Besides acute brain injuries, of which stroke is themost common and harmful for normal brain function-ing, there are several forms of slowly progressive braindisorders, which cause degeneration of specific popula-tion of neurons. In Amyotropic Lateral Sclerosis (ALS),Alzheimerís (AD), Parkinsonís (PD) and Huntingtonís(HD) diseases, cholinergic motor neurons, pyramidalcortical and hippocampal neurons, nigrostriatal dopa-minergic and striatal GABA(γ-aminobutyric acid)-con-taining neurons are affected first, respectively.Incidence of neurodegenerative diseases in human pop-ulation (excluding their genetic forms) is increasingwith ageing. Similar disturbances afflict normal func-tion of all neurons but only the most vulnerable of themare lost. Why some population of neuronal cells aremore prone to die, it is still not fully explained. Somehypotheses point to excitotoxicity and dysregulation incalcium homeostasis, and consequent perturbations inmitochondrial function, activation of proteolyticenzymes, increased production of deleterious reactiveoxygen species (ROS) and impaired cell defense sys-tem, especially in large neurons with long axons in cer-tain brain regions. Cytoskeletal disruptions, proteinaggregations, deficiency of neurotrophic factors andaccompanying inflammation, which are observed in allkinds of neurodegenerative disorders are results andequally triggers of vicious circle of neuronal cell death(1). Moreover, the presence of free iron and other tran-sition metals could also facilitate neurodegeneration inspecific brain areas (e.g. substantia nigra in PD) byincreasing production of reactive oxygen species (2).Therefore, there have been two different strategies forpreventative and therapeutic interventions in neurode-generative disorders. One strategy is to block the dis-ease-specific events that are believed to initiate the neu-rodegenerative process (e.g. anti-amyloid strategy inAD or increasing dopamine level in PD) (3-5), howev-er, these new symptomatic therapies still do not addressthe fundamental problem underlying neurodegenerativediseases, the continual loss of specific populations ofneurons. Alternatively, neuroprotective strategies aim tohalt or slow the progressive degenerative processes. Aplethora of different agents have been studied in cellu-lar and animal models of neurodegenerative diseases,but only few of them reached clinical trials. Putative

neuroprotectants are searched among substances whichcould delay or stop degeneration of neurons at a partic-ular stage of neuronal cell death by inhibiting for exam-ple excitotoxicity, apoptosis, ROS production, mito-chondrial perturbations or inflammation. Not fullyunderstood mechanism of neuronal cell death (6), het-erogeneity of animal models of neurodegenerative dis-eases, which are often inadequate to represent a humandisease and impropriate design of clinical trials, are themain reasons of the lack of efficiency of a preclinicallystudied neuroprotective substances in human treatment(7-9). In spite of big effort devoted to explaining bio-chemical and molecular changes during neuronal celldeath and to constructing better animal models ofhuman brain disorders, there is a constant search forsubstances with multidirectional action on several stepsof neuronal cell death or for a combined treatment withfew neuroprotective drugs (10, 11). Besides a pharma-cological intervention, genomic approaches are inten-sively studied that aim to delay or stop neuronal celldeath or to strengthen cellular defense system, e.g. anti-sense strategy, gene silencing methods, overexpressionof endogenous neuroprotective proteins. Restorativetherapy is another way to combat neurodegenerativediseases and could involve cell therapy or methodswhich increase neurogenesis, and as a result, normalneuronal activity could be restored in the damaged brainregions. These new neuroprotective pharmacologicalapproaches as well as genomic and cellular restorativemethods are the subject of the article presented herein.

Double-faced neuronal cell death and implications

for neuroprotection

Acute brain injuries and stroke are accompaniedby a rapid, necrosis-like cell death in the damagedregion and slowly progressing, apoptosis-like neuronaldegeneration in the neighboring area (12). Neuronal lossin age-related, chronic neurodegenerative diseases isalso evoked by induction of the apoptotic pathway byseveral endogenous or exogenous factors. In somecases, it is difficult to unequivocally define the type ofcell death, and a continuum of necro-apoptotic process-es is observed, which has implications to a proper treat-ment strategy. Furthermore, the inhibition of one formof cell death (necrosis) could accelerate another (apop-

NEW TARGETS FOR PUTATIVE NEUROPROTECTIVE AGENTS

DANUTA JANTAS-SKOTNICZNA, MA£GORZATA KAJTA and W£ADYS£AW LASO—

Department of Experimental Neuroendocrinology, Institute of Pharmacology, Polish Academy of Sciences, SmÍtna 12, 31-343 KrakÛw, Poland

Abstract: A better understanding of intracellular pathways engaged in the neuronal cell death afforded us newtargets for developing putative neuroprotectants. Also pharmacological or genomic intervention aimed to mod-ulate the expression of endogenous neuroprotective or toxic agents is a very promising strategy. Taking intoaccount enormous complexity of biochemical cascades involved in neurodegenerative processes, multipoten-tial or combined pharmacological approaches seem to be more efficient in combating degenerative brain dis-eases. Moreover, an improved cell transplantation also adds to a plethora of methods, which are used to affordneuroprotection and promote neurorestoration. All those strategies are reviewed in the presented article.

359

tosis), so combined treatment with anti-necrotic andanti-apoptotic agents also has been postulated (13).Excessive or moderate stimulation of glutamate receptorby this agonist (the main excitatory neurotransmitter), inthe process named excitotoxicity, leads to neuronal celldeath by necrosis or apoptosis, respectively, especiallyin those neurons, which possess a high number ofionotropic, calcium-permeable glutamate receptors(NMDA (N-methyl-D-aspartate) or AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid) receptors(14). Moreover, adverse conditions, such as the reducedlevel of oxygen or glucose, increased level of oxidativestress, exposure to toxins and disease-specific abnor-malities (e.g. beta-amyloid, mutant huntingtin, mutantCu/Zn SOD) were reported to sensitize neurons to exci-totoxic death (1). After brain trauma (e.g. epilepticseizures, traumatic brain and spinal cord injuries) orenergy deprivation (stroke), a large amount of glutamateis released from injured cells, thus activating voltage-dependent sodium and calcium channels and/or ligand-gated ionotropic glutamate receptors, thereby causingneuronal cell depolarization, activation of proteolyticenzymes, cell swelling and lysis. In contrast to necrosis,apoptosis could be evoked not only by an overactivationof glutamate receptors and calcium overload but also bycytokines (e.g. tumor necrosis factor, FAS ligand),insufficiency of trophic factors, DNA damage and mis-folded proteins can trigger this kind of cell death.Apoptotic stimuli initiate a cascade of intracellularevents by activating several intracellular enzymes, suchas a nitric oxide synthase (NOS), proteases (caspases,calpains, cathepsins), protein kinases (p38 MAPK,JNK), protein phosphates (calceneurin) and increasingmitochondrial permeability, thus causing the release ofpro-apoptotic agents, such as cytochrome c and apopto-sis inducing factor (AIF) into cytoplasm. The releasedfactors then activate several effector proteolyticenzymes or directly influence chromatin structure, giv-ing characteristic morphological changes, such as theapoptotic nuclear fragmentation, cellular blebbing andformation of apoptotic bodies (6). Recently, it has beenshown that the released cytochrome c could interactwith endoplasmic reticulum receptors (IP3R) and in thisway can accelerate calcium-dependent apoptosis (15),which suggests the existence of complex interactionsbetween proteins in the cell death pathways and couldbe a target for searching new anti-apoptotic substances.The level of pro-apoptotic (Bax, Bad, Bid) and anti-apo-totic (Bcl-2, Bcl-xl) proteins from the Bcl-2 proteinfamily, and the presence of endogenous inhibitors ofapoptosis (XIAPs, C-IAPs) decide whether neuronalcell would die by apoptosis or not (1, 16). In differentkinds of neurodegenerative disorders, the level of anti-apototic proteins was observed to be decreased, anti-oxidative cell defense system was impaired while thelevel of pro-apoptotic proteins was elevated, whichfacilitated the apoptotic neuronal cell death upon harm-ful stimuli. Moreover, inflammatory processes, whichare a source of deleterious ROS, nitric oxide andcytokines, also accompanies degenerative brain disor-ders. The formation of free radicals is frequently associ-ated with ischemia, reperfusion and intracranial hemor-rhages and also accompanies chronic neurodegenerativedisorders with dysfunctional mitochondria (17-19). The

risk of free radical-induced damage is very high in thebrain, which is due to its high oxygen demand, highlevel of iron and unsaturated fatty acids, which are sub-strates for production of reactive lipid radicals (20).

Neuroprotective strategies

Anti-excitotoxic approaches. So far, classical anti-excitotoxic pharmacological intervention in rapid celldeath mechanisms was obtained by the inhibition ofCa2+ influx mainly through NMDA receptors. Although,different kinds of NMDA receptor antagonists haveshown neuroprotective action under experimental con-ditions, they were useless in clinical practice because oftheir negative influence on normal brain function (com-petitive and noncompetitive antagonists of NMDAreceptor impair memory and learning processes) orstrong adverse effects (NMDA channel blockers elicitpsychomimetic and amnesic action) (21, 22). AmongNMDA antagonists, only memantine, a low-affinityvoltage- and use- dependent open channel blocker, iswidely used in the treatment of moderate to severe ADwith combination with acetylcholinesterase inhibitors,but still there is no clear evidence of its neuroprotectiveaction in AD patients (only symptomatic improvement).Potency of memantine as neuroprotective drug is cur-rently also studied in clinical trials for stroke, glaucoma,vascular and HIV-related dementias and second-genera-tion memantine derivatives, which will have better neu-roprotective activity have been proposed (e.g. nitrome-mantine, a combination of memantine with nitroglyc-erin, which also down regulates NMDA receptor activi-ty) (21). Moreover, it was shown that memantine pro-longed survival in the amyotropic lateral sclerosismouse [SOD1(G93A)] model (23). An abundant num-ber of modulatory sites on NMDA receptor and its spe-cific subunit composition could also be a target for drugtreatment, and new specific agents with such profile aresearched for (e.g. polyamine and redox site modulators,NR2B antagonists, glycine B receptors modulators) (24-26). Furthermore, AMPA receptors have regainedrecently attention, since their new allosteric modulatorswere proposed, and some new data on the mechanism ofAMPA receptor activation, which could contribute toneuroprotection, were reported (27-29). AMPA receptoragonists are tested in AD, as agents which improvelearning and memory (30). Furthermore, an AMPAreceptor antagonist, talampanel (LY-300164) has beenreported to have beneficial effect in non-human primatemodels of Parkinsonís disease and in ischemic rats (10,31). Modulation of AMPA receptor could also beachieved by influencing proteins anchoring AMPAreceptor in synapses (PSD-95, GRIP, PICK, NSF) andis also preclinically investigated (31). Experimentalstudies concern also metabotropic glutamate receptors(mGluR) which modulate activity of ionotropic gluta-mate receptors (group I mGluR) and glutamate release(group II and III mGluR). Negative and positiveallosteric modulators of mGluR I (e.g. MPEP, MTEP)and mGluR II, III prevent calcium overload and exces-sive glutamate release, respectively, and are very prom-ising agents as putative neuroprotectants, at least in ani-mal studies (32-35). To date, only one agonist ofmGluR4, L-AP4 has been tested in clinical trials for PDand development of potent, selective allosteric modula-

360

tors with appropriate pharmacokinetic properties isunder way (36). Acute brain injuries could also beexperimentally attenuated by blocking the voltagedependent calcium and sodium channels or by usingpotassium channel activators. Riluzole, the onlyapproved drug for ALS, acting as an anti-glutamatergicagent by blocking voltage-dependent sodium channelsdelays progress of the motor neuron degeneration (37).Another ñ Na+ blocker and K+ activator (lubelozole andBMS-204352) were studied in clinical trials for stroke,but they showed no efficacy despite promising pre-clin-ical studies (7). Recently, it has been reported that dur-ing ischemia, the calcium permeable acid-sensing ionchannels (ASICs) are activated, and blocking thesereceptors could be a new potential therapeutic target foralleviation of stroke-induced damage (38, 39).

