Parkinson's disease (PD) is characterized by the loss of dopaminergic neurons of the substantia nigra pars compacta (SNpc). Although dopamine (DA) loss is not the only important factor in the etiology of the disease, and several other neurotransmitter systems contribute to its initiation, progression and symptoms, particularly in late-stage PD, the consensus is that DA loss constitutes both the primary and major cause for the most incapacitating motor symptoms of the disease. As a result of this, pharmacological drugs used in PD patients target the dopaminergic system, aiming to replace the lost DA [1]. In this context, fetal tissue transplantation carried out over the last two decades has aimed at the ectopic placement (striatum) of a relevant number of functional DA neurons which, by the release of DA in the caudate and putamen, can offer substantial therapeutic benefit to the patient [2]. Clinical trials using fetal tissue have provided proof-of-principle that neuron replacement works for some motor symptoms (bradykinesia and rigidity), but it does so in a variable degree among patients, does not ameliorate other advanced-stage symptoms (e.g., postural imbalance, freezing and falling, and dementia), and in some cases results in the appearance of disabling dyskinesias. Some suggest that the reported controversial results on efficacy between trials can be explained on the basis of differences in tissue preparation, methodological procedures, immunosuppression, age and disease stage of the transplant recipients, and graft location [3?5]. In this scenario of hot debate on the future of cell transplantation in PD [6,7], recently awarded EU funds will enable a new clinical trial. By bringing together different clinical centers, and at the same time unifying methodologies, it is expected ? since the transplants do work in some patients ? that this new trial will give a more coherent picture of the actual therapeutic benefit.
Leaving aside from the present editorial, the ethical concerns regarding the use of fetal tissue and local legal frameworks regulating fetal tissue research and use in some countries, the major hurdle to developing cell transplantation into a routine clinical practice is the limitation imposed by fetal tissue procurement, which is difficult to standardize and requires the use of tissue from several fetuses for a single patient. This is the main reason why, over a decade ago, a major research effort was undertaken to find alternative sources of tissue for transplantation in PD, and when stem cells (SCs) emerged as a fundamental player in the cell therapy scenario. One concept learnt from previous clinical trials is the need for a standardized cell source for transplantation in order to homogenize procedures and outcomes, and, in this manner, SCs may offer the possibility for the reproducible generation of (unlimited numbers of) functional human DA neurons. However, in spite of the hope that the new trial will clarify the procedures to be used for an effective therapeutic use of DA neuron replacement, there are still major roadblocks in the way to the implementation of the use of SCs as donor tissue for PD patients. Whereas a complete account of individual research papers describing developments in the field can be found elsewhere [2,5,8,9], here we will focus on some challenging aspects that remain to be solved before SCs, or their derivatives, can be used in a safe and competitive manner when compared with fetal tissue.
Neural stem cells (NSCs) of fetal human origin (hNSCs) were the first type of SC explored for PD cell therapy in the mid-1990s, soon after the culture of neurospheres was described [10]. Although highly safe (in vivo tumor formation has never been reported for true hNSCs, see [11]), a decade since, hNSCs still pose problems for implementation in clinical use for PD. Although sometimes not evident from the literature, hNSCs ? particularly those derived from the ventral mesencephalon (VM) ? grow poorly in culture, change their properties with time [passages], loose their capacity for the generation of neurons (and DA neurons in particular), and survive poorly in the brain when grafted [5,12?14]. This results in modest behavioral impact in animal models when compared with fresh fetal tissue (in the best case, a 50% compensation in drug-induced rotation in hemiparkinsonian rats). Genetic modification of either forebrain or midbrain NSCs with transcription factors (e.g., Ascl1, Nurr1, Lmx1a or Foxa2) may enhance their neurogenic potential in vitro [2,5,15,16], but does not result in fully efficient transplants. Even though not considered at present for clinical use, immortalized lines of hNSCs from the VM, which overcome some of the limitations described above, have been generated. These lines are immortalized but not transformed by any criteria examined and do not generate tumors in vivo, they show great stability with time in culture, but, in the case of VM cells, tend to lose their phenotype or their neurogenic capacity. As a result, efficient transplants have not yet been reported. In our group, we have recently generated a cell line of VM origin (hVM1) cells, which fulfill all the criteria required as a source of human DA neurons for transplantation in PD (see [17] for a description and discussion of the aforementioned properties of this and other cell lines). The cells fully differentiate into true SNpc human DA neurons of the A9 subtype [17]. The problem found with these hVM1 cells is that they progressively lose their neurogenic potential upon passaging ( 1 year of continuous culture), as occurs with primary neurospheres. However, similar to the results obtained in previous studies using forebrain hNSCs [18,19], we have reverted this situation using Bcl-XL [20]. Bcl-XL did not only enhance survival of the cells through a canonical antiapoptotic action, but also induced neurogenic genes and genes specifically related to dopaminergic development and maturation. As a result, hVM1 cells became a stable producer of human DA neurons that survived well in the grafted hemiparkinsonian rat brain (2 months survival time examined), without forming tumors, and matured to exert behavioral recovery [20]. This study constitutes the first demonstration that VM-derived hNSCs can constitute a source of useful human DA neurons for transplantation. Bcl-XL has also been recently used to enhance the survival of human DA neurons from human embryonic SCs (hESCs) [21]. Obviously, future molecular and cell biology studies should aim at replacing the genetic modifications in the cells (v-myc and Bcl-X L) to generate stable NSCs suitable for use in humans. In the particular case of v-myc, a detailed investigation of its actions at the molecular level is urgently needed. Understanding the mechanism(s) by which v-myc allows for stable proliferation of hNSCs, together with deeper knowledge of the factors that regulate the hNSC niche are essential in order to manage their healthy culture, and thus develop hNSCs into a cellular product for clinical therapies. In addition, both for primary (unmodified) and for immortalized hNSCs, thorough safety studies should be conducted before planning any implantation in humans. We would like to stress at this point that, even when immortalized NSCs carry a gene that controls the cell cycle, after almost two decades of work no evidence for unsafe behavior has ever been reported and that their status should be the same as that granted to their primary (unmodified) counterparts (exemplified by neurosphere cultures of hNSCs).
