WO1999023113A2 - Rho family antagonists and their use to block inhibition of neurite outgrowth - Google Patents

Abstract

This invention provides for the use of antagonists of Rho, or proteins related to Rho as therapeutic targets for agents designed to block growth inhibition by myelin or myelin proteins. One embodiment pertains to the use of Rho antagonists that foster axon regeneration in the central nervous system. The therapeutic agent or antagonist can be small molecules, proteins or peptides, or any agent that binds to Rho or its family members to inactivate this patheway. Embodiments include: the use of the Rho regulatory pathway as a target for Rho antagonists; the use of GDP dissociation inhibitors (GDIs) inhibit the dissociation of GDP from Rho, and thereby prevent the binding of GTP necessary for the activation of Rho; the use of Rho specific GTPase activating protein (GAPs) as targets for the regulation of Rho activity; the use of agents that promote Rho binding to GDI, and block Rho binding to the plasma membrane are also considered within the scope of this invention; the use of C3 transferase and related toxins such as A and B, with related Rho-inhibitory activity to inactivate Rho and stimulate axon growth; the use of dominant negative forms of Rho, used to inactivate Rho, to foster axon growth.

Description

RHO FAMILY ANTAGONISTS AND THEIR USE TO BLOCK INHIBITION OF NEURITE OUTGROWTH

FIELD OF INVENTION

This invention relates to the regulation of growth of neurons in the Central Nervous System.

BACKGROUND

Followώglimimamlheadultcentralnervous system(CNS) ofmammals, injuredneurons do otregenerate theirtransected axons. An important barrier to regeneration is the axon growth inhibitory activity that is present in CNS myelin and that is also associatedwith theplasmamembrane of oligodendrocytes, the cells that synthesize myelin in the CNS (see Schwab M.E., etal, (1993) Ann. Rev. Neurosci, 16, 565-595, forreview) . The growth inhibitory properties of CNS myelin have been demonstratedin anumber of differentlaboratories by awide variety of techniques, including plating neurons onmyelin substrates or cryostat sections of white matter, and observations of axon contact with mature oligodendrocytes (Schwab, M.E., et al, (1993) Annu. Rev. Neurosci. 16, 565-595). Therefore, it is well documented that adult neurons cannot extend neurites over CNS myelin in vitro.

Ithas also been well documented thatremovingmyelin in vivo improves the success of regenerative growth over the nativeterrain of the CNS. Regeneration occurs after irradiation of newbomrats, aprocedurethat kills oligodendrocytes and prevents the appearance of myelin proteins (Savio and Schwab, ( 1990) Neurobiology l, 4130-4133). After such aprocedure in rats and combined with a corticospinal tract lesion, some corticospinal axons regrow long distances beyαndthe lesions. Also, in a chick odel of spinal cordrepair, the onset of my ehnation correlates with aloss of its regenerative ability of cutaxons(Keirstead, et al, (1992) Proc. Nat. Acad. Sci. (USA) 89, 11664-1-1668). The removal of myelin with anti-galactocerebroside and complement inthe embryonic chick spinal cord extends the permissive period foraxonalregeneration. These experiments demonstrate agoodcorrelationbetweenmyelinationandthe failure of axons to regenerate in the CNS.

Myelin inhibits axon growth because it contains atleastseveraldifferentgrowth inhibitory proteins. Ithas been well documentedby us and by others thatmyeHn-associatedglycoprotein (MAG) has potent growth inhibitory activity, both in vitro and in vivo (McKerracher, L., et al, (1994) Neuron 13. 805-811; Mukhopadhyay, G, et al, (1994) Neuron 13, 805-811; Li, M., et al, (1996) /. Neurosci. Res. 46, 404414; Schafer,M., etal, (1996) Neuron lβ, 1107-1113.). Ahighmolecularweightinhibitory activity has been characterized by Schwab and collaborators, and neutralization of this activity with the IN-1 antibody allows some axons to regenerate in white matter (S chwab, M.E. , et al , ( 1993^") Ann. Rev. Neurosά., 16, 565-595; Bregman,B., etal, (1995) .fare378,498-501.). We also have evidence that thereisanadώtionalgrowminhibitoryproteininmyelin(Xiao,Z., etal, (1991) Soc. Neurosci. Absts. 23, 1994). Clearly, there are multiple inhibitory proteins thatstopaxonregeneration in j ammahanCNS myelin.

In addition to the myelin-derivedinhibitors there are also other growth inhibitory molecules expressed in the d tmammahanCNS. Tenacinisagrowthinljibitoryproteinthatisexpressedinsomeurmyelinated regions of the CNS (Bartsch,U., et al, (1994) J. Neurosci. 14, 4756 -4768) and after lesion tenascin is expressedby astrocytes that border the lesion site (Aj emai and David ( 1994) J. Comp.Neurol 34*0. 233 -242) . Also growth inhibitory proteins that are proteogly cans are expressedby reactive astrocytes , and these proteins form abaπier to regeneration at the glial scar (McKeon and Silver ( 1995) Exp. Neurol 136, 32 - 43).

While axons damagedintheCNS invivo do not typically regrow, there have been some reports oflong distoceaxonextensioninadultwhitematter.Suchgrowthhasbeenobservedfollσw^ graftedneuraltissue(Wictorin,K., e^ Neurosci. 14, 1596-1612.; Isacson, O. and Deacon, T.W. (1996) Neuroscience 75, 827-837), suggesting that embryonicneurons primed forrapidextensionofaxons may be less susceptible to growth inhibition. S ome embryonic neurons are not sus ceptible to MAG (Mukhopadhyay, G., etal, (\ 994) Neuron tt, 805-81 l),butmost embryonic eurons are inhibitedbytheothermyelin inhibitors (Schwab, M.E., etal, (1993) Ann. Rev. Neurosci, 16, 565-595). Therefore, in the cases when axons are able to extend onmyelin, signalingthroughintracellularpathways may play animportantrole in stimulating, or blocking the inhibition of axon growth. For example, it is known that laminin is able to stimulate rapid neurite growth (Kuhn, T.B., et al, (1995) Neuron U, 275-285), and we have documented that when laminin is present in sufficient concentration, neurites can extend directly onmyelin substrates . These fmdkgssiiggestthepossibihtytø^^ for laminin, is sufficientto allow axon growth onmyelin. Similarly,ithasbeendocumentedthatwhentheadhesionmor culeLl is express ed ectopically on astrocytes, it can partially overcome theirnon-permissive substrate properties (Mohajeri, M.H., et al, (1996) Eur. J. Neurosci _8, 1085-1097). Therefore, neurons can, under appropriate conditions, grow axons oninhibitory substrates, suggesting that the balance of positive to negative growth cues is a critical determinant for the success or failure of axon regrowth after injury.

Growth inhibitory proteins typically cause growth cone collapse, a process that causes dramatic rearrangementstothegrowthconecytoskeleton(Bandtlow,C.E.,etal.,(1993) Science259.80-83; Fan, J., etal, (1993) J. CellBiol U\, 867-878; Li, M., etal., (1996) J. Neurosci. Res. 46, 404-414). One family of proteins that has been imphcatedin receptor-mediated signaling to the cytoskeleton is the small . GTPases oftheRhofamily(Hall,A. (1996) Λrøι. Rev. CellBiol 10,31-54). In non-neuronal cells it has been clearly documentedthat mutations in Rho family members that include Rho, Rac and cdc42, affect adhesion, actin polymerization, andώeformation of 1-ιmempodiaandfιlopoda- which -rre all process importanttomotiUty(Nobes,C.D. andHall,A.R. (1995) Cell , 53-62.). There is now good evidence thatmembers of the Rho family regulate axon outgrowth in development. Mutations inRho-relatedfamily members blocktheextensionofaxonsinDrosophila(Luo,L.,etal.,(1994) Genes Dev.8, 1787-1802) -fflddisraptaxonalpathfindinginC. elegans (Zipkin,I.L., etal., (1997) Ce//90, 883-894.). Morerecently ithasbeenshown that theguidancemoleculecollapsin acts throughaRac-dependentmechanism(Jin,Z. andStrittmatter, S.M. (1997) J. Neurosci. 17, 6256-6263). Intransgenic mice that express constitutively active Rac in Purkinj e cells , there are alterations in the development of axon terminals and dendritic arborizations (Luo,L., etal, (1996) Nature379, 837-840.). Consistentwith the observations invivo, itwas found that dominantnegative Rac expressed inPC12 cells disrupts neurite outgrowth in response toNGF(Hutchens,J.A., etal, (\991) Molec. Biol Cell 8, 481-500.). Also, treatment of PC12 cells with lysophosphatidic acid, amitogenicphospholipid, causesneuriteretractionthatismediatedbyRho (Tigyi,G, etal, (1996) J.Neurochem. 66, 537-548.). Therefore, differentmembersofthe Rho family can exert distinct effects on neurite growth, and in PC 12 cells the activation of Rho is correlated with growth cone collapse. Innon-neuronal cells, Rhoparticipates inintegrin-dependent signalling (Laudanna, C.. ef α/., (1996) Sc/ence 27^ ^

12542- 12548). The possibility that Rho might play arole within the my elin-derived growth inhibitory system has been studied (Jin, Z. andStrittmatter, S.M. (1997) /. Neurosci. 1 7, 6256-6263). Itwas concluded, however, that the inhibitory effects of myelin are not mediated by Rho family members.

Aneedremains for ameans of inactivating the multiple inhibitory proteins presentinmyelinthatprevent axonal regrowth after injury in the CNS.

