Journal of Neuroimmunology 153 (2004) 108 – 121
`
`www.elsevier.com/locate/jneuroim
`
`Sphingosine 1-phosphate receptor agonists attenuate relapsing–remitting
`experimental autoimmune encephalitis in SJL mice
`
`Michael Webb*, Chui-Se Tham, Fen-Fen Lin, Karen Lariosa-Willingham, Naichen Yu,
`Jeffrey Hale, Suzanne Mandala, Jerold Chun, Tadimeti S. Rao
`
`Molecular Neuroscience Laboratory, Merck Research Laboratories, 3535 General Atomics Court, San Diego, CA 92121, USA
`
`Received 3 March 2004; received in revised form 23 April 2004; accepted 26 April 2004
`
`Abstract
`
`FTY720 is a prodrug for FTY-phosphate, an agonist at four of the five known receptors for sphingosine-1-phosphate (S1P). We show that
`administration of either FTY720 or FTY-P to SJL mice with established relapsing – remitting experimental autoimmune encephalitis (EAE)
`results in a rapid and sustained improvement in their clinical status, and a reversal of changes in expression of mRNAs encoding some myelin
`proteins and inflammatory mediators. EAE produced by adoptively transferring lymph node cells from immunized mice to naı¨ve hosts is
`similarly ameliorated by FTY-P. Treatment with FTY-P is accompanied by a dose-responsive peripheral lymphopoenia.
`D 2004 Elsevier B.V. All rights reserved.
`
`Keywords: Relapsing – remitting EAE; FTY720; Lymphopoenia; Gene expression
`
`1. Introduction
`
`FTY720 is a structural analog of myriocin, a metabolite
`of the ascomycete fungus Isaria sinclairia, with some
`structural resemblance to sphingosine, an endogenous lyso-
`lipid. Sphingosine undergoes phosphorylation via sphingo-
`sine kinase leading to the formation of sphingosine-1-
`phosphate (S1P), the cognate ligand for the family of S1P
`receptors (S1PR). Activation of S1P receptors results in a
`plethora of physiological actions such as chemotaxis, cel-
`lular differentiation, survival and growth, and regulation of
`actin-based cytoskeletal reorganization which leads to cell
`adherence and cell shape changes (Goetzl and An, 1998;
`Chun, 1999; Fukushima et al., 2001).
`FTY720 is a novel immunosuppressive agent which is
`active in various animal models of graft rejection and
`autoimmune disease, including graft versus host disease,
`type 1 diabetes, and rheumatoid arthritis. (Chiba et al., 1996;
`Suzuki et al., 1996a,b, 1998; Masubuchi et al., 1996;
`Matsuura et al., 2000). It is currently under development
`as an immunosuppressive agent for transplantation. It was
`thought initially that the mechanism of action was through
`
`* Corresponding author. Tel.: +1-858-638-7957.
`E-mail address: xhua@san.rr.com (M. Webb).
`
`0165-5728/$ - see front matter D 2004 Elsevier B.V. All rights reserved.
`doi:10.1016/j.jneuroim.2004.04.015
`
`induction of apoptosis in T lymphocytes (Nagahara et al.,
`2000; Suzuki et al., 1996a,b, 1997), and one effect of in
`vivo treatment with FTY720 is a profound lymphopoenia in
`the peripheral blood, with lymphocyte cell counts falling to
`as low as 5 – 10% of control levels at therapeutic doses of
`the compound. However, therapeutic effects are achieved in
`rats at doses of < 1 mg/kg, at which the plasma concen-
`trations are about two orders of magnitude lower than those
`required to drive T cell apoptosis in vitro (Yanagawa et al.,
`1998). A further argument against a role of apoptosis in
`most in vivo situations is provided by the observation that
`adoptively transferred fluorescently labeled lymphocytes
`disappear from the peripheral circulation on FTY720 treat-
`ment but reappear when drug treatment
`is discontinued
`(Pinschewer et al., 2000). It now appears that at least one
`of the mechanisms by which FTY720 achieves its effects in
`vivo is by a sequestration of circulating lymphocytes in
`peripheral lymph nodes (Pinschewer et al., 2000; Brink-
`mann et al., 2000, 2001a,b; Mandala et al., 2002, Xie et al.,
`2003).
`FTY720 is a substrate for sphingosine kinase-2 (Sanchez
`et al., 2003) and phosphorylation in vivo (Mandala et al.,
`2002) has been demonstrated. The resultant ester (FTY-P)
`has a structure similar to sphingosine-1-phosphate, which is
`the preferred ligand at a group of G protein coupled
`
`Apotex v. Novartis
`IPR2017-00854
`NOVARTIS 2014
`
`

