WO1993010798A1

WO1993010798A1 - Monoclonal antibody directed against g-cam and method of using it

Info

Publication number: WO1993010798A1
Application number: PCT/US1992/010453
Authority: WO
Inventors: Eldon E. Geisert, Jr.
Original assignee: Uab Research Foundation
Priority date: 1991-12-04
Filing date: 1992-12-04
Publication date: 1993-06-10

Abstract

A monoclonal antibody directed against a novel cell adhesion molecule of glial cells has been developed. The glial cell adhesion molecule is characterized by a molecular weight of around 106 kDa. The antibody is specific for the glial cell adhesion molecule and is marked by an ability to disrupt the formation of gliotic scars in damaged nerve tissue.

Description

MONOCLONAL ANTIBODY DIRECTED AGAINST G-CAM AND METHOD OF USING IT

FIELD OF THE INVENTION

The present invention is directed to a monoclonal antibody (hereafter called AMPl) against a glial cell adhesion molecule and a method of using AMPl to regenerate nerve growth in damaged nerve tissue.

BACKGROUND OF THE INVENTION

Glial cells play an important role in the development and maintenance of the central nervous system. When injury occurs to the nervous system, glial cells are known to form gliotic scars around the damaged area. The gliotic scars are thought to be responsible for blocking subsequent growth of axons across the scarred area. Such a condition can result in the permanent loss of nerve function, including paralysis or brain damage depending on the location of the scarred region.

SUMMARY OF THE INVENTION

The present invention is able to disrupt gliotic scar formation, thereby permitting regeneration of nerve growth in damaged nerve tissue. The invention is based on the discovery of a novel glial cell adhesion molecule (hereafter called G- CAM) which promotes gliotic scar formation. By developing an antibody directed against G-CAM, it was discovered that scar formation could be blocked by adding the antibody to damaged nerve tissue, thus allowing the nerve tissue to regenerate. BRIEF DESCRIPTION OF THE FIGURES

Figure 1. Immunofluorescence micrographs of confluent monolayers of rat type I astrocytes labeled with: mouse anti- rat G-CAM monoclonal antibody AMPl (A) ; rabbit anti-bovine GFAP (B) ; rabbit anti-rat N-CAM antibody 161 (C) ; and rabbit anti- human α₅B-_| integrin (D) . Micrographs A and B are of the same field. Cells in A were briefly reacted with 40 ug/ l AMPl in Eagle's basal medium with 10% fetal calf serum, pH 7.3 at 25°C, fixed with 4% paraformaldehyde, and labeled with FITC-conjugated rat anti-mouse IgG secondary antibody (Boehringer Mannheim) . Cells were then permeabilized with 95% ethanol:5% acetic acid at -20°C, followed by anti-GFAP (Accurate Chemical and Scientific Corp.) and RITC-conjugated goat anti-rabbit IgG adsorbed to normal rat IgG (Hyclone) . Cells in C and D were fixed with 4% 5. paraformaldehyde prior to i munostaining. Scale bar, 50 urn.

Figure 2. Im unoprecipit tion and immunoblots of astrocytic proteins under reducing conditions. Lanes A, B, and C: 7.5% SDS- PAGE fluorograph of i munoprecipitated ³⁵S-methionine-labeled proteins using different rabbit antisera. Lane A: precipitation 0 with pre-immune serum. Lane B: precipitation with polyclonal antibody 4080 against the conserved intracellular domain of the β-, integrin subunit. Lane C: precipitation with a polyclonal antibody against the human fibronectin receptor. Also present are β-, (115 kDa) and pre-β (100 kDa) bands and larger associated α subunits. Lanes D, E, and F: immunoblots from 4-16% SDS-PAGE of astrocyte-proteins reacted with antibodies against different adhesion molecules. Lane D: rat astrocyte proteins reacted with rabbit antiserum 161 against N-CAM. Lane E: chicken astrocyte proteins reacted with anti-N-cadherin mAb, NCD2. Lane F: rat astrocyte proteins reacted with AMPl. Relative molecular weights are indicated to the left.

