Oral administration of human or murine interferon alpha suppresses relapses and modifies adoptive transfer in experimental autoimmune encephalomyelitis

Journal of Neuroimmunology

ELSEVIER

Journal of Neuroimmunology 58 (1995) 61-69

Oral administration of human or murine interferon alpha suppresses relapses and modifies adoptive transfer in experimental autoimmune encephalomyelitis Staley A. Brod a,., Mohammed Khan a, Ronald H. Kerman u, Miguel Pappolla c a Department of Neurology 7.044, University of Texas Health Science Center at Houston, PO Box 20708, Houston, TX 77225, USA b Division of Immunology and Organ Transplantation, University of Texas Health Science Center at Houston, PO Box 20708, Houston, TX 77225, USA c Department of Surgery and the Department of Pathology and Laboratory Medicine, University of Texas Health Science Center at Houston, PO Box 20708, Houston, TX 77225, USA

Received 24 October 1994; revised 30 November 1994; accepted 30 November 1994

Abstract

Chronic relapsing experimental autoimmune encephalitis (CR-EAE) is an inflammatory process of the central nervous system (CNS) that closely resembles the human disease multiple sclerosis (MS). EAE was induced in SJL/J mice and following recovery from the initial attack, animals were fed varying doses of human or murine interferon alpha (IFN-a), or mock IFN three times per week. After relapse, concanavalin A-activated spleen cells were transferred adoptively from orally fed animals into recipient animals. Oral administration of human or murine IFN-a suppressed relapse in actively immunized animals, modified adoptive transfer of EAE, and decreased mitogen/antigen proliferation and IFN-3, secretion in both donors and recipients. IFN-a acts orally by modifying the encephalitogenicity of donor spleen T cells.

Keywords: Interferon-a; Experimental autoimmune encephalomyelitis; Oral feeding; Adoptive transfer; Interferon-y

I. Introduction

Chronic relapsing experimental autoimmune encephalitis (CR-EAE), a T cell-mediated disease, provides an animal model to assess interventions that may modify the course of a human autoimmune disease (Raine and Stone, 1977; Wisniewski and Keith, 1977; Feuer et al., 1985). Previous work has demonstrated that immunomodulatory cytokines can modify EAE. Parenteral (IV) natural rat interferon (105 units) can suppress partially acute E A E in male Lewis rats and inhibit passive hyperacute localized E A E when administered on the same day of immunogen inoculation (Abreu, 1982; Abreu et al., 1983). Parenterally administered natural human IFN-a can decrease T cell function and T cell-dependent antibody production in humans (Balkwill, 1985).

* Corresponding author. Phone (713) 792 5777 ext. 44; Fax (713) 745 0768 0165-5728/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0 1 6 5 - 5 7 2 8 ( 9 4 ) 0 0 1 8 8 - X

We have recently demonstrated that orally administered type 1 interferons decrease clinical score, inflammatory loci and the secretion of interferon (IFN)-y in 5000 unit (U) IFN-fed compared to mock-fed and 5000 U subcutaneously (s.c.) treated rats with E A E (Brod et al., 1995). 100 U of murine species-specific IFN-a//3 administered orally three times a week suppresses clinical relapse in murine C R - E A E (Brod and Burns, 1994). E A E can be adoptively transferred by using in vitro antigen or concanavalin A (ConA)-activated spleen cells from actively immunized mice and provides a model to examine the effect of orally administered cytokines on T cells (Peters and Hinrichs, 1982; Lublin, 1985; Whitham et al., 1991). Available natural murine type 1 IFN are manufactured by viral induction in murine fibroblasts and could contain non-IFN immunoactive proteins other than IFN that could be responsible for disease modification. Human IFNs demonstrate a wide range of antiviral activity (1-50%) in murine systems (Weber et al., 1987; McInnes et al., 1989) and human recombinant IFNs do not contain

S.A. Brod et al. /Journal of Neuroimmunology 58 (1995) 61-69

62

extraneous immunoactive proteins. Therefore, suppression of actively induced relapses and modifying of adoptive transfer with pure human preparations would eliminate the possibility that non-IFN murine proteins modify disease, but might require higher doses relative to murine IFN to show effects. Accordingly we examined whether the oral administration of human recombinant and murine species specific IFN-a would modify adoptive transfer of EAE.

2. Materials and methods

2.1. Induction cephalomyelitis

of experimental

autoimmune

en-

A chronic relapsing form of EAE was induced in 7-10-week-old female S J L / J mice using the method of Brown and McFarlin as modified by Miller (Brown and McFarlin, 1981; Miller et al., 1987), Briefly, each mouse received an s.c. injection over the shaved right flank of 0.3 ml of an emulsion containing 1 mg of syngeneic mouse spinal cord homogenate (MSCH) in 0.15 ml of phosphate-buffered saline and 0.03 mg of Mycobacterium tuberculosus hominis H37Ra (MT) (Difco Labs, Detroit, M1) in 0.15 ml of incomplete Freund's adjuvant (IFA) (Difco Labs, Detroit, MI). 7 days later, the mice received a similar injection in the left flank. Initial clinical signs of disease were seen between days 13 and 25 post immunization. Clinical severity of the initial attack was graded as follows: 0, no disease; 1, minimal or mild hind limb weakness (associated with limp tail); 2, moderate hind limb weakness or mild ataxia (waddling gait a n d / o r poor righting ability); 3, moderate to severe hind limb weakness; 4, severe hind limb weakness or moderate ataxia; 5, paraplegia with no more than moderate four limb weakness; 6, paraplegia with severe four limb weakness or severe ataxia. Animals were scored in a blinded fashion for 12 days to 5 weeks, and a daily or weekly score was computed for each group of animals. Animals were maintained in accordance with the guidelines of the Committee on Care of Laboratory Animals of the Laboratory Research Council (DHEW: NIH, No. 85-23, revised 1985).

