Human neural progenitor cells

Human first trimester embryonic and fetal tissue, retrieved from elective routine abortions after informed consent, is a source of stem and progenitor cells that may be used in experimental and clinical regenerative medicine as described later. However, there are ethical, practical, and immunological challenges and concerns due to the origin and derivation of these human cells. Human cell therapy application in the laboratory or clinic necessitates adherence to ethical guidelines, evaluation and approval of reproducible and transparent study protocols by regulatory authorities and the regional human ethical committee, biobank regulations, and informed consent from tissue donor and recipient host. A close and stable collaboration between the laboratory and clinic in charge is a prerequisite, including set protocols for collection, dissection, cell culture, and expansion, viral, bacterial, donor cell test batteries, and reporting procedures.

In this study, we will review the usage of neural cells derived from the central nervous system (CNS) in experimental studies for developing treatment approaches for diseases and insults of the CNS, with a particular focus on spinal cord injuries (SCIs). We will discuss a few clinical trials including these cells as treatment, the possible therapeutic effects of transplanted cells, and the immunological aspects of clinical application of cell therapy in the CNS.

Diseases and insults of the CNS

Diseases of the CNS and brain and spinal cord injuries arising from different types of insults have some unique characteristics. Neurons in the CNS may be very large in size; they play a pivotal role in the key functions of the CNS, the input of sensory information, information processing, and output of motor function control (including autonomic functions such as respiration and blood pressure). The primary motor neurons in the human cerebral cortex have axons of length ≥1 m. It has been suggested that such features may render neurons susceptible to disease. Small groups of these highly specialized neurons have specific functions. As a consequence, loss of even a rather limited number of cells, for example, after a stroke may result in pronounced neurological symptoms. Neurons lost due to pathological processes will not be replaced by neuron regeneration (see below), with few exceptions. The same limitation probably applies for oligodendrocytes, the cells responsible for the myelination of axons. However, the CNS is also characterized by a high degree of plasticity, and even significant neurological symptoms and deficits often improve for several years after an insult.

An increasing number of studies have shown that neurodegenerative disorders such as Parkinson’s disease (PD), Alzheimer’s disease (AD), Huntington’s disease (HD), and amyotrophic lateral sclerosis (ALS) represent disease processes that progress during decades, with a slow but steady degeneration of neurons. Eventually, this neuronal loss results in clinical symptoms, especially with the loss of affected neurons, and plasticity and other compensatory mechanisms do not suffice. In the case of PD, loss of 30% of the affected dopamine neurons in the ventral mesencephalon is estimated when symptoms start to occur . Postmortem analyses showed a 60–70% reduction of the dopaminergic nerve terminals and the dopamine levels in the innervated caudate nucleus and putamen at this stage ; this was later verified in vivo using PET (positron emission tomography) imaging . Due to the late onset of symptoms, the possible time for treatment is limited to a disease stage when most of the affected neural cells are already lost. This could change if presymptomatic patients could be identified with biomarkers and imaging techniques. At present, identification of presymptomatic patients with AD, PD, and ALS is practically possible only among members of families with known inherited variants of these disorders, typically only a few percent of the patients.

Another important aspect of neurodegenerative disorders is that although the degeneration may be limited to a small population of neurons at the early stages of disease, a more extensive degeneration often occurs at later stages. The involvement of different types of neurons in several locations has obvious implications for the possible strategies for cell therapy in these diseases. In AD, loss of cholinergic neurons in the subcortical forebrain is an early pathological change associated with the loss of spatial memory, but global brain atrophy always occurs as the disease progresses. During the first years of disease, PD patients mainly suffer from symptoms caused by the loss of dopamine neurons in the substantia nigra and often develop symptoms of dementia associated with cortical atrophy . ALS is characterized by the loss of lower motor neurons in the ventral spinal cord and the upper motor neurons in the cerebral cortex. However, in some familial forms of ALS, the motor symptoms are accompanied by frontal lobe dementia. This has major consequences with regard to cell therapy strategies, as described later.

