- Email: [email protected]
Parenteral Administration of an Activating Monoclonal Antibody to the α1β1 Integrin in Dogs
Immunobiology (2000) 202, pp. 239-253 © 2000 Urban & Fischer Verlag http://www.urbanfischer.de/journals/immunobiol
1 Department of Medicine F, 2 Laboratory for Immunoregulation, 3 Hematology Institute, 4 Department of Pathology, 5Neuroimmunology Unit, and 6Neufeld Cardiac Research Institute, Chaim Sheba Medical Center, Tel Hashomer, Sackler Faculty of Medicine Tel Aviv University, Israel, and 7 School of Biological Sciences, University of Manchester U.K.
Parenteral Administration of an Activating Monoclonal Antibody to the a 1~ 1 Integrin in Dogs 1
3
4
ILAN BANK 1, 2, IZHAR HARDAN , EVGENIA LOKSHIN , DEVORA NAS , 5
6
7
SHMUEL MIRON , DANIEL OHAD , SUZANNE SPONG , and DAVID R.GARROD
7
Received September 22, 1999 . Accepted in revised form March 22, 2000
Abstract In mice, monoclonal antibody (mAb) to the al integrin abrogate gastro-intestinal damage during graft-versus-host-disease (GVHD), suggesting anti al mAb as candidates for treatment in humans as well. Our current data show that one such reagent, mAb 1B3.1, when immobilized to plastic wells via rabbit- anti murine (ram) immunoglobulin (Ig) induces a protein kinase-dependent spreading of activated human T cells. Furthermore, it significantly increases the proliferative response, and expression of interleukin-2 (IL-2) receptors (R) and CD69, of resting T cells, expressing minimal integrin on the cell surface, to sub-optimal stimulation by anti-C03 mAb. We found, in addition, that mAb IB3.1 a) immuno-precipitates a 1~ 1 integrins from cell-surface iodinated canine epithelial cells b) is highly reactive with canine T cells after their activation and c) inhibits adhesion of canine T cells to collagen IV: Despite the potential ability of the mAb to co-activate T cells in vitro, two dogs that received 4 injections of 0.5-0.3 mg/Kg of mAb 1B3.1 remained healthy, developing only marginal transient lymphopenia. Injection of 0.75mg/Kg in a third dog induced a more marked lymphopenia, and an additional dose of 1.0 mg/Kg 2 weeks later was followed by gastrointestinal hemorrhage. Importantly, the lymphopenia was associated with a greater and more persistent decrease of C08+ than of C04+ T cells, leading to an increase in the C04/C08 ratio 24 hours after the injection. Thus, despite it's co-activating effects in vitro, administration of this mAb in vivo is feasible when appropriately dosed, and may have immuno-modulatory effects.
Abbreviations: BSA = bovine serum albumin, ECM = extracellular matrix, OMEM = Oulbecco's
Modification of Eagles Medium, FCS = fetal calf serum, FM = final medium, FITC = fluoresceine isothiocyanate, FACS = fluorescence activated cell sorter, gam = goat anti-mouse, GVHO = graft-versushost-disease, Ig = immunoglobulin, IL-2 = interleukin-2, IU = international units, IMOM = Iscoves modified Oulbecco medium, MOCK = Madin Darby Canine Kidney, mAb = monoclonal antibody, NAC = non-adherent cells, 00 = optical density, PBMC = peripheral blood mononuclear cells, PBS = phosphate-buffered saline, PHA = phytohemagglutinin, ram = rabbit-anti-murine, VLA = very late antigen, SOS-PAGE = sodium-dodecyl sulfate polyacrylamide gel-electrophoresis, TBS = tris-buffered saline. 0171-2985/00/202/03-239 $ 15.00/0
240 . 1.
BANK
et al.
Introduction Six of the ~ 1 integrins, consisting of unique a peptides (a l-a6) in non- covalent association with a common ~1 peptide, are expressed on leukocytes (1, 2). T cell ~1 integrins have been studied almost exclusively in humans and mice (2-5). Resting T lymphocytes express high levels of the a4~1 and a5~1 fibronectin receptors (also known as the very late antigens (VLA) 4 and 5), and of a6~1, a laminin receptor, whose adhesive functions are markedly and transiently enhanced by activation with phorbol-esters, or triggering of the T cell receptor (TCR)-CD3 complex by antigen or mAb ("inside-out" signalling) (3). On the other hand, cross linking ofT cell surface membranes with extracellular matrix (ECM) proteins or mAb directed against VLA-4, 5 or 6 in the presence of a sub-mitogenic concentration of mAb to CD3 augments T cell proliferation ("outside-in" signalling) (4). The VLA-l (al~l) and VLA-2 (a2~1) heterodimers, receptors for collagens and laminin, are found at low levels on human and murine resting T cells (2, 5-7). Increased expression of these integrins occurs on the surface membrane after relatively prolonged activation periods and they are used by activated human and murine T cells to adhere and spread on collagens (2, 5-10). A recent study demonstrated that intraperitoneal (IP) administration of mAb to the a 1~ 1 integrin prevents the mucosal damage of murine GVHD, which suggests that anti al integrin mAb could be used in larger mammals as well for the therapy of this disease (11). We have addressed this possibility by employing a mAb specific for human a 1 integrins which was found previously, on the one hand to block adhesion of T cells to collagen, and on the other-to activate certain T cells (8, 9). In these studies we further analyzed the activation of T cells by the mAb and, using novel data concerning cross-reactivity of this mAb with the canine integrin, report
for the first time preliminary studies of effects of it's parenteral administration.
Materials and Methods Peripheral blood mononuclear cells (PBMC) and
T cell
cultures
PBMC were isolated from normal consenting volunteers and from dogs by Ficoll-Hypaque density centrifugation (12). Dog PBMC were activated with 1% phytohemagglutinin (PHA) (Biological Industries, Beit Haemek, Israel) then expanded and passaged in 24-well tissue culture plates (Costar Corporation, Cambridge, MA, USA) in Iscoves modified Dulbecco medium (IMDM) supplemented with 100/0 heat-inactivated fetal calf serum (FCS), 1% glutamine, and 1% penicillin streptomycin solution (Biological Industries, Beit Haemek, Israel) (final medium (FM)) and 100 international units (IU)/ml of recombinant interleukin 2 (IL-2), (Proleukin, Chiron Corp., Amsterdam, the Netherlands) (8). For isolation of resting T cells from the PBMC macrophages were first removed by incubating PBMC in tissue culture treated plastic dishes (Costar corporation Cambridge, MA, USA) for 1 hour in a 60/0 CO 2 humidified 37°C incubator in FM and the non-adherent cells (NAC) were harvested. To remove residual macrophages, B cells and in vivo activated T cells, the NAC were treated with anti HLA-DR mAb OKI2 for 30 minutes, washed X3 in phosphate-buffered saline (PBS) pH = 7.2, then panned on culture plates coated with 10 flg/ml of affinity purified ram-Ig diluted in Tris-buffered saline (TBS), pH = 9.5 as previously described (13). The NAC population contained >980/0 CD3+HLA-DR- lymphocytes (not shown).
VLA-1 expression and function in dogs .
241
MAb MAbs 1B3.1 (IgA, anti-a1B1 integrin), OKT11 (IgG1, anti C02), OKT3 (IgG2, anti C03), and OKI2 (IgG 1, anti HLA-OR) were prepared from ascites fluid from Balb/C mice bearing the respective hybridomas (12, 14). TEPC-15 (IgA, anti phosphorylcholine) was purchased from Sigma Chemical Co. St. Louis, MO, USA (12). Mabs OKT3 and 1B3.1 were further purified by affinity chromatography and dialyzed against PBS. MAb AIIb2 against 1 integrin chains was kindly supplied by Dr. C. DAMSKY (14). MAb MOPC-167 (irrelevant murine IgA) was a gift from Prot Z. ESHAR, Weizmann Institute, Rehovoth, Israel. MAb to canine CD18, CD45RA, CD4 and CD8 were kindly supplied by Dr. ~ MOORE, University of California, Davis, USA (15,16). Directly conjugated mAb to IL-2R and CD69 (Leu 23) were obtained from Dako, Denmark, and Becton Dickinson, San Jose, CA, USA, respectively.
B
Fluorescence-activated cell sorter (FACS) analysis
2 x 10 5 cells were stained with saturating concentrations of mAb at 4°C, washed, then incubated with goat anti-mouse (gam)-Ig conjugated to fluoresceine iso-thiocyanate (FITC) (gam-Ig-FITC) (Sigma) as previously described (8). The level of expression of alB 1 integrin quantitated from data supplied by Coulter FACS software is expressed as mean fluorescence intensity (MFI, arbitrary units) of analyzed cells. Dogs
The study was approved by the Tel Aviv University Helsinki committee. Male mongrel dogs were housed in the Sackler Medical School facility for medical research. Dogs were sedated by intramuscular injection of a short acting barbiturate. MAb were administered into a peripheral vein after extensive dialysis into PBS and 0.45 ~m sterile filtering. Hematological parameters
Blood was drawn into calcium citrate containing tubes and hematological parameters evaluated using standard clinical Coulter counters. Values were validated in some experiments by direct visualization of stained blood smears. Adhesion assays
Assays were performed as previously described (17,18). Dissolved collagen IV (Sigma) or fibronectin were diluted in H 2 0 to 10 ~g/ml and 50 ~l added to 96 micro-wells tissue culture plates (Costar), which were air dried at room temperature overnight. The ECM protein coated wells and control non coated wells were blocked with 50 ~l of 0.20/0 bovine serum albumin (BSA) (Sigma) in PBS for 30 minutes, then washed X3 with PBS. Cultured cells were washed twice in fresh IMDM by centrifugation at 1000 r.p.m, and resuspended in IMDM. 150,000 cells/well were added to triplicate wells, in a final volume of 200 ~l/well of IMDM. For inhibition studies, mAb (final concentration 1:200 of ascites fluid) were added to the medium. The plates were placed in a humidified atmosphere at 37°C, 6% CO 2 , for 40 minutes, and non-adherent cells discarded. Adherent cells were fixed in 700/0 ethanol for 10 minutes, and stained with 0.1 % crystal violet solution in H 2 0 for 20 minutes. Plates were washed extensively in H 2 0, and 100 ~l of 0.1 % Triton X-I 00 solution added for 30 minutes to release the dye. The optical density (OD) at 550 nm was recorded using an automated ELISA plate reader. Direct counting of adherent cells established that OD was proportional to the number of adhering cells (18). Results are reported as mean OD of triplicate wells containing ECM protein minus mean OD of triplicate control wells coated with BSA alone. Proliferation assays
Flat bottom ninety-six well tissue culture plates (Costar) were coated overnight at room temperature with 10 ~g/ml of affinity purified ram-Ig diluted in TBS, pH = 9.5, then washed X3 in PBS
242 . 1.
BANK et
al.
