Abstract
Irreversible and permanent damage develop immediately adjacent to the region of reduced cerebral blood perfusion in stroke patients. Currently, the proven thrombolytic treatment for stroke, tissue plasminogen activator, is only effective when administered within 3 h after stroke. These disease characteristics should be taken under consideration in developing any therapeutic intervention designed to widen the narrow therapeutic range, especially cell-based therapy. Over the past several years, our group and others have characterized the therapeutic potential of human umbilical cord blood cells for stroke and other neurological disorders using in vitro and vivo models focusing on the cells' ability to differentiate into nonhematopoietic cells including neural lineage, as well as their ability to produce several neurotrophic factors and modulate immune and inflammatory reaction. Rather than the conventional cell replacement mechanism, we advance alternative pathways of graft-mediated brain repair involving neurotrophic effects resulting from release of various growth factors that afford cell survival, angiogenesis, and anti-inflammation. Eventually, these multiple protective and restorative effects from umbilical cord blood cell grafts may be interdependent and act in harmony in promoting therapeutic benefits for stroke.
Keywords
Introduction
A report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee ranks stroke third among all other causes of death behind heart diseases and cancer in the US (96). Each year about 600,000 new patients experience first stroke attacks, and 180,000 are recurrent attacks (96). Among ischemic stroke survivors who were at least 65 years of age, 50% suffered from hemiparesis and 26% were dependent on others to assist in their daily activities (96). Currently, the only Food and Drug Administration-approved thrombolytic drug for acute ischemic stroke is recombinant-tissue plasminogen activator (r-tPA), which acts by dissolving the blood clot in the vessel and restoring proper blood flow. However, there is only a 30–35% success rate for recovery, even though r-tPA is administered within 3 h after the initial onset of an embolic stroke (75, 84), and the drug effectiveness decreases with time compounded by considerable risks such as intracranial hemorrhage and recurrent stroke (84). Eventually, only 3–5% of stroke patients reach the hospital within this therapeutic range and of those only 29% are eligible for r-tPA (80), indicating that only a few stroke patients have an opportunity at full restoration with rapid thrombolytic management. This unfortunate clinical status clearly solicits new treatment modality with a wider therapeutic spectrum for stroke management.
The ischemic stroke cell death cascades involve two major pathophysiological events. Glutamate excitotoxicity develops in the ischemic tissue due to the sudden drop in oxygen and nutrients that begins immediately upon loss of blood flow (106). Subsequently, indirect inflammatory responses, such as generation of free oxygen radicals and nitric oxide (24, 71), activation of microglia within tissue hypoxia and necrosis (68, 107), and invasion of other inflammatory cells by way of chemotaxis and leakage across the compromised blood–brain barrier, contribute to disease progression (34, 41, 137). Thus, a more successful and effective treatment over current thrombolytic therapy must have a multifaceted anti-inflammatory and neuroprotective ability to respond to all of these insults (80) as well as a longer range of opportunity that increases the survival rate and improves functional recovery. This demand leads to exploration of new therapeutic strategies targeting the restorative stage over the acute narrow therapeutic window, such as cell-based therapeutics designed to rescue against cell loss and simultaneously provide anti-inflammatory and/or neurotrophic factors to combat the secondary neurodegeneration inherent in stroke.
The advent of stem cells catapulted the field of cell therapy into the forefront of possible treatments for neurological disorders, including stroke. Of particular interest is the availability of adult stem cells as donor cells for transplantation. Tissue-specific adult autologous stem cells such as neural stem cells (NSCs), bone marrow (BM), or peripheral blood stem cells stand as attractive alternative sources for cell-based therapies for stroke (15) because they have the capacity for self-renewal and can generate cells other than themselves, resembling the multipotent features of primitive stem cells (i.e., embryonic or fetal origin). For example, adult NSCs can differentiate into non-central nervous system derivatives, such as blood (9) or skeletal muscle cells (38) and vice versa. This plasticity of adult stem cells opens new approaches for their application in the treatment of various neurodegenerative disorders. Nonetheless, clinical application of adult NSCs, like embryonic or fetal stem cells, requires a careful preclinical evaluation of their safety and efficacy, as well as bioethical issues (27, 50). Additional technical difficulties exist, such as ensuring purity of neural cultures and the harvesting of needed populations for transplantation (91, 99). Accumulating evidence has revealed the therapeutic potential of hematopoietic stem cells in the treatment of not only blood-related diseases but also various unrelated disorders, even neurodegenerative diseases. Of these, BM is the current gold standard source of hematopoietic progenitor cells used to reconstitute blood lineages after myeloablative therapy in a number of malignant and nonmalignant blood diseases (90). Several reports have demonstrated that BM stem cells can give rise to non-blood lineage tissues including neural cells (36, 86, 97). Some disadvantages of BM as donor cells, however, include the relative prolonged timing between procurement of and treatment with BM cells (61), the considerable cost of the harvesting procedure (61), and the low availability of human leukocyte antigen (HLA)-matched donors with BM, which is crucial to avoid graft-versus-host disease (GvHD) (3). Although these limitations can be avoided by transplantation of autologous BM or peripheral blood with granulocyte colony-stimulating factor stimulation, patients with malignancies are often unable to receive autografts due to the risk of reinfection with tumorgenic cells (81). The “timing” of intervention is critical for successful stroke outcome, and therefore it may not be feasible to recover, manipulate, and process quality controlled autologous BM cells in a good manufacturing practice (cGMP) environment in time.
