FTY720 Immunosuppression Impairs Effector T Cell
`Peripheral Homing Without Affecting Induction, Expansion,
`and Memory1
`
`Daniel D. Pinschewer,* Adrian F. Ochsenbein,* Bernhard Odermatt,† Volker Brinkmann,‡
`Hans Hengartner,* and Rolf M. Zinkernagel2*
`FTY720 (2-amino-2-(2-[4-octylphenyl]ethyl)-1,3-propanediol hydrochloride) prolongs survival of solid organ allografts in animal
`models. Mechanisms of FTY720 immunomodulation were studied in mice infected with lymphocytic choriomeningitis virus
`(LCMV) to assess T cell responses or with vesicular stomatitis virus to evaluate Ab responses. Oral FTY720 (0.3 mg/kg/day) did
`not affect LCMV replication and specific CTL and B cells were induced and expanded normally. Moreover, the anti-viral humoral
`immune responses were normal. However, FTY720 treatment showed first a shift of overall distribution of CTL from the spleen
`to peripheral lymph nodes and lymphocytopenia was observed. This effect was reversible within 7–21 days. Together with un-
`impaired T and B cell memory after FTY720 treatment, this finding rendered enhancement of lymphocyte apoptosis by FTY720
`in vivo unlikely. Secondly, the delayed-type hypersensitivity reaction to a viral MHC class I-presented peptide was markedly
`reduced by FTY720. These results were supported by impaired circulation of LCMV specific TCR transgenic effector lymphocytes
`in the peripheral blood and reduced numbers of tissue infiltrating CTL in response to delayed-type hypersensitivity reaction.
`Thirdly, in a CD8ⴙ T cell-mediated diabetes model in a transgenic mouse expressing the LCMV glycoprotein in the islets of the
`pancreas, FTY720 delayed or prevented disease by reducing islet-infiltrating CTL. Thus, FTY720 effectively reduced recirculation
`of CD8ⴙ effector T cells and their recruitment to peripheral lesions without affecting the induction and expansion of immune
`responses in secondary lymphoid organs. These properties may offer the potential to treat ongoing organ-specific T cell-mediated
`immunopathologic disease. The Journal of Immunology, 2000, 164: 5761–5770.
`
`sible mechanism of action (7, 8). However, the concentrations
`needed for lymphocyte apoptosis in vitro were several orders of
`magnitude greater than the blood concentrations in rats given
`FTY720 at the therapeutic dose of 1 mg/kg (9). In vivo FTY720
`has been reported to shift the distribution of lymphocytes from
`spleen to Peyer’s patches and peripheral lymph nodes and accel-
`erate homing of naive lymphocytes to these organs (9). Another
`study suggested that low blood lymphocyte counts may reflect re-
`duced emigration of effector cells to the periphery (10), but they
`could not distinguish effects on the circulation of naive mature
`lymphocytes from specific effector cells. Whether FTY720 inter-
`feres with the generation of specific effector T cells has not been
`shown. Therefore, the mechanisms responsible for the reported
`reduced graft infiltration under FTY720 treatment (10) remain un-
`clear. In the present study we analyzed the immunosuppressive
`mechanisms of 0.3 mg/kg FTY720 daily as monotherapy on T cell
`responses to lymphocytic choriomeningitis virus (LCMV) and B
`cell responses to vesicular stomatitis virus (VSV) in the mouse.
`This dose reveals biological effects in vivo and may be given to
`transplantation patients in combination therapy with other immu-
`nosuppressants. An in vivo half-life around or below 24 h (data not
`shown) excludes the possibility of major drug accumulation.
`Primary infection with LCMV, a noncytopathic RNA virus, is
`controlled almost exclusively by CTL (11, 12). Immunopathology
`in this infection is also T cell-mediated (11, 13, 14), and ␣␤TCR
`transgenic CD8⫹ T cells are available against a major CTL target
`LCMV glycoprotein peptide 33–41 (GP33); so LCMV infection
`offers a good model to assess beneficial or detrimental effects of
`immunomodulation. On the other hand, VSV, a highly cytopatho-
`genic RNA virus, was used to study the effects of FTY720 on B
`cell responses. An early neutralizing Ab response against VSV is
`
`F TY720
`
`(2-amino-2-(2-[4-octylphenyl]ethyl)-1,3-propanediol
`hydrochloride)3 is a chemical derivative of myriocin (ther-
`mozymocidin), a substance found in the fungi Myriococ-
`cum albomyces and Isaria sinclairii (1, 2). Despite the strong sup-
`pressive activity of myriocin in mixed lymphocyte reactions, this
`property is lost upon chemical modification to FTY720. FTY720
`has been shown to prevent graft rejection in various animal allo-
`transplantation models (3–5). Good synergy with cyclosporin and
`sirolimus has been reported (5, 6). The current strategy of immu-
`nosuppressive regimens is to combine different substances with
`distinct mechanisms of action to minimize the dose and toxicity of
`each individual therapeutic constituent.