(12). The mAbs diluted in PBS were added to the ram-Ig coated wells for 2 hours at 4°C which were then washed X3 with PBS. 5 x 10 4 cells/well were added, and plates placed in a 60/0 CO 2 970/0 humidified incubator at 37°C. The cultures were pulsed overnight for 12 hours with 1 mC/well of 3 [H] thymidine (Amersham) after 60 hours of incubation then harvested onto nylon mesh filter paper. Incorporation of thymidine was measured in a ~-counter after placing the dried filters in vials containing scintillation fluid as previously described (12). All experiments were performed in triplicate.
Induction of IL-2R and CD69 Cells were cultured as described above for 24 hours, then washed X2 in PBS and stained with mAb to CD25 or CD69 that were directly conjugated to FITC.
Spread ing assays 24 well tissue culture plates (Costar) were coated overnight at room temperature with 10 J,Lg/ml of affinity purified ram-Ig diluted in TBS pH=9.5 then washed 3x in PBS (12). MAb (1:200 of ascites in PBS) were added for 2 h at 4°C and the wells were then washed 3 x with PBS. T cells removed from growing cultures were placed in the CO 2 incubator in FM alone or FM with 100 ng/ml staurosporine for 1 hour then washed 3 x in IMDM prior to layering in the wells (19). 5 x 10 5 T cells in 500 J,Llof FM were added in each well and plates placed in the CO 2 humidified incubator for 3 h, after which wells were rinsed with IMDM to remove non-adhering cells. The number of adhering cells per inverted phase microscope field, exhibiting elongated and or spherical (round) morphology was recorded (9).
Immunoprecipitation Madin Darby Canine Kidney (MDCK) cells (European Collection of Animal Cell Cultures, Porton Down, Salisbury, U. K.) were grown in Dulbecco's Modification of Eagles Medium (DMEM) supplemented with 10% (v/v) FCS or Newborn Calf Serum, penicillin (100u/ml)) and streptomycin 1OOllg/ml at 37°C in 5% CO 2 , Confluent cells were trypsinized and seeded into 9 em petri dishes (1 x 10 7 / dish) precoated with a layer of collagen gel. The collagen substratum was digested with 0.1 % (w/v) collagenase in DMEM. Cells released from the gel were washed in ice cold PBS and resuspended to 1.5 - 2 x 10 8 cells/ml in PBS. Two hundred microliters of lactoperoxidase (lmg/ml in PBS), 1.7-2 mCi Na 125 1 (lCN Pharmaceuticals Ltd., Thame, Oxon. or Amersham International pIc., Little Chalfont, Bucks, U.K.) in 0.2 ml PBS and 2 III of 0.12% (v/v) H 2 0 2 in PBS were added to 1 ml of cells. After a 5 minute incubation on ice, 20 III of 0.12% H 2 0 2 were added for an additional 5 minutes. This step was repeated and the reaction stopped by adding 10 ml of PBS (20). Cells were then washed and lysed in 2 ml of extraction buffer (1 % (v/v)Triton X-100, 10 mM TrisHCl pH 7.4, 0.13 M NaCl, 5 J,Lg/ml BSA, 1mM MgCI 2 , 1mM MnCI 2 , 10 J,L/mlleupeptin and 1 mM phenylmethylsulfonyl fluoride) on ice for 30 minutes. Lysates were centrifuged (3,000 x g for 5 minutes) and supernatants passed down a PD-10 column (Pharmacia Biotech, St Albans, Herts, U.K.). Labelled proteins were eluted in 3.5ml of extraction buffer and extracts precleared by adding 10% (v/v) Protein G-Sepharose and mixing for 1.5 hrs at room temperature. Beads were removed by centrifugation (14,000 x g for 10 minutes) and the cleared extract added to an equal volume of PBS+ containing 2.5 mg/ml BSA and immune Igs or control mouse IgG, and rotated overnight at 4°C. Immunoprecipitates were collected by the addition of 20 J,Ll protein GSepharose/tube and rotating for a further 1.5 hr at 4°C. Beads were harvested by centrifugation (7,000xg for 3 min), washed 3 times in extraction buffer and twice in PBS+, and resuspended in sodium-dodecyl sulfate polyacrylamide gel-electrophoresis (SDS-PAGE) sample buffer (with or without mercaptoethanol) and boiled for 3 min. Immunoprecipitated proteins were separated by SDS-PAGE using the method of LAEMMLI (21). Gels were soaked in 100/0 (v/v) acetic acid, 250/0 (v/v) methanol, dried on a BIO-RAD gel drier, and exposed to x-ray film or analyzed using a Molecular Dynamics phosphorimager. Digitized images were obtained using a Fujix Bas 2000 Bio imaging analyser.
VLA-I expression and function in dogs . 243
Results a 1~ 1 integrin-mediated
T cell activation
It was previously shown that a proliferative response is induced in cultured or in vivo activated but non-proliferating T cells expressing high levels of the a 1~ 1 integrin, by cross linking the cells with solid phase bound mAb 1B3.1 in the presence of a non-mitogenic dose of anti-CD3 mAb (12). Since the goal of our current study was to evaluate possible effects of the mAb on T cells in vivo, we extended these observations by studying the effects of the mAb on freshly isolated and highly purified resting T cells from normal volunteers, which expressed a very low level of VLA-1 on the cell surface (MFI of cells stained with mAb 1B3.1=1.16, control MFI=1.05, MFI of cells stained with mAb TEPC-15 =1.16, not shown) (12). When such cells were coated with mAb 1B3.1, placed in culture wells after washing of excess mAb, and triggered with a previously determined minimally mitogenic concentration of mAb OKT3, their proliferative response was not significantly different from that of non-treated or isotype control (mAb TEPC-15, anti-phosphoryl choline) treated cells, in marked contrast to the augmented proliferation of cells that were pre-treated with mAb OKT11 (directed against the costimulatory CD2 molecule) (Fig. 1A). However, when the resting T cells were placed in culture in micro-wells containing immobilized mAb 1B3.1, a markedly enhanced proliferation occurred relative to that of cells cultured in wells coated with the control mAb TEPC-15 or with no cross linking mAb (Fig. 1B). Furthermore, the enhancement of proliferation induced by cross linking VLA-1 molecules with immobilized anti-integrin mAb was dependent upon the concentration of mAb 1B3.1 used to coat the wells (Fig. 1C). In addition, cells activated by mAb 0 KT3 in the presence of cross linking of the VLA-1 integrin, markedly upregulated the surface expression of IL-2R and of CD69 on the cell surface (Fig. 1D). These results suggest that prolonged and efficient crosslinking of sparsely expressed VLA-1 molecules on the surface membrane of resting T cells significantly augments CD 3 mediated activation signals in previously resting T cells. Protein-kinase-dependent signalling
by VLA-l
To further investigate the nature of "outside-in" signals transmitted by a 1~ 1 integrins, VLA-1 + T cells cultured in IL-2 containing medium were placed on plastic wells coated with ram-Ig alone, or with ram-Ig to which mAb 1B3.1, OKT11 or the non reactive control mAb MOPC-167 had been added. Only wells coated with mAb 1B3.1 and OKT11 contained adherent cells after three h of incubation (Fig. 2A). Cells adherent to OKT11 coated wells had a predominantly spherical morphology, whereas cells adherent to mAb 1B3.1 were either spherical or spread (elongated). The T cells were next incubated for 1 h in either medium, or in medium containing 100 ng/ml of staurosporine, an inhibitor of protein kinase activity, prior to seeding in the wells (19). Adherent elongated cells, but not spherical cells, were significantly decreased in wells containing cells pretreated with staurosporine (Figs. 2B 1 and 2B2). Thus, T cell spreading induced by cross linking of the alB 1 integrin requires the activation of intracellular proteinkinases.
244 .
I. BANK et al.
7ססoo
RAM
eoooo
RAM+OKT3
50000
~~j
RAM+ 183.1+OKT3
O:J
RAM+TEPC+OKT3
1~j oj
RAM+OKT11 +OKT3
a
20000
40000
CPM
60000
---_._..
medUn+OKT3
1B3.1.0KT3
-------
60000
60 C
D
50
50000
~ 40000
j
30000
j
20000
j
10000
rnedIlm
.!/}. 1)
~ 40
a
••
111600
11800 11400 Dilution of ascites
11200
~ ~30
j j
•
Il-2R
~
CD69
11 20
1: j
.....
.-....om
TlPC40KT3
Figure 1. Enhanced activation in response to immobilized anti VLA-l mAb. A. Resting human T cells in suspension were coated with no mAb, with mAb IB3.1, TEPC, or OKTll and excess mAb washed away. The cells were then cultured (5 X 104/well) in triplicate wells that were pre-coated with ram-Ig, and triggered with 10 ng/ml of anti-CD3 mAb. 3[H] thymidine incorporation was measured after 72 hours of culture. Results represent the mean CPM of triplicate experiments B. Non pre-treated T cells were added to culture wells that had been precoated with ram-Ig alone (medium) or with 1/200 of ascites fluid ofmAbs IB3.1 orTEPC-15, then triggered with 10 ng/ml ofanti-CD3 mAb. C. Dose response of the mitogenic effect of increasing concentrations of immobilized mAb 1B3.1 (squares) or TEPC-15 (black circles) using the same experimental protocol. D. T cells activated as in B were removed from culture wells after 24 hours, stained with FITC conjugated mAb to IL-2R or CD69 and analyzed by flow cytometry.
Reactivity of mAb 1B3.1 with canine cells To develop an animal model in which in vivo effects of mAb IB3.1 could be evaluated, we determined the reactivity of this mAb with canine tissues. Immunoprecipitation of cell-surface iodinated molecules of the MDCK line, which, in screening flow-cytometry experiments was found to react with mAb IB3.1 (not shown), using mAb's IB3.1 and AIIB2 were performed (Fig. 3). MAb IB3.1 immuno-precipitated peptides of200 kDa and 110 kDa, corresponding to al and pI integrin chains respectively (lane 2) (8). MAb AIIB2, directed against the 1 integrin peptide, immuno-precipitated both 200 kDa a 1
B
VLA-l expression and function in dogs· 245 A
ram
•
ram +OKT11
~herical
ram + 183.1
elongated
ram + MOPC-167
o
30 60 90 120150
elongated
spherical
cells. 81
rlm+ MOPV1fr7
cells. 82
rlm+ MOPC-1fr7
•
Cf,um+staurospor,ne
rlm+183.1 • • • •JlI
•
Cf,um+staurosporlne
rlm+1B3.1 , - - - - - - - - - . ,
medium
o
100
200
300
400
cenl per field
500
medium
o
20
40
eo
eo
celli per tIIId
100
Figure 2. Inhibition of T cell spreading by staurosporine. A. 10 5 human alp 1+ T cells, previously cultured in IL-2, were allowed to adhere to wells pre-coated with ram-Ig or with ram-Ig to which mAbs IB3.1, OKTll or MOPC-167 had been immobilized. Cells exhibiting spherical or elongated morphology were counted using an inverted phase microscope. Bars represent mean ± 1 std of three fields. B. The cells were allowed to adhere to wells pre-coated with ram- Ig or with ram- Ig to which mAb IB3.1 or mAb MOPC-167 had been immobilized. A parallel aliquot of cells was incubated with staurosporine (100 ng/ml) for 1 hour prior to addition to identical wells. Cells exhibiting spherical (B 1) or elongated morphology (B2) were counted. Bars represent mean ± 1 std of three fields.
and 160 kDa (corresponding to a6 peptides) as well as 110 kDa peptides corresponding to ~ 1 integrin chains (lane 3) (14). These results indicated that mAb 1B3.1 specifically recognizes a 1~ 1 integrins expressed by canine epithelial cells. Next, we investigated patterns of reactivity of mAb 1B3.1 with canine T cells. Flow cytometry of resting and in vitro activated PBMC isolated from 4 mongrel dogs (in 5 experiments) demonstrated that the reactivity of the activated cells with mAb 1B3.1 was consistently and significantly higher, suggesting that expression of the a 1~ 1 integrin is enhanced on the surface of canine T cells after their activation, as previously demonstrated for human T cells (Fig. 4) (12). Finally, adhesion assays of activated dog T cells expressing a 1~ 1 integrins to ECM proteins collagen IV and fibronectin were performed (Materials and Methods). An isotype control mAb (MOPC-167) did not inhibit adhesion to either protein (Fig. 5). In contrast, adhesion of the cells to collagen IV, (but not to fibronec-
246 . I. BANK et al.
