We are in the early stages of evaluating iPSC therapy in stroke and relatively few studies have been done (Table 9.1). Early efforts at iPSC transplantation for stroke involved direct injection of undifferentiated iPSC into the region of the infarct. Transplantation of mouse undifferentiated iPSC into the lesioned striatum and cortex 24 h after transient (30 min) middle cerebral artery occlusion (MCAo) in mice treated with cyclosporine resulted in teratoma formation by 28 days and worse functional outcome [17, 18]. The tumors were much larger in the postischemic brain than in the unlesioned cortex in sham-operated mice. Conversely, Jiang et al. transplanted human undifferentiated iPSC into the area of the striatum after MCA stroke in female rats [19]. Seven days after stroke, they treated animals with iPSC, fibroblast (cell control), or phosphate-buffered saline (PBS) control. They found iPSC treatment resulted in an improvement in functional outcome as early as 4 days post transplant and continuing through 16 days post transplant. iPSC transplantation also reduced infarct size. The iPSC had migrated from the injection site along the corpus callosum toward the infarct, and they showed differentiation into “neural cells.” However, they only showed staining for the astrocyte marker glial fibrillary acidic protein (GFAP) and the progenitor marker nestin, which is not specific for neurons. Surprisingly, they found no tumors or teratomas in their animals, which were all immunosuppressed with corticosteroids. Similarly, Chen et al. transplanted iPSC with fibrin glue (FG) cells into the subdural space of rats in the area near the infarct after MCAo [20]. The iPSC-FG cells reduced pro-inflammatory and increased anti-inflammatory cytokines in the brain, reduced infarct size, and improved functional outcome. There was no evidence of teratoma formation. However, when iPSC without FG were transplanted into the brain before the cerebral infarct, they improved functional outcome but all animals at 1 month showed evidence of teratoma formation. The subdural delivery method avoided potential injury of the brain parenchyma during injection of the cells and the subdural transplantation method with iPSC-FG reduced teratoma formation. It is unclear what factors limited the potential of iPSCs to form tumors in these studies, yet it would be of significant interest as future studies of iPSC-derived therapies progress. There is precedent for limited tumor formation upon transplantation of iPSCs into ischemic tissue. In the heart, mouse iPSC engrafted and improved functional outcome (echocardiography) after myocardial infarction in the immunocompetent host without forming teratomas. However, teratomas did form when these cells were transplanted in immunodeficient hosts, and this was associated with worsening cardiac function [21]. Immunocompetence in the host recipients ensured controlled iPSC engraftment with tissue integration without perturbing electrical homeostasis and without teratoma formation.

Clearly, transplantation of undifferentiated iPSC is highly unlikely to be a viable option as a cell therapy due to the high levels of tumorigenicity associated with iPSC and the significant risk of tumor formation in patients. On the other hand, terminally differentiated neurons do not engraft and survive after transplantation so that an “intermediate” stage of differentiation is required. Therefore, most transplantation strategies of iPSC-derived cells in stroke have employed cells developmentally restricted to the neural pathway, alternatively termed “neural progenitor cells„ (NPC), “neural stem cells„ (NSC) or “neuroepithelial-like cells„ (NES). These cells give rise to neurons and glial cells.

Chang et al. were among the first to demonstrate that transplantation of iPSC–NPC after ischemic stroke can promote functional recovery. They transplanted human iPSC–NPC into the contralateral striatum 7 days after an MCAo in a rat suture occlusion model. They tracked these cells with MRI and found they migrated toward the peri-infarct area on the contralateral side [22]. The migrating cells expressed C-X-C chemokine receptor type 4 (CXCR4), a ligand for stromal cell-derived factor 1 (SDF-1) and this SDF-1 CXCR4 axis is known to play a key role in stem cell migration. The iPSC-NPC differentiated into multiple types of mature neurons, astrocytes, and oligodendrocytes and integrated into damaged tissues. In addition, cell transplantation led to reduced inflammation and gliosis and improved functional outcomes on a battery of tests. Transplantation of iPSC-NPC also led to increased proliferation and migration of endogenous NSCs derived from the subventricular zone and led to an overall increase in neurogenesis in the ischemic brain. No evidence of tumor formation was found.