This backgroundjMormationisprovidedforthepurpose ofmaking known informationbeUeved by the applicanttobeofpossiblerelevancetothepresentinvention. No admission is necessarily intended, nor . shouldbe∞nstaed,tl anyoftheprecedinginformationconstiω

SUMMARY OF THE INVENTION

Thepresentinventionrelates to antagonists andinhibitors to members of the Rho family of proteins and diagnostic, therapeutic, andresearchuses for each of these aspects. In particular, members oftheRho family of proteins serve as atherapeutic target to foster regrowth of injured or degenerating axons inthe

CNS.

In accordance with thepresent invention, apreferred embodimentrelates to antagonists and inhibitors of members of the Rho family of proteins and their use as ameans ofblocking a common signalingpathway usedby the diverse growth inhibitory molecules . The antagonists andinhibitors may be mutated forms of Rho andbiologically active (Rho family-inhibitory) fragments, peptides, C3 andbiologically active (Rho family-inhibitory) fragments, or small molecules such as Y-27632.

In yet afurther aspect of the present invention, Rho family member proteins can be used to design small molecules that antagonize and inhibit Rho family proteins, to blockinhibition of neurite outgrowth. In another aspect of the present invention Rho family members can beusedto design antagonist agents that suppress the myelin growth inhibitory system. These antagonist agents can be used to promote axon regrowth and recovery from trauma or neurodegenerative disease.

This invention provides for the use of Rho, or proteins related to Rho as therapeutic targets for agents designedtoblockgrowth ibitionbymyelinormyehnproteins. Oneembodimentpertainstotheuseof Rho antagonists that foster axon regeneration in the central nervous system. The therapeutic agent or antagonist can be small molecules, proteins or peptides, or any agent that binds to Rho or its family - members to inactivate this pathway.

Another embodiment pertains to the us e of the Rho regulatory pathway as a target for Rho antagonists . This pathway involves the GDP/GTP exchangeproteins(GEPs). Rho has two interconvertible forms, GDP-boundinactive, and GTP-boundactive forms. The GEPs promote the exchange of nucleotides and thereby constitute targets for regulating the activity of Rho.

In another embodiment GDP dissociation inhibitors (GDIs)inhibitthe dissociation of GDP fromRho, and thereby preventthe binding of GTP necessary forthe activation of Rho. Therefore, GDIs are targets for agents thatregulate Rho activity. The GTP-bound active Rho can be convertedto the GDP-found inactive form by a GTPase reaction that is facilitated by its specific GTPas e activating protein (GAP) . Thus, another embodiment pertains to the use of GAPs as targets forthe regulation of Rho activity. Such inhibitors could block exchange of the GTP/GDP cycle of Rho activation/inactivation.

Another embodiment pertains to the factthat Rho is foundinthe cytoplasm complexedwith a GTPase inhbitingprotein(GDI). Tobecomeactive,Rhobinds GTPandistranslocatedtothemembrane. Thus, agents that promote Rho binding to GDI, and block Rho binding to the plasma membrane are also considered within the scope of this invention.

Yet another embodiment pertains to the observation that abacterialmon-ADP ribosyltransferase, C3 transferase, ribosylates Rho to inactivate the protein. Thus this embodiment pertains to the use of C3 transferaseto inactivate Rho andstimulate axon growth. Likewise, other bacterial toxins, such as toxins A and B , with related Rho-inhibitory activity are considered to be within the scope of this invention.

Moreover, various mutations of the Rho protein can create dominantnegative Rho, which can interfere with the biological activity of endogenous Rho inneurons. Thus, yetafiirther embodimentof this invention pertains to the use of dominant negative forms of Rho, used to inactivate Rho, to foster axon growth.

In accordance with another aspect of thepresent invention, there is provided an assay methoduseful to identify Rho family member antagonist agents that suppress inhibition of neuron growth, comprising the steps of: a) culturingneurons onagrowthpermissivesubsfratethatincorporates a growth-inhibiting amount of a Rho family member; and b) exposingthe culturedneurons of step a) to acandidate Rho family member antagonist agentin an amount and for a period sufficient prospectively to permit growth of the neurons; thereby identifying as Rho family antagonists the candidates of step b) which elicitneurite outgrowth from the cultured neurons of step a).

In accordance with another aspect of the present invention, there is provided amethod to suppress the inhibition ofneuron, comprising the steps of delivering, to the nerve growth environment, aRho family antagonist in an amount effective to reverse said inhibition.

In another embodiment, kinases activated by Rho, such as Rho-associated kinase, are antagonist candidates. Thus, compounds such as Y-27632 (U.S. PatentNo.04997834), thatblockRho-associated kinase activity, thereby inactivating the Rho signaling pathway, are also embodiments of thi- invention. Thus, the use other compounds within this family of compounds as describedinU.S. PatentNo.04997834 that inhibit Rho kinase are also considered within the scope of this invention.

In yet another embodiment, a kit is provided comprising components necessary to conduct the assay method useful to screen Rho family antagonist agents .

Various other objects and advantages of the presentinvention will become apparent from the detailed description of the invention.

BRIEF DESCRIPTION OF FIGURES

Figure 1 shows results oftreatmentwith C3 to stimulate neurite outgrowth on inhibitory MAGsubstrates. A) PC12 ceUsplatedonMAGremainedroundedanddidnotextendneurites. B) Cells plated on MAG in the presence of C3 grew neurites . C) PC 12 cells plated onpolylysine (PLL) substrates as apositive control. Figιπ^,e2showstheroleofintegrinskoverriώnggrowthMιibiti Theanti- αl integrin function blocking antibody, 3A3 , was used to determine if integrin functionis necessary for laminin to override growminhibitionbymyelin or MAG. For experiments onmyelinsubstrates (A-D), cellswerefluorescently labeMwithDil, andplated onmyehn (A), polyly^^^ was added tosamplesA-C,the3A3 antibody toD. Neurites donotextendonmyelin but grow onlaminin ormixedlaminin myehnsubstrates. When3A3 is added, lamininno longer overrides growthinhibitionby myelin. Panels (E-H) show by phase contrast cells plated on recombinant MAG (E), laminin (F), or recombinantMAGplus laminin (G andH), with control antibody (E-G) or with 3 A3 (H). Integrin function is needed to override growth inhibition by MAG.

Figure3 presents theresults of studies inwhichPC12 cells transfectedwith dominantnegativeRhoextend short neurites on MAG substrates. Mock-transfected PC12 cells (a,c,e) or cells transfected with dominant-negative Rho (b,d,f) wereplated on laminin (a,b) or MAG (c-f) . MAG inhibits neurite outgrowth (c), but dominantnegative Rho cells spread on MAG andsome cells extend shortneurites (d). Treatment with C3 further stimulates neurite outgrowth on MAG from both lines of cells (e,f).

Figure 4 shows activation of Rho on MAG substrates . Activated Rho is associated with the plasma membrane. To determine if activated Rho was detectedunder conditions where PC 12 cells do not grow neurites, cells were grown in suspension or plated on MAG or collagen substrates. Two hours later the plasmamembranes were purified, the proteins separatedby SDS PAGE, and the proteins transferedto nitrocellulose and stainedwith Ponceau S . Rho A was detected on the blots by immunoreactivity with anti- RhoA antibody. Immunoreactivity was strongest when cells were grown in suspension or when cells were plated on MAG. Therefore, Rho A is more active when cells are keptin suspension orplated on MAG than when plated on growth-permissive collagen.

Figure 5 shows treatment of retinal neurons with C3 stimulates neurite growth onpolylysine andMAG substrates. On nMAG substrates neurite growth is inhibited (a) , but after C3 treatment retinal neurons plated onnMAGsubstrates extendneurites (b). Growth of neurites fromretinalneurons plated onPLL(c). Bar, 50 μm.

Figure 6 demonstrates ADP-ribosylation of Rho by C3 detected in cultured cells. PC12 cells orretinal neurons were cultured in the presence (+) or absence of C3 (-) for two days. The cells were lysed, and 10 μg of protein from eachsample was separatedona 11% acrylamidegel. Theproteins weretransfered to nitrocellulose, probed withmouse anti-RhoA antibody and anti-mouse-HRP antibody, andrevealed by a chemiluminescent reaction (top panel). Themembranes werethenreprobedwithrabbbit anti-Cdc42 and anti-rabbitalkalinephosphatase andrevealed withNTB/BCTP colorreaction. Treatment of cells with C3 resulats inanADP-ribosylation-induceddeα-easeinthemobility of RhoA. Themobility of Cdc42 does not change with C3 treatment.

Figure 7 illustrates methods usedto study the effect of C3 on injured opticnerve. Figure7ashows the optic nerve was removed from the sheath prior to crushing with 10.0 sutures (top) and C3 was applied in Gelfoam and Elvaxtubes (redbars) immediately following opticnerve crush (middle). Theretinal ganglion cell axons were detected by anterograde labeling with cholera toxin and timmunodetection of the cholera toxininlongitudinal sections of the opticnerve (bottom). Figures 7c, 7d, 7e, and7f showtreatment of crushed optic nerve with C3 stimulates regenerative growth of retinal ganglion cell axons. (c) Longitudinal 15 μmsection of a buffer-treated control optic nerve showing the failure of the RGC axons . to cross theinjuredregion; (d,e ) Longitudinal 15 μmsections of two different optic nerves treatedwith C3 showing anterogradely-labeled axons extending past the crush (arrows) . The site of crushis indicated with arrowheads; (f) Highermagnification view of (e) showingthe twisted growth of regenerating axons. Bar, 100 μm (c,d,e) and 50 μm in f . Figure 7b shows quantitation of axon regeneration across the site of lesion. Representationofregenerationobservedin different animals. For each animal, themaximumnumberof axons observedinasingle 14 μmsectionwas countedatdifferentdistancesfromthesiteofthecrushEach point represents one animal, but animals with growth past 500 μm are also represented at the shorter distances. Largenumbers ofregenerating fibers (>10/section)wereobservedto cross the lesion after C3 treatment compared to treatment with PBS.