1 2 3 200 -
<4I-al
100
Figure 3. Autoradiogram of integrins immuno-precipitated from MOCK cells. SOS-PAGE analysis (non-reduced) of extracts of 25 I] cell surface iodinated MOCK cells immunoprecipitated with non immune mouse Ig (lane 1), with mAb IB3.1 (anti-al integrin) (lane 2), or mAb Allb2 directed against the common ~I integrin chain (lane 3). Location of200 kOa and 100 kOa molecular weight markers run simultaneously are indicated on the left. Expected locations of the indicated integrin peptides are designated on the right. The a6 integrin, immunoprecipitated in lane 3, migrates at 160kOa.
e
tin) was strongly inhibited by mAb 1B3.1. These results indicate that the a 1~ 1 integrin in activated dog T cells, as in humans, functions as a specific receptor for collagen
I\Z Effects of anti VLA-l in vivo Preliminary experiments were conducted to test whether the mAb, shown above to co-activate T cells under specific experimental conditions in vitro, could be safely administered to live dogs (Table 1). Minor gastro-intestinal side effects were noted after injections of 0.3- 0.5 mg/Kg mAb 1B3.1 in dogs 1 and 2. Injections of this dose of mAb 1B3.1 were also followed by a non-significant decrease of lymphocytes in the PB after 6 h, with almost complete recovery occurring after 168 h (Fig. 6a). This transient reduction of lymphocytes was not significantly different from the mild lymphopenia induced by injection of the anti phosporyl-choline control mAb TEPC15. In dog 3, in contrast, into which 0.75 mg/Kg of mAb were administered, a more
VLA-l expression and function in dogs .
1 ~
+oJ
"in
5j 0.8
+oJ
c:
• activated cells
resting cells
---t> p
----[>
g 0.6 Q)
p
oU) o
p
::3
i;:
C
eu
0.2\ =-
E
•
:
'---
o
• ••
------iC>
•• •••
•
• -.-•
-• -
•
Q)
~ 0.4
247
• •
=
p
cont.
TEPe-15
183.1
cont.
'fEPC-IS
IB3.1
Figure 4. Increased al~1 integrin on activated T cells. MFI of 5 samples of resting and dog PBMC activated in vitro for 3 weeks as described in materials methods, after staining with gam-Ig-FITC alone (cont.) or with mAbs TEPC-15 or IB3.1 followed by gam-Ig-FITC. Origin and tips of arrows indicate groups of values the means (horizontal lines) of which were compared using a one tailed Student's t-test. Resulting p values are indicated.
0.06
0.03
0.05
0.025
0.04
0.02
collagen p
ci
·0
:I 0.03
0 0.015 :I
-8 0.02
0.01
0
-8
0.01
0.005
0
0
Figure 5. Inhibition of adhesion to ECM proteins by mAb 1B3.1. Cultured a 1~ 1 integrin expressing canine T cells were allowed to adhere to plastic wells coated with either fibronectin or collagen IV blocked with BSA, or with BSA alone, for 40 min in the presence of medium (black bars), medium containing 1:200 dilution of ascites of mAb's 1B3.1 (white bars) or isotype control mAb MOPC-167 (obliquely hatched). Adherent cells were stained with crystal violet, and dye released with 0.1 % triton. Results, on the Y-axis, are given as delta 00 = 00 of wells coated with ECM and blocked with BSA - 00 of wells with no ECM protein and blocked with BSA. Results are representative of 3 experiments. P values obtained from the Student's t-test, are shown.
profound and delayed lymphopenia was induced, while the neutrophil count was unaffected, but no adverse clinical effect was observed (Fig. 6b). Additional administration of a higher dose (1 mg/Kg) of mAb 1B3.1 two weeks after the first injection, resulted in massive hematemesis after 26 hours and subsequent death of this dog.
248 . 1.
BANK
et al. 140%~------------.
0
130%
0120% Q)
Q)
:§ 100%
E ;; 100%
.9
.9
>
a;
Q)
Q)
.~
li 70%
-~
'B
80°A»
neutrophils
--lymphocytes
~ 60°A» C 400/0
~
c:
~
~ 40%
~ 20%
Q) ~
A
?
10% - ' - - - - + - - - - + - - - - + - - - - + - - - - - ' o 6 24 168 hours after injection
B
o
6 24 168 336 342 hours after injection
Figure 6. Effect of mAb IB3.1 on lymphocyte counts. A. Lymphocytes per cubic mm were enumerated in the PB before (time 0) and after injections of PBS (open squares), 0.5 mg/Kg - 0.3 mg/Kg of mAb IB3.l (black squares, representing mean and 1 std of 3 experiments in dog 1 (Table 1) or 0.5 mg/Kg of mAb TEPC-15 (triangles). Results are presented as percent of the lymphocytes at the time points indicated on the X-axis relative to values at time o. B. Lymphocyte and neutrophil counts after injection of 0.75 mg/Kg in dog 3. A second dose of 1 mg/Kg was given 336 hours after the first injection.
Table l. In vivo administration of reagents to dogs: dosages, intervals, and clinical outcome.
Dog 1
Dog 2
Dog 3
day reagent dose symptom
0 mAb IB3.1 0.5 mg/Kg vomiting
day reagent dose symptom
0 mAb IB3.1 0.5 mg/Kg none
day reagent dose symptom
0 mAb IB3.1 0.75 mg/Kg none
66 mAb IB3.1 0.5 mg/Kg none
123 mAb TEPC-15 0.5 mg/Kg none
141 PBS none
183 mAb IB3.1 0.3 mg/Kg none
7
sacrificed for autopsy 14 mAb IB3.1 1.0 mg/Kg hematemesis after 6 hours, death after 24 hours
Effects of injected mAb on T cell subsets
To further evaluate the effects of parenterally injected mAb, CD18+, CD45RA+, CD4+ and CD8+ lymphocytes were enumerated in the PB. Injection of mAb 1B3.1 resulted in marked reduction in cells expressing these lymphocyte markers after 3-6 hours, whereas injection of the isotype control mAb TEPC-15 had only a marginal effect (Figs. 7A and 7B). Although the percentage ofPB lymphocytes expressing CD45RA, CD18 and
VLA-I expression and function in dogs 8
A
140%
14QOk
.! 1200k
CD18 CDS
-.
'0 600/0 C ~ 4QOA. 8. 200/0 0%
CD4
~
CD45RA
0
...
'0 6OOA.
CDS -.
CD18
180%
-B-
80%
g 100°A.
.! 120%
...
, 1000/0
I
·249
CD4
{40%
~
CD45RA
20%
0%
24 6 hours after injection
-B-
0
24 3 6 hours after injection
C
10
....
8
8
~0
01,1ntlVlA-1
6
D3,1nti-VlA-1 ~
01, TEPC-15
4
~
01,P8S
2
0
0
168 24 hours after injection
Figure 7. Effect of mAb 1B3.1 on T cell subsets: Percent of subsets among the PBMC (relative to time o = 100%) after injection of mAb TEPC-15 (A) or mAb 1B3.1 (B, representative of three experiments). C: CD4/CD8 ratio in PBMC at the indicated time points. Results for dog 1 (01) are representative of three separate experiments, for dog 3 (03) - of a single experiment.
CD4 returned to pre-injection values 24 hours after injection of mAb IB3.1, the percentage of CD8± cells in the PB remained suppressed. As a result, the ratio of CD4 to CD8 cells was markedly increased in the PB 24 hours after the injection (Fig.7c). Evaluations of these T cell subsets 1 week after the injections, revealed that the ratio had returned to the pre-injection values.
Discussion MAb 1B3.1 was previously found to recognize an epitope within the I domain of the human al integrin chain, which contains its collagen and laminin binding sites as well as cation binding sites important for integrin functions, suggesting that this mAb is a potential reagent for treating diseases T cell-mediated diseases, such as GVHD (22). The studies described here, showing that this anti-human al mAb recognizes the same integrin in canine cells, together with the demonstration that the mAb functionally blocks the adhe-
250 . 1.
BANK
et al.
sion of activated canine T cells to collagen I~ indicate that there is an homology of expression and functions of this molecule between human and canine T cells. These results thus suggest, for the first time, that dogs could serve as a model for studying the role of a 1~ 1 integrins in the pathogenesis of diseases (e.g. atherosclerosis, sarcoidosis and rheumatoid arthritis) in which al~l-integrin positive activated T cells are involved (12, 23, 24). Since the a 1~ 1 integrin is broadly distributed on fibroblasts, liver cells, endothelial cells, natural killer (NK) cells and other cellular elements, parenteral administration of a mAb to this molecule could affect many physiological systems (25-29). Furthermore, our results showing that persistent cross-linking of resting T cells, that express very low levels of the integrin, via the a 1 integrin chain using immobilized mAb significantly augments their activation, raised additional serious concerns regarding possible activating effects during repetitive administration of the mAb in vivo, when immune complexes may have formed and settled in tissues. However, our results indicate that intravenous injection of anti-integrin mAb 1B3.1 at up to 0.75 mg/Kg to healthy mongrel dogs is clinically well tolerated, and caused no apparent histo-pathological alterations in liver, kidney, or gut tissues as determined in an autopsy study done on dog 2 (Table I, and data not shown). Moreover, even repeated injections of up to 0.5 mg/Kg, spaced over greater than 1 month intervals, were clinically well tolerated. Such a dose would be expected to yield 5-10 f.1g/ml of the mAb in the dog's serum prior to distribution to extracellular fluids from the vasculature which is equivalent to the saturating concentration of the mAb used for cytofluorometric assays, and for the blocking experiments shown in Figure 5. Thus, our results suggest that short term health is unimpeded by blocking al integrins on circulating PBMC. Interestingly, increasing the dose by only 500/0, to 0.75 mg/Kg, resulted in quite a profound and prolonged decrease of circulating lymphocytes, although with no apparent clinical effect (Figure 6B). This higher concentration of the mAb may have induced a redistribution of lymphocytes among body fluids and tissues by enhancing their ability to avoid adhesion to collagen IV in basement membranes (9, 18). The gastro-intestinal hemorrhage and subsequent death of this dog after a second injection of an even higher dose of the mAb, may have been due to toxicity of the mAb to cells expressing the a 1~ 1 integrin, or to pathological effects of circulating immune-complexes, unrelated to the antigenic specificity of mAb 1B3.1. An additional mechanism could involve co-activating effects in T cells simultaneously engaged by antigens, and even direct activation of mucosal yb T cells which in vitro become induced to express cytokine receptors in response to cross linking with mAb IB3.1 even in the absence of antigenic triggering (9). The in vitro results demonstrating markedly enhanced expression ofIL-2R and CD69 in cells cross linked by mAb IB3.1, suggest that in vivo IB3.1 bound T cells could be further triggered by IL-2 in their micro-environment and may become primed for enhanced cytolytic activity mediated by CD69 (30). Since signals generated by cross linking of al integrins on T cells (as well as fibroblasts) are dependent upon activation of protein-kinases (Figure 2), inhibitors of protein kinases might be sufficent to abrogate such potentially harmful T cell activating effects of the parenterally administered mAb (31,32). While further experiments, including autopsies and in situ immunohistological studies of tissues are under way in our laboratory to determine conclusively the mechanism of the lymphopenia and gastrointestinal toxicity induced by the parenteral administration of high doses of mAb 1B3.1, the potential harmful effects should be taken into consideration when such reagents are considered for treatment of intestinal-mucosa manifestations of GVHD.