`In vitro studies on FTY720 suggested that selective apoptotic
`cell death of lymphocytes, mainly of the T cell subset, was a pos-
`
`*Institute of Experimental Immunology and †Laboratory for Special Techniques, De-
`partment of Pathology, University Hospital, Zurich, Switzerland; and ‡Novartis
`Pharma AG, Basel, Switzerland
`Received for publication May 7, 1999 Accepted for publication March 21, 2000
`The costs of publication of this article were defrayed in part by the payment of page
`charges This article must therefore be hereby marked advertisement in accordance
`with 18 U S C Section 1734 solely to indicate this fact
`1 This work was supported by the Swiss National Science Foundation, the Kanton of
`Zurich, and Novartis Pharmaceuticals (Basel, Switzerland)
`2 Address correspondence and reprint requests to Dr Rolf M Zinkernagel, Institute of
`Experimental Immunology Schmelzbergstrasse 12, 8091 Zurich, Switzerland E-mail
`address: rolf zinkernagel@pty usz ch
`3 Abbreviations used in this paper: FTY720, 2-amino-2-(2-[4-octylphenyl]ethyl)-1,3-
`propanediol hydrochloride; LCMV,
`lymphocytic choriomeningitis virus; CFSE,
`5-(and 6-)carboxyfluorescein diacetate succinimidyl ester; LCMV-WE, LCMV strain
`WE; LCMV-NP, nucleoprotein of LCMV; LCMV-GP33/GP33, peptide 33–41 of the
`LCMV glycoprotein; vaccinia-WR, vaccinia virus strain WR; VSV, vesicular stoma-
`titis virus; VSV-IND, VSV strain Indiana; VSV-NJ, VSV strain New Jersey; VSV-
`NP52/NP52, peptide 52–59 of the VSV nucleoprotein; tg, transgenic; DTH, delayed-
`type hypersensitivity reaction
`
`Copyright © 2000 by The American Association of Immunologists
`
`0022-1767/00/$02 00
`
`Apotex v. Novartis
`IPR2017-00854
`NOVARTIS 2049
`
`

`

`5762
`
`FTY720 IMPAIRS EFFECTOR T CELL PERIPHERAL HOMING
`
`required to prevent lethal encephalitis, and VSV infection induces
`lifelong protective B cell memory (15, 16).
`We now show that FTY720 does not impair efficient priming of
`T and B cell/Ab responses, but inhibits immune responses by pre-
`venting the homing of effector T cells to lesions in peripheral
`organs.
`Materials and Methods
`Mice
`C57BL/6 mice (H-2b), 318 TCR transgenic mice, expressing a TCR spe-
`cific for LCMV-GP33 together with H-2Db, and RIP-GP transgenic mice
`expressing the LCMV-GP on pancreatic islet cells were obtained from the
`breeding colony of the Institut fu¨r Labortierkunde, Veterinary Hospital
`(Zurich, Switzerland) Breeding was performed under specific pathogen-
`free conditions, and experiments were performed under conventional ani-
`mal facility conditions
`Viruses, virus detection, and inactivation
`Viruses. LCMV-WE was originally obtained from Dr F Lehmann-
`Grube (Heinrich Pette Institute, Hamburg, Germany) and was grown on
`L929 cells VSV Indiana (VSV-IND; Mudd-Summers isolate) and VSV
`New Jersey (VSV-NJ; Pringle isolate) were obtained from Dr D Kolak-
`ovsky (University of Geneva, Geneva, Switzerland) and were grown on
`BHK21 cells Vaccinia-WR was generously provided by Dr B Moss (Na-
`tional Institutes of Health, Bethesda, MD) and was propagated on BSC40
`cells
`Infectious LCMV in the organs was detected by focus-forming assays
`performed as previously described (17)
`For UV inactivation a small volume of high titer VSV preparation was
`exposed as a thin layer in a petri dish to a UV lamp (Philips, 15W) for 3
`min at a distance of 8 centimeters (18)
`Cell analysis and staining
`Cell analysis and counting of absolute numbers of transferred cells circu-
`lating in the blood were performed by FACS by mixing a defined number
`of autofluorescent beads with a known volume of blood to permit the
`calculation of positive cell number per microliter, using Becton Dickinson
`Trucount Tubes (Mountain View, CA) In addition to the manufacturer’s
`instructions, the absolute number of autofluorescent beads per sample was
