JOURNAL OF TISSUE ENGINEERING AND REGENERATIVE MEDICINE J Tissue Eng Regen Med (2015) Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/term.2052
REVIEW
Regenerative therapy for hippocampal degenerative diseases: lessons from preclinical studies Chaitra Venugopal†, Shashank Chandanala†, Harish Chandra Prasad, Danish Nayeem, Ramesh R. Bhonde and Anandh Dhanushkodi* School of Regenerative Medicine, Manipal University, Bangalore, India
Abstract Increase in life expectancy has put neurodegenerative diseases on the rise. Amongst these, degenerative diseases involving hippocampus like Alzheimer’s disease (AD) and temporal lobe epilepsy (TLE) are ranked higher as it is vulnerable to excitotoxicity induced neuronal dysfunction and death resulting in cognitive impairment. Modern medicines have not succeeded in halting the progression of these diseases rendering them incurable and often fatal. Under such scenario, regenerative studies employing stem cells or their by-products in animal models of AD and TLE have yielded encourageing results. This review focuses on the distinct cell types, such as hippocampal cell lines, neural precursor cells, embryonic stem cells derived neural precursor cells, induced pluripotent stem cells, induced neurons and mesenchymal stem cells, which can be employed to rescue hippocampal functions in neurodegenerative diseases like AD and TLE. Besides, the divergent mechanisms through which cell based therapy confer neuroprotection, current impediments and possible improvements in stem cell transplantation strategies are discussed. Authors are aware of the voluminous literature available on this issue and have made a sincere attempt to put forth the current status of research in the field of cell based therapy concurrently discussing the promise it holds for combating neurodegenerative diseases like AD and TLE in the near future. Copyright © 2015 John Wiley & Sons, Ltd. Received 20 November 2014; Revised 8 April 2015; Accepted 29 April 2015
Keywords hippocampal degeneration; mesenchymal stem cells; conditioned medium; neurogenesis; neural progenitors; neural stem cell niche
1. Introduction Since the seminal study by Scoville and Milner (1957) revealing that damage to the hippocampal formation causes severe impairment of declarative memory in humans, a substantial volume of literature has been accumulated in an effort to define the specific role of the hippocampus in mammalian behaviour. In rodents, damage to the hippocampal formation produces striking impairment in both acquiring and recalling spatial information (Best et al., 2001; Dhanushkodi et al., 2007; Eichenbaum et al., 1999; *Correspondence to: A. Dhanushkodi, School of Regenerative Medicine, GKVK Post, Bellary Road, Allalasandra, Yelahanka, Bangalore 560065, India. E-mail: [email protected]; [email protected] † These authors contributed equally to this study.
Copyright © 2015 John Wiley & Sons, Ltd.
O’Keefe and Dostrovsky, 1971; Rekha et al., 2009). Further interest in the hippocampus arose from the revelation of hippocampal atrophy in neurodegenerative diseases such as temporal lobe epilepsy (TLE) (Dhanushkodi and Shetty, 2008) and Alzheimer’s disease (AD) and cognitive impairments associated with these diseases (Dhanushkodi and McDonald, 2011; Dhanushkodi and Shetty, 2008; Flanigan et al., 2014). Neurodegenerative diseases are devastating ones in which neurons undergo progressive degeneration in several brain regions. The excessive release of excitatory neurotransmitter glutamate and subsequent hyperactivation of glutamate receptors is one of the underlying causes of neuronal damage in many neurodegenerative diseases (Doble, 1995; Dong et al., 2009). The currently prescribed drugs merely provide symptomatic relief and do not influence the course of these diseases. Besides, these drugs cause severe side-effects ranging from
C. Venugopal et al.
psychological to cognitive to motor disturbances. Cellbased therapy is emerging as a potential therapeutic alternative for treating various CNS diseases (Acharya et al., 2009; Chen et al., 2014; Garcia et al., 2014; Liu et al., 2014; Rosenblum et al., 2014; Srivastava et al., 2009; Tfilin et al., 2010; Tsupykov et al., 2014). Although in neurodegenerative diseases, such as amyotrophic lateral sclerosis and Parkinson’s disease, hippocampal atrophy and memory impairments have been reported at the very advanced stages, early degeneration in the hippocampus is considered to be the principal cause of neurodegenerative diseases such as AD and TLE. Accordingly, in this review we focus on the recent advances made in the field of cell therapy for hippocampal neurodegenerative conditions such as TLE and AD, with emphasis on possible mechanisms of cellular and functional recovery following cell transplantation. We also discuss the challenges to be tackled in order to successfully translate the benefits of cell therapy from bench to bedside.
2. Hippocampus: structure and function The hippocampus is a fundamental constituent of the limbic system in the brain. The basic circuitry of the hippocampus has been well appreciated since the time of Ramon Cajal (Baratas Diaz, 1997). The hippocampal formation is not a unitary structure, but instead comprises numerous subdivisions, such as the dentate gyrus (DG), hippocampus proper (CA1–CA4), entorhinal cortex (EC) and subiculum (Figure 1A). These structures are intimately connected to each other by the well-known trisynaptic pathways (Figure 1B). The entorhinal cortex receives highly processed information from large areas of primary and association cortices and acts as a major input component to the hippocampus proper. The entorhinal cortical neurons from layers I–III project towards the DG via a perforant pathway, while the mossy fibres originating from DG cells project to the CA3 subfield of the hippocampus. CA1 neurons receive synaptic connections from CA3 neurons through Schaffer’s collaterals. The subiculum is the primary target of CA1 pyramidal cells, projecting back to the deeper layers of the entorhinal cortex and thus completing the circuit (Amaral and Witter, 1989; Witter et al., 1989). The hippocampus plays an important role in consolidating short-term memories into long-term memories, and is well known for its ability to form cognitive maps, a neural representation of space and position (O’Keefe and Dostrovsky, 1971). Hippocampal neurons are highly vulnerable to neuronal alterations associated with ageing, stress (Esch et al., 2002), depression (Warner-Schmidt and Duman, 2006) and neurodegenerative diseases such as TLE (Dhanushkodi and McDonald, 2011; Dhanushkodi and Shetty, 2008) and AD (Fukutani et al., 2000), in which cognitive impairment is one of the prominent clinical manifestations. In the next two sections, we summarize the hippocampal pathology in TLE and AD. Copyright © 2015 John Wiley & Sons, Ltd.
