Stem cells injected into the cerebral ventricles-or even injected intravenously-find their way to sites of inflammation in an animal model of multiple sclerosis (MS), according to a report in today’s Nature. Once there, they replace the damaged myelin, clean up astrocytic scarring, and help reduce inflammation. Most importantly, the treatment abolishes the motor deficits of the disorder.

Focal transplantation of stem cells offers hope in a disease such as Parkinson's, where much of the disease can be traced to a well-delineated area of neurodegeneration in the substantia nigra; however, disorders such as MS and Alzheimer's disease involve much more widespread neurodegeneration. MS and similar demyelinating disorders also have other interesting points of intersection with Alzheimer's disease, including inflammatory processes in the areas of neurodegeneration.

Gianvito Martino, Angelo Vescovi, and their colleagues at the San Raffaele Hospital in Milan, Italy, injected mice with antibodies to a component of myelin to produce the well-known model experimental autoimmune encephalomyelitis (EAE). They then injected the mice with neurospheres-small aggregates of stem cells and their progeny derived from the cerebral ventricular zones of adult mice.

Whether introduced via the cerebral ventricles or the blood stream, these cells spread throughout the neural axis and differentiated into both myelin-producing oligodendrocytes and neurons. Interestingly, most of the transplanted cells were drawn to largely demyelinated areas of inflammation, where they gave rise to new myelin-producing cells. Concomitant with this repopulation, mice appeared to regain almost normal movement, beginning at about day 15 after transplantation.

A further benefit was that these transplanted cells mediated the reduction of "scarring" caused by overgrowth of astrocytes in the demyelinated areas. This step can itself be beneficial, because reactive astrogliosis hampers remyelination. The stem cells also appear to have played a role in reducing inflammatory processes by decreasing levels of inflammatory molecules (e.g., tumor-necrosis-factor-α) and metalloprotesases.

"One point of particular interest here is that these cells hitch a ride into damaged sites by using α4 integrin-the very molecule that mobilizes the immunological attack," notes Lawrence Steinman of Stanford University in California in his News and Views comment. α4 integrin is an adhesion molecule found on the surface of the immune cells that attack myelin in EAE. Apparently, the presence of this molecule allows the stem cells to cross the blood-brain barrier and move to areas of active inflammation in the CNS.

"To the best of our knowledge, this work provides the initial evidence showing how neural precursors may represent a renewable source of cells which, when transplanted into the cerebroventricular system or into the blood stream, can reach multiple areas of a chronically injured adult CNS, enter the brain tissue and seek damaged areas where they promote structural and functional recovery," conclude the authors.

For Alzheimer's researchers, a central question is whether this same method can deliver stem cells to areas of putative inflammation surrounding neurodegeneration in models of AD. And if so, can they work the same restorative magic that they did for this model of MS?—Hakon Heimer

Q & A with Gianvito Martino:

Q: Could this be applied to AD?
A: To make neural stem cells travel into the right areas where demyelination was ongoing, we took advantage of the inflammation occurring in the experimental MS model that we have been using. Thus, I do not see how can we apply this experimental therapeutic model "tout court" to AD. As I understand it, inflammation in AD takes place in the later stages of the disease and is milder than in MS. To me, inflammation in AD is just a physiological reaction to neurodegeneration; in MS the scenario is the opposite, whereby inflammation causes neurodegeneration. Thus, I do not know if inflammation in AD is enough to make cells travel to the right place and replace nonfunctioning neurons.

Q: Will you take this work into AD mouse models?
A: We have not, to date, planned to work in AD mice; however, it would be extremely interesting to see such experiments. We have been impressed by what we found in EAE; we frankly did not expect these results in the very beginning. Therefore, it is time to go in AD, as well, and see what can happen, but I would suggest someone who is more deeply involved in AD research try this. I would be delighted to share my experience on that.

Q: Are current mouse models adequate for such an experiment?
A: I think that AD models in mice are suitable. The time frame when you inject the cells is crucial. The mode of action of our cells in EAE mice is bimodal. They not only differentiate into ensheathing oligodendrocytes, but they also mediate the proliferation of endogenous oligodendrocyte precursors and down-regulate proliferation of endogenous astroglia possibly via a humoral mechanism.


  1. Comment by Kiminobu Sugaya This study proves our concept that neurospheres injected into ventricle migrate and incorporate into the host CNS (Qu et al., 2001). Since this transplantation method causes minimum damage compared with direct transplantation into brain tissue, immune attack to donor stem cells by the host will be reduced, and this results in better survival and efficacy of the transplanted stem cells. Although our group has repeatedly used this stem cell injection method (Kim et al., 2002), this article sets a milestone for the migration of stem cells through the ventricle wall—even the blood-brain barrier. At the same time, I would not recommend intravenous injection, because a large part of the neural stem cell may differentiate into blood or other peripheral-type cells before they reach the target area.

    The next question would be whether this type of stem cell transplantation is useful for Alzheimer’s disease therapy. For the time being, I have to say: Not yet. We have transplanted neural stem cell into the lateral ventricle of AβPP-transgenic mice, and we found that the donor cells are differentiated into only astrocytes, not into neurons (in preparation for publication). This may be due to physiological functions of AβPP on stem cell differentiation (see ARF live discussion). Although we have to consider effects of the pathological disease environment on stem cell biology, intraventricle injection would be the best transplantation method for CNS stem cell therapies.


