The Neural Stem Cell Response To Hypoxicischemic Injury

Little is actually known about the response of NSCs to CNS injury in general, let alone HI brain injury in particular. Is it possible to repopulate an "ablated" CNS with NSCs in the way hematopoietic stem cells reconstitute lethally irradiated bone marrow? HI brain injury was initially viewed as an injury that is not only of importance in its own right but also might serve as a prototype for other large, acquired brain in-juries.27 It occurred to us that to help answer this question we might be able to use one of our prototypical NSC clones, clone C17.2,28-30 as "reporter cells." This well-characterized clone is just one of several with stem cell features that exist in the literature — multipotent, self-renewing, self-maintaining, nestin-positive, responsive to various stem cell trophins. As one would demand of a putative stem cell, NSCs from clone C17.2 are able to participate in the development of the CNS throughout the neuraxis and across developmental periods, from fetus to adult.17,18,23,25,26,28,29 Engrafted and integrated NSCs are visible because they have been transduced also with a reporter gene, lacZ, that allows the cells to stain blue when processed with Xgal histochemistry, or to appear brown of fluorescent following reaction with an antibody against E. coli P-galactosidase (Pgal) in immunoperoxidase and immunofluorescence protocols, respectively.26,28 This ability to identify progeny of a donor NSC is important because, by their nature, NSCs integrate and intermingle seamlessly into the host following transplantation, do not form a discernible graft-host border, and actually come to resemble host neural cells of the same phenotype. When we talk about using clone C17.2 NSCs as "reporter" cells, we mean using well-characterized, indelibly marked cells with known ancestry, potential, and clonal relationships that are traceable, abundant, and homogenous, that intermingle imperceptibly with host cells in vivo and that can, therefore, be used as a tool for mirroring, probing, and tracking — i.e., "reporting" on — the behaviors of neighboring endogenous progenitors that are otherwise invisible to such monitoring and whose own clonal relationships and degree of homogeneity are much less certain. Such cells would also allow well-controlled experiments to proceed with minimal variability in cell population under study from experiment to experiment, animal to animal, condition to condition. The type of injury in which NSCs would be investigated in these preliminary experiments would be focal HIE engendered by permanent liga-tion of the right common carotid artery of a week-old mouse followed by exposure of the animal to 8% ambient oxygen. This combination of ischemia and hypoxia results in extensive injury to the hemisphere ipsilateral to the carotid ligation while leaving the contralateral hemisphere as an intact control.

In the first set of pilot experiments,27 we wondered what might be observed if we took a normal animal in which "reporter" NSCs had become stably integrated throughout the brain during a critical period of its development (creating virtually a chimeric brain of host and reporter cells) and then exposed that animal to unilateral HI injury. The experimental paradigm, therefore, was as follows: clone C17.2 NSCs were transplanted into the cerebral ventricles of mice on the day of birth (P0), allowing the NSCs access to the subventricular germinal zone (SVZ) that lines the ventricular system running the length of the neuraxis; this results in widespread migration, stable integration, and intermixture of donor NSCs with host cells throughout the parenchyma.17 The right hemisphere was then subjected to HI injury at 1 week of age (P7). The brains were analyzed 2-5 weeks later. The resulting picture in these preliminary studies was complex but intriguing. In contrast to the intact side, where the reporter NSCs remained widely and evenly interspersed throughout the intact parenchyma, the reporter NSCs in the HI-injured hemisphere appeared to be densely and preferentially clustered around the infarction cavity. The heavy accumulation and number of cells in that location suggested either that many NSCs had migrated to that particular area, or that the cells near there had proliferated, or both. In addition, in the penumbra of the infarction, an increased number of donor-derived cells were identified immunocytochemically as oligodendrocytes and neurons. Neurons and oligodendrocytes are the two neural cell types that are most susceptible to HI injury and that are least likely to regenerate spontaneously in the "post-developmental" mammalian cortex. Furthermore, in the intact hemisphere, as might be expected from NSCs implanted after the completion of embryonic cortical neurogenesis, no donor-derived neurons and many fewer oligodendrocytes were noted. Therefore, following HI brain injury, NSCs appeared to evidence components of altered proliferation, migration, and differentiation. This is precisely the type of behavior one might expect of a stem cell. We decided to start examining each of these components in a systematic fashion.27

First we asked whether there was new transient proliferation by quiescent NSCs, both reporter and host cells. To answer this question, a transplant of reporter NSCs was performed at P0 into the cerebral ventricles; unilateral HI injury was induced at P7 (after the cells had stably integrated, differentiated, and become quiescent); the mice were then pulsed with bromodeoxyuridine (BrdU), a nucleotide analogue, at various post-HI injury time points. The preliminary analysis revealed that, before injury, donor-derived cells were completely quiescent; however, after HI injury, the percentage of reporter (JacZ+) cells that became mitotic (i.e., incorporated BrdU) increased rapidly, peaked at about 3 days after induction of HI, and then fell back to 0 by 1 week after HI. Host cells did precisely the same thing; their pattern of proliferation was virtually superimposable upon that of donor cells, peaking at ~3

days after HI and then returning to 0, also suggesting an induction of transient proliferation.