Figure 1. (A) Coronal section of mouse brain, indicating various subfields of the hippocampus. (B) Schematic diagram illustrating the intrinsic neural circuitry of the hippocampus proper: SUB, subiculum; DG, dentate gyrus; DH, dentate hilus; CA1, CA3, cornu ammonis 1 and 3; PP, perforant pathway; MF, mossy fibres; SC, Schaffer collaterals
3. Hippocampal pathology in temporal lobe epilepsy Epilepsy affects nearly 65 million people worldwide. Among different forms of epilepsy, TLE is the most common and well-characterized. As the name implies, TLE chiefly originates in the temporal lobe region of the brain and, more specifically, in the hippocampus. Approximately 40% of patients with epilepsy are considered to have TLE (Engel, 2001; McKeown and McNamara, 2001). Archetypally, a well-defined sequela is noticed in TLE, that an initial precipitating injury (IPI), due to acute seizures, stroke, head trauma or high fever, is followed by a silent/latent period with absence of seizures and then the genesis of spontaneous recurrent motor seizures that persist for the rest of the patient’s life (Acharya et al., 2008; Dhanushkodi and Shetty, 2008). Various animal models that were developed by electrical stimulation, by chemo-convulsion drugs such as pilocarpine/kainic acid, or by gene manipulation, provided valuable insight into the neurobiology of epileptogenesis, i.e. from IPI to the development of chronic epilepsy. The pathological features in animal models following IPI, such as hippocampal sclerosis (Sloviter, 1999), neuroinflammation (Oprica et al., 2003), loss of hippocampal interneurons (Kuruba et al., 2011) and abnormal neural plasticity (mossy fibres sprouting in the CA3 subfield and abnormal neurogenesis in the dentate gyrus) (Rao et al., 2006; Parent et al., 2006), closely mimic the clinical manifestation of human TLE. As epileptogenesis progresses, hippocampal J Tissue Eng Regen Med (2015) DOI: 10.1002/term
Cell-based therapy for hippocampal degenerative diseases
neurogenesis declines drastically (Hattiangady et al., 2004) and is manifested as severe depression and cognitive impairment. Nearly 30% of patients with TLE are refractory to anti-epileptic drugs. Besides patients, those who respond to anti-epileptic drugs display severe sideeffects, including cognitive impairments and motor disturbances. Thus, identifying alternative therapeutic strategies, such as stem cell therapy, to curtail TLE is highly justified. In this context, animal models of TLE are valuable prototypes to explore the possibilities of stem cell therapy for TLE and to understand the cellular–biochemical–molecular changes that take place following cell transplantation.
4. Hippocampal pathology in Alzheimer’s disease In 1907, Dr Alois Alzheimer first reported the clinical manifestations of a new neurodegenerative disease called Alzheimer’s disease (AD), a progressive neurodegenerative disease clinically characterized by decline in memory and cognitive abilities (Alzheimer.,1907). It is the most common cause of irreversible dementia, initially manifesting as impairment in episodic memory that gradually progresses towards a more crippling global decline in cognitive functions, incapacitating the individual and ultimately culminating in death. With a global increase in life expectancy and a rise in the ageing population, it is estimated that the incidence of AD will quadruple in the next 40 years, positioning it on the verge of attaining a pandemic status and concurrently posing an ever-increasing financial and emotional burden on patients, caregivers and society. Two prominent neuropathological features of AD are the appearances of amyloid plaques and neurofibrillary tangles. Amyloid plaques, also called neuritic plaques, are extracellular deposits of aggregated amyloid-β (Aβ) peptides, 40–42 amino acids-long proteolytic derivatives of the parent transmembrane protein, amyloid precursor protein (APP). On the other hand, neurofibrillary tangles are intracellular aggregates, often present in neuronal cell bodies and dendrites, containing hyperphosphorylated microtubule-associated TAU protein as the principal proteinaceous component. Apart from abnormal deposits of protein aggregates, the brains of AD patients and AD transgenic mice show evidence of oxidative damage (Tayler et al., 2010) and neuroinflammation (Graeber, 1999; McGeer et al., 2000), representative of their participation in the initiation and progression of AD pathogenesis. Broadly, two forms of AD exist, familial AD (FAD) and sporadic AD (SAD). Although the aetiology of SAD is unknown, several risk factors, such as ageing, diabetes, cardiovascular disease, apolipoprotein allele status and environmental toxins such as aluminium, have been suggested. On the other hand, genetic mutations in APP, presenilin-1 and presenilin-2 have been proposed as a primary cause for FAD. As the cholinergic neurons in the nucleus basalis of Meynert (NBM) are most Copyright © 2015 John Wiley & Sons, Ltd.
vulnerable in AD, initial attempts were made to mimic the animal prototype of AD by specifically destroying cholinergic neurons in the NBM (Fine et al., 1985; Emerich et al., 1992). Such animal models demonstrated profound cognitive impairments resembling human AD, albeit lacking amyloid plaques and neurofibrillary tangles. Later on, the neuropathological features of FAD were successfully mimicked in various transgenic animal models of AD by manipulating three crucial genes involved in FAD, i.e. APP, presenilin-1 and presenilin-2 (Games et al., 1995; Hsiao et al., 1996; Holcomb et al., 1998). Transgenic animal models of AD mimic the classical neuropathological hallmarks associated with human AD, such as abnormal accumulation of extracellular amyloid plaques and intraneuronal neurofibrillary tangles in the hippocampus and neocortex at the early stage of the disease and, as the disease progress, a widespread accumulation of amyloid plaques and neurofibrillary tangles recognizable in association cortices and subcortical structures such as the thalamus and amygdala. A salient feature of memory impairment associated with AD dementia, suggestive of hippocampal degeneration, is the rapid loss of recently acquired information, coupled with difficulty in consolidating new memories, with absolute preservation of the ability to recollect past memories. Similar to human AD, transgenic animal models of AD develop cognitive deficits, and thus serve as valuable tools to decipher the neurobiology of AD and also to explore possible therapeutic strategies (Kobayashi and Chen, 2005; Chen et al., 2000). Although the majority of the characteristic features of AD are imitated in transgenic animal models, not all of them demonstrate neuronal losses and enlargement of the ventricles, as was noticed in human AD.