DETAILED DESCRIPTION OF THE INVENTION

This invention arises frommediscovery atRhofamilymembers are key molecules in regulating inhibition by myelin proteins, andby MAG Thus,this inventionprovides eadvantageofiden inganintracellular target, Rho family members, for all of the multiple inhibitory proteins that must be inactivatedto allow for growth on myelin. This invention provides antagonists of Rho family members, that permit axon regeneration. The method of this invention provides for inactivation of Rho family members , thereby stimulatingneurite growth on growth inhibitory substrates. Therefore, antagonists thatinactivate Rho family members in vivo allow axon regeneration in the injured or diseased CNS.

This invention provides for the use of Rho, or proteins related to Rho as therapeutic targets for agents designedto block growthinhibition by myelin or myelin proteins. One embodiment pertains to the use of

Rho antagonists that foster axon regeneration in the central nervous system. The therapeutic agent or antagonist can be small molecules, proteins or peptides, or any agent that binds to Rho or its family members to inactivate this pathway. Another embodiment pertains to the use of the Rho regulatory pathway as atargetforRho antagonists. This pathway involves the GDP/GTP exchangeproteins(GEPs). Rho has two interconvertible forms , GDP-bound inactive, and GTP-bound active forms . The GEPs promote the exchange ofnucleotides and thereby constitute targets for regulating the activity of Rho. In another embodiment GDP dissociation inhibitors (GDIs) inhibitthe dissociation of GDP fromRho, and thereby prevent the binding of GTP necessary forthe activation of Rho. Therefore, GDIs are targets for agents thatregulate Rho activity. The GTP-boundactiveRho canbe convertedto the GDP-boundinactive form by a GTPase reaction that is facilitated by its specific GTPase activating protein (GAP). Thus, another embodiment pertains to the us e of GAPs as targets for the regulation of Rho activity . Another embodiment pertains to the factthat Rho is found inthe cytoplasm complexedwith a GTPase inhibiting protein (GDI). Tobecome active, Rho binds GTP andis translocated to the membrane. Thus, agents that promote Rho bindingto GDI, andblockRho bindingto theplasmamembrane are also considered within the scope of this invention. Yet another embodiment pertains to the observation that abacterial ADP ribosylfransferase,C3lransferase_Jribosylates Rho to inactivate totheuseofC3transferasetoinactivateRhoandstimulateaxongrowth. Likewise, otherbacterialtoxins, such as toxins A andB , withrelatedRho-inhibitory activity are considered to be within the scope of this invention. Moreover, various mutations of the Rho protein can create dominantnegative Rho, which can mterfere with me biologic-d activity of endogenous Rhoinneurons. Thus etafurtherembodimentofthis invention pertains to me use of dominantnegative forms of Rho, used to inactivate Rho, to foster axon growth.

"Antagonisf refers to apharmaceutical agent which in accordance with the present invention which inhibits at le-st one biological activity normally -.ssoci^ the inhibition of neuron growth. Antagonists whichmay beused in accordance with thepresentinvention include without limitation, one or more Rho family members fragment, a derivative of Rho family members or of a Rho family members fragment, an analog of Rho family members or of a Rho family members fragment or of said derivative, and apharmaceutical agent, and is further characterizedby the property of suppressing Rho family members-mediated inhibition of neurite outgrowth. Preferred antagonists include: mutated forms ofRho, such as Rho wherein the effector domain, A-37, has been mutatedto prevent GTP . exchange; meADP-ribosyltansferaseC3andbiologicallyeffectivefragments thatantagoniseRho family members in one of the assays of this invention; and compounds such as Y-27632 that antagonise Rho- associatedkinase(Somiyo, 1997,Nature,389:908-910;Uehata,etal.,1997,Nature,389:990-994;U.S. Patent No. 4,997,834). As described above, other antagonists include GDP dissociation inhibitors (GDIs), such as Rho GDP-dissociationinhibitor 1 (RhoGDIfromHomo sapiens) inhibitthe dissociation of GDP fromRho, and thereby prevent the binding of GTP necessary forthe activation of Rho (see, for example, Takahashi, K., J. Biol. Chem, (1997), 272:23371-5; Gosser, Y.Q., et al., Nature (1997) 387:814; Adra, et al., (1997) Proc. Natl. Acad. Sci., 94:4279-4284. The antagonist of Rho family members in accordance with thepresent invention is not limitedto Rho family members or its derivatives, butalso includes the therapeutic application of all agents, referredherein as pharmaceutical agents, which alter the biological activity of the Rho family members protein such that inhibition of neurons or their axon is suppressed.

The term "effective amount" or "growth-promoting amount" refers to the amount of pharmaceutical agent required to produce a desired antagonist effect of the Rho family members biological activity. The precise effective amountwill vary with the nature of pharmaceutical agentused andmay be determined by one or ordinary skill in the art with only routine experimentation.

As used herein, the Rho family of proteins comprises, but is not limited to rho, rac, cdc42 and their isotypes, such as RhoA, RhoB, RhoC, as well as Rho-associatedkinase that are expressedin neural tissue. Othermembers ofthe Rho family that are determined andwhose inhibition of activity allows forneurite outgrowth are comtemplated to be part of this invention. (See, for example, Katoh, H., et al., J. Biol. Chem, 273:2489-2492, 1998; vanLeeuwen, F., etal, J. CellBiol, 139:797-807, 1997; Matsui etal., EMBOJ. 15:2208-2216, 1996; Amanoetal., Science, 275:1308; Ishizakι,T. etal,(1997)FEBSLett, 404: 118-124).

As usedherein, the terms "Rho family member biological activity" refers to cellular events triggeredby, beingofeitherbiochemicalorbiophysicalnatui-e. Thefollowingj^tisprovided,withoutHmitation, which discloses some ofthe known activities associated with contact-mediated growth inhibition of neurite outgrowth, adhesion to neuronal cells, andpromotion of neurite outgrowth fromnewbom dorsalroot ganglion neurons.

As usedherein, the term "biologically active" , orreference to the biological activity ofRho family members or, or polypeptide fragment thereof, refers to apolypeptide that is able to produce one ofthe functional characteristics exhibitedby Rho family members or its receptors describedherein. In one embodiment, biologically activeproteins are those that demonstrate inhibitory growth activities central nervous system neurons. Such activity may be assayed by any method known to those of skill in the art.

The term C3 refers to C3 ADP-ribosyltransferase, a specific Rho inactivator . A pref erredrepres entative example is C3 ADP-ribosyltransferase, a23 KDaexoenzyme secreted from certainstrains oftypes Cand D fromClostiidiumbotulinum, which specifically ADP-ribosylates therho family of these GTP-binding proteins. This ADP-ribosylation occurs at aspecific asparagine residue in their putative effector domain, andpresumably interferes with their interactionwithaputative effector molecule downstream insignal transduction. Numerous references describing these compounds can be foundin Methods in Enzymology, Vol256,PartB,Eds.: W.E.Balch, C.H.Der, andA. H-dl;AcademicPress, 1995,foreg.Pgs?196-206, 207 et seq, 184-189, and 174 et seq..

B ased on the present evidence that Rho family members can affect growth inhibitory protein signals in myelin, the means exist to identify agents and therapies that suppress myelin-mediatedinhibition of nerve growth. Further, one can exploit the growth inhibiting properties ofRho family members, or Rho family members agonists, to suppress undesired nerve growth. Without the critical finding that Rho family members has growth inhibitory properties, these strategies would not be developed.

Rho Family Member Antagonists and Assay Methods to Identify Rho family members Antagonists

In one embodiment, Rho family member antagonists will be inhibitors of GTPase activity. The GTP/GDP cycle of Rho family members activation/inactivation is regulated by a number of exchange factors . Compounds that block exchange, thereby inactivatingRho family members are preferred embodiments of this invention.

In another embodiment suitable Rho family member antagonist candidates are developed comprising fragments, analogs andderivatives of ho family members. Sequences for Rho family members are known, such as those described: Chardin,P., etal., (1988) Nucleic Acids Research, 16:2717; Yeramian, etal., (1987) Nucleic Acids Research, 15: 1869). Such candidates mayinterferewithl ofamπymembers- mediated growth inhibition as competitive butnon-functional mimics of endogenous Rho family members. From the amino acid sequence ofRho family members and from the clonedDNA coding for it, it will be appreciated that Rho family members fragments can be produced either by peptide synthesis or by recombinantDNA expression of either atruncated domain ofRho family members , or of intact Rho family members couldbe preparedusing standard recombinantprocedures, that can then be digested enzymically in either arandom or a site-selective manner. Analogs of ho family members or Rho family members fragments can be generated als o by recombinant DNA techniques or by peptide synthesis , and will incorporate one or more, e.g. l-5,L-orD-aminoacidsubstitutions. Derivatives of Rho family members, Rho family members fragments and Rho family members analogs can be generated by chemical reaction ofthe parent substance to incorporate the desired derivatizing group, such as N-teiminal, C-terminal and intra-residue modifying groups that have the effect of masking or stabilizing the substance or target amino acids within it.