Antiapoptotic agents. Beside anti-glutamatergicapproaches in treating neurodegenerative diseases, thereare a lot of studies with agents which act on intracellu-lar steps of neuronal cell death. One class of those sub-stances include anti-apoptotic agents, which act mainlyto inhibit caspase-3, which is an effector caspase in bothintracellular and extramitochondrial cell death pathways(40, 41). Indeed, peptide-based caspase inhibitors pre-vent neuronal loss in animal models of head injury andstroke, suggesting that finding of non-peptide smallmolecules, that halt the apoptotic process, could be agood strategy in the treatment of neurodegenerative dis-orders. In addition, preclinical developments includeadvances in protein therapy based on the use of naturalinhibitors of caspases, which possess the advantage oftargeting synergistic neuroprotective pathways. Thisstrategy uses peptidic vectors to carry large moleculesthrough the blood-brain barrier and the membrane ofbrain cells (42). Discovery that the activation of cas-pase-3 could be also important for cell survival (43) hasraised some doubts about use of its inhibitors in clinicalpractice. Therefore, other methods of stopping apopto-sis were taken into consideration, and inhibition of pro-tein kinases activated during the neuronal programmedcell death are intensively studied. There are evidencessuggesting that c-Jun-N-terminal kinase (JNK) pathwayis one the signaling cascades that mediates apoptoticneuronal cell death under a variety of stressful stimuli.The mixed-lineage kinase (MLK) family of kinases,which are key activators of the JNK pathway is impor-tant in this signaling pathway. CEP-1437 is a potent andselective inhibitor of members of the MLK family, sinceit is able to inhibit the activation of the JNK pathwayand to prevent neuronal cell death. It is under clinicaltrials for PD (44) Moreover, a cell-permeable peptidedesigned specifically to inhibit JNK was reported toreduces neuronal degeneration during ischemia via sup-pressing the extrinsic and intrinsic pathways of apopto-sis (45-47). Another serine threonine kinase, Rhokinase, plays a fundamental role in migration, prolifera-tion and survival of cells. Brain and spinal cord injury inadult mammals is accompanied by the activation of Rhokinases and inhibition of growth and sprouting of neu-rites. On the other hand, the inhibition of Rho kinaseactivity was reported to accelerate regeneration andrestore function of the nervous system in animal modelsof ischemia and inflammatory neurodegenerative dis-eases, thus constituting a new target for neuroprotection

(48). With respect to other kinases having a potential ofneuroprotection, one should mention the AMPñactivat-ed kinase (AMPK), which is the main sensor of intra-cellular energetic balance and is activated and phospho-rylated during hypoxia. An inhibitor of AMPK, C75 hasdecreased ischemic brain injury, whereas an activator ofthis enzyme showed opposite effect (49). Also variouscell-permeable calpain inhibitors have been synthesizedfor pharmacological inhibition of calpain activity andsome of them have shown significant neuroprotection inanimal models of the CNS injuries and diseases, indi-cating their therapeutic potential (50, 51). Since alter-ations in the expression of Bcl-2 family members occurin several neurodegenerative diseases (AD, PD, ALS,HD, TBI, stroke), methods to increase the overallexpression and/or function of anti-apoptotic Bcl-2 fam-ily members, and thus promote neuronal survival, areextensively studied. Most treatment efforts focus eitheron the targeted delivery of anti-apoptotic Bcl-2 familymembers into the affected brain regions of interest viaviral vectors, the generation of direct interactions of lowmolecular weight inhibitors with pro-apoptotic Bcl-2family members, or the induced expression of Bcl-2family members secondary to pharmacological manipu-lation (safe and efficacious Bcl-2 family mimetics) (52).

Antioxidants and mitochondrial stabilizers. Thebest known neuroprotectants and free radical-neutraliz-ing agents include polyphenols, e.g. quercetin, vitaminE, lazaroids and endogenously produced melatonin andestrogens (53, 54). Since melatonin, a hormone secretedfrom the pineal gland has been reported to have antiox-idant and metalñchelating properties, its dietary supple-mentation seems to be beneficial to slowing neurode-generative changes associated with the brain aging (55).A combination of iron chelation and antioxidant therapymay be one of significant approaches to neuroprotec-tion. Multifarious neuroprotective activities of green teacatechins, have brain-permeable, nontoxic, transitionmetal (iron and copper)-chelatable/radical scavengerproperties, that were shown in a wide array of cellularand animal models of neurological diseases (56). Newgeneration of synthetic antioxidants, such as tirilazad,ebselen, edarawon and NXY-059 have also additionalmetal-chelating properties, and they have been inten-sively studied in both animal models of ischemia and inclinical trials. Only ederawon has been accepted forstroke treatment in Japan, whereas NXY-059 is now inphase III clinical trials (57). It should be mentioned herethat some antioxidants such as vitamin C may in thepresence of iron via Fentonís reaction increase hydrox-yl radical formation and promote oxidative lipid damage(58). Beside antioxidants, agents restoring mitochondri-al function or preventing permeabilization its mem-branes are also extensively tested. Such agents includecoenzyme Q, creatine, inhibitors of the permeabilitytransition pore complex (in particular ligands ofcyclophilin D), openers of mitochondrial ATP-sensitiveor Ca2+-activated K+ channels, and membrane stabilizers(proteins from the Bcl-2 family, cytidine-5í-diphospho-choline) (59, 60). Also ketogenic strategy was proposedin the treatment some brain degenerative diseases sincethe sodium salt of the physiological ketone, D-beta-hydroxybutyrate (KTX 0101) was reported to protectneurons against damage caused by dopaminergic neuro-

361

toxins, beta-amyloid or ischemia by improving ATPproduction (61).

Anti-inflammatory strategies. Considering theinvolvement of inflammatory processes in mechanismof neurodegeneration, one should keep in mind thatactivated microglia and astrocytes are a rich source ofoxygen radicals, nitric oxide and neurotoxic and proin-flammatory cytokines. Nitric oxide has both neurotoxicand neuroprotective properties. It may attenuate neu-ronal damage through inhibition of NMDA receptoractivity. On the other hand, nitric oxide can be trans-formed to peroxynitrite, which inactivates numerousproteins including glutamate transporter, destroys DNAand deenergizes mitochondria, ultimately leading tonecrotic and apoptotic cell death (20). Clinical trialswith inhibitors of nitric oxide synthase in patients withstroke have not been successful, whereas some nitricoxide donors and its precursor L-arginine increase per-fusion and decrease ischemic brain damage. Althoughthe above data give rationale for further trials with nitricoxide donors, there are some doubts about their clinicalutility, since these substances possess a rather narrowtherapeutic time window (62, 63). Cyclooxygenases,COX-1 and COX-2 mediate the first step of prosta-glandin synthesis from arachidonic acid. Some data sug-gest that inducible COX-2 and the main product of itsenzymatic activity, prostaglandin E2, participate inischemia-related neuronal damage. In vitro studiesshowed that the stimulation of EP2 prostanoid recep-tors, in cAMP-dependent manner protected neuronalcells against damage evoked by NMDA or oxygen/glu-cose deprivation. These encouraging data have beenconfirmed in an animal model of transient brainischemia, which is in agreement with high level of EP2receptor expression in the neocortex, hippocampus andstriatum (64). Several epidemiological studies haveindicated that the long-term use of NSAIDs, most ofwhich are cyclooxygenase (COX) inhibitors, mayreduce the risk of Alzheimerís disease. For this reason,anti-inflammatory COX-inhibiting NSAIDs havereceived increased attention in experimental and thera-peutic trials for Alzheimerís disease. However, severalrecent efforts attempting to demonstrate a therapeuticeffect of NSAIDs in Alzheimerís disease have largelyfailed (65). Immunophilin ligands, such as FK506 andcyclosporin A, used in immunosuppression, were foundto be neuroprotective in several models of neuronal cellinjury (12). FK506 has been shown to interfere with theapoptotic pathway of neuronal cells, since it inhibitedJNK activity, cytochrome c release, caspase-3 activa-tion, and CD95 ligand expression. Those observationswere a clue to synthesis of new immunophilin ligands,such as GPI-1046 and V-10,367, which are nonim-munosuppressive derivatives of FK506. They have beenshown to exert neuroprotective and neuroregenerativeactions in several systems and could serve as therapiesfor neurodegenerative disorders (66, 67). Inflammatorymediators, such as TNFalpha and FasL play an impor-tant role in apoptosis in ALS and their expression is ele-vated in lumbar spinal cord section of both ALS patientsand transgenic animal model of this disease. It has beenfound that thalidomide and lenalidomide prolong sur-vival in the transgenic mouse model of ALS (G93Amice) by destabilizing DNA coding for TNFalpha and

other cytokines (68), thus being a potential drug fortreatment ALS (69). Interestingly, the same group ofinvestigators reported that also pioglitazone, a peroxi-some proliferator-activated receptor-gamma agonistwith anti-inflammatory properties, improved motor per-formance, delayed weight loss, attenuated motor neuronloss, and extended survival of G93A mice. Pioglitazoneappears to have therapeutic potential for human ALS,since it also reduces iNOS, NFkB, and 3-nitrotyrosineimmunoreactivity in the spinal cords of G93A trans-genic mice (70). Some antibiotics could also be benefi-cial in some brain diseases, for example, minocycline,which possesses anti-apoptotic and p38 MAPKinhibitory activity, was neuroprotective in acute braininjuries and ALS animal models and is currently studiedin clinical trials (71-74). Given the pathogenic impact ofoxidative stress and neuroinflammation, therapeuticstrategies aimed to blunt these processes are consideredan effective way to confer neuroprotection. Recently,the nuclear transcription factor Nrf2, that binds to theantioxidant response element (ARE) in gene promoters,has been reported to constitute a key regulatory factor inthe co-ordinate induction of a battery of endogenouscytoprotective genes, including those coding for bothantioxidant- and anti-inflammatory proteins suggestingthat targeting the Nrf2/ARE pathway may represent anovel therapeutic approach to the development of a newclass of neuroprotective drugs (75).