5. Cell-based therapies for combating neurodegenerative diseases Neurodegenerative diseases are complex pathophysiological conditions in which the existing drugs in practice provide only symptomatic relief, without mitigating the course of the disease. Thus, there is a pressing need to identify alternative therapeutic strategies to arrest the progression of neurodegeneration. Cell therapy is emerging as a potential therapeutic modality against neurodegenerative diseases (Li et al., 2009a, 2014; Maisano et al., 2009; Rekha et al., 2009). Two possible avenues through which functional recovery could be achieved include stimulation, followed by mobilization of endogenous stem cells towards the vicinity of degenerating sites (Diederich et al., 2009; Li et al., 2012; Sehara et al., 2007), or transplanting exogenous neural cells that can proliferate and replace the lost neurons (Maisano et al., 2009; Rekha et al., 2009, 2011). The concept of adult neurogenesis is now widely acknowledged and is known to occur in two brain regions, the subventricular zone (SVZ) and the subgranular zone (SGZ) of the DG (Deng et al., 2009; Li et al., 2009b; Suh et al., 2009; Thuret et al., 2009; van Praag et al., 1999). J Tissue Eng Regen Med (2015) DOI: 10.1002/term
C. Venugopal et al.
Despite hippocampal neurogenesis being a continuous process, the rate at which it generates new neurons in the SGZ is not sufficient to compensate for the extensive neuronal losses observed in many hippocampal neurodegenerative conditions. A promising approach for replacing the lost neurons would be to transplant neural cells at the site of hippocampal injury (Acharya et al., 2009; Rekha et al., 2009). In the following sections, we review several preclinical investigations that provide evidence of cellular and functional recovery following local or systemic transplantation of neural precursor/adult stem cells in hippocampus-associated neurodegenerative conditions. For the cell transplantation approach, several sources of neuronal cells, such as conditionally immortalized hippocampal cells (Figure 2A), fetal neural precursor cells, embryonic stem cell-derived neural precursor cells (Figure 2B, C), induced pluripotent stem cells, terminally differentiated neurons (induced neurons) and adult stem cells such as mesenchymal stem/stromal cells (Figure 2D, E), have been investigated and will be discussed in the following sections.
6. Hippocampal cell line as a source for cell therapy Initial studies involving the transplantation of immortalized hippocampal cells in experimentally induced hippocampal neurodegenerative models provided early proofof-principle that transplanted cells could assist in
reconstructing the hippocampal circuitry and promoting behavioural recovery (Gray et al., 2000; Hodges et al., 2000; Martinez-Serrano et al., 1995; Sinden et al., 1997; Virley et al., 1999). We have demonstrated that in a rat model of subicular damage, an output component of the hippocampus that is affected in the very early stage of AD, transplantation of a hippocampal cell line derived from postnatal day 1 H-2Kb ts A58 transgenic mice into the hippocampal CA1 subfield exhibited greater migration through the corpus callosum and subsequent integration into the CA1, CA3 and DG subfields of the hippocampal formation. Furthermore, the transplanted cells prevented retrograde and anterograde neurodegeneration by releasing various growth factors, thus contributing to improved spatial learning and memory performance (Rekha et al., 2009). Although conditionally immortalized neural cells provide the benefits of off-the-shelf availability of neural cells for transplantation purposes, the integration of viral genes/oncogenes to transform neural cells into immortal cells put patients at higher risk of unwanted side-effects, such as teratoma formation, following transplantation.
7. Fetal neural precursor cells as a source for cell therapy Neural precursor cells (NPCs) freshly derived from fetus hippocampus are an excellent source of donor cells that can be used to repopulate the degenerated regions of
Figure 2. Potential sources of cells for cell therapy. (A) Conditionally immortalized hippocampal cell line. (B) In vitro embryoid body formation from human embryonic stem cells. (C) Neural stem cells derived from embryoid body. (D) Mesenchymal stem cells derived from human bone marrow. (E) Mesenchymal stem cells derived from human dental pulp. (F) Embryonic stem cells are characterized by their self-renewal and pluripotent properties to give rise to three germ cell types Copyright © 2015 John Wiley & Sons, Ltd.
J Tissue Eng Regen Med (2015) DOI: 10.1002/term
Cell-based therapy for hippocampal degenerative diseases
the hippocampus in TLE or AD. In support of this notion, NPCs obtained from the fetus hippocampus of transgenic mice [129-Gt (ROSA) 26Sortm1 (EGFP) Luo/J] encoding green fluorescent protein, when engrafted into the hippocampus of aged rats following kainic acid-mediated hippocampal injury resulted in the differentiation of 3–5% of engrafted cells into neurons, 28% into astrocytes and 6– 10% into oligodendrocytes (Shetty et al., 2008). Likewise, NPCs derived from the anterior subventricular zone of postnatal F344 rat pups expressing placental alkaline phosphatase, transplanted into kainic acid-mediated hippocampus-damaged adult F344 rats, displayed excellent survival and migration and differentiated into diverse neuronal subtypes, such as inhibitory interneurons, astrocytes, oligodendrocytes and oligodendrocyte progenitors. The grafted cells also expressed various neurotropic factors, such as glial-derived nerve growth factor (GDNF), brain-derived nerve growth factor (BDNF), fibroblast growth factor (FGF) and vascular endothelial growth factor (VEGF), which are well known for their mitogenic and regeneration capacities (Hattiangady and Shetty, 2012). In the rat model of TLE, transplantation of fetus GABAergic neurons into the hippocampus or substantia nigra restrained seizure development through restoration of GABA levels (Hattiangady et al., 2008; Loscher et al., 1998; Waldau et al., 2010). A recent study by Hunt et al. (2013) demonstrated that the medial ganglionic eminence progenitor cells derived from fetuses of green fluorescent protein-positive mice, upon transplantation into the hippocampus of epileptic mice, significantly suppressed seizure intensity and restored spatial learning and memory functions and the reduced aggressive and hyperactive behaviours that are the most common comorbidities observed in TLE subjects. Post-mortem analysis of brain samples from recipient mice revealed that the majority of the grafted cells were differentiated into functional GABAergic interneurons with prominent afferent and efferent connectivity. Given the complexity of AD, in which the pathology is widespread throughout the brain in advanced stages of the disease, cell therapy at the early stage of the disease could be beneficial in mitigating disease progression. As cholinergic transmission is compromised in AD, initial cell transplantation studies in animal models of AD, with lesions to cholinergic neurons in the NBM, demonstrated that transplantation of fetal cholinergic neurons derived from the septum or diagonal band of Broca into the neocortex improved acetyl choline levels and passive avoidance memory, whereas fetal hippocampal cells (noncholinergic) transplantation failed to restore cognitive function in this animal model (Fine et al., 1985). Consistent with this observation, another study by Emerich et al. (1992) demonstrated that transplantation of cholinergic neurons derived from fetal septum/diagonal band of Broca into the hippocampus of a rat model of AD improved cholinergic transmission and enhanced spatial learning and memory performance in an eight-arm radical maze task. However, when the complexity of the learning task was increased, the graft-recipient rats were not Copyright © 2015 John Wiley & Sons, Ltd.