`

`M. Webb et al. / Journal of Neuroimmunology 153 (2004) 108–121
`
`109
`
`receptors with five known members (S1P receptors, S1PR).
`Mandala et al. (2002) showed that FTY720 itself has weak
`or no activity at any of these receptors but that the phosphate
`ester is an agonist with low nanomolar potency at four of the
`five receptors, and these observations were further substan-
`tiated by Brinkmann et al. (2002). A nonhydrolysable
`phosphonate analogue of FTY-P retained sufficient potency
`at the four S1P receptors to have in vivo efficacy in a
`lymphopoenia assay. FTY720, sphingosine-1-phosphate,
`and the phosphonate analogue of FTY-P all caused a rapid
`and reversible peripheral lymphopoenia in rats and mice,
`reaching a nadir at 4 h postinjection. The potency of
`compounds in this assay reflected their intrinsic affinities
`at the four receptors, indicating that one or more of these
`receptors is indeed the molecular target of these compounds
`(Mandala et al., 2002). Recent reports using selective
`agonists demonstrate that S1P1 is the target for lymphope-
`nia (Forrest et al., 2004; Sanna et al., 2004). Moreover,
`lymphocytes genetically deleted for S1P1 have thymic
`emigration and recirculation defects similar to that achieved
`with the receptor agonists, suggesting that S1P1 is required
`for egress (Matloubian et al., 2004) and that the agonists
`induce lymphopenia by downregulating S1P1 on lympho-
`cytes (Graler and Goetzl, 2004).
`Two papers have appeared indicating that FTY720 is
`active in rodent models of experimental autoimmune en-
`cephalitis (EAE), an animal model of multiple sclerosis.
`Brinkmann et al. (2002) treated Wistar rats with FTY720
`(0.3 mg/kg/day) from the day of induction of EAE (day 0)
`in Wistar rats. They showed that rats thus treated did not
`develop EAE in this monophasic model. In a different acute
`monophasic rat model, in which myelin basic protein is
`used as the immunogen in Lewis rats, Fujino et al. (2003)
`showed once again that dosing rats orally from day 0
`almost completely suppressed the development of disease.
`This was associated with a marked reduction in the number
`of T lymphocytes infiltrating the spinal cord and a reduc-
`tion in the levels of the TH1 cytokines IL-2, IL-6, and
`interferon gamma. While these data are impressive,
`the
`efficacy of FTY720 when given at the onset of clinical
`symptoms or on established EAE is unknown. In the
`present investigation, we have examined the efficacy of
`FTY720 and its phosphate ester on an established disease
`state in the relapsing – remitting EAE model in SJL mouse
`(SJL rr-EAE), a model which mimics several features of
`human MS.
`EAE can be induced in SJL strain mice by immunizing
`them with a peptide fragment of the myelin protein,
`proteolipid protein (PLP),
`together with pertussis toxin
`(PTX) treatment (McRae et al., 1992). The resulting
`EAE shares many features with human multiple sclerosis.
`It is a chronic disease from which the mice never recover
`and has a relapsing – remitting pattern similar to that of the
`major form of human multiple sclerosis (McRae et al.,
`1992, 1995). In addition to the infiltration of T lympho-
`cytes into the brain and the spinal cord (and unlike the
`
`common rat models), these mice also exhibit both demy-
`elination and also axonal damage (Sobel et al., 1990;
`Marracci et al., 2002). Demyelination has for many deca-
`des been regarded as the hallmark of multiple sclerosis, but
`in recent years, the importance of the axonal damage and
`eventual neuron loss as the key processes underlying the
`relentless progression of the disease has been recognised
`(Trapp et al., 1998, 1999; Wujek et al., 2002). In addition
`to confirming these features of the SJL-rr model, we also
`show that in this EAE model, significant changes in the
`quantitative expression of mRNAs encoding both myelin-
`related proteins and mediators such as granulocyte-macro-
`phage colony stimulating factor (GM-CSF) and inducible
`nitric oxide synthase occur.
`In this study, we characterise the effects of FTY720 and
`its phosphorylated derivative on the clinical state, and levels
`of circulating lymphocytes in the SJL mouse rr-EAE model.
`Initiating dosing just prior to the onset of clinical signs
`delays and blunts the first phase of the disease as previously
`reported in the rat models. More significantly for the clinical
`setting, if treatment with these compounds is delayed until
`the peak of the first phase of disease, there is an immediate
`and rapid improvement in the clinical status of the animals
`which is maintained for as long as dosing is continued.
`Treatment with FTY-P is associated with a partial reversal in
`the gene expression changes seen in untreated animals with
`the disease. We also show that there is a correlation between
`the magnitude of clinical improvement and the levels of
`lymphopoenia achieved at any given dose of FTY-P, but this
`correlation is incomplete, indicating that nonselective S1P
`receptor agonism may exert its effects in this model by
`additional mechanisms. FTY-P is also effective in improv-
`ing clinical status and preventing mortality in an adoptive
`transfer version of the disease.
`
`2. Materials and methods
`
`2.1. Experimental autoimmune encephalitis
`
`2.1.1. SJL mouse
`Female SJL mice (6 – 9 weeks old) were obtained from
`the Jackson laboratory (Bar Harbor, ME). They were
`housed in a 12-h light / dark cycle with access to food
`and water ad lib. All animal procedures were conducted in
`accordance with protocols approved by the local animal
`care committee.
`
`2.1.2. Active immunization
`EAE was induced following the methods of McRae et al.
`(1992). A 20-mer peptide based on the mouse proteolipid
`protein (PLP) sequence 139 – 151 (His – Ser – Leu – Gly –
`Lys – Trp – Leu – Gly – His – Pro – Asp – Lys – Phe (custom
`synthesized by American Peptide, Sunnyvale, CA, 98%
`purity) was used to induce EAE. Peptide was dissolved at
`5 mg/ml in distilled water and emulsified with an equal
`
`