Multipotential agents and combined pharma-cotherapy. It appears that the successful treatment ofneurodegenerative diseases can be achieved by combin-ing a few drugs, which have neuroprotective properties.Such combinations could include for example antago-nists of glutamate receptors with inhibitors of apoptosis,or with antioxidants or anti-inflammatory agents (7, 10,21). Since dysfunction in acetylation homeostasis couldevoked neurodegeneration, agents which restore bal-anced acetylation status could diminish cell death (76).For example, the histone deacetylase inhibitor phenyl-butyrate (PBA) was effective in the G93A transgenicmouse model and a combination of this drug with thecatalytic antioxidant AEOL 10150 accelerated neuro-protection (77). In the same ALS model, the combina-tion of creatine with COX-2 inhibitors produced addi-tive neuroprotective effects and extended survival (78).Moreover, new, anti-apoptotic properties of drugs clini-cally used in symptomatic treatment of neurodegenera-tive diseases (memantine, selegiline, antiepilepticdrugs) are being discovered and are used in constructingnew multipotential drugs (11, 21, 79) For exampleladostigil, a derivate of rasagaline (monoaminooxidaseB inhibitor) with acetycholinesterase inhibitory activityis tested in AD and PD (11). In addition, a multipoten-tial drug, such as dexanabinol (synthetic cannabinoidwith NMDA antagonist, antioxidant and anti-TNF prop-erties) and AM36 (Na+ blocker, NMDA receptor antag-onist and anti-apoptotic agent) are tested in stroke (7).Other drugs, statins (HMG-CoA reductase inhibitors],exert cholesterol-independent pleiotropic effects thatinclude anti-thrombotic, anti-inflammatory, and anti-oxidative properties, and specific anti-excitotoxiceffects independent of 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA) reductase inhibition, which haspotential therapeutic implications (80). Beside exoge-

362

nous substances with the multipotential neuroprotectiveaction, there are several endogenously produced factorswhich also act as neuroprotectants at several steps of theneuronal cell death. Increasing their synthesis by phar-macological or genetic approaches could be also benefi-cial to protection against neurodegenerative diseases.

Endogenous neuroprotectants

Estrogens. Estrogens exhibit neuroprotectiveproperties in vivo and in vitro. They protect neuronsagainst different insults, such as anoxia, oxidativestress, hydrogen peroxide, iron or amyloid-β peptide(81), however, their effects seem to depend on the typeof neuronal tissue, stage of its development and neuro-toxic factors. Although estrogens have been found to beespecially important regulators of mitochondrial apop-totic pathway (81-84), they appear also to be involved incytokine-mediated apoptosis. Recent data suggest thatestrogens interfere with cytokine effects on the centralnervous system (CNS) by affecting cytokine signalingcomponents, such as nitric oxide (NO) synthesis andactivation of transcription factor NFκB (85-87).Astrocytes, which play a key role in estrogen-inducedsynapse formation, plasticity, and neuronal morpholo-gy, are the main target of estrogen in the brain (88, 89).It is possible that by releasing neuroprotective factors,such as transforming growth factor (TGF-β1, TGF-β2)or glial cell-derived neurotrophic factor (GDNF), astro-cytes, potentiate neuroprotection attributed to estrogen(90, 91). Under physiological conditions, IL-1β mayhave a neuroprotective action through positive modula-tion of nerve growth factor (NGF)) (92). Since the hip-pocampus is more abundant in Il-1 receptors than theneocortex, it could also be more effectively protected byestradiol, which is known to interact with growth fac-tors, such as brain-derived neurotrophic factor (BDNF),insulin-like growth factor-I (IGF-I) or NGF and theirreceptors (93). An inflammatory reaction that con-tributes to the tissue damage after cerebral ischemia orto the progression of neurodegenerative disease is medi-ated in part by toxic amounts of nitric oxide (NO) andexaggerated production of pro-inflammatory cytokines.Estrogens were found to inhibit inflammatory reactionin astroglial and microglial cells, possibly via estrogenreceptor-dependent activation of MAP kinase (94).Estrogens are neuroprotective in a variety of in vitro andin vivo models of cerebral ischemia (95). There is abody of evidence indicating that during brain ischemiathe physiological estrogen stimulation results in anincreased release of vasolidating substances, whichimprove brain metabolism and cerebral blood flow.Estrogen-mediated neuroprotection against ischemialikely involves the activation of gamma PKC throughthe G-protein-coupled estrogen receptors on the plasmamembrane (96). However, the non-estrogen receptormechanism has also been suggested (97). There is anumber of evidences suggesting a direct interference ofestrogens with apoptotic processes, especially thosetriggered by mitochondrial pathway. Estradiol has beenfound to stimulate the expression of anti-apoptotic pro-teins, like Bcl-2 or Bcl-xl, which prevent cytochrome cefflux from mitochondria to cytoplasm that, finally,results in the inhibition of caspase-dependent apoptoticcell death. Estrogen-induced Bcl-2 up-regulation was

found in arcuate nucleus neurons of female rats in vivo(82) and in NT2 neuronal cells in vitro (98), while estra-diol effects on Bcl-xl were observed in vivo and in vitroin hippocampal and cortical cells (99). An anti-apoptot-ic action of estrogens may also depend on suppressionof pro-apoptotic gene transcription, such as Nip-2 orBad (100-103), possibly through the AP-1 site down-stream of c-Jun amino N-terminal kinases (JNKs) andcaspase-3 activation, as indicated in nigral dopaminer-gic neurons in rat primary cultures (104). Recently,estradiol has been found to prevent caspase-6-mediatedneuronal cell death, possibly by inducing a caspaseinhibitory factor (CIF) through a receptor-mediated, butnon-genomic, pathway (105). Moreover, estrogens cancounteract apoptotic insults by affecting the basal levelsof tau phosphorylation at a site known to be phosphory-lated by glycogen synthase kinase-3β (GSK-3β), there-by acting as antioxidants and interacting directly withcellular and mitochondrial membranes (106-109).

Finding the most suitable and effective protectivestrategy needs evaluation of risk-to-benefit ratio. Thus,issues of thrombotic and neoplastic risks must be fac-tored into the design of estrogen alternatives for hor-mone replacement therapy. Selective estrogen receptormodulators (SERMs) may represent an alternative toestrogen for the treatment or the prevention of neurode-generative disorders. SERMs represent a class compris-ing a growing number of compounds that act either asestrogen receptor agonists or antagonists in a tissue-spe-cific manner distinct from that of estradiol.Neuroprotection attributed to estradiol is associatedwith a strong down-regulation of reactive astroglia,while SERMs do not affect reactive gliosis (110).SERMs presently in use are tamoxifen and raloxifene.Next-generation SERMs taken into clinical studiesinclude idoxifene, droloxifene, ospemifene, arzoxifene,acolbifene/EM-800, levormeloxifene, lasofoxifene (CP-336,156), bazedoxifene (TSE-424) and HMR 3339(111).

Multiple issues regarding the timing, formulationand duration of the estrogen therapy intervention remainunresolved. The results from hormone replacementstudies in humans are thus far inconclusive. A possiblealternative to hormonal replacement therapy is toincrease local steroidogenesis by neural tissues, whichexpress enzymes for steroid synthesis and metabolism.The rate-limiting step in steroid biosynthesis is theintramitochondrial transport of cholesterol, a processthat is mediated by the steroidogenic acute regulatoryprotein (StAR). The importance of StAR has been illus-trated by analyses of patients with lipoid congenitaladrenal hyperplasia, a disorder that markedly disruptssteroidogenesis (112). Although StAR plays an essentialrole in peripheral endocrine glands, its presence and rolein the brain had been previously questioned (113).However, a number of studies have confirmed a localincrease in the levels of pregnenolone and progesteronefollowing traumatic injury in the brain and spinal cord.The expression and activity of aromatase, which syn-thesizes estrogen, was also found to be elevated ininjured brain (114). Thus, StAR becomes an attractivepharmacological target to promote neuroprotection.

Neuropeptides. A number of peptides prevent neu-ronal damage in both in vitro and in vivo experiments.

363

The most potent are pituitary adenylate cyclase activat-ing polypeptide (PACAP), vasointestinal peptide (VIP)and activity dependent neurotrophic factor (ADNF)which ameliorate glutamate-mediated toxicity at as lowas femtomolar concentrations. The molecular mecha-nism of action of those neuropeptides has not been fullyelucidated, however, activation of adenylate cyclase andMAP kinases, inhibition of JNK kinases and inductionof some cell protecting proteins are likely to participatein this process (115, 116). PACAP also inhibits oxida-tive stress-induced, caspase 3 activity and attenuatesinterleukin-6 and chemokine release from astrocytes.Intravenous administration of PACAP prevents hip-pocampal pyramidal cell death in animal models ofbrain ischemia. VIP protects neurons by regulatingcytokine release and stimulating synthesis of ADNF inglial cells. ADNF and its derivatives (ADNF-9, ADNF-14) at femtomolar concentrations exert long-lasting pro-tective effects on neurons against oxidative stress, exci-totoxicity and hypoglycemia. These peptides alsoincrease production of anti-apoptotic transcription fac-tor NFκB and heat shock protein hsp-60. Recently, ahybrid peptide composed of ADNF and another bioac-tive peptide, humanin was shown to suppress amyloid-beta toxicity and memory deficit in mice (117). A sub-stantial evidence points also to neuroprotective proper-ties of other neuropeptides, e.g. thyrotropin releasinghormone (TRH), neuropeptide Y, and galanin. Thesepeptides acting on their specific receptors protect cellsvia presynaptic inhibition of glutamate release, and/orpostsynaptically hyperpolarizing neuronal membrane(118, 119). Derivates of TRH, novel dipeptides havemultipotential actions that make them candidates for thetreatment of both acute and chronic neurodegeneration(120). N-Acetylaspartylglutamate (NAAG) is the mostabundant and widely distributed peptide transmitter inthe mammalian nervous system. NAAG activates themetabotropic glutamate mGlu(3) receptor at presynapticsites, inhibiting the release of neurotransmitters, includ-ing glutamate, and activates mGlu(3) receptors on glialcells, stimulating the release of neuroprotective growthfactors from these cells. NAAG is inactivated by specif-ic peptidases following its synaptic release. Novel com-pounds that inhibit these enzymes prolong the activityof synaptically released NAAG and have significanttherapeutic efficacy in animal models of stroke, trau-matic nervous system injury, amyotrophic lateral scle-rosis (121).

Neurosteroids. Neurosteroids, which are a sub-class of steroids including dehydroepiandrosterone(DHEA), dehydroepiandrosterone sulfate (DHEAS),pregnenolone, pregnenolone sulfate (PREGS), proges-terone and allopregnanolone, are found in abundance inthe developing central nervous system (CNS). This sug-gests that they have trophic factor-like activity or caninteract with various neurotransmitter systems to pro-mote neuronal remodeling. The decreased levels ofDHEAS have been hypothesized to contribute to theincreased vulnerability of the ageing or stressed humanbrain to ischemia and other neurodegenerative diseases.It was reported that there was a correlation between highlevels of key proteins implicated in the formation ofplaques and neurofibrillary tangles and decreased brainlevels of PREGS and DHEAS in humans, suggesting a

possible neuroprotective role of these neurosteroids inAD (122). Neurosteroids are multifunctional owning tothe modulation of activity and expression of severalreceptors (NMDA, GABA, sigma receptors). They weredemonstrated experimentally to inhibit excitotoxicityand apoptosis, so their mechanism of action in neuro-protection is under further elucidation (123-125). Inorder to avoid side effects of neurosteroids in the non-damaged areas in the brain and in peripheral tissues,indirect methods of increasing neurosteroid synthesisare proposed. This approach was applied mainly to lig-ands of peripheral-type benzodiazepine receptors,which upon activation regulate brain dysfunction byaccelerating peripheral-type benzodiazepine receptor-dependent cholesterol transport into mitochondria andincreasing formation of neuroactive steroids. InAlzheimerís disease, neurosteroid biosynthesis wasreported to be altered, namely there was a decrease inthe level of the intermediate 22R-hydroxycholesterol,whereas this steroid was found to exert neuroprotectiveaction against beta-amyloid neurotoxicity. Moreover,basal expression of the peripheral-type benzodiazepinereceptor was up-regulated in a number of neuropatholo-gies, including gliomas and neurodegenerative disor-ders, as well as in various forms of brain injury andinflammation. Hence, the use of a stable spirostenolderivative which was found to be neuroprotective, sug-gests that compounds developed based on critical inter-mediates of neurosteroid biosynthesis could offer novelmeans for neuroprotection (126).