`re-evaluated by FACS with the instrument settings used for the respective
`cell analysis Relative cell numbers varied within a range of ⫾2%; truly
`absolute numbers are, however, less precisely measurable
`For counting of transferred specific CTL present in secondary lymphoid
`organs single-cell suspensions were prepared as described for cytotoxicity
`assays The total number of lymphocytes per organ was assessed in a
`Neubauer chamber (improved bright-line, Brand Wertheim, Germany)
`The percentage of CD8⫹V␣2⫹V␤8⫹ triple-positive cells in the lympho-
`cyte population was evaluated by FACS and multiplied by the total number
`of lymphocytes to obtain the absolute number of CD8⫹V␣2⫹V␤8⫹ indi-
`cator CTLs per organ
`The fluorescent dye CFSE from Molecular Probes (Eugene, OR) was
`used to label cells before transfer as previously described (19)
`TCR transgenic CD8⫹ T cells were detected with anti-mouse CD8␣-
`TRI Ab from Caltag (Burlingame, CA), anti-mouse V␤8 1/8 2-FITC, and
`anti-mouse V␣2-PE Abs (PharMingen, San Diego, CA) CFSE-labeled
`CD4⫹ and CD8⫹ T cells were stained with anti-mouse CD4-TRI and anti-
`mouse CD8-TRI (both from Caltag), respectively
`Assessment of the primary footpad swelling reaction and of
`DTH/footpad infiltrating CTL
`LCMV-WE (200 PFU) in 30 ␮l of MEM supplemented with 2% FCS and
`GP33 (3 ␮g/ml) or NP52 (1 mg/ml) in 30 ␮l of balanced salts solution was
`injected into both hind footpads The footpad thickness was measured with
`a spring-loaded caliper (Kroeplein, Schluchtern, Hessen, Germany) Val-
`ues were taken as the mean of both hind footpads (20, 21)
`CTL assay and peptides
`Virus-specific cytotoxic T cells were assayed as described previously (22)
`Briefly, single-cell suspensions were prepared from the spleens and lymph
`nodes of mice at the indicated time points and used directly in a 51Cr
`release assay Target cells were either GP33-coated EL4 or vaccinia-WR-
`infected MC57 cells (multiplicity of infection of 3 for 2 h) The LCMV
`peptide GP33 and the VSV peptide NP52–59 were purchased from Neo-
`system Laboratoire (Strasbourg, France)
`
`Ab detection
`VSV-NJ and VSV-IND neutralizing Abs were determined in a neutraliza-
`tion assay as described previously (23) IgG was determined by its resis-
`tance to reduction by ␤-ME The difference between total Ig and IgG rep-
`resents IgM LCMV-NP-specific Abs were measured by ELISA using the
`following steps as described previously (24): 1) coating with baculovirus-
`derived LCMV-NP, 2) blocking with 2% BSA in PBS, 3) 10-fold predi-
`luted mouse serum was titrated in 3-fold dilutions over 12 steps, 4) IgM-
`or IgG-specific horseradish peroxidase-labeled goat anti-mouse Abs (0 5
`␮g/ml; Southern Biotechnology Associates, Birmingham, AL), 5) substrate
`2,2⬘-azinobis-3-ethylbenzthiazolazine sulfonic acid (Roche, Mannheim,
`Germany) and H2O2 (Fluka Buchs, Switzerland) Ab titers were deter-
`mined as the serum dilutions yielding an absorption (OD405) of twice back-
`ground levels
`Diabetes induction and assessment of blood glucose levels
`For the induction of an autoimmune diabetes, RIP-GP transgenic mice
`were infected i v with 200 PFU of LCMV-WE This elicits diabetes in
`RIP-GP transgenic but not in C57BL/6 mice (25) (see also Results) Glu-
`cose levels in the blood were measured using the Haemo-Glucotest color
`reaction sticks and were quantified with the reflection photometer Reflolux
`II (Roche) according to the manufacturer’s instructions
`Immunohistology
`Freshly removed organs were immersed in HBSS and snap-frozen in liquid
`nitrogen Tissue sections of 5-␮m thickness were cut in a cryostat, placed
`on siliconized glass slides, air-dried, fixed with acetone for 10 min, and
`stored at ⫺70°C For the staining of lymphocyte differentiation markers,
`rehydrated sections were incubated with primary rat mAbs against CD4
