IMMUNOLOGY
`
`Induction of CD4ⴙCD25ⴙ regulatory T cells by
`copolymer-I through activation of transcription
`factor Foxp3
`
`Jian Hong*, Ningli Li†, Xuejun Zhang‡, Biao Zheng‡, and Jingwu Z. Zhang*†‡§¶
`
`†Joint Immunology Laboratory of Institute of Health Sciences and Shanghai Institute of Immunology, Shanghai Institutes for Biological Sciences, Chinese
`Academy of Sciences and Shanghai Second Medical University, Shanghai 200025, China; Departments of *Neurology and ‡Immunology, Baylor College
`of Medicine, Houston, TX 77030; and §E-Institute of Shanghai Universities, Shanghai 200025, China
`
`Communicated by Zhu Chen, Shanghai Second Medical University, Shanghai, China, March 27, 2005 (received for review December 3, 2004)
`
`Copolymer-I (COP-I) has unique immune regulatory properties and
`is a treatment option for multiple sclerosis (MS). This study re-
`vealed that COP-I induced the conversion of peripheral CD4ⴙCD25ⴚ
`to CD4ⴙCD25ⴙ regulatory T cells through the activation of tran-
`scription factor Foxp3. COP-I treatment led to a significant increase
`in Foxp3 expression in CD4ⴙ T cells in MS patients whose Foxp3
`expression was reduced at baseline. CD4ⴙCD25ⴙ T cell lines gen-
`erated by COP-I expressed high levels of Foxp3 that correlated with
`an increased regulatory potential. Furthermore, we demonstrated
`that the induction of Foxp3 in CD4ⴙ T cells by COP-I was mediated
`through its ability to produce IFN-␥and, to a lesser degree, TGF-␤1,
`as shown by antibody blocking and direct cytokine induction of
`Foxp3 expression in T cells. It was evident that in vitro treatment
`and administration with COP-I significantly raised the level of
`Foxp3 expression in CD4ⴙ T cells and promoted conversion of
`CD4ⴙCD25ⴙ regulatory T cells in wild-type B6 mice but not in IFN-␥
`knockout mice. This study provides evidence for the role and
`mechanism of action of COP-I in the induction of CD4ⴙCD25ⴙ
`regulatory T cells in general and its relevance to the treatment
`of MS.
`
`IFN-␥ 兩 multiple sclerosis
`
`Copolymer-I (COP-I) is a random polymer of four amino
`
`acids (glutamic acid, lysine, alanine, and tyrosine) enriched
`in myelin basic protein (MBP) and has proven treatment efficacy
`for multiple sclerosis (MS) (1–3). Although COP-I has been
`approved as a treatment option for MS since 1995, the mecha-
`nism of action of COP-I in relation to its treatment efficacy is
`unclear (4). COP-I has been found to activate T cells in human
`and animal experimental systems (5–7). The relevance of this
`unique property of COP-I to its efficacy in MS is thought to
`partially involve a biased induction of T helper (Th)2 immunity.
`There is evidence suggesting that human T cell lines generated
`by COP-I initially secrete Th1 (IL-2 and IFN-␥) and Th2
`cytokines (IL-4, IL-6, and IL-10) in response to COP-I (8).
`However, repeated in vitro stimulation of these T cell lines
`progressively shifts cytokine production toward the Th2 re-
`sponse (9, 10). Similarly, repeated COP-I injections may lead to
`deviation from Th1 to Th2 response in patients with MS (11, 12).
`Studies reported by other investigators, however, indicate that
`the effect of COP-I on the induction of T cell activation is not
`entirely selective for Th2 cells and that it consistently activates
`the production of Th1 and Th2 cytokines in MS (13). Other
`plausible mechanisms have been proposed that include its
`inhibitory property on the T cell responses to MBP containing
`the four frequently appearing amino acids that COP-I comprises
`(14, 15). This effect is thought to involve competition between
`COP-I and MBP for binding sites on MHC class II molecules (16,
`17). However, recent studies suggest that the inhibition of COP-I
`on T cells is not entirely specific for MBP because COP-I also
`affects the activation of T cells specific for two other myelin
`antigens, proteolipid protein and myelin oligodendrocyte glyco-
`
`protein, as well as the binding of these antigens to MHC class II
`molecules (18). Alternatively, this inhibitory effect of COP-I is
`attributed to ‘‘bystander’’ suppression of unknown mechanism.
`To date, the exact mechanism of action of COP-I remains
`elusive.
