Experimental Autoimmune Panencephalitis and Uveoretinitis Transferred to the Lewis Rat by T ...

13 Pages·2003·4.13 MB·English

Experimental Autoimmune Panencephalitis and
Uveoretinitis Transferred to the Lewis Rat by T
Lymphocytes Specific for the $100~/Molecule, a
Calcium Binding Protein of Astroglia
By Kirnikazu Kojirna,* Thomas Berger,r Hans Lassmann,*ll
Dunja Hinze-Selch,* Yiping Zhang,* Jochen Gehrrnann,(cid:1)82
Konrad Reske,**Hartrnut Wekerle,* and Christopher Linington*
From the *Department of Neuroimmunology, Max-Planck Institute for ,yrtaihcysP 82152
,deirsnitraM Germany; eht lacigolorueN* ;etutitsnI eht tnemtrapeDg of ,ygolorueN University of
;anneiV eht hcraeseRII Unit for Experimental ,ygolohtaporueN Austrian Academy of ,secneicS
A-1090 Vienna, Austria; eht IDepartment of ,ygolohpromorueN Max-Planck Institute for
,yrtaihcysP 82152 ;deirsnitraM and eht **Institut fiir Immunologie, Johannes grebnetuG D
o
,tiitisrevinU D-55101 Mainz, Germany wn
lo
a
d
e
d
fro
m
yrammuS h
ttp
The pathogenic potential of autoimmune T cell responses to nonmyelin autoantigens was investigated ://ru
p
ianu toantigen. the Lewis rat The using Lewis the rat astrocyte-derived mounts a vigorous calcium RT1B binding 1 (major protein histocompatibility $100/3, as a model complex nonmyelin class ress.org
II) restricted autoimmune response to an immunodominant $100/3 epitope (amino acid residues /jem
/a
i7n6 -91). the nervous The adoptive system, transfer but only of S100/3-specific minimal neurological T cell lines dysfunction induced a in severe naive inflammatory syngeneic recipients. response rticle-p
d
The inability of cificeps-3j001S T cell transfer to induce severe disease was associated with a decreased f/1
8
0
recruitment of ED1 § macrophages into the central nervous system (CNS) in comparison with /3
/8
that seen in severe experimental autoimmune encephalomyelitis (EAE) induced by the adoptive 17
/1
transfer of myelin basic protein (MBP)-specific T line cells. Moreover, unlike encephalitogenic 39
6
MBP-specific T cell lines, S100/3-specific T cell lines exhibited no cytotoxic activity in vitro. 02
5
Histopathological analysis also revealed striking differences in the distribution of inflammatory /81
7
lesions in MBP- and S100/3-specific T cell-mediated disease. In contrast to the MBP paradigm, .pd
f b
S100/3-specific T cell transfer induces intense inflammation not only in the spinal cord, but y g
u
throughout the entire CNS and also in the uvea and retina of the eye. In view of the distribution e
of lesions throughout the grey and white matter of the CNS we propose to term this new model st on
2
experimental autoimmune panencephalomyelitis (EAP) to differentiate it from EAE. These 0 Ja
experiments demonstrate for the first time that nonmyelin CNS autoantigens can initiate a nu
a
pathogenic autoimmune T cell response, although the nature of the target autoantigen profoundly ry 2
0
influences the clinical and histopathological characteristics of the resulting autoimmune disease. 23
This is not simply a consequence of the distribution of the autoantigen, as both MBP and 3/001$
are coexpressed in many areas of the CNS, but reflects differences in the capacity of different
regions of the CNS to process and present specific autoantigens. This new model of T cell-mediated
autoimmune CNS disease exhibits a number of similarities to multiple sclerosis (MS), such sa
its mild clinical course and the involvement of areas of the brain and eye, which are absent in
myelin-mediated models of EAE. Nonmyelin autoantigens may therefore play an unexpectedly
important role in the immunopathogenesis of inflammatory diseases of the CNS.
T wo basic tenets in cellular immunology were radically
1 snoitaiverbbA desu ni siht :repap ,BBB niarb-doolb ;reirrab CNS, lartnec
revised during the past decade as a direct consequence suovren ;metsys ,EAE latnemirepxe enummiotua ;sitileymolahpecne ,PAE
of studies of experimental autoimmune encephalomyelitis latnemirepxe enummiotua ;sitilahpecnenap ,GOM nileym etycordnedogilo
(EAE) .1 First, the concept of "immunoprivilege" with re- ;nietorpocylg ,SM elpitlum ;sisorelcs PPD, deifirup protein .evitavired
718 J. Exp. Med. (cid:14)9 The rellefekcoR ytisrevinU sserP (cid:12)9 0022-1007/94/09/0817/13 00.2$
emuloV 081 rebmetpeS 1994 928-718
spect to the central nervous system (CNS) was reappraised Table .1 Amino Acid secneuqeS of eht citehtnyS Rat SlO0~
after the demonstration that not only can activated T lympho- seditpeP Used ni This Study
cytes migrate through the endothelial blood-brain barrier
Designation Residues Amino acid sequence
(BBB) regardless of their antigen specificity ,1( 2); but also
that the CNS contains numerous cellular elements (astrocytes,
microglia, cerebral endothelial cells) that can function sa APCs A4 1-14 FVDILAVMAKELES
(3-5). Second, the isolation of encephalitogenic T cell lines 4B 4-18 SYQHFVDILAVMAKE
from both myelin-primed (6) and naive rats (7) demonstrated C4 10-24 KDGERGSYQHFVDIL
that potentially autoaggressive T cell clones are not neces-
D4 15-20 KKLKHKDGERGSYQH
sarily silenced by either deletion or anergy, but are normal
E4 20-34 EKLESKKLKHKDGER
components of the healthy immune repertoire (8, 9).