able to perform efficiently as compared to naïve rats, suggesting that the cholinergic transplantation-mediated behavioural recovery depends upon the brain region where the graft cells are transplanted and is also limited to the intellectual demands of the testing circumstances. Regardless of the positive outcome observed in preclinical studies using rodent fetal cells for treating animal models of TLE/AD, translation of utilizing human fetus-derived NPCs for treating human TLE/AD is nearly impossible, due to serious ethical concerns involved in obtaining fetal samples and the huge volume of fetal tissues required for subsequent transplantations. In order to circumvent these obstacles, embryonic stem cells were recognized as an inexhaustible source to produce NPCs for cell therapy in neurodegenerative diseases.
8. Embryonic stem cells as a source for cell therapy Embryonic stem cells (ESCs) can be derived from the inner mass at the blastula stage and possess the unique characteristic features of self-renewal ability with simultaneous generation of lineage-specific progenitor cells of three germ layers (Figure 2F). Embryonic stem cells are highly versatile cells that can be properly guided towards the neuronal lineage under in vitro conditions. Neural precursor cells derived from ESCs can be cultured successfully in vitro as neurospheres (Figure 2B) and can be effectively differentiated into specific neuronal phenotypes for transplantation purposes (Vescovi et al., 1993). In a mouse model of TLE, in which hippocampal sclerosis plays a major role in epileptogenesis, NPCs derived from mouse ESCs, upon transplantation into the DG, exhibited enhanced survival and differentiation within the granular cell layer. Furthermore, this study also demonstrated that upregulation of chemoattractants, such as stromalderived factor-α1, as a consequence of hippocampal injury following status epilepticus attracted the transplanted NPCs preferentially towards the vicinity of the lesion (Hartman et al., 2010). As TLE is mainly caused by imbalance in excitatory and inhibitory functions, transplantation of NPCs with ventral forebrain identity derived from a mouse ES cell line, upon transplantation into the hippocampus of epileptic mice, could differentiate into various subtypes of GABAergic neurons with the typical electrophysiological properties of inhibitory neurons (Maisano et al., 2012). Although there have been relatively few studies, ESderived NPCs were also investigated in animal models of AD. Wang et al. (2006) compared the effects of transplantation of pure mouse ES cells or ES cell-derived neurospheres into a mouse model of AD with NBM lesions. Transplantation of pure ES cells into the frontal association cortex and barrel cortex induced extensive teratoma formation with further decline in cognitive function, highlighting the caution needed in using undifferentiated ES cells for cell therapy purposes. On the other J Tissue Eng Regen Med (2015) DOI: 10.1002/term
C. Venugopal et al.
hand, transplantation of neurospheres derived from ES cells differentiated into cholinergic neurons and, to some extent, into serotonergic neurons, with significant improvement in cognitive function. Akin to fetal neural precursor cells, utilizing ES-derived NPCs for clinical practice also presents several hurdles. This is mainly due to serious ethical and political issues associated with obtaining human embryonic tissue, potential teratoma formation and graft-versus-host immune reactions following transplantation.
9. Induced pluripotent stem cells as a source for cell therapy To bypass the ethical/political issues associated with using fetal or embryonic tissue for transplantation purposes, autologous induced pluripotent stem cells (iPSCs) can be used as an alternative cell source. Induced pluripotent stem cells are adult-derived somatic cells that can be reverted to embryonic stage by ectopic expression of transcription factors such as Oct4, Sox2, c-Myc and Klf4 (Takahashi and Yamanaka, 2006). Induced pluripotent stem cells have been extensively used to model human diseases in Petri plates (Okano and Yamanaka, 2014). Nevertheless, our understanding of iPSCs is still in its infancy and the potential of teratoma formation (Takahashi and Yamanaka, 2006) following iPSCs transplantation hamper their use for cytotherapy. Therefore, comprehensive investigations are necessary to evaluate the behaviour of iPSCs following transplantation before they can be used clinically.
10. Terminally differentiated neurons (induced neurons) as a source for cell therapy Contemporary studies highlight the feasibilities of direct conversion of adult somatic fibroblasts or brain astrocytes into terminally differentiated neurons called ’induced neurons’ by integrating neural fate determinant transcriptional factors, such as achaete-scute complex-like 1, brain-2 and myelin transcription factor-like 1 (ABM) (Pfisterer et al., 2011a; Vierbuchen et al., 2010; Addis et al., 2011). This direct conversion procedure bypasses the need for a transient proliferation stage and generates terminally differentiated neurons that do not form teratomas. Consequently, induced neurons provide the added advantages of producing tailor-made, autologous and terminally differentiated and phenotype-specific neurons to treat specific neurodegenerative diseases. In addition to ABM, a cocktail of other transcriptional factors and small molecules could be incorporated to produce specific subtypes of neurons, such as dopaminergic, cholinergic and spinal cord motor neurons (Pfisterer et al., 2011b; Son et al., 2011; Caiazzo et al., 2011; L. Liu et al., 2013). Copyright © 2015 John Wiley & Sons, Ltd.