That there are so many changes peaking at 3 days following injury is intriguing. The literature on stroke — and, in fact, on other injuries — has suggested that the interval of 3-7 days after insult is a very metabolically, biochemically, and molecularly active temporal "window" during which a variety of mitogens, trophins, extracellular matrix molecules, and other factors are uniquely elaborated. We shall return to this "window" and its impact on neural stem cell biology later in the review.

Next, we began to approach the question whether reporter NSCs (and, by extension, host NSCs), in fact migrated to the areas of neurodegeneration. NSCs (clone C17.2) were transplanted into only the left intracerebroventricular space at P0. At 1 week of age, unilateral HI injury was induced in the contralateral right hemisphere in some animals, while in others the right hemisphere was left intact. In animals with an intact right hemisphere, engrafted stem cells simply remained stably distributed and densely integrated throughout the parenchyma of only the transplanted left hemisphere. But in animals in which the right hemisphere had been infarcted, cells at multiple levels throughout the cerebrum dramatically appeared to migrate across the corpus callosum and any available interhemispheric commissure to the infarcted region (Figure 3.1). With high magnification under light and electron microscopy, one could appreciate leading processes of NSCs migrating along interhemispheric connections toward the damaged areas. Even within the infarct, one could see reporter cells migrating into the heart of the necrotic area.

figure 3.1 Migration by transplanted "reporter" stem cells to the ischemic area of a mouse brain subjected to unilateral, focal hypoxic-ischemic brain injury. Clone C17.2 neural stem cells were injected into the left cerebral ventricle of a mouse on the day of birth (postnatal day 0 [P0]). At 1 week of age (P7), the animal was subjected to contralateral right-sided hypoxic-ischemic injury. The animal was analyzed at maturity with Xgal histochemistry to identify lacZ-express-ing donor-derived cells (which stain blue). Some cells appeared to migrate along the corpus cal-losum (arrowhead) throughout the cerebrum toward the highly ischemic area (arrow).

figure 3.1 Migration by transplanted "reporter" stem cells to the ischemic area of a mouse brain subjected to unilateral, focal hypoxic-ischemic brain injury. Clone C17.2 neural stem cells were injected into the left cerebral ventricle of a mouse on the day of birth (postnatal day 0 [P0]). At 1 week of age (P7), the animal was subjected to contralateral right-sided hypoxic-ischemic injury. The animal was analyzed at maturity with Xgal histochemistry to identify lacZ-express-ing donor-derived cells (which stain blue). Some cells appeared to migrate along the corpus cal-losum (arrowhead) throughout the cerebrum toward the highly ischemic area (arrow).

Therefore, there seems to be evidence that NSCs already integrated into the CNS will migrate to an area of subsequent infarction. Will reporter NSCs implanted after HI injury also be drawn to areas of damage? To investigate this question, the following paradigm is followed: Unilateral (right) HI injury is induced at P7, and reporter NSCs are transplanted into the contralateral (left) cerebral ventricle 3 days later (at P10). As a control, some animals not subjected to right HI are also transplanted on the left at P10. As before, in intact animals, the NSCs remain nicely but stably integrated on the transplanted left side. However, in animals that have been in-farcted on the right before transplantation on the left, reporter NSCs migrate avidly across the corpus callosum and other interhemispheric commissures to the area of infarction throughout the length of the cerebrum. Furthermore, they integrate into those infarcted areas as if drawn or directed by a tropism for the region. When reporter NSCs are injected directly into the infarcted area on the right, they never migrate in the other direction to the contralateral intact side, in these pilot studies.

This last manipulation, that of injecting NSCs directly into the infarct, suggests what our next set of experiments entailed. NSCs (clone C17.2) are transplanted directly into the degenerating infarcted region at various time points following the induction of unilateral HI. When implantation is performed shortly after an HI (e.g., the following day), robust engraftment is seen throughout the infarcted area. If

figure 3.2 Robust engraftment by transplanted neural stem cells within the ischemic region of a mouse brain subjected to unilateral focal hypoxic-ischemic injury (HI). This mouse was subjected to right hypoxic-ischemic injury on postnatal day 7 (P7). Three days later (P10), the animal received a transplant of clone C17.2 neural stem cells within the region of infarction. The animal was analyzed at maturity with Xgal histochemistry. A representative coronal section is shown. Robust engraftment was evident within the ischemic are (arrow). Similar engraftment was evident throughout the hemisphere. Even cells that implanted outside the region of infarction appeared to migrate along the corpus callosum toward the ischemic area. The most exuberant engraftment was evident 3-7 days after HI. Immunocytochemical and ultrastructural analysis revealed that a subpopulation of donor-derived cells, especially those in the penumbra, differentiated into neurons and oligodendroglia, the two neural cell types most characteristically damaged by HI and the cell types least likely to regenerate spontaneously in the postnatal brain.