`

`110
`
`M. Webb et al. / Journal of Neuroimmunology 153 (2004) 108–121
`
`volume of Freund’s adjuvant containing 5 mg/ml of H37Ra
`M. tuberculosis (Difco) following the procedure of Stevens
`et al. (1999). On the day of initiation of EAE, referred to as
`day 1 of the experiment, mice received a single injection of
`100 ul of emulsion subcutaneously in a skin fold at the back
`of the neck. On day 1, they were also given 250 ng of
`pertussis toxin (List Biological Labs, Campbell, USA) from
`a solution of 50 ug/ml in 0.01 M sodium phosphate buffer
`pH7.0 with 0.9% NaCl intraperitoneally. This injection was
`repeated on day 3. Control animals received an injection of
`water/Freund’s emulsion and both pertussis injections.
`Mice were weighed daily and assessed using a clinical
`scale as follows: 0 = healthy mouse, 1 = flaccid tail, 2 = hind
`limb weakness, 3 = paralysis of one or both hind limbs,
`4 = forelimb paralysis, and 5 = death. Animals were eutha-
`nased if they reached a score of 4. Mice which appeared to
`be dehydrated were given 1 ml of saline ip per day, and
`cages containing mice with clinical scores of above 2 were
`provided with wet chow on the cage base to ensure that
`animals could reach their food.
`Drugs and experimental compounds were given by
`intraperitoneal injection or by oral gavage as appropriate.
`FTY-P was synthesized at the Merck Research Laboratories,
`Rahway. Mitoxantrone was purchased from Calbiochem as
`a 10-mg/ml solution and was diluted for use in saline.
`In some experiments, animals were bled retroorbitally at
`a frequency not greater than once per week to provide blood
`samples for lymphocyte counting. At the termination of
`experiments, tissues (brain, spinal cord, liver) were taken
`and snap frozen on dry ice for biochemical analysis. In some
`cases, tissues to be used for RNA preparation were collected
`and stored overnight at 4 jC in an RNA preservative
`solution (RNAlater, Qiagen), and then placed at 70 jC
`
`until RNA preparation.
`
`2.1.3. EAE induction in SJL/J mice via adoptive transfer of
`PLP-reactive lymphocytes
`EAE was passively induced by adoptive transfer of
`PLP-reactive lymphocytes. Thoracic and popliteal lymph
`nodes were collected 11 – 12 days after immunization of
`donor mice with PLP in CFA. Free lymph node cells were
`obtained by mechanically dissociating the nodes, and red
`blood cells were hypotonically lysed by a 2-min treatment
`in 0.144 M (0.95%) ammonium chloride, 17 mM Tris –
`HCl, pH7.4. These cells were cultured in Click’s medium
`(EHAA) supplemented with 10% foetal calf serum and 50
`uM beta mercaptoethanol. Cells were stimulated for 48 –
`72 h in the presence of 25 ug/ml PLP139 – 151 or BSA as a
`negative control. Proliferation of the stimulated cells was
`confirmed by [3H]- thymidine incorporation and micro-
`scopically by the appearance of blasts. Recipient SJL
`mice were injected intraperitoneally with PLP-reactive
