J Neuroimmune Pharmacol (2014) 9:182–194 DOI 10.1007/s11481-014-9533-5
INVITED REVIEW
Neurotrophin Strategies for Neuroprotection: Are They Sufficient? Joseph P. Steiner & Avindra Nath
Received: 4 December 2013 / Accepted: 13 February 2014 / Published online: 8 March 2014 # Springer Science+Business Media New York (outside the USA) 2014
Abstract As people are living longer, the prevalance of neurodegenerative diseases continues to rise resulting in huge socio-economic consequences. Despite major advancements in studying the pathophysiology of these diseases and a large number of clinical trials currently there is no effective treatment for these illnesses. All neuroprotective strategies have either failed or have shown only a minimal effect. There has been a major shift in recent years exploring the potential of neuroregenerative approaches. While the concept of using neurotropins for therapeutic purposes has been in existence for many years, new modes of delivery and expression of this family of molecules makes this approach now feasilble. Further neurotropin mimetics and receptor agonists are also being developed. The use of small molecules to induce the expression of neurotropins including repurposing of FDA approved drugs for this approach is another strategy being pursued. In the review we examine these new developments and discuss the potential for such approaches in the context of the pathophysiology of neurodegenerative diseases. Keywords Degeneration . Neurons . Neurotrophin . Neuroprotection Degeneration of neurons is the final common underlying cause in most neurological disorders. However, multiple processes and pathways lead to neuronal injury. Neurodegeneration may J. P. Steiner (*) : A. Nath NINDS Translational Neuroscience Center, National Institutes of Health, Room 7C-105; Bldg 10, 10 Center Drive, Bethesda, MD 20892, USA e-mail: [email protected] A. Nath Section of Infections of the Nervous System, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, USA
occur rapidly as a result of an acute insult, such as traumatic brain injury or a stroke. Alternatively, neurodegeneration occurs much more slowly and progressively over many years and decades, such as in HIV-associated neurocognitive disorders (HAND), Amytrophic Lateral Sclerosis (ALS), Huntington’s Disease (HD) Parkinson’s (PD) and Alzheimer’s Diseases (AD). A common feature in these neurodegenerative disorders is loss of neuronal connections, axonal injury and dendritic simplification. These axodendritic injuries result in loss of synaptic plasticity, and may ultimately result in neuronal cell death. The development of potential therapeutics to intervene in these neurodegenerative processes is an ongoing task in both the pharmaceutical industry and academic drug development groups. Some neurodegenerative disorders are the result of genetic mutations in one or more target genes. However, even with the genetic mutations resulting in these neurologic disorders, therapeutic development is still not rapidly forthcoming. These familial forms of neurodegeneration for ALS, PD and AD only account for a minority of reported cases of these neurologic disorders. Thus, we need to look at the different pathways involved in neurodegenerative disorders to identify therapeutic targets. Multiple targets exist to elicit different forms of neurodegeneration. Some of these targets include excitatory glutamate, mitochondrially mediated reactive oxygen species production and loss of antioxidant proteins, intracellular protein aggregates from the endoplasmic reticulum in the unfolded protein response, extracellular aggregated proteins from amyloid-β and synucleinopathies, the ubiquitin-proteasome, autophagy and programmed cell death. Additional toxins are also derived from neurotransmitter analogs, mitochondrial electron transport inhibitors and overexpressed forms of mutant disease specific proteins such as amyloid-β and αsynuclein. Inflammatory cytokines and oxidative stress induced by glial cells also can play a key role in the
J Neuroimmune Pharmacol (2014) 9:182–194
neurodegenerative processes. Over the past few decades, there have been many failed neuroprotective interventions, targeting different proteins or pathways, including the NMDA receptor, reactive oxygen species, L-type calcium channels, AChE, antioxidants, anti-inflammatory agents, calcium-activated proteases, kinases, phosphatases among others. There have been a number of clinical trials for neurotrophins, including CNTF for ALS, BDNF for ALS, GDNF for ALS and Parkinson’s Diseases and IGF-1 for ALS, all of which have been unsuccessful, very likely because of poor brain penetration of these neurotrophins (reviewed in Géral et al. 2013). While some of these clinical trials have provided some preliminary signs of a neuroprotective effect, causing a delay in disease progression as assessed by imaging of neurons, biomarker production or neurotransmitter levels, no definitive phase II clinical trial has yet demonstrated an appreciable and robust neuroprotective therapeutic effect. It is likely that a successful therapeutic approach needs to be able to counteract multiple molecules and pathways. In this review we explore the extent to which neurotrophins may serve that role.
Neurotrophin Hypothesis In 1981, Appel hypothesized that there was a unifying hypothesis for the cause of ALS, PD and AD (Appel 1981). It was suggested that these neurologic disorders resulted from the lack of a disease specific hormone or neurotrophic factor secreted by the target tissue neurons. This disease specific hormone would be released by the postsynaptic cell and then exert its effect in a retrograde fashion following reuptake up by the presynaptic terminal. (Appel 1981). Thus a hormone or growth factor would be critical to the growth and development of many types of neurons. Specifically, growth factors such as nerve growth factor (NGF), brain derived neurotrophic factor (BDNF) and glial cell-line derived neurotrophic factor (GDNF) may act to promote the survival and function of sensory and motor neurons, as well as cholinergic, serotonergic, cortical and hippocampal neurons. (Wu 2005; Nagahara and Tuszynski 2011) Reductions in these neurotrophic factors have been associated with a number of neurodegenerative disorders, such as AD, HAND, PD, ALS among others. (Allen et al. 2013) and references therein). Thus preclinical drug development directed toward enhancing neurotrophin signaling was suggested to provide a novel therapeutic approach for neurodegenerative disorders.
Neurotrophin Receptors The first of the neurotrophic or nerve survival factors described was nerve growth factor (NGF, (Levi-Montalcini and
183
Hamburger 1951), a soluble factor secreted from mouse tumor that promoted neuronal fiber outgrowth in the treated chick embryos sensory and sympathetic neurons. (Levi-Montalcini 1952; Cohen et al. 1954; Levi-Montalcini and Cohen 1960) NGF is initially translated as a precursor polypeptide, proNGF which is cleaved extracellularly by plasmin and matrix metalloproteases (Lee et al. 2001) to produce the mature form of NGF. The homodimeric NGF exerts its actions by interacting with two receptors, the tropomyosin receptor tyrosine kinase (TrkA) and the low affinity p75 pan neurotrophin receptor (p75NTR). (Aloe et al. 2012) In the presence of TrkA, NGF promotes neuronal survival, differentiation and cell growth by utilizing the shc/Grb2, phospholipase Cgamma/PIP2, PI3K, ras/ERK, Akt and PKC pathways. (reviewed in (Allen et al. 2013)). The p75NTR assists the selective binding of NGF to TrkA (Chao and Hempstead 1995) to achieve these neuronal growth effects. The p75NTR alone in the absence of TrkA and NGF can promote neuronal apoptosis via activation of the TNF receptor associated factor (Trafs), NFkB and ceramide (Blochl and Blochl 2007). In addition to NGF, the other members of the neurotrophin family of survival factors and neuronal growth promoting proteins includes brain derived neurotrophic factor (BDNF), neurotrophin 3 (NT3) and neurotrophin 4 (NT4). (Reichardt and Mobley 2004; Allen et al. 2013). BDNF binds with high affinity to TrkB, but with lower nanomolar affinity to p75NTR (Chao 1994). The third neurotrophin discovered, designated NT3 (Maisonpierre et al. 1990), has the highest affinity for the Trk C protein (Lamballe et al. 1991), but also potently interacts with Trk B, and also with lesser affinity to the p75NTR. The fourth neurotrophin identified, NT4 (also designated NT5 or NT4/5) signals mainly through the Trk B receptor kinase (Berkemeier et al. 1991). While there may be a role for NGF, BDNF, NT3 and NT4 signaling through the Trk A, B and C, each of these neurotrophins may also signal through the p75NTR with nanomolar affinity (Chao 1994). Interestingly, in the absence of Trks and neurotrophins, p75NTR may elicit apoptosis via ceramide and NFkB activation through TNF receptor associated factors (Gruss and Dower 1995). The p75NTR can bind to both the mature and pro- forms of the neurotrophins, along with other proteins to promote neuronal outgrowth, proliferation and also cell death (Chao et al. 2006). The p75NTR protein, when bound by the precursor of NGF, pro-NGF, and a co-receptor protein sortilin induced apoptosis (Lee et al. 2001). Similarly, the precursor of BDNF, proBDNF, also binds to sortilin in the presence of p75NTR when released by cultured neurons (Teng et al. 2005). This pro-BDNF induced apoptosis required the cell surface interaction of proBDNF with sortilin and p75NTR (Teng et al. 2005).