An efficacious substitute to BM cells is human umbilical cord blood (hUCB), which has already been used clinically for several years to reconstitute BM and blood cell lineages in children with various hematological malignant and nonmalignant diseases (105). Since the first hUCB cell transplantation to treat a 5-year-old child who contracted Fanconi anemia in 1988 (44), a number of hUCB transplants have been performed worldwide, many of them with unrelated donors (69, 118, 124). Like BM, hUCB use has thus far been limited to hematopoietic malignancies, marrow failure, and immunodeficiency disorders, but numerous reports reveal the many advantages of hUCB cells for cellular therapies compared to BM (81). This includes the fact that hUCB cells, even from allogeneic origin, are immature, and produce a lower incidence of graft rejection, GvHD, and posttransplant infections (64, 118). These characteristics have been attributed to higher numbers of immature progenitors in hUCB compared to those of adult marrow stem cells, on the basis of telomere length (120). Moreover, the immaturity of cord blood stem cells and their immune naïveté have been implicated in the optimal effects of these cells for hematopoietic and somatic organ therapy.
The following sections present recent findings on the specific “stemness” properties and therapeutic potential of hUCB as a unique, safe, and efficacious donor cell for transplantation in stroke.
Unique Features of hUCB
Cell Composition
As a source of pluripotent stem cells, hUCB has been reported to be a source of hematopoietic stem cells, endothelial cell precursors, mesenchymal progenitors, and multipotent/pluripotent lineage stem cells (7, 35, 63). Hematopoietic stem cells from hUCB are rich in the most primitive cells and are capable of long-term reconstitution of blood lineages (11, 79, 115). While the number of myeloid progenitors in hUCB is similar to the number in BM (11), hUCB cells have a greater colony-forming ability (78) and can be expanded in long-term cultures in vitro using different growth factors and have longer telomeres than adult cells (120). Moreover, it has been revealed that hUCB transplants are better at restoring the host hematopoietic progenitor cell reservoir than adult BM stem cells (37). Even a single hUCB sample can supply enough hematopoietic stem cells to provide both short- and long-term engraftment (105).