F4 25-39 ENNILEKLESKKLKH
However, these results are all derived from studies in which
EAE is mediated by an autoaggressive T cell response against A5 31-45 ELFHSLENNILEKLE
a single myelin antigen, the myelin basic protein (MBP), which 5B 36-50 QEKIEELFHSLENNI
until very recently appeared to be the predominant, if not C5 41-55 KDVVEQEKIEELFHS
the only, encephalitogenic protein present in the CNS. Analysis D5 46-60 LTEMVKDVVEQEKIE
of the encephalitogenic, MBP-specific T cell response revealed 5E 51-65 DGDEDLTEMVKDVVE
that it exhibited several unusual properties. In at least two D
5F 56-70 FDCEGDGDEDLTEMV ow
species, the Lewis rat and mice with the H-2 u haplotype n
A6 61-75 AMFEQFDCEGDGDED loa
(PL/J and B10.PL strains), the primary MBP-specific T cell d
e
response is directed against single, strongly immunodominant 6B 66-80 VMSVFAMFEQFDCEG d fro
epitopes, the identity of which is species and strain depen- C6 71-85 HCATTVMSVFAMFEQ m h
dent. Furthermore, encephalitogenic T cells specific for these D3 76-91 EHEFFEHCATTVMSVF ttp://ru
immunodominant epitopes use a highly restricted repertoire p
re
Vo~2f/V ~4 TCK genes, elements generally (10-13). TCR The V~8.2 potential in combination therapeutic witrhele - eni v2oB6 ,)reS/ulG( ~001$ 87protein ,)alA/reS( sreffid dna from 08 tar .)ell/laV( B001S ta seudiser 7 ,)laV/teM( ss.org/je
vance of these observations was rapidly demonstrated in EAE m/a
(14-18); however, the molecular characteristics responsible their pathogenic potential. S100B-specific, class II MHC-re- rticle
for the encephalitogenicity of the MBP-specific T cell response stricted, CD4 § Thl T line cells were found to mediate an -pd
have still to be elucidated. In particular, it has to be demon- intense inflammatory response throughout the CNS that differs f/180
strated whether EAE mediated by myelin-specific T cells pro- dramatically from MBP-mediated EAE with respect to le- /3/8
1
vides a general paradigm to study inflammatory responses sion distribution, structure and cellular composition. Addi- 7/1
3
in the CNS, or is a special case of limited usefulness. tionally, S100B-specific T cells induce inflammatory changes 960
To approach this problem and identify the structural/func- in both the uvea and retina of the eye, a pathology reported 25/8
tional characteristics of CNS proteins necessary to induce EAE, to occur in a subgroup of patients with multiple sclerosis 17.p
we elected to study the encephalitogenic potential of non- (MS) (26). df b
myelin CNS proteins that differ from MBP with respect to These observations demonstrate that MBP and other my- y gu
e
their site of synthesis and physicochemical characteristics. One elin proteins are not the only CNS autoantigens able to in- st o
of these proteins was $100/3, a calcium binding protein that duce extensive autoimmune-mediated encephalomye- n 2
0
was first isolated from the CNS. S100B is a small, highly litis/panencephalomyelitis in mammals. A potential role for Ja
n
Liskoleu ble, MBP, acidic S100B protein is a major consisting component of 19 of amino the nervous acids (19, system, 20). ahave utoimmune to be considered responses the to pathnogoennmesyiesl in autoantigens of inflammatory will now dis- uary 20
2
accounting for (cid:127)0.4% of the total soluble protein in the CNS eases of the CNS, in particular MS. 3
(21), but unlike MBP which is only present in oligodendro-
cytes and myelin, expression of 3/001$ in the CNS is restricted
to astrocytes. Elsewhere, $100~ is also expressed by Schwann Materials and Methods
cells in the peripheral nervous system and Mtiller cells in the Animats dna Antigens. Inbred Lewis rats aged 6-8 wk were ob-
retina (22, 23). Astrocytes also secrete a biologically active tained from the animal breeding seitilicaf of the Max-Planck Insti-
form of SIOOB and the protein can be detected in the tute (Martinsried, Germany). Bovine 3/001$ protein saw obtained
cerebrospinal fluid (24, 25). from Sigma Chemical Co. (St. Louis, MO). Synthetic rat B001S
To provide a direct comparison with the encephalitogenic seditpep were prepared by Dr. N. Groome (Oxford sekoorB Univer-
MBP-specific T cell response a panel of S100/3-specific T cell ,ytis Oxford, UK) on a multiple peptide synthesizer (Abimed, Dus-
seldorf, Germany) using standard Fmoc chemistry (Table .)1 Pep-
lines was generated from both bulk, and limiting dilution
tide purity was monitored by amino acid ,sisylana esahp-esrever
cultures of lymph node cells obtained from S100B-primed
HPLC and, when appropriate, mass spectrometry.
animals. These T cell lines were then analyzed in vitro with
mAbs dna .ares-itnA The following mAbs were used in this
respect to their epitope specificity, phenotype, TCR V0 gene
study: OX6 (RT1.B, rat I-A), OX17 (RT.1D, rat I-E), and OX18
usage, cytokine production, and cytotoxicity, and in vivo after (MHC class I), W3/13 (pan T cell/granulocyte); W3/25 (CD4);
adoptive transfer into naive syngeneic recipients to determine OX8 (CDS); OX39, II.-2 receptor; OX22 (CD45RO). The suc-
818 Experimental Autoimmune sitilahpecnenaP dna sitiniteroevU
ceeding rat T cell receptor-specific mAbs were used: R73 (con- tions. Transfectants were selected using Hygromycin B at 200/xg/ml
stant region of TCR-ce//3 27); R78 (TCR Vo 8.2); B73 (TCR added 48 h after transfection.
V08.5); and G101 (TCR Val0) (28) and HIS42 (TCR Vol6). Rat Cytokine assays were performed on the culture supernatants of
macrophages and microglia were differentiated using the mAbs the various T cell lines as described previously (37) after 48 h of
TRPM3 (29), ED1, ED2, ED3 (30), and MUC102 (31). Rabbit antigen-specific restimulation. IFN-3' levels were measured using
anti-S100~ and anti-GFAP antisera were purchased from Sigma a rat IFN-3' ELISA kit (Holland Biotechnologies, Leiden, The
Chemical Co. W3/13, W3/25, OXS, OX39, and MUC102 were Netherlands). The IL-2-dependent CTLL cell line was used to assay
purified in our laboratory from the culture supernatants of the ap- IL-2 activity using mouse rlL-2 as a standard. II--6 activity was de-
propriate hybridoma cell lines; ED mAbs, OX22, OX42, and HIS42 termined using the IL-6--dependent 7TD1 hybridoma cell line. TCR-
were purchased from Serotec (Oxford, UK) and TR.PM3 from Di- oe//3 activity was assayed using TNF-oe//3--sensitive LM cells and
anova (Hamburg, Germany). a polyclonal TNF-ot-specific goat antimurine antibody (Genzyme
cificeps-negitnA T Cell Lines. Lewis rats were immunized in the Corp., Cambridge, MA). The 7TD1 and LM ceils were generously
hind foot pads with either 50/ig native antigen ($100~ protein, provided by Dr. K. Frei (Universit~itespital Zurich, Switzerland).