`lymphocytes (6 – 12.5 million cells/mouse in a volume of
`0.5 ml phosphate buffered saline per mouse). Clinical
`evaluation of EAE was conducted using the methods
`described above.
`
`2.2. Lymphopoenia assays
`
`The lymphopoenia assay was performed using an in vitro
`microprocessor controlled automatic blood analyzer instru-
`ment (H-2000 Hematology Analyzer (Hospitex Diagnostics
`LP, Webster, TX) using whole blood samples.
`Whole blood from EAE animals was collected by
`retroorbital bleeding, using EDTA as an anticoagulant.
`The blood was used within 4 h at room temperature for
`cell counting. Lymphopoenia was defined in our assay as
`the reduction of lymphocyte cell density (cells/Al) as
`compared to PLP control vehicle. Data were expressed by
`plotting cell density as a function of time (days in vivo) and
`analyzed using software Prismk 3.0 (San Diego, CA) All
`values are expressed as mean F standard error of mean
`(S.E.M.).
`
`2.3. Molecular biology
`
`2.3.1. Isolation of total RNA and reverse transcription
`Total RNA from mouse spinal cord, brain, lung, and
`heart (LPS-treated) was isolated using an RNeasyR Protect
`Midi Kit (Qiagen). Total RNA of PLP-treated mouse lym-
`phocytes was isolated using an RNeasyR Protect Mini Kit
`(Qiagen). RNA was treated with RNase-free DNase I to
`remove genomic DNA contamination. Total RNA (1.2 Ag)
`from each was reverse transcribed using a RETROscriptk
`Kit (Ambion). In each reverse transcription reaction, a
`reaction-omitting reverse transcriptase was included for
`the assessment of genomic DNA contamination.
`
`2.3.2. Cloning, construction, and DNA sequencing of mini
`genes
`The cDNA encoding myelin components (MAG, PLP,
`MBP, MOG, and CNPase) and h actin was amplified from
`mouse brain with polymerase chain reaction (PCR) techni-
`ques. The cDNA encoding GM-CSF and iNOS was ampli-
`fied from mouse lung and LPS-induced heart, respectively.
`For cloning of mini genes for myelin components, iNOS
`and h actin, 3 Al of cDNA was amplified with 2.5 U of
`PfuTurboR DNA polymerase (Stratagene), 200 nM dNTP,
`and 10 pmol of each primer in a total volume of 50 Al for
`30 cycles in GeneAmpR 9700 thermocycler (ABI). Each
`cycle consisted of 30 s at 95 jC, 30 s at 55 jC, and 45 s at
`75 jC. GM-CSF mini gene was amplified from 3 Al of
`cDNA using 2.5 U of PlatinumR Taq DNA polymerase
`High Fidelity (Invitrogen), 200 nM dNTP, 2 mM MgSO4,
`1 High Fidelity PCR buffer, and 10 pmol of each primer
`in a total volume of 50 Al. PCR was performed using a
`GeneAmpR 9700 thermocycler (ABI) with incubation at 94
`jC for 30 s, 55 jC for 30 s, and 74 jC for 45 s for 30
`cycles. The PCR-amplified fragments were subcloned into
`PCR cloning vectors, pCRR – BluntII-TOPOR, or
`pCRRII – TOPOR (Invitrogen) using a PCR Cloning Kit
`(Invitrogen). The integrity of these genes was confirmed by
`sequencing using an ABI 3100 automated fluorescence
`
`