Within the mononuclear fraction of hUCB, the component is primarily composed of lymphocytes and monocytes (87). In an electron microscopic comparison of hUCB, peripheral blood, and BM cells, hUCB cells had the more immature morphology of myelomonocytic cells with small numbers of mature neutrophils (77). Compared with adult peripheral blood, hUCB resembles the B-cell population but with a lower absolute number of T cells (CD3+) and a higher CD4+/CD8+ ratio (49, 87). Moreover, when comparing the characteristics of B-cell differentiation in vitro from hematopoietic progenitors (CD34+) between cord blood and peripheral blood, B-cell precursors differentiated from cord blood are more immature (51). The relative immunological immaturity of hUCB reveals a higher proportion of immature T cells (CB45RA+), but decreased numbers of mature memory (CD45RO+) (29, 49) and CD56+ cytotoxic (29) T cells compared with other adult sources. Moreover, cord blood lymphocytes express cytokines and their receptors such as interleukin (IL)-2, IL-6, IL-7, tumor necrosis factored (TNF-α), and interferon-γ (IFN-γ) at lower levels than in adult blood cells (45, 139) and produce greater amounts of the anti-inflammatory cytokine IL-10 (45, 89), which downregulates expression of CD86 on dendritic cells (DCs), and may prevent initiation of the T-cell-mediated immune reaction (12). These increased levels of IL-10 may stimulate regulatory T cells, which in turn inhibit antigen-specific immune responses (5). Monocytes in hUCB are also immature because hepatocyte growth factor (HGF) activates adult monocytes, but cord blood monocytes have no response to HGF, indicating they cannot induce integrin-α5 subunit or intercellular adhesion molecule-1, both of which are necessary for normal function (59). Moreover, hUCB monocytes have a reduced expression of HLA-DR compared to adult cells (113). In addition, cell subpopulations differ between adult blood and cord blood; for example, cord blood has higher numbers of natural killer (NK) cells (29), which are able to inhibit T-cell proliferation and reduce TNF-α production (33). On the contrary, cytotoxicity from NK cells is lower in hUCB compared to adult blood (28). DCs are the sentinel cells of the body and initiate immune responses within the lymph nodes. The DCs in cord blood reflect lymphoid DCs that are most likely to be responsible for colonizing neonatal tissue, while there are more myeloid DCs in adults (128). The lymphoid DCs induce anti-inflammatory T-helper 2 cell responses, which, along with the naive T cells, may promote down-regulation of immune responses (4, 128). These immature immunological properties in hUCB cells appear to result in a prolonged immunodeficient state after hUCB transplantation (42, 114). This leads to a low incidence of GvHD and viral transmission. Additionally, such cellular components could also allow for relatively flexible donor-recipient matching requirements, hence leading to a shorter preparatory period for treatment (81). Rocha et al. found that GvHD incidence was significantly lower in children receiving hUCB transplants compared to BM recipients, when both sources were from an HLA-identical sibling (95). Furthermore, even unrelated HLA-mismatched UCB recipients have a lower GvHD incidence than HLA-identical BM recipients (94).
Recently, we characterized in vitro two different subpopulations of mononuclear hUCB cells: adherent and floating (19). A significant number of progenitor/stem and neural cell antigens were found to be expressed on cells in the floating population, whereas the adherent cell population mainly contained lymphocytes (~53%) expressing hematopoietic antigens. These results suggest that a nonhematopoietic subpopulation of cells exists within mononuclear hUCB cells and seems to have the potential to become neural cells.
The mesenchymal stem cell (MSC) has also been found as a small fraction of UCB cells (30, 46, 133). It was recently confirmed that MSCs and MSC-like progenitors can be isolated from amniotic fluid, placenta, and Wharton's jelly (30). MSCs derived from hUCB show impressive plasticity and differentiate into cells of all germ-line derivatives (57, 70, 133). Under appropriate conditions for culturing and differentiation, hUCB MSCs could differentiate into osteogenic cells and chondrogenic cells (66). After incubation in neural differentiation medium, hUCB MSCs expressed the neural cell surface antigen A2B5, the neurofilament polypeptide NF 200, the oligodendrocyte precursor marker 04, intermediate filament proteins characteristic of neural differentiation (nestin and vimentin), as well as the astrocyte marker glial fibrillary acidic protein (GFAP) and the neural progenitor marker TuJ-1 (32). Additionally, supernatants from cultured hUCB MSCs promoted survival of NT2N neural cells and peripheral blood mononuclear cells cultured under conditions designed to induce cell stress and limit protein synthesis (32). We also showed immunomodulatory effects of the hUCB MSCs after coculture with murine splenocytes (32). Taken together, these results indicate that hUCB MSCs also possess multiple utilities that may contribute to their therapeutic potency in the treatment of neurological diseases.
Ex Vivo Expansion and Surface Antigens
There is only a single chance to collect hUCB cells, and the number of mononuclear cells extracted from that isolated donation is finite (81). Therefore, many researchers are currently interested in the ex vivo expansion of cord blood progenitor cells to provide a sufficient amount of stem cells for adult transplantation (72). Expansion is essential in providing adequate dosage for both children and adults, reducing the time to engraftment, thereby increasing success rate, as well as for recurrent therapy in the event of initial graft failure (81). Unfortunately, current methods of expanding hUCB cells do not preserve the quality of the hematopoietic progenitor cells through to the end product (56, 93), but also cannot make up for the cells lost in the storage process, let alone augment them to a suitable dosing regimen (52, 73, 117, 134). Moreover, simply globally multiplying the cells has been proven to not reduce the time to engraftment (56). Instead, relative proportions of cell subpopulations that home to marrow, form colonies, and repopulate blood lineages are required (81). Thus, recently, several studies about ex vivo expansion of different cell subpopulations of hUCB have been reported. Huang et al. (54) found that ex vivo expansion of hematopoietic stem cells (CD34+) of hUCB with hUCB MSCs was more effective than without MSCs. Wei et al. (127) reported that T cells and NK cells as well as CD34+ cells could be expanded from hUCB in the same medium with the combination of cytokines. The expanded mononuclear cells could reconstitute hematopoiesis and eliminate minimal residue leukemia disease in transplanted mice. Shirvaikar et al. (103) showed that the combination of thrombopoietin with IL-3 used for expansion of megakaryocytic progenitors from cord blood not only resulted in optimal proliferation but also protected these cells from apoptosis. Thus, these cells could be employed to supplement cord blood grafts and speed up platelet recovery in transplant recipients (103).