guinea pig MBP, or purified protein derivative PPD), or 35 #g cihpargoroulfotyC Analysis. The surface phenotype of the T cell
of synthetic peptide, in complete Freund's Adjuvant (CFA) con- lines was investigated by indirect immunofluorescence staining. 2
taining 4 mg/ml muiretcabocyM sisolucrebut H37a (Difco Laborato- x 10 s viable cells were washed with PBS containing 0.2% BSA
ries, Inc., Detroit, MI). and 01 mM NaN3 and incubated with the primary mAb for 1 h
Two different protocols were used to establish antigen specific on ice. The cells were washed to remove the primary antibody and
T cell lines. 01 d after immunization single cell suspensions were stained with fluorescein conjugated goat anti-mouse IgG (Dianova,
D
prepared from the draining lymph nodes. The primed cells were Hamburg, Germany) for 1 h on ice. Cells were then washed and ow
either cultured at a concentration of 701 cells/ml to establish bulk immediately analyzed using a FACScan | (Becton Dickinson, Heidel- nlo
a
T cell lines (LS) (6), or alternatively serially diluted to provide a berg, Germany). A live gate was obtained by incubating the cells de
d
large number of pauciclonal T cell lines (prefixed; 1S 32). In the in PBS containing propidium iodide. fro
m
latter case, lymph node cell suspensions were cultured in 96-well noitcudnI of EAE. Lewis rats were injected into the tail vein h
5r01ound-bottom to 100 cells/well. microtiter The plates wells at were cell supnpulemmbeenrtse d ranging with from antigen 2 x wtiigtenh vardyriinvegn restimulationnu. mbers of T Animals cell blasts were harvested cared for after in 72 accordance h of an- ttp://rup
cell and numirbraediarte d of (4,000 2 x 10Vwell. rad) syngeneic After 3 thymus d, the medicells a to was give removed a final wclinical ith German signs of federal EAE on regulations a daily basis. and EAE weighed severity and was examined scored for as: ress.org
and the remaining T cells expanded for a further 7-10 d in media 0, no disease; 0.5, partial loss of tail tonus; ,1 complete tail atony; /jem
asntuipgpenl emented in the wpresence ith IL-2. of The 2 x T l0 cells s irradiated were then (4,000 restimulated rad) syngeneic with d2e, ath. hind Demyelination limb paraparesis; was 3, induced hind limb 5 d after paralysis; the injection 4, moribund; ofT cells 5, /article-p
thymus cells/well. After visual inspection of the cultures, those by the intravenous injection of 4 mg of the mouse mAb 8-18C5 df/1
positive for growth were expanded in the presence of IL-2 and specific for the myelin oligodendrocyte glycoprotein (MOG) (38). 80
/3
antigen-specific 1S T cell lines selected as described for the bulk Histology. Rats were perfused through the heart with 4% /8
1
culture-derived LS cell lines. The LS and 1S T cell lines were then paraformaldehyde in PBS and tissue postfixed in the same fixative 7/1
3
subsequently propagated as described previously (6, 33). for 3 h at 4~ Alternatively, tissues were removed and immedi- 96
0
noitarefilorP and L ymphokine Assays. Lymph node cells (2 x ately snap frozen. Multiple sections were stained with hematox- 25
/8
10S/well), or 2 x 401 T line cells plus 0.9 x 601 irradiated (4,000 ylin-eosin and Luxol fast blue. Immunocytochemistry was per- 17
rad) syngeneic thymocytes were cultured in 0.2 ml medium in 96- formed on both paraffin embedded and frozen sections as described .pd
f b
well round-bottomed microtiter plates in the presence or absence previously (39). y g
of the appropriate antigens. After 56 h, 1 #Ci/well 3Hthymi- The frequency of inflammatory infiltrates in the cortex, periven- ue
dine was added and the cells were harvested 61 h later. The mean tricular white matter, diencephalon, mesencephalon, cerebellum, st on
cpm were determined for triplicate cultures using a solid phase gas medulla oblongata, and spinal cord was determined by counting 20
scintillation counter (Matrix 96 Direct Beta Counter; Packard, the number of perivenous inflammatory cuffs in multiple hema- Jan
u
bFyra nkfurt, this method Germany). are approximaItte ly should be noted five thatt imes the lower counts than obtained with toofxy lin-eosin-stained 375 mmVanimal. sections The results in standardized are expressed regions as the of numba er total area of ary 20
2
methods using scintillation fluid. inflammatory infiltrates/mm .2 The total number of T cells and 3
The ability of either syngeneic astrocytes or the mouse L cell macrophages was determined in sections immunostained with the
line, RT3.3, which is transfected with functional RTIB 1 to act appropriate antibodies using a 100-point morphometric lattice placed
as APCs was determined using irradiated cultures (5,000 tad) seeded in the microscope's ocular (ocular x 10, objective x 100). The
into flat bottomed 96-well plates at a density of 2 x 401 cells/weU. total area quantitated was 0.6 mm 2 in the spinal cord and 0.2 mm 2
T line cells (2 x 104/well) were added to give a total volume of in the sciatic nerves and spinal roots and the results are presented
0.2 ml media in the presence or absence of antigen. T cell prolifera- as the number of cells/mm 2 of tissue area.