`In this study, we examined the proposed hypothesis that the
`unique properties of COP-I in the activation of T cells may
`induce CD4⫹CD25⫹ regulatory T cell responses. The hypothesis
`was prompted based on our initial discovery that COP-I was able
`to induce the expression of transcription factor Foxp3 in CD4⫹
`T cells, which is associated with CD4⫹CD25⫹ regulatory T cells
`(19–21). A potential role of COP-I in the induction of
`CD4⫹CD25⫹ regulatory T cell response is particularly relevant
`to MS because they have been recognized recently as an
`important regulatory component that keeps autoreactive T cells
`in check (19–22). Significant deficiencies in the number or
`function of these regulatory T cells have been found to correlate
`with several autoimmune conditions,
`including MS (23–25).
`CD4⫹CD25⫹ regulatory T cells can be distinguished from other
`CD4⫹ activated T cells of nonregulatory functions present in
`the CD4⫹CD25⫹ T cell pool by the expression of transcription
`factor Foxp3 (26). Gene transfer of Foxp3 converts naive T cells
`toward a regulatory T cell phenotype similar to that of naturally
`occurring CD4⫹ regulatory T cells (19, 27). Experiments were
`performed here to investigate whether COP-I was able to induce
`conversion of peripheral CD4⫹CD25⫺ T cells to CD4⫹CD25⫹
`regulatory T cells through the activation of Foxp3 in human and
`animal systems. Foxp3 expression and regulatory function of T
`cells were also analyzed ex vivo in MS patients with or without
`COP-I treatment and in mice administered with COP-I. Human
`COP-I-specific, short-term T cell
`lines were generated and
`characterized. The study described here has provided evidence
`indicating the role of COP-I in the induction of CD4⫹CD25⫹
`regulatory T cells through the activation of Foxp3.
`
`Materials and Methods
`Cell Stimulation. Fresh peripheral blood mononuclear cells
`(PBMC) were isolated from blood specimens by Ficoll hypaque
`separation. PBMC (2 ⫻ 105) or purified T cells (1 ⫻ 105) were
`cultured in the presence or absence of 40 ␮g兾ml COP-I alone or
`40 ␮g兾ml COP-I with irradiated T cell-depleted PBMC as a
`source of antigen-presenting cells (2 ⫻ 105) in complete RPMI
`medium 1640 in U-bottom 96-well plates at 37°C in 5% CO2. On
`day 4 of the culture, an aliquot of T cells was harvested for Foxp3
`gene expression by real-time PCR analysis. Supernatants were
`collected for cytokine detection, and 1-␮Ci (1 Ci ⫽ 37 GBq)
`
`Abbreviations: COP-I, copolymer-I; MBP, myelin basic protein; MS, multiple sclerosis; PBMC,
`peripheral blood mononuclear cells; Th, T helper.
`¶To whom correspondence should be addressed at: Department of Neurology, Baylor
`College of Medicine, 6501 Fannin Street, NB302, Houston, TX 77030. E-mail: jzang@
`bcm.tmc.edu.
`
`© 2005 by The National Academy of Sciences of the USA
`
`www.pnas.org兾cgi兾doi兾10.1073兾pnas.0502187102
`
`PNAS 兩 May 3, 2005 兩 vol. 102 兩 no. 18 兩 6449 – 6454
`
`Page 1 of 6
`
`YEDA EXHIBIT NO. 2073
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`[3H]thymidine (Amersham Pharmacia) was added to the re-
`maining cultures in the last 6 h ofculture before cell harvesting.
`
`Foxp3 mRNA Expression by Real-Time PCR. Quantitative real-time
`RT-PCR was performed on a Prism 7000 sequence detection
`system (Applied Biosystems). Hypoxanthine phosphoribosyl-
`transferase was used as a reference for sample normalization.
`Total RNA isolated from PBMC or purified T cells were
`reverse-transcribed into cDNA by using random hexamer. Hu-
`man Foxp3 primers (forward, 5⬘-CAC CTG GCT GGG AAA
`ATG G-3⬘; reverse, 5⬘-GGA GCC CTT GTC GGA TGA T-3⬘)
`and TaqMan minor groove binder probe (5⬘-FAM-ACT GAC
`CAA GGC TTC AT-3⬘) sequences were designed with the
`PRIMER EXPRESS application program (Applied Biosystems).
`Mouse Foxp3 primers and probe and all hypoxanthine phospho-
`ribosyltransferase primers and probe were purchased as forms of
`Assay-on-Demand (Applied Biosystems). The detailed amplifi-
`cation protocol used here is described in Supporting Materials
`and Methods, which is published as supporting information on
`the PNAS web site. A representative specific amplification of
`Foxp3 mRNA derived from a human CD4⫹CD25⫹ T cell
`preparation is shown in Fig. 8, which is published as supporting
`information on the PNAS web site.