Meanwhile, there are major problems arising from the choice of an optimal stem cell population for ex vivo expansion and definition of the desired characteristics of the expanded stem cells to be used for transplantation (102). Expanded cells need to maintain telomerase length continuously in order to preserve their pluripotent capacity (134). The telomerase activity of hematopoietic progenitor cells is high and tends to be most active during proliferation (20). Determining which cells should be expanded is also tricky (81). The level of maturity of a cell is identified by the presence of, or a lack of, a combination of cell surface antigens. These are potential markers for the selection of stem cells important for expansion and implantation. For example, the mononuclear fraction of hUCB contains roughly 1% CD34+ cells (81). This CD34+ population (CD34 being a marker designating the cell as having a role in early hematopoiesis) in hUCB can be defined as more primitive than those found in BM because a higher proportion of them are negative for CD38, a marker for prelymphoid cells (14, 25). Another human antigen, CD133, has also been identified as a marker for primitive hematopoietic progenitor/stem cells that may prove to be a substitute to CD34 for the selection and expansion of hematopoietic cells for transplantation (65, 135). It has been shown that about 80% of CD34+ cells express CD133 and more than 97% of CD133+ cells are CD133+/CD34+ in fresh cord blood (48). Although CD133+ cells comprised 0.7% of the total mononuclear hUCB cells (74), expansion of CD133+ and CD133+/CD34+ cells was significantly higher than that from the CD34+ cells (48). These findings suggest that CD133+ cells may be a more primitive hematopoietic progenitor/stem cell than CD34+ cells (39). Furthermore, CD133+ cells have been identified in fetal brain and are considered to be NSCs (110, 119), even though it is not yet known whether the CD133+ cells found in hUCB are phenotypically and functionally identical to the NSCs found in fetal brain (81). In the meantime, determining a definitive phenotype for hUCB MSC has proven to be controversial, unlike that of BM. Robinson et al. (93) defined the MSC as positive for CD16, CD73, CD90, and CD105 and negative for CD31, CD34, CD45, CD80, and HLA-DR, whereas Yang et al. (133) characterized the MSC as having positive markers for CD13, CD29, CD44, and CD90 and negative markers for CD14, CD31, CD34, CD45, CD51/61, CD64, CD106, and HLA-DR. On the other hand, we showed that hUCB MSCs were negative for CD3, CD11b, CD19, CD34, and CD45 in culture (32).