184
Neurotrophin Mediated Therapeutics Preclinical studies in rodent models of AD, PD, HD and ALS with both NGF and BDNF have demonstrated significant efficacy (Allen et al. 2013). For example, in animal models of Huntington’s Disease with increased CAG repeats, BDNF mRNA and protein levels reduced by more than 70 % in the striatum (Gharami et al. 2008) and 60 % in hippocampus (Lynch et al. 2007). Strategies to deliver BDNF via transplanted BDNF secreting cells (Perez-Navarro et al. 2000), or BDNF expression driven by CaMKII and GFAP promoters (Gharami et al. 2008; Giralt et al. 2010) normalized the BDNF levels, protected neurons from neurotoxicity and ameliorated HD phenotypes in the mice. However, the inability to deliver the peptide neurotrophic factors systemically has limited the direct clinical utility of BDNF and other neurotrophic factors. Successful delivery of peptidic neurotrophic factors to the CNS targets is challenging, because the neurotrophins have low bioavailability. In general, therapeutic application of BDNF proteins for neurologic disorders is difficult because they have a short half-life, poor bioavailability due to proteolysis and low permeability through the blood brain barrier (Géral et al. 2013) BDNF has a very short plasma half-life, determined to be less than 10 min (Pardridge et al. 1994). Second, the compound has poor blood brain barrier permeability (Wu 2005). It is interesting that a report has suggested that there may be active transport through the blood brain barrier (BBB) by direct administration of BDNF (Pan et al. 1998). Direct injection of recombinant human BDNF to ALS patients had little clinical impact (Group 1999). While the therapy was well tolerated by the patients, the BDNF had an extremely short half-life and limited CNS exposure through the BBB. Similarly, intrathecal administration of recombinant human BDNF up to 150 ug/day was well tolerated, with dose dependent CSF levels of BDNF (Ochs et al. 2000). The trial was too small to allow conclusions on clinical efficacy of the treatment. In preclinical models of neurodegenerative disorders, BDNF treatment has been used successfully to block neuropathology resulting from ischemic, ALS, HD, PD, AD and spinal cord injuries, following direct administration of BDNF (Allen et al. 2013). BDNF has been dosed via intravenus administration after focal cerebral ischemia (Schabitz et al. 2000), with reduced cortical infract volume and neurological deficits. Bax positive neurons were decreased, while Bcl-2 positive neuronal staining was increased within the penumbra of BDNF treated rats. In a model of cerebral venous injury, BDNF was also delivered via intraventricular infusion pumps (Takeshima et al. 2011), with the result of reduced infarct size and fewer TUNEL-positive apoptotic cells in BDNF treated rats, along with no effect on cerebral blood flow. In addition,
J Neuroimmune Pharmacol (2014) 9:182–194
direct intrahippocampal injection of BDNF was also effective at inducing long term memory persistence in rats in the absence of new protein synthesis. (Bekinschtein et al. 2008). Direct intracerebroventricular administration of BDNF was effective in a rat modified forced swim test, where the BDNF produced a long lasting antidepressant effect, but no changes in locomotor activity (Hoshaw et al. 2005). While delivery of large protein molecules like BDNF across the blood brain barrier is not efficient, other routes of administration have been employed to deliver BDNF to the target tissue in the CNS. Intranasal delivery of BDNF to rats resulted in nanomolar brain concentrations of the neurotrophin within 25 min of dosing. These levels of BDNF were sufficient to activate prosurvival PI3K/Akt pathway, resulting in nearly 2-fold increase in phosphoAkt from 24 through 96 h after a single intranasal dose (Alcala-Barraza et al. 2010).
Viral Vector Delivery of BDNF Viral delivery of BDNF has proven to be an effective neuroprotective strategy for delivery of BDNF to the CNS. Adenoassociated viral vector delivery of BDNF and GDNF was neuroprotective in a quinolinic acid rat model of HD. (Kells et al. 2004) Instriatal delivery the neurotrophins protected against striatal neuron death in the this model, as well as protecting NOS-positive interneurons. In another series of experiments, adenoviral vector delivery of BDNF promoted recovery of spinal cord axons following spinal cord transection (Koda et al. 2004). The regeneration of rubrospinal axons was also accompanied by significant locomotor function in the hindlimbs of the BDNF treated rats. Lentiviral expression of BDNF into multiple animal models of Alzheimer’s Disease provided significant neuroprotection, even when administered after disease onset (Nagahara et al. 2009). Lenti-BDNF reversed synaptic loss and cognitive decline, improved agerelated changes in gene expression and restored cell signaling. Another example of virally delivered BDNF was provided with Sendai virus-delivered BDNF, which when injected into subcortical white matter, induced BDNF levels and ameliorated memory deficits and synaptic degeneration in a mouse model of AD (Iwasaki et al. 2012).
Cell-Based Delivery of BDNF Delivery of BDNF to the CNS has also been successfully achieved via stem cell delivery of the neurotrophin. In one study, bone marrow stromal cells expressing BDNF were utilized to treat rats with transected spinal cord injuries. (Lu et al. 2005a, b). There was significant expression of the BDNF protein and regrowth of sensory and motor axons into the
J Neuroimmune Pharmacol (2014) 9:182–194
lesion site, but functional recovery was not observed. Bone marrow stem cells (BMSC) engineered to express BDNF were also used in a more model of experimental autoimmune encephalomyelitis (Makar et al. 2009). Mice receiving the BDNF expressing BMSC displayed delayed EAE symptom onset and reduced clinical severity score. They also had decreased demyelination and increased remyelination, along with inhibition of brain levels of TNF-α and IFN-γ. In aged triple transgenic mice (3xTg-AD) expressing pathogenic forms of amyloid precursor protein, presenilin, and tau, hippocampal neural stem cell transplantation rescues the spatial learning and memory deficits in aged 3xTg-AD mice. (Blurton-Jones et al. 2009). Interestingly, cognitive function was improved without altering A or tau pathology. Instead, the improved cognition involves BDNF-mediated enhancement of hippocampal synaptic density. Additionally, BDNF secreting human mesenchymal stem cells have also been transplanted into the substantia nigra of 6-OHDA lesioned rats, resulting in increased nigral and striatal tyrosine hydroxylase, accompanied by attenuation of amphetamine induced motor activity. (Somoza et al. 2010). In addition to stem cells, BDNF has also been successfully delivered from fibroblasts expressing BDNF. Levivier and colleagues (Levivier et al. 1995) utilized the intrastriatal 6hydroxydopamine (6-OHDA) model of PD, and treated the animals with intrastriatal grafts of BDNF producing fibroblasts. They found that the BDNF-producing fibroblasts prevented loss of nigral neuronal cell bodies and partially blocked loss of striatal nerve terminals, compared to control fibroblast transplants. Similar neuroprotective observations were made in a MPP+ lesioned rat following intranigral implantation of BDNF-producing fibroblasts (Frim et al. 1994).