`
`Isolation of Human CD4ⴙCD25ⴙ Regulatory T Cells. Isolation of
`human CD4⫹, CD4⫹CD25⫹, and CD4⫹CD25⫺ T cells was
`performed by using a human regulatory T cell isolation kit
`(Miltenyi Biotec, Auburn, CA) according to manufacturer’s
`instructions. Briefly, CD4⫹ T cells were first isolated through
`negative selection by removing all other cell types. Preisolated
`CD4⫹ T cells were incubated with 10 ␮l of magnetic beads
`conjugated with anti-CD25 antibody (for 107 cells) to separate
`CD4⫹CD25⫹ and CD4⫹CD25⫺ T cell populations. The purity of
`the resulting T cell populations was confirmed to be ⬎97% by
`flow cytometry.
`
`Generation of Specific T Cell Lines. PBMC were initially seeded out
`in 96-well round-bottom plates at 20,000 cells per well with
`irradiated autologous PBMC (100,000 cells per well) as acces-
`sory cells in 10% FBS RPMI medium 1640 in the presence of 40
`␮g兾ml COP-I. Cultures were supplemented with IL-2 at 50
`units兾ml after 48 h. After 7 days, all cultures per wells were
`assayed for specific reactivity to COP-I in proliferation assays.
`Briefly, each well was split into four aliquots (⬇104 cells per
`aliquot) and cultured in duplicate with 105 irradiated autologous
`PBMC in the presence and absence of COP-I (40 ␮g兾ml).
`Cultures were kept for 3 days and pulsed with [3H]thymidine at
`1 ␮Ci per well during the last 16 h of culture. A positive T cell
`line was defined as specific for COP-I when cpm were at 1,500
`and exceeded the reference cpm (in the absence of
`COP-I) at least 3-fold. COP-I reactive T cell lines were expanded
`by restimulation with COP-I under the same conditions de-
`scribed above.
`
`Inhibition Assay. CD4⫹CD25⫺ T cells (responder) and
`CD4⫹CD25⫹ T cells (inhibitor) were cocultured at 5 ⫻ 103 per
`well in the presence or absence of anti-CD3 and anti-CD28
`monoclonal antibodies in U-bottom 96-well plates at a responder
`to inhibitor ratio of 1:1. All cells were cultured in the presence
`of 105 irradiated accessory cells. At day 7, cell proliferation was
`measured as described above. The inhibition was calculated as
`follows: [1 ⫺ (experimental cpm per control cpm)] ⫻ 100%.
`
`Antibody Blocking Experiments. PBMC were plated out in 96-well
`plates at 105 cells per well and stimulated with 40 ␮g兾ml COP-I
`in the presence or absence of the indicated blocking antibodies
`to various cytokines. The final concentrations of the antibodies
`as suggested by the manufacturers were as follows: 1 ␮g兾ml for
`
`Induction of Foxp3 mRNA expression by COP-I as a function of time
`Fig. 1.
`and proliferation rate of PBMC in response to COP-I. (A) PBMC preparations
`derived from 10 randomly selected healthy individuals were cultured, respec-
`tively, in the presence or absence of 40 ␮g兾ml COP-I in complete RPMI medium
`1640. Cells were collected at the indicated induction time over a span of 7 days
`and analyzed for mRNA expression of Foxp3 by real-time PCR. (B) CD4⫹ T cell
`preparations purified from the same PBMC specimens described above were
`cultured in the presence or absence (medium control) of COP-I, respectively,
`for 4 days. Mixed irrelevant peptides (see Materials and Methods) were used
`at the same concentration as the control. Cells were harvested for Foxp3
`expression by real-time PCR. Data are given as relative mRNA expression. *,
`Significant statistical differences between COP-I-treated cells and controls
`(P ⬍ 0.01).
`
`antibodies to IL-8, TNF-␣, IFN-␥, and IL-10, 50 ng兾ml for IL-1␤
`antibody and 10 ␮g兾ml for TGF-␤1 antibody. Cells were har-
`vested on day 4 for Foxp3 expression.