Inspecific embodiments of the invention, candidate Rho family member antagonists include thosethat are derivedfromadetermination ofthe functionally active region(s) of aRho family member. Antibodies ca be preparedusing techniques known inthe against epitopes in Rho family members, which, when found to be function-blocking in z^'n vitro assays, canbeusedtomaptheactiveregions ofthe polypeptide as has been reported for other proteins (for example, seeFahrig, et al, (1993) Europ. J. Neurosci, 5, 1118- 1126; Tropak, et al, (1994) J. Neurochem., 62, 854-862). Thus, it canbe determined hichregions ofRho familymembersGTPases recognized by substrate molecules that are involvedminhibitionofneurite outgrowth. When those are known, synthetic peptides can be prepared to be assayed as candidate antagonists ofthe Rho family members effect. Derivatives ofthese canbe prepared, includingthosewith selected amino acid substitutions to provide desirable properties to enhance their effectiveness as antagonists of the Rho family members candidate functionalregions of Rho family members can also be determinedby the preparation of altered forms ofthe Rho family members domains using recombinant DNA technologies to produce deletion or insertionmutants that canbe expressed in various cell types as chimericproteins. All ofthe above forms ofRho family members, andfoιτns thatmay be generatedby technologies not limited to the above, can be tested for the presence of functional regions that inhibit or suppress neurite outgrowth, and can be used to design and prepare peptides to serve as antagonists.

In accordance with an aspect ofthe invention, the Rho family member antagonist is formulated as a pharmaceutical composition which contains the Rho family member antagonists an amount effective to suppress Rho family member-mediated inhibition of nerve growth, in combination with a suitable pharmaceutical carrier. Such compositions are useful, in accordance with another aspect ofthe^* invention, to suppress Rho family member-inhibitednerve growth inpatients diagnosedwith avariety of neurological disorder, conditions andailments of thePNS and the CNS where treatmentto increase neurite extension, growth, or regeneration is desired, e.g., inpatients with nervous system damage. Patients suffering from traumatic disorders (including but not limited to spinal cordinjuries, spinal cordlesions, surgicalnerve lesions or other CNS pathway lesions) damage secondary to infarction, infection, exposure to toxic agents, malignancy, paraneoplastic syndromes, or patients with various types of degenerative disorders ofthe central nervous system can be treated with such Rho familymembers antagonists. Examples of such disorders include but are notlimitedto Strokes, Alzheimer's disease, Down's syndrome, Creutzfeldt- Jacob disease,kura, Gerstrnan-Sfraιιsslersyndrome,SCTapie. disease, Riley-Day familial dysautonomia, multiplesystematrophy, amylotrophic lateral sclerosis or Lou Gehrig's disease, progressive supranuclear palsy, Parkinson's disease and the like. The Rho family members antagonists may be usedto promote the regeneration of CNS pathways, fiber systems andtracts. In aparticular embodiment ofthe invention, the Rho family members antagonist is usedto promote the regeneration of nerve fibers over long distances following spinal cord damage.

In another embodiment, the invention provides an assay method adapted to identify Rho family member antagonists, that is agents that block or suppress the growth-inhibiting action of Rho family members. In its most convenient form, the assay is a tissue culture assay that measures neurite out-growth as a convenient end-point, and accordingly uses nerve cells that extendneurites when grown on apermissive substrate. Nerve cells suitable in this regardinclude neuroblastoma cells ofthe NG108 lineage, such as NG108-15, as well as otherneuronal cell lines such as PC12 cells (American Type Culture Collection, 12301 Parklawn Drive, Rockville, MD 20852 USA, ATCC Accession No. CRL 1721), human neuroblastomacells, and primary cultures of CNS or PNS neurons taken from embryonic, postnatal or adult animals. The nerve cells, for instance about 10³ cells-microwell or equivalent, are cultured on a growth permissive substrate, such as polylysine or laminin, that is over-layed wimagrow -inhibiting amount ofRho family members . The Rho family members incorporatedin the culture is suitably my elin- extractedRho family members, although forms ofRho family members otherthan endogenous forms can be usedprovided they exhibit the Rho family members property of inhibiting neuron growth when added to a substrate that is otherwise growth permissive.

In this assay, candidate Rho family member antagonists, i. e. , compounds thatblockthe growth-inhibiting effect ofRho fatnilymembers, are addedto the Rho f-imily membei'-containing tissue culture preferably in amount s ufficientto neutralize the Rho family member growth-inhibiting activity, that is between 1.5 and 15 μg of Rho family members antagonist per well containing a density of 1000NG108-15 cells/well cultured for 24 hr. inDulbecco' s minimal essential medium. After culturing for aperiod sufficient for neurite outgrowth, e.g. 3-7 days, the culture is evaluated for neurite outgrowth, and antagonists are thereby - revealedas those candidates which eUcitneurite outgrowth. Desirably, candidates selectedas Rho family member antagonists are those which elicit neurite outgrowth to astatistically significant extent, e.g. , in at least 50%, more desirably at least 60%, e.g. 70%, per 1,000 cultured neurons.

Otherassayteststhatcouldbeusedincludewithoutlimitation the following: 1) The growth cone collapse assay that is usedto assess growth inhibitory activity of collapsin (Raper, J.A., andKapfhammer, J.P.,

(1990) Neuron, 2, 21-29; Luo, L., et al., (1993) Cell 75. 217-227) and of various other inhibitory molecules (Igarashi, M.. , et al. , ( 1993) Science 259, 77-79) whereby the test substance is addedto the culture medium and a loss of elaborate growth cone morphology is scored. 2) The use of patterned substrates to assess substrate preference (Walter, J. et al, (1987) Development 101.909-913; Stahl, et al, (1990) Neuron 5,735-743) or avoidance of test substrates (Ethell, D.W,. et al, (1993) Dev. Brain Res.72, 1-8). 3) The expression of recombinant proteins on aheterologous cell surface, and the transfected cells areusedin co-culture experiments. The ability oftheneurons to extendneurites on the transfectedcells is assessed (Mukhopadhyay et al., ( 1994) Neuron 13 , 757-767) . 4) The use of sections oftissue, such as sections of CNS whitematter, to assess molecules that may modulate growth inhibition (Carbonetto, S., et al, (1987) J. Neuroscience 1 610-620; Savio, T. and Schwab, M.E., (1989) J. Neurosci, 9: 1126-1133). 5) Neurite retraction assays whereby test substrates are applied to differentiatedneural cells fortheirability to induce or inhibitthe retr-action of previously extendfedneurites (Jalnink, et al, (1994) J. Cell Bio. 126, 801-810; Sudan, H.S., t . (1992) Neuron 8, 363-375; Smalheiser,N., (1993) J.Neurochem. 6 340-342). 6) The repulsion ofcell-cell interactions by cell aggregation assays (Kelm, S., et al, (1994) Current Biology 4, 965-972; Brady-Kainay, S., et al, (1993) J. CellBiol. 4, 961-972). 7) The use ofnitrocellulose to prepare substrates for growth assays to assess the ability of neural cells to extendneurites on the test substrate (Laganeur, C. andLemmon, V., (1987) PNAS 84, 7753-7757; Dou, C-L and Levine, J.M, (1994) J. Neuroscience 14. 7616-7628).

Diagnostic, Therapeutic and Research Uses for Rho Family Member Antagonists

Rhofamilymemberantagonists have uses in diagnostics. Suchmolecules canbeusedinassaystodetect, prognose, diagnose, or monitor various conditions, diseases, and disorders affecting neurite growth extension, invasiveness, andregeneration. Alternatively, the Rho family member antagonists may be used tomonitortherapies for diseases andconditions which ultimatelyresultinneive damage; such diseases and conditions include but are not limited to CNS trauma, (e.g. spinal cordinjuries), infarction, infection, malignancy, exposure to toxic agents, nutritional deficiency, paraneoplastic syndromes, and degenerative nerve diseases (includingbutnothmited to Alzheimer's disease, Parkinson's disease, Huntington's Chorea, amyotrophic lateral sclerosis, progressive supra-nuclear palsy, and other dementias). In a specific embodiment, suchmolecules may be usedto detect anincreaseinneurite outgrowth as an indicator of CNS fiberregeneration. For example, in specific embodiments, alteredlevels ofRho family members activity in apatient sample containing CNS myelin canbe adiagnosticmarkerforthepresence of amalignancy, including butnot limitedto ghoblastoma, neuroblastoma, andmelanoma, or a condition involvingnerve growth, invasiveness , or regeneration in a patient.

Useful for nerve growth suppression are pharmaceutical compositions that contain, inan amount effective to suppress nerve growth, Rho family member antagonist in combination with an acceptable carrier. Candidate Rho familymembers antagonists include fragments ofRho family members that incorporate the ectodomain, including the ectodomainper se and other N- and/or C-terminally truncated fragments ofRho family members orthe ectodomain, as well as analogs thereof in which amino acids, e.g. from 1 to 10 residues, are substituted, particularly conservatively, and derivatives ofRho familymembers or Rho family members fragments in which theN- and/or C-terminal residues are derivatizedby chemicalstabilizing groups.

In apreferred embodiment, mutated forms ofRho family members are used as antagonists. One key exampleis Rho with amutated effector domain, A-37, which prevents GTP exchange. Various other mutations ofthe Rho protein that create dominate negative Rho which can interfere with the biological activity of endogenous Rho inneurons are consideredas antagonists within the scope of this invention to inactivate Rho, thereby fostering growth of neurons.