Although it is not practically conceivable to obtain autologous human astrocytes, emerging evidence suggests that, like somatic fibroblasts, brain astrocytes can also be transdifferentiated into specific subtypes of neurons in in vitro (Addis et al., 2011). Advancing further, Torper et al. (2013) recently demonstrated that, following stereotaxic injection of Cre-inducible lentivirus that carried ABM into the striatum of GFAP-Cre transgenic mice, endogenous striatal astrocytes were efficiently transdifferentiated into neurons, thus emphasizing the likelihood of converting the endogenous terminally differentiated glial cells into mature neurons. While the production of induced neurons opens up the prospect of making patientand disease-specific neurons, the current protocols employed yield only a negligible percentage (
REVIEW
Regenerative therapy for hippocampal degenerative diseases: lessons from preclinical studies Chaitra Venugopal†, Shashank Chandanala†, Harish Chandra Prasad, Danish Nayeem, Ramesh R. Bhonde and Anandh Dhanushkodi* School of Regenerative Medicine, Manipal University, Bangalore, India
Abstract Increase in life expectancy has put neurodegenerative diseases on the rise. Amongst these, degenerative diseases involving hippocampus like Alzheimer’s disease (AD) and temporal lobe epilepsy (TLE) are ranked higher as it is vulnerable to excitotoxicity induced neuronal dysfunction and death resulting in cognitive impairment. Modern medicines have not succeeded in halting the progression of these diseases rendering them incurable and often fatal. Under such scenario, regenerative studies employing stem cells or their by-products in animal models of AD and TLE have yielded encourageing results. This review focuses on the distinct cell types, such as hippocampal cell lines, neural precursor cells, embryonic stem cells derived neural precursor cells, induced pluripotent stem cells, induced neurons and mesenchymal stem cells, which can be employed to rescue hippocampal functions in neurodegenerative diseases like AD and TLE. Besides, the divergent mechanisms through which cell based therapy confer neuroprotection, current impediments and possible improvements in stem cell transplantation strategies are discussed. Authors are aware of the voluminous literature available on this issue and have made a sincere attempt to put forth the current status of research in the field of cell based therapy concurrently discussing the promise it holds for combating neurodegenerative diseases like AD and TLE in the near future. Copyright © 2015 John Wiley & Sons, Ltd. Received 20 November 2014; Revised 8 April 2015; Accepted 29 April 2015
Keywords hippocampal degeneration; mesenchymal stem cells; conditioned medium; neurogenesis; neural progenitors; neural stem cell niche
1. Introduction Since the seminal study by Scoville and Milner (1957) revealing that damage to the hippocampal formation causes severe impairment of declarative memory in humans, a substantial volume of literature has been accumulated in an effort to define the specific role of the hippocampus in mammalian behaviour. In rodents, damage to the hippocampal formation produces striking impairment in both acquiring and recalling spatial information (Best et al., 2001; Dhanushkodi et al., 2007; Eichenbaum et al., 1999; *Correspondence to: A. Dhanushkodi, School of Regenerative Medicine, GKVK Post, Bellary Road, Allalasandra, Yelahanka, Bangalore 560065, India. E-mail: [email protected]; [email protected] † These authors contributed equally to this study.
Copyright © 2015 John Wiley & Sons, Ltd.
O’Keefe and Dostrovsky, 1971; Rekha et al., 2009). Further interest in the hippocampus arose from the revelation of hippocampal atrophy in neurodegenerative diseases such as temporal lobe epilepsy (TLE) (Dhanushkodi and Shetty, 2008) and Alzheimer’s disease (AD) and cognitive impairments associated with these diseases (Dhanushkodi and McDonald, 2011; Dhanushkodi and Shetty, 2008; Flanigan et al., 2014). Neurodegenerative diseases are devastating ones in which neurons undergo progressive degeneration in several brain regions. The excessive release of excitatory neurotransmitter glutamate and subsequent hyperactivation of glutamate receptors is one of the underlying causes of neuronal damage in many neurodegenerative diseases (Doble, 1995; Dong et al., 2009). The currently prescribed drugs merely provide symptomatic relief and do not influence the course of these diseases. Besides, these drugs cause severe side-effects ranging from
C. Venugopal et al.
psychological to cognitive to motor disturbances. Cellbased therapy is emerging as a potential therapeutic alternative for treating various CNS diseases (Acharya et al., 2009; Chen et al., 2014; Garcia et al., 2014; Liu et al., 2014; Rosenblum et al., 2014; Srivastava et al., 2009; Tfilin et al., 2010; Tsupykov et al., 2014). Although in neurodegenerative diseases, such as amyotrophic lateral sclerosis and Parkinson’s disease, hippocampal atrophy and memory impairments have been reported at the very advanced stages, early degeneration in the hippocampus is considered to be the principal cause of neurodegenerative diseases such as AD and TLE. Accordingly, in this review we focus on the recent advances made in the field of cell therapy for hippocampal neurodegenerative conditions such as TLE and AD, with emphasis on possible mechanisms of cellular and functional recovery following cell transplantation. We also discuss the challenges to be tackled in order to successfully translate the benefits of cell therapy from bench to bedside.
2. Hippocampus: structure and function The hippocampus is a fundamental constituent of the limbic system in the brain. The basic circuitry of the hippocampus has been well appreciated since the time of Ramon Cajal (Baratas Diaz, 1997). The hippocampal formation is not a unitary structure, but instead comprises numerous subdivisions, such as the dentate gyrus (DG), hippocampus proper (CA1–CA4), entorhinal cortex (EC) and subiculum (Figure 1A). These structures are intimately connected to each other by the well-known trisynaptic pathways (Figure 1B). The entorhinal cortex receives highly processed information from large areas of primary and association cortices and acts as a major input component to the hippocampus proper. The entorhinal cortical neurons from layers I–III project towards the DG via a perforant pathway, while the mossy fibres originating from DG cells project to the CA3 subfield of the hippocampus. CA1 neurons receive synaptic connections from CA3 neurons through Schaffer’s collaterals. The subiculum is the primary target of CA1 pyramidal cells, projecting back to the deeper layers of the entorhinal cortex and thus completing the circuit (Amaral and Witter, 1989; Witter et al., 1989). The hippocampus plays an important role in consolidating short-term memories into long-term memories, and is well known for its ability to form cognitive maps, a neural representation of space and position (O’Keefe and Dostrovsky, 1971). Hippocampal neurons are highly vulnerable to neuronal alterations associated with ageing, stress (Esch et al., 2002), depression (Warner-Schmidt and Duman, 2006) and neurodegenerative diseases such as TLE (Dhanushkodi and McDonald, 2011; Dhanushkodi and Shetty, 2008) and AD (Fukutani et al., 2000), in which cognitive impairment is one of the prominent clinical manifestations. In the next two sections, we summarize the hippocampal pathology in TLE and AD. Copyright © 2015 John Wiley & Sons, Ltd.