`
`Immunization and Mouse T Cell Preparation. Wild-type and IFN-␥
`gene knockout C57BL兾6 mice were obtained from The Jackson
`Laboratory. Wild-type and IFN-␥ gene knockout mice were
`injected s.c. with 1.6 mg of COP-I per injection at day 0, day 2,
`and day 4 for a total of three injections. Mice were killed at a
`4-day interval for the isolation of CD4⫹CD25⫹ T cells. Spleens
`were gently minced in complete medium containing 10% FBS
`and CD4⫹ T cells were isolated by using a mouse CD4⫹ T cell
`negative selection kit (Miltenyi Biotec). T cell-depleted spleno-
`cytes of wild-type mice were used as antigen-presenting cells as
`indicated. To isolate CD4⫹CD25⫺ T cells, anti-CD25 antibody
`(10 ␮l for 107 T cells) conjugated with microbeads was incubated
`with preselected CD4⫹ T cells before separation, to yield a purity
`of ⬎95% of CD4⫹CD25⫺ T cells. For recovery of CD4⫹CD25⫹
`T cells, the column bound T cells were flushed off with cold
`medium. The purity of CD4⫹CD25⫹ T cells was always ⬎95%.
`
`Supporting Information. For further information, see Supporting
`Materials and Methods and Fig. 8; see also Figs. 9 and 10, which
`are published as supporting information on the PNAS web site.
`
`Results
`Induction of Transcription Factor Foxp3 Expression in Human CD4ⴙ T
`Cells by COP-I. After the initial discovery that COP-I induced the
`expression of transcription factor Foxp3 in human PBMC, the
`time course of induction of Foxp3 expression in PBMC derived
`from healthy individuals was evaluated in response to COP-I
`stimulation by real-time PCR analysis. As shown in Fig. 1A, the
`peak expression of Foxp3 occurred ⬇4 days after exposed to
`COP-I. The effect of COP-I on the induction of Foxp3 expres-
`sion correlated closely with the rate of cell proliferation in
`PBMC induced by COP-I (data not shown). Furthermore, the
`observed effect appeared specific for COP-I as the control
`peptides, although stimulatory to some PBMC preparations, did
`not alter Foxp3 expression in the same PBMC preparations (Fig.
`1 A). Similar results were obtained when purified CD4⫹ T cells
`were treated with COP-I under the same experimental condi-
`tions (Fig. 1B). It was shown that COP-I induced the expression
`
`6450 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0502187102
`
`Hong et al.
`
`Page 2 of 6
`
`YEDA EXHIBIT NO. 2073
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`IMMUNOLOGY
`
`Immunoblot analysis of Foxp3 expression in CD4⫹CD25⫹ T cells after
`Fig. 3.
`COP-I treatment. PBMC were treated with 40 ␮g兾ml COP-I or control peptides
`for 4 days. The resulting CD4⫹CD25⫹ T cells were subsequently purified and
`lysed. Lysate was then subjected to 10% SDS兾PAGE and followed by immu-
`noblot analysis with an anti-human Foxp3 antibody. An antibody to human
`␤-actin was used as a control.
`
`end, human CD4⫹CD25⫺ T cells were purified by magnetic
`bead separation and exposed to COP-I. CD4⫹CD25⫹ T cells
`rose from background value of 7.99 ⫾ 0.28% to 23.36 ⫾ 0.69%
`after CD4⫹CD25⫺ T cell preparations were exposed to COP-I
`compared with 16.08 ⫾ 0.31% in those treated with the control
`(P ⬍ 0.05, data not shown). In vitro treatment of CD4⫹CD25⫺
`T cells with COP-I led to a significantly elevated level of Foxp3
`expression in resulting CD4⫹CD25⫹ T cells (Fig. 2A, P ⬍ 0.05).
`CD4⫹CD25⫹ T cells converted from the CD4⫹CD25⫺ T cell
`fractions by COP-I exhibited considerable inhibitory activities
`on T cell proliferation induced by anti-CD3兾CD28 antibodies
`(Fig. 2B). The increased Foxp3 expression was confirmed in
`parallel by immunoblot analysis (Fig. 3). The results indicate that
`COP-I induced conversion of CD4⫹CD25⫺ to CD4⫹CD25⫹ T
`cells of regulatory function. Furthermore, human T cell lines
`were generated from healthy individuals by repeated stimulation
`with COP-I and characterized for the reactivity, Foxp3 expres-
`sion and the inhibitory rate. As shown in Table 1, T cell lines
`reactive to COP-I had high expression of Foxp3 that was roughly
`10 times higher than that seen in control T cell lines or T cells
`stimulated one time with COP-I and displayed considerable
`inhibitory properties compared to control T cell lines generated
`by irrelevant peptide control. When compared to control T cell
`lines, the cytokine profile of COP-I reactive T cell lines was
`noticeably biased toward high production of IFN-␥ (Table 1).
