Stroke is one of the most common causes of neurological diseases related death in the worldwide [25]. Extensive ischemic injury is a neurological disorder caused by multiple factors including hypoxia and severe damage to the cerebral parenchyma that result in the formation of a cystic cavity and consequential loss of neural cells and their connections. This leads to the death of oligodendrocytes, astrocytes, and endothelial cells [26]. In the past, fetal brain tissue transplants have been shown to promote limited recovery in animal models of stroke, but ethical considerations and a scant supply of human fetal tissue have limited this approach [27]. Successful isolation and transplantation of adult NSCs has demonstrated the feasibility of using autologous NSCs transplantation as a regenerative strategy after stroke. Most animal models of ischemic stroke involve occlusion of the arterial blood supply to the brain using surgical techniques [28]. Jeong et al. reported that human NSCs delivered intravenously are beneficial in an animal model of stroke [29]. Intravenously transplanted NSCs can enter the brain of rats with intracerebral hemorrhage (ICH), survive, migrate, and improve functional recovery. Transplanted NSCs selectively migrate to the perihematomal areas and differentiated into neurons (approximately 10%) and astrocytes (approximately 75%). Kelly et al. also studied the effects of human neurospheres derived from CNS stem cells transplanted into the ischemic cortex of rats 7 day after distal middle cerebral artery occlusion [30]. The neurospheres continued to survive in both naïve and ischemic brains 4 weeks after transplantation. Additionally, the microenvironment was found to influence neurosphere migration and fate.

NSCs are a relatively quiescent cell population, and stem cell proliferation in the SVZ is tightly controlled under physiological conditions. Stroke upregulates the expression of mitogens, including epidermal growth factor (EGF) and bFGF, which may contribute to increased NSC numbers [31]. Zhang et al. found that after stroke, neuroblasts derived from NSCs proliferate and migrate into the ischemic striatum [32]. This might have important implications for targeting endogenous NSCs and progenitor cells in the areas of damaged brain tissues undergoing repair. Moreover, increased neurogenesis and migration of NSCs to the site of injury indicates that stromal derived factor-1 (SDF-1) and angiopoietin-1 contribute to the targeting of newly formed neural progenitors to the injury sites [33].

Genetically modified NSCs have also been shown to improve function in a mouse model of stroke. This might be accomplished by overexpressing brain-derived neurotrophic factor (BDNF), nerve growth factor (NGF), or Akt which is known as a general mediator of cell survival [34]. Akt, a serine/threonine kinase, exerts antiapoptotic effects against a variety of pro-apoptotic factors including withdrawal of extracellular signaling molecules, oxidative and osmotic stress, and ischemic shock [35]. This protein prevents cerebellar granule cells from undergoing apoptotic cell death, and promotes the survival of hippocampal neurons under hypoxic conditions [36]. In mouse ICH models, human NSCs genetically modified to express Akt1 improve motor performance as determined by rotarod and limb placement tests, and increase the survival of grafted NSCs or differentiation into neurons and astrocytes [37]. Taken together, the results of several previous studies suggest that NSC transplantation is a potential regenerative therapy for treating stroke. Currently, the development of stem cell therapy for stroke patients is in its infancy.

PD is common neurodegenerative disease characterized by an extensive loss of dopamine (DA) neurons in the substantia nigra pars compacta (SNpc) and DA neuronal terminals in the striatum [38]. Another major pathological feature is the presence of Lewy bodies (LBs) which are intraneronal proteinacious cytoplasmic inclusions in the surviving neurons [39]. Clinically, patients with PD exhibit rigidity, bradykinesia, resting tremors, and postural instability [40]. One of the earliest biochemical changes seen in these individuals is decreased levels of reduced glutathione (GSH), a major component of cellular antioxidant defenses [40]. Reduced GSH levels are also associated with incidental LB diseases thought to be asymptomatic precursors to PD [41].

