Background

Parkinson’s disease (PD) is the second most prevalent neurodegenerative disorders characterized by marked depletion of dopamine caused by the loss or degeneration of DA neurons in the substantia nigra (SN) of the midbrain [1,2,3]. Thus far, there is still no curative treatment for PD although existing treatments such as deep brain stimulation and pharmacotherapies (including levodopa monoamine oxidase B) can alleviate some of the symptoms; they tend to lose efficacy over time and thus do not prevent irreversible disabilities indefinitely [4,5,6,7]. Based on the molecular pathological properties of PD, cell replenishment therapy, especially for DA neurons, is considered the most attractive approach to treating PD patients. The advent of stem cells and cellular reprogramming has brought the gospel for PD patients, opening the way for potential applications of cell-based therapy. Among various types of cells, neural stem cells (NSCs) have long been demonstrated to be a promising candidate for treating PD. Nevertheless, the clinical application of NSCs is still challenging due to limited cell sourcing and ethical concerns. Therefore, other types of stem cells that are readily available appear to hold enormous promise for cell replacement therapy in regenerative diseases.

In various studies, reprogrammed somatic cells or stem cells including induced pluripotent stem cells (iPSCs), embryonic stem cells (ESCs), and mesenchymal stem cells (MSCs), e.g., from individual sources, have exhibited the unique characteristics of self-renewal capacity and multipotential differentiation. Nevertheless, each type has pros and cons as an available source for regenerative cell-based therapies. In particular, there are several serious concerns including ethical issues, tumor formation, and genetic instability, thus undermining their therapeutic application [8,9,10]. These are mainly attributable to the generation of iPSCs with several exogenous genetic manipulations, destroying the embryo in harvesting ESCs and generating teratoma arising from intermediate pluripotent cell state after transplantation [10, 11]. Additionally, current protocols for differentiating these types of stem cells into lineage-specific neurons such as DA neurons usually have poor conversion efficiency and are highly involved in terms of time consumption, technical complexity, and induction materials [12,13,14]. Therefore, it is imperative to seek an alternative stem cell source that can generate DA neurons while being free of evident disadvantages such as ethical issues and tumor formation. Meanwhile, it is of great importance to establish a highly efficient and cost-effective differentiation strategy to generate functional DA neurons from stem cells.

Spermatogonial stem cells (SSCs) are a subpopulation of type A spermatogonia [15, 16]. SSCs were previously regarded as unipotent stem cells, differentiating into sperm only. With further studies on SSCs, however, the previously prevailing orthodoxy has been challenged. To date, a growing number of studies have demonstrated that SSCs from both mouse and human testes can dedifferentiate into ES-like cells that are capable of giving rise to various cell lineages of all three germ layers [17,18,19,20,21,22,23], implying that SSCs may be an alternative cell source for regenerative medicine. In addition, SSCs have several merits over other types of stem cells, including lack of ethical issues, lower tumorigenesis, and no immunogenicity [24]. More notably, SSCs can transdifferentiate into uterine, prostatic, and skin epithelium cells in vivo after transplantation into these tissues, mainly due to the strong sensitivity of SSC to living microenvironments [21]. Thus, SSCs could serve as another potential source for replacing damaged or degenerated cells in various neurodegenerative diseases. However, direct conversion of hSSCs to functional DA neurons without forced expression of transcription factors or reprogramming to a pluripotent state has not been achieved in vitro.

Here, we describe a straightforward two-step induction strategy for differentiating hSSCs into DA neurons based on unique characteristics of SSCs. Using this protocol, we achieved the direct conversion of hSSCs into DA neurons that recapitulated the morphological, key biochemical, and functional features of primary midbrain DA neurons. Notably, the acquisition of hSSC-derived DA neurons using OECCM and a set of defined cell-extrinsic factors as well as several small molecules could be much safer than other options. On the other hand, the newly established differentiation protocol obviates a complicated manipulation procedure and reduces the overall cost of the differentiation process. The present groundbreaking study may open the possibility of generating human DA neurons from hSSCs for PD therapy.