Figure 1. (A) Coronal section of mouse brain, indicating various subfields of the hippocampus. (B) Schematic diagram illustrating the intrinsic neural circuitry of the hippocampus proper: SUB, subiculum; DG, dentate gyrus; DH, dentate hilus; CA1, CA3, cornu ammonis 1 and 3; PP, perforant pathway; MF, mossy fibres; SC, Schaffer collaterals
3. Hippocampal pathology in temporal lobe epilepsy Epilepsy affects nearly 65 million people worldwide. Among different forms of epilepsy, TLE is the most common and well-characterized. As the name implies, TLE chiefly originates in the temporal lobe region of the brain and, more specifically, in the hippocampus. Approximately 40% of patients with epilepsy are considered to have TLE (Engel, 2001; McKeown and McNamara, 2001). Archetypally, a well-defined sequela is noticed in TLE, that an initial precipitating injury (IPI), due to acute seizures, stroke, head trauma or high fever, is followed by a silent/latent period with absence of seizures and then the genesis of spontaneous recurrent motor seizures that persist for the rest of the patient’s life (Acharya et al., 2008; Dhanushkodi and Shetty, 2008). Various animal models that were developed by electrical stimulation, by chemo-convulsion drugs such as pilocarpine/kainic acid, or by gene manipulation, provided valuable insight into the neurobiology of epileptogenesis, i.e. from IPI to the development of chronic epilepsy. The pathological features in animal models following IPI, such as hippocampal sclerosis (Sloviter, 1999), neuroinflammation (Oprica et al., 2003), loss of hippocampal interneurons (Kuruba et al., 2011) and abnormal neural plasticity (mossy fibres sprouting in the CA3 subfield and abnormal neurogenesis in the dentate gyrus) (Rao et al., 2006; Parent et al., 2006), closely mimic the clinical manifestation of human TLE. As epileptogenesis progresses, hippocampal J Tissue Eng Regen Med (2015) DOI: 10.1002/term
Cell-based therapy for hippocampal degenerative diseases
neurogenesis declines drastically (Hattiangady et al., 2004) and is manifested as severe depression and cognitive impairment. Nearly 30% of patients with TLE are refractory to anti-epileptic drugs. Besides patients, those who respond to anti-epileptic drugs display severe sideeffects, including cognitive impairments and motor disturbances. Thus, identifying alternative therapeutic strategies, such as stem cell therapy, to curtail TLE is highly justified. In this context, animal models of TLE are valuable prototypes to explore the possibilities of stem cell therapy for TLE and to understand the cellular–biochemical–molecular changes that take place following cell transplantation.
4. Hippocampal pathology in Alzheimer’s disease In 1907, Dr Alois Alzheimer first reported the clinical manifestations of a new neurodegenerative disease called Alzheimer’s disease (AD), a progressive neurodegenerative disease clinically characterized by decline in memory and cognitive abilities (Alzheimer.,1907). It is the most common cause of irreversible dementia, initially manifesting as impairment in episodic memory that gradually progresses towards a more crippling global decline in cognitive functions, incapacitating the individual and ultimately culminating in death. With a global increase in life expectancy and a rise in the ageing population, it is estimated that the incidence of AD will quadruple in the next 40 years, positioning it on the verge of attaining a pandemic status and concurrently posing an ever-increasing financial and emotional burden on patients, caregivers and society. Two prominent neuropathological features of AD are the appearances of amyloid plaques and neurofibrillary tangles. Amyloid plaques, also called neuritic plaques, are extracellular deposits of aggregated amyloid-β (Aβ) peptides, 40–42 amino acids-long proteolytic derivatives of the parent transmembrane protein, amyloid precursor protein (APP). On the other hand, neurofibrillary tangles are intracellular aggregates, often present in neuronal cell bodies and dendrites, containing hyperphosphorylated microtubule-associated TAU protein as the principal proteinaceous component. Apart from abnormal deposits of protein aggregates, the brains of AD patients and AD transgenic mice show evidence of oxidative damage (Tayler et al., 2010) and neuroinflammation (Graeber, 1999; McGeer et al., 2000), representative of their participation in the initiation and progression of AD pathogenesis. Broadly, two forms of AD exist, familial AD (FAD) and sporadic AD (SAD). Although the aetiology of SAD is unknown, several risk factors, such as ageing, diabetes, cardiovascular disease, apolipoprotein allele status and environmental toxins such as aluminium, have been suggested. On the other hand, genetic mutations in APP, presenilin-1 and presenilin-2 have been proposed as a primary cause for FAD. As the cholinergic neurons in the nucleus basalis of Meynert (NBM) are most Copyright © 2015 John Wiley & Sons, Ltd.
vulnerable in AD, initial attempts were made to mimic the animal prototype of AD by specifically destroying cholinergic neurons in the NBM (Fine et al., 1985; Emerich et al., 1992). Such animal models demonstrated profound cognitive impairments resembling human AD, albeit lacking amyloid plaques and neurofibrillary tangles. Later on, the neuropathological features of FAD were successfully mimicked in various transgenic animal models of AD by manipulating three crucial genes involved in FAD, i.e. APP, presenilin-1 and presenilin-2 (Games et al., 1995; Hsiao et al., 1996; Holcomb et al., 1998). Transgenic animal models of AD mimic the classical neuropathological hallmarks associated with human AD, such as abnormal accumulation of extracellular amyloid plaques and intraneuronal neurofibrillary tangles in the hippocampus and neocortex at the early stage of the disease and, as the disease progress, a widespread accumulation of amyloid plaques and neurofibrillary tangles recognizable in association cortices and subcortical structures such as the thalamus and amygdala. A salient feature of memory impairment associated with AD dementia, suggestive of hippocampal degeneration, is the rapid loss of recently acquired information, coupled with difficulty in consolidating new memories, with absolute preservation of the ability to recollect past memories. Similar to human AD, transgenic animal models of AD develop cognitive deficits, and thus serve as valuable tools to decipher the neurobiology of AD and also to explore possible therapeutic strategies (Kobayashi and Chen, 2005; Chen et al., 2000). Although the majority of the characteristic features of AD are imitated in transgenic animal models, not all of them demonstrate neuronal losses and enlargement of the ventricles, as was noticed in human AD.