Polymer Systems to Encapsulate, Deliver and Release BDNF Another approach to deliver the BDNF is to modify the protein, or encapsulate it in polymers to deliver it to the target tissue. One method utilized was to covalently add polyethyleneglycol polymers to the carboxyl residues of glutamate and aspartate. (Sakane and Pardridge 1997) PEGylation of BDNF with PEG2000 and PEG5000 significantly reduced the systemic clearance of the BDNF-PEG following Intravenous administration, with only minimal loss of BDNF. In order to enhance the delivery of BDNF to the spinal cord, PEGylated BDNF was administered intrathecally, resulting in improved half life of the BDNF in CSF and enhanced penetration into the spinal cord tissue (Soderquist et al. 2009). N-terminal PEGylated BDNF was administered after spinal cord injury and resulted in axonal sprouting and some locomotor recovery (Ankeny et al. 2001), although these
185
improvements were similar to high doses of native BDNF treatment. Synthetic polymeric systems to encapsulate and then release BDNF have been devised and successfully employed. Bertram et al. (Bertram et al. 2010) synthesized polymers based on poly (lactic co-glycolic) acid (PLGA), poly Llysine and PEG to load and deliver BDNF. They found that a polymer of PLGA, PLL, and PEG led to the delivery of BDNF for periods of time greater than 60 days and that the BDNF delivered was bioactive. Other naturally occurring polymers, like those containing collagen, have been applied to spinal cord injury (SCI) in rats (Han et al. 2009). They developed a linear ordered collagen scaffold to generate a collagen binding domain-BDNF system for use in a rat hemisection model of SCI. The collagen-BDNF system improved neuron survival and recovery from SCI recovery. Another natural polymer, hyaluronic acid based hydrogel containing a matrix metalloproteinase peptide, IKVAV peptide derived from laminin and BDNF, with human mesenchymal stem cells was engineered for treatment of spinal cord injury (Park et al. 2010). Intrathecal injection of rats with these hydrogels containing BDNF showed the greatest improvement on locomotor tests. It was hypothesized that the hyaluronic acid-based hydrogels containing IKVAV and BDNF created microenvironments that facilitated differentiation of stem cells along the neural cell lineage, resulting in a regenerative response.
BDNF Peptidomimetics Numerous studies have demonstrated the utility of the BDNF protein in in vitro and in vivo models of neurodegeneration. (Allen et al. 2013; Lu et al. 2013) and references therein) However, the utility of BDNF is hampered by its lack of CNS penetration and distribution. The idea to derive low molecular weight peptides that retain the neurotrophic actions of BDNF but have improved pharmacokinetic parameters and brain penetration was developed. (O’Leary and Hughes 1998) From a homology model of the BDNF homodimer, O’Leary and Hughes designed small peptides that mimicked loop 2 of BDNF and found active amino acid residues linked to the activation of TrkB in a monocyclic 10 amino acid peptide. In subsequent work, they used that 10mer to create a series of bicyclic dimeric peptides with pairs of 10 amino acid peptides (O’Leary and Hughes 2003). The neuroprotective optimization process led to a tricyclic dimeric peptide that was very potent in promoting survival of chick sensory ganglion neurons, with an Ec50 of 11pM, only about 2 fold less active than BDNF itself. Additional BDNF peptidomimetics were synthesized from loops 1, 2 and 4 of BDNF as peptide monomers and dimers, with many of these compounds displaying nanomolar potency at promoting sensory neuronal survival
186
J Neuroimmune Pharmacol (2014) 9:182–194
Table 1 Inducers of BDNF Class
System(s)
References
Antidepressants fluoxetine Sertraline
rats in vivo Human patients
(Dwivedi et al. 2006; Molteni et al. 2006)l (Nibuya et al. 1995; Peng et al. 2008b; Matrisciano et al. 2009) (Cattaneo et al. 2010)
escitalopram
human patients
Paroxetine Imipramine Desimipramine
human astrocyte-glioma cells in vitro in vitro NSCs mice in vivo
Nortriptyline Amitriptyline
mice in vivo mice in vivo, human patients
(Martinez-Turrillas et al. 2005; Angelucci et al. 2011) (Peng et al. 2008a) (Nibuya et al. 1995; Jacobsen and Mork 2004; Dwivedi et al. 2006) (Cattaneo et al. 2010) (Xu et al. 2003; Lee et al. 2010)
rat cortical neurons in vitro hippo in vitro; rats in vivo mice in vivo mice in vivo rats in vivo rats in vivo
(Yagasaki et al. 2006) (Cooke et al. 2009) (Calabrese et al. 2007; Mannari et al. 2008) (Ikenouchi-Sugita et al. 2009) (Rogoz et al. 2005a) (Nibuya et al. 1995)
Dosed with sertraline-human patients rats in vivo rats in vivo
(Yoshimura et al. 2008) (Czubak et al. 2009) (Bai et al. 2003)
mouse astrocytes in vitro rat PC12 cells in vitro In vitro mouse neurons, rat cortical cells, mice in vivo rats in vivo
(Mizuta et al. 2000) (Mizuta et al. 2000) (Youdim et al. 2005) (Kupershmidt et al. 2009; Avramovich-Tirosh et al. 2010; Youdim 2012) (Coppell et al. 2003)
human patients, mice in vivo mice in vivo
(Leyhe et al. 2008; Autio et al. 2011) (Autio et al. 2011; Gil-Bea et al. 2011)
mice in vivo mice in vivo
(Wang et al. 2006) (Francis et al. 2012)
rat astrocytes, DA neurons in vitro, rats in vivo rat cortical neurons, rats in vivo mice in vivo DA neurons in vitro mice in vivo rats in vivo
(Rahvar et al. 2011; Zhang et al. 2012a, 2012b)
Fluvoxamine Venlafaxine Duloxetine Milnacipran Mirtazapine Mianserin Antipsychotics Risperidone Olanzapine Clozapine MAO inhibitors Selegiline Rasagiline TVP1022 M30 Tranylcypromine AChE inhibitors Donepezil Galantamine Huperzine A Rivastigmine Antioxidants Resveratrol Curcumin J147 Caffeic Acid Phenethyl ester Ferulic acid Ginko biloba EGb761 Activities Voluntary Exercise Intermittant Fasting Mood stabilizers Ampakines Lithium Valproate
