**6. GDNF treatment in Parkinson's disease**

The degeneration of DA-ergic neurons in the substantia nigra of the midbrain is the main pathological characteristic of PD. Based on this characteristic, the most predominant therapeutic strategies are to administer DOPA agents to supply dopamine to the brain, or to employ dopamine agonists to activate the DA-receptor directly. However, the efficacies of these drugs tend to wane because the residual DA-ergic neurons decrease during the progression of PD. In 1993, GDNF, as a growth factor that promotes the survival and growth of DA-ergic neurons in the midbrain, was extracted and qualified (Lin *et al.*, 1993). Early studies demonstrated that GDNF elevates the affinity of neurons to DA, increases the uptake of DA and the expression of tyrosine hydroxylase and promotes the growth of neurapophysis in cultured midbrain DA-ergic neurons. In addition, GDNF was demonstrated to act as neurotrophic factor in spinal motor neurons (Henderson *et al.*, 1994) and noradrenergic neurons (Arenas *et al.*, 1995) in the CNS. Further, GDNF was found to maintain a low level in PD patients and in elder rodents (Yurek & Fletcher-Turner, 2000; Jenner and Olanow, 1998). Thus, this neurotrophic factor is viewed as a potential treatment for neurodegenerative disease.

In cultured embryonic DA-ergic neurons, the paucity of GDNF initiates a novel non-classical death pathway in which the mitochondria does not release cytochrome to the cytoplasm, Bax is not activated and over-expressed Bcl-xL does not inhibit death. However, death is inhibited by the caspase inhibitor BAF, and mutants of caspase-9, caspase-3 and caspase-7, indicating that caspases are essential for the process. Another study demonstrated that in the MLK-c-Jun pathway, caspase-2 and caspase-7 are essential for GDNF-paucity-induced death. In neuroblastoma cells, the GDNF response is similar.

Based on this principle, researchers have adopted different models to supply GDNF in PD animal models. The results demonstrated that GDNF inhibits the variation of Bcl-2 and Bax, up-regulates Bcl-XL, inhibits the combination of CYT-C and Apaf-1 as well as the activity of caspase-3 (Ghribi *et al.*, 2001) reverses the death of DA-ergic neurons induced by

neurons (Chao *et al.*, 2003). Some reports have demonstrated that the membrane lipid rafts provide a perfect environment for the activation of integrins, mediated by growth factors, in the system of survival for oligodendrocytes. The shift of activated integrin to membrane lipid rafts is an important cue, which provides rationale for the actions of functional T cells

N-cadherin, a calcium dependent cell adhesion molecule similar to NCAM140 and integrin β1, is a trans-membrane protein that mediates the axonal growth of astrocytes and Schwann cells. As mentioned above, the intracellular domain of N-cadherin is similar to the extradomain of RET. Additionally, N-cadherin can activate the PI3K/Akt and Ras/Raf/MAPK signal pathways (*et al.*,) (73), which provides the rationale for further delving into GDNF and

It is closely related to signal transduction in the membrane and provides a platform for anchoring proteins. A: GFRα1 and integrin β1 integrate into the membrane lipid raft of DAergic neurons, but RET, N-cadherin, and NCAM-140 are outside of the raft. B: A GDNF dimer combines with GFRα1 and forms a GDNF-GFRα1 complex, which binds to RET and gathers in the membrane lipid raft. The combination of the GDNF dimer and NCAM dimer requires another substance. N-cadherin and integrin do not gather in the membrane lipid

The degeneration of DA-ergic neurons in the substantia nigra of the midbrain is the main pathological characteristic of PD. Based on this characteristic, the most predominant therapeutic strategies are to administer DOPA agents to supply dopamine to the brain, or to employ dopamine agonists to activate the DA-receptor directly. However, the efficacies of these drugs tend to wane because the residual DA-ergic neurons decrease during the progression of PD. In 1993, GDNF, as a growth factor that promotes the survival and growth of DA-ergic neurons in the midbrain, was extracted and qualified (Lin *et al.*, 1993). Early studies demonstrated that GDNF elevates the affinity of neurons to DA, increases the uptake of DA and the expression of tyrosine hydroxylase and promotes the growth of neurapophysis in cultured midbrain DA-ergic neurons. In addition, GDNF was demonstrated to act as neurotrophic factor in spinal motor neurons (Henderson *et al.*, 1994) and noradrenergic neurons (Arenas *et al.*, 1995) in the CNS. Further, GDNF was found to maintain a low level in PD patients and in elder rodents (Yurek & Fletcher-Turner, 2000; Jenner and Olanow, 1998). Thus, this neurotrophic factor is viewed as a potential treatment

