Conceived and designed the experiments: LDF. Performed the experiments: LRN DF FG MB VRM LDF. Analyzed the data: LRN DF FG MB VRM ALV LDF. Contributed reagents/materials/analysis tools: GT ALV. Wrote the paper: LRN DF DD LDF.
The authors have declared that no competing interests exist.
Understanding the physiology of human neural stem cells (hNSCs) in the context of cell therapy for neurodegenerative disorders is of paramount importance, yet large-scale studies are hampered by the slow-expansion rate of these cells. To overcome this issue, we previously established immortal, non-transformed, telencephalic-diencephalic hNSCs (IhNSCs) from the fetal brain. Here, we investigated the fate of these IhNSC's immediate progeny (i.e. neural progenitors; IhNSC-Ps) upon unilateral implantation into the corpus callosum or the hippocampal fissure of adult rat brain, 3 days after global ischemic injury. One month after grafting, approximately one fifth of the IhNSC-Ps had survived and migrated through the corpus callosum, into the cortex or throughout the dentate gyrus of the hippocampus. By the fourth month, they had reached the ipsilateral subventricular zone, CA1-3 hippocampal layers and the controlateral hemisphere. Notably, these results could be accomplished using transient immunosuppression, i.e administering cyclosporine for 15 days following the ischemic event. Furthermore, a concomitant reduction of reactive microglia (Iba1+ cells) and of glial, GFAP+ cells was also observed in the ipsilateral hemisphere as compared to the controlateral one. IhNSC-Ps were not tumorigenic and, upon in vivo engraftment, underwent differentiation into GFAP+ astrocytes, and β-tubulinIII+ or MAP2+ neurons, which displayed GABAergic and GLUTAmatergic markers. Electron microscopy analysis pointed to the formation of mature synaptic contacts between host and donor-derived neurons, showing the full maturation of the IhNSC-P-derived neurons and their likely functional integration into the host tissue. Thus, IhNSC-Ps possess long-term survival and engraftment capacity upon transplantation into the globally injured ischemic brain, into which they can integrate and mature into neurons, even under mild, transient immunosuppressive conditions. Most notably, transplanted IhNSC-P can significantly dampen the inflammatory response in the lesioned host brain. This work further supports hNSCs as a reliable and safe source of cells for transplantation therapy in neurodegenerative disorders.
The isolation of multipotent neural stem cells (NSCs) from the human central nervous system (CNS) has spurred the investigation of new cell-therapy approaches for brain injuries and neurodegenerative diseases. NSCs, which reside in specialized regions of the adult CNS, in particular in the subventricular zone (SVZ)
Also owing to the resilience of hNSCs (human neural stem cells) to expansion ex vivo, a relatively limited number of studies has investigated the use of hNSCs for the experimental treatment of cerebral ischemia
In this paper, we demonstrate that the IhNSC's immediate progeny, represented by neural progenitors undergoing early differentiation phases (IhNSC-Ps) exhibit widespread integration ability and long-term survival when transplanted into the brain of adult rats lesioned by transient global ischemia. IhNSC-Ps generated both glial cells and mature neurons, both in the cortex and the corpus callosum. We also found that IhNSC-P-derived neuronal cells were able to establish heterotypic synaptic junctions with the host tissue after 4 months from transplantation.
Although several studies have reported a weak host' immunogenic response against transplanted hNSCs and their progeny in the brain, this issue has never been unraveled
All animal experimental protocols were approved by the Ethics Review Committee for Animal Experimentation of the Italian Ministry of Health (protocol number 37/2007-B). Adult male Sprague-Dawley rats (350–400 gr) were anesthetized with ketamine (60 mg/Kg) and Xylazine (10 mg/Kg). The common carotid arteries were exposed bilaterally by means of a ventral midline incision and occluded with microvascular clips for 10 minutes. The body temperature of the rats was mantained at 37°±0.5°C by a heating pad provided with a rectal probe. All physiological parameters were monitored and recorded throughout the surgery with BIOPAC Data Acquisition System. During the 10 minutes of carotid occlusion, mean blood pressure was maintained at 50 mmHg by withdrawal of blood from the femoral artery previously exsposed and incannulated with PE50 tubing connected to a BIOPAC system and to a collector. After the removal of the clips from the carotid arteries, the blood was reinjected into the femoral artery. After the surgery, the rats were daily treated with subcutaneous injections of antibiotics (Enrofloxacin 10–15 mg/Kg) and painkillers (Carprofen 5 mg/Kg) for one week.