Neurotrophic factors. Among the most widelystudied growth factors in relation to neurodegenerativedisorders is insulin-like growth factor I (IGF-I). IGF-I isimportant for fetal and postnatal development, but italso controls tissue homeostasis throughout life via theregulation of cell proliferation and apoptosis. IGF-I hasmultiple actions at different control points of apoptosis,including the Bcl-2 family proteins, inhibitors of cas-pases, and signaling of death-inducing receptors acti-vates prosurvival PI3K/Akt or MAPK/Erk signalingpathways (127, 128). Evidence that IGF-I rescues motorneurons has led to clinical trials of human recombinantIGF-I in ALS patients (74). However, systemic deliveryof human recombinant IGF-I in these trials did not leadto beneficial clinical effects in ALS patients that mighthave been due to the inactivation of IGF-I by binding toIGF binding proteins (IGFBPs), and/or limited deliveryof IGF-I to motor neurons. The IGF analogues with lowaffinity for IGFBPs and analogues that are able to dis-place IGF-I from IGFBPs are better candidates for newclinical trials. Another possibility is to find a way ofIGF-I transport without hindering the circulating andtissue-specific IGFBPs, namely IGF-I delivery based ongene therapy. Other polypeptide trophic factors, likebasic fibroblast growth factor (bFGF), osteogenic pro-tein-1 (OP-1), vascular endothelial growth factor (Veg-f), granulocyte colony stimulating factor (G-CSF), stemcell factor (SCF) and erythropoietin show beneficialeffects in experimental model of acute brain ischemiaand in restoring the central nervous function (129-131).However, only bFGF and erythropoietin have beenadmitted to further clinical trials in patients with acutestroke (132). Erythropoietin is an oxygen-dependentcytokine, which enhances red blood cell production by

364

inhibition of progenitor cell apoptosis. Besides regulat-ing proliferation and differentiation of erythroidal cells,erythropoietin and its receptor seem to be involved inthe central nervous system functions. It ameliorates neu-ronal injury because of antiapoptotic, antiinflamatory,antioxidant, angiogenic, neurogenic and neurotrophicproperties (133, 134). Intracellular mechanism of bene-ficial effects of erythropoietin on neurons involves theactivation of Jak2/Stat3 and P13/Akt pathways andstimulation of antiapoptotic protein (Bclxl) synthesis(135). Both erythropoietin and darbepoietin, a recombi-nant protein with erythropoiesis-stimulating activity,decrease ischemia-related brain damage and neurologi-cal deficit (131, 134, 136, 137). On the other hand,asialoerythropoietin, a neuroprotective variant of ery-thropoietin that lacks erythropoietic action was noteffective in the R6/2 transgenic Huntingtonís diseasemice (138). Granulocyte-colony stimulating factor (G-CSF) is a growth factor that orchestrates the prolifera-tion, differentiation, and survival of hematopoietic pro-genitor cells but there is a growing body of evidencefrom experimental studies suggesting that G-CSF alsohas important nonhematopoietic functions in the centralnervous system. The neuroprotective actions of G-CSFhave mainly been attributed to its anti-inflammatory andanti-apoptotic effects, induction of neurogenesis andangiogenesis. G-CSF is a potential new agent for neuro-protection but precise mechanisms of G-CSF-inducedneuroprotection must be further elucidated before itenters clinical trials (139). The main obstacle for wideclinical use of trophic factors is their polypeptide struc-ture and inability to penetrate through blood-brain bar-rier. To solve this problem, a chimerical brain-derivedneurotrophic factor (BDNF) has been synthesized, inwhich the BDNF moiety was bound with monoclonalantibody against transferrin receptors expressed in theblood-brain barrier. This chimerical peptide after intra-venous administration easily penetrates to the centralnervous system and exerts strong protective effects onneurons in models of brain ischemia (140). Infusion ofBDNF into the basal ganglia and the use of viral vectorsare also tested (141-144). However, these interestingand promising approaches have to be verified in furtherexperiments. Constructing low molecular weight pep-tides with neuroprotective and neurotrophic properties(e.g. cerebrolysin), which possess wider therapeutictime window could be also beneficial in neurodegener-ation (145). Alternatively, the stimulation of synthesisof endogenous neuroprotective factor has been proposedas another method of treatment of neurodegenerativedisorders. From this perspective, the conditional pre-conditioning seems to be particularly interesting. It hasbeen found that exposure of neurons to short-lastinghypoxia or sublethal concentration of NMDA enhancesexpression of thioredoxin, transcription factors, likeNFκB, neurotrophins and heat shock proteins (HSP),which can prevent subsequent neuronal damage result-ing from lethal insult (20, 146). Alternatively, neu-rotrophin synthesis can be stimulated by some pharma-cological agents such as vitamin D3 and some of itsanalogs which may ameliorate postischemic brain injury(147). In line with these data, it has been reported thatsome low-calcium analogues at nanomolar concentra-tions protect neuronal cells against NMDA- and hydro-

gen peroxide-induced damage in vitro, however, theymay be less efficient in seizure-induced brain damage(148). Recently, it has been often underlined that phys-ical exercises, cognitive stimulation and diet restrictioncould be a promising method to prevent or delay occur-rence of neurodegenerative diseases. Evidence suggeststhat increased physical activity and low caloric dietscould increase expression of endogenous neuroprotec-tants such as IGF-1 and BDNF (1, 5, 149).

Endocannabinoids. Evidence has accumulatedover the last few years suggesting that endocannabi-noid-based drugs may be potentially useful to reduce theeffects of neurodegeneration. In fact, exogenous andendogenous cannabinoids were shown to exert neuro-protection in a variety of in vitro and in vivo models ofneuronal injury, and the release of endocannabinoidsduring neuronal injury may constitute a protectiveresponse. The inhibitory effect of cannabinoids on reac-tive oxygen species, glutamate and tumor necrosis fac-tor and activation of the phosphatidylinositol 3-kinase/protein kinase B pathway, induction of phospho-rylation of extracellular regulated kinases and theexpression of transcription factors and neurotrophinspoint to its multipotential activity. Dexanabinol (HU-211), a synthetic cannabinoid, is currently beingassessed in clinical trials for traumatic brain injury andstroke (150, 151).

Neuroprotective amino acids. Some amino acids,such as taurine and kynurenic acid possess neuroprotec-tive activity. Taurine is released from neuronal tissueupon depolarization and prevents calcium uptake, thusinhibiting excitotoxicity. Kynurenic acid is an endoge-nous broad-action antagonist of EAA receptors. It blocksa majority of ionotropic glutamate receptors, but prefer-entially acts on glycine binding to NMDA receptor com-plex (152). Recently, it has been demonstrated that ener-getic deficit decreases the activity of kynurenic acid syn-thesizing enzyme, and that insufficient production of thisacid may participate in neurodegeneration induced by anoxidative phosphorylation inhibitor (153).

Gene- and cell-based therapies

Genetic approaches to the treatment of neurode-generative disorders involve mainly constructing ofviral vectors for targeted neurotrophic protein deliveryas mentioned before [142]. Also viral vector-mediatedgene delivery of a dopamine-synthesizing enzyme intothe striatum was reported to restore local dopamine pro-duction and allowed for behavioral recovery in animalmodels of PD (154). Another genomic strategy, whichcould be beneficial in the treatment of neurodegenera-tive disorders, is the RNA interference (RNAi) methodin silencing the expression of specific toxic genes, but itis still only experimentally studied. Recent studies haveshown that cultured neurons can be efficiently transfect-ed with siRNA and the targeted genes (e.g. p75 neu-rotrophin receptor, pro-apoptotic proteins from Bcl-2family, caspases) were effectively silenced (155).Research efforts in the RNAi field aim to gain a betterunderstanding of how its underlying machinery isorchestrated, to define the biological role of this con-served pathway, to determine how to effectively manip-ulate RNAi in the laboratory and to integrate all thisknowledge to develop novel therapies for human dis-

365

eases (156, 157). It is a more promising method, sinceclinically relevant RNAi-mediated gene silencing innon-human primates was demonstrated (158).

Alternative strategy in combating neurodegenera-tion is transplantation of specific neurons to the damagedbrain regions. The most advanced progress has beenmade in Parkinsonís disease. So far, cell replacementtherapy in Parkinsonís disease (PD) has been based onthe use of primary dopaminergic (DA) neuroblastsobtained from the brain of aborted human fetuses.Clinical trials show that intrastriatal DA neuron trans-plants can give substantial symptomatic relief inadvanced PD patients but there are also several prob-lems. First, graft survival and clinical outcome has beentoo variable so far, suggesting that DA neuron grafts maynot be equally effective in all PD patients. Secondly, ithas become clear that immune mechanisms leading toslowly developing inflammatory responses may compro-mise long-term graft survival and function. Third, theproblems associated with the use of tissue from abortedfetuses make it necessary to develop alternative sourcesof cells for transplantation. Recent progress in the gener-ation of DA neuroblasts from neural progenitors andembryonic stem cells suggests that these kinds of cellsmay offer more accessible, defined and standardizedsources of cells for clinical transplantation in PD (159).The hematopoietic system is one of the best character-ized human cellular system in which a multipotent adultstem cell is the origin of all the cells of a tissue. In thelast years, the use of stem cells has given rise to manyhopes in regenerative medicine, especially in diseaseswithout efficient therapies. However, the hematopoieticsystem in which stem cell transplantation has enteredclinical practice since many years has also shown thelimits of these approaches (160).