While HD is an inherited familial disease, the sporadic forms predominate among the other mentioned neurodegenerative disorders. Patients with familial forms show relatively similar symptoms and pathological changes as the sporadic form, but the disease typically affects younger people. In familial AD, the mutated gene is, in most cases, a part of the so-called γ-secretase complex, an enzymatic complex that cleaves a large number of membrane proteins such as amyloid precursor protein (APP) and notch. Regarding familial ALS, the most commonly mutated gene encodes another enzyme known as the superoxide dismutase 1 (SOD1). However, the loss-of-function gene mutation does not contribute to the occurrence of these diseases. The underlying mechanism of all these neurodegenerative disorders is rather believed to be the aggregation and accumulation of misfolded mutated proteins, or in the case of AD, aggregation of longer variants of the Aβ protein, which is produced in higher amounts due to cleavage of APP by the mutated γ-secretase. HD, PD, AD, and ALS as well as other related diseases are therefore often referred to as “conformational disorders” or “misfolded protein diseases” (for a review, see Soto & Estrada ).

However, neurodegenerative disorders are also caused by inherited loss-of-function gene mutations. These are very rare, severe inherited diseases affecting cell metabolism or intracellular storage, with severe effects on the brain of the affected children. Some of these fatal diseases such as infantile and late-infantile neuronal ceroid lipofuscinosis (NCL) and the leukodystrophy disorder Pelizaeus–Merzbacher disease (PMD) have been targeted for gene and cell therapy to replenish the host CNS with cells carrying the functional enzymes.

Multiple sclerosis (MS) is a progressive disease affecting the CNS. In contrast to other disorders, it is primarily a neuroinflammatory disease and has traditionally not been included among the primary neurodegenerative disorders. However, it has recently been recognized that neurodegeneration also occurs in MS, probably involving degenerative mechanisms such as oxidative injury and mitochondrial damage , processes that have also been implicated in neurodegeneration after ischemia and trauma.

In contrast to the neurodegenerative disorders, CNS insults such as stroke (focal brain ischemia), traumatic brain injury, and SCI have sudden onset, and the primary necrotic degenerative process affects all cell types in the affected region, often creating a cavity. Ischemic and traumatic insults also initiate secondary degeneration in the tissue surrounding the primary lesion leading to apoptotic cell death. A number of biochemical and cellular mechanisms have been suggested to be involved in the secondary degeneration, including increased extracellular concentrations of excitatory amino acids such as glutamate, release of free iron after local hemorrhage, formation of reactive oxygen species, collapse of ion gradients, energy crisis, and inflammation (see Lai et al. ). Several of these mechanisms interact and can amplify each other, and the possibilities for the so-called vicious circles are numerous. While the primary degeneration leads to cell death in hours to days, the secondary degenerative processes can probably be active for weeks to months in human patients. Consequently, the potential therapeutic window for cell therapy (or other treatments) is much longer if the secondary degeneration is targeted.

Degenerative processes in the CNS, particularly extensive necrosis after traumatic or ischemic insults, will activate the immune system and trigger inflammatory mechanisms in response, for example, to remnants of dead cells. However, whether inflammation in neurodegenerative disorders such as HD and AD is part of driving the ongoing neurodegeneration or is a secondary event is still an unanswered question. Even though inflammation is a reactive secondary process and not a primary event (except for MS), the resulting bystander injury of neurons may contribute significantly to the neuronal loss . These processes have been extensively studied with regard to CNS trauma and ischemia. The abundantly available experimental data show that primary necrosis elicits inflammatory reactions with potentially harmful effects on the surrounding tissue. The role of inflammation is, however, very complex . A reasonable hypothesis considering this fact is that the neuroprotective effect of transplanted donor human neural stem/progenitor cells (hNPCs) and other types of cell therapies may be due to an inhibitory effect of inflammatory processes.

