September 30th, 2002
Evan Y. Snyder and Kook I. Park
from Nature Medicine
Two papers in this issue address the problem of neuron birth and replacement in the adult brain following injury.[1,2] The old saw, 'the devil is in the details', was never as true and of such importance to patients and practitioners as in dealing with the prospects of brain repair. Given how radically the stem-cell concept has altered our formerly deterministic view of the central nervous system (CNS), it is not unexpected that the public and even the scientific community have become intoxicated by an almost unbridled euphoria over the prospects of cell rebirth. But now we must deal with the fine print in our contract with developmental biology.
Though ostensibly unrelated, or even contradictory, both studies should be viewed as voicing cautionary notes on the ease and efficiency with which neuron replacement may be accomplished. Taken together, the reports alert us to the limitations of neuron replacement. One study places blame for actual pathology on the neural progenitor cell (NPC) itself, and, more importantly, the milieu in which the NPC must function, making neuronal replacement via endogenous or exogenous NPCs problematic. The other study, despite a deceptively upbeat title, actually finds neuron replacement to be meager, very restricted, short-lived, and non-functional even after extensive injury. Both studies, in their own way, are sobering and challenge the regeneration/stem-cell field to acknowledge these limitations and focus on circumventing them. However, we should be left not feeling that these obstacles are insurmountable, but simply that much work remains, principally in understanding both fundamental neurodevelopmental principles and basic pathophysiological processes. We must understand how these two forces interact before cell replacement will become tractable for most neurological diseases.
The repository of much neural plasticity is the NPC. Many NPCs reside throughout life within secondary germinal zones of the brain, for example, the subventricular zone (SVZ) of the fore-brain or dentate gyrus of the hippocampus. Indeed, it was the implantation of exogenous NPCs to 'probe' abnormal microenvironments that first suggested the existence of spontaneous compensatory mechanisms for genetic[5,6] or acquired deficiencies,[7,8] including by neurogenesis beyond its normal confines. However, these compensations alone do not redress deficits in the most devastating of cases. Thus interest has focused on how these natural processes might be augmented by supplying exogenous stimulants[9,10] and/or NPCs (ref. 11).
Irradiation -- both total body as preconditioning for bone-marrow transplantation and cranial for CNS malignancies -- is a common and largely unavoidable intervention fraught with significant cognitive and motor side effects. Monje et al. attempt to place these deficits in the context of our growing knowledge of NPC biology and adult neurogenesis. The authors observed that irradiation of the brains of adult rats produced NPC dysfunction within the neurogenic zones of the hippocampus, regions plausibly implicated in cognitive deficiencies. One month after irradiation, animals were injected systemically with a thymidine analog, bromodeoxyuridine (BrdU), which incorporates into the nuclei of proliferating cells and can subsequently be detected with an antibody. The presumption that cells that give rise to new neurons are cycling is a commonly accepted (albeit non-specific) working criterion for NPCs (refs. 3,9,10). The number of surviving BrdU-labeled cells and the type of cells they became after one month was then quantified. Not only was hippocampal proliferation blunted and neuron-generating capacity ablated, but any BrdU+ cells present adopted glial fates.
To determine whether the damage was to the NPC itself or to the neurogenic milieu in which the NPC resided, the authors performed classic complementation experiments. When placed in culture, NPCs from an irradiated animal, though not entirely normal, nevertheless yielded normal numbers of neurons. In the converse experiment, when NPCs cultured from a non-irradiated brain were transplanted into an irradiated hippocampus, their differentiation into neurons was blunted by nearly 90%. These data suggested that this disease state seemed not to be an NPC-intrinsic problem, but rather a dysfunction of the neurogenic microenvironment; that is, irradiation altered regulators of NPC survival, number and fate (Fig. 1). Defects included a striking increase in the number of activated microglia and a threefold increase in the distance between NPCs and their microvascular supply. Such defects are not compatible with ready neuronal replacement, whether by endogenous neurogenesis or transplanted NPCs. That does not preclude changing the milieu, for example, via anti-inflammatory and pro-angiogenic maneuvers. However, indiscriminate transplantation of exogenous NPCs or stimulation of endogenous NPCs here would likely prove futile and potentially harmful.
Figure 1. An inhospitable milieu can thwart neuronal replacement by NPCs (neural progenitor cells) or stem cells. This limitation is demonstrated by the effects of irradiation on neural NPCs in the SGZ of the hilus of the hippocampus (areas in inset progressively enlarged). In the normal adult, NPCs are thought to give rise to neuroblasts that yield mature neurons or to glioblasts that yield mature glial cells (Glia) in appropriate ratios. Signals emanating from the microvasculature (red arrows) might take part in instructing this NPC differentiation. Following irradiation, however, the neurogenic milieu is likely impaired by the invasion of activated microglia and macrophages that secrete cytokines (yellow dots) that may lead to one of the following: 1) altered NPC progression through the cell-cycle; 2) impaired neuronal differentiation; 3) neuronal apoptosis; and/or 4) blocking of instructive or trophic cues emanating from the microvasculature. NPC recruitment and guidance that might typically derive from microvasculature-elaborated signals are further abrogated by a substantially increased distance between NPCs and their supporting vasculature. This situation results in a decrease in cycling NPCs and an increase in the number of glia at the expense of new neurons. Generation of new neurons is thus virtually eliminated. Transplanted unirradiated NPCs have the same fate as the native irradiated NPCs in this milieu.