Animals models of PD have been generated with toxins including neurotoxins used to induce dopaminergic neurodegeneration, 6-hydroxydopamine (6-OHDA), 1-methyl-4-phenyl-1,2,3,6,-tetrahydropyridine (MPTP), paraquat, and rotenone [42]. 6-OHDA is relatively selective for monoaminergic neurons, resulting from preferential uptake by DA and noradrenergic transporters. Since this compound cannot permeate the blood-brain barrier (BBB), it must be administered by local stereotaxic injection into the SNpc, median forebrain bundle (MFB) which forms ascending dopaminergic and serotonergic projections into the forebrain, or striatum to target the nigrostriatal dopaminergic pathway [43]. Yasuhara et al. transplanted green fluorescent proteinlabeled NSCs into lesions in the striatum of a 6-OHDA-treated rat model of PD. The grafted NSCs survived in the brain lesions which are positive for the neuronal marker, mitogen-activated protein 2 (MAP2), and synaptophysin-positive terminals. Furthermore, endogenous neurogenesis was observed in the rat SVZ. In another rat PD model generated with 6-OHDA, Kim et al. transplanted NSCs expressing genes encoding tyrosine hydroxylase (TH) and GTP cyclohydrolase 1 (GTPCH1) in order to create dopamine-producing NSCs [44]. Marked improvement was observed in the PD rats that received NSCs expressed TH and GTPCH.

In humans and monkeys, treatment with MPTP produces an irreversible and severe Parkinsonian syndrome characterized by features of PD including tremor, rigidity, slowness of movement, postural instability, and freezing [42]. The susceptibility to MPTP increases with age in both monkeys and mice [45]. In a previous study, a small number of human NSC progeny were found to differentiate into TH and/or dopamine transporter (DAT)-positive cells, suggesting that the microenvironment within and around the SNpc lesions in adult monkeys permits the development of a DA phenotype in responsive progenitor cells [46]. These results indicated that naïve or genetically modified NSCs have a great potential for use in cell replacement therapy for patients suffering from PD.

Brain tumors are the leading cause of cancer mortality in children, and remain difficult to cure despite advances in surgical techniques and adjuvant therapy [47]. Malignant brain tumors, including glioblastoma multiforme, remain virtually untreatable and lethal. Currently available treatments for brain tumors include radical surgical resection followed by radiation or chemotherapy, and have substantially improved the survival rate in patients suffering from these lesions. However, brain tumors remain incurable in large proportion of patients. Therefore, there is an urgent need for effective and minimally toxic therapies for treating these tumors.

Most current research on human brain tumors is focused on the molecular and cellular analysis of bulk tumor masses. A widely used molecular approach is suicide gene-based therapy that relies on the conversion of non-toxic prodrugs into toxic anticancer drugs via the expression of exogenous enzymes. Additionally, genetic immunotherapy involving the transfer of genes expressing immune-stimulating cytokines has also been developed [48]. Genetically modified NSCs may also be used to treat various human brain tumors. For example, Aboody et al. suggested that the prodrug 5-fluorocytosine (5-FC) along with NSCs able to express cytosine deaminase (CD), a bioactive factor, could dramatically reduce tumor burden in vivo [49]. Transplanted NSCs expressing CD can convert 5-FC into 5-fluorouracil (5-FU) and induce tumor regression. NSCs expressing the CD gene can also generate an agent that kills tumor cells and undergo self-elimination should the NSCs themselves become mitotic [49]. In another study, a significant decrease of glioma tumoral mass (approximately 50%) was observed in rat neural progenitor cells expressing CD gene treated with 5-FC [50]. The 5-FC itself had no effect on the tumor in the absence of cells expressing CD.

In athymic nude mice, the ability of NSCs secreting the pro-apoptotic protein TRAIL to treat human gliomas was investigated [51]. High levels of TRAIL secretion was observed within the main tumor mass as well as the tumor pockets and satellites, indicating that NSCs expressing TRAIL migrated into the tumor outgrowths. TRAIL-induced cell death led to a highly significant decrease in tumor volume compared to transplantation with NSCs expressing LacZ or the saline-inoculated control.

In an murine intracranial medulloblastoma model, human NSCs expressing the CD gene were injected into the contralateral hemisphere of the brain [52]. The mice were then treated systemically with 5-FC. Histologic analyses showed that the NSCs migrated to the tumor bed and lesion boundary, resulting in a 76% reduction of tumor volume [52]. This finding provides a rationale for further evaluating NSC-based cellular delivery systems for treating human brain tumors, including gliomas or medulloblastomas.