To generate IhNSC-Ps for transplantation, IhNSC neurospheres, cultured as described in De Filippis et al. 2007, were mechanically dissociated and transferred onto laminin (Roche, Base, Switzerland,
Characterization was performed by immunostaining assays with primary antibodies β-TubulinIII (β-Tub, TUJ-1, 1∶400, Covance), Galactocerebroside C (GalC, 1∶100, Chemicon), Glial fibrillary acidic protein (GFAP, 1∶500, Chemicon), Green Fluorescent Protein (GFP, 1∶500, Sigma), Microtubule-associated protein 2 (MAP2, 1∶200, Sigma), Doublecortin (DCX, 1∶200, Santa Cruz) and Neural Cell Adhesion Molecule (NCAM, 1∶100, Santa Cruz).
Transduction of IhNSC with a lentiviral vector carrying the gfp gene was carried as described in
Experimental design (
In the paralel a set of lesioned animals was transplanted with GFP+ IhNSC-Ps in the periventricular region next to cc (n = 3 each time point) or in the hippocampal fissure (n = 3 each time point), constitutively immunosuppressed and sacrificed 1, 3 and 4 months later.
Rats were anesthetized with an intraperitoneal injection of ketamine (60 mg/Kg) and Xylazine (10 mg/Kg), placed in a sterotactic frame (David Kopf Instruments, Tujunga, CA) and injected with 2 µL of cell suspension (1×105 cells/µL control medium) using a Hamilton syringe to the hippocampal fissure (anteroposterior: −5.3; lateral: +3.0; dorsoventral: −3.0) or to the posterior periventricular region in the cc (anteroposterior: −5.3; lateral, +3; dorsoventral: −2). All animals were immunosoppressed with Cyclosporine A (15 mg/Kg; Sandimmun, Novartis) administered subcutaneously starting 2 days before transplantation and for the duration of the study or for 14 days for transient immunosuppression experiments.
Rats were anestethized with an intraperitoneal injection of Avertin (300 mg/Kg) and transcardially perfused-fixed with 4% paraformaldehyde. Brains were fixed overnight in 4% paraformaldeyde at 4°C, then sequentially transferred in 10%, 20% and 30% sucrose solutions. Brains were then cryopreserved (Killik, Bio-Optica, Italy), frozen and stored at −80°C. Coronal sections (18 µm thick) were obtained using a cryostat, transferred onto Super Frost/Plus object glasses (Menzel-Glaser, Braunschweig, Germany) and stored at −20°C. Sections were let dry at room temperature for 1 hour, rehydrated in phosfate-buffered saline and blocked with phosphate-buffered saline containing 10% Normal Goat Serum and 0,3% Triton X-100 for 90 minutes at room temperature. The following primary antibodies and dilutions were used: Human Specific Nuclei (HuN, 1∶100, Chemicon), β-TubulinIII (TUJ-1, 1∶400, Covance), Gamma-aminobutyric acid (GABA, 1∶500, Sigma), Glial fibrillary acidic protein (GFAP, 1∶500, Chemicon), Green Fluorescent Protein (GFP, 1∶500, Sigma), Glutamate (GLUTA, 1∶500, Sigma), Microtubule-associated protein 2 (MAP2, 1∶200, Sigma), Doublecortin (DCX, 1∶200, Santa Cruz), Neural Cell Adhesion Molecule (NCAM, 1∶100, Santa Cruz), human specific Ki67 (Ki67, 1∶200, Novocastra), Iba1 (1∶100, Wako). The fluorescent secondary antibodies used were labelled with Cy3 (1∶800, Jackson), Cy2 (1∶200, Jackson), Alexa Fluor 546 and 488 (1∶800, Molecular Probes). DAPI (ROCHE) was used as nuclear marker. Immunofluorescence-labeled sections were viewed under a fluorescence microscope (Zeiss Axioplan 2 imaging) and a confocal microscope (Leica Dmire2).
The percentage of dying cells was assessed by counting the pyknotic nuclei over total nuclei into the CA1 layer of lesioned and healthy control animals in serial brain sections (each 200 µm) as described below (n = 3 rats/time point).