figure 3.2 Robust engraftment by transplanted neural stem cells within the ischemic region of a mouse brain subjected to unilateral focal hypoxic-ischemic injury (HI). This mouse was subjected to right hypoxic-ischemic injury on postnatal day 7 (P7). Three days later (P10), the animal received a transplant of clone C17.2 neural stem cells within the region of infarction. The animal was analyzed at maturity with Xgal histochemistry. A representative coronal section is shown. Robust engraftment was evident within the ischemic are (arrow). Similar engraftment was evident throughout the hemisphere. Even cells that implanted outside the region of infarction appeared to migrate along the corpus callosum toward the ischemic area. The most exuberant engraftment was evident 3-7 days after HI. Immunocytochemical and ultrastructural analysis revealed that a subpopulation of donor-derived cells, especially those in the penumbra, differentiated into neurons and oligodendroglia, the two neural cell types most characteristically damaged by HI and the cell types least likely to regenerate spontaneously in the postnatal brain.

transplantation is postponed until 5 weeks after HI, virtually no, if any, engraftment is achieved. Engraftment is most exuberant 3-7 days after HI (Figure 3.2).

Is there, indeed, a change in differentiation fate by these reporter NSCs in these areas of degeneration compared with what might be seen in intact brain? Immunocytochemical and ultrastructural examination of the engrafted regions, particularly in the penumbra of the infarct, suggests that indeed there is. Donor-derived cells (recognized by an anti-Pgal antibody) are assessed for the expression of neural cell type-specific antibodies (e.g., NeuN, neurofilament, MAP-2 for neurons, CNPase for oligodendrocytes, GFAP for astrocytes, nestin for immature, undiffer-entiated progenitors). A subpopulation of donor NSCs in the injured postnatal neo-cortex differentiate into neurons (~5%) and oligodendrocytes (~4%). Other cell types are astroglial — though no scarring seems apparent — and undifferentiated progenitors. As noted below, these numbers contrast significantly with what one finds in an intact age-matched recipient neocortex. The presence of donor-derived neurons can be detected as much as 1 mm away from the heart of the infarction cavity on the side of the lesion, suggesting a relatively large "sphere of influence" exerted by the injured tissue. (Interestingly, occasionally we would note host-derived neurons in an otherwise severely destroyed cortex; this type of finding is consistent with our belief that some host NSCs, just as the reporter NSCs do, try to shift their differentiation toward compensation for neuronal cell death, a phenomenon that we are perhaps augmenting with our transplants; more on this phenomenon later.) Examination of the penumbra under the electron microscope in these preliminary studies supports the immunocytochemical assessments. A significant number of donor-derived oligodendrocytes and neurons are appreciated. Some donor-derived pyramidal neurons receive synaptic input from the host.

Quantification of the differentiation pattern by transplanted reporters NSCs in the injured neocortex compared with that in the intact neocortex is dramatic and illuminating. Whereas 5% of engrafted NSCs on the injured side differentiate into neurons, no neuronal differentiation by NSCs is seen at all in the intact neocortex, consistent with both the normal absence of neurogenesis in the postnatal mammalian cortex and with our own prior findings.25,31 There is a 5-fold increase in the number of donor-derived oligodendrocytes in the injured neocortex compared with the intact neocortex. The number of astrocytes does not significantly differ between the two sides. Also, there is an upregulation of nestin in donor NSCs in response to injury (almost three times as many donor cells are nestin-positive in the injured cortex compared with the intact cortex, suggesting that they may become activated or primed to make a differentiation choice).

These preliminary quantitative data are presented to make a qualitative point. On the intact side of the infarcted animal, there is no neuronal differentiation at all; on the injured side, 5% of donor-derived cells are now neurons. The magnitude of that number is less significant than the phenomenon of qualitatively going from consistently no neurons to neurons of any number at a stage in development when no cortical neurons should normally be born. As mentioned previously, oligodendrocytes and neurons are the two neural cell types most damaged by HI injury. It appears from these preliminary data that NSCs may be attempting to repopulate and reconstitute that area of injury — particularly within a certain temporal window — by "shifting" their normal differentiation fate to compensate for the loss of those particular cell types, especially neurons. It seems indeed likely that, as a consequence this type of neurodegeneration, signals are elaborated to which NSCs (donor and probably host) are able to respond in a reparative fashion. Precisely what those signals are is an area of ongoing active investigation. They no doubt are a complex mix of various mito-gens, neurotrophins, adhesion molecules, cytokines, and so forth.

Although the preliminary numerical data cited above are presented principally to illustrate the "shift" toward neuronal differentiation by NSCs in response to injury, it is instructive to note that, given the vast number of NSCs that engraft into the in-farcted region, a differentiation of even 5% of such cells into neurons translates into tens of thousands of replacement neurons supplied to that degenerating region. We don't actually know how many neurons and how much circuitry is required to functionally reconstruct a damaged mammalian system. We do know that, fortunately, 100% restoration is not needed; older lesion data would suggest that as little as 10% may be sufficient.

Was this article helpful?

0 0

Post a comment