2.2. Oral administration of IFN Following the initial attack (day 30 post active immunization), animals were fed varying doses (10-1000 U) of murine natural IFN-a (mlFN) (Cytimmune mouse IFN-a, 4.0 × 105 IRU/ml, Lee Biomolecular Research, Inc., San Diego, CA), recombinant human IFNt~ (hrlFN) (100-1000) (Schering IFN-oqi b, 3 × 10 6 IU/ml, Schering Pharmaceuticals, Kenilworth, N J), or mock murine IFN-a (Cytimmune, < 2 IRU/ml, Lee

Biomolecular Research, Inc., San Diego, CA) using a syringe fitted with a 20-gauge ball point needle (Thomas Scientific, Swedesboro, N J) three times per week (Monday, Wednesday, Friday) for 5 weeks. We had previously established that 1 and 10 U of orally administered hrlFN did not have any significant effect on clinical relapse. PBS or mock IFN was used as control for hrlFN-a and mlFN-a, respectively, since denatured IFN may retain some immunological activity. Mock murine IFN-te control is a preparation identical to the IFN preparation except that the fibroblast cultures are not induced with Newcastle disease virus (according to Lee Biomolecular Research Inc., Virology and Immunology Catalog, 1989). Recombinant human IFNt~nb was used in these experiments since (i) hrlFN-a does not contain potential Newcastle virus-induced immunoactive molecules arising during manufacture of fibroblast murine natural IFN-a; (ii) there is a high degree of homology between human and murine IFN-a gene products (Kawade, 1987); and (iii) type 1 IFNs can induce viral resistance in cultured heterologous mouse cells thus demonstrating trans-species activity (Blalock and Barons, 1977).

2.3. Adoptive transfer Five weeks after clinical relapse attack, all spleens were aseptically removed and pooled in groups from each set of donors, and single-cell suspensions were prepared. Suspensions were washed three times in PBS and then resuspended at 2 × 106 cells/ml in standard media. Cells were cultured with ConA (2.5 /zg/ml) (Sigma Chem Co., St. Louis, M O ) + 1, 10, or 100 U hrlFN-a in vitro in 75-cm 2 tissue culture flasks for 72 h in a humidified 5% COa/95% air incubator at 37° C. Following incubation, cells were collected, washed twice in PBS, and viability determined by standard Trypan blue exclusion. Viable concentrations were adjusted to 107 cells/0.5 ml Dulbecco's PBS immediately prior to i.p. injection into naive recipient mice.

2.4. Lymph node cell and spleen preparation Animals were sacrificed by CO 2 narcosis after relapse (4 weeks after immunization and 5 weeks after initiation of feeding) or post i.p. transfer attack (12-23 days post transfer after peak disease was attained) and pooled draining popliteal lymph node and pooled spleen cells from all animals in each group were removed and single-cell suspensions were made through 90-/zm stainless wire meshes. Red cell lysis was performed in the spleen cell suspensions with 2 ml of red cell lysis solution (ACK) added to the pellet and the reaction allowed to continue at 5 min at room temperature.

S.A. Brod et al. /Journal of Neuroimmunology 58 (1995) 61-69 2.5. In uitro T cell proliferation Following clinical attack (4 weeks after active immunization and 5 weeks after initiation of feeding during relapse or 12-23 days post i.p. transfer in acute attack) mice were sacrificed, draining inguinal lymph nodes were pooled and cultured in vitro to determine antigen-specific T cell proliferative responses. Antigen stimulation was carried out with antigen at either 0 or 1 0 / ~ g / m l (guinea pig myelin basic protein (GP-MBP), proteolipid protein peptide 139-151 (PLP 139-151 H S L G K W L G H P D K F synthesized by the Howard Hughes Institute Biopolymers Facility, U T Southwestern Medical Center, Dallas, TX) or MT) and mitogen stimulation with C o n A at 2.5 / z g / m l by incubating lymph node cells at 2 × 105 cells/well in R P M I (Gibco, G r a n d Island, NY) supplemented with 10% fetal calf serum (FCS) (Whittaker Bioproducts, Walkersville, MD), 1% sodium pyruvate (Gibco, G r a n d Island, NY), 1% glutamine (Gibco), 1% penicillin/streptomycin, and 50 /zM 2-mercaptoethanol (standard medium). GPMBP was p r e p a r e d according to methods in (Brod and Burns, 1994). The plates were incubated at 5% CO 2 and humidified at 37°C for 4 days. At that time the cells were pulsed with 2 /zCi of tritiated [3H]dTd and harvested 18 h later on an automated harvester. [3H]dTd uptake was measured in a Beckman (liquid) scintillation counter. Cultures were run in triplicate and the results expressed as Acpm.

2.6. Histology

1/20. Background fluorescent reactivity was determined using either isotype control phycoerythrin or FITC-conjugated m A b as control. Flow cytometric analysis was performed (FACscan, Becton-Dickinson Cell Analysis, Elmhurst, IL).