Materials and methods

Preparation of primary hSSCs

Testicular tissues from obstructive azoospermic (OA) patients at the age of 25–45 years who had normal spermatogenesis and needed microdissection for testicular sperm extraction were fully washed in DMEM/F12 (Gibco) with antibiotics (Gibco). Importantly, all experiments were performed in accordance with the relevant guidelines and regulations of the Institutional Ethical Review Committee of Hong Hui Hospital affiliated to ** in the cylinder, while the reduced activity was inhibited by engrafted hSSCs. Statistical analysis indicated that there was a significant difference between the mock and hSSC transplantation groups, and no significant difference was found between the normal and hSSCs transplantation groups (Fig. 8a). Moreover, alterations in spontaneous activity to remove the label were reliably observed after hSSC transplantation. Strikingly, MPTP-treated animals engrafted with hSSCs showed significant improvement in adhesive removal performance compared with mock animals, displaying a significantly shortened time to remove the label using their forepaws. Additionally, there was a statistical difference between engrafted hSSCs of MPTP-treated and normal animals (Fig. 8b), indicating the therapeutic potential of hSSCs. Consistent with the spontaneous activity test, the gait analysis indicated that MPTP-treated mice showed a progressive decline in the walking distance during the unit time (2 min), whereas engraftment of hSSCs remarkably impeded this declining trend. Nevertheless, no significant difference was found at any of the time points posttransplantation (Fig. 8c). Likewise, the walking velocity of MPTP-treated animal was also increased by hSSC engraftment and showed an increasing tendency with prolongation of transplantation time. Strikingly, the significant increase occurred mainly beyond 4 weeks posttransplantation; however, no significant difference was observed between the hSSC transplantation and normal groups although there was an increasing trend in the walking velocity in hSSC transplantation group (Fig. 8d). In addition, analysis of challenging beam assays indicated that MPTP-treated mice showed a significant increase in errors per step compared to normal mice, while the parameter regarding errors per step was remarkably decreased by hSSC engraftment at 6 weeks posttransplantation (Fig. 8e). When animals were tested on a beam that tapered from wide to narrow, similar results were obtained (Fig. 8f), demonstrating their propensity to make errors was reduced. Additionally, the number of amphetamine-induced rotations following MPTP injection in animals transplanted pre-induced hSSCs decreased (at 1, 2, and 3 weeks posttransplantation: 899 ± 185 turns/90 min, 756 ± 215 turns/90 min, and 560 ± 235 turns/90 min, respectively) compared to control group (1033 ± 243 turns/90 min). Notably, the number of amphetamine-induced rotations significantly decreased in the animals transplanted pre-induced hSSCs for 3 weeks compared to that in control group (p < 0.05) (Fig. 8g).

Fig. 8
figure 8

Behavioral deficit improvement after hSSC transplantation. a Hindlimb step** in the cylinder in normal (n = 10), hSSC-transplanted (n = 8), and mock (saline-injected, n = 6) mice at 7 weeks postoperation. b Contact time in normal (n = 10) and hSSC-transplanted (n = 8), and mock (saline-injected, n = 5) mice at 7 weeks postoperation. “*” and “#” represent p < 0.05 compared to mock and normal, respectively. c, d Distance traveled in unit time and movement velocity in normal (n = 10), hSSCs-transplanted (n = 8), and mock (saline-injected, n = 6) mice at postoperation. e Errors per step in the challenging beam of the abovementioned mice at 7 weeks postoperation. f Mean errors at different beam widths. g The number of amphetamine-induced rotation in hSSCs groups, compared to that of the control group (mock). “*” and “**” represent p < 0.05 and p < 0.01, respectively, compared to mock. Each animal received at least five trials. “#” and “##” represent p < 0.05 and p < 0.01, respectively, compared to normal