The acute response in the CNS after a neural lesion includes infiltration of peripherally derived immune cells (such as neutrophils, monocytes, macrophages, and lymphocytes), activation of glial cells (microglia and astrocytes), and release of cytokines . The activated microglia/macrophages are considered to have at least two phenotypes, an M1 type that produces neurotoxins and an M2 type with an immunoregulatory function that supports wound healing. Inflammation is a critical defense mechanism that limits the spread of lesion and clears debris but also exacerbates the injury and damage in the surrounding healthy tissue. Inflammation and activated microglia may persist for months, and inflammation-related genes are often chronically upregulated. Thus, treatments with immunomodulatory effects, although a complex process to interfere with, can have an extended therapeutic window, which would allow start of a treatment several weeks after a CNS insult.

Diseases and insults of the CNS

Diseases of the CNS and brain and spinal cord injuries arising from different types of insults have some unique characteristics. Neurons in the CNS may be very large in size; they play a pivotal role in the key functions of the CNS, the input of sensory information, information processing, and output of motor function control (including autonomic functions such as respiration and blood pressure). The primary motor neurons in the human cerebral cortex have axons of length ≥1 m. It has been suggested that such features may render neurons susceptible to disease. Small groups of these highly specialized neurons have specific functions. As a consequence, loss of even a rather limited number of cells, for example, after a stroke may result in pronounced neurological symptoms. Neurons lost due to pathological processes will not be replaced by neuron regeneration (see below), with few exceptions. The same limitation probably applies for oligodendrocytes, the cells responsible for the myelination of axons. However, the CNS is also characterized by a high degree of plasticity, and even significant neurological symptoms and deficits often improve for several years after an insult.

An increasing number of studies have shown that neurodegenerative disorders such as Parkinson’s disease (PD), Alzheimer’s disease (AD), Huntington’s disease (HD), and amyotrophic lateral sclerosis (ALS) represent disease processes that progress during decades, with a slow but steady degeneration of neurons. Eventually, this neuronal loss results in clinical symptoms, especially with the loss of affected neurons, and plasticity and other compensatory mechanisms do not suffice. In the case of PD, loss of 30% of the affected dopamine neurons in the ventral mesencephalon is estimated when symptoms start to occur . Postmortem analyses showed a 60–70% reduction of the dopaminergic nerve terminals and the dopamine levels in the innervated caudate nucleus and putamen at this stage ; this was later verified in vivo using PET (positron emission tomography) imaging . Due to the late onset of symptoms, the possible time for treatment is limited to a disease stage when most of the affected neural cells are already lost. This could change if presymptomatic patients could be identified with biomarkers and imaging techniques. At present, identification of presymptomatic patients with AD, PD, and ALS is practically possible only among members of families with known inherited variants of these disorders, typically only a few percent of the patients.

Another important aspect of neurodegenerative disorders is that although the degeneration may be limited to a small population of neurons at the early stages of disease, a more extensive degeneration often occurs at later stages. The involvement of different types of neurons in several locations has obvious implications for the possible strategies for cell therapy in these diseases. In AD, loss of cholinergic neurons in the subcortical forebrain is an early pathological change associated with the loss of spatial memory, but global brain atrophy always occurs as the disease progresses. During the first years of disease, PD patients mainly suffer from symptoms caused by the loss of dopamine neurons in the substantia nigra and often develop symptoms of dementia associated with cortical atrophy . ALS is characterized by the loss of lower motor neurons in the ventral spinal cord and the upper motor neurons in the cerebral cortex. However, in some familial forms of ALS, the motor symptoms are accompanied by frontal lobe dementia. This has major consequences with regard to cell therapy strategies, as described later.

While HD is an inherited familial disease, the sporadic forms predominate among the other mentioned neurodegenerative disorders. Patients with familial forms show relatively similar symptoms and pathological changes as the sporadic form, but the disease typically affects younger people. In familial AD, the mutated gene is, in most cases, a part of the so-called γ-secretase complex, an enzymatic complex that cleaves a large number of membrane proteins such as amyloid precursor protein (APP) and notch. Regarding familial ALS, the most commonly mutated gene encodes another enzyme known as the superoxide dismutase 1 (SOD1). However, the loss-of-function gene mutation does not contribute to the occurrence of these diseases. The underlying mechanism of all these neurodegenerative disorders is rather believed to be the aggregation and accumulation of misfolded mutated proteins, or in the case of AD, aggregation of longer variants of the Aβ protein, which is produced in higher amounts due to cleavage of APP by the mutated γ-secretase. HD, PD, AD, and ALS as well as other related diseases are therefore often referred to as “conformational disorders” or “misfolded protein diseases” (for a review, see Soto & Estrada ).