Animal models

Taking advantage of significant progress in molec-ular genetics, a number of new animal models of neu-rodegenerative diseases have been recently proposed.With respect to ALS, the transgenic mice harboring ahuman Cu2+/Zn2+ superoxide dismutase 1 (SOD1) trans-gene containing the G93A mutation show impaired thecapacity of spinal cord high-affinity glutamate uptake.Also transgenic rats expressing human SOD1 G93Awith ALS-like phenotype, motor neuron degeneration inthe spinal cord and reduced sensitivity to riluzole havebeen generated (161, 162). Another model of ALS, i.e.transgenic mice overexpressing a mutant SOD1 with ahistidine to arginine substitution at position 46 (H46RSOD1), was successfully used for demonstrating of neu-roprotective activity of oxidized recombinant humangalectin-1 (163). For studying mechanism ofAlzheimerís disease and putative drugs, the PDAPPmice demonstrating age-dependent accumulation ofamyloid beta (42)-containing plaques and decreasedadult hippocampal neurogenesis seem to be an appro-priate model (164). Much effort has been put into devel-oping models of Huntingtonís disease, which is causedby a polyglutamine expansion in the gene encoding thehuntingtin protein (165). The rat malonate model of thisdisease may be useful for testing putative drugs. Forexample, neuroprotective effects of intrastriatal injec-tion of M826, a reversible caspase-3 inhibitor was evi-

denced in this model (166). However, other models onyeast, Caenorhabditis elegans, Drosophila melano-gaster and mouse have also been developed (167). Animpressive progress has been made in the generation ofParkinsonís disease models. It is firmly established thatthe increased level of iron, MAO-B activity, oxidativestress, inflammatory processes, glutamatergic excitoxic-ity, nitric oxide synthesis, abnormal protein folding andaggregation, reduced expression of trophic factors mayall contribute to the mechanism of substantia nigradopaminergic cell death. In animal models, some neuro-toxins, like MPTP, 6-OHDA and met-amphetamine arecommonly used as inducers of SN dopaminergic celldamage (168). Familial forms of Parkinsonís disease areassociated with mutation of several genes coding foralpha-synuclein, parkin, DJ-1, UCHL1, Pink1 andLRRK2. For constructing trangenic models ofParkinsonís disease, a member of the nuclear receptorsuperfamily (NURR1) and a homebox transcription fac-tor (PITX3) involved in differentiation and developmentof SN neurons are also employed. The genetic modelshave generally good construct, face, and predictivevalidity. However, the alpha-synuclein overexpressed,parkin knockout and DJ-1 knockout mouse phenotypesare not as profound as the Nurr 1 heterozygotes andPITx3 aphokia mice, but they may provide insight intoearly stages of this disease, whereas the Nurr1 het-erozygotes and Pitx3 are good models to study the laterstages of Parkinsonís disease (169).

CONCLUSIONS

Despite enormous progress in biological and medici-nal sciences, no efficient neuroprotective drug has beendesigned so far. Neurochemical, genomic and proteomicstudies on the mechanism of neurodegeneration provide uswith a plethora of new data, however, there are still seriousproblems with their straightforward interpretation. Takinginto account more and more complex biochemical cascadesof neurodegenerative processes, it has been postulated thatpharmacological multipotential or combined approachesmay be most appropriate in neuroprotective strategies.Furthermore, better animal models of neurodegenerativediseases are required, and in vitro models should be basedon human rather than animal neuronal cell lines, since thecomposition of protein targets for new drugs may showsome species differences. A number of neuroprotective sub-stances possess a narrow time-window, have dual activity(pro- or anti-apoptotic, depending on the concentration, thetype of a neuronal tissue and the stage of its development),or unfavorable pharmacokinetic properties.

It appears that further improvement of the outcomeof neurodegenerative disease treatments may dependnot only on a better understanding of molecular mecha-nism of neuronal injury, but also on biopharmaceuticalproperties of new drugs and on properly designed clini-cal studies.

REFERENCES

1. Mattson M.P. and Magnus T.: Nat. Rev. Neurosci.7, 278 (2006).

2. Zecca L. et al.: J. Clin. Path. Mol. Pathol. 54, 414(2001).

366

3. Longo F.M., Massa S.M.: NeuroRx 1, 117 (2004).4. Aisen P.S.: CNS Drugs 19, 989 (2005).5. Fumagalli F. et al.: Pharmacogenom. J. 6, 95 (2006).6. Yakovlev A.G., Faden A.I.: NeuroRx 1, 5 (2004).7. Muir K.M., Teal P.A.: J. Neurol. 252, 1011 (2005).8. Dauer W., Przedborski S.: Neuron 39, 889 (2003).9. Dib M.: Drugs 65, 2463 (2005).

10. Wu S.S., Frucht S.J.: CNS Drugs 19, 723 (2005).11. Youdim M.B.H. et al.: Nat. Rev. Neurosci. 7, 295

(2006).12. Kaminska B., Zawadzka M.: in Neuroprotection (in

Polish), manual of XX Winter School of Institute ofPharmacology, Polish Academy of Sciences,KrakÛw 2003, p. 41.

13. Loetscher H. et al.: Drug Discov. Today 6, 671 (2001).14. Mattson M.P.: Neuromolecular Med. 3, 65 (2003).15. Boehning D. et al.: Nat. Cell Biol. 5, 1051 (2003).16. Scorrano L. et al.: Science 300 (5616), 135 (2003).17. Tabner B.J. et al.: Biochem. Soc. Trans. 33, 1082

(2005).18. Potashkin J.A. et al.: Antioxid. Redox Signal. 8,

144 (2006).19. Aracena P. et al.: Biol. Res. 39, 157 (2006).20. Lazarewicz J., Salinska E.: in Neuroprotection (in

Polish), manual of XX Winter School of Institute ofPharmacology, Polish Academy of Sciences,KrakÛw 2003, p. 9.

21. Lipton S.A.: NeuroRx 1, 101 (2004).22. Wood P.L.: Drugs 8, 229 (2005).23. Wang R., Zhang D.: Eur. J. Neurosci. 22, 2376 (2005).24. Nikam S.S., Meltzer L.T.: Curr. Pharm. Des. 8, 845

(2002).25. Javitt D.C.: Mol. Psychiatry 9, 984 (2004).26. Danysz W., Parsons C.G.: Neurotox. Res. 4, 119

(2002).27. Marenco S., Weinberger D.R.: CNS Drugs 20, 173

(2006).28. Jayakar S.S., Dikshit M.: Int. J. Neurosci. 114, 695

(2004).29. Weiser T.: Curr. Drug Targets CNS Neurol. Disord.

4, 153 (2005).30. Danysz W.: Curr. Opin. Investig. Drugs 3, 1081

(2002).31. Erdo F. et al.: Brain Res. Bull. 68, 269 (2006).32. Marino M.J. and Conn P.J.: Curr. Opin. Pharmacol.

6, 98 (2006).33. Baskys A. et al.: Neuropharmacology 49 (Suppl 1),

146 (2006).34. Baskys A. et al.: Ann. N.Y. Acad. Sci. 1053, 55

(2005).35. Chong Z.Z. et al.: Curr. Neurovasc. Res. 2, 197 (2005).36. Marino M.J. et al: Curr. Top. Med. Chem. 5, 885

(2005).37. Mathew S.J. et al.: Rev. Bras. Psiquiatr. 27, 243 (2005).38. Xiong Z.G. et al.: J. Membr. Biol. (2006) Apr 17;

[Epub ahead of print].39. Xiong Z.G. et al.: Cell 118, 687 (2004).40. Legos J.J. et al.: Expert Opin. Emerg. Drugs 6, 81

(2001).41. Charriaut-Marlangue C.: Therapie 59, 185 (2004).42. Onteniente B.: Curr. Drug Targets CNS Neurol.

Disord. 3, 333 (2004).43. McLaughlin B.: Apoptosis 9, 111 (2004).44. Silva R.M. et al.: Mov. Disord. 20, 653 (2005).

45. Gao Y. et al.: J. Cereb. Blood Flow Metab. 25, 694(2005).

46. Bonny C. et al.: Rev. Neurosci. 16, 57 (2005).47. Guan Q.H. et al.: Brain Res. (2006) May 1 [Epub

ahead of print].48. Mueller B.K. et al.: Nat. Rev. Drug Discov. 4, 387

(2005).49. McCullough L.D. et al.: J. Biol. Chem. 280, 20493

(2005).50. Zalewska T.: in Neuroprotection (in Polish), manual

of XX Winter School of Institute of Pharmacology,Polish Academy of Sciences, KrakÛw 2003, p. 51.

51. Ray S.K. et al.: Curr. Drug Targets CNS Neurol.Disord. 2, 173 (2003).

52. Shacka J.J. et al: Curr. Drug Targets CNS Neurol.Disord. 4, 25 (2005).

53. Hardeland R. et al: Int. J. Biochem. Cell Biol. 38,313 (2006).

54. Mandel S., Youdim M.B.: Free Radic. Biol. Med.37, 304 (2004).

55. Lahiri D.K. et al.: Ann. N.Y. Acad. Sci. 1035, 216(2004).

56. Mandel S. et al.: Mol. Nutr. Food Res. 50, 229 (2006).57. Green A.R. et al.: Curr. Drug Targets CNS Neurol.

Disord. 4, 109 (2005).58. Bailey D.M. et al.: Free Radic. Biol. Med. 40, 591

(2006).59. Mattson M.P., Kroemer G.: Trends Mol. Med. 9,

196 (2003).60. Hurtado O. et al.: Neurobiol. Dis. 18, 336 (2005).61. Smith S.L. et al.: CNS Drug Rev. 11, 113 (2005).62. Szczudlik A.: in Neuroprotection (in Polish), manual

of XX Winter School of Institute of Pharmacology,Polish Academy of Sciences, KrakÛw 2003, p. 99.

63. Willmot M. et al: Free Radic. Biol. Med. 39, 412(2005).

64. McCullough L. et al.: J. Neurosci. 24, 257 (2004).65. Ho L. et al.: CNS Drugs. 20, 85 (2006).66. Klettner A., Herdegen T.: Curr. Drug Targets CNS

Neurol. Disord. 2, 153 (2003).67. Ganel R. et al.: Neurobiol. Dis. 21, 556 (2006).68. Kiaei M. et al.: J Neurosci. 26(9), 2467 (2006). 69. Greig N.H. et al.: Ann. N.Y. Acad. Sci. 1035, 290

(2004).70. Kiaei M. et al.: Exp. Neurol. 191, 331 (2005).71. Elewa H.F. et al.: Pharmacotherapy 26, 515 (2006).72. Festoff B.W. et al: J. Neurochem. (2006) Apr 21;

[Epub ahead of print].73. Pontieri F.E. et al.: Neurol. Sci. 26, 285 (2005).74. Nirmalananthan N., Greensmith L.: Curr. Opin.

Neurol. 18, 712 (2005).75. van Muiswinkel F.L., Kuiperij H.B.: Curr. Drug

Targets CNS Neurol. Disord. 4, 267 (2005).76. Saha R.N., Pahan K.: Cell Death Differ. 13, 539

(2006).77. Petri S. et al.: Neurobiol. Dis. 22, 40 (2006).78. Klivenyi P. et al.: J Neurochem. 88, 576 (2004).79. Willmore L.J.: Epilepsy Behav. 7 (Suppl 3), 25

(2005).80. Bosel J. et al.: J. Neurochem. 92, 1386 (2005).81. Garcia-Segura L.M. et al.: Prog. Neurobiol. 63, 29

(2001).82. Garcia-Segura L.M. et al.: Neuroreport 9, 593 (1998).83. Kajta M. et al.: Neuroscience 123, 515 (2004).

367

84. Kajta M., Beyer C.: Endocrine 1, 3 (2003).85. Cerillo G. et al.: J. Steroid Biochem Mol. Biol. 67,

79 (1998).86. Dodel R.C. et al.: J. Neurochem. 73, 1453 (1999).87. Drew P.D., Chavis J.A.: J. Neuroimmunol. 111,

77 (2000).88. Garcia-Segura L.M. et al.: J. Neurobiol. 40, 574 (1999).89. Mong J.A. et al.: J. Neurosci. 19, 1464 (1999).90. Dhandapani K.M. et al.: Endocrinology 146, 2749

(2005).91. Platania P. et al.: Neurobiol. Dis. 20, 461 (2005).92. Di Loreto S. et al.: J. Neuroimmunol. 106, 32. (2000).93. Amantea D. et al.: Pharmacol. Res. 52, 119 (2005).94. Bruce-Keller A.J. et al.: Endocrinology 141, 3646

(2000).95. Simpkins J.W. et al.: Cell Mol. Life Sci. 62, 271

(2005).96. Hayashi S. et al.: J. Neurochem. 93, 883 (2005).97. Cordey M., Pike C.J.: J. Neurochem. 96, 204 (2006).98. Singer C.A. et al.: Neuroreport 9, 2565 (1998).99. Pike C.J.: J. Neurochem. 72, 1552 (1999).