`
`Induction of Foxp3 Expression and Regulatory Function in CD4ⴙ T Cells
`by COP-I Is Mediated Through IFN-␥. We further evaluated whether
`the effect induced by COP-I was mediated through its ability to
`induce the production of certain cytokines. It was evident that T cell
`activation induced by COP-I resulted in the significantly increased
`production of IFN-␥, TGF-␤1, and TNF-␣but not IL-10 and IL-4
`(Fig. 4A), which was in agreement with a number of the previous
`reports, including our own (5, 28, 29). A panel of six monoclonal
`antibodies to the selected cytokines that were predominantly pro-
`duced in response to COP-I described here and in other reports (30)
`was analyzed for its potential blocking effect on the property of
`COP-I. As illustrated in Fig. 4B, only IFN-␥ antibody significantly
`blocked the effect of COP-I on the induction of Foxp3 (73.3%
`inhibition, P ⬍ 0.05), whereas the observed effect of TGF-␤1
`antibody did not reach a statistically significant level (36.1% inhi-
`
`Conversion of CD4⫹CD25⫺ T cells to CD4⫹CD25⫹ regulatory T cells.
`Fig. 2.
`CD4⫹CD25⫺ T cell preparations were purified from six healthy individuals and
`cultured with COP-I or the peptide control, respectively, in the presence of
`irradiated autologous PBMC as a source of antigen-presenting cells. Cells were
`collected on day 4 and analyzed for Foxp3 expression in refractionated
`CD4⫹CD25⫺ and CD4⫹CD25⫹ T cell subsets by real-time PCR (A) and the
`inhibition on autologous T cell proliferation induced by anti-CD3兾CD28 anti-
`bodies in inhibition assays (B). *, Significant statistical differences between T
`cells treated with COP-I and those treated with the peptide control (P ⬍ 0.05).
`
`of Foxp3 predominantly in CD4⫹ T cell population of both
`CD45RA and CD45RO phenotypes (Fig. 9).
`The potential in vivo effect of COP-I on the expression of
`Foxp3 in T cells was evaluated in MS patients treated with COP-I
`by using untreated MS patients and healthy volunteers as
`controls. To this end, PBMC prepared from MS patients with or
`without standard COP-I treatment and healthy controls were
`analyzed ex vivo for the expression of Foxp3. As illustrated in Fig.
`10, Foxp3 expression was significantly decreased in untreated
`MS patients compared with that of healthy individuals. The
`expression of Foxp3 was significantly higher in PBMC prepara-
`tions derived from nine different MS patients that had been on
`COP-I treatment for 10–12 months, even though the Foxp3
`expression level did not reach that seen in healthy controls.
`Among nine patients examined, three pretreatment specimens
`were available for self-paired analyses that indicated a significant
`high level of Foxp3 expression in posttreatment PBMC.
`
`Conversion of Peripheral CD4ⴙCD25ⴚ T Cells to CD4ⴙCD25ⴙ Regula-
`tory T Cells by COP-I Through the Activation of Foxp3. We then
`examined whether the induction of Foxp3 expression in CD4⫹ T
`cells by COP-I
`represented conversion of peripheral
`CD4⫹CD25⫺ T cells to CD4⫹CD25⫹ regulatory T cells. To this
`
`Table 1. Foxp3 mRNA expression and cytokine productions in COP-I specific T cell lines and control T cell lines
`Cytokine production, pg兾ml
`
`Specificity
`
`SI
`
`Inhibition, %
`
`IFN-␥
`
`TGF-␤1
`
`TNF-␣
`
`IL-4
`
`IL-10
`
`No. of
`T cell lines
`
`Foxp3
`mRNA
`
`15
`10
`
`COP-I
`Peptide control
`
`26.2 ⫾ 20.4
`2.1 ⫾ 3.3
`
`45.5 ⫾ 18.4
`4.9 ⫾ 1.9
`5.2 ⫾ 3.5 ⫺55.5 ⫾ 12.6
`
`1,650.12
`56.45
`
`368.91
`385.91
`
`56.33
`64.65
`
`29.71
`32.17
`
`149.65
`122.34
`
`T cell lines were cultured with irradiated autologous PBMC in the presence of COP-I or the peptide control, respectively. Specific
`reactivity of the T cell lines to COP-I was determined as stimulation index (experimental cpm兾background cpm). SI, Stimulation index.
`
`Hong et al.
`
`PNAS 兩 May 3, 2005 兩 vol. 102 兩 no. 18 兩 6451
`
`Page 3 of 6
`
`YEDA EXHIBIT NO. 2073
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`The effect of IFN-␥on the expression of Foxp3 in human PBMC. PBMC
`Fig. 5.