tion was determined at the end of 3 d culture as described above. etycortsA Culture Astrocytes were prepared from newborn lewis
The cell line RT3.3 was established using the mulcos system (34) rat cerebra. The meninges were removed and a single cell suspen-
(a kind gift from Dr. I. Saito, National Institutes of Health, Toyko, sion prepared by trypsinization and seeded into plastic culture flasks
Japan). Briefly, cDNA clones pLRB118 (35) and pLR~x6 (36) were (4). The nonadherent cells were removed after 4 h and cultured
inserted into EcoRI and HindlII sites respectively of the cassette in 5% CO2 at 37~ The medium was first replaced after 24 h
plasmid pmoRH, excised by Sill digestion and ligated with the and subsequently every 3-4 d. After 01 d the primary cultures were
Sfil linearized cosmid vector dCHD2L (34). MHC class II and in- shaken for 3-5 h to remove microglial cells. The remaining ad-
variant chain negative rat-2 cells )601( were transfected with 5/ig herent astrocytes were removed by trypsinization and transferred
of cosmid DNA employing the DOTAP procedure (Boehringer to flasks at a density of 3 x 104/cm .2 After a further 2-3 wk in
Mannheim, Mannheim, Germany) following the supplier's instruc- culture, the cells were again harvested by trypsinization before use
819 Kojima et .la
Table 2. Antigen yticificepS of cificeps-~/OOlS T Cell Lines
Response to antigens
Line No. Ag D3 $100/3 PPD Con A
naem malc
LS1 28 10,517 12,651 13 10,876
LS2 100 5,642 10,764 82 11,440
SP-D3 524 10,583 472,8 36 10,722
S1-D1 3 4,310 017,7 2 8,869
$1-E12 410 2,968 5,212 171 9,242
S 1-F5 5 2,643 7,082 2 10,176
$1-D2 6 2,457 10,354 3 9,955
$1-G5 7 2,133 5,533 3 9,663
$1-F12 5 2,096 003,7 3 11,699
D
o
$1-D12 21 1,160 371,7 7 9,336 wn
lo
$1-E8 6 1,037 4,111 4 8,749 ad
e
d
$1-E2 6 1,096 8,450 8 10,284 fro
m
$1-G12 4 755 4,788 6 780,01 h
ttp
$1-F4 124 552 4,512 7 11,233 ://ru
p
$1-E3 6 452 2,590 0 9,70i re
$1-F6 28 375 6,746 8 11,288 ss.org
$1-F8 4 318 10,205 3 12,203 /jem
/a
$1-C7 3 218 6,049 3 9,219 rticle
$1-G7 8 131 9,079 5 10,614 -pd
f/1
$1-G6 3 301 10,574 2 12,408 8
0
/3
$1-G9 1 54 2,588 2 8,281 /8
1
7
$1-G8 6 8 3,232 3 5,817 /13
9
6
0
2
5
/8
Antigen specific T line cells 2( x 104/well) were cultured in triplicate with irradiated (4,000 rad) syngeneic tbymocytes (0.9 x )11ew/601 in the 17
.p
presence or absence of either the synthetic rat $1003 peptides (Table ,)1 bovine $1003, DPP (final concentrations 01 #g/ml), or Con A (2.5/~g/ml) df b
for 3 d. Cultures were pulsed with 3HTdR for the last 61 h of the 72-h culture and the incorporation of isotope determined sa described. The y g
results are presented sa the mean cpm obtained, standard deviations were no more than 16%. In addition to the response to peptide D3 the following ue
responses to peptides C6 (LS1, 503 cpm; LS2, 014 cpm; $1-F4, 462 cpm) and A5 ($1-E2, 423 cpm) were observed. st o
n
2
0
Ja
n
u
a
Table 3. Surface epytonehP of LS dna 1$ cificeps-lfOOlS T Cell seniL ry 20
2
3
Line W3/13 W3/25 OX8 OX39 OX22 R73 R78 B73 G101 HIS42
LS1 99.0 96.9 2.5 98.7 0.3 99.0 4.1 4.4 11.3 15.7
LS2 98.9 98.1 4.5 95.8 0.4 99.7 1.8 2.5 3.0 7.8
S1-D1 99.7 99.2 1.2 99.6 0.4 99.9 0.3 0.4 9.8 7.8
$1-D2 99.4 98.3 1.8 98.8 0.47 100 0.2 0.4 4.8 82.0
$1-C7 99.3 97.4 1.1 97.3 0.4 99.9 0.4 0.4 0.3 50.0
$1-G7 98.1 99.3 1.4 100 0.6 99.6 0.4 0.4 0.4 7.8
$1-G9 99.6 99.3 2.9 99.9 0.5 99.9 0.2 0.5 0.2 98.8
SP-D3 99.7 96.3 2.2 96.4 0.3 99.9 1.8 8.9 3.5 14.4
Viable lymphoblasts (2 x 01 )s were incubated with the appropriate mouse mAb and subsequently stained with FITC-conjugated goat anti-mouse
IgG and analyzed by FACScan | Controls were stained with the second Ab alone. The results are indicated sa the percent of positive cells.
820 Experimental Autoimmune Panencephalitis and Uveoretinitis
Table 4. Cytokine Profiles of SlOOfl- and MBP-specific T Cell Lines
Supernatant activities
Cell line ~3-NFI IL-2 IL-6 TNF
mean _+ SE
1SL >10,000 6.8 + 0.3 8531 +_ 216 803 +_ 28
2D-1$ >10,000 8.1 +_ 0.1 5231 +_ 236 851 + 93
3D-PS >10,000 2.2 +_ 0.2 3181 +_ 057 801 +_ 32
1PC >10,000 2.9 +_ 0.4 721 +_ 36 391 _+ 78
58Z >10,000 ND 094 +_ 072 507 +_ 97
34LMR >10,000 0.9 +_ 1.0 3131 +_ 575 022 +_ 77
APC control <250 ND 051 ND
D
o
w
n
in cytotoxicity and proliferation .syassa These cultures contained lines specific for either S100fl (LS1, S1-D2, SP-D3) or MBP lo
a
d
>97% GFAP + cells of which >80% were lf001S .+ Microglia erew (Z85 37 CP1 34, 34LMR) was analyzed during the first ed
not detected. 48 h of an antigen-specific restimulation. All T cell lines from
etycortsA yticixototyC Assays. Astrocytes harvested by trypsini- secreted IFN-% IL-6, and TNF-o~ into the medium and IL-2 http
medium sation were )selgaE( cultured (BME) in llew-69 )llew/401( flat-bottom in the setalppresence in complete or ecnesba lasab of was detected at low levels in the supernatants with the ex- ://rup
ception of Z85 (Table 4). The production of IFN-3r indicates re
001 U/ml of rat rlFN-7 (Holland Biotechnologies) for 84 h before that these cell lines are of the Thl subtype. ss.o
esu in either cytotoxicity or proliferation .syassa rg
The MHC restriction of the Lewis rat T cell response to /je
Cytotoxicity saw determined by rC15 release (40). The sllec m
rC15were sawlabeled removed with rC15by washing (1/~Ci/10 the 4 ccells ells) three for 2 times, h at 37~ followed ssecxE by bTovhien e anti-rat S100fl I-A was mAb established OX6 (125/~g/ml) using the reduced T cell litnhee prolifer- $1-D2. /article-p
incubation at 37~ for 5 h and a second round of washing. Freshly ative response of this T cell line to S100fl by 70%, whereas df/1
8
detsevrah T cell blasts were added to the labeled astrocytes in com- the same concentration of mAb reduced the proliferative re- 0/3
plete DME to give different E/T ratios in the ecneserp or ecnesba sponse of a control MBP-specific T cell line, C1-C9, by 97%. /81