There is a publication bias for “positive” preclinical studies in experimental stroke and for other disease states, while negative studies are underreported in the literature [23, 24]. A “negative study” with iPSC–NSC was reported by Jensen et al. [25]. They transplanted 250,000 iPSC–NSC or PBS (group size of ten) into rats 7 days post MCAo. Grafting was observed in eight of ten animals. The cells displayed a neuronal phenotype and unbiased stereology of the tissue showed double the number of injected cells with many of the cells being positive for the proliferation marker Ki67. Although there was no tumor formation, there was graft overgrowth in some of the animals. They found no functional improvement out to 20 days using a battery of tests that included the elevated body swing test, adhesive removal, and the cylinder test and found no reduction in infarct size. These negative results may be related to the differentiation stage of NSC transplanted. Given the high proliferation of the NSC in vivo, the NSC may have been “too undifferentiated”. Previous studies with NSC derived from ESC showed that graft overgrowth in the host was related to the in vitro differentiation state and proliferative rate in the transplanted cells and not the ischemic environment of the host tissues [26].

Oki et al. and the Lund group reported their work with iPSC differentiated in vitro into long-term neuroepithelial-like stem cells (lt-NES) [27]. They transplanted the lt-NES cells into C57 BL mice striatum 1 week after a 30 min MCAo. The mice were treated with cyclosporine to reduce graft rejection. At 10 weeks, 7 of 12 mice showed engraftment of the cells with only 10 % of cells surviving engraftment. Within 1 week after transplantation, mice receiving the lt-NES performed better on the staircase test than mice receiving vehicle injection but they found no difference in the corridor test. They also found no association with long-term engraftment of cells and performance on this test. This lack of association and the early improvement at 1 week made it unlikely that the improved functional performance was related to neuronal cell replacement and more likely due to the production of regenerative or neuroprotective factors. This was supported by the observed increase in vascular endothelial growth factor (VEGF) a known stimulator of angiogenesis, neurogenesis, neuroprotection, and glial growth [28] near the lt-NES transplant. However, they were not able to detect any increased blood vessel growth indicating the VEGF effect was not via increased angiogenesis, but through an alternative mechanism. They also demonstrated through flurogold studies that lt-NES-derived neuronal cells were capable of producing long axonal extensions from the striatum into the globus pallidus suggesting that these cells are capable of complex integrative activity.

Oki et al. also transplanted NES into the striatum of nude rats 48 h after 30 min MCAo and found that about 50 % of the engrafted cells survived at 4 months although not every rat showed engraftment. The majority of the cells (72 %) at 4 months expressed neuronal nuclei (NeuN), indicating a mature neuronal phenotype, while only a minority (6 %) were positive for the astrocyte marker GFAP. They also transplanted lt-NES into the cortex of 10 nude rats 48 h after they underwent distal MCAo. Seven survived and two were sacrificed at 2 months and five at 4 months. Graft survival was about 80 % at 2 months and 60 % at 4 months. The majority of cells (77 %) expressed NeuN at 4 months but about 5 % expressed dopamine- and cAMP-regulated neuronal phosphoprotein (DARPP)-32, a marker of striatal neurons, and similar to the percentage found after striatal transplantation. This finding that the site where the cells were transplanted (e.g., cortex vs. striatum) did not influence their final fate led the authors to conclude that the in vitro differentiation of cells was a more important determinant of fate than the in vivo environment in which they were transplanted. Perhaps the most significant finding of this study was that electrophysiological recordings of transplanted lt-NES in brain slices showed functional neuronal activity. In two of the ten engrafted cells, excitatory α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor-mediated currents could be evoked by stimulating a cortical region remote from the transplant. These data suggest that iPSC-derived NES were capable of successfully integrating into the brain synaptic circuitry.