Full size image

Discussion

The progressive degeneration and further loss of DA neurons in the midbrain is the main pathological process of PD, generally causing motor and sensory dysfunction [41, 42]. Regardless of the temporary amelioration of symptom by current therapeutic approaches (drug, gene, surgery and deep-brain stimulation, etc.), no effective treatments hitherto can cure the disease. On the basis of the properties of PD, cell replenishment therapy has been proposed as a promising treatment strategy for PD. Despite the therapeutic potential of a variety of stem cells including NSCs, their availability is extremely limited, mainly due to their individual shortcomings, including low accessibility and expansion and ethical and immune concerns [39]. Although ESCs, iPSCs, and MSCs hold great potential in regenerative medicine, their clinical application could not circumvent the risk for undesired genotoxicity, mutagenesis, and tumorigenesis besides differentiation uncertainty and ethical controversies [9,10,11, 43]. For this reason, finding a desirable cell-based therapy for PD is essential. Recently, advances in stem cell reprogramming studies have revealed that SSCs can transdifferentiate into neural cells, cardiovascular cells, and prostatic, uterine, and skin epithelial cells in addition to sperm lineage cells [17, 19, 22, 24, 44, 45]. This unique differentiation property is mainly attributed to both SSC-intrinsic and SSC-extrinsic factors [21]. In addition, several previous reports together with our recent studies demonstrated that SSCs have substantial advantages over other types of stem cells, such as rapid expansion, strong multipotential, and high amenability to the extrinsic environment [24, 46, 47]. More importantly, in comparison to several types of primary stem cells, SSCs have several attractive features, including no ethical concerns, lower tumorigenicity and host immune response, and wide availability [48], thus avoiding current obstacles and risks of autologous or allogenetic transplantation of SSCs. Therefore, we speculate that SSCs are likely to be the best prospective candidates for clinical applications. Nonetheless, there is not a highly successful and efficient method of reprogramming hSSCs to DA neurons.

In the present study, we developed an efficient and straightforward protocol for the conversion of hSSCs into functional DA neurons. The converted SSCs exhibited typical neuronal morphology accompanied by biochemical phenotypes of midbrain DA neurons, as well as the ability to secrete dopamine and initiate neuron-specific electrophysiology activity and calcium imaging. Following the in vitro study, the therapeutic potential of the converted cells from hSSCs was specifically supported in our in vivo experiment by behavioral improvement in the MPTP-lesioned mice after striatal cell transplantation. Safety, efficacy, and practicability will be of critical importance in develo** a novel strategy for PD clinical therapy. Based on this, we tried to avoid the deficits and undesired side effects in the conventional protocol and further optimize the tedious procedure by relying on the high amenability of SSCs to the extrinsic environment. To develop a highly efficient approach for DA neuron conversion in hSSCs, apart from the delivery of traditional induction factors, such as GDNF, RA, and forskolin, OECCM and several defined components (SHH, TGFβ3, FGF8α, VPA, and SB) that effectively promote reprogramming events, including epigenetic dynamics, neurogenesis, survival, neuron subtype specification, and maturation, were supplemented. This transdifferentiation protocol for converting hSSC to DA neurons circumvents conventional drawbacks: inefficiency, tedious multistep processes, and the high risk of exogenous gene delivery. In summary, we sought to develop the hSSC induction system for future clinical consideration.