7
of the antigen (10/~g/ml). Radioactivity saw measured in tripli- The mAbs OX17 and OX18 had no significant effect on the /13
9
cate using a gamma counter (Cobra Auto-gamma 5003; )drakcaP 6
proliferative response of either T cell line. These results were 0
2
after a 12-h incubation. Maximal lysis saw determined after incu- confirmed using the transfected rat-2 mouse L cell line RT3.3, 5/8
bation of the astrocytes with 2.5% Triton .001-X 17
which expresses both the ~c and fl chains of Lewis rat I-A. .p
d
This mouse fibroblast call line was capable of acting as an f by g
APC in vitro inducing an antigen-specific proliferative re- ue
sponse in both T cell lines tested (Table 5). st o
n
Results noitacifitnedI of eht tegraT sepotipE dna (cid:127)V TCR ~gasU A 20
noitaziretcarahC of cificeps-lfOOIS T Cell Lines. Lewis rats panel of sixteen overlapping synthetic rat S100fl peptides elbaT( Jan
u
a
immunized with bovine S100fl mounted a vigorous T cell )1 was used to identify T cell epitopes recognized by the -lf001S ry 2
response to this antigen and of a total of 16 S100fl-specific specific T cell lines. The T cell lines LS1 and LS2 both recog- 02
3
T cell lines generated during this study, 12 were investigated nized an immunodominant epitope common to both rat and
in more detail; 2 T cell lines derived from bulk cultures (LS1 bovine S100fl located within the COOH-terminal sequence
and LS2); and a further 91 S100fl-specific 1$ T cell lines de- of the protein, amino acid 76-91, peptide D3 (Fig. .)1 Anal-
rived from microcultures. After three restimulation cycles the ysis of the 1$ T cell lines sdected from serial dilutions of
T cell lines were specific for bovine S100fl protein and ex- the original lymph node cell preparations confirmed that the
hibited no residual response to PPD (Table 2). rat S100fl sequence 76-91 was a dominant epitope with re-
The eight S100fl-specific T cell lines analyzed by FACS | spect to the T cell response to bovine S100fl (Table 2). How-
all exhibited the classical W3/13 +, OXS-, CD25 + and ever, although all 1$ T cell lines proliferated strongly in re-
TCR-odfl + phenotype of CD4 + rat T cell lines. However sponse to native bovine S100fl, the proliferative response to
they are clearly differentiated from encephalitogenic CD4 § this epitope was significantly lower than that seen in the bulk-
MBP-specific rat T cell lines and clones by their TCR Va derived T cell lines 1SL and LS2. Proliferation in response
gene usage (Table 3). Unlike the MBP-specific T cell lines, to peptide D3 ranged from 20 to 55% of the response to
S100fl-specific T cell lines do not preferentially use TCR. bovine S100fl for eight of the 1$ T cell lines analyzed, while
Va8.2. the remaining T cell lines responded poorly to this rat epi-
The spectrum of cytokines secreted by representative T cell tope. In the case of the T cell hnes such as $1-G8 and $1-G9
821 Kojima et al.
Table 5. MHC Restriction of eht clficeps-~0OIS T Cell Line S1-D2 dna eht cificeps-PBM T Cell Line CI-C9
mAb added
T cell line APC Ag 0 OX6 OX17 OX18
125 ttg/ml
$1-D2 Thy + 9534 +- 380 2832 +. 364 (70.3) 8268 +- 413 7846 + 471
Thy - 4 _+ 2 ND ND ND
RT3.3 + 1085 +_ 430 ND ND ND
RT3.3 - 6 +- 2 ND ND ND
C1-C9 Thy + 4602 +- 920 134 +_ 105 (97.1) 5425 + 109 3272 +_ 262
Thy - 4 _ 2 ND ND ND
RT3.3 + 1136 +. 511 ND ND ND
RT3.3 - 4 +_ 4 ND ND ND
D
o
w
n
The MHC restriction of the T cell lines $1-D2 )cificeps-B001S( and C1-C9 )cificeps-PBM( was determined using the mAbs OX6, OX17, dna OX18. loa
d
Confirmation that these responses were I-A restricted was obtained using the transfected mouse L cell line, KT3.3, that expresses Lewis rat KT1B .1 ed
Expression was confirmed by both FACS | analysis and biosyntheic labelling. The data era presented sa the mean cpm of triplicate cultures +_ .DS fro
m
The percent inhibition obtained in the presence of the mAb OX6 is given in brackets. h
ttp
there was no response to any of the peptides tested indicating Antigen Added ://rup
re
that they recognize bovine S100/~-specific epitopes (Table 2). Media only ss.o
The T cell line, SP-D3, established from Lewis rats immunized SIOOB rg/je
with the peptide D3 (amino acid 76-91), as anticipated, 1-14 m/a
proliferated vigorously in response to both the immunizing 4-18 rticle
peptide and native bovine $100/~ (Table 2). All subsequent -p
studies were performed using the T cell lines LS1, LS2, $1- 1015--229 4 df/180
D2, $1-C7, $1-G7, $1-G9, and SP-D3, which exhibit differing 20-34 /3/81
degrees of reactivity with the immunodominant peptide amino 25-39 7/13
acid 76-91. 96
31-45 0
FACS | analysis using a panel of four murine mAbs recog- 25
36-50 /8
nizing defined rat TCR V0 products revealed that only 17
41-55 .p
2lin-es 4% expressed of the LS1 V08.2 and LSo2 n their S100~-specific surface (Table bulk 3). culture The extT ent cell 46-60 df by g
51-65 ue
woafs Va8.5 slightly expression higher (3-11%)w,a s also as very was low, usage whereas of V016 V~10 (8-16%). usage 56-70 st on 2
The SP-D3 T cell line raised against the immunodominant 61-75 0 Ja
peptide D3 also expressed low levels of V~8.2 (2%), V010 66-80 nua
(4%), and V~8.5 (9%), and again a significant proportion 71-85 ry 2
0
of Va16 (14%). None of the $1 T cell lines tested expressed 76-91 23
I _ I I i _ i ~ i
either V~8.2 or V~8.5, whereas V~16 usage was (cid:127)50% in 0 2 4 6 8 10 12 14
$1-C7, 80% in $1-D2, and 100% in $1-G9. Within the limits c,p.m. (thousands)
of the technique used these results demonstrate that unlike
Figure 1. Epitope yticificeps of the cificeps-BOOlS T cell line, .1SL 1SL
the T cell response to MBP, the Lewis rat T cell response
T line cells were restimulated with either native bovine ,lf001S or syn-
to $100~ does not preferentially use TCR V~8.2.