Proliferation and Differentiation into Neural and other Lineages
While embryonic primitive cells have highly proliferative potential and resistance to rejection and apoptosis, which therefore may contribute to deleterious fibrosis and malignancies, cord blood stem cells are relatively quiescent under normal circumstances (116). Thus, the intrinsic proliferative and differentiatable potential of these quiescent stem cells needs to be stimulated by factors within the environment (23), without which the cells do not proliferate unrestrained. Several in vitro and in vivo reports have been published showing the multipotent nature of hUCB cells, including neural cells, thus helping to advance the potential usefulness of them in treating neurodegenerative diseases. Bicknese et al. (8) and Buzanska et al. (13) demonstrated that hUCB cells could differentiate to express βIII-tubulin, GFAP, and galactosylceramide (GalC). Of particular interest, Buzanska et al. (13) used the CD34–/CD45– nonhematopoietic mononuclear cell fraction of hUCB to obtain neural stem cell-like cells. From these, a clonogenic line of hUCB NSCs, which expressed nestin and GFAP, was derived. In the presence of neuromorphogen/retinoic acid, 40% of the hUCB NSCs expressed βIII-tubulin and microtubule-associated protein (MAP)-2 (neuronal marker), 30% expressed GFAP and S100β (astrocytic marker), and 11% expressed GalC (oligodendrocytic marker). Retinoic acid plus brain-derived neurotrophic factor has also been shown to induce neural differentiation in hUCB NSC culture. During 7 days of coculture with neuromorphogens, rat astrocytes, or hippocampal slices, 80% of cells expressed βIII-tubulin and 64% coexpressed MAP-2 (60). Goodwin et al. (46) demonstrated a subset of mononuclear cells from hUCB, which had been maintained in continuous culture for more than 6 months without antigen expression for hematopoietic differentiation. When these cells were exposed to osteogenic, adipogenic agents, or epidermal and basic fibroblast growth factors (EGF and bFGFs), they expressed bone, fat, and neural markers, respectively. These data suggest that hUCB contains a cell population that is capable of expressing antigens of multiple lineages, demonstrating that the plasticity of these cells makes them very important for cellular therapeutics with the goal of system repair (39). Confirming the presence of a cell population in hUCB with multipotent ability, McGuckin et al. (76) developed a negative immunomagnetic selection method that depletes hematopoietic lineage marker-expressing cells from hUCB, thereby isolating a discrete lineage-negative stem cell population (0.1% of mononuclear hUCB). These selected lineage-negative hUCB cells expanded into primitive nonadherent hematopoietic progenitors and simultaneously produced adherent cells with neuroglial progenitor cell morphology over 8 weeks. Gene expression analysis showed upregulation of primitive neuroglial progenitor cell markers for GFAP, nestin, musashi-1, and necdin.
Our group also revealed that the mononuclear hUCB cells under specific conditioned media including retinoic acid and nerve growth factor (NGF) expressed various neural markers for early neural precursors (musashi-1, nestin, TuJ1), mature neurons (NeuN, MAP-2), and astrocytes (GFAP) (98). Moreover, cells exposed to retinoic acid and NGF treatment increased TuJ1 and GFAP expression by approximately twofold (98). Furthermore, Zigova et al. (138) showed that TuJ1 and GFAP immunopositive cells from hUCB mononuclear cells after treatment with retinoic acid and NGF survived in the subventricular zone of the rat neonatal forebrain. Additionally, we have demonstrated that in standard growth medium such as Dulbecco's modified Eagle medium (DMEM), mononuclear hUCB cells also express neural markers, such as nestin, TuJ1, MAP-2, and GFAP (40). Interestingly, at 2 weeks after plating, colocalization of nestin, MAP-2, and various cytoplasmic expressions of TuJ1 by the cells was observed. Moreover, the increased number of cells expressing CD133 antigen after 7 days in culture probably gives rise to cells that show immature and mature neural characteristics at the same time (39). Recently, similar to the above studies, we have examined the ability of mononuclear hUCB cells to express neural antigens after in vitro exposure to defined media supplemented with a cocktail of growth and neurotrophic factors (20). When floating mononuclear hUCB cells were exposed to a neurogenic environment through treatment with growth factors and neurotrophins, they had increased proliferation, expression of neural antigens, and also survived longer in vivo (20).
Once the cell type of interest has been determined and an adequate population of these cells has been obtained, the next step would be to transplant the cells. This raises the question of what happens to the cells following transplantation and this is discussed in the next section.
Fate of Grafted Cells: What and Why?
The cells used for transplantation into humans are preferred to be of human origin. Although xenografting of human cells into the rodent is a widely accepted research paradigm that often yields promising results (1, 125), this presents a dilemma in in vivo research with human-originated stem cells. While hUCB cells show the inability to generate cytotoxic T cells that respond to allogenic antigens, and to produce the proinflammatory cytokines such as IFN-γ and TNF-α, transplantation of hUCB cells into adult or aged rat brains was associated with vigorous rejection and strong immunosuppression was required to protect the graft (126). The limited survival of the grafts in in vivo models possibly results from immune rejection. Therefore, recently we performed studies using hUCB cells to determine their ability to survive in vivo and the effect of the immune response on the survival by transplantation into the normal striatum of immunodeficient NOD SCID mice (125). First, in vitro, long-term culture (60–211 days) of hUCB cells resulted in several different populations of cells including neuron-like cells that were positive for TuJ1 and nestin. Five days after transplantation of cells of this neuronal phenotype, many differentiated hUCB cells expressing characteristic neuronal proteins were detected. However, at 1 month postgrafting, hUCB cells were no longer detected without evidence of T-cell-mediated rejection, such as CD4 and CD8 lymphocytes and activation of microglia and astrocytes (125). These findings suggest that the disappearance of implanted cells was not due to a T-cell-mediated immune response. The hUCB cells can survive for long periods of time in vitro and they can also exhibit neural characteristics dependent on their environmental conditions, but the same does not appear to be true in vivo even though the cause of this discrepancy is unclear (125). A number of factors, such as the donor cell's condition and the recipient's environment, might simultaneously affect survival of the transplanted cells.