`and CD8 (YTS 191 and YTS 169) (26) Insulin was detected using affinity-
`purified guinea pig primary Abs (Dako, Glostrup, Denmark) Primary Abs
`were revealed by sequential incubation with either goat anti-rat Ig Abs
`(Tago, Burlingame, CA) and alkaline phosphatase-labeled donkey anti-
`goat Ig Abs or alkaline phosphatase-labeled rabbit anti-guinea pig Ig Abs
`and alkaline phosphatase-labeled goat anti-rabbit Ig Abs (Jackson Immu-
`noResearch Laboratories, West Grove, PA) Dilutions of the affinity-puri-
`fied secondary Abs were made in Tris-buffered saline containing 5% nor-
`mal mouse serum Alkaline phosphatase was visualized using naphthol
`AS-BI (6-bromo-2-hydroxy-3-naphtholic acid-2-methoxy anilide) phos-
`phate and new fuchsin as substrate Endogenous alkaline phosphatase was
`blocked by levamisole Color reactions were performed at room tempera-
`tures for 15 min with reagents from Sigma (St Louis, MO) Sections were
`counterstained with hemalum, and coverslips were mounted with glycerol
`and gelatin
`FTY720 administration
`FTY720 was obtained from Novartis Pharmaceuticals (Basel, Switzerland)
`and was dissolved in distilled water A dose of 0 3 mg/kg was given daily
`by gavage in a volume of 100 ␮l/10 g body weight
`Statistical analysis
`We performed repeated measures ANOVA with two between factors
`(group and trial) and one within factor (time point or E:T cell ratio, re-
`spectively) In experiments with two groups, post-hoc comparisons at dif-
`ferent time points were performed using unpaired t tests with Bonferroni
`correction In experiments with three groups, the Bonferroni-Dunn test was
`used for pairwise between-group comparisons Between-group compari-
`sons on different days were performed using a two-way ANOVA with
`post-hoc Bonferroni-Dunn test and an additional Bonferroni correction for
`tests on various days Cell counts were log-transformed to achieve an ap-
`proximate normal distribution and compared using a two-way ANOVA
`(factors trial and group) with Bonferroni correction for testing of multiple
`organs The figures show the data of only one representative experiment,
`while two or three comparable sets of experiments were used for statistical
`analysis as stated in the figure legends
`Results
`Reduced primary swelling reaction after LCMV infection
`Infection of C57BL/6 mice with 200 PFU of LCMV into the hind
`footpad causes a primary swelling reaction starting on day 6 and
`peaking around days 7–9. The first peak of the swelling reaction
`from days 6 to 8 is caused by CD8⫹ cells, whereas the following
`lesser swelling from days 9 to 14 is mainly CD4⫹ T cell dependent
`
`

`

`The Journal of Immunology
`
`5763
`
`FIGURE 1.
`Impairment of the primary footpad swelling reaction by
`FTY720 C57BL/6 mice were either treated from day ⫺1 until day 15 daily
`with 0 3 mg/kg FTY720 orally (f) or left untreated (䡺; control) On day
`0 they were immunized with 200 PFU of LCMV-WE diluted in a volume
`of 30 ␮l of MEM/2% FCS into the hind footpad Footpad swelling was
`monitored as described in Materials and Methods Each symbol represents
`a group of three mice One representative of two statistically analyzed
`experiments is shown Values indicate the mean ⫾ SD ⴱ, p ⬍ 0 005
`
`(21). This simple in vivo readout offers a reliable assessment of
`overall T cell-mediated immune responses. Immunosuppression
`by any mechanism modulating this response is readily revealed
`(27). One group of mice was treated daily with 0.3 mg/kg FTY720
`orally starting 1 day before infection, and the control group was
`left untreated. FTY720 treatment significantly reduced the swell-
`ing during the entire period ( p ⫽ 0.001; Fig. 1). This suggested
`that the early CD8⫹ as well as the CD4⫹ effector T cell activity
`were impaired at least as assessed locally in the footpad.