`

`M. Webb et al. / Journal of Neuroimmunology 153 (2004) 108–121
`
`111
`
`sequencer (ABI). The sequence was analyzed using soft-
`ware Sequencher.
`
`3. Results
`
`2.3.3. TaqManR PCR
`Quantitative real-time PCR was carried out using an
`ABI Prism 7900 sequence detector on 1 Al of cDNA
`samples using 900 nM each primer, 250 nM TaqMan
`probe, and 25 Al of TaqManR Universal PCR Master
`Mix, in a total volume of 50 Al. PCR was carried out with
`incubation at 50 jC for 2 min, 95 jC for 10 min followed
`by 40 cycles of 15 s at 95 jC and 1 min at 60 jC. Primers
`used for these genes were designed from sequences in the
`Genebank database (Table 1). Additional reactions were
`performed on each 96-well plate using a known dilution of
`DNA from mini genes or cDNA as PCR template for
`constructing a standard curve relating threshold cycle to
`cDNA concentration. Data were analyzed using software
`SDS2.0. All data were normalized to h actin and expressed
`as %control.
`
`2.4. Statistics
`
`All values are expressed as mean F standard error of
`mean (S.E.M., n = 4 individual animals each with 2 – 3
`replicates). Data were analyzed by analysis of variance
`(ANOVA) followed by post hoc analysis (Neuman – Keul’s
`test; Prizm 3.0, GraphPad, San Diego, CA) and statistical
`significance inferred at a p V 0.05.
`
`Table 1
`Primer and probe sequences used for mouse Taq-Man experiments
`
`3.1. EAE in SJL mice
`
`Mice injected subcutaneously with a single dose of
`PLP emulsified in CFA,
`together with intraperitoneal
`doses of PTX on the first and third days, began to lose
`body weight and manifest clinical signs at about day 7.
`The decline in body weight and increase in clinical scores
`occurred rapidly, reaching their maxima at about day 14.
`At this point, mice had typically lost about 30 – 35% of
`their body weight, and mean clinical scores were usually
`about 3. In our hands,
`the subsequent evolution of the
`disease showed some interexperiment variations. In some
`experiments, the first phase of disease would remit to a
`low clinical score by about day 25, with a slow gain of
`body weight
`to about 90% of control values. In those
`cases where such a clear remission from first phase was
`observed, mice spontaneously relapsed with subsequent
`bouts of severe disease, separated by intervening remis-
`sions (Fig. 1A).
`In other experiments,
`the relapsing –
`remitting pattern was less obvious, and after a clear first
`phase with a lower degree of remission, the mean clinical
`scores showed a chronic secondary type of progression
`(Fig. 1B).
`the clinical scores (during
`In both of these patterns,
`remission in the case of clearly rr-EAE) became progres-
`sively higher as the disease advanced. After the first phase,
`
`Gene
`
`PLP
`
`MBP
`
`MAG
`
`MOG
`
`CNPase
`
`iNOS
`
`GM-CSF
`
`h actin
`
`INFg
`
`Accession #
`
`M15442
`
`M11291; XM_129053; M15062
`
`NM_010758
`
`U64572
`
`M31810
`
`U43428
`
`X02333
`
`X03672
`
`K00083
`
`Forward primer (F)/TaqManR probe (TM)/Reverse primer (R)
`
`F: AGCGGGTGTGTCATTGTTTG
`TM: 5VAAACTTGTCGGGATGTCCTAGCCA
`R: 5VACAACAGTCAGGGCATAGGTGAT
`F: 5VGACCCAAGATGAAAACCCAGTAGT
`TM: 5VCATTTCTTCAAGAACATTGTGACACCT
`R: 5VTTGGGATGGAGGTGGTGTTC
`F: 5VCGCACGGTGGAGCTGAGT
`TM: 5VTCATGTATGCACCTTGGAAGCCC
`R: 5VCCACCACCGTCCCATTCA
`F: 5VTGTAGGCCTTGTATTCCTCTTCCT
`TM: 5VACGAAGTTTTCCTCTCAGTCTGTGCTG
`R: 5VGTCCGATGGAGATTCTCTACTTCTG
`F: 5VTGTGCTGCACTGTACAACCAAAT
`TM: 5VCACCACCTCCTGCTGGGCGTATTCT
`R: 5VAGGCCTTGCCATACGATCTCT
`F: 5VAAATCCCTCCTGATCTTGTGTTG
`TM: 5VACTCGTACTTGGGATGCTCCATGGTCA
`R: 5VCAACCCGAGCTCCTGGAA
`F: 5VAGAAGTCGTCTCTAACGAGTTCTCCTT
`TM: 5VCGGGTCTGCACACATGTTAGCTTCTT
`R: 5VGTAGACCCTGCTCGAATATCTTCAG
`F: 5VCGATGCCCTGAGGCTCTTT
`TM: 5VCCAGCCTTCCTTCTTGGGTATG
`R: 5VTTTCATGGATGCCACAGGATT
`F: 5VGCATAGATGTGGAAGAAAAGAGTCTC
`TM: 5VCATCCTTTTGCCAGTTCCTCCAGA
`R: 5VGCTCTGCAGGATTTTCATGTCA
`
`