5. Cell-based therapies for combating neurodegenerative diseases Neurodegenerative diseases are complex pathophysiological conditions in which the existing drugs in practice provide only symptomatic relief, without mitigating the course of the disease. Thus, there is a pressing need to identify alternative therapeutic strategies to arrest the progression of neurodegeneration. Cell therapy is emerging as a potential therapeutic modality against neurodegenerative diseases (Li et al., 2009a, 2014; Maisano et al., 2009; Rekha et al., 2009). Two possible avenues through which functional recovery could be achieved include stimulation, followed by mobilization of endogenous stem cells towards the vicinity of degenerating sites (Diederich et al., 2009; Li et al., 2012; Sehara et al., 2007), or transplanting exogenous neural cells that can proliferate and replace the lost neurons (Maisano et al., 2009; Rekha et al., 2009, 2011). The concept of adult neurogenesis is now widely acknowledged and is known to occur in two brain regions, the subventricular zone (SVZ) and the subgranular zone (SGZ) of the DG (Deng et al., 2009; Li et al., 2009b; Suh et al., 2009; Thuret et al., 2009; van Praag et al., 1999). J Tissue Eng Regen Med (2015) DOI: 10.1002/term
C. Venugopal et al.
Despite hippocampal neurogenesis being a continuous process, the rate at which it generates new neurons in the SGZ is not sufficient to compensate for the extensive neuronal losses observed in many hippocampal neurodegenerative conditions. A promising approach for replacing the lost neurons would be to transplant neural cells at the site of hippocampal injury (Acharya et al., 2009; Rekha et al., 2009). In the following sections, we review several preclinical investigations that provide evidence of cellular and functional recovery following local or systemic transplantation of neural precursor/adult stem cells in hippocampus-associated neurodegenerative conditions. For the cell transplantation approach, several sources of neuronal cells, such as conditionally immortalized hippocampal cells (Figure 2A), fetal neural precursor cells, embryonic stem cell-derived neural precursor cells (Figure 2B, C), induced pluripotent stem cells, terminally differentiated neurons (induced neurons) and adult stem cells such as mesenchymal stem/stromal cells (Figure 2D, E), have been investigated and will be discussed in the following sections.
6. Hippocampal cell line as a source for cell therapy Initial studies involving the transplantation of immortalized hippocampal cells in experimentally induced hippocampal neurodegenerative models provided early proofof-principle that transplanted cells could assist in
reconstructing the hippocampal circuitry and promoting behavioural recovery (Gray et al., 2000; Hodges et al., 2000; Martinez-Serrano et al., 1995; Sinden et al., 1997; Virley et al., 1999). We have demonstrated that in a rat model of subicular damage, an output component of the hippocampus that is affected in the very early stage of AD, transplantation of a hippocampal cell line derived from postnatal day 1 H-2Kb ts A58 transgenic mice into the hippocampal CA1 subfield exhibited greater migration through the corpus callosum and subsequent integration into the CA1, CA3 and DG subfields of the hippocampal formation. Furthermore, the transplanted cells prevented retrograde and anterograde neurodegeneration by releasing various growth factors, thus contributing to improved spatial learning and memory performance (Rekha et al., 2009). Although conditionally immortalized neural cells provide the benefits of off-the-shelf availability of neural cells for transplantation purposes, the integration of viral genes/oncogenes to transform neural cells into immortal cells put patients at higher risk of unwanted side-effects, such as teratoma formation, following transplantation.
7. Fetal neural precursor cells as a source for cell therapy Neural precursor cells (NPCs) freshly derived from fetus hippocampus are an excellent source of donor cells that can be used to repopulate the degenerated regions of
Figure 2. Potential sources of cells for cell therapy. (A) Conditionally immortalized hippocampal cell line. (B) In vitro embryoid body formation from human embryonic stem cells. (C) Neural stem cells derived from embryoid body. (D) Mesenchymal stem cells derived from human bone marrow. (E) Mesenchymal stem cells derived from human dental pulp. (F) Embryonic stem cells are characterized by their self-renewal and pluripotent properties to give rise to three germ cell types Copyright © 2015 John Wiley & Sons, Ltd.
J Tissue Eng Regen Med (2015) DOI: 10.1002/term
Cell-based therapy for hippocampal degenerative diseases
the hippocampus in TLE or AD. In support of this notion, NPCs obtained from the fetus hippocampus of transgenic mice [129-Gt (ROSA) 26Sortm1 (EGFP) Luo/J] encoding green fluorescent protein, when engrafted into the hippocampus of aged rats following kainic acid-mediated hippocampal injury resulted in the differentiation of 3–5% of engrafted cells into neurons, 28% into astrocytes and 6– 10% into oligodendrocytes (Shetty et al., 2008). Likewise, NPCs derived from the anterior subventricular zone of postnatal F344 rat pups expressing placental alkaline phosphatase, transplanted into kainic acid-mediated hippocampus-damaged adult F344 rats, displayed excellent survival and migration and differentiated into diverse neuronal subtypes, such as inhibitory interneurons, astrocytes, oligodendrocytes and oligodendrocyte progenitors. The grafted cells also expressed various neurotropic factors, such as glial-derived nerve growth factor (GDNF), brain-derived nerve growth factor (BDNF), fibroblast growth factor (FGF) and vascular endothelial growth factor (VEGF), which are well known for their mitogenic and regeneration capacities (Hattiangady and Shetty, 2012). In the rat model of TLE, transplantation of fetus GABAergic neurons into the hippocampus or substantia nigra restrained seizure development through restoration of GABA levels (Hattiangady et al., 2008; Loscher et al., 1998; Waldau et al., 2010). A recent study by Hunt et al. (2013) demonstrated that the medial ganglionic eminence progenitor cells derived from fetuses of green fluorescent protein-positive mice, upon transplantation into the hippocampus of epileptic mice, significantly suppressed seizure intensity and restored spatial learning and memory functions and the reduced aggressive and hyperactive behaviours that are the most common comorbidities observed in TLE subjects. Post-mortem analysis of brain samples from recipient mice revealed that the majority of the grafted cells were differentiated into functional GABAergic interneurons with prominent afferent and efferent connectivity. Given the complexity of AD, in which the pathology is widespread throughout the brain in advanced stages of the disease, cell therapy at the early stage of the disease could be beneficial in mitigating disease progression. As cholinergic transmission is compromised in AD, initial cell transplantation studies in animal models of AD, with lesions to cholinergic neurons in the NBM, demonstrated that transplantation of fetal cholinergic neurons derived from the septum or diagonal band of Broca into the neocortex improved acetyl choline levels and passive avoidance memory, whereas fetal hippocampal cells (noncholinergic) transplantation failed to restore cognitive function in this animal model (Fine et al., 1985). Consistent with this observation, another study by Emerich et al. (1992) demonstrated that transplantation of cholinergic neurons derived from fetal septum/diagonal band of Broca into the hippocampus of a rat model of AD improved cholinergic transmission and enhanced spatial learning and memory performance in an eight-arm radical maze task. However, when the complexity of the learning task was increased, the graft-recipient rats were not Copyright © 2015 John Wiley & Sons, Ltd.