At different time points, the rate of survival of IhNSC-Ps was evaluated by counting GFP+ or HuN+ cells in serial brain sections (each 200 µm apart) spanning the graft area of n = 3 rats per time point. The total number of surviving transplanted cells was calculated for the whole graft using Abercrombie formula
The percentage of Iba1+ or GFAP+ cells was counted by sampling three field in the hippocampal region of healthy or lesioned rats at 3, 7, 14 and 30DAI.
For all the quantifications an average number of 3 sections was counted per rat spanning about 500 µm along the antero-posterior axis.
Statistical analysis was performed by one-way ANOVA. Data is reported as means±SEM. (*P<0.05. **P<0.01, ***P<0.001).
Animals were perfused with 4% paraformaldehyde in phosphate buffer (0.12 M, pH 7.4). Brain samples were then cut using a vibratome (section thickness 30–40 µm). Free-floating brain slices were washed in Tris-buffered saline (TBS) and pre-incubated in 3% goat normal serum (NGS) in TBS for 30 min. Cells were also fixed with 4% formaldehyde and rinsed in TBS. Sections and cells were subsequently incubated with a rabbit anti-GFP antibody (1∶250 for brain slices, 1∶650 for cell cultures; Chemicon International) overnight at 4°C in 1% NGS/TBS. A secondary antibody (goat anti-rabbit HRP-labeled, 1∶250 dilution; PerkinElmer, Boston, MA) was used for 1 h at room temperature before developing the immunoreactive signal determined by the reaction of 3,3′-diaminobenzidine tetrahydrochloride (DAB; Sigma, St. Louis, MO) with H2O2.
Immunolabeled samples were post-fixed in 1% OsO4 in cacodylate buffer (0.12 M, pH 7.4) for 45–60 min, dehydrated and embedded in Epoxy resin. Ultramicrotome (Ultracut E, Reichert-Jung) 60 nm sections (both rat hippocampus and cultured cells) were then examined by a Philips CM 10 transmission electron microscope. Images were taken with a Mega View II digital camera (Soft Imaging System).
We have previously shown that IhNSC transplantated into the immunodficient SCID mice brain can survive for as long as 6 months
(A–B) Damaged cells at 14DAI (pyknotic nuclei, arrows) versus normal CA1 cells (B arrowhead). (C) Quantification of damaged cells in control and lesioned animals. (D–F) Iba1+ microglial cells with ameboid (lesioned hippocampus D–E) or stellate (control F) morphology. (G–H) Increase of microglia (Iba1+ cells) in the lesioned hippocampus (G, 7DAI) with respect to control (H). (I–L) Morphology of astroglial cells in the lesioned hippocampus (I, 7DAI) and control (L). Scale bars: A, B: 5 µm, D, E and F: 10 µm, G, H, I and L: 50 µm.
The lesion generated by transient global ischemia in the central nervous system is widespread and involves most brain districts. Notwithstanding, cortical areas and the CA1 layer of the hippocampus (
In order to contribute to the neural regeneration in the early phases following tissue damage, IhNSC-Ps cells were transplanted nearby the CA1 layer soon after lesioning. Previous results with various transplantation paradigms in several animal models have shown that transplantation of undifferentiated NSCs cells from neurospheres generate mainly glial progenitors upon engraftment
The IhNSC-Ps were injected into the posterior periventricular region, next to the corpus callosum or in the hippocampal fissure of rats at 3 DAI (
(A–C″) IhNSC-Ps into the cc at 14 (A) and 30 (B) days post transplant. Long processes from donor cells directed toward the cc (C′) and the upper cortical layer (C″). (D–F) Distribution of IhNSC-Ps along the SGZ at 14 days (D) and migrating to the lower SGZ at 30 days (E). Single GFP-IhNSC-Ps with tipical stem cell phenotype in the SGZ (F, arrow). Confocal analysis of colocalization of HuN with GFP (inset in F). cc: corpus callosum, GZ: granular zone, SGZ: subgranular zone, DG: dentate gyrus. Scale bars: A, F, F inset: 10 µm, C′ and C″: 20 µm, B, D, E: 50 µm.
At 14 DAI, IhNSC-Ps injected into the hippocampus were found to integrate into the DG (n = 3) (
This analysis demonstrates that IhNSCs efficiently survive in vivo and that their engraftment and migration capacities are improved in a lesioned brain, which is consistent with previous results showing that injury generates a local environment permissive for the integration of xenotransplanted cells
Next, we evaluated the differentiation of IhNSC-Ps into specific neuronal and glial phenotypes following transplantation into the ischemic environment by analyzing the colocalization of selective markers for neurons, astroglial and oligodendroglial cells with the anti-human specific antibody anti-huN. This was carried out on IhNSC-Ps that were not tagged with GFP, in order to rule out possible effects on their differentiation properties, as consequence of viral transduction with the GFP expression construct.