`

`112
`
`M. Webb et al. / Journal of Neuroimmunology 153 (2004) 108–121
`
`peaking with a mean clinical score of 2 at day 15, and this
`was followed by a relatively small remission to a mean
`score of about 1.25 by day 20, at which point the animals
`remained stable until day 30. At day 30, the clinical scores
`worsened, but
`in this experiment, a chronic secondary
`phase was seen without further evidence of clear relapses
`and remissions.
`We studied the expression of the myelin-related proteins
`PLP, MBP (Fig. 2B and C), and also CNPase, MAG, and
`MOG (data not shown). The change in expression of the
`latter three mRNAs was very similar to that shown for MBP
`and PLP. The first phase of the disease was marked by a
`profound reduction in the level of these mRNAs, all of
`which fell to 26% or less than control levels at day 14. There
`was a subsequent recovery of expression at day 27, some-
`what prior to the minor improvement in clinical score that
`began at day 30. Thereafter, expression levels again de-
`clined at day 44, at which time the clinical status was slowly
`worsening.
`The expressions of the three inflammatory mediators
`IFNg, iNOS, and GM-CSF were similarly markedly elevat-
`ed during the first phase (Fig. 2D – F) and significantly
`reduced in the first small remission. In all three cases, they
`remained somewhat higher during the later phases of the
`disease, with GM-CSF showing the largest elevation
`(>1000-fold at all time points) over control levels at these
`later phases.
`
`3.3. Clinical effect of mitoxantrone, FTY720, and FTY-P
`
`We examined the effect in this model of mitoxantrone,
`a clinically used drug. Because of inherent toxicity, we
`were able to dose only for 6 days. When the drug was
`given for 6 days from day 6 – 12, it delayed the onset of
`disease and blunted the severity of the first phase, but at
`later time points, there appeared to be little difference in
`the condition of these animals from that of untreated
`controls (Fig. 3).
`We initiated once-daily dosing of mice with FTY720 at 3
`mpk, beginning at day 7, 14, or 25 after immunization.
`Animals dosed from day 7 showed a delay in onset of
`clinical signs of a few days, and the severity of the first
`phase was blunted compared with control mice (Fig. 4A).
`The degree of weight loss in this group was also reduced as
`compared with vehicle-treated mice. After resolution of this
`mild first phase, the mice remained at a very low and stable
`clinical score for the duration of the compound dosing.
`Much more dramatic effects were seen if doing was initiated
`at the peak of the first phase, when clinical scores were at
`their highest (Fig. 4B). A rapid improvement in the clinical
`score was initiated and maintained over several days, until
`by day 22, the mice had reached a similar low score as the
`mice treated from day 7. In this experiment, the remission at
`day 25 in the untreated control animals was relatively
`modest, and the clinical scores of mice which received
`FTY from this point declined to a level significantly below
`
`Fig. 1. Representative time course of clinical evolution of EAE in two
`separate experiments. (A) A typical relapsing – remitting time course is seen
`for animals induced for EAE by immunization with PLP and treatment with
`pertussis toxin. (B) A less clear remission is seen after phase 1, and the
`animals remain chronically sick at a clinical score of about 2.
`
`body weight was little affected during subsequent relapses.
`Mortality was seen at levels which varied between experi-
`ments from between about 10% and 30%, but almost all
`mortality occurred in the first phase, and animals surviving
`this phase were unlikely to die subsequently.
`
`3.2. Gene expression changes during EAE
`
`We used the Taq-Man technique to examine quantitative
`changes in the expression of several myelin-related and
`inflammation-related genes during the course of EAE. In
`these studies, gene expression was initially expressed as a
`percentage of the expression of the h actin-encoding gene,
`whose expression was essentially constant between differ-
`ent samples. We then derived a final value for the expres-
`sion in EAE samples by expressing the values as a
`percentage of the levels observed in disease-free CFA/
`PTX-treated controls taken at
`the same time. Fig. 2A
`represents the clinical evolution of the disease in the
`EAE experiment from which these time course samples
`were derived. There was, as was always observed in this
`model, a clear first phase, beginning at about day 7 and
`
`