thetic rat B001S peptides (see Table )1 sa described in the text. Only the
SlOOfl-specific CD4 + T Cell Lines Induce an Autoimmune overlapping COOH-terminal peptides amino sdica 71-85 and amino sdica
Panencephalitis with Uveoretinitis, Rather Than a Pure En- 76-91 induced a significant proliferative response of the 1SL T cell line
cephalomyelitis. Naive Lewis rats injected with 5 x 10 6, in vitro locating the immunodominant T cell epitope(s) to this domain
of the molecule. No other peptides were found to initiate a tnacifingis
107, or 2 x 701 LS1 and LS2 lymphoblasts failed to exhibit
evitarefilorp response in either this, or yna other of the T cell lines tested.
any dose-related clinical signs of EAE. All animals began to
lose weight 4 d post T cell transfer and exhibited a transient
loss of tail tone that lasted a further 3-4 d. After this time ical hallmarks of MBP-induced EAE, were absent. Similar
the animals began to recover weight (Fig. 2). No other clin- losses of weight and tail tone were also observed in animals
ical signs of disease were noted, in particular hind limb parapa- injected with 701 T cell blasts obtained from the S100fl-
resis, paralysis, and urinary incontinence, which are the clin- specific T cell lines SP-D3, S1-D1, and $1-D2. The adoptive
822 Experimental Autoimmune sitilahpecnenaP and Uveoretinitis
It b
20
01 4
0 3
-I0 2
-20 -'
Figure 2. ehT lacinilc esruoc of esaesid decudni
by eht adoptive refsnart of -negitnaotua-SNC
, , I I A , ,
cificeps T llec .senil evitatneserpeR stluser deniatbo
C d
after eht evitpoda refsnart of )a( desolc ,selcric 701
20
ylhserf detavitca cificeps-~001S T sUec enil( ;)1SL
nepo ,selcric 701 ylhserf detavitca ciiiceps-DPP T
;sllec )b( 01 ~ ylhserf detavitca 3D-PS T line sllec
cificeps rof eht citehtnys B001S editpep D3, onima
dica seudiser 76-91; )c( 701 freshly detavitca
cificeps-B001S T sllec enil( )1SL in noitanibmoc
-10 ~ / 2
with na suonevartni noitcejni of 4 gm of eht bAm
D
5C81-8 5 d retfa T llec refsnart .)worra( eud ot eht o
w
i , ! I i i i i I t l l re fytsinreavrest slaminaof the esaesidwere desufrep decudni on after yad ;8the )d( bAm 701 nloade
S I0 DAYS POST T CELL TRANSFER $ 10 . y)lh1sePrfC detavitca cificeps-PBM T sllec T( llec enil d from h
ttp
://ru
p
ptroanosrfelr y if of at 701 all $1-G7 to the or peptide $1-G9 D3 T call (Table blasts, 2) failed which to respond induce atnhd e S10m0e3s-sepnecciefipch alon, T both cell-mediated of which are disease strongly (Table involved 6). This in ress.org
any clinical changes in naive recipients. Controls injected with S1003-specific T cell-mediated inflammatory disease of the /jem
/a
tohfe weight, same doses or tail of tone, PPD-specific whereas Lewis T cell rats blasts injected showed with no com- loss cCephNalSiti s thus (EAP). qualifies as an experimental autoimmune panen- rticle-p
d
parable numbers of MBP-specific T line cells developed se- The Clinical Course and Histopathology of cificeps-pOOlS T f/1
8
0
vere clinical signs of acute EAE by day 4 (Fig. 2). Cell-mediated CNS Inflammation Can Be Modulated yb the /3
/8
Rats were perfused 6 d after T cell transfer and the CNS refsnartoC of MOG-specific Antibody. In view of the exten- 17
/1
(brain, spinal cord, optic nerve), peripheral nervous system sive inflammation seen in animals receiving S100/5-specific 39
6
(PNS) (nerve roots, sciatic nerve, trigeminal nerve), eyes, heart, T cell lines it was of interest to determine whether these clin- 025
lungs, tongue, thymus, kidney, spleen, bowel, adrenal glands, ically silent lesions could be modulated by cotransfer of the /817
and salivary glands removed for histopathology. mAb 8-18C5. This mAb recognizes an epitope of the MOG .pdf b
Surprisingly, despite the absence of obvious signs of CNS exposed on the surface of the myelin membrane. Intravenous y g
u
ldiinseeas e $1-G9), this analysis all the revealeS100B-specific d that, with T cell one lines exception induced (T a cell se- iennjheacntcieosn of the this clinical antibody severity in animals of the with disease MBP-mediated and induces ElAoEca l est on
2
vere inflammatory response in the CNS (Fig. 3, a, c, d, and demyelination (41, 42). 0 Ja
J). Inflammatory infiltrates were also seen in the peripheral In the present study, intravenous injection of 4 mg of nua
nerve roots in 60% of animals, but only very rarely in the purified 8-18C5 5 d after the transfer of S100/3-specific T cells ry 2
0
sciatic nerve. In contrast, uveitis was very common, occur- rapidly induced the classical signs of EAE, paraparesis/paral- 23
ring in 49 of the 60 animals examined and was in some cases ysis within 24 h (Fig. 2). This was associated with the for-
associated with scleritis, iridocyclitis, and periphlebitis retinae mation of large plaques of demyelination throughout the CNS
(Fig. 3, g, h, and i). Uveitis was not observed in any animal (Fig. 3 e). Injection of the same dose of a control mouse IgG
with MBP-induced EAE. No inflammatory responses were preparation failed to induce any enhancement in either din-
seen in any other of the tissues examined. ical status or demyelination in animals with S100B-induced
The intensity of the inflammatory response induced in the CNS inflammation. The intense inflammatory response seen
CNS by the S100~-spedfic T cell lines was qualitatively similar in the CNS of animals injected with S100~-specific T cell
to that seen in animals with severe (grade 2+), EAE induced lines is therefore associated with an enhanced permeability
by a MBP-specific T cell line (Figure 3, a and b). However, of the BBB sufficient to permit the 8-18C5 mAb to enter
unlike MBP-induced disease, in which the inflammatory le- the CNS and initiate demyelination.