Figure 3. Phase contrast micrographs of confluent cultures of rat astrocytes illustrating the effect of AMPl treatment on astrocytic cell-cell junctions. Treatment with 150 ug/ml AMPl in Ca^""^" and Mg^** free Hank's balanced salts with 10 mM HEPES buffer (pH 7.5) for 15 minutes resulted in disruption of the monolayer (A) . Similar treatment with the mAb 13-38 directed against an extracellular domain of N-CAM had little effect (B) . Scale bar_r 50 urn.

DETAILED DESCRIPTION OF THE INVENTION

In one embodiment, the invention is a glial cell adhesion molecule, G-CAM. Another embodiment of the present invention is the monoclonal antibody directed against G-CAM, referred to as AMPl. A third embodiment of the present invention is a hybridoma capable of producing AMPl. A fourth embodiment is a method of regenerating nerve growth in damaged nerve tissue comprising administering an effective amount of AMPl to a mammal in need thereof. A fifth embodiment is a pharmaceutical composition containing a pharmacologically effective amount of AMPl.

The mAB AMPl was produced from mice immunized with cultured rat glial cell plasma membrane proteins (M.W. 100-160 kDa) cut from a Coomassie blue-stained polyacryla ide gel, according to the culturing method of Geisert et al., "Changing interactions between astrocytes and neurons during CNS maturation," Dev. Biol.. 143(1990): 335-345. Following a final boost of whole rat brain white matter, the mice were anesthetized with Ketalar (50 mg/kg) , the lymph nodes were dissected, and the resulting lymphocytes were fused with AG8 myeloma cells using PEG 1,000 (Boehringer Mannheim) . During the screening of the resulting hybridomas by immunohistochemical and immunoblot methods, a mAB, AMPl, was identified that labeled rat glia. The hybridoma producing AMPl has been deposited with the American Type Culture Collection (at 12301 Parklawn Drive, Rockville, Maryland, USA) and given the deposit number ATCC .

When confluent cultures of rat neonatal astrocytes were stained with the mAB AMPl (for immunohistochemical methods see Geisert et al., "Changing interactions between astrocytes and neurons during CNS maturation," Dev. Biol. , 143(1990) : 335-345) , regions of cell-cell contact were intensely labeled (Fig. 1) , indicating that the AMPl antigen is most likely a cell adhesion molecule. Since astrocytes .are known to express N-cadherin, N- CAM, and members of the β-, integrin family, the following series of experiments was designed to determine if the AMPl antigen was a previously described adhesion molecule.

The cell adhesion molecule N-CAM has a variety of isoforms with three different molecular weights. Astrocytes predominantly express the 140 kDa and 120 kDa isoforms of N-CAM with minimal amounts of the 180 kDa form (Noble et al., "Glial cells express N-CAM/D2-CAM-like polypeptides in vitro," Nature. 316(1985) : 725- 728) . Based simply on its molecular weight, the 106 kDa AMPl antigen appeared not be N-CAM (Fig. 2) . When adjacent strips of an immunoblot of rat astrocyte proteins (for detailed methods, see Geisert et al. , "Expression of microtubule-associated protein 2 by reactive astrocytes," Proc. Natl. Acad. Sci.. 87(1990): 3967-3971) were probed with antibodies directed against N-CAM, including two different monoclonal antibodies, 3F4 and 3G6 (Akeson et al. , "Smooth muscle cells transiently express N-CAM, 5. Mol. Brain Res.. 4(1988): 107-120; and Williams et al., "Individual neural cell types express immunologically distinct N-CAM forms," J. Cell Biol.. 101(1985): 36-42), and one rabbit antiserum, 161 (Small et al. , "Polypeptide variation in an N- CAM extracellular immunoglobulin-like fold is developmentally 10 regulated through alternative splicing," Neuron, 1(1988): 639- 655) , labeled bands corresponding to the three molecular weights of N-CAM were identified (Fig. 2) . The 140 kDa isoform was the most prominently labeled band. On an adjacent strip of the immunoblot probed with AMPl, a single 106 kDa protein was labeled 15 (Fig. 2) . The protein recognized by AMPl was smaller than any of the N-CAM isoforms. Furthermore, AMPl did not blot to protein samples from mouse L-cells that were transfected with the rat gene coding for 180 kDa isoforms and 140 kDa isoforms of N-CAM, nor did AMPl label these cells by immunofluorescence. These data 20 demonstrate that the AMPl antigen is not N-CAM.