While the study by Arvidsson et al. might seem, on its surface, to provide a counter-example to that of Monje et al., in fact, it reinforces this cautionary theme. The authors used essentially the same BrdU pulsing strategy but applied it to adult rats subjected to experimental middle cerebral artery stroke. When examined after two weeks, a small number of the proliferating cells expressed neuronal markers within the striatum. That structure does not normally generate neurons in adulthood. Most noteworthy was not the appearance of such cells, but their small, almost trivial number despite the massive degree of destruction. Not only were they few in number, but most (80%) were short-lived ( 2 wk), suggesting they made no contact with targets. Although some cells stained for markers consistent with a more mature striatal neuronal phenotype, the subpopulation was so very small (< 0.2% of the striatal neuronal population) that it is unlikely that they have functional significance. There is no additional evidence to suggest they were functional, sent appropriate projections or made a behavioral impact. Hence, the experiments may suggest that stimuli (perhaps to the SVZ) elicit the emergence of neuroblasts, but we must deduce that there was no replacement or repair. Also notable was the total absence of neuroblast addition to the cortex despite the degree of its impairment.
The small number and low survival of incipient neurons -- and that only in one of the damaged CNS regions -- likely reflects unfavorable environmental or developmental conditions for neurogenesis and/or neuron survival. The most salient message from this study, therefore, is that even if beckoned to re-enter the cell cycle and yield neuron-like cells, the NPCs are typically insufficient in number, distribution, longevity or integrative capacity to be effective. Therefore they must be optimized, possibly by altering the milieu. This might be accomplished by supplying additional cells in particular developmental states or by engineering NPCs to express particular neurotrophic, neuroprotective or detoxifying molecules. Whether exogenous NPCs, in the absence of a manipulated milieu, would suffer the same fate as the endogenous neuroblasts is unknown.
A critical reading of these papers should compel the neuroregeneration field both to exercise caution and to recognize the enormous challenges we face before NPC-mediated repair -- whether by transplanting exogenous cells or by using endogenous cells -- becomes feasible. As a partial blueprint for the future, a number of caveats would seem to be in order.
First, it is important to know how disease biology itself impinges on NPC biology. One cannot simply 'dump' cells into a defect without this knowledge. Not only do different diseases have different needs, but the same disease has different needs at different phases within the same patient. The requirements for one brain region (for example, striatum) may be inappropriate for another (for example, cortex).[2,14] Cellular therapies alone might be ill suited for some cell non-autonomous disease processes (that is, problems extrinsic to the NPC) such as aspects of radiation and ischemic encephalopathy.
Second, it is important to know your cell. Is what we are asking of the NPC within its biological repertoire? To direct NPCs to different regions to yield reciprocally interacting cells of the right type -- in the right ratio, in the right layer, making the right connections with the right partners and to shield non-targeted cells and regions from such influences -- presupposes a level of understanding we will not have for years.
Third, the pathophysiological complexity of most disorders calls for multi-faceted solutions. Restitution of function likely requires more than the replacement of a single cell type but rather reconstitution of the entire milieu, including glia,[15,16] angioblasts, and extracellular matrix. Indeed, one may want to take advantage of the ability of a stem cell itself to give rise to multiple types of both neurons and glia. Furthermore, NPC-mediated interventions must go hand-in-glove with adjunctive interventions, orchestrated to 'overcome' synergistically restrictions imposed by the milieu, such as pro-angiogenic and/or anti-excito-toxic, anti-inflammatory and anti-apoptotic manipulations.
Such challenges demand that neuron replacement be viewed as an expression of developmental principles. We may more intelligently promote NPC compensation in adulthood by first examining injury to the developing (that is, pediatric) brain to discern the mechanisms by which the CNS shifts developmental patterns to achieve homeostasis. Those signals might be the ones to recapitulate outside the typical windows of neurogenesis.
Finally, it is important to avoid glib statements regarding the ease of neuron replacement. Patients are ill served. Circumspection is not a retreat from the promise of NPCs, simply an acknowledgement of the sophistication that will be required to exploit them properly. Until the above issues are resolved, brain repair will not become a reality. Nor, in fact, will the use of stem cells from any organ for any neurological or non-neurological disease.
Evan Y. Snyder, Department of Neurology & Division of Newborn Medicine Harvard Medical School Boston, Masscusetts, USA; and Kook I. Park, Department of Neurology & Division of Newborn Medicine Harvard Medical School Boston and Department of Pediatrics, Pharmacology, & Brain Yonsei University College of Medicine Seoul, Korea, USA
Nat Med 8(9):928-930, 2002.
© 2002 Nature Publishing Group