VLA-l expression and function in dogs . 251
An additional specific effect of injection of mAb 1B3.1 is the alteration of the CD4/CD8 ratio irrespective of the degree of lymphopenia induced. This shift was due to the more persistent depleting effect on CD8+ rather than CD4+ cells, which is perhaps the consequence of the higher level of expression of a 1~ 1 integrins on activated CD8+ T cell clones (9). Differential effects on CD8+ and CD4+ T cells could potentially play an important role in immune mediated diseases such as GVHD. For example, preferential migration of CD8+ T cells from the vasculature or lymph into tissues could affect the outcome of local immune responses. In summary, these preliminary studies indicate for the first time that parenteral administration of a mAb to the a 1 integrin to a large mammal is clinically feasible, when given at appropriate dosages and intervals. Furthermore, the results suggest that this potential therapy could induce differential shifts in specific T cell populations, resulting in altered ratios among the subsets in various anatomic locations, in addition to the potential blockade of specific interactions with the extra cellular matrix, and augmentation of activation induced via the TCR-CD3 complex. Whether and how this mAb, or other similar reagents, can be safely used at doses and intervals sufficient for modulation ofT cell mediated diseases in a large animal model is an important subject for future investigations. Acknowledgement This work was supported by a grant from the Israel Ministry of Health to I. BANK and I. HARDAN.
References 1. HYNES R. 1992. Integrins: versatility, modulation, and signaling in cell adhesion. Cell 69: 11. 2. HEMLER M. E. 1990. VLA proteins in the integrin family: structures, functions and their role on leukocytes. Annu. Rev. Immunol. 8: 365. 3. SHIMIZU Y., G. A. VAN SEVENTER, K. J. HORGAN, and S. SHAW. 1990. Regulated expression and binding of three VLA (~1) integrin receptors on T cells. Nature 345: 250. 4. SHIMIZU Y., G. A. VAN SEVENTER, K. J. HORGAN, and S. SHAW. 1990. Co-stimulation of proliferative responses of resting CD4+ T cells by the interaction of VLA-4 and VLA-5 with fibronectin or VLA-6 with laminin. ]. Immunol. 145: 59. 5. MIYAKE S., T. SAKURAI, K. OKUMURA, and H.YAGITA. 1994. Identification of collagen and laminin receptor integrins on murine T lymphocytes. Eur. ]. Immunol. 24: 2000. 6. HEMLER M. E., J. ]ACOBSON, M. B. BRENNER, D. MANN, and]. L. STROMINGER. 1985. VLA-1: a T cell surface antigen which defines a novel late stage of human T cell activation. Eur. ]. Immunol. 15: 502. 7. GOLDMAN R., J. HARVEY, and N. HOGG. 1992. VLA-2 is the integrin used as a collagen receptor by leukocytes. Eur. ]. Immunol. 22: 1109. 8. BANK I., M. E. HEMLER., M. B. BRENNER, D. COHEN, V. LEVY, J. BELKO, C. CROUSE, and L. CHESS. 1989. A novel monoclonal antibody, 1B3.1, binds to a new epitope of the VLA-1 molecule. Cell. Immunol. 122: 416. 9. BANK I., M. BOOK, and R. WARE. Functional role ofVLA-l (CD49a) in adhesion, cation dependent spreading, and activation of cultured human T lymphocytes. 1994. Cell Immunol. 156: 424. 10. CHAN B. M. C., J. G. r WONG, A. RAo, and M. E. HEMLER. 1991. T cell receptor-dependent, antigen-specific stimulation of a murine T cell clone induces a transient, VLA protein-mediated binding to extracellular matrix. J. Immunol. 147: 398. 11. TANAKA T., Y. OHTSUKA., H.YAGITA, Y. SHIRATORI, M. OMATA, and K. OKUMURA. 1995. Involvement of a 1 and a4 integrins in gut mucosal injury of graft versus host disease. Int. Immunol. 7: 1183.
252 . I. BANK et ai. 12. BANK I., D. ROTH, M. BOOK., R.BLOCK, ~LANGEVITZ, M.EHRENFELD, and M. PRASe 1991. Expression and functions of very late antigen 1 in inflammatory joint diseases. J. Clin. Immunoi. 11: 29. 13. BANK I., A. TANAY, A. MIGDAL, M. BOOK, and A. LIVNEH. 1995.Vy9-V02+ yO T cells from a patient with Felty syndrome that exhibit aberrant responses to triggering of the CD3 molecule can regulate immunoglobulin secretion by B cells. Clin. Immunol. and Immunopathol. 74: 162. 14. SONNENBERG A., H. JANSSEN, E HOGERVORST, H. CALAFAT, and]. HILGERS. 1987. A complex of platelet glycoproteins Ic and IIa identified by a rat monoclonal antibody. J. BioI. Chern. 262: 10376. 15. MOORE ~ E, T. OLIVRY, and D. NAYDAN. 1994. Canine cutaneous epitheliotropic lymphoma (mycosis fungoides) is a proliferative disorder of CD8+ T cells. Am. J. Pathoi. 144: 421: 12. 16. TIPOLD A., ~ E MOORE, T,W JUNGI, V. SAGER, and M.VANDEVELDE. 1998. Lymphocyte subsets and CD45RA positive T cells in normal canine cerebrospinal fluid. J. Neuroimmunoi. 82: 90. 17. LANGE T. S., A. K. BIELINSKY, K. KIRCHBERG, I. BANK, K. HERMANN, T. KRIEG, and K. SCHARFETTER-KoCHANEK. 1995. Divalent cations Mg2 +and Ca2 +differentially regulate ~1 integrin mediated adhesion of dermal fibroblasts and keratinocytes to different extracellular matrix proteins. J. Exp. Dermatoi. 4: 130. 18. BANK I., E. !{APMAN, R. SHAPIRO, G. SCHIBY, I. GOLDBERG, A. BARZILAI, H. TRAU, and H. GUR. 1999. The epidermotropic mycosis fungoides associated a1~1 integrin (VLA-1, CD49a/CD29) is primarily a collagen IV receptor on malignant T cells. J. Cutan. Pathol. 26: 65. 19. DEAS 0., C. DUMONT., M. MACFARLANE, M. ROULEAU, C. HEBIB, E HARPER, E HIRSCH, B. CHARPENTIER, G. M. COHEN, and A. SENIK. 1998. Caspase-independent cell death induced by anti CD2 or staurosporine in activted human peripheral T lymphocytes. J. Immunoi. 161: 3375. 20. LEBIEN T. W, D. R. BOUE, J. G. BRADLEY, and]. H. KERSEY. 1982. Antibody affinity may influence antigenic modulation of the common acute lymphoblastic leukemia antigen in vitro. ]. Immunol. 129: 2287. 21. LAEMMLI U. K. 1970. Cleavage of structural proteins during the assembly of the head ofbacteriophage T4. Nature (Lond.) 227: 680. 22. KERN A., R. BRIESEWITZ, I. BANK, E. E. MARCANTONIO. 1994. The role of the I domain in ligand binding of the human integrin a 1~ 1. J. BioI. Chern. 269: 22811. 23. HANSSON G. K., J. HOLM., and L. JONASSON. 1989. Detection of activated T lymphocytes in the human atherosclerotic plaque. Am. ]. Pathoi. 135: 169. 24. SALTINI C., M. E. HEMLER., and R. G. CRYSTAL. 1988. T lymphocytes compartmentalized on the epithelial surface of the lower respiratory tract express the very late activation antigen complex VLA-1. Clin. Immunoi. Immunopathoi. 46: 221. 25. PEREZ-VILLAR J. J., I. MELERO, A. GISMONDI, A. SANTONI, and M. LOPEZ-BoTET. 1996. Functional analysis of alpha 1 beta 1 integrin in human natural killer cells. Eur. J. Immunol. 26: 2023. 26. KONO Y" S. YANG, M. LETARTE, and E. A. ROBERTS. 1995. Establishment of a human hepatocyte line derived from primary culture in a collagen gel sandwich culture system. Exp. Cell. Res. 221: 478. 27. SOBEL R. A., J. R. HINOJOZA., A. MAEDA, and M. CHEN. 1998. Endothelial cell integrin laminin receptor expression in multiple sclerosis lesions. Am. J. Pathoi. 153: 405. 28. RACINE-SAMSON L., D. C. ROCKEY., and D. M. BISSEL. 1997. The role of alpha1betal integrin in wound contraction. A quantitative analysis of liver myofibroblasts in vivo and in primary culture. J. BioI. Chern. 272: 30911. 29. TAWIL N. J., M. HOUDE, R. BLACHER et ai. 1990. Alpha 1 beta 1 integrin heterodimer functions as a duallaminin/collagen receptor in neural cells. Biochemistry. 29: 6540. 30. BORREGO E, M. J. ROBERTSON, J. RITZ, J. PENA, and R. SOLANA. 1999. CD69 is a stimulatory receptor for natural killer cell and its cytotoxic effect is blocked by CD94 inhibitory receptor. Immunology 97: 159.
VLA-1 expression and function in dogs.
253
31. POZZI A., K. K. WARY, F. G. GIANCOTTI, and H. A. GARDNER. 1998. Integrin alpha1beta1 mediates a unique collagen-dependent proliferation pathway in vivo. ]. Cell. BioI. 142: 587. 32. RAVANTI L., ]. HEINO., C. LOPEZ-OTIN, and V. M. KAHARI. 1999. Induction of collagenase-3 (MMP-13) expression in human skin fibroblasts by three dimensional collagen is mediated by p38 mitogen activated protein kinase. ]. BioI. Chern. 274: 2446. Dr. ILAN BANK, Dept. Medicine F, Chaim Sheba Medical Center, Tel Hashomer, Israel 52621. Fax: ++97235302114, e-mail: [email protected]
1 Department of Medicine F, 2 Laboratory for Immunoregulation, 3 Hematology Institute, 4 Department of Pathology, 5Neuroimmunology Unit, and 6Neufeld Cardiac Research Institute, Chaim Sheba Medical Center, Tel Hashomer, Sackler Faculty of Medicine Tel Aviv University, Israel, and 7 School of Biological Sciences, University of Manchester U.K.