N-cadherin is also present in regions of astrocyte-astrocyte contacts. Cadherins characteristically require Ca^**. Therefore, the calcium sensitivity of the AMPl antigen was tested. Rat astrocytes were placed in serum-free medium containing a Ca^** 25 chelator (5 mM EDTA) and 0.005% trypsin for 15 minutes. The cultures were fixed and immunostained with AMPl. No decrease in immunofluorescence was observed when these astrocytes were compared to control cultures in which EDTA was omitted. When similar cultures were examined by immunoblotting procedures, 30 there was no obvious difference in the intensity of the 106 kDa band recognized by AMPl between cultures incubated with or without EDTA. In parallel cultures of chick astrocytes treated in a similar manner, there was na near complete loss of N- cadherin immunolabeling by mAb NCD2 directed against chicken N- 35. cadherin after treatment with 5 mM EDTA and 0.005% trypsin. Furthermore, immunoblot analysis demonstrated that chick N- cadherin has a relative molecular weight of 127 kDa, considerably larger than the AMPl antigen (Fig. 2) . These data indicate that the AMPl antigen is not a cadherin, for it is not susceptible to proteolysis in low calcium and has a significantly lower molecular weight than chick N-cadherin. 5. The third family of known adhesion molecules found on astrocytes is the integrins. These receptors are heterodimers, composed of one a and one β subunit, and are primarily involved in heterophilic binding to extracellular matrix components such as laminin, fibronectin, and collagen, although recent evidence 0 indicates that some members of the β, integrin family are involved in cell-cell interactions (Larjava et al., "Novel function for B₁ integrins in keratinocyte cell-cell interactions," J. Cell Biol.. 110(1990): 803-815). Integrin subclasses expressed by astrocytes also appear to be functioning 5 in cell-cell binding (Fig. 1) . Therefore, the possibility that AMPl was recognizing a β-, integrin was investigated. Triton X- 100 extracts of cultured astrocytes were immunoprecipitated with two different polyclonal antisera directed against the β-, integrin subunit: a rabbit polyclonal antiserum 4080 (Darribere et al., "In vivo analyses of integrin β-, subunit function in fibronectin matrix assembly," J. Cell Biol. , 110(1990): 1813- 1823) , directed against the highly conserved intracellular domain of the β-_! integrin, and a rabbit antiserum developed against the human fibronectin receptor (o;₅B₁ integrin) (Larjava et al., "Novel function for β_, integrins in keratinocyte cell-cell interactions," J. Cell Biol.. 110(1990): 803-815). When these immunoprecipitated proteins were run on 8% polyacrylamide gels, transferred to nitrocellulose and reacted with the polyclonal antibody to the β_, subunit, the β-, subunit was labeled on the blots; however, on adjacent strips of the same immunoblot, AMPl did not recognize any protein. In addition, repetitive immunoprecipitation with antibody 4080 failed to deplete the AMPl antigen from the Triton X-100 extract, as determined by immunoblotting of SDS-denatured samples. To confirm that the a subunits were present in the sample of immunoprecipitated proteins, cultures of astrocytes were radiolabeled with ³⁵S- methionine (ICN) and immunoprecipitated. SDS-PAGE fluorograrhy revealed radiolabeled proteins at the molecular weights of the β-, and the pre-β subunits (Fig. 2) . Additional higher molecular weight bands were also observed, corresponding to the associated α subunits. Attempts to directly immunopreσipitate AMPl antigen 5. from the same radiolabeled Triton X-100 extracts were not successful, indicating that the antibody did not recognize the antigen in solution with this detergent. Finally, when the rabbit anti-human fibronectin receptor antiserum was added to the living cultures of astrocytes, the labeled integrins reorganized on the surface of the cells as patches, while subsequent staining of the same culture with AMPl demonstrated that its antigen remained in the astrocyte intercellular junctions. Taken together these data demonstrate that the AMPl antigen is not a member of the β., integrins. To better define the role of the AMPl antigen in astrocyte intercellular junctions, confluent cultures of neonatal astrocytes were treated with AMPl. The culture dishes were first rinsed in Hank's balanced salts and placed on a Nikon inverted microscope. When 150 ug/ml of AMPl in HEPES buffered Hank's balanced salts were added to the cultures, there was a rapid change in the regions of contact between the astrocytes. Within five minutes of the addition of AMPl, membranes at the regions of cell-cell contact appeared to ruffle. By ten minutes after the addition of the antibody, the cells began to pull apart and continued incubation resulted in gaps in the previously confluent monolayer (Fig. 3) . this disruption was not observed in when the cells were maintained in either Hank's balanced salts or following the addition of 150 ug/ml of irrelevant mAbs directed against either the 160 kDa neurofilament protein (mAb 8-152) or against the extracellular domain of rat N-CAM (mAb 13-38, Fig. 3) . When the AMPl treated cultures were returned to normal culture medium, the cells returned to their normal appearance within two hours, indicating that the antibody was not toxic to the cells. Taken together these data show that the protein recognized by AMPl is not N-CAM, N-cadherin, or a member of the β., integrin family. The AMPl antigen is not present on cultured embryonic neurons, but is expressed on cultured astrocytes and cultured oligodendrocytes as demonstrated by indirect immunohistochemistry. The AMPl antigen is involved in stabilizing glial cell intercellular junctions, and is expressed by cultured 5. rat glial cells but not neurons.