The efficacy of modalities using NSCs based on gene direct enzyme/prodrug therapy (GEPT) has been examined in various animal models of human cancer [53,54]. Hepatocellular carcinoma (HCC) is the sixth most common cancer and third most common cause of cancer-related death in the world [55]. For treating HCC in a xenograft SCID mouse model, Yi et al. compared NSCs with HB1.F3.CD and HB1.F3.CD.IFN-β cells expressing the CD and/or interferon-beta (IFN-β) gene [56]. Results of this experiment showed that NSCs expressing the therapeutic genes have the potent advantage of selective migration toward HCC cells in vivo. After 8 weeks, mice in the negative control group (without stem cells or prodrug) bearing tumors reached the endpoint of ethical death. However, tumor growth in animals treated with the HB1.F3.CD or HB1.F3.CD.IFN-β cells was inhibited approximately 40-50% compared to the control group [56]. Additionally, fluorescence prestained NSCs were detected in HCC tumor mass of animal models. These data indicate that NSCs possess tumor-specific tropism in cases of HCC as well as brain lesions. Yi et al. also studied the activities of NSCs expressing therapeutic genes in mice bearing tumors arising from MDA-MB-231 human breast carcinoma, colorectal cancer, or Ishikawa endometrial cancer cells [57-59]. Results of these experiments demonstrated that tumor volume is regulated in animals treated with the NSCs and prodrug. Aggressive behavior of tumor cell masses is also substantially reduced by injection with stem cells and prodrug treatment in mice [60].

Several researchers have confirmed the therapeutic effect of NSCs in animal models of metastatic tumors. The brain receives 15-20% of the body's blood flow, thereby increasing the chance of circulating tumor cells reaching the brain. The reported incidence of brain metastases ranges from 12-35% [61]. Many patients with brain metastasis harbor two or more metastases. Treatment of these individuals is hampered by the fact that intact blood-brain barrier (BBB) is largely impermeable to most chemotherapeutic drugs.

The most common origins of brain metastasis include primary cancers of the lung, breast, and skin [62]. Primary lung cancer has the highest incidence of brain metastasis with approximately 40% of all lung cancer patients developing brain metastasis. Breast cancer is the second most common cancer associated with brain metastasis [63]. Animal models of metastatic tumors have been produced using two methods: direct tumor cell implantation into the brain or blood-born brain metastasis [64]. In an animal model of metastatic breast cancer, NSCs expressing the CD or carboxyl esterase (CE) suicide genes were used to treat tumors [65]. CE can convert the prodrug CPT-11 to the toxic compound SN-38 [66]. Yi et al. also investigated the effect of NSCs expressing the yeast CD (yCD) gene in an animal model of metastatic lung cancer [67]. Expression of yCD appears to be far more effective for converting 5-FC into 5-FU compared to bacterial CD, both in vitro and in vivo [68]. NSCs expressing yCD in the presence of 5-FC reduced the density and aggressive behavior of lung cancer cells compared to the negative control or NSCs without 5-FC. The cytotoxic drug 5-FU is effective for treating brain metastases but cannot penetrate the BBB. However, 5-FC readily crosses the BBB or is transported into the brain parenchyma [69]. Therefore, this prodrug may have great potential for treating brain metastases.

The adult brain has a capacity for self-repair and replacing lost neurons in several CNS regions such as the olfactory bulb, hippocampus, subependymal zone, and cortex [70]. NSCs within these neurogenic regions can proliferate and differentiate into neurons or glia, thus providing a reservoir for the replacement of cells lost during normal cell turnover and after brain injury [71]. There have been a number of recent reports describing the effects of NSC transplantation for achieving functional recovery from CNS damage. Evidence from these investigations suggests that NSCs may be a suitable component for treating neurological diseases such as stroke, PD, brain tumors, primary tumors, and metastatic lesions (Figure 1). In summary, naïve or genetically modified NSCs may represent an effective new modality for treating various human brain diseases without inducing injurious effects commonly associated with more conventional therapies.