]_αanotherprefeιτedembodimentGDPdissociati^ from Rho, and thereby prevent the binding of GTP necessary for the activation ofRho are used as antagonists.

Inyet another preferred embodiment, GTPase activating protein (GAP) which f acilitates the conversion ofthe GTP-bound active Rho to the GDP-bound inactive form forms the target for regulation ofRho activity. Thus, compounds thatactivate GAP, thereby facilitating the conversion of active Rho into inactive Rho would be candidates for promoting neuronal growth.

In still another preferred embodiment, compounds that affect Rho bindingto the plasmamembrane, thereby decreasing the activity of ho are also considered Rho antagonists of this invention. In this case, thetarget design is basedontiieknowledge that RhoisfoundinthecytoplasmcomplexedwithaGTPaseinhibiting protein (GOT). To become active, Rho binds GTP andis translocatedto the membrane. Thus, agents that promote GDI activity andblockRho bindingto the plasmamembrane would decrease Rho activity, thereby serving as Rho antagonists that would permit neuron growth.

Inspecific embodiments of the invention, candidate Rho family members antagonists includespecificregions ofthe Rho family members molecule, and analogs or derivatives of these. These canbe identified by using the same technologies described above for identification ofRho family members regions that serve as inhibitors of neurite outgrowth.

The Rho family members related derivatives, analogs, and fragments ofthe invention can be producedby various methods known in theart. Themanipulationswhichresultin their production can occur at the gene orproteinlevel. For example, Rho family members-encoding DNA canbemodifiedby any of numerous strategies known inthe art (Maniatis et al. , Molecular Cloning, ALaboratoiy Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. , 1982), such as by cleavage at appropriate sites with restriction endonuclease(s), subjected to enzymatic modifications if desired, isolated, and ligated in-vitro.

Additionally, the Rho family members-encoding gene can be mutated in-vitro or in-vivo for instance in the manner applied for production ofthe ectodomain, to create and/or destroy translation, initiation, and/or termination sequences, orto create variations in codingregions and or formnew restriction endonuclease sites or destroy preexisting ones, to facilitate further in-vitro modification. Any technique formutagenesis jmownintheartcabbeuseo^includingbutnotUmitedto, in-vitro site directedmutagenesis (Hutchinson, et al, (1978) J. Biol. Chem. 253, 6551), use of TAB™ linkers (Pharmacia), etc.

For delivery ofRho familymembers antagonists, various known delivery systems can be used, such as encapsulation posomesorsermpermeablememb^ glial cells, oligodendroghal cells, fibroblasts, etc. accordingtotheprocedureknowntothoseskilledinthe are(Lindvall, etal, (1994) Curr. Opinion Neurobiol 4, 752-757). Linkage to ligands such as antibodies can be us edto target delivery to myelin and to other therapeutically relevant sites in-vivo. Methods of introduction include, but are not limited to, intradermal, intramuscular, intraperitoneal, intravenous, subcutaneous, oral, andintranasal routes, and transfusion into ventricles or asite of operation (e.g. for spinal cordlesions) ortumorremoval. Likewise, cells secreting Rho family members antagonist vity, for example, andnot by way oflimitation,hybridoma cells encapsulatedinasuitablebiologicalmembranemay be implanted in a patient so as to provide a continuous source of Rho family members inhibitor.

Therapeutic Uses of Rho family Antagonists

In an embodiment, antagonists, derivatives, analogs, inhibitors ofRho familymembers can be used in regimens where anincreaseinneurite extension, growth, or regeneration is desired, e.g., inpatients with nervous system damage. Patients suffering from traumaticdis orders (including butnot limitedto spinal cord injuries, spinal cordlesions, or other CNS pathway lesions), surgicalnerve lesions, damage secondary to - infarction, infection, exposure to toxic agents, malignancy, paraneoplasticsyndromes, orpatients with various types of degenerative disorders ofthe central nervous system can be treated with such inhibitory protein antagonists. Examples of such disorders include but are not limited to Alzheimer's Disease, Parkinsons' Disease, Huntington's Chorea, amyotrophiclateralsclerosis,progressivesupranuclearpalsy and other dementias. Such antagonists may be usedto promote the regeneration of CNS pathways, fiber systems and tracts . Administration of antibodies directedto an epitope of, (or the binding portion thereof, or cells secreting such as antibodies) can also beusedtoinMbitRhofamtiymembersprotem function in patients . In aparticular embodiment ofthe invention, antibodies directed to Rho family members may be usedto promote the regeneration of nerve fibers over long distances following spinal cord damage.

Various delivery systems are known andcanbeusedfor delivery of antagonists or irMήtor ofTho family members andrelatedmolecules, e.g., encapsulationinliposomes orsemipeimeablemembranes, expression by bacteria, etc. Linkage to ligands such as antibodies can be used to target myelin associated protein-relatedmolecules to therapeutically desirable sites in vivo. Methods of introduction include but are not mitedto intradermal, intramuscular, intraperitoneal, intravenous, subcutaneous, oral, andintranasal routes, and infusion into ventricles or a site of operation (e.g. for spinal cord lesions) or tumor removal.

In addition, any method which results in decreased synthesis ofRho family members may" be usedto diminish theirbiological function. For example, andnotby way oflimitation, agents toxicto the cells which synthesize Rho family members and/or its receptors (e.g. oligodendrocytes) may be used to decrease the concentration of inhibitory proteins to promote regeneration of neurons.

EXAMPLES

EXAMPLE 1

This example demonstrates in vitro evidence that Rho family members are respons ible for regulating the . neuronal response to MAG. In particular, this demonstratesthatthesmallGTPaseRhoregulatesthe responsetoMAG PC12 cells were plated onpolylysine (PLL), 1-ιminin, or MAG substrates and exposed toNGFtostimulateneurite growth. PC12 cells differentiatedneurites onPLLandlamininsubstrates,but on MAG substrates the cells remained rounded and did not grow neurites.

The addition ofthe ADP-ribosyltransferase C3 from Clostridiumbotulinum, that efficiently inactivates Rho familymembers withoutaffecting Rac and cdc42 (Udagawa, T. andMcIntyre, B.W. (1996) J. Biol Chem. TJX, 12542-12548), allowed the cells to extendneurites on MAG substrates. In addition this example demonstrates neurite growth fromPC 12 cells transfected with a dominantnegativeN19RhoA construct. On laminin and PLL substrates theNl 9 RhoA PC12 cells grewneurites thatwere longerthan the mock-transfected controls. Moreover, N19 RhoAPC12 cells were ableto extend neurites when plated on MAG substrates . Therefore, the inactivation ofRho stimulates neurite outgrowth and allows neurite extension on MAG substrates . These results imphcate Rho in signaling growth inhibition by MAG

Cell Culture

We obtained PC 12 cells from three different sources: from Dr. Phil Barker (Montreal Neurological Institute); from the ATCC (obtained from W. Mushinsky, McGill), andfromGaborTigyi, (University of Tennessee) andwe found that all lines of cells were inhibitedby bothmyelin andMAG. PC12 cells were grown in Dulbecco'smodifiedeagle's medium (DMEM) with 10 % horse serum and 5 % fetal bovine serum. PC 12 cells stably transfectedwith constitutively active and dominant negative RhoA constructs were kindly providedby Dr. G. Tigyi (University of Tennessee, Memphis, USA) . The three cell lines used included amock transfected cell tine, a constitutively active RhoA (V 14GRhoA) cell line, and a dominant negative RhoA (N19TRhoA) cell line. TransfectedPC12 cell lines weremaintainedin the growthmedium containing 400 mg/L G418. For cell differentiation experiments, cells were plated on appropriate substrates in DMEM with 1 % fetal bovine serum and 100 ng/ml nerve growth factor. For experiments on mixed subsfratø(lamininMAGorlamimn myehn),PC12wereρlatedinDMEMwit^ in the presence or the abs ence of 50μg/ml of an irrelevant antibody or of a purified function blocking antibody (clone 3 A3) against the rat α 1 β 1 integrin (a gift of S .Carbonetto) . PC12 cell differentiation experiments were done in 96-well plates in duplicate, andeachexperimentwasrepeatedaminimumof three times.

To culture cerebellar granule cells, 3 - 4 rats from P3 to P7 were decapitated. The cerebellum was removed and placed in MEM-HEPES where underlying tissue and the meninges was removed. The cerebellumwas cutinto small pieces andtreatedwithO.125% tiypsinin MEM-HEPES for 20' at 37°C. The tissue was thentrituratedwith a fire polishedpasteur pipette to breakup any clumps of tissue. The cells were spun down at 1500rpmfor 10', andthepelletwas resuspendedin MEM-HEPES with 2mM EDTA. The cell suspensionwas placed on an iso-osmotic percoll gradientwith 60% and35% percoll, centrifugedfor 15' at2300φm, and the interface between the 60% and35%ρercollwas collected. Cells werewashedonce, andresuspendedinDMEMwitii 10% FBS, vitamins, andpenicillin/streptomycinin the presence or absence of 20 μg/ml C3 transferase. Cells were placed in 4-chamber, chamber slides coated withpoly-1-lysine or laminin andtreatedwith spots of MAG ormyelin. 200,000 cells per chamber were plated.