`

`M. Webb et al. / Journal of Neuroimmunology 153 (2004) 108–121
`
`113
`
`Fig. 2. Gene expression changes during development of EAE. (A) Time course of EAE experiment yielding samples for gene expression studies. (B – F) Taq-
`Man quantitation of changes in spinal expression of the myelin-related genes for proteolipid protein (PLP; B), myelin basic protein (MBP; C), and for the
`inflammatory mediators interferon gamma (D), iNOS (E), and GM-CSF (F). Individual samples were initially expressed as a percentage of h actin expression
`determined in the same Taq-Man reaction and then finally expressed as percentages of the values for the same genes obtained from CFA and PTX treated non-
`EAE control samples at the same time point. Values are means F S.E.M., n = 4. Statistical significance is calculated by reference to the day 7 time point in each
`data set.
`
`Fig. 3. Effect of mitoxantrone (0.25 mg/kg, ip once daily between days 6 and 12) on clinical evolution of EAE.
`
`

`

`114
`
`M. Webb et al. / Journal of Neuroimmunology 153 (2004) 108–121
`
`Fig. 4. Effect of once-daily intraperitoneal dosing with FTY720 initiated at different times in the development of the disease. (A) FTY720 (3 mg/kg) was given
`daily by ip from day 7 until day 60. Dosing was resumed at day 70. (B) FTY720 (3 mg/kg) was given daily by ip injection between days 14 or 25 until day 60.
`Dosing was resumed at day 70.
`
`the lowest achieved by nondosed vehicle-treated controls.
`Once again, on continued dosing, the mice were stably
`maintained at this low level of clinical sign. FTY720 dosing
`was interrupted for a 10-day period at day 60. After a delay
`of a few days, clinical signs reappeared in mice in all
`treatment groups, which relapsed over a period of days to
`similar disease scores to those seen in untreated controls. On
`resumption of dosing,
`these signs once again began to
`ameliorate. When dosing was initiated at day 25, at which
`point the animals were expected to have fully remitted from
`the first phase, the clinical scores were again reduced to a
`level much below that of untreated controls and then
`maintained (Fig. 4B), although the eventual maintained
`clinical scores were not quite as low as when dosing was
`initiated at day 14 or earlier.
`It has been shown (Mandala et al., 2002) that the effects
`of FTY720 are a result of the generation of the metabolite
`FTY-P, which is an agonist at four of the five known S1P
`
`receptors. We therefore used FTY-P in further experiments
`to examine the dose response for clinical efficacy and
`peripheral
`lymphopoenia, and the relationship between
`these two phenomena. In initial experiments, FTY-P was
`given from day 14 to mice induced for EAE and cumulative
`clinical scores were determined.
`The pattern of clinical response was similar to that
`described for FTY720 itself and was dose responsive. There
`was a rapid initial improvement in clinical scores of the
`mice in the highest dosage group, with lower doses giving
`intermediate effects. On cessation of dosing at day 25, there
`was an initial delay of about 2 days, after which the mice
`rapidly relapsed to disease scores similar to or more severe
`than those seen in untreated controls (Fig. 5A). Clinical
`scores were added cumulatively during the dosing period,
`and the data indicate that
`the level of clinical efficacy
`obtained was proportional to the dose of FTY-P adminis-
`tered (Fig. 5B).
`
`

