.,
`2004).
`Indeed, recent studies have demonstrated that
`FTY720-P promotes the survival of oligodendroglial lineage
`cells in vitro (Jung et al., 2007). Moreover, FTY720-P ligation
`of S1P receptors on astrocytes (Osinde et al., 2007) could
`contribute to its known enhancement of endothelial barrier
`function (Abbott et al., 2006; Baumruker et al., 2007) and
`possibly to myelination (Talbott et al., 2005; Ishibashi et al.,
`2006). Further studies are needed to directly elucidate the in
`vivo consequences of FTY720-P signaling on CNS cells, which
`is underscored by the recent in vitro observation that S1P
`activation of S1P1 and S1P3 receptors can inhibit gap junc-
`tions in astrocytes (Rouach et al., 2006).
`The signature feature of FTY720 is its ability to rapidly
`reduce blood lymphocytes as a consequence of S1P1-mediated
`retention in the peripheral lymph nodes (Lo et al., 2005). It is
`notable that FTY720 not only spares CD4⫹CD25⫹ T-regula-
`tory cells (Treg) but also induces their functional activity
`(Daniel et al., 2007). Other mechanisms that act indepen-
`dently of S1P receptors may also be part of the activity of
`FTY720, such as suppression of eicosanoid production due to
`inhibition of cytosolic phospholipase A2 (Payne et al., 2007).
`Importantly, FTY720 neither inhibits the activation of lym-
`phocytes at therapeutically relevant concentrations nor
`overtly alters their effector function, including antibody re-
`
`Fig. 7. Light microscopic [14C]FTY720 autoradiography in rats. Semithin
`Epon-embedded sections of spinal cord, counterstained with toluidine
`blue, at 24 h (A) and 7 days (B) after the last dose of [14C]FTY720. Black
`autoradiography granules (arrows) are primarily localized along the my-
`elin sheets. Neurons (N) and axons (A) are free of grains, except for
`occas