Preparation of growth substrates

Poly-1-lysine was obtained from Sigma (St. Louis, Mo). Laminin was prepared from EHS tumors (Paulsson and Lindblom ( 1994). Cell biology: A laboratory handbook, Academic Press, pp589-594) and collagen fromrattails (Greene, etal. (1987) Meth. Enzymology 147,207-216). Myelinwas made from bovine brain corpus callosum, and native MAG was purified from myelin after extraction in 1% octylglucoside and separation by ion exchange chrOmatography (McKerracher, L , et al. , ( 1994) Neuron 13, 805-811). This native MAG has some additional proteins, including some tenascin (Xiao, Z., et al., (1997) Neurosci. Abstr. 23, 1994). Recombinant MAG was made in baculovirus as described (McKerracher, L., et al, (1994) Neuron 13. 805-811).

Test substrate were prepared as uniform substrates in96-well plates or4-chambered slides, or as spots on 18 mm glass coverslips . First, poly-L-ly sine was coated by incubation of 100 μg/ml for 3 hours at 37°C, andthewells or coverslips were washedwith water anddried. Laminin substrates were prepared by incubating 25 μg/ml laminin on poly-L-lysine coated dishes for 3 hours at 37°C. SolidMAG ormyelin substrateswerepreparedbydrymgdownMAGovemight^orincubatingalOmg/πjimyelinsolutionfor 3 hours onpolylysine coatedsubstrates. For 96-well plates, 14 μg of eitherrecombinant]V G(rMAG) orofnativelv-AGpei-wellwasuseάFornixedlamini myelinorlam proteins and lO μg oflaminin were dried down on 96-well plates precoated with polylysine. For 4- chambered chamber slides, 40 μgMAGperchamberwasused, andfor lOOmmplates 0.6-1 mgofMAG was dried down. Spots of MAG on coverslips were generatedby plating of 2 mg/ml recombinant MAG on polylysine for 3 -4 hours in ahumid chamber at 37°C. Collagen substrates were made by mcubating 10-15 μg/ml of rat tail collagen for 3 hours at 37°C.

Immunocytochemistry

PC12 cells were visualized by phase contrast microscopy, or following labelling with the lipophilic fluorescent dye, Dil (McKerracher, L. , etal, ( 1994) Neuron 13, 805-811). Granule cells were visualized by immunocytochemistry. Following 12-24hours in culture, cellswerefixedfor30'atroomtemperature in4%paraformaldehyde,0.5% glutaraldehyde, 0.1 M phosphate buffer. Following fixation, cellswere washed3 X5' withPBS andthen blocked for 1 hour at room temperature in 3 %BS A, 0.1%Triton-X 100. Granulecellculhireswei-emcubatedovernightwithapolyclonalanti-rM^ to label MAG spots. The MAG antibody was detected using anFITCconjugatedsecondaryantibody. Rhodamineconjugatedphalloidin as diluted 1:200 with thesecondary antibody to label granide cell -ictin filaments.

C3 transferase preparation and use

TheplasmidpGEX2T-C3 coding forthe GST-C3 fusion proteinwas obtained from A. Hall (London). Recombinant C3 was purified as described by Dillon and Feig (Met. Enzymology, ( 1994), 256, pp 174- 184). After fusionprotein cleavage by thrombin,thrombin was removedbyincubatingtheproteinsolution 1 hour onicewith lOOμlof p-aminobenzamidine agarose-beads (Sigma). The C3 solutionwas desalted onPDIO column (Pharmacia)withPBS, andsterilizedtiιiOughaO.22 μm filter. The C3 concentration was evaluatedby Lowiy assay (DC protein assay, Bio-Rad) and toxin purity was controlledby SDS-PAGE analysis. Totestthe effect of C3 on the outgrowth on PC12 cells, C3 transferasewas scrape loadedinto the cells before plating on appropriate substrates. Cells were grown to confluenceinse mcontainingmediainό well plates. Cells werewashedonce withscraping buffer (114mMKCl, 15mMNaCl, 5.5mMMgCl₂, 1 OmM Tris-HCl). Cells were then scrapedwith arubber policeman into 0.5 ml scraping buffer in the presence or absence of 20 μg/ml C3 fransferase. The cells were pelleted, and resuspended in 2 ml DMEM, l%FBS, and50ng/mlnervegrowthfactorbeforeρlatrng. 10 μg/ml C3 was addedto scrape loaded cells. Cells were differentiated for 48 hours then fixed in 4 % paraformaldehyde, 0.5 % glutaraldehyde, 0.1 M P0₄ buffer.

Membrane Translocation Assay for RhoA

PC12 cellswere collected andresuspendedinDMEM, 0.1 %BSA, 50ng/mlNGF, then plated on 100 mm dishes coated with collagen or MAG, or left in suspension. Two hours later, cells were washed with ice cold PBS + protease inhibitors (1 μg/ml aprotinin, 1 μg/ml leupeptin, 1 μg/ml antipain, 1 μg/ml pepstatin). Cells were then scraped into SmlPBS+proteaseinhibitors, and the cells were pelleted and washedwith PBS ÷ protease inhibitors. The cell pellets were mechanically homogenizedby 25 strokes in aglass-teflonhomogenizer,thehomogenatecenmTugedfor20minat8,000φm,andthecell debris in the pellet was discarded. The supernatant was centrifugedfor 1 hour at 100,000 xg to separate membrane and cytosolic fractions. Membrane pellets were washed 1 X with PBS + protease inhibitors and resuspended in PBS with0.5%SDS, and 50-100 μgofmembrane protein was analyzedby SDS-PAGE on 12 % gels . Gels were transferred to Protran nitrocellulose membrane and stained with Ponceau S . Blots were blockedfor 1 hour in 5 % skimmilk in TBS , andprobed overnight with Rho A antibody diluted 1 :200 in 1.5 % skim milk in TBS . Rho A antibody was detected by using an alkaline phosphatase conjugated secondary antibody and an alkaline phosphatase detection kit (Gibco-BRL).

Growth inhibition of PC12 cells and its modulation by NGF and laminin PC 12 cells typically extend neurites in respons e to NGF, but when plated on myelin substrates the cells remainroundanddonotextendneurites(Moskowitz,P.F.,etal.,(1997) J. NeurosciRes. 34, 129-134.) (Fig.2). MAGis apotentinhibitor of axon growth present in myelin. We observed thatPC12 cells plated on substrates ofMAGalsoremainedrounded(Fig.1), a finding in contrastto areportthatPC12 cells are not responsive to MAG (Bartschj U., et al, (1995) Neuron \5_, 1375-1381). To further examine the response of PC12 cells to MAG, we plated three different lines of PC12 cells on both native and recombinant MAGsubstrates in thepresenceofNGF. All oftheϋnesofPC12 cells showedreducedcell spreading, andmost cells remained rounded withoutneuiites. However, with inCTeasingtime,someneurites were able to extend on MAG substrates (see below). We also observedthat different preparations of MAG can differ in their potency to inhibit neurite growth, and thatthe activity of MAG is red ced or lost upon freeze-thaw.

Laminin is known to override completely, growth inhibition ofNGl 08 cells by myelin (David, S . , et al. , (1995) J. Neurosci Res. 42, 594-602). Similarly, we found that PC12 cells areableto extendneurites onmixedmyelinandlamininsubstrates or on mixed laminin/MAG substrates (Fig.2). To determine if signalling t-iroughintegrins is responsible for overriding growth inhibition by myelin, weusedtheintegrin function blocking antibody 3A3raisedagainstthe αl subunit extracellular domain. Previous studies have documented thatαlBl integrin is the dominant integrin expressedby PC12 cells, andthatthe3A3 antibody blocks PC12 cell neurite growth on laminin (Tomaselli, K.J., etal., (1990) Neuron 5, 651-662). We. platedPC12 cells onmixedmyelinandlamininsubstrates,inthepresenceofthe3A3 antibody, or with a non-specific IgG antibody as a control. The 3A3 antibody blockedneurite extension on both andlaminin and the mixed my elin/laminin substrates (Fig.2). On MAG or on myelin substrates the cells remained rounded. The observation that the 3A3 antibody restores growth inhibition on mixed substrates demonstrates that laminin does not override growthinhibition by maskingthe inhibitory domain of MAG, but that signals elicited through integrins receptors are responsible.

Effect of C3 Transferase on PC12 cells To investigatepossible intracellular targets thatmay override growth inhibition by myelin and by MAG, we focused on the small GTPase Rho which is known to play arole in convergent signalhng pathways that affectmoιphologyandmotility(Hall,A.,(1996) Ann. Rev. CellBiol. 10,31-54). We inactivatedRho in PC12 cells by scrape loading them with the bacterial toxin C3 before platingthe cells on the test substrates. C3 is known to inactivate Rho through ADP ribosylation (Udagawa, T. andMcIntyre, B. W. (1996) J.Biol. Chem. 271, 12542-12548). On control substrates of polylysine andlaminin, treatment with C3 potentiatedboththenumberof cells withneurites andthe length of neurites fromcells (Fig.3). On MAG andmyelinsubstrates where neurite formation is inhibited, C3 has adramatic effect on theabilityto extend neurites (Fig 3). When treated with C3, abouthalfofthePC12 cells plated on eitherrMAG or native MAG hadneurites of approximately 1 cell body diameter. In contrast, the untreated cells remainedrounded and clumped. Similarly, PC12 cells plated onmyelinremainedrounded, butthe addition of C3 allowedneurites to extend directly on the myelin substrate. These results demonstrate that C3 treatment elicits neurite growth from PC12 cells plated on growth inhibitory myelin or MAG substrates.