2.8. Measurement of cytokine secretion Spleen and lymph node cells generated as described above were stimulated with ConA at 2.5 ~ g / m l or PLP 139-151 at 1 0 / x g / m l in standard media. Supernatants were collected at 48 h after C o n A or PLP 139-151 activation of mice T cells and frozen at - 7 0 ° C after centrifugation. Interleukins were measured using solid phase E L I S A assay. Anti-IL-2, anti-IL-4 and anti-IFN-7 (PharMingen, San Diego, CA) were incubated on polyvinyl plastic 96-well microtiter plates with 0.01 M carbonate buffer (pH 9.6) overnight at 4° C. The plates

Tj

20

/

1.5

0

S7-" >

Following sacrifice, randomly selected spinal cords (mock, n = 5; 10 U n = 5; 100 U n = 5) were removed and immersion-fixed in 2% neutral-buffered formalin for a minimum of 2 weeks. After fixation, cords were sectioned in entirety in the longitudinal plane at approximately 3 m m intervals and processed to paraffin. Paraffin blocks were sectioned at 6 - 8 /xm, and were stained with Hematoxylin and Eosin and examined by light microscopy, Cord sections were evaluated independently for foci of inflammation by a blinded observer, without knowledge of the treatment status of the animals prior to sacrifice. Spinal cord tissue was sampled in an identical fashion for each animal and numbers of inflammatory foci per section ( > 20 inflammatory cells) in the parenchyma were counted.

2. 7. Phenotyping Cytofluographic analysis of murine T cells was performed by means of direct immunofluorescence with fluorescein-conjugated anti-CD3 m A b (Caltag Labs, South San Francisco, CA), anti-L3T4 mAb, and antiLy-2 m A b phycoerythrin-conjugated m A b (Collaborative Research Products, Bedford, MA) at a dilution of

63

5 u c u

o c. CO

1.#

4/

.

s.I

\

--D--

PBS

. . . . . . . .

1OO u n i t s h r I F N

--o--

1OOOunits

hrlFN

o5 ooL 0

Week

relopse

Fig. 1. Experiments l + 2. Oral administration of human recombinant IFN-a in murine CR-EAE suppresses relapse attacks. Three groups of six SJL/J 6-8-week-old female mice were immunized as described in Materials and methods. On day 40 post-immunization, one group was fed PBS, another group 100 units hrIFN-a and a third group was fed 1000 units hrIFN-a three times per week over the following 5 weeks. All animals were scored by a blinded observer for clinical disease until day 75 after immunization. Values represent combined data of two separate experiments of mean weekly group clinical scores+S.E.M. (mock PBS fed vs. 100 U for weeks 2-5, P < 0.001; vs. 1000 U hrIFN-a fed animals for weeks 4-5, P < 0.01 by nonpaired t-test).

64

S.A. Brod et al. /Journal of Neuroimmunology 58 (1995) 61-69

were blocked with 3% BSA in phosphate-buffered saline for 3 h. 100 /zl of supernatants were added at various dilutions that were titered to the linear portion of the absorbance/concentration curve in triplicate and incubated for 1 h at room temperature. The plates were washed five times with Tween (0.05%) in phosphate-buffered saline, and 100 /zl peroxidase-conjugated interleukin monoclonal antibody (with a different epitopic determinant than the first antibody used to coat the polyvinyl plate) at a 1:1000 concentration was added for 60 min. Following an additional wash step, the peroxidase substrate o-phenylenediamine dihydrochloride was added, and the optical density measured at 450 nm; OD reading in ng/ml was derived from standard curves of IL-2, IL-4 and IFN-y.

2.9. Statistical analysis Statistical analysis was performed using a two-tailed non-paired Student's t-test.

3. Results

3.1. Orally administered human recombinant IFN-a can suppress clinical disease and inhibit proliferative responses to MBP (experiments 1 and 2) Three groups of six S J L / J 6-8-week-old female mice that were immunized with MSCH subsequently had an attack beginning by day 16 and ending by day 30. Each group of animals had comparable scores after the initial clinical attack had subsided. On day 30 post

immunization, groups were either mock-fed PBS, 100 U human recombinant IFN-a (hrIFN-a) or 1000 U hrIFN-a three times per week over the following 5 weeks. Mean weekly clinical scores demonstrated significant differences in outcome in the mock PBS-fed versus 100 U and 1000 U hrIFN-a-fed animals (Fig. 1). The mock PBS-fed group incurred increasing disease severity over the course of 5 weeks as shown by increasing neurological deficit (group clinical score) over time. The 1000 U-fed hrIFN-a underwent a mild single attack with decreasing neurological deficit. Overall, oral hrIFN-a decreased the group scores during relapse. Thus, orally administered hrIFN-a is active by the oral route and suppresses clinical relapses. Following clinical relapse, mice were sacrificed, spleen and draining inguinal lymph node cells were pooled into hrIFN-a-fed (n = 6) and mock PBS-fed (n = 6) groups, respectively, and cultured in vitro to determine antigen-specific T cell proliferation. There was a significant decrease in proliferation in draining inguinal lymph nodes to GP-MBP in mice fed 1000 U hrIFN-a compared to mock PBS-fed control (P < 0.05), and in spleen cells in mice fed 100 U hrIFN-a to GP-MBP and MT compared to mock PBS-fed animals (GP-MBP, P < 0.05; MT, P < 0.05) (Table 1).