`preparations were obtained from 10 healthy individuals and cultured in the
`presence or absence of the indicated recombinant cytokines at a concentra-
`tion of 25 ng兾ml for 72 h. (A) The resulting cells were analyzed for the
`expression of Foxp3 by real-time PCR. The data are presented as relative
`expression of Foxp3. CD4⫹CD25⫺ T cells were purified from the same PBMC
`cultured in the presence of recombinant IFN-␥at the indicated concentrations
`for 72 h. (B) mRNA expression of Foxp3 in the resulting T cells was measured
`under the same experimental conditions. CD4⫹CD25⫺ T cells were cultured in
`the presence of human recombinant IFN-␥ (25 ng兾ml) and a monoclonal
`antibody to human IFN-␥(10 ␮g兾ml). An isotype-matched antibody was used
`at the same concentration as a control. (C) The resulting T cells were analyzed
`for mRNA expression of Foxp3 by real-time PCR. The results were reproducible
`in at least three independent experiments. *, Statistical difference between
`IFN-␥ and medium control (P ⬍ 0.01); **, statistical difference between
`anti-IFN-␥ and the controls (P ⬍ 0.05).
`
`indicating a trait of CD4⫹CD25⫹ regulatory T cells (Fig. 7B) and
`significant inhibitory activities at day 16 (Fig. 7C). Taken
`together, the results indicate that the activation of Foxp3 ex-
`pression and the conversion of CD4⫹CD25⫺ to CD4⫹CD25⫹
`regulatory T cells by COP-I required IFN-␥ as a mediator.
`
`Discussion
`In this study, we demonstrated that COP-I, an immunoregulatory
`agent of unknown mechanism, has a unique property in the
`conversion of CD4⫹CD25⫺ T cells to CD4⫹CD25⫹ regulatory T
`
`Induction of Foxp3 expression in T cells by in vitro treatment with
`Fig. 6.
`COP-I in wild-type and IFN-␥-deficient mice. Splenocytes were derived from
`wild-type mice (A) and IFN-␥ knockout mice (B) of the B6兾C57 background.
`CD4⫹CD25⫺ T cells were subsequently isolated by magnetic bead separation
`and cultured in the presence or absence of 40 ␮g兾ml COP-I and irradiated
`splenocytes predepleted for T cells as a source of antigen-presenting cells. The
`resulting T cells were collected for Foxp3 mRNA expression on day 4. *,
`Statistical significance between the COP-I untreated group and the COP-I-
`induced group (P ⬍ 0.01).
`
`Cytokine profile of PBMC in response to COP-I stimulation and
`Fig. 4.
`antibody blocking experiments. (A) PBMC preparations derived from healthy
`individuals were cultured in the presence of COP-I or the peptide control at a
`concentration of 40 ␮g兾ml. Culture supernatants were collected on day 4 and
`analyzed for the production of the indicated cytokines by ELISA. Data are
`presented as the mean cytokine concentrations of six individual samples. (B)
`In parallel experiments, the same PBMC preparations were cultured with 40
`␮g兾ml COP-I in the presence of the indicated purified monoclonal antibodies
`used at concentrations of 50 ng兾ml to 10 ␮g兾ml as instructed, respectively.
`Cells were collected on day 4 and analyzed for Foxp3 expression by real-time
`PCR. *, Statistical differences (P ⬍ 0.05).
`
`bition, P ⫽ 0.095). The results indicated that the effect of COP-I was
`mediated by IFN-␥and, to a lesser degree, by TGF-␤1. TGF-␤1 in
`the presence of anti-CD3 antibody was recently described to have
`an effect on Foxp3 expression in T cells (31), whereas IFN-␥had not
`been known for a similar ability to induce Foxp3 expression. The
`role of IFN-␥ in the induction of Foxp3 expression by COP-I was
`confirmed in further characterization with recombinant IFN-␥. As
`shown in Fig. 5, IFN-␥but not other cytokines (including TGF-␤1)
`used at a similar concentration in the absence of a T cell stimulus
`induced the expression of Foxp3 in CD4⫹CD25⫺ T cells.
`The role of IFN-␥ in the induction of Foxp3 expression and
`regulatory function in T cells was further investigated in IFN-␥
`knockout mice. As shown in Fig. 6,
`in vitro treatment of
`CD4⫹CD25⫺ T cells with COP-I resulted in increased expression
`of Foxp3 in wild-type mice but not IFN-␥ knockout mice of the
`same C57兾BL6 background. It was evident that although ad-
`ministration of COP-I induced an increase in CD4⫹CD25⫹ T
`cells in wild-type mice and IFN-␥knockout mice, Foxp3 expres-
`sion was found in CD4⫹CD25⫹ T cells obtained from wild-type
`mice but not those of IFN-␥ gene-knockout mice (Fig. 7A).