`
`Unimpaired induction and expansion of cytotoxic T cells
`To evaluate potential effects of FTY720 on T cell induction and
`expansion (28, 29) vs T cell effector function in the footpad, the
`following experiments were performed. Mice were immunized
`with 200 PFU of LCMV-WE i.v., and CTL activity was measured
`at its peak on day 8. Because a difference in the distribution of
`naive T cells had already been shown (9), not only splenocytes but
`also lymph node cells were tested as effectors in a 51Cr release
`assay (Fig. 2). An obvious decrease in cytolytic activity in the
`spleen ( p ⫽ 0.004; Fig. 2A) after FTY720 treatment contrasted
`with an increase in cytotoxic activity in peripheral lymph nodes
`( p ⫽ 0.01; Fig. 2B). This finding rendered a general reduction in
`the cytotoxic T cell population unlikely. Cells from pooled popli-
`teal, inguinal, mesenteric, para-aortic, axillary, and submandibular
`lymph nodes and spleen yielded the same cytotoxic activity in
`treated and untreated mice ( p ⫽ 0.74; Fig. 2C). The same exper-
`iment was repeated with vaccinia-WR with similar results (Fig. 2,
`D–F). FTY720-treated mice exhibited lower cytotoxic activity in
`the spleen ( p ⫽ 0.0001) and moderately more cytotoxicity in
`lymph nodes ( p ⫽ 0.01), but activity in pooled secondary lym-
`phoid organs was comparable to that in untreated control mice
`( p ⫽ 0.29). Despite an only moderate difference in lymph nodes
`(Fig. 2, B and E), statistical analysis of more than one comparable
`experiment resulted in overall significantly higher activity in
`lymph nodes of FTY720-treated mice. This finding should be
`viewed in the context of additional data presented below. Absolute
`numbers of lymphocytes found in pooled secondary lymphoid or-
`gans were about the same as those in controls (not shown). Taken
`together, these data show that a dose of 0.3 mg/kg FTY720 did not
`interfere significantly with the induction or expansion of specific
`cytotoxic T cells, but FTY720 apparently changed the distribution
`pattern of specific activated CD8⫹ effector T cells in a manner
`similar to that found for naive lymphocytes (9). To exclude effects
`
`FIGURE 2. Unimpaired induction and expansion of cytotoxic T cells
`under FTY720 treatment A–C, C57BL/6 mice were either treated with 0 3
`mg/kg FTY720 orally daily from day ⫺1 until the end of the experiment
`(f) or left untreated (䡺; control) After immunization with 200 PFU of
`LCMV-WE i v on day 0, mice were sacrificed on day 8, and cells from
`spleen and lymph nodes were tested in a primary in vitro cytotoxicity
`assay Specific cytotoxic activity against GP33-labeled EL-4 target cells
`was measured for spleen cells (A), for inguinal lymph node cells (B), and
`for pooled cells from popliteal, inguinal, mesenteric, para-aortal, axillary,
`and submandibular lymph nodes and spleen (C) Each symbol represents a
`group of three mice One representative of two statistically analyzed ex-
`periments is shown Spontaneous release was always ⬍20% Values indi-
`cate the mean ⫾ SD D–F, Mice were treated as described in A–C, but were
`immunized on day 0 with 2 ⫻ 106 PFU of vaccinia-WR i v On day 6 mice
`were sacrificed, and spleen and lymph node cells were tested in a primary
`in vitro cytotoxicity assay against vaccinia-WR-infected MC57 target cells
`(D, spleen cells; E, pooled lymph node cells; F, lymph node and spleen
`cells pooled) One representative of two or three statistically analyzed ex-
`periments is shown Spontaneous release was always ⬍20% Values indi-
`cate the mean ⫾ SD
`
`of FTY720 on viral replication, mice were infected with 200 PFU
`of LCMV i.v. Four days later viral titers were measured in the
`spleen and inguinal lymph nodes and were comparable in treated
`and untreated mice (data not shown).
`
`Induction of lymphocytopenia and its reversibility
`In vitro studies had suggested that specific apoptotic death of T
`cells was responsible for the immunosuppressive properties of
`FTY720 (7, 8, 30). Apoptosis might have been expected to reduce
`the cytotoxic effector population, but the results of overall unim-
`paired induction of CTL responses did not readily support this
`idea. In addition, previous studies showed a transient lymphocy-
`topenia after a single dose of FTY720 (3, 9), but did not evaluate
`
`

`

`5764
`
`FTY720 IMPAIRS EFFECTOR T CELL PERIPHERAL HOMING
`
`FIGURE 4.