At one and three months post transplantation, IhNSC-Ps migrating through the corpus callosum and localizing into the DG were found to be relatively immature neuronal cells, expressing NCAM protein (corpus callosum,
(A–F) IhNSC-Ps at 3 months post transplantation differentiate into both NCAM+HuN+ (A, cc) and Dcx+/GFP+ (B, dentate gyrus) neuronal progenitors and β-Tub+HuN+ (C) and MAP2+HuN+ (D) mature neurons in the cortex. (E) Sporadic proliferating cells (Ki67+HuN+, arrow and inset magnification). (F) GFAP+HuN+ astroglial cells in the cortex. (G–H) IhNSC-Ps at 1 month from transplantation generate GABAergic (GABA+HuN+, arrow in G and inset magnification) and Glutamatergic neurons (GLUTA+HuN+, arrow in H and magnification). Scale bars: A, B, C, E, F: 20 µm, D, inset in E, G, H, inset in H: 10 µm, inset in G: 5 µm.
Altogether, the results above show that IhNSC-Ps can differentiate towards the neuronal and astroglial lineages in the ischemic brain. Both immature migratory neuroblasts and more mature β-Tub+ and MAP2+ neuronal cells are produced throughout this process.
We have previously shown that IhNSCs
At the later time tested, i.e. 4 months (
(A) Map of the brain areas colonized by IhNSC-Ps in transiently immunosuppressed ischemic rats 4 months following transplantation. Letters in boxed area refer to the figures B–E. IhNSC-Ps migrate extensively through the cc (B–C) and along the dentate gyrus (D, E and H) to the underling SVZ (D, E). F and G magnifications of boxed areas in E. Scale bars: B,C,E,F,G and H: 50 µm; D: 100 µm.
The CA1 layer of the hippocampus is one of the areas most prominently damaged in transient global ischemia. In addition, neurons newly generated by adult neurogenesis in the CA1 pyramidal layer also die, due to the persistence of inflammatory conditions
(A) Schematic map of the hyppocampal layers colonized by IhNSC-Ps in transiently immunosuppressed ischemic rats. The letters are positioned next to the regions referred to the figures B–E. (B–D) At 4 months following transplantation IhNSC-Ps were found integrated into the CA3 layer and in the underling SVZ (B), emitting long processes toward the dentate gyrus (boxed area in B, shown at higher magnification in (C) and CA1 layer (B–D). (E) At this time IhNSC-Ps were also found distributed along the CA1 layer. Scale bars: in B–D: 50 µm, in E: 30 µm.
Since neither electrophysiological recordings on adult rat brain slices nor high resolution immunofluorescence analysis could be performed because of the ischemia-induced decay of tissue cytoarchitecture, we assessed the ultrastuctural features of GFP-expressing IhNSC-Ps progeny located in the CA1 layer, 4 months after transplantation (
Representative electron micrographs of rat hippocampus (A–E) and IhNSC-Ps differentiated
This supports the notion that surviving IhNSC-Ps progeny have the ability to integrate into the lesioned CA1 area, therein establishing heterotypic synaptic junctions with host cells.
NSCs can act as immunomodulators in pathological, inflammatory brain environments
(A–B) Charts showing the effect of IhNSC-Ps transplantation on the number of microglia (Iba1+, A) and astroglial cells (GFAP+, B) in the hippocampal region at 7, 14 and 30 DAI. (C–H) Representative images of the hippocampal regions, showing the morphology and density of Iba1+ cells (green) and GFAP+ cells (red) in not transplanted (C, E, and G) and transplanted (D, F and H) brain hemispheres at 7, 14 and 30 DAI. Abbreviations: cc = corpus callosum. Scale bars: 50 µm.
It is worth noting that immunosuppression by cyclosporine, be it administered transiently or even continuously, did not affect inflammatory response in lesioned animals, as assessed by immunofluorescence using anti-Iba1 antibody (not shown).