`CD4⫹CD25⫹ T cells derived from wild-type mice but not IFN-␥
`knockout mice exhibited reduced proliferation in response to
`stimulation induced by anti-CD3兾CD28 monoclonal antibodies,
`
`6452 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0502187102
`
`Hong et al.
`
`Page 4 of 6
`
`YEDA EXHIBIT NO. 2073
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`IMMUNOLOGY
`
`Induction of CD4⫹CD25⫹ regulatory T cells in wild-type and IFN-␥knockout mice in response to administration of COP-I. Wild-type and IFN-␥knockout
`Fig. 7.
`mice were administered COP-I at 5 mg per mouse, which represented an effective dosage demonstrated in previous studies (39). Mice were killed on the indicated
`days. Splenocytes were obtained and analyzed ex vivo for Foxp3 expression in purified CD4⫹CD25⫺ and CD4⫹CD25⫹ T cell subsets by real-time PCR (A), the
`proliferative response to anti-CD3兾CD28 antibodies (B), and the inhibition on syngeneic T cell proliferation induced by anti-CD3兾CD28 antibodies in inhibition
`assays (C). *, Significant statistical differences between posttreated and pretreated CD4⫹CD25⫹ T cells (P ⬍ 0.05).
`
`cells. There is compelling evidence presented in this study, indi-
`cating that the induction of CD4⫹CD25⫹ regulatory T cells by
`COP-I is mediated through the activation of transcription factor
`Foxp3 in CD4⫹ T cells. The conclusions are based on the following
`findings. First, we demonstrated direct evidence that COP-I in-
`duced in vitro conversion of unprimed human CD4⫹CD25⫺ T cells
`to CD4⫹CD25⫹ T cells that acquired high expression of Foxp3 and
`the inhibitory functions. These CD4⫹CD25⫹ regulatory T cells
`appeared to expand in response to COP-I from the CD4⫹ T cell
`pool of CD45RA and CD45RO phenotypes. Secondly, COP-I T
`cell lines generated by repeated stimulation cycles exhibited very
`high expression levels of Foxp3 and the inhibition rate when
`compared with single stimulation of CD4⫹CD25⫺ T cells with
`COP-I and control T cell lines. This finding has not only strength-
`ened the key conclusion mentioned above but also indicated that a
`large proportion of COP-I reactive T cell
`lines represent
`CD4⫹CD25⫹ regulatory T cells. Our experiments have confirmed
`correlation between the expression of Foxp3 and the inhibitory
`function in CD4⫹ T cells. Furthermore, standard treatment with
`COP-I in MS patients and administration of COP-I in mice resulted
`in significant increase of Foxp3 expression in T cells, providing the
`compelling in vivo evidence for the role of COP-I in the induction
`of Foxp3. It is conceivable, however, that not all CD4⫹CD25⫹ T
`cells induced by COP-I represent regulatory T cells. As discussed
`below, COP-I activates CD4⫹ effector secreting IFN-␥that renders
`conversion of a proportion of CD4⫹CD25⫹ T cells into regulatory
`T cells.
`The findings described here are highly significant in the
`understanding of the mechanism of action of COP-I in relation
`to its treatment efficacy in MS. In this regard, COP-I may act,
`through its ability to induce CD4⫹CD25⫹ regulatory T cell
`response, to compensate a functional deficit in this important
`regulatory mechanism in MS (23). It is conceivable that this
`unique property of COP-I in the induction of CD4⫹CD25⫹
`regulatory T cells may attribute, at least in part, to the treatment
`efficacy of COP-I in MS. The observation is also likely to offer
`a reasonable explanation for, or it may reconcile with, some
`previously described regulatory functions of COP-I of unknown
`mechanism, including so-called ‘‘bystander’’ inhibitory effect of
`COP-I on T cell activation, which is a frequently reported
`phenomenon (8). The antigen nonspecific inhibitory effect of
`CD4⫹CD25⫹ regulatory T cell response induced by COP-I is
`consistent with the spectrum of inhibition induced by COP-I that
`is often not limited to one antigen (i.e., MBP) and includes
`various other myelin antigens (16, 32). Furthermore,
`if the
`induction of CD4⫹CD25⫹ regulatory T cells is closely associated
`with the treatment effect of COP-I in MS, one potentially
`important aspect of the clinical significance of the study may
`
`involve the role of Foxp3 as a surrogate biomarker in measure-
`ment of treatment efficacy. Currently, the treatment efficacy of
`COP-I can only be measured ⬇9 months after the treatment by
`using the standard clinical and magnetic resonance imaging
`techniques (33). However, it should be cautioned that this study
`does not completely exclude the role of other regulatory prop-
`erties of COP-I in the treatment of MS, especially its ability to
`induce Th2 immunity through repeated administration (5). It is
`conceivable that these regulatory mechanisms induced by COP-I
`may work in concert to achieve sufficient immune regulation
`ultimately beneficial to the clinical course of MS.