100. Garnier M. et al.: J. Neurosci. 17, 4591 (1997).101. Chan S.L. et al.: J. Neurochem. 73, 502 (1999).102. Gollapudi L., Oblinger M.M.: J. Neurosci. Res.

56, 99 (1999).103. Stolzner S.E. et al.: Neuroreport 12, 2797 (2001).104. Sawada H. et al.: FASEB J. 14, 1202 (2000).105. Zhang Y. et al.: J. Neurosci. 21, RC176 (2001).106. Behl C. et al.: Mol. Pharmacol. 51, 535 (1997).107. Green P.S. et al.: Neuroscience 84, 7 (1998).108. Liang Y. et al.: Brain Res. Bull. 54, 661 (2001).109. Goodenough S. et al.: Neuroscience 132, 581 (2005).110. Ciriza I. et al.: J. Neurobiol. 61, 209 (2004).111. Vogelvang T.E. et al.: Drugs 66, 191 (2006).112. Manna P.R., Stocco D.M.: Curr. Drug Targets

Immune Endocr. Metabol. Disord. 5, 93 (2005).113. Kim H.J. et al.: Arch. Histol. Cytol. 67, 383 (2004).114. Garcia-Ovejero D. et al.: Brain Res. Brain Res.

Rev. 48 273 (2005).115. SÍdkowska P., Nowak J.: in Neuroprotection (in

Polish), manual of XX Winter School of Instituteof Pharmacology, Polish Academy of Sciences,KrakÛw 2003, p. 197.

116. Dogrukol-Ak D., Tore F.: Curr. Pharm. Des. 10,1325 (2004).

117. Chiba T. et al.: J. Neurosci. 25, 10252 (2005).118. Shi T.J. et al..: Eur. J. Neurosci. 23, 627 (2003).119. Silva A.P. et al.: Curr. Drug Targets CNS Neurol.

Disord. 4, 331 (2005).120. Faden A.I. et al.: J. Alzheimers Dis. 6(6 Suppl), 93

(2004).121. Neale J.H. et al.: Trends Pharmacol. Sci. 26, 477

(2005).122. Weill-Engerer S. et al.: J. Clin. Endocrinol. Metab.

87, 5138 (2002).123. Compagnone N.A., Mellon S.H.: Front. Neuro-

endocrinol. 21, 1 (2002).124. Lapchak P.A., Araujo D.M.: Int. Rev. Neurobiol.

46, 379 (2001).125. Kaasik A. et al.: J. Steroid Biochem. Mol. Biol.

87, 97 (2003).126. Papadopoulos V. et al.: Neuroscience 138, 749 (2006).127. Kooijman R.: Growth Factor Rev. 2006 Apr 15;

[Epub ahead of print].

128. Aberg N.D. et al.: ScientificWorld Journal. 18, 53(2006).

129. Dhandapani K.M. et al.: J. Neurochem. 95, 9 (2005).130. Zachary I.: Neurosignals 14, 207 (2005).131. Ehrenreich H. et al.: Metab. Brain Dis. 19, 195

(2004).132. Ren J.M., Finklestein S.P.: Curr. Drug Targets

CNS Neurol. Disord. 4, 121 (2005).133. Genc S. et al.: Restor. Neurol. Neurosci. 22, 105

(2004).134. Gassmann M. et al.: Adv. Exp. Med. Biol. 543,

323 (2003).135. Kretz A,. et al.: Mol. Cell Neurosci. 29, 569

(2005).136. Belayev L. et al.: Stroke 36, 1071 (2005).137. Lewis LD.: Semin. Hematol. 41, (Suppl 7), 17 (2004).138. Gil J.M. et al.: Neurosci. 10, 5 (2004).139. Solaroglu I. et al.: Stroke 37, 1123 (2006).140. Pardridge W.M.: Adv. Exp. Med. Biol. 513, 397

(2002).141. Kilic U. et al.: Stroke 34, 1304 (2003).142. Thanos C., Emerich D.: Expert Opin. Biol. Ther.

5, 1443 (2005).143. Wu D.: NeuroRx 2, 120 (2005).144. Dowd E. et al.: Eur. J. Neurosci. 22, 2587 (2005).145. Schauer E. et al.: J. Neural Transm. 2005 Dec 14;

[Epub ahead of print].146. Meloni B.P. et al.: Proteomics. 5, 4743 (2005).147. Losem-Heinrichs E. et al.: Arch. Biochem.

Biophys. 439, 70 (2005).148. Tetich M. et al.: Brain Res. 1039, 1 (2005).149. Carro E. et al.: J. Neurosci. 21, 5678 (2001).150. Croxford J.L.: CNS Drugs. 17, 179 (2003).151. van der Stelt M., Di Marzo V.: Neuromolecular

Med. 7, 37 (2005).152. UrbaÒska E.M. et al.: Brain Research 878, 210

(2000).153. Luchowska L. et al.: Acta Neurobiol. Exp. (Wars).

63, 101 (2003).154. Li X.G. et al.: Mol. Ther. 13, 160 (2006).155. Milhavet O. et al.: Pharmacol. Rev. 55, 629 (2003).156. Rodriguez-Lebron E., Paulson H.L.: Gene Ther.

13, 576 (2006).157. Rodriguez-Lebron E., Gonzalez-Alegre P.: Expert

Rev. Neurother. 6, 223 (2006).158. Zimmermann T.S. et al.: Nature 441(7089), 111

(2006). 159. Bjorklund A.: Novartis Found. Symp. 265, 174 (2005).160. Corti S. et al.: J. Cell. Mol. Med. 8, 329 (2004).161. Howland D.S. et al.: Proc. Natl. Acad. Sci. USA

99, 1604 (2002).162. Dunlop J. et al.: J. Neurosci. 23, 1688 (2003).163. Chang-Hong R. et al.: Exp. Neurol. 194, 203 (2005).164. Donovan M.H. et al.: J. Comp. Neurol. 495, 70

(2006).165. Handley O.J. et al.: Clin. Sci. (Lond). 110, 73

(2006).166. Toulmond S. et al.: Br. J. Pharmacol. 141, 689

(2004).167. Bates G.P., Hockly E.: Curr. Opin. Neurol. 16,

465 (2003).168. Mandel S. et al.: CNS Drugs 17, 729 (2003).169. Fleming S.M. et al.: NeuroRx 2, 495 (2005).

368

369

CANCER CHEMOPREVENTIVE AGENTS-DRUGS FOR THE 21ST CENTURY?

WANDA BAER-DUBOWSKA

University of Medical Sciences in PoznaÒ, Department of Pharmaceutical BiochemistryGrunwaldzka 6, 60-780 PoznaÒ

Abstract: After a quarter of a century of rapid advances in cancer research, the focus of oncological drug devel-opment has shifted from cytotoxic chemotherapy to rationally designed agents that target specific moleculesassociated with malignant cells or their environment. Carcinogenic process is driven by mutation, but there aremany epigenetic variables which could be the targets of early intervention before invasion and metastasisoccur. Chemoprevention is the inhibition, retardation or reversal of carcinogenic processes by pharmacologi-cal or natural agents targeting these pathways in high-risk individuals. This approach was developed more than30 years ago and its credibility was enhanced by the positive results of clinical trials involving subjects withrisk of developing breast cancer and colon tumors. So far however, not many clinical trials provided satisfyingresults, not only because of the lack of efficacy or side toxic effects of chemopreventive agents, but also thelack of precise biomarkers monitoring their effects. In spite of all these obstacles, the field of cancer chemo-prevention is very active, not only because of its accelerating scientific base, but also because is vitally need-ed. New information from molecular studies has identified specific molecular targets for chemopreventiveagents. These include regulatory molecules such as Nrf2, epidermal growth factor receptor kinases, compo-nents of the Janus kinase-signal transducers and activators of transcription (JAK-STAT) pathway, nuclear fac-tor-κB, and cyclin D. The development of new drugs for the control of these targets that are both safe and effec-tive will be important for the future of cancer chemoprevention.

Keywords: chemoprevention, COX-2 inhibitors, NrF2, signaling pathways, transcription factors

The continuing magnitude of cancer problemclearly indicate that the more intensive approach to theprevention of this disease is the only way to eliminatethe cancer burden.

Currently the basic and clinical research are drivenby the elusive goal of cure the advanced disease. Giventhe genotypic and phenotypic heterogeneity of advancedmalignant lesions as they occur in individual patients,indication of specific molecular and cellular targets forthe putative cure is simply impossible. Furthermore, themisperception of cancer as a disease whose most funda-mental characteristic is excessive cell proliferation, hasled to an over-emphasis of testing and development ofcytotoxic drugs that kill cancer cells. Unfortunately,most cytotoxic drugs used in cancer chemotherapy arealso highly toxic to wide spectrum of normal tissues andiatrogenic failure of the damaged organs is a frequentcause of death from cancer (1). Moreover, three-quarterof the new more specific drugs approved by U.S. FDAfrom 1990 through the end of 2002, did not demonstratethat they increased patient survival, but significantlyincreased the cost of treatment (2).

As an alternative approach, we need to consider thatcancer is ultimately the end stage of a chronic diseaseprocess characterized by abnormal cell and tissue differ-entiation. Thus we need to focus more effort on the con-trol of carcinogenesis rather than attempting to cure end-stage disease. This approach, called chemoprevention,was proposed already in the 1960s by Wattenberg andformal definition was introduced by M. Sporn in 1976(3). Chemoprevention is pharmacological approach tointervention by use the natural or synthetic substances inorder to arrest, reverse or delays the process of carcino-genesis. It seems that this approach toward cancer hasnever before held such great scientific promise and at thesame time skepticism about its clinical practicality.

Despite the fact that definitive proof of principle ofchemoprevention has been established in very largeclinical trials including thousands of women at risk fordeveloping breast cancer (4,5), issues of safety and lia-bility continue to hamper the development of the field.

Despite the obstacles the field of cancer chemo-prevention is very active, not only because of its accel-erating scientific base, but also because such anapproach to cancer is vitally needed.

The advances in prevention of cardiovascular dis-eases dramatically reduced the death rate from myocar-dial infarction and stroke and provided important argu-ments that prevention of fatal diseases is possible. So itis time to catch up with the cardiologists.

In 2002 the chemoprevention problem was partlyreviewed in Acta Poloniae Pharmaceutica (6). The pres-ent article will emphasize some other aspects and morerecent progress that has occurred since then.