Growth of dominant-negative Rho-transfected cells on MAG substrates

PC 12 cells transfectedwith constitutively active RhoA (VI 4GRhoA), andPC 12 cells transfectedwith dominantnegativeRhoA(N19TRhoA), andthemock-transfected cells, wereexaminedfortheir ability to extendneurites on differenttestsubstrates. Cellswiththe constitutively activemutation, V14GRhoA cells, differentiatedpoorly on all substrates, including poly-L-lysine andlaminin. Thetreatment of the V 14GRhoA cells with C3 allowed the growth of some shortneurites on all ofthe test substrates, including MAG.

Inthesame series of experiments the response of dominantnegative Rho- transfected cells, N19TRhoA cells, to MAG and myelin substrates was examined. WhenN19TRhoA cells were plated on MAG substrates, they spread and did not remain rounded as did the mock transfected PC12 cells. A small number of cells had short neurites, an effect that was observed on both the rMAG and native MAG substrates (Fig.3).

C3 treatment of mock transfected andN 19TRhoA cells had a dramatic effect of neurite outgrowth as most cells had extensive neurites (Fig.3). Theeffectof C3 onN19TRhoA cells was muchmoremarkedthan the effectonthemocktransfected cells. Therefore, the combination of C3 treatment and transfection of dominant negative Rho elicited excellent outgrowth of neurites from PC 12 cells plated on inhibitory MAG (and myelin) substrates.

Effect of C3 on Primary Cells

To test the involvement ofRho in the response of primary neurons to MAG and to myelin substrates, cerebellar granule neurons were plated on test substrates andtreatedwith C3. Neurite outgrowth from thesecellswas known to be inhibitedby MAG(Li, M., etal, (1996) J. Neurosci. ites. 46, 404-414) and the C3 stimulated growth of neurites from the granule cells on both permissive and inhibitory substrates .

The growth substrate influences the cellular location ofRho

Rho is associated with the plasmamembrane when it is in an activatedstate, andit moves into the cytosolic fraction when it is in the GDP-bound inactive state. To determine if the growth substrate influences the cellular localization ofRho, cells were either left in suspension or platedonMAGor collagensubstrates, and preparedmembranes from the cells twohours later. Itwas shown that Rho was principally localized in the cytosolic fraction when cells wereplated on collagen, agrowthpermissive substrate. However, Rho was associated witiitheplasmamembranewhen cells where grown in suspension andwhen cells were plated on MAG (Fig. 4).

EXAMPLE H: IN VIVO DEMONSTRATIONS 1. Effect of C3 on cultured retinal neurons

To test the involvement ofRho in the respons e of primary neurons to MAG and to myelin substrates, we purifiedretinalneurons and freatedthem with C3. Neuiite outgrowth from these cells was inhibitedby MAG(Fig.5a). AswithPC12cells,treatmentofretinalneurons cells with C3 allowedneurite extension on thegrowthinhibitoiyMAGsubstrates to an extentsimilarto that observed on control substrates (Fig. 5b and 5)

To ensure that the effect of C3 treatment resulted from uptake of C3 into the cells, we examined by Western blot the electrophoretic mobility ofRho inPC 12 cells and retinal neurons treated with C3 (Fig. 6). It has previously been shown that ADP-ribosylation ofRho results in decreasedmobility ofRho on SDS-acrylamide gels (MethodEnzymol. Vol 256, Chapter22 pgl 98). For our experiments, PC 12 cells were scrape-loadedwith C3 orwith scrape-loading buffer as a control, and cell lysates were prepared after 48 hours in culture. Western blots ofthe lysates with anti-RhoA antibody revealedan increasein the apparentmolecularweightin cells treated with C3. As a control forthe specificity ofthe effect, weprobed the same blots for another small GTPase ofthe Rho family, Cdc42. Cdc42 didnotshow any change in mobility upon treatment with C3.

Tocultureretinalneurons,retinaswereremovedfromPl-P5ratpups,andthecells weredissociatedwith 12.5 U papain/ml in Hanks balanced salts solution, 0.2 mg/ml DL cysteine and 20 μg/ml bovine serum albumin. The dissociated cells were plated on test substrates in the presence of 50 μg/ml BDNF in DMEM with 10% FBS, vitamins, andpenicillin/sfreptomycin in the presence or abs ence of 50 μg/ml C3 transferase. Neurons were visualized by fluorescent microscopy with anti-βlH tubulin antibody.

2. Effects of C3 on retinal ganglion cell axon growth in vivo To explore the possibility that treatment of damaged axons with C3 might foster regeneration in vivo, we examinedregeneration of retinal ganglion cell (RGC) axons inthe opticnerve 2 weeks after optic nerve crush. Recently, it has been shown thatmicrolesions in the CNS reduce the extent ofthe glial scar and allow axons access to CNS white matter distal to the lesion (Davies, S.J.A., etal. (1997) Nature390, 680-683). To make microlesions of optic nerve, 10.0 sutures were used to axotomize RGC axons by constriction (Fig.7a). Retrograde labeling of RGCs from the superior colliculus (not shown), as well as anterograde labeling techniques (eg., Fig.7a) verified that RGC axons were effectively axotomized. To apply C3 to crushed nerves , Gelfoam soaked with 2 mg/ml C3 was wrapped around the left optic nerve atthe crush site, andtwoElvaxtubes, each loadedwith20 μgofC3 were positioned for sustained slow release (Fig.7a) . Twelve animals were treated with C3 , and a further 8 animals were treated with PBS as controls . Crushed and regenerating axons were visualized by anterograde labeling with cholera toxin injected into the eye 12 days after optic nerve crush (Fig.7a). Fourteen days after optic nerve crush, longitudinal ciyostat sections of the optic neives were examined by fluorescent microscopy for immunoreactivity to cholera toxin to detect anterogradely labeled RGC axons.

In control optic nerves that received optic nerve crash alone, no RGC axons extendedpastthe crushsite (n=3 animals). In control animals treated with PBS-Elvaxpellets and gelfoam, the crush site was easily detected where most anterogradely labeled axons stopped abruptly (Fig.7c). However, in these animals, afew axons did extend pastthe crash (Fig.7c, arrows), andthenumber of axons thatregenerated varied - fromanimalto animal. The application of GelfoamandElvaxtubes may have altered theresponseto injury. Nonetheless, the response to C3 treatment applied with this lesion paradigm was dramatic.

We observedthat C3 treatment allowed many RGC axons to grow past the region ofthe lesion. In 7 of 12 C3-treated animals, the lesion site was not clearly defined because ofthe large numbers of axons that extendedthroughthesite(Fig.7dande).Manyoftheaxonsthatextendedpastthelesionsite showeda twisted path of growth, supporting their identification as regenerating axons (Fig. If). A quantitative comparison of C3 andPBS treated animals revealed thatmore fibers grew past thelesionsite after C3 treatmentthan after PBS treatment (Fig.7b). For this analysis wemade aeons ervative estimate ofthe lesion site based on morphology , and counted the number of fibers in the distal optic nerve in 14 μm sections. Seven of 12 C3-treated animals showed at least one section with 10-20 axons extending250 μm past the crush, compared with 1 of 8 of the PBS-treated controls (Fig. 7). In some animals regenerating axons were observedup to 1 mmfromthe crash, an extent of regeneration similarto that observedinmouse opticnerve after treatment with IN- 1 antibody to block myelin inhibitors where fibers extended up to 750 μm ((Bartsch, U., et al, (1995) Neuron 15. 1375-1381).

C3 treatment of crushed optic nerve in adult rats Rats were anesthetized with 0.6 mlkg hypnom, 2.5 mg/kg diazepan and 35 mg/kg ketamin. The left optic nerve was exposedby asupraorbital approach, the optic nerve sheath slit longitudinally, the optic nerve liftedoutandcrushed 1 mmfromthe globeby constriction with alO.OsutureheldforόOseconds (Fig.4a). For C3 treatment andbuffer controls, Gelfoam soaked in PBS or 2 mg/ml C3 transferasewas placed on the nerve atthe lesion site. Two 3 mmlong tubes of Elvax (Sefton, et al, ( 1984)) loaded withbuffer or 20μg C3wereirιsertedmtheGelfoamneaι-thenerveforcontinuedslowreleaseofC3 (Fig.4b). Twelve days after crash, 5 μl of 1 % cholera toxin βsubunit (ListBiological laboratories, Inc., Cambell, CA) was injectedinto the vitreous to anterogradely label retinal ganglion cell axons (Fig.4c). Twoweeks after optic nerve crash the animals were fixedby perfusion with 4% paraformaldehyde, and the eyewith attached optic nerve was removed andpostfixed in 4% paraformaldehyde . Longitudinal cry ostat sections were processed for immunoreactivity to choleratoxin with goat anti-choleratoxin at 1 : 12,000 (ListBiol. Labs Inc, CA),followedbyrabbitanti-goatbiotinylated antibody (1:200, Vector Labs, Burlingame,CA), and DTAF-streptavidin (1: 500, Jackson Immunoresearch Laboratories).

Discussion HerewereportthattiiesmallGTPbindingproteinlΛoislikelytobeakeyinterπiediatemtheneuronal responseto neurite growth inhibitory signals. Treatment of cultured PC12 cells, retinal neurons, and cerebellar granule cells with C3 enzyme to inactivate Rho allowedneurites to extend directly on inhibitory substrates of MAG or myelin . Also, PC12 cells transfectedwith dominantnegative RhoA extended neurites on MAG substrates. Therefore, inactivation of ho was sufficient to allow neurite growth on MAG or myelin substrates when neurons were grown in the presence of neuiOtrophic factors.