`Another important aspect of the study is related to defining
`the mechanism of action potentially responsible for the induction
`of Foxp3 expression by COP-I in CD4⫹ T cells, which leads to
`the conversion of CD4⫹CD25⫺ T cells to CD4⫹CD25⫹ regula-
`tory T cells. It was evident in the study that the observed effect
`of COP-I is largely mediated through IFN-␥. It is known from
`this and other reports that COP-I has the ability to stimulate the
`production of Th1 and Th2 cytokines, including IFN-␥ (13, 28).
`The role of IFN-␥ in the induction of Foxp3 expression and
`CD4⫹CD25⫹ regulatory T cell response is supported by the
`following experimental evidence: (i) Blocking of IFN-␥(but not
`other cytokines also induced by COP-I) by specific antibody
`resulted in significant inhibition of the effect of COP-I. (ii)
`Treatment of CD4⫹ T cells with recombinant IFN-␥ led to
`increased Foxp3 expression. (iii) COP-I failed to induce Foxp3
`expression in T cells of IFN-␥ knockout mice in in vitro and in
`vivo settings. The role of IFN-␥ appears different from that of
`TGF-␤1 in the induction of Foxp3 and conversion of
`CD4⫹CD25⫺ T cells to CD4⫹CD25⫹ regulatory T cells. TGF-␤1
`requires costimulation of T cells with an anti-CD3 antibody to
`induce Foxp3 expression (34). As described here and by other
`investigators (31, 34), in contrast to IFN-␥, TGF-␤1 is insuffi-
`cient when used alone to directly induce Foxp3 expression in T
`cells. Our observation that the induction of Foxp3 expression by
`COP-I is partially blocked by antibody to TGF-␤1 supports the
`possibility that the involvement of TGF-␤1 in the induction of
`Foxp3 expression requires a stimulatory signal provided by
`COP-I.
`The finding that IFN-␥ has the effect of mediating the
`induction of Foxp3 expression and CD4⫹CD25⫹ regulatory T
`cells is consistent with a recent report indicating that signal
`transducer and activator of transcription-1 (STAT1), a signaling
`molecule closely associated with the IFN-␥signaling pathway, is
`critical to the induction of CD4⫹CD25⫹ regulatory T cells (24).
`The authors demonstrated that STAT1-deficient mice express-
`ing a transgenic T cell receptor against MBP spontaneously
`developed experimental autoimmune encephalomyelitis, which
`
`Hong et al.
`
`PNAS 兩 May 3, 2005 兩 vol. 102 兩 no. 18 兩 6453
`
`Page 5 of 6
`
`YEDA EXHIBIT NO. 2073
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`was attributable to a functional impairment of CD4⫹CD25⫹
`regulatory T cells in STAT1-deficient mice (24). Furthermore,
`the study described here has also raised new questions regarding
`the role of IFN-␥ in T cell regulation. This is another example
`adding to the recent debate on the functional role of Th1 and Th2
`cytokines in autoimmune conditions, such as MS (35–38). The
`traditionally held Th1 paradigm is being challenged by mounting
`evidence that not all Th1 cytokines (e.g., IFN-␥and TNF-␣) are
`necessarily the culprits for MS; in some aspects, they may be
`beneficial. In conclusion, it is clear from the present study that
`COP-I acts as an inducer for CD4⫹CD25⫹ regulatory T cell
`
`response through the activation of Foxp3 expression. This agent
`has a potent and selective property for the induction of Foxp3
`expression and CD4⫹CD25⫹ regulatory T cells in human and an-
`imal experimental systems, making it an excellent tool for the
`study of CD4⫹CD25⫹ regulatory T cells in future investigations.
`
`This work was supported by Chinese Ministry of Science and Technology
`Grant 863, Projects 04DZ1920, 2002AA216121 and 202CCCD2000;
`Shanghai Commission of Science and Technology Grants 01JC14036,
`20014319207, 04DZ14902, and 03XD14015; National Natural Science
`Foundation of China Grant 30430650; and National Institutes of Health
`Grants NS41289 and NS48860 (to J.Z.Z.).
`
`1. Bornstein, M. B., Miller, A., Slagle, S., Weitzman, M., Crystal, H., Drexler, E.,
`Keilson, M., Merriam, A., Wassertheil-Smoll