mice in vivo, human patients mice and rats in vivo rat cortical and hippocampal slices in vitro cortical neurons in vitro; mice and rats in vivo DA neurons in vitro
(Wu et al. 2006; Wang et al. 2008) (Prior et al. 2013) (Kurauchi et al. 2012) (Sultana et al. 2005; Yabe et al. 2010) (Zhang et al. 2012d) (Adlard et al. 2004, 2005; Rasmussen et al. 2009; Lee et al. 2013) (Lee et al. 2002; Duan et al. 2003) (Mackowiak et al. 2002; Lauterborn et al. 2003, 2009) (Hashimoto et al. 2002; Jacobsen and Mork 2004; Young 2009; Chiu et al. 2013) (Chen et al. 2006; Wu et al. 2008b)
J Neuroimmune Pharmacol (2014) 9:182–194
187
Table 1 (continued) Class
System(s)
References
carbamazepam Lamotrigine Dopamine agonists
rats in vivo rats in vivo
(Chang et al. 2009) (Chang et al. 2009)
Pramipexole Ropinerole NMDA receptor antagonists MK801 Ketamine Memantine Flavonoids Anthocyanins Quercetin/kaempferol Hyperoside Green Tea Catechins L-theanine EGCG Miscellaneous inducers Simastatin Cystamine/Cysteamine Butaprost Deltamethrin
DA neurons in vitro DA neurons in vitro
(Du et al. 2005; Lauterbach 2012) Du et al. 2005; Lauterbach 2012)
mice in vivo mice in vivo rats in vivo
(Autry et al. 2011) (Autry et al. 2011) (Reus et al. 2012)
rats in vivo mice and rats in vivo rat PC12 cells in vitro mice in vitro, rats in vivo SH-SY-5Y cells in vitro, rats in vivo hippocampal and cortical neurons in vitro
(Rendeiro et al. 2013) (Hou et al. 2010; Yao et al. 2012) (Zheng et al. 2012) (Li et al. 2009; Rendeiro et al. 2013) (Cho et al. 2008; Takeda et al. 2011) (Nath et al. 2012)
rats in vivo mice in vivo human microglia and astrocytes in vitro rat cortical cultures in vitro
(Wu et al. 2008a) (Borrell-Pages et al. 2006) (Hutchinson et al. 2009) (Imamura et al. 2006; Ihara et al. 2012) (Visanji et al. 2008; Zhang et al. 2012c) (Mizuta et al. 2000; Katoh-Semba et al. 2002)
Agomelatine Nicotine Metyrapone Rolipram Catalpol Fingolimod
rat DA neurons in vitro; mice in vivo mouse cultured astrocytes, hippocampal neurons rats in vivo mice in vivo rats in vivo, synergy with imipramine rats in vivo, synergistic with desimipramine rat cortical neurons in vitro, mouse in vivo cortical neurons in vitro, mice in vivo
(Ladurelle et al. 2012) (Czubak et al. 2009) (Rogoz et al. 2005b; Rogoz and Legutko 2005) (Fujimaki et al. 2000) (Wang et al. 2009) (Deogracias et al. 2012; Doi et al. 2013)
CEP 1347 BW373U86
mice in vivo rats in vivo
(Apostol et al. 2008; Conforti et al. 2008) (Torregrossa et al. 2004)
PYM50028 (Cogane) Riluzole
(Fletcher and Hughes 2006). Further optimization of these peptides would need to include decreased size, improved solubility, membrane permeability and resistance to proteolytic degradation. Fletcher and colleagues (Fletcher et al. 2008) then designed a BDNF loop 4 mimetic, a proteolytically stable cyclic pentapeptide cyclo[dPAKKR] with these improved pharmaceutic properties, as well as an effective survival factor for sensory ganglion neurons. Interestingly, these functional BDNF mimetics work independently of TrkB and downstream MAPK signaling. They have optimized these cyclic pentapeptides in a second generation of compounds that showed improved biological activity, stability in plasma, and an ability to cross model biological membranes. (Fletcher and Hughes 2009). A most recent study used tetrapeptides synthesized to 5 different regions of the BDNF protein,
and found neurotrophic effects on hippocampal neuronal survival. One of the compounds, B3 potentiated the neuronal survival promoting activity of BDNF (Cardenas-Aguayo Mdel et al. 2013).
Small Molecule Mimetics of BDNF (TrkB Agonists) To overcome issues involved in delivering the protein as a therapeutic, small molecule mimetics of BDNF, which had superior biopharmaceutic properties including pharmacokinetic potential to reach the CNS, were sought and identified. Massa and Longo identified numerous structures from in silico screening of the loop II region of BDNF and by low throughput screening of hippocampal neurons (Massa et al. 2010) These four prototypic compounds were neuroprotective
188
J Neuroimmune Pharmacol (2014) 9:182–194
in a concentration dependent manner with high picomolar activity, selectively activated TrkB with high potency and specificity and promoted neuronal survival against multiple neurotoxins, differentiation, and synaptic function. When LM-22A-4 was administered in vivo, it caused hippocampal and striatal TrkB activation in mice and improved motor learning after traumatic brain injury in rats. (Massa et al. 2010).
Small Molecule Inducers of BDNF Use of small molecule drugs to upregulate BDNF mRNA and increase cellular BDNF levels in the target tissue is another approach to elicit a neuroprotective effect. This approach utilizes FDA approved drugs and other small molecules with good systemic bioavailability, chemical and metabolic stability, membrane permeability and CNS penetration through the blood brain barrier, improved biopharmaceutic properties to induce cellular BDNF levels. A number of agents have been
reported to elicit increased BDNF protein, including antidepressants, antipsychotics, MAO inhibitors, AChE inhibitors, antioxidants, mood stabilizers, dopamine agonists, NMDA receptor antagonists, flavonoids and additional miscellaneous compounds. This list of compounds is presented in Table 1. Of the compounds listed in Table 1, only riluzole has been approved for its neuroprotective effects in treatment of ALS. While the use of small molecules has the advantage of ease of delivery and repurposing of FDA approved drugs has the advantage of quick translation to clinical trials, off target effects may be of concern particularly if the dosage needed is greater than it’s approved dose range.