In the adult brain, ischemia and brain trauma increase neurogenesis in the SVZ and migration of newly generated NSC-derived progenitors to the sites of injury
IhNSCs can generate significant percentages of mature neurons and oligodendrocytes
We found that, even upon the transient immunosuppression conditions used here, IhNSC-Ps integrated into the cortex, corpus callosum and DG of the hippocampus as early as 14 DAI, also migrating along preferential neurogenic pathways
At least 19% of the grafted human cells survived 1 month after transplantation, a percentage similar to that observed after 4 months. This would suggest a quite stable profile of integration and survival of the transplanted cells over a quite long period. Nonetheless, it is also possible that this apparent stability in the overall number of grafted cells may be the consequence of a dynamic balance between two competing processes, the death of engrafted cells and the birth of new ones through cell proliferation. In fact, it is well known that stroke-associated hypoxia enhances the proliferation of neuronal precursors
Only 7 out of 28 of the transplanted control (unlesioned) animals showed appreciable cell engraftment and survival with respect to 37 out of 61 in lesioned animals. Furthermore, in control animals, engrafted cells were mainly localized next to the injection site (not shown). These results are in good agreement with previous studies, showing that both IhNSC-Ps survival and migration are enhanced by the presence of a brain lesion as compared to the healthy CNS tissue
Others and us have previously shown that NSCs undergo prevalent glia differentiation after transplantation in neurodegenerative disease animal models such as metachromatic leukodystrophy (MLD)
A wide array of studies have shown that NSCs are not susceptible to immunological rejection
The succesful use of transient immunosuppression described here, supports the twofold notion of limiting toxicity in an experimental model plagued by high animal mortality and of preventing the bias introduced by the known neuroprotective effects of cyclosporine following hypoxia-ischemia
Our ultrastructural analysis determined the full maturation of IhNSCs progeny by detecting the presence of newly established synaptic junctions between rat axonal terminals and IhNSC-Ps progeny dendritic spines in the CA1 layer, 4 months after transplantation. This is consistent with previous observations, showing the ability of IhNSCs to generate post-synaptic structures and to fire spontaneous action potentials in culture
Besides neurodegeneration
Transient global ischemia is a commonly accepted model of vascular dementia, since it resembles the pathological features of Alzheimer's Disease. In this view, the findings presented in this manuscript lend to the idea of using IhNSCs as a suitable tool to model transplantation of hNSCs in pre-clinical settings. This is particularly relevant in view of the fact that the first phase I clinical trial exploiting cell therapy has been authorized and is currently underway. The trial uses non-immortalized neural cells similar to those described here, which may thus be considered for a prospective use in clinical settings. This is particularly true, considering the suitable migration and differentiation pattern of our IhNSCs in the ischemic brain, their negligible rejection, their ability to establish synaptic interaction with host cells and their capacity to generate appropriate neurotransmitter phenotypes in ischemia target areas, such as the hippocampus and cortex. The ability of transplanted IhNSC-Ps to dampen reactive astrogliosis and microglia activation provide an extra positive element when considering hNSCs for therapeutic purposes in neurodegenerative disorders.
Analysis of the lesioned brain at 3DAI. (A–C) Hematoxylin-eosin showing the pyknotic nuclei present in the lesioned (A, arrows) respect to the control cortex (B), and in the lesioned CA1 layer (C). (D–E) Microglial (Iba1+, D) and astroglial (GFAP+, E) reaction in the hippocampal region of lesioned animals at 3DAI. (F) Quantification of pyknotic nuclei in the CA1 layer and of Iba1+ and GFAP+ cells in the hippocampal region. Scale bar: A and B: 50 µm, C: 5 µm, D and E: 75 µm.
(5.38 MB TIF)
(5.00 MB TIF)
Experimental design. (A) Schematic representation showing the experimental plan with transplanted animals undergoing transient or constitutive immunosuppression. Healthy not transplanted animals (n = 4) have been excluded. (B) Table showing the numerosity of the transplanted animal groups. Abbreviations: cc: corpus callosum, hf: hippocampal fissure, AP: anteroposterior, L: lateral, DV: dorsoventral.
(9.30 MB TIF)
We thank David Della Morte, Cristina Zalfa and Elena Fusar Poli for technical support in obtaining and analyzing the global ischemia animal model. We also thank Pietro De Filippis, Patrizia Karoschtiz, Cesare Rota Nodari, Loredana Turani, Antonio Tomaino, Maurizio Gelati for precious suggestions.