In summary, on the basis of several previous studies in vitro and in vivo, hUCB cells, due to their unique primitive nature and ability to differentiate or become various lineages of cells including neural, may be a useful target for cell-based therapies requiring either the replacement of individual cell types and/or provision of missing materials in neurological disorders.
Preclinical Stroke Research on Transplantation of Human Umbilical Cord Blood Cells
As mentioned, although the timing to begin treatment for stroke is crucial (39), the therapeutic window of current stroke treatment is still restrictive. Many patients outside of this window suffer from irreversible sequela. This current circumstance has brought serious consideration to the exploration of new therapeutic interventions such as a cell-based therapy. Using adult BM mesenchymal stem cells, functional improvement has been demonstrated in a rodent stroke model (16, 18, 22, 100). While it has been reported that transplanted cells could successfully migrate to areas of ischemic infarcts and differentiate into neuronal and glial cell types, recently Chopp and collegues suggested that recovery mechanisms are likely to be due to trophic factors released by cells that may promote endogenous neurogenesis, angiogenesis, and axogenesis rather than a result of neuronal replacement (26, 88, 101, 136). Transplanted hematopoietic stem cells (CD34+), which were isolated from peripheral blood, also could differentiate into neural lineage cells and induce angiogenesis, in conjunction with improvement of the neurological impairment in a rodent stroke model (31, 104).
The In Vivo Potential of Human Umbilical Cord Blood Cells
hUCB is another promising source of multipotential stem cells that has shown affirmative effects in in vivo studies for the treatment of stroke. Chen et al. (17) showed that the intravenous administration of mononuclear hUCB cells at 24 h or 7 days after middle cerebral artery occlusion (MCAO) significantly improved neurological function in a rat model. On histological examination, mononuclear hUCB cells were found mainly in the cortex and striatum of the injured hemisphere in the ischemic boundary zone, whereas few cells were observed in the contralateral side. Some of the implanted mononuclear hUCB cells were immunoreactive for the endothelial cell marker FVIII (8%), as well as GFAP (6%), MAP-2 (3%), and NeuN (2%) by immunohistochemistry (17). Using a nonhematopoietic cell line derived from hUCB, Xiao et al. (132) also observed a decrease in infarct volume after intravenous transplantation into rats with ischemic brain injury. Histological examination showed that some of the transplanted cells were colocalized for human nuclei and NeuN. On the other hand, our group compared the efficiency of intravenous versus intrastriatal transplantation of mononuclear hUCB cells to evaluate which produced the greatest behavioral improvements in rats with permanent MCAO (129). Spontaneous activity (MCAO-lesioned animals become significantly hyperactive) significantly decreased when cells were transplanted 24 h after stroke compared with control animals. Behavioral improvement, as measured by spontaneous activity, was similar with both cell delivery routes. However, in the step test at 2 months after transplantation, significant improvements were found only after intravenous delivery of the hUCB cells. Also, in the passive avoidance test, transplanted rats learned the task faster than nontransplanted ones. These results suggest that intravenous injection of mononuclear hUCB cells may be more effective than direct striatal implantation in producing long-term functional benefits to the stroke animal. Next, we focused on the dose effect of mononuclear hUCB cells on the behavioral recovery and infarct volume in MCAO rats (121). Twenty-four hours after MCAO, rats were intravenously infused with 104 up to 3–5 × 107 mononuclear hUCB cells. At 4 weeks after transplantation, there was a significant recovery in behavior, when 106 or more hUCB cells were delivered. Infarct volume measurements showed an inverse relationship between cell dosage and injured volume, which reached significance at the higher doses of mononuclear hUCB cells. These results revealed an important dose relationship between introduced implanted cells, behavioral improvement, and neuronal sparing, using intravenous mononuclear hUCB cell injection in the MCAO model.