Application of BDNF and Neurotrophin Therapeutics to HIVAssociated Neurocognitive Disorders HIV infection into the CNS results in synaptodendritic damage in neurons, and these synaptodendritic degenerative
Pro-BDNF BDNF Tat+ 1 EGCG Tat+0.1 EGCG
Tat+1Epi
Tat+0.1Epi Tat
Media
Tat+ 1 EGCG Tat+0.1 EGCG
Tat+1Epi Tat+0.1Epi Tat Media
Fig. 1 Catechins reverse Tat mediated changes in BDNF and proBDNF expression in mixed rat neuronal cultures. Rat mixed cortical cultures (106 cells per well on six-well plates) were treated with culture media plus 0.1 % DMSO vehicle, 500 nM Tat plus vehicle, or Tat plus 0.1 or 1 μM epicatechin or Tat plus 0.1 or 1 μM EGCG for 24 h. After 24 h, cell lysates were generated, proteins separated bySDS–polyacrylamide gel electrophoresis, transferred to PVDF membranes and probedwith antibodies to BDNF. The BDNF antibody recognized the precursor of BDNF
(proBDNF) at 32 kDa as well as the mature BDNF protein at 13 kDa. Two replicate experiments are depicted. The data are expressed as percent expression of the media/vehicle treated controls The data from three independent replicates were utilized for each treatment group, and were evaluated by ANOVA for significance. Group-wise post hoc comparisons were assessed by Newman–Keuls multiple comparison test. *p
INVITED REVIEW
Neurotrophin Strategies for Neuroprotection: Are They Sufficient? Joseph P. Steiner & Avindra Nath
Received: 4 December 2013 / Accepted: 13 February 2014 / Published online: 8 March 2014 # Springer Science+Business Media New York (outside the USA) 2014
Abstract As people are living longer, the prevalance of neurodegenerative diseases continues to rise resulting in huge socio-economic consequences. Despite major advancements in studying the pathophysiology of these diseases and a large number of clinical trials currently there is no effective treatment for these illnesses. All neuroprotective strategies have either failed or have shown only a minimal effect. There has been a major shift in recent years exploring the potential of neuroregenerative approaches. While the concept of using neurotropins for therapeutic purposes has been in existence for many years, new modes of delivery and expression of this family of molecules makes this approach now feasilble. Further neurotropin mimetics and receptor agonists are also being developed. The use of small molecules to induce the expression of neurotropins including repurposing of FDA approved drugs for this approach is another strategy being pursued. In the review we examine these new developments and discuss the potential for such approaches in the context of the pathophysiology of neurodegenerative diseases. Keywords Degeneration . Neurons . Neurotrophin . Neuroprotection Degeneration of neurons is the final common underlying cause in most neurological disorders. However, multiple processes and pathways lead to neuronal injury. Neurodegeneration may J. P. Steiner (*) : A. Nath NINDS Translational Neuroscience Center, National Institutes of Health, Room 7C-105; Bldg 10, 10 Center Drive, Bethesda, MD 20892, USA e-mail: [email protected] A. Nath Section of Infections of the Nervous System, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, USA
occur rapidly as a result of an acute insult, such as traumatic brain injury or a stroke. Alternatively, neurodegeneration occurs much more slowly and progressively over many years and decades, such as in HIV-associated neurocognitive disorders (HAND), Amytrophic Lateral Sclerosis (ALS), Huntington’s Disease (HD) Parkinson’s (PD) and Alzheimer’s Diseases (AD). A common feature in these neurodegenerative disorders is loss of neuronal connections, axonal injury and dendritic simplification. These axodendritic injuries result in loss of synaptic plasticity, and may ultimately result in neuronal cell death. The development of potential therapeutics to intervene in these neurodegenerative processes is an ongoing task in both the pharmaceutical industry and academic drug development groups. Some neurodegenerative disorders are the result of genetic mutations in one or more target genes. However, even with the genetic mutations resulting in these neurologic disorders, therapeutic development is still not rapidly forthcoming. These familial forms of neurodegeneration for ALS, PD and AD only account for a minority of reported cases of these neurologic disorders. Thus, we need to look at the different pathways involved in neurodegenerative disorders to identify therapeutic targets. Multiple targets exist to elicit different forms of neurodegeneration. Some of these targets include excitatory glutamate, mitochondrially mediated reactive oxygen species production and loss of antioxidant proteins, intracellular protein aggregates from the endoplasmic reticulum in the unfolded protein response, extracellular aggregated proteins from amyloid-β and synucleinopathies, the ubiquitin-proteasome, autophagy and programmed cell death. Additional toxins are also derived from neurotransmitter analogs, mitochondrial electron transport inhibitors and overexpressed forms of mutant disease specific proteins such as amyloid-β and αsynuclein. Inflammatory cytokines and oxidative stress induced by glial cells also can play a key role in the
J Neuroimmune Pharmacol (2014) 9:182–194
neurodegenerative processes. Over the past few decades, there have been many failed neuroprotective interventions, targeting different proteins or pathways, including the NMDA receptor, reactive oxygen species, L-type calcium channels, AChE, antioxidants, anti-inflammatory agents, calcium-activated proteases, kinases, phosphatases among others. There have been a number of clinical trials for neurotrophins, including CNTF for ALS, BDNF for ALS, GDNF for ALS and Parkinson’s Diseases and IGF-1 for ALS, all of which have been unsuccessful, very likely because of poor brain penetration of these neurotrophins (reviewed in Géral et al. 2013). While some of these clinical trials have provided some preliminary signs of a neuroprotective effect, causing a delay in disease progression as assessed by imaging of neurons, biomarker production or neurotransmitter levels, no definitive phase II clinical trial has yet demonstrated an appreciable and robust neuroprotective therapeutic effect. It is likely that a successful therapeutic approach needs to be able to counteract multiple molecules and pathways. In this review we explore the extent to which neurotrophins may serve that role.
Neurotrophin Hypothesis In 1981, Appel hypothesized that there was a unifying hypothesis for the cause of ALS, PD and AD (Appel 1981). It was suggested that these neurologic disorders resulted from the lack of a disease specific hormone or neurotrophic factor secreted by the target tissue neurons. This disease specific hormone would be released by the postsynaptic cell and then exert its effect in a retrograde fashion following reuptake up by the presynaptic terminal. (Appel 1981). Thus a hormone or growth factor would be critical to the growth and development of many types of neurons. Specifically, growth factors such as nerve growth factor (NGF), brain derived neurotrophic factor (BDNF) and glial cell-line derived neurotrophic factor (GDNF) may act to promote the survival and function of sensory and motor neurons, as well as cholinergic, serotonergic, cortical and hippocampal neurons. (Wu 2005; Nagahara and Tuszynski 2011) Reductions in these neurotrophic factors have been associated with a number of neurodegenerative disorders, such as AD, HAND, PD, ALS among others. (Allen et al. 2013) and references therein). Thus preclinical drug development directed toward enhancing neurotrophin signaling was suggested to provide a novel therapeutic approach for neurodegenerative disorders.
Neurotrophin Receptors The first of the neurotrophic or nerve survival factors described was nerve growth factor (NGF, (Levi-Montalcini and
183
Hamburger 1951), a soluble factor secreted from mouse tumor that promoted neuronal fiber outgrowth in the treated chick embryos sensory and sympathetic neurons. (Levi-Montalcini 1952; Cohen et al. 1954; Levi-Montalcini and Cohen 1960) NGF is initially translated as a precursor polypeptide, proNGF which is cleaved extracellularly by plasmin and matrix metalloproteases (Lee et al. 2001) to produce the mature form of NGF. The homodimeric NGF exerts its actions by interacting with two receptors, the tropomyosin receptor tyrosine kinase (TrkA) and the low affinity p75 pan neurotrophin receptor (p75NTR). (Aloe et al. 2012) In the presence of TrkA, NGF promotes neuronal survival, differentiation and cell growth by utilizing the shc/Grb2, phospholipase Cgamma/PIP2, PI3K, ras/ERK, Akt and PKC pathways. (reviewed in (Allen et al. 2013)). The p75NTR assists the selective binding of NGF to TrkA (Chao and Hempstead 1995) to achieve these neuronal growth effects. The p75NTR alone in the absence of TrkA and NGF can promote neuronal apoptosis via activation of the TNF receptor associated factor (Trafs), NFkB and ceramide (Blochl and Blochl 2007). In addition to NGF, the other members of the neurotrophin family of survival factors and neuronal growth promoting proteins includes brain derived neurotrophic factor (BDNF), neurotrophin 3 (NT3) and neurotrophin 4 (NT4). (Reichardt and Mobley 2004; Allen et al. 2013). BDNF binds with high affinity to TrkB, but with lower nanomolar affinity to p75NTR (Chao 1994). The third neurotrophin discovered, designated NT3 (Maisonpierre et al. 1990), has the highest affinity for the Trk C protein (Lamballe et al. 1991), but also potently interacts with Trk B, and also with lesser affinity to the p75NTR. The fourth neurotrophin identified, NT4 (also designated NT5 or NT4/5) signals mainly through the Trk B receptor kinase (Berkemeier et al. 1991). While there may be a role for NGF, BDNF, NT3 and NT4 signaling through the Trk A, B and C, each of these neurotrophins may also signal through the p75NTR with nanomolar affinity (Chao 1994). Interestingly, in the absence of Trks and neurotrophins, p75NTR may elicit apoptosis via ceramide and NFkB activation through TNF receptor associated factors (Gruss and Dower 1995). The p75NTR can bind to both the mature and pro- forms of the neurotrophins, along with other proteins to promote neuronal outgrowth, proliferation and also cell death (Chao et al. 2006). The p75NTR protein, when bound by the precursor of NGF, pro-NGF, and a co-receptor protein sortilin induced apoptosis (Lee et al. 2001). Similarly, the precursor of BDNF, proBDNF, also binds to sortilin in the presence of p75NTR when released by cultured neurons (Teng et al. 2005). This pro-BDNF induced apoptosis required the cell surface interaction of proBDNF with sortilin and p75NTR (Teng et al. 2005).