Parenteral Administration of an Activating Monoclonal Antibody to the a 1~ 1 Integrin in Dogs 1
3
4
ILAN BANK 1, 2, IZHAR HARDAN , EVGENIA LOKSHIN , DEVORA NAS , 5
6
7
SHMUEL MIRON , DANIEL OHAD , SUZANNE SPONG , and DAVID R.GARROD
7
Received September 22, 1999 . Accepted in revised form March 22, 2000
Abstract In mice, monoclonal antibody (mAb) to the al integrin abrogate gastro-intestinal damage during graft-versus-host-disease (GVHD), suggesting anti al mAb as candidates for treatment in humans as well. Our current data show that one such reagent, mAb 1B3.1, when immobilized to plastic wells via rabbit- anti murine (ram) immunoglobulin (Ig) induces a protein kinase-dependent spreading of activated human T cells. Furthermore, it significantly increases the proliferative response, and expression of interleukin-2 (IL-2) receptors (R) and CD69, of resting T cells, expressing minimal integrin on the cell surface, to sub-optimal stimulation by anti-C03 mAb. We found, in addition, that mAb IB3.1 a) immuno-precipitates a 1~ 1 integrins from cell-surface iodinated canine epithelial cells b) is highly reactive with canine T cells after their activation and c) inhibits adhesion of canine T cells to collagen IV: Despite the potential ability of the mAb to co-activate T cells in vitro, two dogs that received 4 injections of 0.5-0.3 mg/Kg of mAb 1B3.1 remained healthy, developing only marginal transient lymphopenia. Injection of 0.75mg/Kg in a third dog induced a more marked lymphopenia, and an additional dose of 1.0 mg/Kg 2 weeks later was followed by gastrointestinal hemorrhage. Importantly, the lymphopenia was associated with a greater and more persistent decrease of C08+ than of C04+ T cells, leading to an increase in the C04/C08 ratio 24 hours after the injection. Thus, despite it's co-activating effects in vitro, administration of this mAb in vivo is feasible when appropriately dosed, and may have immuno-modulatory effects.
Abbreviations: BSA = bovine serum albumin, ECM = extracellular matrix, OMEM = Oulbecco's
Modification of Eagles Medium, FCS = fetal calf serum, FM = final medium, FITC = fluoresceine isothiocyanate, FACS = fluorescence activated cell sorter, gam = goat anti-mouse, GVHO = graft-versushost-disease, Ig = immunoglobulin, IL-2 = interleukin-2, IU = international units, IMOM = Iscoves modified Oulbecco medium, MOCK = Madin Darby Canine Kidney, mAb = monoclonal antibody, NAC = non-adherent cells, 00 = optical density, PBMC = peripheral blood mononuclear cells, PBS = phosphate-buffered saline, PHA = phytohemagglutinin, ram = rabbit-anti-murine, VLA = very late antigen, SOS-PAGE = sodium-dodecyl sulfate polyacrylamide gel-electrophoresis, TBS = tris-buffered saline. 0171-2985/00/202/03-239 $ 15.00/0
240 . 1.
BANK
et al.
Introduction Six of the ~ 1 integrins, consisting of unique a peptides (a l-a6) in non- covalent association with a common ~1 peptide, are expressed on leukocytes (1, 2). T cell ~1 integrins have been studied almost exclusively in humans and mice (2-5). Resting T lymphocytes express high levels of the a4~1 and a5~1 fibronectin receptors (also known as the very late antigens (VLA) 4 and 5), and of a6~1, a laminin receptor, whose adhesive functions are markedly and transiently enhanced by activation with phorbol-esters, or triggering of the T cell receptor (TCR)-CD3 complex by antigen or mAb ("inside-out" signalling) (3). On the other hand, cross linking ofT cell surface membranes with extracellular matrix (ECM) proteins or mAb directed against VLA-4, 5 or 6 in the presence of a sub-mitogenic concentration of mAb to CD3 augments T cell proliferation ("outside-in" signalling) (4). The VLA-l (al~l) and VLA-2 (a2~1) heterodimers, receptors for collagens and laminin, are found at low levels on human and murine resting T cells (2, 5-7). Increased expression of these integrins occurs on the surface membrane after relatively prolonged activation periods and they are used by activated human and murine T cells to adhere and spread on collagens (2, 5-10). A recent study demonstrated that intraperitoneal (IP) administration of mAb to the a 1~ 1 integrin prevents the mucosal damage of murine GVHD, which suggests that anti al integrin mAb could be used in larger mammals as well for the therapy of this disease (11). We have addressed this possibility by employing a mAb specific for human a 1 integrins which was found previously, on the one hand to block adhesion of T cells to collagen, and on the other-to activate certain T cells (8, 9). In these studies we further analyzed the activation of T cells by the mAb and, using novel data concerning cross-reactivity of this mAb with the canine integrin, report
for the first time preliminary studies of effects of it's parenteral administration.
Materials and Methods Peripheral blood mononuclear cells (PBMC) and
T cell
cultures
PBMC were isolated from normal consenting volunteers and from dogs by Ficoll-Hypaque density centrifugation (12). Dog PBMC were activated with 1% phytohemagglutinin (PHA) (Biological Industries, Beit Haemek, Israel) then expanded and passaged in 24-well tissue culture plates (Costar Corporation, Cambridge, MA, USA) in Iscoves modified Dulbecco medium (IMDM) supplemented with 100/0 heat-inactivated fetal calf serum (FCS), 1% glutamine, and 1% penicillin streptomycin solution (Biological Industries, Beit Haemek, Israel) (final medium (FM)) and 100 international units (IU)/ml of recombinant interleukin 2 (IL-2), (Proleukin, Chiron Corp., Amsterdam, the Netherlands) (8). For isolation of resting T cells from the PBMC macrophages were first removed by incubating PBMC in tissue culture treated plastic dishes (Costar corporation Cambridge, MA, USA) for 1 hour in a 60/0 CO 2 humidified 37°C incubator in FM and the non-adherent cells (NAC) were harvested. To remove residual macrophages, B cells and in vivo activated T cells, the NAC were treated with anti HLA-DR mAb OKI2 for 30 minutes, washed X3 in phosphate-buffered saline (PBS) pH = 7.2, then panned on culture plates coated with 10 flg/ml of affinity purified ram-Ig diluted in Tris-buffered saline (TBS), pH = 9.5 as previously described (13). The NAC population contained >980/0 CD3+HLA-DR- lymphocytes (not shown).
VLA-1 expression and function in dogs .
241
MAb MAbs 1B3.1 (IgA, anti-a1B1 integrin), OKT11 (IgG1, anti C02), OKT3 (IgG2, anti C03), and OKI2 (IgG 1, anti HLA-OR) were prepared from ascites fluid from Balb/C mice bearing the respective hybridomas (12, 14). TEPC-15 (IgA, anti phosphorylcholine) was purchased from Sigma Chemical Co. St. Louis, MO, USA (12). Mabs OKT3 and 1B3.1 were further purified by affinity chromatography and dialyzed against PBS. MAb AIIb2 against 1 integrin chains was kindly supplied by Dr. C. DAMSKY (14). MAb MOPC-167 (irrelevant murine IgA) was a gift from Prot Z. ESHAR, Weizmann Institute, Rehovoth, Israel. MAb to canine CD18, CD45RA, CD4 and CD8 were kindly supplied by Dr. ~ MOORE, University of California, Davis, USA (15,16). Directly conjugated mAb to IL-2R and CD69 (Leu 23) were obtained from Dako, Denmark, and Becton Dickinson, San Jose, CA, USA, respectively.
B
Fluorescence-activated cell sorter (FACS) analysis
2 x 10 5 cells were stained with saturating concentrations of mAb at 4°C, washed, then incubated with goat anti-mouse (gam)-Ig conjugated to fluoresceine iso-thiocyanate (FITC) (gam-Ig-FITC) (Sigma) as previously described (8). The level of expression of alB 1 integrin quantitated from data supplied by Coulter FACS software is expressed as mean fluorescence intensity (MFI, arbitrary units) of analyzed cells. Dogs
The study was approved by the Tel Aviv University Helsinki committee. Male mongrel dogs were housed in the Sackler Medical School facility for medical research. Dogs were sedated by intramuscular injection of a short acting barbiturate. MAb were administered into a peripheral vein after extensive dialysis into PBS and 0.45 ~m sterile filtering. Hematological parameters
Blood was drawn into calcium citrate containing tubes and hematological parameters evaluated using standard clinical Coulter counters. Values were validated in some experiments by direct visualization of stained blood smears. Adhesion assays
Assays were performed as previously described (17,18). Dissolved collagen IV (Sigma) or fibronectin were diluted in H 2 0 to 10 ~g/ml and 50 ~l added to 96 micro-wells tissue culture plates (Costar), which were air dried at room temperature overnight. The ECM protein coated wells and control non coated wells were blocked with 50 ~l of 0.20/0 bovine serum albumin (BSA) (Sigma) in PBS for 30 minutes, then washed X3 with PBS. Cultured cells were washed twice in fresh IMDM by centrifugation at 1000 r.p.m, and resuspended in IMDM. 150,000 cells/well were added to triplicate wells, in a final volume of 200 ~l/well of IMDM. For inhibition studies, mAb (final concentration 1:200 of ascites fluid) were added to the medium. The plates were placed in a humidified atmosphere at 37°C, 6% CO 2 , for 40 minutes, and non-adherent cells discarded. Adherent cells were fixed in 700/0 ethanol for 10 minutes, and stained with 0.1 % crystal violet solution in H 2 0 for 20 minutes. Plates were washed extensively in H 2 0, and 100 ~l of 0.1 % Triton X-I 00 solution added for 30 minutes to release the dye. The optical density (OD) at 550 nm was recorded using an automated ELISA plate reader. Direct counting of adherent cells established that OD was proportional to the number of adhering cells (18). Results are reported as mean OD of triplicate wells containing ECM protein minus mean OD of triplicate control wells coated with BSA alone. Proliferation assays
Flat bottom ninety-six well tissue culture plates (Costar) were coated overnight at room temperature with 10 ~g/ml of affinity purified ram-Ig diluted in TBS, pH = 9.5, then washed X3 in PBS
242 . 1.
BANK et
al.
(12). The mAbs diluted in PBS were added to the ram-Ig coated wells for 2 hours at 4°C which were then washed X3 with PBS. 5 x 10 4 cells/well were added, and plates placed in a 60/0 CO 2 970/0 humidified incubator at 37°C. The cultures were pulsed overnight for 12 hours with 1 mC/well of 3 [H] thymidine (Amersham) after 60 hours of incubation then harvested onto nylon mesh filter paper. Incorporation of thymidine was measured in a ~-counter after placing the dried filters in vials containing scintillation fluid as previously described (12). All experiments were performed in triplicate.