It is contemplated that AMPl can be used in connection with other agents known to promote nerve growth. For example, calcium chelators, such as EDTA (ethylenediaminetetraacetic acid) are known to be inhibitors of N-cadherin, another adhesion molecule 0 present in regions of astrocyte-astrocyte contacts.

In addition, it has been found that inserting collagen- containing support matrices into the site of injury is useful for providing a structure through which nerve cells can grow. See Gelderd, "Evaluation of blood vessel and neurite growth into a collagen matrix placed within a surgically created gap in rat spinal cord," Brain Research. 511(1990): 80-92.

Suitable collagen-containing support matrices which can be used according to the present invention include collagen and mixtures of collagen and Matrigel (produced by Collaborative Research Incorporated) . Matrigel contains mostly laminin and type IV collagen and includes small amounts of nidogen, heparin sulphate, and entactin. Preferred collagens are type I and type IV collagen. The collagens may be taken from any mammalian source, but to insure compatibility, the source of the collagen should preferably match the subject being treated. For example, a collagen-containing support matrix for a human patient should comprise human collagen.

Optionally, stabilizers such as heparin sepharose beads may be added to the collagen-containing support matrix. In addition, cultured astrocytes may be added to the collagen to provide a supplemental source of nerve tissue at the site of injury.

A preferred method is to first administer AMPl and EDTA at the site of injury in an amount sufficient to disrupt scar formation, and thereafter introduce the collagen-containing support matrix to provide support for the neurons that are now able to grow freely as a result of the prior administration of AMPl. It is further contemplated in this method that AMPl can be delivered to the site of injury in a carrier solution, comprising conventional physiologically acceptable carriers, excipients, diluents, and mixtures thereof. A preferred carrier solution is a phosphate buffered saline solution having a pH of 7.4.

The use of AMPl in treating nerve tissue injury is described more fully by the examples below, although the present invention is in no way limited to these particular examples.

Example 1

Nerve tissue injury and AMPl treatment

Rats were anesthetized and a large lesion was made in the brain using a scalpel blade, according to the procedure of Geisert et al., "Antiserum-Induced Growth of Axons Across Lesions of the Adult Rat Brain," Brain Research Bulletin. 15(1985) : 19- 28. The location of the lesion was 3.0 mm posterior to bregma and 0.5 mm lateral to the midline and extends through the lateral boundary of the brain. The lesion passed through regions of gray and white matter. A cannula was lowered into the center of the injured area (4 mm lateral to the midline and 3.5 mm below the cortical surface) and cemented to the skull using dental cement. This provided access to the injured area without a second surgery and was meant to mimic the lowering of an injection needle into the site of injury as would be the case with human treatment. The animal was allowed to survive for a minimum of two months before the treatment began.