184
Neurotrophin Mediated Therapeutics Preclinical studies in rodent models of AD, PD, HD and ALS with both NGF and BDNF have demonstrated significant efficacy (Allen et al. 2013). For example, in animal models of Huntington’s Disease with increased CAG repeats, BDNF mRNA and protein levels reduced by more than 70 % in the striatum (Gharami et al. 2008) and 60 % in hippocampus (Lynch et al. 2007). Strategies to deliver BDNF via transplanted BDNF secreting cells (Perez-Navarro et al. 2000), or BDNF expression driven by CaMKII and GFAP promoters (Gharami et al. 2008; Giralt et al. 2010) normalized the BDNF levels, protected neurons from neurotoxicity and ameliorated HD phenotypes in the mice. However, the inability to deliver the peptide neurotrophic factors systemically has limited the direct clinical utility of BDNF and other neurotrophic factors. Successful delivery of peptidic neurotrophic factors to the CNS targets is challenging, because the neurotrophins have low bioavailability. In general, therapeutic application of BDNF proteins for neurologic disorders is difficult because they have a short half-life, poor bioavailability due to proteolysis and low permeability through the blood brain barrier (Géral et al. 2013) BDNF has a very short plasma half-life, determined to be less than 10 min (Pardridge et al. 1994). Second, the compound has poor blood brain barrier permeability (Wu 2005). It is interesting that a report has suggested that there may be active transport through the blood brain barrier (BBB) by direct administration of BDNF (Pan et al. 1998). Direct injection of recombinant human BDNF to ALS patients had little clinical impact (Group 1999). While the therapy was well tolerated by the patients, the BDNF had an extremely short half-life and limited CNS exposure through the BBB. Similarly, intrathecal administration of recombinant human BDNF up to 150 ug/day was well tolerated, with dose dependent CSF levels of BDNF (Ochs et al. 2000). The trial was too small to allow conclusions on clinical efficacy of the treatment. In preclinical models of neurodegenerative disorders, BDNF treatment has been used successfully to block neuropathology resulting from ischemic, ALS, HD, PD, AD and spinal cord injuries, following direct administration of BDNF (Allen et al. 2013). BDNF has been dosed via intravenus administration after focal cerebral ischemia (Schabitz et al. 2000), with reduced cortical infract volume and neurological deficits. Bax positive neurons were decreased, while Bcl-2 positive neuronal staining was increased within the penumbra of BDNF treated rats. In a model of cerebral venous injury, BDNF was also delivered via intraventricular infusion pumps (Takeshima et al. 2011), with the result of reduced infarct size and fewer TUNEL-positive apoptotic cells in BDNF treated rats, along with no effect on cerebral blood flow. In addition,
J Neuroimmune Pharmacol (2014) 9:182–194
direct intrahippocampal injection of BDNF was also effective at inducing long term memory persistence in rats in the absence of new protein synthesis. (Bekinschtein et al. 2008). Direct intracerebroventricular administration of BDNF was effective in a rat modified forced swim test, where the BDNF produced a long lasting antidepressant effect, but no changes in locomotor activity (Hoshaw et al. 2005). While delivery of large protein molecules like BDNF across the blood brain barrier is not efficient, other routes of administration have been employed to deliver BDNF to the target tissue in the CNS. Intranasal delivery of BDNF to rats resulted in nanomolar brain concentrations of the neurotrophin within 25 min of dosing. These levels of BDNF were sufficient to activate prosurvival PI3K/Akt pathway, resulting in nearly 2-fold increase in phosphoAkt from 24 through 96 h after a single intranasal dose (Alcala-Barraza et al. 2010).
Viral Vector Delivery of BDNF Viral delivery of BDNF has proven to be an effective neuroprotective strategy for delivery of BDNF to the CNS. Adenoassociated viral vector delivery of BDNF and GDNF was neuroprotective in a quinolinic acid rat model of HD. (Kells et al. 2004) Instriatal delivery the neurotrophins protected against striatal neuron death in the this model, as well as protecting NOS-positive interneurons. In another series of experiments, adenoviral vector delivery of BDNF promoted recovery of spinal cord axons following spinal cord transection (Koda et al. 2004). The regeneration of rubrospinal axons was also accompanied by significant locomotor function in the hindlimbs of the BDNF treated rats. Lentiviral expression of BDNF into multiple animal models of Alzheimer’s Disease provided significant neuroprotection, even when administered after disease onset (Nagahara et al. 2009). Lenti-BDNF reversed synaptic loss and cognitive decline, improved agerelated changes in gene expression and restored cell signaling. Another example of virally delivered BDNF was provided with Sendai virus-delivered BDNF, which when injected into subcortical white matter, induced BDNF levels and ameliorated memory deficits and synaptic degeneration in a mouse model of AD (Iwasaki et al. 2012).
Cell-Based Delivery of BDNF Delivery of BDNF to the CNS has also been successfully achieved via stem cell delivery of the neurotrophin. In one study, bone marrow stromal cells expressing BDNF were utilized to treat rats with transected spinal cord injuries. (Lu et al. 2005a, b). There was significant expression of the BDNF protein and regrowth of sensory and motor axons into the
J Neuroimmune Pharmacol (2014) 9:182–194
lesion site, but functional recovery was not observed. Bone marrow stem cells (BMSC) engineered to express BDNF were also used in a more model of experimental autoimmune encephalomyelitis (Makar et al. 2009). Mice receiving the BDNF expressing BMSC displayed delayed EAE symptom onset and reduced clinical severity score. They also had decreased demyelination and increased remyelination, along with inhibition of brain levels of TNF-α and IFN-γ. In aged triple transgenic mice (3xTg-AD) expressing pathogenic forms of amyloid precursor protein, presenilin, and tau, hippocampal neural stem cell transplantation rescues the spatial learning and memory deficits in aged 3xTg-AD mice. (Blurton-Jones et al. 2009). Interestingly, cognitive function was improved without altering A or tau pathology. Instead, the improved cognition involves BDNF-mediated enhancement of hippocampal synaptic density. Additionally, BDNF secreting human mesenchymal stem cells have also been transplanted into the substantia nigra of 6-OHDA lesioned rats, resulting in increased nigral and striatal tyrosine hydroxylase, accompanied by attenuation of amphetamine induced motor activity. (Somoza et al. 2010). In addition to stem cells, BDNF has also been successfully delivered from fibroblasts expressing BDNF. Levivier and colleagues (Levivier et al. 1995) utilized the intrastriatal 6hydroxydopamine (6-OHDA) model of PD, and treated the animals with intrastriatal grafts of BDNF producing fibroblasts. They found that the BDNF-producing fibroblasts prevented loss of nigral neuronal cell bodies and partially blocked loss of striatal nerve terminals, compared to control fibroblast transplants. Similar neuroprotective observations were made in a MPP+ lesioned rat following intranigral implantation of BDNF-producing fibroblasts (Frim et al. 1994).