In addition to route and dosage, determining the therapeutic window of cell transplantation for stroke is a major concern in the preclinical research field as well as for clinical application. Our research group attempted to determine the optimal time to administer hUCB cells after stroke. Initially, in an in vitro assay system, the migration capability of mononuclear hUCB cells according to the time after stroke was investigated (82). This assay revealed an increased migratory activity of mononuclear hUCB cells towards both the hippocampal and striatal extracts harvested at 24–72 h after stroke. Furthermore, the extracts possessed increased levels of cytokine-induced neutrophil chemoattractant-1 (CINC-1) and monocyte chemoattractant protein-1 (MCP-1) at 48 h after MCAO, suggesting participation of these substances in the cell migration. Further analysis of the ischemic extracts showed that growth-regulated oncogene/CINC-1 (the rat equivalent of human IL-8) and MCP-1 were expressed in a time-dependent pattern similar to that of the migration assays. The results from this study are promising in that the current 3-h therapeutic window for the treatment of stroke victims, using approved anticoagulant treatment, may be extended with the use of mononuclear hUCB cell therapy to 24–72 h poststroke event. Also, the chemokines present in the supernatant could provide a sound starting point for examining the mechanisms responsible for the in vivo migration of mononuclear hUCB cells after stroke induction (82). More recently, we demonstrated that, in vivo, the hUCB treatment window is also narrow even though it is longer than thrombolytic agents. When we systemically administered the hUCB cells at times ranging from 3 h to 30 days post-MCAO, maximal improvements were observed with treatment at 48 h (80). These results might suggest that it is not critical for the cells to be present indefinitely after transplantation, but that they are actually only needed within a very short time frame to either appropriately prime the immune system or provide other secretory products to cells within the brain in order to confer therapeutic effects (80).
Mechanisms Underlying Recovery Beyond Neural Differentiation
Cell replacement had been believed to be the main mechanism responsible for the functional recovery after transplantation on the basis of accumulating results of hUCB cells expressing neural phenotypes and providing behavioral improvement (40, 47, 98, 132, 138), even though few cord blood cells are present in the ischemic region compared to the number of infused cells (21, 121, 122, 129, 130). However, there has been increasing evidence of the multifaceted therapeutic effects of hUCB such as neurotrophic, angiogenic, and anti-inflammatory actions, beyond the ability of transdifferentiation into neural lineage, in the MCAO stroke model (Fig. 1). These effects more likely come from the trophic and other factor-releasing capabilities of hUCB (10). Recently, it has been revealed that mononuclear hUCB-derived neuronal progenitor cells decreased free radicals and induced antioxidants and neurotrophic factors such as NGF, vascular endothelial growth factor (VEGF), and bFGF in an oxygen and glucose deprivation insult model (2).

Schematic diagram depicting how human umbilical cord blood (hUCB) grafts can beneficially influence the stroke patients. Transplanted hUCB cells may exert their beneficial effects with respect to stroke either directly on hypoxic cell death by neurogenesis (upper side) or indirectly via release of a number of factors that modulate the secondary inflammation and promote tissue repair and angiogenesis (lower side).
Initially, we focused on the anti-inflammatory action of hUCB cells. Vendrame et al. (122) demonstrated that there was an increase in the number of CD45+/CD11b+ (microglia/macrophage) and CD45+/B220+ (B cell) cells in the brain of rats with permanent MCAO, whereas mononuclear hUCB cell transplantation significantly decreased the number of these cells. The reduction in the number of CD45+/CD11b+ cells after transplantation of hUCB cells is particularly interesting because chronic microgliosis is thought to mediate neuronal damage, not only in ischemic injury but also in other neurodegenerative diseases such as Parkinson's disease (107, 111). The hUCB treatment also decreased the proinflammatory cytokines such as TNF-α and IL-1β (122). Therefore, the potential anti-inflammatory effects of UCB therapy may be one of the mechanisms that protect against neuronal death. In continuing studies, we found that, following MCAO, the rat's spleen size was reduced concomitantly with their CD8+ T-cell counts (123). Interestingly, MCAO-induced spleen size reduction correlated with the extent of ischemic damage; however, hUCB cell treatment rescued the spleen weight, splenic CD8+ T-cell counts, as well as significantly reducing the amount of brain injury. Additionally, splenocyte proliferation assays demonstrated that hUCB cell treatment opposed MCAO-associated T-cell proliferation by increasing the production of IL-10 while decreasing IFN-γ (123). Recently, we investigated the expression of cytokines and chemokines produced by hUCB cells under various culturing conditions (83). The heterogeneous cells from mononuclear fractions of hUCB consistently produced a variety of chemokines and cytokines when grown in various culture conditions, but in particular, IL-8, MCP-1, and IL-1α. They are produced more extensively than any other chemokines in the human body, especially in the brain, and have been implicated as the first line of defense in the inflammatory reaction. These findings support the idea that inflammatory modulation from hUCB transplantation may be partially responsible for the functional improvements seen in animal models of injury, including stroke.