3.2. Activated donor cells from animals administered human recombinant IFN-a orally are less effective in transferring clinical disease compared to cells from mock PBS-fed mice (experiments 3 and 4) Adoptive transfer experiments were performed to determine if activated spleen cells from hrlFN-a-fed

Table 1 Oral IFN-a inhibits proliferation to GP-MBP, PLP 139-151, M T and C o n A Experiments 1 + 2

G P - M B P LN

GP-MBP sp

M T sp

Mock PBS 100 U hrIFN 1000 U h r I F N

6 105 _+ 705 3 330 _+ 595 1 182 _+ 121 *

11357 + 948 5 841 ± 1846 * ND

14086 _+ 1 385 8 396 + 948 * ND

Experiments 3 + 4

C o n A sp

PLP 139-151 LN

Mock PBS 100 U h r I F N

40822 ± 1 803 2406 + 436 *

46187 5:2.836 25966 ± 929 *

Experiments 5 + 6

C o n A LN

PLP 139-151 sp

Mock m I F N 10 U m I F N

94 770 _+ 2 783 31943 ± 7840 *

5 360 ± 262 2166 _+ 93 *

Experiments 7 + 8

C o n A LN

Mock m I F N 10 U m I F N

126 625 ± 10 743 85 900 + 7153 *

Following clinical acute (experiments 3, 4 and 7, 8) or relapse (experiments 1, 2 and 5, 6) attack, mice were sacrificed, spleen (sp) and draining inguinal lymph nodes (LN) were pooled and cultured in vitro to determine antigen-specific T cell proliferative responses. All antigen stimulation was carried out by incubating whole spleen/popliteal draining lymph node populations at 2 × 105 cells/well with antigen at 1 0 / x g / m l (GP-MBP, PLP 139-151 or MT) in standard media and cultured as described in Materials and methods. Cultures were run in triplicate and results are expressed as cpm minus background with cells alone ___S.E.M. ND, not done. * Significantly different from mock control ( P < 0.01). Values represent combined data of two separate experiments.

S.A. Brod et al. / Journal of Neuroimmunology 58 (1995) 61-69

65

Table 2 Oral IFN-a inhibits IFN- 7 and IL-2 secretion

3, O u

/ l.I

i

--D--

"

:3

Mock

.... '"" 1OO u n i t s

T

hrtFN

i

_

i if i I..... Days

post

ip.

Experiments 3 + 4

PLP 139-151: IFN-3,

Mock PBS 100 U h r l F N

30 + 2 12_+2 *

Experiments 5 + 6

C o n A spleen: IFN-7

ConA: IL-2

Mock m l F N 10 U m l F N

42_+ 1 14_+ 1 *

112_+3 61 _+3 *

Experiments 7 + 8

ConA LN: IFN-7

Mock m I F N 10UmlFN

26 _+ 1 6_+1 *

Following clinical acute (experiments 3, 4 and 7, 8) or relapse (experiments 5, 6) attack, mice were sacrificed, spleen cells were pooled and cultured in vitro for cytokine production. Antigen stimulation was carried out by incubating whole spleen populations at 2 x 1 0 5 cells/well with antigen at 100 p . g / m l (PLP 139-151) in standard media or C o n A at 2.5 / z g / m l and cultured for 48 h. Supernatants were collected at 48 h after C o n A activation and frozen at - 7 0 ° C after centrifugation. Interleukins were measured using solid phase ELISA assay. * Significantly different from mock control ( P < 0.01). Values represent combined data of two separate experiments.

transfer

Fig. 2. Experiments 3 + 4. H u m a n recombinant IFN-a fed animals do not adoptively transfer disease. 1 0 x 106 3-day ConA-activated T cells from mock PBS fed or 100 U h r l F N - a fed immunized donor S J L / J mice, which had u n d e r g o n e relapse attacks 5 weeks earlier, were transferred adoptively i.p. to naive S J L / J mice (n = 6, mock donor) (n = 6, 100 U h r l F N - a donor) and followed for evidence of disease. Values represent combined data of two separate experiments of m e a n group daily blinded clinical scores + S.E.M. (mock PBS fed donors vs. h r l F N - a fed donors for days 5-12, P < 0.01 by non-paired t-test).

20 I

~ - -

8u~

Mock

.... 0 - ' " 1 O u n i t s --o--1OOunits

animals could transfer disease. ConA-activated T cells from mock PBS-fed or 100 U hrlFN-a-fed immunized S J L / J mice, followed for 5 weeks after initiation of feeding during relapse (see Fig. 1), were transferred adoptively i.p. to recipient mice that were then followed for evidence of disease. Mice that received ConA-activated mock PBS T cells had a significant clinical attack starting at day 5, while recipients of ConA-activated hrlFN-a T cells had a much less severe clinical attack (Fig. 2). In contrast, ConA-activated spleen cells from mock PBS animals treated with hrlFN-a in vitro as opposed to in vivo did not prevent adoptive disease transfer (data not shown). ConA-induced spleen cell proliferation from mock PBS-fed recipients was significantly greater compared to hrlFNa-fed recipients (Table 1). Since PLP is the major encephalitogen in the S J L / J mouse (Sobel et al., 1990), we demonstrated a significant decrease in proliferation in draining inguinal lymph nodes to PLP 139-151 in mice fed 100 U hrlFN-a compared to PBS-fed control (Table 1). Adoptive transfer of donor cells from IFN-a-