In cultured embryonic DA-ergic neurons, the paucity of GDNF initiates a novel non-classical death pathway in which the mitochondria does not release cytochrome to the cytoplasm, Bax is not activated and over-expressed Bcl-xL does not inhibit death. However, death is inhibited by the caspase inhibitor BAF, and mutants of caspase-9, caspase-3 and caspase-7, indicating that caspases are essential for the process. Another study demonstrated that in the MLK-c-Jun pathway, caspase-2 and caspase-7 are essential for GDNF-paucity-induced

Based on this principle, researchers have adopted different models to supply GDNF in PD animal models. The results demonstrated that GDNF inhibits the variation of Bcl-2 and Bax, up-regulates Bcl-XL, inhibits the combination of CYT-C and Apaf-1 as well as the activity of caspase-3 (Ghribi *et al.*, 2001) reverses the death of DA-ergic neurons induced by

and the basis for the study of integrin in other cells (Leitinger & Hogg, 2002).

raft, but N-cadherin can bind to GDNF outside of the raft.

death. In neuroblastoma cells, the GDNF response is similar.

**6. GDNF treatment in Parkinson's disease** 

for neurodegenerative disease.

its function.

neurotoxins and stimulates functional recovery. In the presence of Aβ peptide, GDNF can inhibit Aβ neurotoxin via blocking the activation of gad153, JNK and ERK induced by ER stress, and promote cell survival. In granular cells of the cerebellum, GDNF inhibits the activation of p38-MAPK via PI3K, and prevents the injury induced by TGF-β in the protoplasmic membrane (Subramaniam *et al.*, 2008).

The summary of research progress on different GDNF supplementation methods is as follows:

Direct injection of GDNF: Some studies have demonstrated that injection of GDNF into the substantia nigra only decreases the apoptosis of DA-ergic neurons, but injection into the striatum corpora repairs the neuronal process of TH-positive neurons and improves the PD symptoms (Bohn *et al.*, 2000). Injection of GDNF into the striatum corpora can decrease the death rate of DA-ergic neurons up to 60%, as well as increase the ispilateral TH-positive immunoreactivity and synaptic growth in the nigrostriatal system. Additionally, injection of GDNF into the striatum corpora can increase the amount, volume and synaptic length of DA-ergic neurons as well as the uptake of dopamine.

Injection of GDNF via virus vehicle: This method can transfer the neurotrophic factors into the CNS (Figure 5). In 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) animal models, a lentivirus encoding GDNF was injected into the caudate nucleus, putamen and substantia nigra; a lentivirus encoding for the β-galactosidase enzyme was injected in the control group. Three months later, the PD system scores in the test groups were improved compared to that of the control group (Kordower *et al.*, 2000). Bowenkamp *et al.* used 6- OHDA animal models and obtained similar results (Bowenkamp *et al*., 1996). These findings indicate that GDNF plays a positive role in neuronal reconditioning and synaptic growth in the substantia nigra. However, over-expressed GDNF may also play a negative role in PD treatment. Eslamboli *et al.* reported that high levels of GDNF expression altered bilateral DA transport and synthesis, and produced toxic or adverse effects (Eslamboli *et al.*, 2003; Eslamboli *et al.*, 2005). Thus, low and sustained expression seems to be most prudent. In Huntington's disease, the repeated gene amplification of the huntingtin protein leads to neurological disorders, which can be attenuated by adenovirus-vehicle mediated GDNF expression.

Injection of GDNF via cell mediation: Researchers are searching for other efficacious ways to import GDNF. For example, the BHK-21 cell, transformed by genetic engineering, can persistently express GDNF in neural stem cells (Behrstock & Svendsen, 2004). Here, we introduce a method of delivering GDNF termed encapsulated cell delivery (ECD). This method requires a catheter-like device with a diameter of no more than 1 mm, which is used to inject agents into the striatum corpora. A mobile device, made up of a hollow fiber membrane (< 1.5 cm), is used to control the diffusion of nutrients and injected GDNF. Cell lines of high quality are selected via this fiber membrane. In the process, the membrane is a barrier preventing immunological rejection of allogenic cells.

GDNF permeable to the blood-brain barrier: Recently, a novel GDNF was developed. It is a fusion protein formed by the combination of GDNF and cell perforin mediated by the transactivator of the HIV virus. Diez *et al.* demonstrated, in the mouse PD model (i.p. injection), that Tat-GDNF could cross the blood-brain barrier and enter into the nigrostriatal system (Dietz *et al.*, 2006). Unfortunately, Tat -GDNF could not protect injured DA-ergic neurons or improve PD symptoms in MPTP models.