A second potential mode of action is neovascularization around ischemic tissue, promoted by hUCB. Angiogenesis is an important healing process of wounds including tissue ischemia, and it can attenuate local tissue ischemia and improve clinical outcomes (67). Furthermore, angiogenesis appears to promote neurogenesis in neurological disease (92, 112). There is accumulating data that transplantation of hUCB cells induce neovascularization, followed by functional improvement in the animal ischemic disease model (53, 62). VEGF was increased in the experimental group (53) even though it was unclear whether transplanted hUCB cells directly released it or stimulated other cells for indirect secretion. UCB not only contains high numbers of CD34+ endothelial progenitors cells (EPCs), which can give rise to mature endothelial cells and induce angiogenesis in ischemic tissues (85), but mononuclear cells (55) or MSCs (30) of UCB also can differentiate into EPCs. Taguchi et al. (109) demonstrated that CD34+ cells derived from hUCB induced angiogenesis, followed by endogenous neurogenesis in a mouse stroke model. Ding et al. (30) revealed that hUCB MSCs differentiated into glial, neuronal, and vascular endothelial cells as well as promoted the formation of new vessel angiogenesis in the rat stroke model. Recently, we found that animals that received the monocyte/macrophage (CD14+)-depleted hUCB preparation performed significantly to a lesser degree than those receiving the other hUCB cells (stem cell-, T-cell-, B-cell-depleted, and whole mononuclear fraction of cord blood) in a rat stroke model (131). These results suggest that monocyte/macrophage is critical to hUCB-induced recovery following stroke. Monocytes have been known to play a key role in inducing angiogenesis under abnormal conditions such as ischemia and tumor (108). Thus, these findings imply that transplantation of monocyte-depleted hUCB does not induce angiogenesis properly, and in turn, cannot improve neurological dysfunction.
A third benefit of hUCB cells is their apparent migration following implantation to ischemic regions of the brain. We recently showed in vivo that several chemokines increase in the ischemic brain area (58). MCP-1 and macrophage inflammatory protein (MIP-1α) are implicated in bloodstream monocyte infiltration into tissues under pathological conditions (6, 43). We found that MCP-1 and MIP-1α expression was significantly increased in the ischemic hemisphere of rat brain, and significantly promoted hUCB cell migration compared to the contralateral side (58). Also chemokine receptors were constitutively expressed on the surface of hUCB cells. The hUCB cell migration was neutralized with polyclonal antibodies against MCP-1 or MIP-1α. These results suggested that the increased host chemokines in the ischemic area can bind surface receptors on exogenous hUCB, and attract systemically delivered hUCB cells into the central nervous system.
In conclusion, stroke involves a complicated cascade of inflammatory events that eventually lead to a significant necrotic area adjacent to the obstructed vasculature. This cascade is time dependent, and when hUCB cells are administered intravenously or intraparenchymally 48 h following stroke onset, they can prevent impending neural cell crisis. The transplantation of hUCB cells most likely inhibits the apoptotic cascade, induces endogenous and/or exogenous neurogenesis, and modulates the immune/inflammatory response to injury both peripherally and locally. Furthermore, another advantage in using hUCB cells for the treatment of stroke is the potential restoration of blood flow to the infarcted area because cord blood contains endothelial progenitor cells, which may be useful in neovascularization therapy. Essentially, if the present problems that hamper the application of hUCB cells to human stroke patients, such as uncertain safety, imperfect expansion processes, and limited graft survival in the recipient, can be overcome in the future, hUCB will be a good candidate for cell-based therapy, providing multiple therapeutic effects in a single transplant that no other pharmacological agent could mimic.
Footnotes
Acknowledgments
We wish to thank Nicole Kuzmin-Nichols for her help. P.R.S. is a cofounder and C.V.B. and A.E.W. are consultants for Saneron CCEL Therapeutics Inc. P.R.S., C.V.B., and D.J.E. were not involved in the editorial and peer review process for this article. This article is based on a presentation to the Pan Pacific Symposium of Stem Cell Research held in Taiwan, June 2008.