Another promising cell source for neuron-replacement therapy are embryonic stem cells (ESCs), as they can provide an unlimited source of self-renewing cells suitable for massive expansion (as opposed to primary hNSCs), and subsequent differentiation to DA neurons for transplantation. Unfortunately, owing to their proliferation potential, a significantly greater effort is required to restrict their differentiation both in vitro and in vivo, in order to halt cell division, achieve the desired phenotype and to avoid tumor formation.
Embryonic stem cells can be efficiently guided to differentiate toward a neuronal lineage, producing enriched populations of cells exhibiting in vitro functions typical of presynaptic DA neurons [21?23]. Increased efficiency of midbrain DA neuron generation can be achieved by the forced expression of specific transcription factors that are important during midbrain development, such as Nurr1, Pitx3, Lmx1a or Foxa2 [24?26] or the antiapoptotic factor Bcl-XL [22]. Inhibition of bone morphogenic protein (BMP) signaling by Noggin can improve the generation of neuroepithelial progenitors generated from human ESCs in culture [27]. These neuroepithelial cells can further develop into midbrain DA neurons by coculture with VM astrocytes [28] or exposure to compounds such as Wnt5a plus fibroblast growth factors [29] or cholesterol derivatives [30]. The generated DA neurons show morphological, molecular and pharmacological (e.g., proper responses to receptor/channel agonists) properties expected from neurons from the midbrain [31]. Although these attributes suggest that hES-derived DA cells may be a desirable source of cells for cell-replacement strategies, the prospects of using hESCs for PD cell therapy remain uncertain, gives inconsistent outcomes reported in PD animal models, including limited survival and function or, at the other end, tumor formation (due to the persistence of proliferating cells) of grafted hESC-derived DA neurons [27,28,32].
A possible solution to homogenize these cultures may be achieved by the enrichment of the cells of interest and/or exclusion of unwanted cells by using fluorescence-activated cell sorting [33,34]. However, for this purpose, good surface markers should be available, or alternatively, safe and specific gene marking with fluorescent reporter genes. A novel strategy that was recently published included the expansion of hESC-derived neural precursor cells (NPCs) of midbrain-type (hESC-NP) in conjunction with Bcl-XL and SHH transgene expression [21]. The advantage of this method seems to be the stable expansion of partially differentiated NPCs, which reduces the presence of tumorogenic cells, along with improved survival of the grafted cells (2 months survival time). As it was the case for the aforementioned hNSCs [20], there are, however, no long-term grafting studies yet published.
The recent development of methods for the generation of ESC-like induced pluripotent SCs (iPSCs) by cellular reprogramming of adult somatic cells by retroviral transduction of defined factors such as Oct3/4, Sox2, c-myc and Klf4 [35,36] offers new possibilities to treat patients with autologous cell grafts. This is a fast-developing field, yielding improvements such as the generation of iPSCs without viral integration [37] and/or with removal of the reprogramming factors [38?40], which are thought to facilitate the use of the cells in future clinical settings. iPSCs can be derived from, for example, skin or fat cells isolated from the own patient or from a healthy immunocompatible individual. Interestingly, iPSCs derived from mouse fibroblasts have been successfully differentiated into midbrain DA neurons; which could alleviate motor asymmetry in 6-OHDA-lesioned rats [41]. The potential advantage of iPSCs is obvious, since they can be obtained without using human embryos, and therefore reducing or eliminating ethical concerns in comparison with hESCs, but they may also represent a potentially histocompatible tissue for autologous transplantation, and may eliminate concerns regarding tissue rejection that plague ESC therapy. However, considerable challenges must still be overcome before these strategies may become a reality in the clinic. First, the exact nature of human iPSCs status and whether they are truly equivalent to hESCs must be demonstrated. It is has been shown that reprogrammed cells quite often keep the genetic/epigenetic memory of adult cells, even after reprogramming into the pluripotent state [42]. Second, it is necessary to establish and standardize efficient differentiation protocols, avoiding culture heterogeneity after directed differentiation, in order to eliminate both the risk of tumor formation, and the presence of cells others than those needed.
In summary, much progress has been recently achieved, providing the neurology field, for the first time, with two sources of true human A9 DA neurons, namely VM hNSCs and hESCs (or h-iPSCs). Enhancing the stable propagation of VM hNSCs, improving the differentiation and at the same time reducing the tumor formation risk of ES and ES-like cells, remain as the major goals. In addition, deriving procedures for the predictable and consistent survival and function of transplanted cells will help the neuron-replacement field to get closer to a realistic clinical application in PD.
See Original Articles, please Click here |