Actions of GDNF on Midbrain Dopaminergic Neurons: The Signaling Pathway 211

the motor functions were elevated by 39%, the quality of life was improved and daily activities were increased by 61%; no adverse effects were reported up to one year later. Results from PET scans demonstrated that GDNF enhanced dopamine in brain. This therapy also attenuated L-dopa induced tremors. Unfortunately, in 2004, Amgen conducted a largerscale trial, and GDNF did not exert effects as expected. Further, the company expressed concern over the safety of GDNF. After that, the direct application of GDNF was all but abandoned. However, exploration into mechanisms of action of GDNF is an alternative

After the identification of GDNF, three other family members (NTN, PSP and ART) were identified, and as further studies are undertaken, more family members may be uncovered. The receptors of these factors are composed of two parts: the GDNF family receptor and the transmembrane tyrosine kinase RET. The GPI-coupled GFR determines their specific signal

Initially, researchers assumed that only one signal pathway for GDNF existed: via GFRα1/2, anchored by receptor tyrosine kinase RET and glycosyl-phosphatidyl inositol. In 2003, *in vitro* studies demonstrated that the adhesion molecule NCAM was a receptor for GFLs that was independent of RET signaling. Recently, studies have indicated that integrin αV/β1 and N-cadherin may also be selective receptors for GDNF. As more evidence comes to light, it is evident that the signal pathway for GDNF protection of DA-ergic neurons is a complex one. In DA-ergic neurons, GDNF activated Ras-MAPK and PI3K/Akt signaling pathways play vital roles in neuronal survival and axon growth. Present studies indicate that NCAM, a new GFI receptor, is also involved in the signaling pathways of GDNF-GFRα1-MET and

NGF-TrkA-RET, which also makes the biological characteristics of GDNF complex.

Some progress has been made in therapeutic trials for neuronal degenerative diseases based on the neurotrophic actions of GDNF. Repeated, direct injection of GDNF is especially valuable for treating myeleterosis such as amyotrophic lateral sclerosis (ALS). Because of the features of the blood-brain barrier and brain tissues, direct injections of GDNF to treat PD and other neurodegenerative diseases have limited use. Thus, direct GDNF gene injection mediated by vehicle or transgenic cellular transplant is a promising alternative. Currently, the challenge lies in the development of a safe, novel vehicle for the delivery of genes that can be specially expressed in tissues, can induce gene expression and can replace the virus

In the management of PD, the application of drugs with low molecular weights is another choice for recombining neurotrophic factors. Other GFLs may be of potential value for treatment. Present studies demonstrate that, in murine PD models, PSPN, which is transferred into brains via neuronal stem cells, exerts effects as well as GDNF, on the survival of DA-ergic neurons in the midbrain (Akerud *et al.*, 2002). As we know, the expression of GFRα4, which is the receptor of PSPN, is more limited than that of GFRα1, the receptor of GDNF. Thus, we can propose that, even if high concentrations of PSPN are used, the adverse effects may be fewer than using GDNF (Lindahl *et al.*, 2000). The next challenge is to identify novel drugs, with low molecular weights and natural neurotrophic properties, to influence intracellular signal transduction. These explorations should be based on the comprehensive recognition of GFLs and the 3D structure of its receptor, as well as

avenue that may aid in the development of new therapies for PD treatment.

**7. Conclusion** 

transduction.

vehicle.

molecules, cells and disease.


#### (Figure source: www.epfl.ch)

Fig. 5. Injection of GDNF into the CNS to transfer neurotrophic factors via virus-vehicle.

Results from animal models provide invaluable information for clinical research. In 2002, Steven Gill and colleagues injected GDNF directly into the brain of PD patients, which significantly attenuated the PD symptoms (Gill *et al.*, 2003). The purpose of this Phase I clinic trial was to evaluate the safety of GDNF injections. Previous data, based on animal models, showed that GDNF can prolong the survival time of DA-ergic neurons and enhance their functions. In this study, researchers pumped GDNF into the striatum corpora of five PD patients at dose of 40 mg/d, for 18 months. These researchers believed that the cells closest to tip of the catheter absorbed the GDNF first. After permeating the brain tissues, GDNF reached the DA-ergic cells. The results demonstrated that the symptoms were attenuated, the motor functions were elevated by 39%, the quality of life was improved and daily activities were increased by 61%; no adverse effects were reported up to one year later. Results from PET scans demonstrated that GDNF enhanced dopamine in brain. This therapy also attenuated L-dopa induced tremors. Unfortunately, in 2004, Amgen conducted a largerscale trial, and GDNF did not exert effects as expected. Further, the company expressed concern over the safety of GDNF. After that, the direct application of GDNF was all but abandoned. However, exploration into mechanisms of action of GDNF is an alternative avenue that may aid in the development of new therapies for PD treatment.