`

`M. Webb et al. / Journal of Neuroimmunology 153 (2004) 108–121
`
`115
`
`lymphocytes when compared with sham-immunized (but
`PTX-treated) controls (Fig. 6A). Because of the profound
`effect of PTX treatment alone in elevating numbers of
`circulating lymphocytes, in subsequent experiments, we
`expressed the peripheral lymphocyte density data as a per-
`centage of the lymphocyte in drug-naı¨ve PLP/PTX-treated
`controls at each time point. After the correction for the effect
`of pertussis toxin, we found an FTY-P dose-dependent
`decrease in the numbers of peripheral blood lymphocytes
`which was first detectable 24 h after the first dose at the
`highest concentration (Fig. 6B). Five days after initiation of
`
`Fig. 5. Effect of FTY-phosphate on clinical state of mice induced for EAE.
`(A) Time course of EAE. FTY-P was given daily by ip injection between
`days 14 and 25 at 0.03, 0.3, or 1 mg/kg. (B) Cumulative clinical scores
`between days 15 and 25 for mice treated daily with 0.03, 0.3, or 1 mg/kg
`FTY-P between days 14 and 25.
`
`3.4. Lymphopoenia in FTY-treated mice
`
`A reversible sequestration of lymphocytes in peripheral
`lymph nodes has been described in animals treated with
`FTY720 (Chiba et al., 1998; Pinschewer et al., 2000; Man-
`dala et al., 2002; Xie et al., 2003). This sequestration has been
`proposed to be at least one of the mechanisms by which
`FTY720 achieves its therapeutic efficacy. We therefore
`measured the changes in numbers of lymphocytes in periph-
`eral blood of the mice receiving the FTY-P throughout the
`dosing regimen. As has been reported previously, (Sewell
`and Andrews, 1989), treating mice with pertussis toxin, as
`was done routinely in the EAE induction protocol, results in a
`large increase in the levels of circulating peripheral lympho-
`cytes, probably because these cells are unable to home to the
`peripheral lymphoid organs. Densities of circulating lym-
`phocytes were approximately 7000/mm3 in naı¨ve animals.
`Fifteen days after PTX treatment, these levels were substan-
`tially elevated and continued to fall until reaching normal
`levels about 30 days after PTX treatment. Actively immu-
`nized animals showed somewhat higher levels of peripheral
`
`Fig. 6. Lymphopoenia produced by FTY-phosphate treatment. (A) Time
`course of effect of pertussis toxin treatment with or without PLP im-
`munization on numbers of peripheral lymphocytes in blood samples. (B)
`Changes in numbers of peripheral blood lymphocytes (expressed as
`percentage of PTX treated but nondiseased controls at the same time point)
`at days 15, 20, 25, 29, and 34 in mice treated daily with FTY-P (0.03, 0.3, 1
`mg/kg) between days 14 and 25. C. Dose – response correlation between
`cumulative clinical scores (days 15 – 25) and lymphopoenia for mice treated
`with FTY-P.
`
`

`

`116
`
`M. Webb et al. / Journal of Neuroimmunology 153 (2004) 108–121
`
`dosing, a dose-dependent lymphopoenia was seen in all
`groups, with levels of peripheral lymphocytes continuing to
`fall in the two highest dose groups until day 24. Lower doses
`produced a lower magnitude of effect at all time points. At all
`doses, the nadir was attained by day 25, where the maximum
`depletion at the highest dose was about 75% of control levels.
`Cumulative clinical scores for the period 14 – 25 days corre-
`lated with the level of lymphopoenia (r2 = 0.91, Fig. 6C).
`When FTY-P dosing was terminated at day 25,
`the
`lymphopoenia was reversed. This occurred with a signifi-
`cant delay compared with the reemergence of clinical signs,
`such that at the highest dose, the numbers of peripheral
`lymphocytes at day 30 remained as low as they had been at
`
`day 20 (Fig. 6B). However, by this day, the clinical scores in
`this group had already reached a mean of about 2 (compare
`with < 0.5 at day 20, Fig. 5A) and thus there was an
`incomplete temporal match between the measures of lym-
`phopoenia and clinical score.
`
`3.5. Effect of FTY-P on gene expression in EAE
`
`We examined the effect of FTY-P on the changes in
`expression of some of the genes which were described
`previously. In this EAE experiment, animals dosed from
`day 14 with 1-mg/kg FTY-P had recovered to a very low
`clinical score by day 25 (Fig. 7A), while untreated animals,
`
`Fig. 7. Effect of FTY-phosphate treatment on gene expression changes at day 25 of EAE. (A) Time course of EAE for study of effect of FTY-P on spinal gene
`expression changes. (B – F) Effect of FTY-P (daily, 1 mg/kg from day 14) on changes in spinal expression of genes related to myelin (B, PLP; C, MBP; D,
`MOG) and inflammation-related (E, iNOS; F, GM-CSF) at day 25. Y axis, mean and S.E.M., n = 4 (MOG, MBP, PLP; n = 4 – 6, iNOS, GMCSF) of expression
`relative to h actin determined in same samples.
`
`

`

`M. Webb et al. / Journal of Neuroimmunology 153 (2004) 108–121
`
`117
`
`although in remission, still maintained a mean score of >1.
`In spinal cord samples, the previously observed reduction in
`expression of myelin-related genes was seen in EAE control
`mice (Fig. 7B – D), but in each case, treatment with 1-mg/kg
`FTY-P significantly increased expression levels. As seen
`previously for iNOS and GM-CSF, expression levels were
`elevated at day 25, relative to controls, but in this case, FTY-
`P treatment significantly attenuated the expression of both
`genes (Fig. 7E,F).
`
`3.6. Effect of FTY-P in adoptive transfer EAE
`
`We examined the effect of FTY-P in EAE induced in
`recipient SJL mice by adoptive transfer of lymph node cells
`from actively immunize