Further, our observations of microlesioned optic nerves after treatment with C3 provide the first evidence that the inactivation ofRho in axons and non-neuronal cells near the site of lesion can help foster regeneration after injury. While the v tro experiments showed that C3 can affect directly the growth of neurites fromretinal cells, itis likely that the effects we observed after application of C3 to the opticnerve in vivo are more complex. C3 may affect other non-neuronal cells, such as macrophages and astrocytes, and these possibilities needto be further examined. Nonetheless, our data provide compelling evidence that C3 can promote neurite growth on inhibitory substrates in vitro, andhelps to overcome growthinhibition in vivo.

Regulation of neurite growth by Rho family members Not all ofthe myelin-derived inhibitory molecules are known to date, and less is known aboutthe neuronal receptors for growth inhibitory molecules. Several different MAG receptors have been identified (Collins et al. 1997; Yang etal. 1996), and additional neuronal receptors to myelin inhibitors are likely to exist. Targeting infracellular signaling mechanisms convergingto Rho rather than individual receptors may be the most practical way to overcome growth inhibition in vivo . The advantage of inactivating Rho to stimulate regeneration is that axons can regenerate directly on the n ative terrain ofthe CNS , and thus may be more likely to find their natural targets.

BothMAGandthe other myelin-derived growth inliibitoiy proteins block axon extension by causing growth cone collapse (Li, M., et al, (1996) J. Neurosci. Res. 46, 404414; Bandtlow, C.E., et al, (1993) Science 259, 80-83). These finding suggested to us that growth cone collapse by the myelin-derived inhibitors might be regulated by Rho. Moreover, in non-neuronal cells, Rho participates in integrin- dependent signaling (Laudanna, C, et al, (1996) Sciencellλ, 981-983.; Udagawa, T. andMcIntyre, B.W. (1996) J.Biol Chem.271, 12542-12548.). Togetherwiththeobservationthatlaminin can override myelin-derivedinhibition,wehypothesizedthat small GTPases ofthe Rho family mightplay arole in integrating signaling from positive andnegative growth cues. To investigate this possibility, we have made use ofthe ADP-ribosyl fransferase C3 from Clostridium botulinum that efficiently inactivates Rho without affecting Rac and Cdc42, two othermembei-s ofthe Rho family (Udagawa.T. andMcIntyre,B.W. (1996) J.Biol. Chem. 271, 12542-12548)andfoundthatC3treatmentfosters neurite growthinthepresenceof growthinhibitors. Moreover, immunocytochemical observations indicate that Rho protein is concentrated atthe filopodial tips of growth cones in adhesion structures calledpoint contacts (Renaudin etal. 1998). Therefore, our in vitro results suggest the Rho signaling pathway is a key target for regulating growth cone motility and stimulating regeneration.

Moreover, this dataisrelevantto the finding ofSong etal (Song etal. Science 281 : 1515-1518 (1998)) who report that growth cone repulsion by MAG can be converted into attraction by elevation of intracellular cAMP levels to activate protein kinase A (PKA). Experiments with non-neuronal cells has implicated cAMP inthe regulation ofRho because elevation of cAMP inhibits Rho activation (Laudanna, C, etal, (1996) Sciencelll 981-983). InPKA deficient PC 12 cells, elevation of cAMP fails to protect from the activation of Rho by lysophosphatidic acid (Tigyi, G., et al, (1996) J. Neurochem. 66, 537-548), afmdingthatsuggests that PKA-dependent regulation ofRho occurs inneural cells as well. Therefore, the cAMP-dependent regulation is likely to be upstream ofRho (Laudanna, C . , et al. , ( 1996) Science 271, 981-983).

The non-neuronal response to optic nerve injury

Remarkably, we observed that RGC axons crossed the lesion site to enter the distal optic nerve after treatment of injured optic nerve with C3. S ome axons grew up to 1 mm past the site of lesion. This distance is comparable to the maximal distances observedfollovvingtreatmentof optic nerve withIN-1 antibody

(Bartsch,U., etal,(l995)Neuron \5, 1375-1381) Themoststriking feature ofourresults was the large number of axons that were able to cross the lesion site compared to PBS-treated controls (see Fig.7). Therefore, it is appears that C3 was also able to promote axon growth on inhibitoiy proteins present atthe ghal scar, indicating that targeting the Rho signaling pathway as widespread efficacy in stimulating axon regeneration after injury.

Claims

We Claim:

1. An antagonist of one or more ofRho family members characterizedby the ability to elicit neurite outgrowth from cultured neurons in an assay method, comprising the steps of: (a) culturmgneurons on a growth permissive substrate that incorporates a growth-inhibiting amount of a Rho family member; and (b) exposing the cultured neurons of step a) to a candidate Rho family member antagonist agent in an amount and for a period sufficient prospectively to permit growth ofthe neurons; thereby identifying as Rho family antagonists the candidates of step b) which eUcitneurite outgrowth from the cultured neurons of step a).

2. The antagonist according to claim 1, wherein said Rho familymembers are selected from the group comprising RhoA, RhoB. RhoC, Rac, cdc42 and Rho-associated protein kinase.

3. The antagonistaccording to claim 1 , wherein saidinteraction with the Rho regulatory pathway is via interaction with GTP/GDP cycle.

4. The antagonistaccording to claim 3 , wherein the interaction with the GTP/GDP cycle involves GTP/GDP exchange protems (GEP's); GDP dissociationinhibitors (GDI's); or GTPase activating protein (GAP) to regulate Rho activity.

5. The use of antagonists of one or more Rho family members to promote neural growth by inhibiting Rho family members in the central nervous system.

6. The us e of ADP-ribosyl transferase C3 , or other closely related toxins, to promote neural growth by inhibiting one or more Rho family members in the central nervous system.

7. TheuseofaGTPaseactivatingpiOteinthatisspecifictoRhoto convert GTP -boundactive Rho to GDP-bound inactive Rho.

8. Theuseof ADP-ribosyl ti-ansferaseC3 accordmgtoclaim6,whereinsaidrelatedtoxirjsaretoxins A or B.

9. Theuseofbiologically active fragments of ADP-ribosyl transferaseC3, analogs and derivatives thereof, to promote neural growth by inhibiting one or more Rho family members in the central neivous system.

10. TheuseofY27632, orrelatedcompounds,topromoteneuralgrowthbyinhibitingRho-associated kinase in the central nervous system.

11. The use of genetically mutated forms ofRho, to promote neural growthby inhibiting one ormore Rho family members in the central neivous system.

12. The use of dominant negative Rho to inactivate Rho, to foster axon growth inthe central nervous system.

13. The genetically mutated forniofRlio according to claim 11, wherein the mutation is inthe effector domain, A-37, thereby preventing GTP exchange.

14. The use of GDP dissociationinhibitors, orstimulation thereof, to inhibitthe dissociation of GDP from Rho and thereby prevent the binding of GTP necessary for the activation of Rho.

15. The use of compounds that promote Rho binding to GTPase inhibiting protein (GDI), thereby antagonizing the ability of Rho to be translocated to the plasma membrane.

16. A method forproducing Rho antagonists from Rho family members, fragments, analogs of derivatives by peptide synthesis or by recombinant DNA expression of either atruncated domain ofRho familymembers,incorporatingoneormoreL- orD-amino acid substitutions, orofintact Rho family members using standardrecombinantprocedures and selecting antagonist characterized by the ability to eUcitneurite outgrowth from culturedneurons in an assay method, comprisingthe steps of: (a) cultiiring neurons onagiOwthpermissivesubstratethatincoiporates

amount of a Rho family member; and (b) exposing the cultured neurons of step a) to a candidate Rho family member antagonist agent in an amount and for a period sufficient prospectively to permit growth ofthe neurons; thereby identifying as Rho family antagonists the candidates of step b) which eUcitneurite outgrowth from the cultured neurons of step a).

17. The antagonist according to claim 1 , wherein derivatives ofRho family members, Rho family members fragments andRho family members analogs canbe generatedby chemical reaction ofthe parent substance to incorporate the desired derivitizing group, such as N-terminal, C-terminal and intra-residue modifying groups that have the effect of masking or stabilizing the substance or target amino acids within it.

18. An antagonist of one or more of Rho family members , characterized by following properties : (a) blocks growth inhibition of neurites by myelin or myelin proteins; and

(b) interferes with Rho family members-mediated growth inhibition as competitive butnon- functional mimics of endogenous Rho family members;

19. A composition comprising a therapeutically effective amount of the composition of claim 1 in a suitable pharmacologic earner.

20. An assay methoduseful to identify Rho family member antagonist agents that suppress inhibition of neuron growth, comprising the steps of:

(a) cultuiing neurons on a growth pei ύssivesubsfrate that incorporates a growth-inhibiting amount of a Rho family member, and

(b) exposing the culturedneurons ofstep a) to a candidate Rho familymember antagonist agent in an amount and for a period sufficient prospectively to permit growth ofthe neurons; thereby identifying as Rho family antagonists the candidates of step b) which eUcitneurite outgrowth from the culturedneurons of step a).

21. A kit to test for Rho family antagonists that can be us ed to promote neurite growth comprising the components necessary to work the method of claim 16, in a suitable container.

22. A method to suppress the inhibition of neuron, comprising the steps of delivering, to the nerve growth environment, a Rho family antagonist in an amount effective to reverse myelin inhibition.