Polymer Systems to Encapsulate, Deliver and Release BDNF Another approach to deliver the BDNF is to modify the protein, or encapsulate it in polymers to deliver it to the target tissue. One method utilized was to covalently add polyethyleneglycol polymers to the carboxyl residues of glutamate and aspartate. (Sakane and Pardridge 1997) PEGylation of BDNF with PEG2000 and PEG5000 significantly reduced the systemic clearance of the BDNF-PEG following Intravenous administration, with only minimal loss of BDNF. In order to enhance the delivery of BDNF to the spinal cord, PEGylated BDNF was administered intrathecally, resulting in improved half life of the BDNF in CSF and enhanced penetration into the spinal cord tissue (Soderquist et al. 2009). N-terminal PEGylated BDNF was administered after spinal cord injury and resulted in axonal sprouting and some locomotor recovery (Ankeny et al. 2001), although these
185
improvements were similar to high doses of native BDNF treatment. Synthetic polymeric systems to encapsulate and then release BDNF have been devised and successfully employed. Bertram et al. (Bertram et al. 2010) synthesized polymers based on poly (lactic co-glycolic) acid (PLGA), poly Llysine and PEG to load and deliver BDNF. They found that a polymer of PLGA, PLL, and PEG led to the delivery of BDNF for periods of time greater than 60 days and that the BDNF delivered was bioactive. Other naturally occurring polymers, like those containing collagen, have been applied to spinal cord injury (SCI) in rats (Han et al. 2009). They developed a linear ordered collagen scaffold to generate a collagen binding domain-BDNF system for use in a rat hemisection model of SCI. The collagen-BDNF system improved neuron survival and recovery from SCI recovery. Another natural polymer, hyaluronic acid based hydrogel containing a matrix metalloproteinase peptide, IKVAV peptide derived from laminin and BDNF, with human mesenchymal stem cells was engineered for treatment of spinal cord injury (Park et al. 2010). Intrathecal injection of rats with these hydrogels containing BDNF showed the greatest improvement on locomotor tests. It was hypothesized that the hyaluronic acid-based hydrogels containing IKVAV and BDNF created microenvironments that facilitated differentiation of stem cells along the neural cell lineage, resulting in a regenerative response.
BDNF Peptidomimetics Numerous studies have demonstrated the utility of the BDNF protein in in vitro and in vivo models of neurodegeneration. (Allen et al. 2013; Lu et al. 2013) and references therein) However, the utility of BDNF is hampered by its lack of CNS penetration and distribution. The idea to derive low molecular weight peptides that retain the neurotrophic actions of BDNF but have improved pharmacokinetic parameters and brain penetration was developed. (O’Leary and Hughes 1998) From a homology model of the BDNF homodimer, O’Leary and Hughes designed small peptides that mimicked loop 2 of BDNF and found active amino acid residues linked to the activation of TrkB in a monocyclic 10 amino acid peptide. In subsequent work, they used that 10mer to create a series of bicyclic dimeric peptides with pairs of 10 amino acid peptides (O’Leary and Hughes 2003). The neuroprotective optimization process led to a tricyclic dimeric peptide that was very potent in promoting survival of chick sensory ganglion neurons, with an Ec50 of 11pM, only about 2 fold less active than BDNF itself. Additional BDNF peptidomimetics were synthesized from loops 1, 2 and 4 of BDNF as peptide monomers and dimers, with many of these compounds displaying nanomolar potency at promoting sensory neuronal survival
186
J Neuroimmune Pharmacol (2014) 9:182–194
Table 1 Inducers of BDNF Class
System(s)
References
Antidepressants fluoxetine Sertraline
rats in vivo Human patients
(Dwivedi et al. 2006; Molteni et al. 2006)l (Nibuya et al. 1995; Peng et al. 2008b; Matrisciano et al. 2009) (Cattaneo et al. 2010)
escitalopram
human patients
Paroxetine Imipramine Desimipramine
human astrocyte-glioma cells in vitro in vitro NSCs mice in vivo
Nortriptyline Amitriptyline
mice in vivo mice in vivo, human patients
(Martinez-Turrillas et al. 2005; Angelucci et al. 2011) (Peng et al. 2008a) (Nibuya et al. 1995; Jacobsen and Mork 2004; Dwivedi et al. 2006) (Cattaneo et al. 2010) (Xu et al. 2003; Lee et al. 2010)
rat cortical neurons in vitro hippo in vitro; rats in vivo mice in vivo mice in vivo rats in vivo rats in vivo
(Yagasaki et al. 2006) (Cooke et al. 2009) (Calabrese et al. 2007; Mannari et al. 2008) (Ikenouchi-Sugita et al. 2009) (Rogoz et al. 2005a) (Nibuya et al. 1995)
Dosed with sertraline-human patients rats in vivo rats in vivo
(Yoshimura et al. 2008) (Czubak et al. 2009) (Bai et al. 2003)
mouse astrocytes in vitro rat PC12 cells in vitro In vitro mouse neurons, rat cortical cells, mice in vivo rats in vivo
(Mizuta et al. 2000) (Mizuta et al. 2000) (Youdim et al. 2005) (Kupershmidt et al. 2009; Avramovich-Tirosh et al. 2010; Youdim 2012) (Coppell et al. 2003)
human patients, mice in vivo mice in vivo
(Leyhe et al. 2008; Autio et al. 2011) (Autio et al. 2011; Gil-Bea et al. 2011)
mice in vivo mice in vivo
(Wang et al. 2006) (Francis et al. 2012)
rat astrocytes, DA neurons in vitro, rats in vivo rat cortical neurons, rats in vivo mice in vivo DA neurons in vitro mice in vivo rats in vivo
(Rahvar et al. 2011; Zhang et al. 2012a, 2012b)
Fluvoxamine Venlafaxine Duloxetine Milnacipran Mirtazapine Mianserin Antipsychotics Risperidone Olanzapine Clozapine MAO inhibitors Selegiline Rasagiline TVP1022 M30 Tranylcypromine AChE inhibitors Donepezil Galantamine Huperzine A Rivastigmine Antioxidants Resveratrol Curcumin J147 Caffeic Acid Phenethyl ester Ferulic acid Ginko biloba EGb761 Activities Voluntary Exercise Intermittant Fasting Mood stabilizers Ampakines Lithium Valproate
mice in vivo, human patients mice and rats in vivo rat cortical and hippocampal slices in vitro cortical neurons in vitro; mice and rats in vivo DA neurons in vitro
(Wu et al. 2006; Wang et al. 2008) (Prior et al. 2013) (Kurauchi et al. 2012) (Sultana et al. 2005; Yabe et al. 2010) (Zhang et al. 2012d) (Adlard et al. 2004, 2005; Rasmussen et al. 2009; Lee et al. 2013) (Lee et al. 2002; Duan et al. 2003) (Mackowiak et al. 2002; Lauterborn et al. 2003, 2009) (Hashimoto et al. 2002; Jacobsen and Mork 2004; Young 2009; Chiu et al. 2013) (Chen et al. 2006; Wu et al. 2008b)