Conceptual basis of chemoprevention

Cancer begins through the multistage process ofcarcinogenesis resulting from the exposure to a wide vari-ety of carcinogenic insults. The major stages of carcino-genesis were deduced over the past 60 years, primarilyfrom animal model studies. These stages are termed initi-ation, promotion and progression (7). Tumor initiationbegins when DNA in a cell or population of cells is dam-aged by exposure to exogenous or endogenous carcino-gens. If this damage is not repaired, it can lead to geneticmutations. The responsiveness of the mutated cells totheir microenvironment can be altered and may give thema growth advantages over the normal cells. In the classictwo-stage carcinogenesis system in mouse skin, a lowdose of initiator (polycyclic aromatic hydrocarbon, 7,12-dimethylbenz[a]anthracene, DMBA) causes permanentDNA damage (the initiating event) but does not give rise

to tumors over the lifespan of the mouse unless a tumorpromoter, such as phorbol ester derivative (naturallyoccurring plant constituent) 12-O-tetradecanoyl-phorbol-13-acetate (TPA), is repeatedly applied (8). The tumorpromotion stage is characterized by selective clonalexpansion of the initiated cells, a result of TPA-inducedoxidative stress, and the altered expression of geneswhose products are associated with hyperproliferation,tissue remodeling and inflammation. During tumor pro-gression, preneoplastic cells develop into tumors througha process of clonal expansion that is facilitated by pro-gressive genomic instability and altered gene expression(Figure 1). Most animal models used in carcinogenesisresearch were developed before the identification of themajor cancer-related genes, the recognition of the impor-tance of host susceptibility to a carcinogenic insult, or therealization that mitogenesis and apoptosis together regu-late cell number. The temporal sequence implied in thisscheme is now thought to be overly simplistic and some-what misleading. Human carcinogenesis rather thanoccurring in three discrete stages in a predictable order isbest characterized as an accumulation of alterations ingenes regulating cellular homeostasis such as oncogenes,tumor suppressor genes, caretakers gene (apoptosis-regu-lating genes and DNA repair genes) and stability genes(BRCA1, BLM, ATM) as well as changes in epigenome(9,10). Nonetheless, these animal models have con-tributed significantly to our understanding of carcinogen-esis and the ways to interfere with that process.Moreover, these models allow convenient categorizationof chemopreventive agents into those that can block initi-ation (blocking agents) and those that suppress promotionand progression (suppressing agents) (7). Early solid can-cers are generally detected as intraepithelial neoplasia orcarcinoma in situ, which correspond to the promotion andprogression stages. Thus anti-promotion and anti-pro-gression agents are therefore of particular clinical inter-est. Ultimately, such agents prevent the growth and sur-vival of cells already committed to become malignant

(secondary and tertiary chemoprevention). Chemo-prevention can be applied at three levels. While second-ary and tertiary chemoprevention is addressed to patientswith preneoplastic lesions and cancer survivors, primarychemoprevention is thought to be applied to generalhealthy population and high-risk individuals and mayrelay to great extent on dietary intervention (11).

Proven clinical efficacy of chemoprovention and the

lesson learned from its practical usage

The credibility of chemoprevention as a serious andpractical approach to the control of cancer has beengreatly enhanced by the results of major randomizedclinical trials in the field of breast cancer. Three differentagents, namely tamoxifen, raloxifene and 4-hydroxy-phenylretiamide (1) have been shown effective agentsfor prevention of breast cancer in women of varyingdegrees of risk. Tamoxifen and raloxifene receptors areexamples of the selective estrogen receptor modulators(SERMs) which bind to both estrogen receptors ER-αand ER-β. SERMs can act selectively in different tissuesand organs either as estrogen antagonists (to suppress theundesirable cancer-promoting effects of estrogen inbreast) or as estrogen agonists (to enhance the desirablegrowth-promoting effects of estrogen in bone).

Just released results from 20,000-women Study ofTamoxifen and Raloxifene (STAR) showed the advan-tages of raloxifene over tamoxifen to prevent breast can-cer in high-risk postmenopausal women, providingequal cancer-preventing benefits with fewer serious sideeffects (12).

Fenretinide is a synthetic retinoid which was shownmany years ago to be a differentiating agent at manyepithelial target sites in experimental animals as well as inhumans (1). On the other hand, the failure of β-carotenetrials provided the evidence that epidemiological dataalone are not sufficient for the selection of a new agent forthe major clinical chemoprevention trial. The appropriateuse of a chemopreventive agent ultimately depends on the

370

Figure 1. Chemoprevention of multistage carcinogenesis.

understanding of its mechanism of action at all levels,namely at the molecular, cellular, tissue and organ levels,as well as in the animal as a whole. The trend in the fieldof chemoprevention has therefore been to develop newagents based on their mechanism of action.

The examples of the chemopreventive agentsdeveloped based on this approach are selective inhi-bitors of inducible cyclooxygenase regulating prosta-glandin synthesis and involved in the inflammation.Chronic inflammation and oxidative stress contribute toexperimental and human carcinogenesis (13). Two sep-arate gene products, COX-1 and COX-2, have similarcyclooxygenase and peroxidase activities, although theyare differently regulated. COX-1 is a constitutive iso-form present in most tissues and mediates the synthesisof PGs required for normal physiological functions.COX-2 is not detectable in most normal tissues, but it isinduced by cytokines, growth factors, oncogenes, andtumor promoters and is involved in several experimen-tal and human cancers (14). Although the concept of therole of persistent inflammation in cancer developmentwas proposed nearly 150 years ago by Virchov based onhis observations that hematopoetic cells frequently infil-trated neoplastic tissue, the immediate relevance of thisnotion to modern molecular genetic was demonstratedby the finding that overexpression of the gene for COX-2 is an early and important event in colon carcinogene-sis (15) and that knockout of the gene for COX-2 couldsuppress colon carcinogenesis in mice that were geneti-cally predisposed to develop this condition (16).

Celecoxib, the first US FDA-approved selectiveCOX-2 inhibitor initially developed for the treatment ofadult rheumatoid arthritis and osteoarthritis, was report-ed to reduce the polyp burden in patients with familialadenomatous polyposis (17).

The demonstration of the overexpression of COX-2in many other forms of epithelial cancer (18) offeredpromise that selective COX-2 inhibitors might be used tosuppress carcinogenesis in organs other than the colon.

Approximately six years after the COX-2inhibitors were approved for use in the United States,the results of three randomized, placebo-controlled tri-als provided new evidence about cardiovascular risk ofrofecoxib, celecoxib and valdecoxib (19,20) The publicannouncement of the APPROVE (Vioxx) results, whichcoincided with Merckís withdrawal of rofecoxib fromthe market in September 2004, prompted researchers toreview the cardiovascular safety results on a similar trialwith COX-2 inhibitors. The results have raised the ques-tion: how did such problem arise, and how can they beprevented in the future?

COX-2 inhibitors not only lack the antiplateleteffects of aspirin; by inhibiting the production of prosta-cyclin, they also disable one of the primary defenses ofthe endothelium against platelet aggregation, hyperten-sion and atherosclerosis (21). COX-2 inhibitors alsopromote an imbalance in favor of vasoconstriction.Thus the combination of reduced prostacyclin synthesisand uninhibited thromboxane synthesis in patients onlong-term coxibs has been proposed as the likely expla-nation for coxib-related cardiovascular toxicity (22). Itis also possible that it is the prolonged plasma concen-trations of the coxibs that are responsible for their dan-gerous cardiovascular consequences.

To avoid such problem in the future, investigatorsmust understand and respect the pharmacology underly-ing drug discovery. Specifically, I/ the pharmacokinet-ics and metabolism of the drug must be studied in rela-tion to different dosing schedules; II/ we do not use can-cer chemotherapeutic drugs as a single agents becausewe know that high-dose single agents induce selectivesurvival, clonal outgrowth and ultimately recurrence. Itis plausible that there is similar effect with high doses ofa single chemopreventive agent, such as β-carotene(22), an effect that seems to be much less of an issuewith the low-dose, multiple agent dietary interventions.At some point, we may conclude that low-to moderatedose multiple agent regimens in the prevention arenamay serve us best; III/ placebo controlled, phase IIchemoprevention trials must extend well beyond 3 to 6months to at least 1 to 2 years if we are to obtain essen-tial long-term safety data. Finally we have to rememberthe critical role that risk-benefit assessments play inagent development for cancer prevention.

New targets and chemopreventive agents

One of the rational and effective at least in experi-mental models strategies for chemoprevention is theblockade of DNA damage caused by carcinogenic insult.However clinical trials that have used exogenous antiox-idants or free-radical scavengers (such as ascorbic acid orβ-carotene) as chemopreventive agents to suppress muta-tion have been markedly unsuccessful in most cases.Therefore, a new approach to control electrophilic xeno-biotics metabolites, ROS and RNO has been developedbased on the intrinsic mechanisms used by the body todeactivate potentially carcinogenic molecules. Thisapproach focuses on nuclear factor erythroid 2-relatedfactor 2 (Nrf2), which is a leucine zipper-type transcrip-tion factor and a key enhancer of a number of genes thatare cytoprotective because of the ability of their respec-tive protein products to deactivate a large number ofpotentially harmful molecules. These proteins, known asphase 2 enzymes deactivate reactive electrophilicmetabolites, directly destroy ROS and stabilize the oxi-dation and reduction potential of the cell. They includeglutathione transferases, quinone reductase, epoxidehydrolase, thioredoxin, catalase, superoxide dismutase,and heme oxygenase. The new attempts to use upregula-tion of Nrf2 activity as a strategy for chemopreventionrely on coordinated upregulation of an entire battery ofcytoprotective macromolecules that are intrinsic to thecell. This new strategy provides marked amplification ofthe action of chemopreventive agents, since the smallmolecules that induce Nrf2 activity are not themselvesconsumed in the deactivation of electrophiles, ROS andRNO. These inducers rather amplify physiological pro-tective mechanisms, and this action occurs within thecell, at sites where it is needed. Nrf2 is known to regulateits various responsive cytoprotective genes through acommon DNA regulatory element, anti-oxidant-responseelement (ARE) Nrf2 is sequestered in cytoplasm by theinhibitory protein, Keap1, until inducers react with cys-teine thiol residues on Keap 1, which causes a conforma-tional change in the Keap1- Nrf2 complex and the releaseof Nrf2, allowing Nrf2 to activate responsive genes.Many naturally occurring and synthetic compounds suchas resveratrol, curcumin, sulforaphane or oltipraz activate

371

the Nrf2-keap1-ARE pathways. Nrf2 can also be activat-ed by phosphorylation, and many signal transduction cas-cades as those involved in mitogen-activated proteinkinases, protein kinase C or phosphatydiloinositol 3-kinase, can also activate Nrf2 (23). Stimulation of Nrf2-ARE signaling pathway offers unique opportunities forfurther development of new agents for chemoprevention.The examples are new synthetic triterpenoids agents,which can stimulate the synthesis of phase 2 enzymes atthe dose levels at below 1nM (24).

While the interaction with Nrf2 may protectagainst DNA damage and initiation of carcinogenesisinterfering or blocking the signal transduction pathwayis useful for the anti-promtion/anti-progression chemo-prevention strategies.

The development of chemotherapeutic agents thattarget specific signal transduction without significantcytotoxicity opened new field also for chemoprevention.One of such approach is the inhibition of epidermalgrowth factor receptor (EGFR) kinase. Erlotinib, lowmolecular EGFR kinase inhibitor was successfully usedfor clinical treatment of advance breast and lung cancersand now is evaluated for the prevention of carcinogen-induced lung tumors in mice (25). The major polyphe-nolic constituent of green tea, (-)-epigallocatechin-3-gal-late (EGCG) and very promising chemopreventive agenthas been shown to modulate multiple signal transductionpathways including EGFR and MAPK cascades both invitro and in vivo (26). Another important signal trans-duction pathway for chemoprevention studies is P13K-protein kinase B (Akt) pathways. Activation Akt hasbeen shown in human bronchial preinvasive lesions (27).