However, neurodegenerative disorders are also caused by inherited loss-of-function gene mutations. These are very rare, severe inherited diseases affecting cell metabolism or intracellular storage, with severe effects on the brain of the affected children. Some of these fatal diseases such as infantile and late-infantile neuronal ceroid lipofuscinosis (NCL) and the leukodystrophy disorder Pelizaeus–Merzbacher disease (PMD) have been targeted for gene and cell therapy to replenish the host CNS with cells carrying the functional enzymes.

Multiple sclerosis (MS) is a progressive disease affecting the CNS. In contrast to other disorders, it is primarily a neuroinflammatory disease and has traditionally not been included among the primary neurodegenerative disorders. However, it has recently been recognized that neurodegeneration also occurs in MS, probably involving degenerative mechanisms such as oxidative injury and mitochondrial damage , processes that have also been implicated in neurodegeneration after ischemia and trauma.

In contrast to the neurodegenerative disorders, CNS insults such as stroke (focal brain ischemia), traumatic brain injury, and SCI have sudden onset, and the primary necrotic degenerative process affects all cell types in the affected region, often creating a cavity. Ischemic and traumatic insults also initiate secondary degeneration in the tissue surrounding the primary lesion leading to apoptotic cell death. A number of biochemical and cellular mechanisms have been suggested to be involved in the secondary degeneration, including increased extracellular concentrations of excitatory amino acids such as glutamate, release of free iron after local hemorrhage, formation of reactive oxygen species, collapse of ion gradients, energy crisis, and inflammation (see Lai et al. ). Several of these mechanisms interact and can amplify each other, and the possibilities for the so-called vicious circles are numerous. While the primary degeneration leads to cell death in hours to days, the secondary degenerative processes can probably be active for weeks to months in human patients. Consequently, the potential therapeutic window for cell therapy (or other treatments) is much longer if the secondary degeneration is targeted.

Degenerative processes in the CNS, particularly extensive necrosis after traumatic or ischemic insults, will activate the immune system and trigger inflammatory mechanisms in response, for example, to remnants of dead cells. However, whether inflammation in neurodegenerative disorders such as HD and AD is part of driving the ongoing neurodegeneration or is a secondary event is still an unanswered question. Even though inflammation is a reactive secondary process and not a primary event (except for MS), the resulting bystander injury of neurons may contribute significantly to the neuronal loss . These processes have been extensively studied with regard to CNS trauma and ischemia. The abundantly available experimental data show that primary necrosis elicits inflammatory reactions with potentially harmful effects on the surrounding tissue. The role of inflammation is, however, very complex . A reasonable hypothesis considering this fact is that the neuroprotective effect of transplanted donor human neural stem/progenitor cells (hNPCs) and other types of cell therapies may be due to an inhibitory effect of inflammatory processes.

The acute response in the CNS after a neural lesion includes infiltration of peripherally derived immune cells (such as neutrophils, monocytes, macrophages, and lymphocytes), activation of glial cells (microglia and astrocytes), and release of cytokines . The activated microglia/macrophages are considered to have at least two phenotypes, an M1 type that produces neurotoxins and an M2 type with an immunoregulatory function that supports wound healing. Inflammation is a critical defense mechanism that limits the spread of lesion and clears debris but also exacerbates the injury and damage in the surrounding healthy tissue. Inflammation and activated microglia may persist for months, and inflammation-related genes are often chronically upregulated. Thus, treatments with immunomodulatory effects, although a complex process to interfere with, can have an extended therapeutic window, which would allow start of a treatment several weeks after a CNS insult.