sions are most prominent in the spinal cord and rare in brain, The Failure of cificeps-pOOlS T Cells ot Induce Severe Clinical
S100~-specific T cell lines induced an inflammatory response esaesiD Is Associated with a desaerceD Activation of ED1 + Mac-
throughout the CNS, in which involvement of both brain segahpor in the CNS. To investigate the cellular/physiolog-
and spinal cord was common (Fig. 3; Table 6). The differ- ical basis responsible for the inability of S100B-specific T cell
ence between the two models was most marked in the cortex lines to induce severe neurological disease, we examined the
823 Kojima et .la
D
o
w
n
lo
a
d
e
d
fro
m
h
ttp
://ru
p
re
ss.o
rg
/je
m
/a
rticle
-p
d
f/1
8
0
/3
/8
1
7
/1
3
9
6
0
2
5
/8
1
7
.p
d
f b
y g
u
e
st o
n
2
0
Ja
n
u
a
ry 2
0
2
3
824 Experimental Autoimmune and Panencephalitis Uveoretinitis
Table 6. Topographic Distribution of eht Inflammatory activation in macrophages was lower in rats with S100B-
Infiltrates in MBP- and SlOOfl-specific T Cell-mediated Disease induced CNS inflammation. ED1 § cells are the major com-
ponent of the classical EAE lesion (Table 7), where they ac-
T cell line specificity count for up to 80% of the infiltrating inflammatory cells,
both in the perivascular space and parenchyma. This is not
Tissue MBP Sl00~/
the case in S100B-specific T cell mediated CNS inflamma-
tion. Although the absolute number of ED1 § cells present
Cortex* 0.004 +_ 0.002 0.746 + 0.088 in the CNS is approximately equal in both MBP- and S100~-
White matter* 0.107 +_ 0.053 0.310 _+ 0.050 specific paradigms, the proportion of this cell type entering
Diencephalon* 0.209 +_ 0.055 0.965 +_ 0.200 the parenchyma is low (Table 7). Moreover, the intensity of
Mesencephalon* 0.499 +_ 0.109 1.500 + 0.064 staining with the ED1 mAb appeared weaker in the $100~
paradigm, suggesting that these cells are poorly activated in
Cerebellum 0.571 + 0.187 0.235 + 0.043
this model. To ensure that this observation was not a staining
Medulla 1.711 +_ 0.380 1.355 + 0.228
artifact, macrophage infiltration was also evaluated using ad-
Spinal Cord 2.040 2.990
ditional markers for rat macrophage populations on fresh
frozen sections. The macrophage-specific mAbs ED3 and
Lewis rats were injected intravenously with activated cificeps-PBM (CP1; TRPM3 also demonstrated that in S100/~-mediated EAP, the
n = 7), or cificeps-3/001S ;1SL( n = )6 T line cells dna perfused 6 d number of activated macrophages present in the perivascular D
o
later sa described in the text. The number of inflammatory cuffs and parenchymal infiltrates was lower than that seen wn
infihrates/mm 2 determined in multiple tissue sections. The distribution in MBP-mediated EAE. In contrast, the mAb ED2, which load
of the lesions si strikingly different in the two models, particularly with ed
respect to the cortex, white matter, diencephalon, dna nolahpecnenesem stains a population of resident perivascular cells in the CNS, fro
m
that are intensively involved in the $1003 paradigm, but only exhibit was found to identify a similar number of cells in both MBP- h
ecnacifmoderate ingis inflammation at 98% in following analysis of the variance). transfer of cificeps-PBM T cells *( apnhde notype S100~/-mediated of ramified diseases. microglia Similarly, in the parenchyma no difference was ifn ound the ttp://rup
re
between the two experimental models. ss.o
composition and distribution of the infiltrating inflamma- Astrocyte Interactions with SIOOff-specific T Cell Lines In rg/je
tory cells in the CNS. This data was compared with that Vitra The cytotoxicity of several pathogenic CD4 + S100/3- m/a
obtained from studies on MBP-specific T cell-mediated EAE. specific T cell lines was assessed in vitro using syngeneic as- rticle
Even after hematotoxylin-eosin staining it was apparent trocytes as targets in a chromium release assay. The lewis -pd
that in the S100~-specific T cell induced disease model, the rat MBP-specific T cell line Z85 (40) and PPD-specific T cell f/18
0
perivascular lesions were very large as compared with those line LP-PPD1 (37) were used as positive and negative con- /3/8
1
seen in MBP-specific T cell-induced EAE. Large numbers trols for T cell mediated cytotoxicity. In agreement with 7/1
3
of cells were tightly packed into clearly delineated perivas- previous studies the Z85 T cell line effectively lysed IFN- 96
0
2
cular cuffs of inflammatory cells. Quantitative immuno- induced, Ia + astrocytes in the presence of its specific antigen, 5
/8
histochemical analysis revealed that the absolute number of MBP (40). In striking contrast, the S1008-specific T cell lines 17
.p
W3/13 + T cells present in the CNS on day 6 was at least LS1, S1-D2 and SP-D3, which induce an inflammatory re- df b
10 x greater in both the perivascular and parenchymal com- sponse in the CNS in vivo, are completely unable to lyse Ia § y g
u
e
partments of animals infected with the S100/3-specific T cell astrocytes in vitro in the presence of S100B (Table 8). Thus, st o
lines, than in MBP-induced EAE (Table 7). unlike classic encephalitogenic MBP-specific T cells (40), n 2
0
However, in contrast to the numbers of T cells found within pathogenic CD4 + S100~-specific T cell lines are not cyto- Ja
n
the CNS, the expression of ED1 as a marker for lysosomal toxic. Moreover, although the astrocytes used in these studies ua
ry 2
0
2
3
Figure 3. Inflammatory snoisel in the rat brain after the adoptive transfer of -3001S or cificeps-PBM T cell lines. )a( ralucsavireP inflammatory setartlifni
are present in the meninges and spinal cord tissue )sworra( 6 d after the adoptive transfer of 701 cificeps-3001S T line .sllec cificeps-~/001S immunocyto-