As a result of research performed over the pastdecade it is clear that there are many fewer pathwaysthan genes (9, 28). Thus, cross-talk of members fromone that of another undoubtedly accounts for much ofthe redundancy, as well as diversity of responses, thatcan result from the surprisingly small finite number oftotal signal pathways. Inhibition of one pathway e.g.EGFR as in case of EGCG results in interference withmultiple signaling cascades including the P13K/Aktpathway, thus blocking the activation of specific genes,DNA synthesis and proliferation (26).

Transcription factors participate in the final stagesof all transduction and stress pathways by causing theup- and down-regulation of specific genes. Thus tran-scription factors and their associated signaling mole-cules also represent good targets for the development ofchemopreventive agents. Transcription factor nuclearfactor-κB (NF-κB) and activator-protein-1 (AP-1) weresuggested as proximate links between inflammation andcarcinogenesis (29, 30). The expression of COX-2 andNOS-2 is controlled by these transcription factors andselective inhibition of these enzymes is often mediatedby either NF-κB or AP-1.

Several synthetic and natural compounds act inthis way. Most of the COX-2 inhibitors including cele-coxib and spices ingredient curcumin act through inhi-bition of AP-1 or NF-κB activation.

Significant inhibition of both transcription factorsand COX-2 was also observed as results of rodentsí treat-ment with resveratrol and its analogue pterostilbene (31).

Urosolic acid, naturally occurring triterpenoidinhibits NF-κB activation, which correlates with sup-

pression of NF-κB-dependent cyclin D1, COX-2 andmatrix metalloprotein 9 expressions (32). The lattergene promoter sequence similarly to cox-2 contains abinding site for NF-κB. Cyclin D1 and its relatives,cyclin D2 and D3 as key regulators of cell cycle and cel-lular differentiation are useful targets for chemopreven-tion. Aberrant expression of cyclin D1 is a frequentoccurrence in bronchial preneoplasia and lung cancerand it was shown that retinoic acid can regulate theexpression of all three cyclins in vitro, suggesting apotential mechanism for chemoprevention by retinoids.

The signals transducers and activators of transcrip-tion (STAT) family of transcription factors and their reg-ulatory Janus kinases (JAKs) also play a critical role in theactivities of both immune cells and tumor cells. Both setsof transcription factors represents of attractive targets fornew research on its application in chemoprevention (25).

Finally, in the post genomic era of cancer biology,it is also becoming increasingly evident that epigeneticcontrols of gene expression play an important role in thedevelopment of cancer. Alterations in gene expressionwithout changes in DNA coding sequences that are her-itable through cell division occur throughout all stagesof carcinogenesis and are involved in silencing tumorsuppressor genes. DNA methylation and modificationof histone protein particularly acetylation might be usedfor chemoprevention strategy (10).

The unique epigenetic fingerprint observed, e.g. inpatients with preinvasive esophageal lesions that pro-gressed to more advanced lesions (33) underscore thepotential use of epigenetic markers in risk assessmentand early detection.

Further considerations and problems to solve

Basically, chemoprevention is addressed to peo-ple/patients without any disease symptoms thus preciseintermediate biomarkers of cancer are essential for itsapplication. Historically, reduced cancer incidence ormortality has been required to show chemopreventivebenefits. These end points make chemoprevention trialslong, large, costly and hence impractical and risky fordrug developers. Obviously even for clinical trials mark-ers of carcinogenesis rather than disease are more rele-vant. Chemoprevention demands the change in the con-cept of ìhealthî in order to consider future events not justimmediate presence. Thus the development of more pre-cise biomarkers of the earliest stages of cancer develop-ment is as important as new chemopreventive agents/drug.

Currently four categories of biomarkers are in use:I/ genetic markers (e.g. micro satellite, p53 mutation, K-ras, FHIT mutation) II/ differentiation biomarkers(squamous markers, mucine gene expression) III/ pro-liferation markers (retinoic acid receptor, proliferatingcell nuclear antigen) IV/ efficacy markers (pharmacody-namics, indicators of oxidative state, COX-2/expres-sion/activity). Recent strategy of chemoprevention effi-cacy evaluation focuses on prevention or regression ofsignificant precancerous lesions, intraepithelial neopla-sia (IEN). Occurring in most epithelial tissue as moder-ate to severe dysplasia, IEN shares phenotypic andgenotypic similarities with invasive disease and is onthe causal pathway leading from normal tissue to can-cer. The IEN was also recommended by AACR TaskForce as an important target for accelerated new agent

372

development (34). Development of microarray tech-nique allowing broad transcriptional profiling adds anovel dimension to identification of molecular markers.

While the development of new synthetic chemo-preventive drugs towards the new molecular targets isexpected, naturally occurring substances should be pro-foundly evaluated. Diet-derived polyphenols, such asresveratrol are particularly attractive, since the long-proven use of their dietary sources suggests low poten-tial for unwanted side effect. However in most preclini-cal studies these compounds are administered in highdoses. Thus, despite the immense number of studies per-formed on resveratrol, the additional efficacy, bioavail-ability and pharmacokinetic data, particularly onnanomolar concentration are still lacking (35).

Overall cancer chemoprevention research baseshould also use more advanced technology like nan-otechnology (36).

The inability to guarantee the long-term safety ofcurrent chemopreventive regiments is perhaps the biggestlimitation to their more widespread application. Althoughtoxic side effects are almost inevitable in the setting ofchemotherapy for advanced disease, they are not accept-able for chemoprevention. One approach to consider asthe conclusion from COX-2 inhibitors clinical trialswould be to use intermittent (ìchemoprevention win-dowî), rather than constant, chronic dosing of chemopre-ventive agents. It should be also emphasize that it isunlikely that a totally safe drug regimen for chemopre-vention will be ever found. As in case of other drugs eval-uation of benefits vs. potential risk should decide aboutthe implementation of chemopreventive agents.

Development of chemopreventive drugs that targetmechanisms such as the inflammatory process or oxida-tive stress offers the possibility of immense clinical ben-efits not only for cancer prevention, but also otherdegenerative diseases where these processes play animportant role in their pathogenesis. The latter shouldbe good argument for pharmaceutical industry, which sofar is reluctant to embrace cancer chemoprevention inthe same way as this industry has successfully embracedthe chemoprevention of cardiovascular disease.

The goal to avert invasive and metastatic malig-nancy by using chemoprevention is rationally, medical-ly and economically justified. Although it is not easy thewill to achieve a cancer treatment paradigm shift mustbe found.

REFERENCES

1. Sporn M.B., Suh N.: Carcinogenesis 21, 525 (2000).2. Johnson J.R., Wiliams G., Pazdur R.: J. Clin. Onco-

logy 21, 1404 (2003).3. Sporn M.B.: Cancer Res. 36, 2699 (1976).4. Fisher B., Costantino J.P., Wickerham D.L., et al.: J.

Natl. Cancer Inst. 98, 1371 (1998).5. Vogel V.G., Costantino J.P., Wickerham D.L., et al.:

JAMA 295, E1 (2006).6. Krzystyniak K.L.: Acta Pol. Pharm 59, 473 (2002).7. Baer-Dubowska W., Ignatowicz E.: in Carcinogenic

and anticarcinogenic food components, Baer-Dubo-wska W, Bartoszek A., Malejka-Giganti D. Eds., p.177, CRC Taylor & Francis Group, Boca Raton Fl2006.

8. DiGiovanni J.: in Carcinogenesis, Waalkes M.P,Ward J.M. Eds., p.265, Raven Press, New York 1994.

9. Vogelstein B., Kinzler K.W.: Nature Medicine 10,789 (2004).

10. Paluszczak J., Baer-Dubowska W.: Post. Biochem.51, 244 (2005).

11. Gescher A.J., Sharma R.A., Steward W.P.: LancetOncol. 2, 371 (2001).

12. Vastag B.: J. Natl. Cancer Inst. 98, 733 (2006).13. Hussain S.P., Hofseth L.J., Harris C.C.: Nature

Rev. 3, 276 (2003).14. Prescott S.M., Fitzpatrick F.A.: Bioch. Biophys.

Acta 1470, M69.15. Sano H., Kawahito Y., Wilder R.L, et al.: Cancer

Res 55, 3785 (1995).16. Oshima M., Dinchuk J.E., Kargman S.L., et al.:

Cell 87, 803 (1996).17. Steinbach G., Lynch P.M., Philips R.K., et al.: N.

Eng. J. Med. 342, 1946 (2000).18. Wilgus T.A., Koki A.T., Zweibel B.S., Kusewitt

D.F., Rubal P.A., Oberyszyn T.A.: Mol. Carcino-gen. 38, 49 (2003).

19. Solomon S.D., McMurray J.J., Pfeffer M.A., et al.:N. Eng. J. Med. 352, 1071 (2005).

20. Bresalier R.S., Sandler R.S., Quan H., et al.: N.Eng. J. Med. 352, 1092 (2005).

21. McAdam B.F., Catella-LawsomF., Mardini I.A.,Kapoor S., Lawson J.A., FitzGerald G.A.: Proc.Natl. Acad Sci USA 96, 272 (1999).

22. Alberts D.S., Potter J.D., Martinez M.E., HessL.M., Stopeck A., Lance P.: What happened to thecoxubs on the way to the cardiologist? CancerEpidemiol. Biomarkers Prev. 14, 555 (2005).

23. Ngyen T., Sherratt P.J., Pickett C.B.: Annu. Rev.Pharmacol. Toxicol. 43, 233 (2003).

24. Dinkova-Kostova A.T., Liby K.T., Stephenson K.K.,et al.: Proc. Natl. Acad. Sci. USA 102, 4584 (2005).

25. Sporn M.B., Liby K.T.: Nature Clinical PracticeOncology 2, 518 (2005)

26. Khan N., Afaq F., Saleem M., Ahmad N., MukhtarH.: Cancer Res. 66, 2500 (2006).

27. Tsao A.S., McDonnell T., Lam S., et al.: CancerEpidemiol. Biomarkers Prev. 12, 660 (2003).

28. Nebert D.W.: Toxicology 181-182, 131 (2002).29. Pikarsky E., Porat R.M., Stein I., et al.: Nature 431,

461 (2004).30. Shaulian E., Karin M.: Oncogene 20, 2390 (2001). 31. Cichocki M., Waszak W., Baer-Dubowska W.:

Acta Biochim. Pol. 52 S1 P9.2 (2005). 32. Shishoda S., Majumdar S., Banerjee S., Aggarwal

B.B.: Cancer Res. 63, 4375 (2003).33. Eads C.A., Lord R.V., Wickramasinghe K., et al.:

Cancer Res. 61, 3225 (2001).34. OíShaughnessy J.A., Kellof G.J., Gordon G.B,

Dannenberg A.J., Hong W.K., Fabian C.J., SigmanC.C., Bertagnolli M.M., Stratton S.P., Lam S.,Nelson W.G., Meyskens F.L., Alberts D.S., FollenM., et al.: Clinical Cancer Res. 8, 314 (2002).

35. Gescher A.J., Steward W.P. Cancer Epidemiol.Biomarkers Prev. 12, 953 (2003).

36. News from the NCI: Centers of cancer nanotech-nology excellence think small to make big changesin cancer prevention, diagnosis and treatment.Cancer Prevention News Letter Issue 7 (2006).

373