u E

mLFN mLFN

10

u o c_ ~9

:..-.-I! %O.OL Week

relapse

Fig. 3. Experiments 5 + 6 . Oral murine species-specific IFN-a suppresses relapse at 1 / 1 0 t h the effective cross-species h r l F N dose. Animals (n = 8 / g r o u p ) were immunized, followed as described in Fig. 1 and treated with mock mlFN, 10, or 100 U mlFN-~. Animals were scored by a blinded observer for clinical disease until 5 weeks after feeding. Values represent combined data of two separate experiments of m e a n weekly group clinical s c o r e s + S . E . M . (mock IFN fed vs. 10 U m l F N - a fed animals for weeks 1-5, P < 0.001; vs. 100 U m l F N - a for weeks 1-5, P < 0.001 by non-paired t-test).

S.A. Brad et al. /Journal of Neuroimmunology 58 (1995) 61-69

66

fed animals decreased PLP-induced IFN-7 secretion in recipients. PLP 139-151 stimulated draining inguinal lymph node cells from hrIFN-a-fed animals demonstrated decreased secretion of IFN-7 (Table 2). There were no changes in IL-2 or IL-4 secretion between mock PBS and hrIFN-a-fed treated animals (data not shown).

3.3. Orally administered murine species-specific IFN-a can suppress clinical disease, decrease inflammation and cytokine secretion (experiments 5 and 6) We have shown previously that 100 U of orally administered murine species-specific IFN-a//3 can suppress relapse attacks (Brad and Burns, 1994). Others have found that human IFN-ce I and IFN-a 2 have 1-50% of antiviral activity in L929 mouse cells compared to human WISH or HEp2 cell lines (Streuli et al., 1981; Weber et al., 1987; McInnes et al., 1989). Therefore we determined if 1 / 1 0 t h the dose (10 U) of per as administered murine species-specific IFN-o~ (mIFN-a) would suppress relapse attacks and prevent adoptive transfer. Animals were immunized and after 30 days at the completion of the initial attack fed with

mock murine IFN, 10 or 100 U mIFN-a. The mock IFN-fed group incurred relapse over the course of 5 weeks. Clinical scores demonstrated significant differences in outcome in the mock mIFN-fed versus 10 and 100 U mIFN-a-fed animals (Fig. 3). Histological examination showed decreased inflammation in animals with decreased clinical scores (mock IFN, 2.2 _+ 0.1; 10 U mIFN-a,: 0.9 + 0.4, P < 0.01; 100 U mIFN-a, 0.8 _+ 0.6, P < 0.05). Orally administered mIFN-a is effective at an order of magnitude less compared to the effective hrIFN dose, consistent with data for cross-species anti-viral activity. Phenotyping at the time of sacrifice demonstrated no significant differences in CD3, CD4 or CD8 cell surface expression in pooled lymph node or pooled spleen cells from mock IFN compared to mIFN-a-fed animals in two separate experiments (data not shown). ConA activation of draining popliteal lymph node cells was inhibited in mIFN-a-fed compared to mock IFN-fed animals (Table 1). There was a significant decrease in proliferation in spleen cells to PLP 139-151 in mice fed 10 U mIFN-ot compared to mock IFN-fed (Table 1). ConA-stimulated spleen cells from mIFN-a-fed animals demonstrated decreased secretion of IFN-7 (Table 2) and IL-2 (Table 2) compared to mock IFN-fed animals.

3.4. Activated donor cells from animals administered murine IFN-a orally are less effective in transferring clinical disease compared to cells from mock IFN-fed mice (experiments 7 and 8)

25

2C

o

1.E

--c~-_

Mock

. . . . o . . . . 10 u n i t s

m[FN

u o

1C

,o.ia i i 1Q 11 12 13 14 Day

i i i i I i 15 16 17 18 19 2Q 21 22 2 3 post

i.p. transfer

Fig. 4. Experiments 7 + 8. Murine IFN-a fed animals do not adaptively transfer acute disease. 10×106 3-day ConA-activated T cells from mock IFN or 10 U mlFN-a fed immunized SJL/J mice that had undergone relapse attacks 35 days earlier were transferred adaptively i.p. to naive SJL/J mice (n ---6, mock IFN donor) (n = 6, 10 U mlFN-a donor) and followed for evidence of disease. Values represent combined data of two separate experimentsof mean group daily blinded clinical scores+ S.E.M. (mock IFN recipients vs. mlFNa treated recipients days 15-22, P < 0.01 by non-paired t-test).

Adoptive transfer experiments were performed to determine if activated spleen cells from 10 U mlFN-afed animals could prevent transfer of disease. ConAactivated mock IFN-fed or 10 U mIFN-a-fed T cells from immunized S J L / J mice, followed for 5 weeks after initiation of feeding during relapse (see Fig. 3), were transferred adaptively i.p. to recipient mice and followed for evidence of disease. Recipient mice of activated mock IFN T cells had significantly more severe clinical attack starting at day 13 compared to mIFN-a recipients (Fig. 4). ConA draining lymph node proliferation was significantly less from mIFN-a recipients compared to mock IFN recipients (Table 1). Lymph node ConA-induced IFN-7 secretion was inhibited in mIFN-a compared to mock IFN recipient animals (Table 2). However, there was no significant difference in the number of lesions between mock recipients versus mIFN-a recipients (mock IFN, 1.8 _+ 0.6 n = 6; 10 U mlFN-a, 1.5 + 0.5, n = 6).

