The ability of the CNS to respond to injury by increasing cell production and attempting regenerative repair has only recently been appreciated.
Still, the inadequate or abortive replacement of CNS cells compares poorly with the regeneration and functional repair seen in other organs.
The haemopoietic system, in particular, maintains circulating populations of cells with short lifespans; it has informed our knowledge of stem-cell biology, yielding the traditional stem-cell model-a hierarchical paradigm of progressive lineage restriction, in which as cells differentiate their fate choices become progressively more limited, and their capacity for proliferation reduced, until fully differentiated, mitotically quiescent cells are generated.
This model has recently been challenged by findings that suggest that adult stem cells have greater plasticity (more differentiation choices) than previously realized.
Of particular relevance to the development of therapeutic interventions for neurological disease is the expression of neuronal or glial markers by cells derived from lineages not previously thought to have neuroectodermal potential.
Importantly, transplantation of these cells produces functional improvements in animal models.
The mechanisms responsible for observed functional benefits have been disputed, and a consensus has not yet been reached.
Pluripotent cells were first isolated from the embryonic inner cell mass.
Ethical concerns, and the potential danger of teratoma formation, have contributed to the limited development of these cells as a therapeutic option: most scientists still talk in terms of "10-15 years".
However, adult stem cells have been used therapeutically for many years in malignant haematological disease and, more recently, for the treatment of inherited storage diseases.
The haemopoietic stem cell has provided the template against which stem cells isolated from other tissues (including skin, liver, and pancreas) have been judged.
Thus, stem cells were traditionally thought to be numerically insubstantial (1 haemopoietic stem cell per 104-105 total bone-marrow nucleated cells), with the innate ability to maintain a stem-cell pool (self-replicate) and generate more committed precursors, which then generate all lineages present in their resident organ.
Evidence from back-transplantation experiments suggested that limits are imposed on differentiation potential early in development.
Traditionally, therefore, differentiation was considered a unidirectional process of sequential, irreversible commitment steps that progressively restrict the range of fate choices available.
Stem cells are relatively quiescent cells, particularly in organs where cell turnover is low; yet, they can respond strikingly to tissue stress.
As cells designed to withstand crises and orchestrate repair, stem cells must be especially resilient.
A remarkable demonstration of their robustness is the isolation of stem cells from post-mortem tissue.
Neural stem cells have been identified in many species, including human beings.
The discovery that progenitors attempt repair in human disease has prompted the development of new theories to explain neurodegenerative disease, as well as offering hope that neurological repair could be enhanced via the supplementation, stimulation, or protection of endogenous precursors.
Several findings challenge the traditional hierarchical model of tissue-specific adult stem cells.
Bone-marrow-derived stem cells may have unexpected differentiation potential and can contribute to nonhaemopoietic tissues.
For example, donor-derived myocytes were found in mice after bone marrow transplantation; likewise, chimerism has been detected in the liver, brain, heart, and lungs of human recipients of bone-marrow transplants.
Under conditions of strong selective pressure, intravenously delivered, highly purified haemopoietic stem cells restore biochemical function in murine tyrosinaemia – an elegant example of transplanted bone-marrow-derived cells restoring function in an adult organ outside the haemopoietic system.
In these studies, hosts had received irradiation or chemotherapy, which could conceivably have alterations, stimulated in cell phenotype, but does not detract from the evidence of restored function.
Somatic stem cells can also contribute to the adult CNS.
After intravenous delivery of bone marrow in adult mice, a proportion of host microglia and macroglia are donor-derived.
These cells distribute widely throughout the brain-an important property of putative reparative cells.
Additionally, bone-marrow cells express neuroectodermal markers when manipulated in culture.
Excitement regarding the potential of bone marrow for regenerative therapy in neurological disease was properly tempered by the knowledge that expression of neural markers alone does not prove functionality, although some studies also provided electrophysiological evidence of function, a disputed property of embryonic stem-cell-derived neurons.
In-vivo findings also suggest that bone marrow can contribute to extra-haemopoietic tissues.
Clonality was not always rigorously demonstrated in in vivo studies, but in the mouse, single bone-marrow-derived cells (isolated by limiting dilution) contribute to several somatic tissues.
Experiments using animal models have also had encouraging results.