`Impaired circulation of CD8⫹ cytotoxic T cells C57BL/6
`mice were either treated from day ⫺1 until day 15 daily with 0 3 mg/kg
`FTY720 orally (f) or left untreated (䡺; control) and infected with 200
`PFU of LCMV-WE i v on day 0 The mice had received 104 318-tgTCR
`V␣2⫹V␤8⫹ spleen cells specific for Db and LCMV-GP33 i v on day ⫺2
`tgTCR⫹ (V␣2⫹V␤8⫹) CD8⫹ T cells in the blood were measured at the
`time points indicated One representative experiment of three is shown
`Values indicate the mean ⫾ SD
`
`of specific effector T cells (31). To follow specific effector T cells
`in the blood of mice, 104 indicator spleen cells from mice with a
`transgenic TCR specific for Db and LCMV-GP33 (318-mouse)
`were transferred to naive recipients that were infected i.v. with 200
`PFU of LCMV-WE. It has previously been shown that this number
`of transferred cells replaces about 50% of the endogenous GP33-
`specific CD8⫹ T cells without changing the overall response
`against LCMV (32, 33). The number of specific circulating trans-
`genic TCR⫹ CD8⫹ T cells per microliter of blood was measured
`by FACS analysis. FTY720 treatment reduced the number of spe-
`cific circulating CD8⫹ T cells by more than a factor of 10 (Fig. 4).
`
`Comparable numbers, but changed distribution pattern, of
`specific CD8⫹ cytotoxic effector T cells
`Comparing the above finding with the data from Fig. 2, it was
`rather unlikely that a reduced population of specific CD8⫹ cyto-
`toxic effector T cells was the underlying reason for their reduced
`circulation in the peripheral blood (Fig. 4). To corroborate this
`conclusion and to further analyze the effect of FTY720 on the
`distribution pattern of specific CD8⫹ cytotoxic effector T cells,
`mice were treated as described in Fig. 4 and were sacrificed at the
`peak of the CTL response on day 8 after infection. Transgenic
`TCR⫹ CD8⫹ T cells were counted in blood, spleen, inguinal
`lymph node, and pooled secondary lymphoid organs (spleen
`pooled with popliteal, inguinal, mesenteric, para-aortal, axillary,
`and submandibular lymph nodes) as described in Materials and
`Methods (Fig. 5). Comparable numbers of indicator CTL were
`present in pooled secondary lymphoid organs of FTY720-treated
`and untreated mice ( p ⫽ 0.52), while drug treatment significantly
`reduced the number of CTL circulating in the blood ( p ⬍ 0.0001).
`Similar to the behavior of naive lymphocytes (9) and compatible
`with the data shown in Fig. 2, A and D, fewer specific CD8⫹
`cytotoxic effector T cells could be found in the spleen of FTY720-
`treated mice compared with untreated control mice ( p ⫽ 0.0003),
`while their number was increased in the inguinal lymph node ( p ⫽
`0.0001). This result taken together with the data from Figs. 1, 2,
`and 4 suggested that FTY720 impairs the circulation of effector
`CD8⫹ T cells, apparently by sequestering them in secondary lym-
`phoid organs, mainly peripheral lymph nodes, without impairing
`their induction or expansion.
`
`FIGURE 3.
`Induction of lymphocytopenia by FTY720 and its revers-
`ibility On day ⫺1, 3 ⫻ 107 naive CFSE-labeled splenocytes were adop-
`tively transferred i v into naive syngeneic C57BL/6 recipients The recip-
`ients were then treated with a single dose of 0 3 mg/kg FTY720 orally on
`day 0 (䡺), treated with the same dose daily from day 0 until day 6 (Œ), or
`left untreated (control ⫽ 100%) The numbers of transferred, CFSE-posi-
`tive circulating CD4⫹ (A) and CD8⫹ (B) T cells per microliter of blood
`were measured by FACS analysis; staining of transferred cells was per-
`formed as described in Materials and Methods Gated living white cells
`were analyzed for CFSE⫹CD4⫹ (A) and CFSE⫹CD8⫹ (B) double-positive
`cells, respectively The values are expressed as a percentage of the mean of
`the control group Each group consisted of four animals One representa-
`tive experiment of two is shown Values indicate the mean ⫾ SD
`
`recovery due to newly generated lymphocytes after FTY720 treat-
`ment was stopped. To address this question, naive spleen cells
`were labeled with the vital dye CFSE (19) and adoptively trans-
`ferred i.v. to naive recipients. The recipients were then treated
`either with a single dose of 0.3 mg/kg FTY720 or with the same
`dose for 7 consecutive days. The numbers of transferred CD4⫹
`(Fig. 3A) and CD8⫹ T cells (Fig. 3B) in the peripheral blood were
`measured by FACS and compared with those in untreated control
`mice. A marked drop in both populations was observed a few
`hours after a single dose of 0.3 mg/kg FTY720; this drop was more
`severe after repetitive treatments. Within 7 or 21 days after treat-
`ment was stopped, respectively, transferred CD4⫹ and CD8⫹ T
`cells in the blood had recovered levels within 10% of the untreated
`control values. The same recovery was found for B lymphocytes
`(data not shown). Additional experiments had shown that the cell
`numbers slowly but continuously recovered during the time span
`of 7 and 21 days, respectively (data not shown). Recovery due to
`proliferation could be excluded, because proliferation and the re-
`sulting CFSE dilution would have been easily detected by alter-
`ation of the fluorescence intensities (19). These results indicate that
`the recovery of circulating lymphocytes after transient FTY720
`application is not due to newly generated cells, suggesting that
`FTY720 probably causes sequestration of lymphocytes rather than
`significantly affecting their viability.