4. Discussion

These data document that orally administered mlFN and hrlFN can suppress relapses and prevent adoptive

S.A. Brod et al. /Journal of Neuroimmunology 58 (1995) 61-69

transfer of EAE. The inhibition of an established and ongoing immune response by oral administration of immunoactive proteins is an important therapeutic issue. Oral administration of myelin proteins improves CR-EAE and decreases inflammation in rats and guinea pigs (Brod et al., 1991) and may decrease the severity of attacks in male DR2- multiple sclerosis patients and the frequency of MBP-specific T cell lines in myelin-treated individuals (Weiner et al., 1993). However, in the human trials and animal experiments, there was only partial suppression of clinical or pathological disease, suggesting that other orally administered immunoactive substances may be superior to myelin antigens. Animals administered 100-1000 U hrlFN-a exhibited less severe relapses than mock-fed animals although 1000 U was marginally less effective compared to 100 U. Thus species-specific mlFN-a was effective at one order of magnitude less than hrlFN-a and decreased the overall neurological deficits from baseline values. This is consistent with ongoing preliminary experiments that suggest a therapeutic dose-response window with 10-100 U species-specific oral mlFN-a the optimal dose, and 0.1, 1 and 1000 U significantly less effective in suppressing clinical relapse disease (data not shown). Human type 1 interferons can augment suppressor cell function in vitro in progressive multiple sclerosis (Noronha et al., 1990). In view of the immunoregulatory properties of the type 1 IFN, its response and production has been assessed in autoimmune diseases. Active rheumatoid arthritis, psoriasis, atopic dermatitis and multiple sclerosis exhibited a significantly reduced IFN-a/fl production compared to normal donors, which parallels the severity of the disease in multiple sclerosis (Seitz et al., 1987; Kapp et al., 1987; Kapp et al., 1988; Maruo, 1988). The reported abnormalities of production or response to type 1 interferon in autoimmune diseases have prompted several small pilot studies using type 1 interferons as therapeutic agents in multiple sclerosis (Panitch, 1987; Knobler, 1988; Kastrukoff et al., 1990). These studies have used systemic or intrathecal administered type 1 interferon in doses of 1 × 106 U or more daily without significant clinical improvement. A recent study of parenterally administered IFN-/3 in relapsing-remitting multiple sclerosis suggests that 8 × 106 U of IFN-/3 s.c. three times per week can decrease relapses by 40-50% and decrease brain inflammation as assessed by serial MRI (IFNB MS Study Group, 1993). A small controlled pilot study showed that similar effects could be obtained at equivalent doses of IFN-a2a given by intramuscular injection (Durelli et al., 1994). The immunomodulatory mechanism of orally administered IFN-a may be inducing anergy or generating suppressor T cells. Modulatory effects of ConA-

67

activated lymphocytes on the mitogen responses of normal responder cells can be abrogated by addition of anti-human leukocyte IFN serum (Kadish et al., 1980). In this setting, removing IFN may prevent the production of inhibitory factors, e.g. soluble immune response suppressor (SIRS) and macrophage derived suppressor factor (Mq~-SF, e.g. TGF-fl) by CD8 ÷ T cells (Aune and Pierce, 1982; Schnapper et al., 1984). Peripheralized T cells may also be required for IFN production after such mitogen stimulation (Stobo et al., 1974). Oral administration of 3 × 105 U IFN-a in mice does not result in detectable levels of IFN-a in the blood in contrast to i.p. administration, nor can its effect be blocked by circulating anti-IFN antibodies (Fleischmann et al., 1992). However, the neutropenic effect of orally administered IFN can be transferred by injection of white blood cells, but not serum, to recipient animals (Fleischmann et al., 1992). Oral administration of 3 × 106/kg IFN-a in dogs (Gibson et al., 1985), 6 × 106/kg in green monkeys (Wills et al., 1984) or in humans up to 150 × 106 U (Witt et al., 1992) does not result in detectable levels of IFN-a in the blood in contrast to parenteral administration and would not be effected by parenteral anti-IFN-a antibodies. Serum levels may not be necessary for clinical effect because recent experiments in our laboratory show that 10 U mlFN-a administered parenterally (s.c.), effective via oral delivery and unlikely to generate significant serum levels, are not clinically efficacious (data not shown). IFN-a may be an immunomodulatory molecule produced by activated CD8 + T and other immune cells that induces suppressor factors, such as SIRS and TGF-/3, which in turn induce hyporesponsiveness to immunized antigens such as MBP and MT. Evidence in humans suggest that 35S-labelled IFN-a can bind in significant amounts in the oropharynx on lymphocytes (Diez et al., 1987) and exhibit immunoinhibitory activity via the gut-associated lymphoid tissue (GALT). Mice receiving T cells adoptively from mIFN-a-fed animals do exhibit inflammation. Others have shown that adoptively transferred myelin oligodendrocyte glycoprotein (MOG) peptide-specific T cells can demonstrate inflammation in the CNS without inducing a neurological deficit (Linington et al., 1993). We might speculate that oral administration of IFN-a generates immunoregulatory CD8 ÷ T cells via the gut immune system that can traffic to the recipient's CNS but may not cause clinical disease by either generating suppressor factors locally, or demonstrating decreased responsiveness to immunized antigens and decreased IFN-3~ secretion. In summary, the oral administration of biological response modifiers such as IFN provides a potentially continuous means of generating immunosuppression of autoreactive T cell populations. Experiments are underway to determine whether actively induced disease