Stem and progenitor cells in the fetal/embryonic CNS

Various immature cells can be isolated from human embryonic and fetal CNS residual tissue and in vitro cultured either as free-floating aggregates of cells or as adherent monolayer cultures in the presence of basic fibroblast growth factor (bFGF) and epidermal growth factor (EGF). There seems to be lack of clarity on how to denominate the different types of cells. In addition to self-renewal, a common feature of all these cells, human neural stem cells (hNSCs) are defined by their multipotency: the ability to differentiate into neurons, astrocytes, and oligodendrocytes, the three major cell types in the CNS (microglial cells have a mesodermal origin). Progenitor cells are more restricted with regard to their differentiation potency, and the terms “neuronal restricted progenitor cells” (NRP) and “glial restricted progenitor cells” (GRP) are occasionally used for progenitor cells specified for neuronal or glial (astrocyte and oligodendrocyte) lineages. Neural progenitor cells thus refer to cells that are more restricted than the hNSCs (hence they do not give rise to the three major cell types) but otherwise not defined. Embryonic/fetal first trimester-derived neural cells cultured as neurospheres are more heterogeneous in nature than the adherently grown cells; spontaneous differentiation of the hNSCs to restricted progenitors occurs. The term “neural progenitor cells” or “neural precursor cells” (NPCs) is therefore often used for neurosphere cells to emphasize the heterogeneity. Importantly, dissociated neurospheres can be used to establish adherent cultures and vice versa.

Some progenitor cells are more specified than only referring to neuronal or glial lineage. Progenitor cells are restricted in the developing neural tube and specifically differentiate into discrete subpopulations of neurons. This is achieved by the combined action of gradients of the so-called morphogens such as sonic hedgehog and bone morphogenetic protein, extensively studied in the developing spinal cord . hNPCs can be pharmacologically and/or genetically modified, thus making them more beneficial. The genetic modification typically applies knowledge about the normal development and lineage specification of neural stem and progenitor cells, achieved through systematic studies of the developing murine and human CNS. Taking into consideration the available information on neural stem cell specification, Hwang et al. used the transcription factor Olig2 to direct hNPCs toward the oligodendrocyte lineage to increase myelination of axons after SCI.

Repair of the adult CNS by endogenous stem cells

The existence of endogenous stem cells and neurogenesis in the mature mammalian and human brain was a finding that contrasted to the general belief that neurogenesis occurs only during development. Endogenous NSCs have recently been shown to reduce the detrimental effects of CNS injury by decreasing secondary degeneration and improving neuronal survival through neurotrophic effects . Endogenous stem cells have also been shown to enhance functional recovery in rodent stroke models . However, in contrast to the insights gained from animal experiments , there is no detectable neurogenesis in the human brain after stroke ; thus it is questionable if endogenous neural stem cells in humans play a role in the recovery from CNS injuries. In the future, pharmacological treatment may be used to activate neurogenesis and replace neurons lost in disease or after insult. One may envisage that this could be achieved in PD, in which a relatively small population of neurons degenerates. However, it may be difficult to recreate the many millions of neurons, astrocytes, and oligodendrocytes that die after stroke or traumatic injury. Transplantation of NPCs with the capacity to replace the lost cells may represent a shorter path to clinical application.

Therapeutic mechanisms of transplanted hNPCs

Transplantation of NPCs is necessary for replacing lost cells, particularly in the CNS with very limited endogenous regenerative capacity. The requirement of new neurons will be very different depending on the type of disease or insult treated. In PD, increased extracellular dopamine levels in the caudate nucleus and putamen, the main regions innervated by the substantia nigra, are adequate for a significant improvement of the motor symptoms without a precise temporal and spatial control of the neurotransmitter release. The requirements on primary motor circuitry affected in ALS or after insults of the brain and spinal cord are very different. Control of muscle function not only requires a high degree of temporal resolution of neuronal activity via the corticospinal tract and the motor neurons but also an intricate organization involving other descending neuronal systems, interactions between central pattern generators, extensor, and flexor muscles and sensory feedback.