`
`Impaired circulation of specific CD8⫹ cytotoxic effector T cells
`The obvious discrepancy between the reduction of the primary
`swelling reaction and the overall efficient induction and expansion
`of cytotoxic T cells required analyses of the circulation properties
`
`

`

`The Journal of Immunology
`
`5765
`
`FIGURE 6. Reduced DTH reaction under FTY720 treatment C57BL/6
`mice were either treated daily with 0 3 mg/kg FTY720 orally starting on
`day ⫺1 (f) and on day 5 after infection (Œ), respectively, until the end of
`the experiment or were left untreated (䡺 control) Thirteen days after in-
`fection with 200 PFU of LCMV-WE i v on day 0, 30 ␮l of GP33 solution
`in a concentration of 3 ␮g/ml dissolved in BSS was injected into both hind
`footpads Footpad swelling was measured as described in Materials and
`Methods Each symbol represents a group of five mice One representative
`of two statistically analyzed experiments is shown Values indicate the
`mean ⫾ SD ⴱ, p ⬍ 0 0016 (for f vs 䡺 and for Œ vs 䡺 only, but not for
`f vs Œ)
`
`control, p ⫽ 0.0001) to the same extent as had been found in mice
`treated from day ⫺1 onward (FTY720 day 1 vs FTY720 day 5,
`p ⫽ 0.51; Fig. 6).
`
`Prevention of a CD8⫹ T cell-mediated autoimmune diabetes
`Interference with the homing of effector T cells to peripheral le-
`sions may offer new therapeutic possibilities against immuno-
`pathological or autoimmune disease. To examine this, we analyzed
`a transgenic mouse expressing the glycoprotein of LCMV in the
`islets of the pancreas (RIP-GP) (25). In this model the transgenic
`GP is ignored by CD8⫹ T cells in the naive mouse, but after
`infection with LCMV, CD8⫹ T cell-mediated autoimmune diabe-
`tes develops because potentially autoreactive T cells have not been
`deleted in the thymus or the periphery, and effector CTLs are ef-
`ficiently induced following LCMV infection of secondary lym-
`phoid organs (25). Consequent perforin-dependent killing by CTLs
`leads to islet destruction and hyperglycemia (34). In a total of three
`experiments, daily treatment with 0.3 mg/kg FTY720 starting 1
`
`FIGURE 7. Prevention of CD8⫹ T cell-mediated autoimmune diabetes
`RIP-GP transgenic mice were either treated with 0 3 mg/kg FTY720 orally
`daily from day ⫺1 until day 20 (f) or left untreated (䡺) On day 0 they
`were immunized with 200 PFU of LCMV-WE i v Glucose levels in the
`blood were measured as described in Materials and Methods Each symbol
`represents an individual mouse One representative experiment of three is
`shown
`
`FIGURE 5. Changed distribution pattern of specific cytotoxic T cells
`C57BL/6 mice were either treated from day ⫺1 until day 8 daily with 0 3
`mg/kg FTY720 orally (f) or left untreated (䡺) and were infected with 200
`PFU of LCMV-WE i v on day 0 The mice had received 104 318-tgTCR
`V␣2⫹V␤8⫹ spleen cells specific for Db and LCMV-GP33 i v on day ⫺2
`Mice were sacrificed at day 8 after infection, and tgTCR⫹ (V␣2⫹V␤8⫹)
`CD8⫹ T cells were counted in blood, spleen, inguinal lymph node, and
`pooled secondary lymphoid organs (pooled spleen and popliteal, inguinal,
`mesenteric, para-aortal, axillary, and submandibular lymph nodes) as de-
`scribed in Materials and Methods One representative of two statistically
`analyzed experiments is shown Values indicate the mean ⫾ SD
`
`Reduced LCMV-specific delayed-type hypersensitivity reaction
`(DTH)
`To study the recruitment of effector CD8⫹ T cells to s.c. sites with
`Ag, mice were immunized with 200 PFU of LCMV-WE i.v. and
`13 days later were treated with the immunodominant Db nonapep-
`tide, gp33–41 of LCMV-glycoprotein (GP33). This peptide elicits
`an exclusively CD8⫹ T cell-mediated DTH in LCMV-infected, but
`not in naı¨ve, C57BL/6 mice (20). Footpad injection of the unre-
`lated immunodominant Db binding octapeptide np52–59 of VSV-
`nucleoprotein (NP52), which causes a similar reaction in VSV-
`infected mice (20), did not lead to any footpad swelling (data not
`shown). Daily treatment with 0.3 mg/kg FTY720 starting 1 day
`before infection significantly reduced the CD8⫹ T cell-mediated
`DTH against GP33 (FTY720 day 1 vs control, p ⫽ 0.0001; Fig. 6).