Induction of IL-2R and CD69 Cells were cultured as described above for 24 hours, then washed X2 in PBS and stained with mAb to CD25 or CD69 that were directly conjugated to FITC.
Spread ing assays 24 well tissue culture plates (Costar) were coated overnight at room temperature with 10 J,Lg/ml of affinity purified ram-Ig diluted in TBS pH=9.5 then washed 3x in PBS (12). MAb (1:200 of ascites in PBS) were added for 2 h at 4°C and the wells were then washed 3 x with PBS. T cells removed from growing cultures were placed in the CO 2 incubator in FM alone or FM with 100 ng/ml staurosporine for 1 hour then washed 3 x in IMDM prior to layering in the wells (19). 5 x 10 5 T cells in 500 J,Llof FM were added in each well and plates placed in the CO 2 humidified incubator for 3 h, after which wells were rinsed with IMDM to remove non-adhering cells. The number of adhering cells per inverted phase microscope field, exhibiting elongated and or spherical (round) morphology was recorded (9).
Immunoprecipitation Madin Darby Canine Kidney (MDCK) cells (European Collection of Animal Cell Cultures, Porton Down, Salisbury, U. K.) were grown in Dulbecco's Modification of Eagles Medium (DMEM) supplemented with 10% (v/v) FCS or Newborn Calf Serum, penicillin (100u/ml)) and streptomycin 1OOllg/ml at 37°C in 5% CO 2 , Confluent cells were trypsinized and seeded into 9 em petri dishes (1 x 10 7 / dish) precoated with a layer of collagen gel. The collagen substratum was digested with 0.1 % (w/v) collagenase in DMEM. Cells released from the gel were washed in ice cold PBS and resuspended to 1.5 - 2 x 10 8 cells/ml in PBS. Two hundred microliters of lactoperoxidase (lmg/ml in PBS), 1.7-2 mCi Na 125 1 (lCN Pharmaceuticals Ltd., Thame, Oxon. or Amersham International pIc., Little Chalfont, Bucks, U.K.) in 0.2 ml PBS and 2 III of 0.12% (v/v) H 2 0 2 in PBS were added to 1 ml of cells. After a 5 minute incubation on ice, 20 III of 0.12% H 2 0 2 were added for an additional 5 minutes. This step was repeated and the reaction stopped by adding 10 ml of PBS (20). Cells were then washed and lysed in 2 ml of extraction buffer (1 % (v/v)Triton X-100, 10 mM TrisHCl pH 7.4, 0.13 M NaCl, 5 J,Lg/ml BSA, 1mM MgCI 2 , 1mM MnCI 2 , 10 J,L/mlleupeptin and 1 mM phenylmethylsulfonyl fluoride) on ice for 30 minutes. Lysates were centrifuged (3,000 x g for 5 minutes) and supernatants passed down a PD-10 column (Pharmacia Biotech, St Albans, Herts, U.K.). Labelled proteins were eluted in 3.5ml of extraction buffer and extracts precleared by adding 10% (v/v) Protein G-Sepharose and mixing for 1.5 hrs at room temperature. Beads were removed by centrifugation (14,000 x g for 10 minutes) and the cleared extract added to an equal volume of PBS+ containing 2.5 mg/ml BSA and immune Igs or control mouse IgG, and rotated overnight at 4°C. Immunoprecipitates were collected by the addition of 20 J,Ll protein GSepharose/tube and rotating for a further 1.5 hr at 4°C. Beads were harvested by centrifugation (7,000xg for 3 min), washed 3 times in extraction buffer and twice in PBS+, and resuspended in sodium-dodecyl sulfate polyacrylamide gel-electrophoresis (SDS-PAGE) sample buffer (with or without mercaptoethanol) and boiled for 3 min. Immunoprecipitated proteins were separated by SDS-PAGE using the method of LAEMMLI (21). Gels were soaked in 100/0 (v/v) acetic acid, 250/0 (v/v) methanol, dried on a BIO-RAD gel drier, and exposed to x-ray film or analyzed using a Molecular Dynamics phosphorimager. Digitized images were obtained using a Fujix Bas 2000 Bio imaging analyser.
VLA-I expression and function in dogs . 243
Results a 1~ 1 integrin-mediated
T cell activation
It was previously shown that a proliferative response is induced in cultured or in vivo activated but non-proliferating T cells expressing high levels of the a 1~ 1 integrin, by cross linking the cells with solid phase bound mAb 1B3.1 in the presence of a non-mitogenic dose of anti-CD3 mAb (12). Since the goal of our current study was to evaluate possible effects of the mAb on T cells in vivo, we extended these observations by studying the effects of the mAb on freshly isolated and highly purified resting T cells from normal volunteers, which expressed a very low level of VLA-1 on the cell surface (MFI of cells stained with mAb 1B3.1=1.16, control MFI=1.05, MFI of cells stained with mAb TEPC-15 =1.16, not shown) (12). When such cells were coated with mAb 1B3.1, placed in culture wells after washing of excess mAb, and triggered with a previously determined minimally mitogenic concentration of mAb OKT3, their proliferative response was not significantly different from that of non-treated or isotype control (mAb TEPC-15, anti-phosphoryl choline) treated cells, in marked contrast to the augmented proliferation of cells that were pre-treated with mAb OKT11 (directed against the costimulatory CD2 molecule) (Fig. 1A). However, when the resting T cells were placed in culture in micro-wells containing immobilized mAb 1B3.1, a markedly enhanced proliferation occurred relative to that of cells cultured in wells coated with the control mAb TEPC-15 or with no cross linking mAb (Fig. 1B). Furthermore, the enhancement of proliferation induced by cross linking VLA-1 molecules with immobilized anti-integrin mAb was dependent upon the concentration of mAb 1B3.1 used to coat the wells (Fig. 1C). In addition, cells activated by mAb 0 KT3 in the presence of cross linking of the VLA-1 integrin, markedly upregulated the surface expression of IL-2R and of CD69 on the cell surface (Fig. 1D). These results suggest that prolonged and efficient crosslinking of sparsely expressed VLA-1 molecules on the surface membrane of resting T cells significantly augments CD 3 mediated activation signals in previously resting T cells. Protein-kinase-dependent signalling
by VLA-l
To further investigate the nature of "outside-in" signals transmitted by a 1~ 1 integrins, VLA-1 + T cells cultured in IL-2 containing medium were placed on plastic wells coated with ram-Ig alone, or with ram-Ig to which mAb 1B3.1, OKT11 or the non reactive control mAb MOPC-167 had been added. Only wells coated with mAb 1B3.1 and OKT11 contained adherent cells after three h of incubation (Fig. 2A). Cells adherent to OKT11 coated wells had a predominantly spherical morphology, whereas cells adherent to mAb 1B3.1 were either spherical or spread (elongated). The T cells were next incubated for 1 h in either medium, or in medium containing 100 ng/ml of staurosporine, an inhibitor of protein kinase activity, prior to seeding in the wells (19). Adherent elongated cells, but not spherical cells, were significantly decreased in wells containing cells pretreated with staurosporine (Figs. 2B 1 and 2B2). Thus, T cell spreading induced by cross linking of the alB 1 integrin requires the activation of intracellular proteinkinases.
244 .
I. BANK et al.
7ססoo
RAM
eoooo
RAM+OKT3
50000
~~j
RAM+ 183.1+OKT3
O:J
RAM+TEPC+OKT3
1~j oj
RAM+OKT11 +OKT3
a
20000
40000
CPM
60000
---_._..
medUn+OKT3
1B3.1.0KT3
-------
60000
60 C
D
50
50000
~ 40000
j
30000
j
20000
j
10000
rnedIlm
.!/}. 1)
~ 40
a
••
111600
11800 11400 Dilution of ascites
11200
~ ~30
j j
•
Il-2R
~
CD69
11 20
1: j
.....
.-....om
TlPC40KT3
Figure 1. Enhanced activation in response to immobilized anti VLA-l mAb. A. Resting human T cells in suspension were coated with no mAb, with mAb IB3.1, TEPC, or OKTll and excess mAb washed away. The cells were then cultured (5 X 104/well) in triplicate wells that were pre-coated with ram-Ig, and triggered with 10 ng/ml of anti-CD3 mAb. 3[H] thymidine incorporation was measured after 72 hours of culture. Results represent the mean CPM of triplicate experiments B. Non pre-treated T cells were added to culture wells that had been precoated with ram-Ig alone (medium) or with 1/200 of ascites fluid ofmAbs IB3.1 orTEPC-15, then triggered with 10 ng/ml ofanti-CD3 mAb. C. Dose response of the mitogenic effect of increasing concentrations of immobilized mAb 1B3.1 (squares) or TEPC-15 (black circles) using the same experimental protocol. D. T cells activated as in B were removed from culture wells after 24 hours, stained with FITC conjugated mAb to IL-2R or CD69 and analyzed by flow cytometry.
Reactivity of mAb 1B3.1 with canine cells To develop an animal model in which in vivo effects of mAb IB3.1 could be evaluated, we determined the reactivity of this mAb with canine tissues. Immunoprecipitation of cell-surface iodinated molecules of the MDCK line, which, in screening flow-cytometry experiments was found to react with mAb IB3.1 (not shown), using mAb's IB3.1 and AIIB2 were performed (Fig. 3). MAb IB3.1 immuno-precipitated peptides of200 kDa and 110 kDa, corresponding to al and pI integrin chains respectively (lane 2) (8). MAb AIIB2, directed against the 1 integrin peptide, immuno-precipitated both 200 kDa a 1
B
VLA-l expression and function in dogs· 245 A
ram
•
ram +OKT11
~herical
ram + 183.1
elongated
ram + MOPC-167
o
30 60 90 120150
elongated
spherical
cells. 81
rlm+ MOPV1fr7
cells. 82
rlm+ MOPC-1fr7
•
Cf,um+staurospor,ne
rlm+183.1 • • • •JlI
•
Cf,um+staurosporlne
rlm+1B3.1 , - - - - - - - - - . ,
medium
o
100
200
300
400
cenl per field
500
medium
o
20
40
eo
eo
celli per tIIId
100
Figure 2. Inhibition of T cell spreading by staurosporine. A. 10 5 human alp 1+ T cells, previously cultured in IL-2, were allowed to adhere to wells pre-coated with ram-Ig or with ram-Ig to which mAbs IB3.1, OKTll or MOPC-167 had been immobilized. Cells exhibiting spherical or elongated morphology were counted using an inverted phase microscope. Bars represent mean ± 1 std of three fields. B. The cells were allowed to adhere to wells pre-coated with ram- Ig or with ram- Ig to which mAb IB3.1 or mAb MOPC-167 had been immobilized. A parallel aliquot of cells was incubated with staurosporine (100 ng/ml) for 1 hour prior to addition to identical wells. Cells exhibiting spherical (B 1) or elongated morphology (B2) were counted. Bars represent mean ± 1 std of three fields.
and 160 kDa (corresponding to a6 peptides) as well as 110 kDa peptides corresponding to ~ 1 integrin chains (lane 3) (14). These results indicated that mAb 1B3.1 specifically recognizes a 1~ 1 integrins expressed by canine epithelial cells. Next, we investigated patterns of reactivity of mAb 1B3.1 with canine T cells. Flow cytometry of resting and in vitro activated PBMC isolated from 4 mongrel dogs (in 5 experiments) demonstrated that the reactivity of the activated cells with mAb 1B3.1 was consistently and significantly higher, suggesting that expression of the a 1~ 1 integrin is enhanced on the surface of canine T cells after their activation, as previously demonstrated for human T cells (Fig. 4) (12). Finally, adhesion assays of activated dog T cells expressing a 1~ 1 integrins to ECM proteins collagen IV and fibronectin were performed (Materials and Methods). An isotype control mAb (MOPC-167) did not inhibit adhesion to either protein (Fig. 5). In contrast, adhesion of the cells to collagen IV, (but not to fibronec-
246 . I. BANK et al.