In this study, we demonstrated that this induction system is particularly amenable to giving rise to a large number of functional DA neurons as demonstrated by significant dopaminergic marker expression and functional properties after 2 and 3 weeks of in vitro induction. To gain novel insights into the molecular mechanisms underlying the neural conversion of hSSCs, we illustrate the principles of establishing this induction system, relying on the roles of each component in the conversion. Herein, OECCM was used as a critical basic component from the OEC culture supernatant. Numerous studies have demonstrated that OECs can secrete a variety of neurotrophic growth factors and cell adhesion molecules, some of which play pivotal roles in directing neural differentiation, survival, maturation, and migration and further maintenance of differentiated neuronal phenotypes [49,50,51,52,53,54]. Similar to the claim, our recent study also revealed that OECCM significantly improved adipose-derived stem cell and bone mesenchymal stem cell transdifferentiation efficiency [55, 56]; therefore, OECCM includes the induction system. The use of a defined set of factors, including SHH, GDNF, FGF8α, and TGFβ3, plays pivotal roles in neurogenesis and differentiation. In particular, some factors (e.g., SHH and FGF8α) are requisite for subtype specification of differentiating neurons during the embryonic brain development stage [57,58,59,60,61], since FoxA2 and Lmxla/b, two important effectors in downstream of the SHH signaling pathway, have been shown to effectively convert differentiating neural cells into DA neurons and maintain the phenotype specification [62, 63]. The presence of GDNF in the induction system mainly attributes its biofunctions to midbrain dopaminergic neurons [64, 65], such as the promotion of proliferation and specification, neurite growth, synaptic and electrophysiological maturation, soma expansion and expression of phenotype-specific proteins as well as regulating downstream effector genes. Of critical importance is our application of three components (e.g., TGFβ3, VPA and SB) required for cell lineage transdifferentiation [66, 67]. Other combined factors may also play synergistical roles in hSSC reprogramming. As a result, epigenetic switching caused by a complex interaction among these distinct molecules eventually achieves the conversion of hSSCs into DA neurons.

Although it is unclear how the minimal set of defined factors induce hSSCs to transdifferentiate into DA neurons in a short time frame, our present results have shown that our approach is efficient for generating a high number of functional DA neurons both in vitro and in vivo. To further elaborate the conversion, we assayed global transcriptome profiles and activation of several crucial proneurogenic factors before and after hSSC transdifferentiation in addition to conventional morphological traits and phenotypic characteristics. More importantly, it is imperative to determine whether the converted cells function as genuine DA neurons (e.g., synapse formation, dopamine release, neuron-specific calcium imaging and electrophysiology, and rescuing deficits after grafted in an animal model of PD). As expected, these converted hSSCs acquire cellular characteristics that assemble those of genuine DA neurons without undergoing an ESC process. Strikingly, the transdifferentiation of hSSCs toward DA neurons exhibits high efficiency due to dramatic epigenetic alterations revealed by a gain of neural cell attributes and loss of SSC fate. In particular, there is a high similarity in epigenetic genes (accounting for 88.9%) and activation of several key determinants (e.g., EN-1, Pitx3, Foxa2, Lmx1a/b, and OTX2) required for the development and specification of DA neurons [62]. The results indicate that the induction process appears to mimic the dopaminergic neurogenesis that occurs during embryonic development, albeit with limited precision. In addition, we traced the fate of engrafted hSSCs up to 4 weeks posttransplantation and found that the engrafted hSSCs could differentiate into DA neuronal lineages without forming tumors. This is likely to due to direct in vivo transdifferentiation of SSCs without ESC or intermediate precursor process, which have been strengthened by double staining for nestin and cyclinD1 in the grafts. Inspiringly, the grafted cells exhibited extensive arborization and irregular somata with robust neurite and remarkably enhanced behavioral improvement in MPTP-lesioned mice as compared to controls. Our results further revealed excellent safety and efficacy in a rodent model of PD using SSC-derived DA neurons, which may be due to direct induction without the introduction of ectopic genes. This finding has also been supported by our previous reports [20]. Thus, the present study may provide an implication for the application of SSC-based therapeutic covering a vast array of regenerative medicine disciplines.

Conclusions

In summary, this results presented here showed that we successfully induced the differentiation of hSSCs to DA neurons using OECCM combined with a set of defined factors. The differentiated cells exhibited morphology and functionality similar to genuine DA neurons, such as spontaneous action potentials, DA release, and rescuing behavioral deficits of PD animals. Of unusual significance, the direct conversion of hSSCs to DA neurons by means of several active molecules without the instruction of exogenous genes is thus relatively safe and circumvents numerous concerns in future clinical applications. Overall, this innovative approach provides an extremely attractive solution to significant deficits in PD cell-replacement therapy. Also, this approach may serve as a general strategy for the generation of many distinct neuronal subtypes.