Rodent studies demonstrated that focal implantation or intravenous delivery of bone-marrow cells enhanced remyelination in the spinal cord.
Transplantation of bone-marrow-derived cells also improved function in models of cerebral ischaemia, Parkinson's disease, Huntington's disease, and trauma.
However, questions have been raised about the mechanism of benefit and effect.
The degree of differentiation and integration of the transplanted cells often correlates poorly with the observed functional benefit.
Indeed, functional benefit has been recently reported in the absence of transdifferentiation.
The finding that bone-marrow cells fused in vitro with embryonic stem cells (although fusion is a recognized property of the latter) raised the possibility that fusion might account for apparent transdifferentiation of adult cells.
Two studies demonstrated the presence of genetic markers derived from both host and donor in individual liver cells, cardiomyocytes, hepatocytes, and Purkinje cells of the cerebellum after transplantation of bone-marrow-derived cells.
However, the picture is complicated-no host-donor fusion was shown in various other sites, including renal glomeruli, buccal epithelium, pancreas, and brain, after transplantation of adult bone-marrow-derived cells.
This complex scenario warrants a more detailed consideration of how adult stem cells might contribute to tissue repair.
How adult stem cells might contribute to repair?
At least six different mechanisms might be proposed: transdifferentiation, de-differentiation, transdetermination, cell fusion, true pluripotent stem-cell behavior, and the production of trophic factors (Figure 1).
These mechanisms are, of course, not mutually exclusive.
Fig. 1. How adult stem cells might contribute to repair
In transdifferentiation, a mature cell assumes the phenotype and function of another fully differentiated cell.
This mechanism occurs during normal oesophageal development, when smooth muscle cells switch to skeletal muscle.
Some reports of transdifferentiation have been based on morphological characteristics and lineage-specific markers alone.
However, to fulfill the criteria for transdifferentiation, multilineage engraftment and functional activity must also be demonstrated.
Transdetermination is the redirection of lineage-committed stem cells or precursors to an alternative lineage.
This happens during development in drosophila, but it is extremely difficult to establish definitively that cells are irreversibly committed to a lineage before transdetermination, either at single cell or population level.
De-differentiation refers to the gain in differentiation potential that would occur if mature cells were pushed back up the hierarchical model of lineage restriction.
After injury, some amphibians regenerate limbs, tail, and even brain and spinal cord by de-differentiation at the injury site.
Normally, primordial germ cells transplanted into blastocysts are fate-restricted to germ cells.
However, environmental manipulation allows these cells to contribute to all somatic tissues.
For instance, oligodendrocyte precursor cells can be induced to generate neurospheres and subsequently neurons.
Whether de-differentiation of mammalian cells occurs under normal circumstances or in repair remains unclear.
Investigators recognized in the 1960s that differentiated cell fate could be altered in the rather extreme experimental conditions used when a somatic cell nucleus was injected into an enucleated egg, and the cloning of mammals provides proof of principle.
Nonetheless, much of our knowledge about cell fusion has been derived from studies involving heterokaryons-cells fused in vitro.
Typically, specialized functions are lost in these cells, although they may be regained after chromosome loss.
Silent genes may be activated; for example, synthesis of human muscle proteins occurs in human amniotic fibroblasts after fusion with mouse muscle cells.
It has been postulated that gene dosage (the relative genetic contribution of the two fused cell types) is important both to novel gene activation in heterokaryons and to the suppression of malignancy in hybrids.
The formation of (non-dividing) heterokaryons indicates that differentiated cell phenotype can be altered without DNA replication and cell division.
This mechanism has particular advantages when considering ways to affect repair in highly specialized and integrated environments, such as the adult CNS.
Is fusion necessarily either an undesirable event or an unwanted capacity?
Although there is a well-recognized association between aneuploidy (abnormal chromosome number and chromosomal rearrangements) and neoplasia, it is increasingly clear that polyploidy (abnormal chromosomal number) has been under-recognized and can occur in the absence of malignancy.
Indeed, in some circumstances polyploidy might be advantageous.
Polyploidy contributes to the development of new species, including the evolution of primates, correlates with metabolic requirement and postnatal growth rate; and occurs during insect embryogenesis; and a substantial proportion of normal neural progenitors have been found to be aneuploid in mice.
In the adult mammal, it is striking that organs with considerable regenerative capacity are frequently populated with cells that are not diploid; megakaryocytes are polyploid cells and aneuploidy often occurs in the liver.