1 2 3 200 -
<4I-al
100
Figure 3. Autoradiogram of integrins immuno-precipitated from MOCK cells. SOS-PAGE analysis (non-reduced) of extracts of 25 I] cell surface iodinated MOCK cells immunoprecipitated with non immune mouse Ig (lane 1), with mAb IB3.1 (anti-al integrin) (lane 2), or mAb Allb2 directed against the common ~I integrin chain (lane 3). Location of200 kOa and 100 kOa molecular weight markers run simultaneously are indicated on the left. Expected locations of the indicated integrin peptides are designated on the right. The a6 integrin, immunoprecipitated in lane 3, migrates at 160kOa.
e
tin) was strongly inhibited by mAb 1B3.1. These results indicate that the a 1~ 1 integrin in activated dog T cells, as in humans, functions as a specific receptor for collagen
I\Z Effects of anti VLA-l in vivo Preliminary experiments were conducted to test whether the mAb, shown above to co-activate T cells under specific experimental conditions in vitro, could be safely administered to live dogs (Table 1). Minor gastro-intestinal side effects were noted after injections of 0.3- 0.5 mg/Kg mAb 1B3.1 in dogs 1 and 2. Injections of this dose of mAb 1B3.1 were also followed by a non-significant decrease of lymphocytes in the PB after 6 h, with almost complete recovery occurring after 168 h (Fig. 6a). This transient reduction of lymphocytes was not significantly different from the mild lymphopenia induced by injection of the anti phosporyl-choline control mAb TEPC15. In dog 3, in contrast, into which 0.75 mg/Kg of mAb were administered, a more
VLA-l expression and function in dogs .
1 ~
+oJ
"in
5j 0.8
+oJ
c:
• activated cells
resting cells
---t> p
----[>
g 0.6 Q)
p
oU) o
p
::3
i;:
C
eu
0.2\ =-
E
•
:
'---
o
• ••
------iC>
•• •••
•
• -.-•
-• -
•
Q)
~ 0.4
247
• •
=
p
cont.
TEPe-15
183.1
cont.
'fEPC-IS
IB3.1
Figure 4. Increased al~1 integrin on activated T cells. MFI of 5 samples of resting and dog PBMC activated in vitro for 3 weeks as described in materials methods, after staining with gam-Ig-FITC alone (cont.) or with mAbs TEPC-15 or IB3.1 followed by gam-Ig-FITC. Origin and tips of arrows indicate groups of values the means (horizontal lines) of which were compared using a one tailed Student's t-test. Resulting p values are indicated.
0.06
0.03
0.05
0.025
0.04
0.02
collagen p
ci
·0
:I 0.03
0 0.015 :I
-8 0.02
0.01
0
-8
0.01
0.005
0
0
Figure 5. Inhibition of adhesion to ECM proteins by mAb 1B3.1. Cultured a 1~ 1 integrin expressing canine T cells were allowed to adhere to plastic wells coated with either fibronectin or collagen IV blocked with BSA, or with BSA alone, for 40 min in the presence of medium (black bars), medium containing 1:200 dilution of ascites of mAb's 1B3.1 (white bars) or isotype control mAb MOPC-167 (obliquely hatched). Adherent cells were stained with crystal violet, and dye released with 0.1 % triton. Results, on the Y-axis, are given as delta 00 = 00 of wells coated with ECM and blocked with BSA - 00 of wells with no ECM protein and blocked with BSA. Results are representative of 3 experiments. P values obtained from the Student's t-test, are shown.
profound and delayed lymphopenia was induced, while the neutrophil count was unaffected, but no adverse clinical effect was observed (Fig. 6b). Additional administration of a higher dose (1 mg/Kg) of mAb 1B3.1 two weeks after the first injection, resulted in massive hematemesis after 26 hours and subsequent death of this dog.
248 . 1.
BANK
et al. 140%~------------.
0
130%
0120% Q)
Q)
:§ 100%
E ;; 100%
.9
.9
>
a;
Q)
Q)
.~
li 70%
-~
'B
80°A»
neutrophils
--lymphocytes
~ 60°A» C 400/0
~
c:
~
~ 40%
~ 20%
Q) ~
A
?
10% - ' - - - - + - - - - + - - - - + - - - - + - - - - - ' o 6 24 168 hours after injection
B
o
6 24 168 336 342 hours after injection
Figure 6. Effect of mAb IB3.1 on lymphocyte counts. A. Lymphocytes per cubic mm were enumerated in the PB before (time 0) and after injections of PBS (open squares), 0.5 mg/Kg - 0.3 mg/Kg of mAb IB3.l (black squares, representing mean and 1 std of 3 experiments in dog 1 (Table 1) or 0.5 mg/Kg of mAb TEPC-15 (triangles). Results are presented as percent of the lymphocytes at the time points indicated on the X-axis relative to values at time o. B. Lymphocyte and neutrophil counts after injection of 0.75 mg/Kg in dog 3. A second dose of 1 mg/Kg was given 336 hours after the first injection.
Table l. In vivo administration of reagents to dogs: dosages, intervals, and clinical outcome.
Dog 1
Dog 2
Dog 3
day reagent dose symptom
0 mAb IB3.1 0.5 mg/Kg vomiting
day reagent dose symptom
0 mAb IB3.1 0.5 mg/Kg none
day reagent dose symptom
0 mAb IB3.1 0.75 mg/Kg none
66 mAb IB3.1 0.5 mg/Kg none
123 mAb TEPC-15 0.5 mg/Kg none
141 PBS none
183 mAb IB3.1 0.3 mg/Kg none
7
sacrificed for autopsy 14 mAb IB3.1 1.0 mg/Kg hematemesis after 6 hours, death after 24 hours
Effects of injected mAb on T cell subsets
To further evaluate the effects of parenterally injected mAb, CD18+, CD45RA+, CD4+ and CD8+ lymphocytes were enumerated in the PB. Injection of mAb 1B3.1 resulted in marked reduction in cells expressing these lymphocyte markers after 3-6 hours, whereas injection of the isotype control mAb TEPC-15 had only a marginal effect (Figs. 7A and 7B). Although the percentage ofPB lymphocytes expressing CD45RA, CD18 and
VLA-I expression and function in dogs 8
A
140%
14QOk
.! 1200k
CD18 CDS
-.
'0 600/0 C ~ 4QOA. 8. 200/0 0%
CD4
~
CD45RA
0
...
'0 6OOA.
CDS -.
CD18
180%
-B-
80%
g 100°A.
.! 120%
...
, 1000/0
I
·249
CD4
{40%
~
CD45RA
20%
0%
24 6 hours after injection
-B-
0
24 3 6 hours after injection
C
10
....
8
8
~0
01,1ntlVlA-1
6
D3,1nti-VlA-1 ~
01, TEPC-15
4
~
01,P8S
2
0
0
168 24 hours after injection
Figure 7. Effect of mAb 1B3.1 on T cell subsets: Percent of subsets among the PBMC (relative to time o = 100%) after injection of mAb TEPC-15 (A) or mAb 1B3.1 (B, representative of three experiments). C: CD4/CD8 ratio in PBMC at the indicated time points. Results for dog 1 (01) are representative of three separate experiments, for dog 3 (03) - of a single experiment.
CD4 returned to pre-injection values 24 hours after injection of mAb IB3.1, the percentage of CD8± cells in the PB remained suppressed. As a result, the ratio of CD4 to CD8 cells was markedly increased in the PB 24 hours after the injection (Fig.7c). Evaluations of these T cell subsets 1 week after the injections, revealed that the ratio had returned to the pre-injection values.
Discussion MAb 1B3.1 was previously found to recognize an epitope within the I domain of the human al integrin chain, which contains its collagen and laminin binding sites as well as cation binding sites important for integrin functions, suggesting that this mAb is a potential reagent for treating diseases T cell-mediated diseases, such as GVHD (22). The studies described here, showing that this anti-human al mAb recognizes the same integrin in canine cells, together with the demonstration that the mAb functionally blocks the adhe-
250 . 1.
BANK
et al.
sion of activated canine T cells to collagen I~ indicate that there is an homology of expression and functions of this molecule between human and canine T cells. These results thus suggest, for the first time, that dogs could serve as a model for studying the role of a 1~ 1 integrins in the pathogenesis of diseases (e.g. atherosclerosis, sarcoidosis and rheumatoid arthritis) in which al~l-integrin positive activated T cells are involved (12, 23, 24). Since the a 1~ 1 integrin is broadly distributed on fibroblasts, liver cells, endothelial cells, natural killer (NK) cells and other cellular elements, parenteral administration of a mAb to this molecule could affect many physiological systems (25-29). Furthermore, our results showing that persistent cross-linking of resting T cells, that express very low levels of the integrin, via the a 1 integrin chain using immobilized mAb significantly augments their activation, raised additional serious concerns regarding possible activating effects during repetitive administration of the mAb in vivo, when immune complexes may have formed and settled in tissues. However, our results indicate that intravenous injection of anti-integrin mAb 1B3.1 at up to 0.75 mg/Kg to healthy mongrel dogs is clinically well tolerated, and caused no apparent histo-pathological alterations in liver, kidney, or gut tissues as determined in an autopsy study done on dog 2 (Table I, and data not shown). Moreover, even repeated injections of up to 0.5 mg/Kg, spaced over greater than 1 month intervals, were clinically well tolerated. Such a dose would be expected to yield 5-10 f.1g/ml of the mAb in the dog's serum prior to distribution to extracellular fluids from the vasculature which is equivalent to the saturating concentration of the mAb used for cytofluorometric assays, and for the blocking experiments shown in Figure 5. Thus, our results suggest that short term health is unimpeded by blocking al integrins on circulating PBMC. Interestingly, increasing the dose by only 500/0, to 0.75 mg/Kg, resulted in quite a profound and prolonged decrease of circulating lymphocytes, although with no apparent clinical effect (Figure 6B). This higher concentration of the mAb may have induced a redistribution of lymphocytes among body fluids and tissues by enhancing their ability to avoid adhesion to collagen IV in basement membranes (9, 18). The gastro-intestinal hemorrhage and subsequent death of this dog after a second injection of an even higher dose of the mAb, may have been due to toxicity of the mAb to cells expressing the a 1~ 1 integrin, or to pathological effects of circulating immune-complexes, unrelated to the antigenic specificity of mAb 1B3.1. An additional mechanism could involve co-activating effects in T cells simultaneously engaged by antigens, and even direct activation of mucosal yb T cells which in vitro become induced to express cytokine receptors in response to cross linking with mAb IB3.1 even in the absence of antigenic triggering (9). The in vitro results demonstrating markedly enhanced expression ofIL-2R and CD69 in cells cross linked by mAb IB3.1, suggest that in vivo IB3.1 bound T cells could be further triggered by IL-2 in their micro-environment and may become primed for enhanced cytolytic activity mediated by CD69 (30). Since signals generated by cross linking of al integrins on T cells (as well as fibroblasts) are dependent upon activation of protein-kinases (Figure 2), inhibitors of protein kinases might be sufficent to abrogate such potentially harmful T cell activating effects of the parenterally administered mAb (31,32). While further experiments, including autopsies and in situ immunohistological studies of tissues are under way in our laboratory to determine conclusively the mechanism of the lymphopenia and gastrointestinal toxicity induced by the parenteral administration of high doses of mAb 1B3.1, the potential harmful effects should be taken into consideration when such reagents are considered for treatment of intestinal-mucosa manifestations of GVHD.