J Neuroimmune Pharmacol (2014) 9:182–194
187
Table 1 (continued) Class
System(s)
References
carbamazepam Lamotrigine Dopamine agonists
rats in vivo rats in vivo
(Chang et al. 2009) (Chang et al. 2009)
Pramipexole Ropinerole NMDA receptor antagonists MK801 Ketamine Memantine Flavonoids Anthocyanins Quercetin/kaempferol Hyperoside Green Tea Catechins L-theanine EGCG Miscellaneous inducers Simastatin Cystamine/Cysteamine Butaprost Deltamethrin
DA neurons in vitro DA neurons in vitro
(Du et al. 2005; Lauterbach 2012) Du et al. 2005; Lauterbach 2012)
mice in vivo mice in vivo rats in vivo
(Autry et al. 2011) (Autry et al. 2011) (Reus et al. 2012)
rats in vivo mice and rats in vivo rat PC12 cells in vitro mice in vitro, rats in vivo SH-SY-5Y cells in vitro, rats in vivo hippocampal and cortical neurons in vitro
(Rendeiro et al. 2013) (Hou et al. 2010; Yao et al. 2012) (Zheng et al. 2012) (Li et al. 2009; Rendeiro et al. 2013) (Cho et al. 2008; Takeda et al. 2011) (Nath et al. 2012)
rats in vivo mice in vivo human microglia and astrocytes in vitro rat cortical cultures in vitro
(Wu et al. 2008a) (Borrell-Pages et al. 2006) (Hutchinson et al. 2009) (Imamura et al. 2006; Ihara et al. 2012) (Visanji et al. 2008; Zhang et al. 2012c) (Mizuta et al. 2000; Katoh-Semba et al. 2002)
Agomelatine Nicotine Metyrapone Rolipram Catalpol Fingolimod
rat DA neurons in vitro; mice in vivo mouse cultured astrocytes, hippocampal neurons rats in vivo mice in vivo rats in vivo, synergy with imipramine rats in vivo, synergistic with desimipramine rat cortical neurons in vitro, mouse in vivo cortical neurons in vitro, mice in vivo
(Ladurelle et al. 2012) (Czubak et al. 2009) (Rogoz et al. 2005b; Rogoz and Legutko 2005) (Fujimaki et al. 2000) (Wang et al. 2009) (Deogracias et al. 2012; Doi et al. 2013)
CEP 1347 BW373U86
mice in vivo rats in vivo
(Apostol et al. 2008; Conforti et al. 2008) (Torregrossa et al. 2004)
PYM50028 (Cogane) Riluzole
(Fletcher and Hughes 2006). Further optimization of these peptides would need to include decreased size, improved solubility, membrane permeability and resistance to proteolytic degradation. Fletcher and colleagues (Fletcher et al. 2008) then designed a BDNF loop 4 mimetic, a proteolytically stable cyclic pentapeptide cyclo[dPAKKR] with these improved pharmaceutic properties, as well as an effective survival factor for sensory ganglion neurons. Interestingly, these functional BDNF mimetics work independently of TrkB and downstream MAPK signaling. They have optimized these cyclic pentapeptides in a second generation of compounds that showed improved biological activity, stability in plasma, and an ability to cross model biological membranes. (Fletcher and Hughes 2009). A most recent study used tetrapeptides synthesized to 5 different regions of the BDNF protein,
and found neurotrophic effects on hippocampal neuronal survival. One of the compounds, B3 potentiated the neuronal survival promoting activity of BDNF (Cardenas-Aguayo Mdel et al. 2013).
Small Molecule Mimetics of BDNF (TrkB Agonists) To overcome issues involved in delivering the protein as a therapeutic, small molecule mimetics of BDNF, which had superior biopharmaceutic properties including pharmacokinetic potential to reach the CNS, were sought and identified. Massa and Longo identified numerous structures from in silico screening of the loop II region of BDNF and by low throughput screening of hippocampal neurons (Massa et al. 2010) These four prototypic compounds were neuroprotective
188
J Neuroimmune Pharmacol (2014) 9:182–194
in a concentration dependent manner with high picomolar activity, selectively activated TrkB with high potency and specificity and promoted neuronal survival against multiple neurotoxins, differentiation, and synaptic function. When LM-22A-4 was administered in vivo, it caused hippocampal and striatal TrkB activation in mice and improved motor learning after traumatic brain injury in rats. (Massa et al. 2010).
Small Molecule Inducers of BDNF Use of small molecule drugs to upregulate BDNF mRNA and increase cellular BDNF levels in the target tissue is another approach to elicit a neuroprotective effect. This approach utilizes FDA approved drugs and other small molecules with good systemic bioavailability, chemical and metabolic stability, membrane permeability and CNS penetration through the blood brain barrier, improved biopharmaceutic properties to induce cellular BDNF levels. A number of agents have been
reported to elicit increased BDNF protein, including antidepressants, antipsychotics, MAO inhibitors, AChE inhibitors, antioxidants, mood stabilizers, dopamine agonists, NMDA receptor antagonists, flavonoids and additional miscellaneous compounds. This list of compounds is presented in Table 1. Of the compounds listed in Table 1, only riluzole has been approved for its neuroprotective effects in treatment of ALS. While the use of small molecules has the advantage of ease of delivery and repurposing of FDA approved drugs has the advantage of quick translation to clinical trials, off target effects may be of concern particularly if the dosage needed is greater than it’s approved dose range.
Application of BDNF and Neurotrophin Therapeutics to HIVAssociated Neurocognitive Disorders HIV infection into the CNS results in synaptodendritic damage in neurons, and these synaptodendritic degenerative
Pro-BDNF BDNF Tat+ 1 EGCG Tat+0.1 EGCG
Tat+1Epi
Tat+0.1Epi Tat
Media
Tat+ 1 EGCG Tat+0.1 EGCG
Tat+1Epi Tat+0.1Epi Tat Media
Fig. 1 Catechins reverse Tat mediated changes in BDNF and proBDNF expression in mixed rat neuronal cultures. Rat mixed cortical cultures (106 cells per well on six-well plates) were treated with culture media plus 0.1 % DMSO vehicle, 500 nM Tat plus vehicle, or Tat plus 0.1 or 1 μM epicatechin or Tat plus 0.1 or 1 μM EGCG for 24 h. After 24 h, cell lysates were generated, proteins separated bySDS–polyacrylamide gel electrophoresis, transferred to PVDF membranes and probedwith antibodies to BDNF. The BDNF antibody recognized the precursor of BDNF
(proBDNF) at 32 kDa as well as the mature BDNF protein at 13 kDa. Two replicate experiments are depicted. The data are expressed as percent expression of the media/vehicle treated controls The data from three independent replicates were utilized for each treatment group, and were evaluated by ANOVA for significance. Group-wise post hoc comparisons were assessed by Newman–Keuls multiple comparison test. *p