68

S.A. Brod et al. /Journal of Neuroimmunology 58 (1995) 61-69

can be suppressed by adoptively transferred cells from IFN-fed animals (suppression) or whether CD4 + T cells from IFN animals are unable to transfer disease to naive recipients (anergy). The evaluation of orally administered type 1 interferon in the experimental animal model murine CR-EAE may be of relevance for the therapy of early autoimmune disease including relapsing-remitting multiple sclerosis and other chronic non-neurological autoimmune diseases.

Acknowledgements Supported in part by a grant from the Clayton Foundation.

References Abreu, S.L. (1982) Suppression of experimental allergic encephalomyelitis by interferon. Immunol. Comm. 11, 1-7. Abreu, S.L., Tondreau, J., Levine, S. and Sowinski, R. (1983) Inhibition of passive localized experimental allergic encephalomyelitis by interferon. Int. Arch. Allergy Appl. Immun. 72, 30-33. Aune, T.M. and Pierce, C.W. (1982) Activation of a suppressor T-cell pathway by interferon. Proc. Natl. Acad. Sci. USA 79, 3808-3812. Balkwill, F.R. (1985) In: J. Taylor-Papadimitriou (Ed.), Interferons: Their Impact in Biology of Medicine. Oxford Medical Publications, Oxford, pp. 61-80. Blalock, J.E. and Barons, S. (1977) Interferon induced transfer viral resistance between animal cells. Nature 269, 422-425. Brod, S.A. and Burns, D.H. (1994) Suppression of relapsing experimental allergic encephalomyelitis in S J L / J mouse by oral administration of type 1 interferons. Neurology 44, 1144-1148. Brod, S.A., AI-Sabbagh, A., Sobel, R.A. and Weiner, H.L. (1991) Suppression of relapsing experimental allergic encephalomyelitis in Lewis rats and strain 13 guinea pigs by oral administration of myelin antigens. Ann. Neurol. 29, 615-622. Brod, S.A., Scott, M. and Burns, D.H. (1995) Modification of acute experimental autoimmune encephalomyelitis in the Lewis rat by oral administration of type 1 interferons. J. Int. Res. 15, 115-122. Brown, A.M. and McFarlin, D.E. (1981) Relapsing experimental allergic encephalomyelitis in the S J L / J mouse. Lab. Invest. 45, 278-284. Diez, R.A., Perdereau, B. and Falcoff, E. (1987) From old results to new perspectives: a look at interferons fate in the body. J, Interferon Res. 7, 553-557. Durelli, L., Bongioanni, M.R., Cavallo, R., Ferraro, B., Ferri, R., Ferrio, M.F., Bradac, G.B., Riva, A., Vai, S., Geuna, M., Bergnami, L. and Bergamasco, B. (1994) Chronic systemic highdose recombinant interferon alpha-2a reduces exacerbation rate, MRI signs of disease activity, and lymphocyte interferon gamma production in relapsing-remitting multiple sclerosis. Neurology 44, 406-413. Feuer, C., Prentice, D.E. and Cammisuli, S. (1985) Chronic relapsing experimental allergic encephalomyelitis in the Lewis rat. J. Neuroimmunol. 10, 159-166. Fleischmann, W.R., Koren, S. and Fleischmann, C.M. (1992) Oral administered interferons exert their white blood cell suppressive effects via a novel mechanism. Proc. Soc. Exp. Biol. Med. 201, 199-207.