VLA-l expression and function in dogs . 251
An additional specific effect of injection of mAb 1B3.1 is the alteration of the CD4/CD8 ratio irrespective of the degree of lymphopenia induced. This shift was due to the more persistent depleting effect on CD8+ rather than CD4+ cells, which is perhaps the consequence of the higher level of expression of a 1~ 1 integrins on activated CD8+ T cell clones (9). Differential effects on CD8+ and CD4+ T cells could potentially play an important role in immune mediated diseases such as GVHD. For example, preferential migration of CD8+ T cells from the vasculature or lymph into tissues could affect the outcome of local immune responses. In summary, these preliminary studies indicate for the first time that parenteral administration of a mAb to the a 1 integrin to a large mammal is clinically feasible, when given at appropriate dosages and intervals. Furthermore, the results suggest that this potential therapy could induce differential shifts in specific T cell populations, resulting in altered ratios among the subsets in various anatomic locations, in addition to the potential blockade of specific interactions with the extra cellular matrix, and augmentation of activation induced via the TCR-CD3 complex. Whether and how this mAb, or other similar reagents, can be safely used at doses and intervals sufficient for modulation ofT cell mediated diseases in a large animal model is an important subject for future investigations. Acknowledgement This work was supported by a grant from the Israel Ministry of Health to I. BANK and I. HARDAN.
References 1. HYNES R. 1992. Integrins: versatility, modulation, and signaling in cell adhesion. Cell 69: 11. 2. HEMLER M. E. 1990. VLA proteins in the integrin family: structures, functions and their role on leukocytes. Annu. Rev. Immunol. 8: 365. 3. SHIMIZU Y., G. A. VAN SEVENTER, K. J. HORGAN, and S. SHAW. 1990. Regulated expression and binding of three VLA (~1) integrin receptors on T cells. Nature 345: 250. 4. SHIMIZU Y., G. A. VAN SEVENTER, K. J. HORGAN, and S. SHAW. 1990. Co-stimulation of proliferative responses of resting CD4+ T cells by the interaction of VLA-4 and VLA-5 with fibronectin or VLA-6 with laminin. ]. Immunol. 145: 59. 5. MIYAKE S., T. SAKURAI, K. OKUMURA, and H.YAGITA. 1994. Identification of collagen and laminin receptor integrins on murine T lymphocytes. Eur. ]. Immunol. 24: 2000. 6. HEMLER M. E., J. ]ACOBSON, M. B. BRENNER, D. MANN, and]. L. STROMINGER. 1985. VLA-1: a T cell surface antigen which defines a novel late stage of human T cell activation. Eur. ]. Immunol. 15: 502. 7. GOLDMAN R., J. HARVEY, and N. HOGG. 1992. VLA-2 is the integrin used as a collagen receptor by leukocytes. Eur. ]. Immunol. 22: 1109. 8. BANK I., M. E. HEMLER., M. B. BRENNER, D. COHEN, V. LEVY, J. BELKO, C. CROUSE, and L. CHESS. 1989. A novel monoclonal antibody, 1B3.1, binds to a new epitope of the VLA-1 molecule. Cell. Immunol. 122: 416. 9. BANK I., M. BOOK, and R. WARE. Functional role ofVLA-l (CD49a) in adhesion, cation dependent spreading, and activation of cultured human T lymphocytes. 1994. Cell Immunol. 156: 424. 10. CHAN B. M. C., J. G. r WONG, A. RAo, and M. E. HEMLER. 1991. T cell receptor-dependent, antigen-specific stimulation of a murine T cell clone induces a transient, VLA protein-mediated binding to extracellular matrix. J. Immunol. 147: 398. 11. TANAKA T., Y. OHTSUKA., H.YAGITA, Y. SHIRATORI, M. OMATA, and K. OKUMURA. 1995. Involvement of a 1 and a4 integrins in gut mucosal injury of graft versus host disease. Int. Immunol. 7: 1183.
252 . I. BANK et ai. 12. BANK I., D. ROTH, M. BOOK., R.BLOCK, ~LANGEVITZ, M.EHRENFELD, and M. PRASe 1991. Expression and functions of very late antigen 1 in inflammatory joint diseases. J. Clin. Immunoi. 11: 29. 13. BANK I., A. TANAY, A. MIGDAL, M. BOOK, and A. LIVNEH. 1995.Vy9-V02+ yO T cells from a patient with Felty syndrome that exhibit aberrant responses to triggering of the CD3 molecule can regulate immunoglobulin secretion by B cells. Clin. Immunol. and Immunopathol. 74: 162. 14. SONNENBERG A., H. JANSSEN, E HOGERVORST, H. CALAFAT, and]. HILGERS. 1987. A complex of platelet glycoproteins Ic and IIa identified by a rat monoclonal antibody. J. BioI. Chern. 262: 10376. 15. MOORE ~ E, T. OLIVRY, and D. NAYDAN. 1994. Canine cutaneous epitheliotropic lymphoma (mycosis fungoides) is a proliferative disorder of CD8+ T cells. Am. J. Pathoi. 144: 421: 12. 16. TIPOLD A., ~ E MOORE, T,W JUNGI, V. SAGER, and M.VANDEVELDE. 1998. Lymphocyte subsets and CD45RA positive T cells in normal canine cerebrospinal fluid. J. Neuroimmunoi. 82: 90. 17. LANGE T. S., A. K. BIELINSKY, K. KIRCHBERG, I. BANK, K. HERMANN, T. KRIEG, and K. SCHARFETTER-KoCHANEK. 1995. Divalent cations Mg2 +and Ca2 +differentially regulate ~1 integrin mediated adhesion of dermal fibroblasts and keratinocytes to different extracellular matrix proteins. J. Exp. Dermatoi. 4: 130. 18. BANK I., E. !{APMAN, R. SHAPIRO, G. SCHIBY, I. GOLDBERG, A. BARZILAI, H. TRAU, and H. GUR. 1999. The epidermotropic mycosis fungoides associated a1~1 integrin (VLA-1, CD49a/CD29) is primarily a collagen IV receptor on malignant T cells. J. Cutan. Pathol. 26: 65. 19. DEAS 0., C. DUMONT., M. MACFARLANE, M. ROULEAU, C. HEBIB, E HARPER, E HIRSCH, B. CHARPENTIER, G. M. COHEN, and A. SENIK. 1998. Caspase-independent cell death induced by anti CD2 or staurosporine in activted human peripheral T lymphocytes. J. Immunoi. 161: 3375. 20. LEBIEN T. W, D. R. BOUE, J. G. BRADLEY, and]. H. KERSEY. 1982. Antibody affinity may influence antigenic modulation of the common acute lymphoblastic leukemia antigen in vitro. ]. Immunol. 129: 2287. 21. LAEMMLI U. K. 1970. Cleavage of structural proteins during the assembly of the head ofbacteriophage T4. Nature (Lond.) 227: 680. 22. KERN A., R. BRIESEWITZ, I. BANK, E. E. MARCANTONIO. 1994. The role of the I domain in ligand binding of the human integrin a 1~ 1. J. BioI. Chern. 269: 22811. 23. HANSSON G. K., J. HOLM., and L. JONASSON. 1989. Detection of activated T lymphocytes in the human atherosclerotic plaque. Am. ]. Pathoi. 135: 169. 24. SALTINI C., M. E. HEMLER., and R. G. CRYSTAL. 1988. T lymphocytes compartmentalized on the epithelial surface of the lower respiratory tract express the very late activation antigen complex VLA-1. Clin. Immunoi. Immunopathoi. 46: 221. 25. PEREZ-VILLAR J. J., I. MELERO, A. GISMONDI, A. SANTONI, and M. LOPEZ-BoTET. 1996. Functional analysis of alpha 1 beta 1 integrin in human natural killer cells. Eur. J. Immunol. 26: 2023. 26. KONO Y" S. YANG, M. LETARTE, and E. A. ROBERTS. 1995. Establishment of a human hepatocyte line derived from primary culture in a collagen gel sandwich culture system. Exp. Cell. Res. 221: 478. 27. SOBEL R. A., J. R. HINOJOZA., A. MAEDA, and M. CHEN. 1998. Endothelial cell integrin laminin receptor expression in multiple sclerosis lesions. Am. J. Pathoi. 153: 405. 28. RACINE-SAMSON L., D. C. ROCKEY., and D. M. BISSEL. 1997. The role of alpha1betal integrin in wound contraction. A quantitative analysis of liver myofibroblasts in vivo and in primary culture. J. BioI. Chern. 272: 30911. 29. TAWIL N. J., M. HOUDE, R. BLACHER et ai. 1990. Alpha 1 beta 1 integrin heterodimer functions as a duallaminin/collagen receptor in neural cells. Biochemistry. 29: 6540. 30. BORREGO E, M. J. ROBERTSON, J. RITZ, J. PENA, and R. SOLANA. 1999. CD69 is a stimulatory receptor for natural killer cell and its cytotoxic effect is blocked by CD94 inhibitory receptor. Immunology 97: 159.
VLA-1 expression and function in dogs.
253
31. POZZI A., K. K. WARY, F. G. GIANCOTTI, and H. A. GARDNER. 1998. Integrin alpha1beta1 mediates a unique collagen-dependent proliferation pathway in vivo. ]. Cell. BioI. 142: 587. 32. RAVANTI L., ]. HEINO., C. LOPEZ-OTIN, and V. M. KAHARI. 1999. Induction of collagenase-3 (MMP-13) expression in human skin fibroblasts by three dimensional collagen is mediated by p38 mitogen activated protein kinase. ]. BioI. Chern. 274: 2446. Dr. ILAN BANK, Dept. Medicine F, Chaim Sheba Medical Center, Tel Hashomer, Israel 52621. Fax: ++97235302114, e-mail: [email protected]