The animals with chronic injury were anesthetized, the treatment cannula was opened, and the treatment was delivered through the cannula directly into the site of injury. Treatment took place over the course of one to three hours, and was delivered in a total volume of 200-300 ul/treatment, of which 2 to 4 mg of AMPl was used and 2.5 to 5 mM of EDTA was used per treatment. The carrier solution was phosphate buffered saline having a pH of 7.4. Control animals were only given the carrier solution.

After the treatment, the animals were sacrificed. However, in order to preserve the brain tissue, the animals were deeply anesthetized, following which they were perfused through the heart with saline and then 4% paraformaldehyde in 0.1 M phosphate buffer. Brains were postfixed overnight then cryoprotected with 30% sucrose before sectioning. Horizontal frozen sections (40 5. um thick) of brain tissue through the lesion area were examined using standard histological stains and immunohistochemical methods to detect reactive astrocytes and other molecules. Results

In the treated animals, the astrocytic scar was disrupted. 0 The astrocytes did not appear to be tightly associated with each other and the external cell layer (ependyma) was removed from the surface of the scar. This left the animal with a disrupted scar and a cavity separating the two sides of the brain. Processes of cells were observed dangling into the lesion cavity. The 5 control animals showed no scar disruption.

Example 2 Introduction of a collaσen-containinq support matrix

In other animals treated according to the previous example, a collagen-containing support matrix was introduced into the cavity at the site of injury 30 minutes to an hour after the initial AMPl treatment. The collagen-containing support matrix was comprised mainly of type I collagen and Matrigel (containing laminin and type IV collagen with small amounts of nidogen, heparin sulphate and entactin) , as in Gelderd, "Evaluation of blood vessel and neurite growth into a collagen matrix placed within a surgically created gap in rat spinal cord," Brain Research, 511(1990): 80-92, and in some animals cultured astrocytes and/or heparin sepharose beads were added to help stabilize the collagen-containing support matrix.

The matrix materials (generally 30 ul solution containing equal amounts of type I collagen and Matrigel with 1-2 X 10⁵ astrocytes and heparin sepharose beads) were injected into the cavity through the implanted cannula. The matrix materials were kept on ice prior to injection. The cannula was capped 30 minutes after injection of the matrix materials. The treated animals were then perfused and processed as set forth above in

Example 1.

Results

Xn many regions, the matrix material had been integrated into the brain regions and virtually no scar was present. Once the scar was disrupted by the initial treatment in Example 1, the scar did not reform even after the matrix was introduced. In some regions, the brain tissue and treatment area were indistinguishable. When the tissue was stained, there was clear evidence of exuberant axonal growth into the regions of injected matrix as demonstrated by the large number of growth cones.

10

Claims

I claim:

1. A monoclonal antibody having specificity for a glial cell adhesion molecule and the ability to disrupt gliotic scar formation. 5

2. A monoclonal antibody according to claim 1, wherein the glial cell adhesion molecule is characterized by a molecular -, weight of about 106 kDa.

3. A hybridoma capable of producing a monoclonal antibody, wherein said monoclonal antibody is specific for a glial cell

10 adhesion molecule and has the ability to disrupt gliotic scar formation.

4. A method for promoting nerve growth in damaged nerve tissue in a mammal comprising administering a composition comprising a pharmacologically effective amount of AMPl and a

15 physiologically acceptable carrier.

5. A method according to claim 4, wherein the composition further comprises a pharmacologically effective amount of a calcium chelator.

6. A method according to claim 5, wherein the calcium 20 chelator is EDTA.

7. A method according to claim 6, wherein the carrier is a phosphate buffered saline solution having a pH of 7.4.

8. A method according to claim 4, wherein said mammal is a human.

25 9. A method for promoting nerve growth in damaged nerve tissue in a mammal comprising administering a composition comprising a pharmacologically effective amount of AMPl, a pharmacologically effective amount of a calcium chelator, and a physiologically acceptable carrier; 30 and introducing a collagen-containing support matrix into the site of damaged nerve tissue, wherein the support matrix comprises equal amounts of type I collagen and Matrigel.

10. A method according to claim 9, wherein said matrix further comprises astrocytes. , 35 11. A method according to claim 10, wherein the calcium chelator is EDTA.

11

12. A method according to claim 11, wherein said matrix further comprises heparin sepharose beads.

13. A method according to claim 9, wherein said mammal is a human.

12