Gibson, D.M., Coffer, S., Spiegel, H.E. and Colburn, W.A, (1985) Pharmokokinetics of recombinant leucocyte A interferon following various routes and modes of administration to the dog. J. Int. Res. 5, 403-408. IFNB MS Study Group (1993) Interferon beta-lb is effective in relapsing-remitting multiple sclerosis. I. Clinical results of a multicenter, randomized, double blind, placebo-controlled trial. Neurology 43, 655-669. Kadish, A.S., Tansey, F.A., Yu, G.S.M., Doyle, A.T. and Bloom, B.R. (1980) Interferon as a mediator of human lymphocyte suppression. J. Exp. Med. 151,637-650. Kapp, A., Gillitzer, R., Kirchner, H. and Schopf, E. (1987) Production of interferon and lymphoproliferative response in whole blood cultures derived from patients with atopic dermatitis. Arch. Dermat. Res. 279 (Suppl.), $55-$58. Kapp, A., Gillitzer, R., Kirchner, H. and Schopf, E. (1988) Decreased production of interferon in whole blood cultures derived from patients with psoriases. J. Invest. Dermatol. 90, 511-514. Kastrukoff, L.F., Oger, J.J., Hashimoto, S.A., Sacks, S.L., Li, D.K., Palmer, M.R., Koopmans, R.A., Petkau, A.J., Berkowitz, J. and Paty, D.W. (1990) Systemic lymphoblastoid interferon therapy in chronic progressive multiple sclerosis. I. Clinical and MRI evaluation. Neurology 40, 479-486. Kawade, Y. (1987) The interferon system in the mouse. In: S. Baron, F. Dianzoni, G.J. Stanton and W.R. Fleischmann (Eds.), The Interferon System. UT Press, Austin, pp. 169-174. Knobler, R.L. (1988) Systemic interferon therapy of multiple sclerosis: the pros. Neurology 38, 58-61. Linington, C., Berger, T., Perry, L., Weerth, S., Hinze-Selch, D., Zhang, Y., Li, H.-C., Lassmann, H. and Wekerle, H. (1993) T cells specific for the myelin oligodendrocyte glycoprotein mediate an unusual autoimmune inflammatory response in the central nervous system. Eur. J. Immunol. 23, 1364-1372. Lublin, F. (1985) Adoptive transfer of murine relapsing experimental autoimmune encephalomyelitis. Ann. Neurol. 17, 188-190. Maruo, Y. (1988) Interferon production and natural killer activity of peripheral blood lymphocytes obtained from patients with multiple sclerosis. Hokkaido J. Med. Sci. 63, 521-533. Mclnnes, M., Chambers, P.J., Cheetham, B.F., Beilharz, M.W. and Tymms, M.J. (1989) Structure function studies of interferons-a: amino acid substitutions at the conserved residue tyrosine 123 in human interferons-a. J. Interferon Res. 9, 305-314. Miller, S.D., Clatch, R.J., Pevear, D.C., Trotter, J.L. and Lipton, H.L. (1987) Class II restricted T cell response in Theiler's murine encephalomyelitis virus induced demyelinating disease. J. Immunol. 138, 3776-3784. Noronha, A., Toscas, A. and Jensen, M.A. (1990) Interferon beta augments suppressor cell function in multiple sclerosis. Neurology 27, 207-210. Panitch, H.S. (1987) Systemic alpha-interferon in multiple sclerosis: long term patient follow-up. Arch. Neurol. 44, 61-63. Peters, B.A. and Hinrichs, D.J. (1982) Passive transfer of experimental allergic encephalomyelitis in the Lewis rat with activated spleen cells: differential activation with mitogens. Cell. Immunol. 69, 175-185. Raine, C.S. and Stone, S.H. (1977) Animal models for multiple sclerosis: chronic experimental allergic encephalomyelitis in inbred guinea pigs. N.Y. State J. Med. 77, 1693-1696. Schnapper, H.W., Pierce, C.W. and Aune, T.M. (1984) Identification and initial characterization of Con A and interferon-induced human suppressor factor: evidence for a human equivalent of murine soluble immune response suppressor (SIRS). J. Immunol. 132, 2429-2435. Seitz, M., Napierski, I., Augustin, R., Hunstein, W. and Kirchner, H. (1987) Reduced production of interferon alpha and interferon gamma in leukocyte cultures from patients with active rheumatoid arthritis. Scand. J. Rheumat. 16, 257-262.

S.A. Brod et aL /Journal of Neuroimmunology 58 (1995) 61-69 Sobel, R.A., Tuohy, V.K., Lu, Z., Laursen, R.A. and Lees, M.B. (1990) Acute experimental allergic encephalomyelitis in the SJL/J mouse induced by a synthetic peptide of myelin proteolipid protein. J. Neuropathol. Exp. Neurol. 49, 468-474. Stobo, J., Green, I., Jackson, L. and Baron, S. (1974) Identification of a subpopulation of mouse lymphoid cells required for interferon production after stimulation with mitogens. J. Immunol. 112, 1589-1593. Streuli, M., Hall, A., Boll, W., Stewart, W.E., Nagat, S. and Weisman, C. (1981) Target cell specificity of two species of human interferon-c~ produced from Escherichia coli and of hybrid molecules derived from them. Proc. Natl. Acad. Sci. USA 78, 2848-2852. Weber, H., Valenzuela, D., Lujber, G., Gubler, M. and Weissman, C. (1987) Single amino acid changes that render human IFN-a 2 biologically active on mouse cells. EMBO J. 6, 591-598. Weiner, H.L., Mackin, G.A., Matsui, M., Orav, E.J., Khoury, S.J., Dawson, D.M. and Hailer, D.A. (1993) Double-blind pilot trial of

69

oral tolerization with myelin antigens in multiple sclerosis. Science 259, 1321-1324. Whitham, R.H., Bourdette, D.N., Hashim, G.A., Herndon, R.M., llg, R.C., Vandenbark, A.A. and Offner, H. (1991) Lymphocytes from SJL/J mice immunized with spinal cord respond selectively to a peptide of proteolipid protein and transfer demyelinating relapsing experimental autoimmune encephalomyelitis. J. Immunol. 146, 101-107. Wills, R.J., Spiegel, H.E. and Soike, K.F. (1984) Pharmacokinetics of recombinant leucocyte A interferon following IV infusion and bolus, IM, and PO administration to african green monkeys. J. Int. Res. 4, 399-409. Wisniewski, H.M. and Keith, A.B. (1977) Chronic relapsing experimental allergic encephalomyelitis: an experimental model of multiple sclerosis. Ann. Neurol. 1, 144-148. Witt, P.J., Goldstein, D., Storer, B.E., Grossberg, S.E., Flashner, M., Colby, C.B. and Borden, E.C. (1992) Absence of biological effects of orally administered interferon-/3so r. J. Int. Res. 12, 411-413.