`To avoid an effect of FTY720 on early phases of the anti-LCMV
`response, a third group of mice was only treated with FTY720
`from day 5 after infection onward. Consistent with earlier findings
`(4, 6), the evoked DTH reaction was reduced (FTY720 day 5 vs
`
`

`

`5766
`
`Fl'Y'lZO IMPAJRS EFFECTOR T CELL PERIPHERAL HOMING
`
`.‘
`‘
`V
`. m‘
`‘
`y
`-
`.
`fa.’§( 6.\.‘.‘; ANN-O, ‘\'“.\‘ :t'
`V
`-’ '-
`2 -
`
`
`‘
`'5'.
`"s-
`!v‘.
`‘ “a “‘ “2‘
`51-.
`337 ' fl’f‘naflhi w’ 3W u‘
`,
`.
`.
`,V 1,;
`. 1.. «.
`‘
`
`
`,
`
`

`

`The Journal of Immunology
`
`5767
`
`tration, the following experiments were performed. RIP-GP mice
`were treated as described above (Fig. 7) and sacrificed on days 9
`and 20 after infection for histological examination. On day 9 con-
`trol mice developed diabetes and exhibited considerable CD8⫹
`infiltrates (Fig. 8B), while only very few CD8⫹ T cells were visible
`around the pancreatic islets of FTY720-treated mice (Fig. 8F).
`Corresponding to the degree of diabetes they had developed (Fig.
`7), a similar reduction of CD8⫹ cytotoxic T cells could also be
`seen 20 days after infection (Fig. 8H, normoglycemic mouse) com-
`pared with dense aggregates in untreated control mice (Fig. 8D).
`Parallel to increased CD8⫹ T cell infiltration staining for insulin
`decreased (Fig. 8, A, C, E, andG).
`C57BL/6 mice were treated with 0.3 mg/kg FTY720 orally daily
`starting 1 day before infection with 200 PFU of LCMV-WE i.v. on
`day 0. On day 13 LCMV-GP33 was injected into the hind footpad,
`and the mice were sacrificed 1, 3, 8, and 18 h later. Histological
`sections of the footpad were analyzed for infiltrating CD8⫹ and
`CD4⫹ T cells. One hour after peptide injection, some perivascular
`CD8⫹ T cells could be found in the s.c. tissue of untreated control
`mice (Fig. 8I), while no infiltration was seen in FTY720-treated
`mice (Fig. 8M). Two hours later massive aggregates of perivascu-
`lar CD8⫹ T cells and a beginning infiltration of approximate sweat
`glands (Fig. 8J) were considerably reduced by FTY720 (Fig. 8N).
`Eight hours after peptide challenge FTY720 treatment reduced the
`extensive infiltration of sweat glands (Fig. 8, K and O) and muscle
`(not shown). Another 10 h later, decreasing numbers of CD8⫹ T
`cells had caused considerable destruction and edema in untreated
`mice (Fig. 8L), while in the footpad of FTY720-treated mice
`CD8⫹ T cells had caused correspondingly less damage (Fig. 8P).
`Only very few to no CD4⫹ T cells were found in the samples
`above (Fig. 8, Q, R, U, andV, and data not shown) independent of
`FTY720 treatment. Present mainly at later time points, CD4⫹ T
`cells were probably attracted by the local inflammatory process
`initiated by CTLs. Footpad injection of the Db binding immuno-
`dominant VSV peptide NP52 did not cause either swelling (not
`shown) or considerable infiltration of CD8⫹ (Fig. 8, S, T, W, and
`X) or CD4⫹ T cells (data not shown) independent of FTY720
`treatment. Similarly, GP33 did not cause infiltration in naive
`C57BL/6 mice (data not shown). This suggests that the infiltrates
`shown in Fig. 8,I—