Polyploid cells have also been reported in the cerebellum, arterial smooth muscle, heart, uterus, and thyroid under normal circumstances or in response to stress.
Donor cells have been identified in the liver, brain, and heart after bone-marrow transplantation and, notably, these are the locations at which fusion has been most convincingly demonstrated.
These findings raise the intriguing possibility that cell fusion might have crucial but hitherto unappreciated physiological roles in development, tissue maintenance, and repair.
It seems likely that in vivo, under normal circumstances, diploid cells have a selective advantage over polyploid cells, but this balance might alter in disease states.
Certainly, cell fusion seems to be more common than previously recognized; polyploidy is not synonymous with disease, and whether fusion compromises or enhances the reparative capacity of stem cells remains to be discovered.
True multipotent cells may persist beyond embryogenesis and, if provided with the appropriate signals, differentiate into cells of multiple lineages.
The multipotent adult progenitor cell is a potential candidate for such a cell, having been differentiated in vitro into cells classically identified as being of mesodermal, neuroectodermal, and endodermal lineages.
Although rare (1 per 106 bone marrow mononuclear cells) these cells proliferate rapidly in vitro (cell doubling time 48-72 h) and have been expanded by more than 60 cell doublings.
It is likely that our current knowledge of cell markers is inadequate to define cell populations accurately, so it is possible that cells with true multipotency will have contaminated experiments previously reported as examples of transdifferentiation.
If the presence of an adult reserve of stem cells is confirmed, the question of why such cells fail to effectively contribute to repair in disease states must be addressed.
This research may provide further insight into the etiology of some diseases.
Stem cells of various origins, including bone marrow, do have tropism for disease suggesting that signals may be released from areas of tissue damage into the circulation.
Indeed, in-vitro studies have shown that bone-marrow-derived cells migrate in response to monocyte chemoattractant protein-1, a chemokine produced during cerebral ischaemia.
Further identification of the mechanisms involved would pave the way for the development of treatment regimens designed to enhance endogenous mobilization of stem cells in disease states, perhaps using small molecules.
Stem cells could also play a part in promoting functional recovery by means other than cell replacement.
The production of trophic factors might confer resistance to disease, or promote the survival, migration, and differentiation of endogenous precursors.
Certainly, bone-marrow cells are known to produce a wide variety of cytokines and exert paracrine effects.
That cells previously thought to be of restricted lineage can give rise to progeny of different germ-cell lineages has questioned not only our understanding of stem-cell biology, but also wider issues pertaining to mechanisms of tissue homoeostasis and repair in the adult mammal.
These issues have profound implications for reparative medicine.
To establish definitively the mechanism responsible for observed changes in cell phenotype, major technical difficulties must be overcome: the starting population must be adequately characterized and purified with minimal in-vitro manipulation, the extent of previous lineage commitment must be demonstrated, and functional integration will have to be proven.
Some terms currently in use (such as transdifferentiation, de-differentiation, and transdetermination) might simply represent attempts to reconcile observed experimental findings with the hierarchical model of stem-cell function, and might be rendered obsolete as our understanding deepens.
At present, the available evidence seems to suggest that the mechanism responsible might be organ-specific or depend on the precise nature of disease and conditioning.
However, the easy access of bone-marrow-derived stem cells to the circulation, together with their tropism for disease and production of cytokines, means that these cells are ideally suited to a physiological role in tissue maintenance and repair.
Studies demonstrating cell fusion were widely interpreted as raising substantial barriers to the further development of treatment regimens using non-organ specific stem cells, often overlooking the fact that the outcome was functional repair.
We, and a growing body of other investigators, have argued that this is not necessarily the case, although there is an obvious need for further clarification and safety data.
Indeed, the great theoretical advantages of adult stem cells, and in particular of bone-marrow-derived cells - their ready accessibility and autologous nature, decades of use in the treatment of haematological malignancy, and ethical robustness-are already yielding dividends.
The transition from laboratory to clinic has commenced, and successful preliminary studies of cell therapy in ischaemic heart disease using bone marrow or skeletal-muscle-derived stem cells have been reported.
We are optimistic that haematologists' extensive clinical experience with stem-cell transplantation is likely to be of future benefit to many patients.