While lt-NES improved functional outcome and showed engrafted and evidence of integration into functional circuitry, there remained a question of whether more differentiated neurons such as progenitors differentiated into cortical neurons in vitro might be a more optimal cell replacement. To test this hypothesis, a distal MCAo in nude rats was used to produce a mostly cortical infarct. Forty-eight hours after stroke, either “fated cells” differentiated into cortical neurons in vitro or “nonfated” lt-NES were transplanted into the cortex. The fated cells had a lower proliferation rate but more efficient conversion into neurons with greater pyramidal cell morphology. Fated cells also possessed a layered cortical structure and survived long term (2 months). Both the fated and nonfated neurons improved bilateral function in the stepping test compared to vehicle-treated controls. Electrophysiological recording in brain slices at 20–25 weeks post transplantation showed that both fated and nonfated cells showed functional properties of mature neurons and were able to fire action potentials induced by either depolarizing current injection or spontaneously. Moreover, following electrical stimulation of the intact cortex adjacent to the transplant of fated cells, a monosynaptic-evoked response was recorded in the fated cells suggesting functional synapses.

A major concern of iPSC is that the reprogramming genes used to generate these cells are often integrated into the genome and many of these genes are tumorigenic. To overcome this limitation, Mohammad et al. used vector-free and transgene-free human iPSC and iPSC–NPC differentiated from these iPSC [29]. These cells were transplanted into mice 7 days after they underwent distal MCAo with bilateral transient common carotid artery (CCA) occlusion resulting in a barrel cortex infarction. iPSC–NPC improved recovery of sensorimotor function as measured by the adhesive removal test and led to increased brain-derived neurotrophic factor (BDNF; a factor known to have neuroprotective effects in stroke and stimulates survival, growth, and synaptogenesis) levels. Intrinsic optical signals (IOS) evoked by whisker stimulation at the barrel cortex were assessed 30 days after transplantation and indicated that transplanted iPSC–NPC restored neurovascular coupling. No tumor formation was found up to 12 months. Liu et al. showed similar results with transplanted neural precursor cells (NPreC) derived from a “novel mouse iPSC” line reprogrammed from mouse fibroblasts using a virus and oncogene-free method with two factors under conditions of hypoxia [30]. These iPSC–NPreCs after transplantation into a mouse stroke model differentiated into neurons and astrocytes and improved functional outcome. These results demonstrate that safe and efficacious cells can be derived utilizing a nonviral and nonintegrating iPSC reprogramming approach.

Since stroke generally occurs in older patients, it is important to test cell therapies in aged animals. Tatarishvili et al. from the Lund group transplanted iPSC–NES in aged rats (24 months) 48 h after distal MCA occlusion [31]. They found that 49 % of the injected cells engrafted at 8 weeks and differentiated into mature gamma γ-aminobutyric acid (GABAergic) neurons. The iPSC–NES improved functional outcome as measured by the cylinder test at 4 and 7 weeks and reduced microglial activation.

One of the important issues in transplant of iPS–NP is the optimal time to transplant in relation to the stroke event. Previous work with NSC (derived from fetal cells) showed that transplantation in the striatum after MCA stroke at 48 h was superior to transplantation at 6 weeks in terms of graft survival [32]. This was thought to be related to the more established and vigorous immune response with microglial activation at 6 weeks. While the 48-h time point was more effective, this is not a “clinically practical” time to transplant in humans. Logistically, it would be difficult to prepare the patient or the cells in such a brief period. Studies of iPSC with transplantation times of 7 days have shown engraftment and improvement in functional outcome indicating that this time point of transplantation is effective [27, 29]. Even 7 days would be logistically challenging for the production of therapeutic cells. Even the most rapid protocols for iPSC production require weeks [33]. Then iPSC must be expanded and differentiated, which again requires weeks before cells reach an NSC fate. Ultimately, these cells must still undergo rigorous testing of plasticity, homogeneity, and karyotype, under good manufacturing practice (GMP) conditions to ensure a safe product.

There is paucity of data on optimized cell dose, and there are no dose response studies of iPSC transplantation in stroke. Most of the doses of iPSC–NSC have been extrapolated from dose response studies done with NSC derived from fetal cells in the rat. Darsila et al. transplanted 300,000 (2 × 150,000), 750,000, and 1.5 million NSCs after stroke in the rat and found that doses above 300,000 were not associated with any increased graft survival suggesting that 300,000 cells were the maximum number in the rat [32]. As can be seen in Table 9.2, the dose range for transplantation in rodents ranges from 2 × 10⁵ to 1 × 10⁶. There are no dose response studies in large animals, and it is not clear if the number of cells in the rat can be multiplied by the ratio of human to rat brain size to find an appropriate dose in man.