chemistry )nworb( slaever that $100~ si confined to astrocytes in the white and grey matter, lmmunocytochemistry for $100~, nuclear staining with
hematoxylin; x70. )b( Inflammatory infiltrates in the spinal cord )sworra( 6 d after the adoptive transfer of 5 x 601 cificeps-PBM T line cells. MBP-
cificeps immunocytochemistry )nworb( slaever that PBM si present in the myelin sheaths of both the white dna grey matter. Note however the distribution
of PBM between white and grey matter si the esrever of that seen for $1003 in a. Immunocytochemistry for MBP, nuclear staining with hematoxylin;
x .07 )c( After the adoptive transfer of cificeps-B001S T cells the majority of the sllec in the inflammatory infiltrates era T .sllec Immunocytochemistry
with W3/13; x360. )d( Adjacent serial section to that in c stained with 1DE to identify macrophages in the lesion, note only very few 1DE + cells
era present in the infiltrate; x360. )e( Co-transfer of cificeps-3001S T cells followed 5 d later by an intravenous injection of monoclonal MOG-specific
antibody. Injection of the antibody induces the formation of extensive confluent plaques of demyelination in the spinal cord white matter. Klfiver
myelin stain; x70. D( Inflammatory infiltrate in the frontal cortex induced by the adoptive transfer of cificeps-3001S T .sllec $100~ immunoreactivity
)nworb( si confined to astrocytes. Immunocytochemistry for $1003, nuclear staining with hematoxylin; x 250. )g( The adoptive transfer of cificeps-B001S
T cells induces a pefivenous inflammatory infiltrate in the retina (periphlebitis retinae); hematoxylin-eosin; x .571 )h( Extensive inflammation in the
uvea dna sclera of the eye induced by the adoptive transfer of cificeps-3001S T cells. Immunocytochemistry for $1003 slaever the presence of ~/001$
in many cells throughout the tissue. Immunocytochemistry for $1003, nuclear staining with hematoxylin; x250. )i( Extensive inflammation in the
iris induced by the adoptive transfer of cificeps-~001S T cells, iris stroma cells era immunoreactive for .~001S Immunocytochemistry for S100~, nuclear
staining with hematoxylin; x .052
825 Kojima et .la
Table 7. Quantitation of Macrophage and T Cell Infiltration into eht CNS
ralucsavireP Parenchym
T cell
specificity ED1 W3/13 Ratio* ED1 W3/13 Ratio
MBP 982 (31) 83 (41) 7.6 308 (52) 921 (4) 6.2
B001S 700 (126) 1,158 )6( 0.6 514 (40) 1,173 )3( 0.4
w3/13 + dna 1DE + sllec were detatitnauq ni multiple lanips cord snoitces sa debircsed ni eht text.
were $100~ § we were unable to demonstrate the presenta- T cell lines were selected for reactivity against bovine $100~
tion of autologous $100~/by astrocytes to syngeneic T cell protein. Epitope mapping using synthetic peptides representing
lines in either proliferation or cytotoxicity assays. rat $100~ sequences proved that most of these T cell lines
are truly self-reactive. This was most dearly seen in the T
cell lines LS1 and LS2 derived from the bulk lymph node
Discussion cell cultures, in which the proliferative response to the COOH- D
o
w
This study demonstrates that in the Lewis rat, autoim- terminal rat $100~ peptide, amino acids 76-91, was virtually n
lo
a
mune T cells specific for an astrocyte-derived autoantigen, equivalent to that obtained with the native bovine protein. d
e
d
S100B, adoptively transfer an unusual type of inflammatory This epitope appears to be immunodominant in the bulk- fro
m
CNS disease, which can be best described sa EAP. Clinical derived S100B-specific T cell lines and its pathogenicity was h
disease and histopathology in this new model of tissue specific confirmed using a T cell line generated using a synthetic rat ttp://ru
autoimmunity differ dramatically from that seen in "classical" 19-67~001$ peptide. In contrast to the MBP-specific T cell re- p
re
MBP-induced EAE, and in some respects may resemble more sponse, recognition of this immunodominant $1003 epitope ss.o
closely the human disease MS. was not however found to be associated with preferential usage rg/je
In their basic characteristics, autoaggressive $100~/- and of the TCR V~8.2 gene. m/a
MBP-specific T cell lines are indistinguishable; both express Adoptive transfer experiments established that this S100~- rticle
the CD4+CD8 - membrane phenotype and use the TCR- specific T cell response is highly pathogenic. However the -pd
ol/~; their antigen specific responses are MHC class II disease induced by S100~-speeific T cell lines differs radically f/18
0
(RT1.B )1 restricted, and, in addition, both MBP- and from that produced by MBP-specific T cells. Despite the in- /3/8
1
S100B-specific T cell lines are Thl-like, synthesizing IL-2 and tense inflammatory response in the CNS, clinical signs of neu- 7/1
3
IFN-3' upon activation. For practical reasons, the cificeps-B001S rological dysfunction were minimal. Moreover, the histo- 96
0
2
5
/8
1
7
.p
d
Table 8. Cytotoxicity of S100fl- or 391 cificeps-editpeP T Cell Lines f by g
u
e
Effector/target (E/T) ratio st o
n
2
0
Line A 1:01 3:1 1:1 0.3:1 0:1 Ja
n
u
a
ry 2
1SL + 0 0 0 0 0 02
3
- 0 0 0 0 0
$1-D2 + 0 0 0 0 0
- 0 0 0 0 0
LS-D3 + 0 0 0 0 0
- 0 0 0 0 0
Z85" + 31.7 + 3.5 6.71 + 6.1 8.7 + 2.5 0 0
- 0 0 0 0 0
LP-PPD1 + 0 0 0 0 0
- 0 0 0 0 0
setycortsA were dedees sa target sllec ni llew-69 mottob-taif setalp ta a ytisned of 401 .llew/sllec After a 48-h noitabucni with IFN- 001( U/ml)
to induce aI ,noisserpxe eht sllec were delebal with rClS 1( #Ci/10 4 cells), dna ylhserf detavitca T stsalbohpmyl dedda ni tnereffid E/T ratios ni
eht ecneserp or ecnesba of eht etairporppa negitna enivob( ,3001$ aeniug gip ,PBM PPD ta 01 #g/ml). sisyL saw denimreted gnisu etacilpirt selpmas
retfa a 12-h .noitabucni The results era given sa eht mean cificeps sisyl +_ .DS
628 Experimental enummiotuA sitilahpecnenaP dna sitiniteroevU