    . Human neural stem cells improve cognitive function of aged brain. Neuroreport. 2001 May 8;12(6):1127-32. PubMed.

    . Reelin function in neural stem cell biology. Proc Natl Acad Sci U S A. 2002 Mar 19;99(6):4020-5. PubMed.

  2. The recent manuscript by Pluchino et al. offers the intriguing possibility that cells can be delivered to all regions of the brain merely by injecting them into the bloodstream of animals. The numbers required to reach the brain for the dramatic improvements seen appear quite small. About a million were injected, which, given blood flow dynamics, can at best be in the range of thousands of cells reaching the brain. This small number appeared to be targeted to the sites of injury and thus had an effect disproportionate to their number. The authors suggest that this is because neural stem cells have a homing tendency, and they show that the cells express CD44 and other candidate homing molecules.

    There were a number of points I found surprising in the manuscript.

    1. In general, neural stem cells are not very migratory and, in normal development, are restricted to stem cell niches in the developing and adult brain. Stem cells, therefore, do not express homing receptors and ,indeed, labeling with CD44 antibodies does not show expression on neural stem cells in vivo or after short-term culture in vitro. CD44 expression is seen on astrocytes and astrocyte progenitors, however, and NSC cultures are known to stochastically differentiate into astrocytes, or at least express astrocytic markers after prolonged culture. In inflammatory disorders in most regions of the brain, damage does not promote stem cell proliferation (our unpublished results). While cell proliferation is observed, it can be attributed to glial progenitor cells, endothelial elements, and astrocytes. Thus, cues to ensure proliferation and direct appropriate differentiation of exogenously transplanted stem cells are unlikely to be present. Indeed, reports of direct transplantation of stem cells into spinal cord injuries has suggested that stem cells themselves do not survive well in damaged tissue. This would suggest that the cultured NSCs contained a significant fraction of astrocytes or astrocyte-like cells, which likely survived in the damaged milieu.

    In the case of Alzheimer’s disease, it is unclear if such homing would direct adequate numbers of cells to appropriate sites in an Alzheimer’s brain and whether appropriate cues or an appropriate milieu would exist to direct site-specific neuronal differentiation.

    2. Inflammation can alter blood-brain barrier kinetics, but that it allows stem cells or any cells other than microglia to cross the vascular endothelium is quite surprising. Local damage is known to enhance macrophage and Schwann cell invasion of the CNS, and these cells can migrate along blood vessels and can myelinate central axons. These Schwann cells, endogenous progenitors, as well as the small number of cells that transited the blood-brain barrier are likely responsible for the improvement seen.

    It is unclear whether the endogenous glial progenitors proliferate in Alzheimer’s disease models or if remyelination offers therapeutic benefit. The degree of neuronal differentiation seen in the present report was small and unlikely to be a significant component of the observed benefit. It is, therefore, unclear whether a similar strategy would offer similar benefit in Alzheimer’s disease.

    3. The recovery observed was dramatic and did not correlate well with the number of transplanted cells, or the degree and rate of remyelination reported. The authors correctly point out that trophic/indirect effects could be important. However, neural stem cells are not known to produce trophic molecules for astrocytes, neurons, or oligodendrocytes, and while the authors showed some production of cytokines in cultures, no quantitative data was shown.

    The cytokines/trophic response required to enhance survival of endangered neurons in Alzheimer’s disease is likely to be different than the cytokine profile required for EAE-damaged cells. It is, therefore, unclear if one can assume the same therapeutic benefit would be seen in Alzheimer’s disease models.

    These concerns in no way detract from the impressive nature of the results reported by Pluchino et al. Their finding that intravenous delivery of cells can deliver cells to regions that are otherwise inaccessible is an exciting finding, as well. At the very least, their results raise the possibility that growth factors and genes can be targeted to an appropriate site in a relatively straightforward manner.

    At the same time, this study does raise questions about the cell population that was effective and the mechanisms underlying the observations. It will be important to determine why the cells delivered intravenously were effective before one can proceed further. It is also important to note that long-term passaged cultures are clearly different from acutely harvested cultures, and results obtained with immortalized stem cell lines have been different from acutely harvested or short-term passaged cells. Indeed, Morsehead and colleagues have suggested that stem cell character is altered in as few as 10 passages (Morshead et al., 2002). Finally, one must determine not only whether human cells behave in the same way, but also whether the blood-brain barrier will be equally permeable.


    . Hematopoietic competence is a rare property of neural stem cells that may depend on genetic and epigenetic alterations. Nat Med. 2002 Mar;8(3):268-73. PubMed.

Make a Comment

To make a comment you must login or register.


No Available References

Further Reading


  1. . New method for transplantation of neurosphere cells into injured spinal cord through cerebrospinal fluid in rat. Neurosci Lett. 2002 Jan 25;318(2):81-4. PubMed.

Primary Papers

  1. . Injection of adult neurospheres induces recovery in a chronic model of multiple sclerosis. Nature. 2003 Apr 17;422(6933):688-94. PubMed.
  2. . Medicine: Collateral damage repaired. Nature. 2003 Apr 17;422(6933):671-2. PubMed.