able to perform efficiently as compared to naïve rats, suggesting that the cholinergic transplantation-mediated behavioural recovery depends upon the brain region where the graft cells are transplanted and is also limited to the intellectual demands of the testing circumstances. Regardless of the positive outcome observed in preclinical studies using rodent fetal cells for treating animal models of TLE/AD, translation of utilizing human fetus-derived NPCs for treating human TLE/AD is nearly impossible, due to serious ethical concerns involved in obtaining fetal samples and the huge volume of fetal tissues required for subsequent transplantations. In order to circumvent these obstacles, embryonic stem cells were recognized as an inexhaustible source to produce NPCs for cell therapy in neurodegenerative diseases.
8. Embryonic stem cells as a source for cell therapy Embryonic stem cells (ESCs) can be derived from the inner mass at the blastula stage and possess the unique characteristic features of self-renewal ability with simultaneous generation of lineage-specific progenitor cells of three germ layers (Figure 2F). Embryonic stem cells are highly versatile cells that can be properly guided towards the neuronal lineage under in vitro conditions. Neural precursor cells derived from ESCs can be cultured successfully in vitro as neurospheres (Figure 2B) and can be effectively differentiated into specific neuronal phenotypes for transplantation purposes (Vescovi et al., 1993). In a mouse model of TLE, in which hippocampal sclerosis plays a major role in epileptogenesis, NPCs derived from mouse ESCs, upon transplantation into the DG, exhibited enhanced survival and differentiation within the granular cell layer. Furthermore, this study also demonstrated that upregulation of chemoattractants, such as stromalderived factor-α1, as a consequence of hippocampal injury following status epilepticus attracted the transplanted NPCs preferentially towards the vicinity of the lesion (Hartman et al., 2010). As TLE is mainly caused by imbalance in excitatory and inhibitory functions, transplantation of NPCs with ventral forebrain identity derived from a mouse ES cell line, upon transplantation into the hippocampus of epileptic mice, could differentiate into various subtypes of GABAergic neurons with the typical electrophysiological properties of inhibitory neurons (Maisano et al., 2012). Although there have been relatively few studies, ESderived NPCs were also investigated in animal models of AD. Wang et al. (2006) compared the effects of transplantation of pure mouse ES cells or ES cell-derived neurospheres into a mouse model of AD with NBM lesions. Transplantation of pure ES cells into the frontal association cortex and barrel cortex induced extensive teratoma formation with further decline in cognitive function, highlighting the caution needed in using undifferentiated ES cells for cell therapy purposes. On the other J Tissue Eng Regen Med (2015) DOI: 10.1002/term
C. Venugopal et al.
hand, transplantation of neurospheres derived from ES cells differentiated into cholinergic neurons and, to some extent, into serotonergic neurons, with significant improvement in cognitive function. Akin to fetal neural precursor cells, utilizing ES-derived NPCs for clinical practice also presents several hurdles. This is mainly due to serious ethical and political issues associated with obtaining human embryonic tissue, potential teratoma formation and graft-versus-host immune reactions following transplantation.
9. Induced pluripotent stem cells as a source for cell therapy To bypass the ethical/political issues associated with using fetal or embryonic tissue for transplantation purposes, autologous induced pluripotent stem cells (iPSCs) can be used as an alternative cell source. Induced pluripotent stem cells are adult-derived somatic cells that can be reverted to embryonic stage by ectopic expression of transcription factors such as Oct4, Sox2, c-Myc and Klf4 (Takahashi and Yamanaka, 2006). Induced pluripotent stem cells have been extensively used to model human diseases in Petri plates (Okano and Yamanaka, 2014). Nevertheless, our understanding of iPSCs is still in its infancy and the potential of teratoma formation (Takahashi and Yamanaka, 2006) following iPSCs transplantation hamper their use for cytotherapy. Therefore, comprehensive investigations are necessary to evaluate the behaviour of iPSCs following transplantation before they can be used clinically.
10. Terminally differentiated neurons (induced neurons) as a source for cell therapy Contemporary studies highlight the feasibilities of direct conversion of adult somatic fibroblasts or brain astrocytes into terminally differentiated neurons called ’induced neurons’ by integrating neural fate determinant transcriptional factors, such as achaete-scute complex-like 1, brain-2 and myelin transcription factor-like 1 (ABM) (Pfisterer et al., 2011a; Vierbuchen et al., 2010; Addis et al., 2011). This direct conversion procedure bypasses the need for a transient proliferation stage and generates terminally differentiated neurons that do not form teratomas. Consequently, induced neurons provide the added advantages of producing tailor-made, autologous and terminally differentiated and phenotype-specific neurons to treat specific neurodegenerative diseases. In addition to ABM, a cocktail of other transcriptional factors and small molecules could be incorporated to produce specific subtypes of neurons, such as dopaminergic, cholinergic and spinal cord motor neurons (Pfisterer et al., 2011b; Son et al., 2011; Caiazzo et al., 2011; L. Liu et al., 2013). Copyright © 2015 John Wiley & Sons, Ltd.
Although it is not practically conceivable to obtain autologous human astrocytes, emerging evidence suggests that, like somatic fibroblasts, brain astrocytes can also be transdifferentiated into specific subtypes of neurons in in vitro (Addis et al., 2011). Advancing further, Torper et al. (2013) recently demonstrated that, following stereotaxic injection of Cre-inducible lentivirus that carried ABM into the striatum of GFAP-Cre transgenic mice, endogenous striatal astrocytes were efficiently transdifferentiated into neurons, thus emphasizing the likelihood of converting the endogenous terminally differentiated glial cells into mature neurons. While the production of induced neurons opens up the prospect of making patientand disease-specific neurons, the current protocols employed yield only a negligible percentage (
