14 October, 2005. On September 22 in New York City, the Institute for the Study of Aging hosted a conference to assess the potential therapeutic benefits of stem cell technology in brain aging and Alzheimer disease. Convened by Howard Fillit, the meeting had three stated aims:
- Review knowledge regarding neurogenesis and how it may apply to AD
- Develop a consensus on whether neurogenesis is a valid target for AD
- Identify existing—and speculate on new—therapeutic strategies to promote neurogenesis for AD
Ira Black, from the Robert Wood Johnson Medical School in Piscataway, New Jersey, started off by reviewing his work on adult stem cells. Whether adult stem cells have the potential to turn into cells of different lineages—for example, hematopoietic progenitors differentiating into neurons—is highly controversial (see ARF related news story and ARF news story), but Black and colleagues have found that bone marrow stromal cells (BMSC), which are derived from the mesoderm, can give rise to ectodermally derived cells, including neurons (see Munoz-Elias et al., 2004). When the scientists injected BMSC into the brain ventricles of rat embryos, they found that the cells migrated and appeared in the developing cortex about four days later. This migration appears to be conventional in that the cells travel along fibers extending out into the parenchyma, and the fibers derive from donor cells that have developed into radial glia.
Black reported that his lab had discovered cells similar to BMSC in the amniotic membrane. These amniotic-derived cells express ecto-, meso- and endodermal markers, suggesting they may have the potential to turn into a variety of different cell types. The advantage of marrow-derived cells, however, is their potential for autologous grafts, which would preclude the need for simultaneous immunosuppressant therapy.
Clearly, the developing brain is a far cry from the aged AD brain, and the audience quizzed Black about how the microenvironment would influence the fate of BMSCs. Black’s lab is focusing on what environmental signals influence the differentiation and migration of progenitors. He emphasized that identifying such factors is going to be extremely important, not only for uncovering ways to manipulate transplanted cells, but also for stimulating stem or progenitor cells that are naturally present in the adult brain. Modifying endogenous cells with small molecules has much better therapeutic potential than periodically injecting people with cells, he suggested. Other scientists including Jeffrey Macklis have argued this point, as well.
These two themes—the role of the environment, both local and external, and the role of signaling factors—appeared foremost on the speakers’ minds at this meeting. Matthew During, from Weill Medical College at Cornell University, broached both topics when he reviewed recent findings on the effects of the external environment on hippocampal neurogenesis (see Cao et al., 2004).
His group has found that both environmental enrichment and training in the Morris water maze, two paradigms that are known to stimulate neurogenesis in the subgranular zone of the hippocampus, lead to increased expression of vascular endothelial growth factor (VEGF) in the brains of adult rats. In fact, this was the only factor that was increased in response to both paradigms. Other signaling molecules known to be associated with enhanced neurogenesis, such as brain-derived neurotrophic factor (BDNF), glial-derived growth factor (GDNF), and phosphorylated cyclic AMP response element binding protein (phospho-CREB), failed to respond to one or both regimens.
For discussion’s sake, During compared the brain to a muscle. In muscle, exercise leads to a local drop in the partial pressure of oxygen, which in turn induces the expression of genes with hypoxic response elements in their promoters. Such genes include VEGF. The analogy is tempting in light of recent reports indicating that “exercising” one’s brain helps to maintain cognitive ability and stimulates adult neurogenesis (see recent ARF related news story and ARF news story).
But is the association between VEGF and adult neurogenesis anything more than a correlation? During argued that it is. His team overexpressed VEGF in the hippocampus by transfecting rat brains with adenoviruses carrying the VEGF gene, and found that the mice performed better in a passive avoidance test, showed improved spatial memory in the Morris water maze, and had increased cell proliferation in the subgranular zone of the hippocampus. Perhaps the strongest evidence comes from the use of RNAi to silence VEGF in the adult brain (VEGF knockouts are lethal, precluding this experimental approach). During and colleagues found that in animals treated with VEGF RNAi, exposure to an enrichment environment does not lead to the increased cell proliferation observed in wild-type mice.
The work suggests that VEGF mimetics might help promote adult neurogenesis and potentially stave off neurodegeneration. In this regard it is worth noting that the factor has already been used with some success to delay disease onset and lengthen survival in mouse models of amyotrophic lateral sclerosis (ALS, see ARF related news story).
The influence of the environment was also the topic of Teresita Briones from the University of Illinois at Chicago. She has found that enriched housing conditions (including, for example, daily replacement of toys, group housing) leads to a higher number of granule neurons in the rat hippocampus. Furthermore, she found that enriched environmental conditions lead to significantly more synapses per neuron (2,200 contacts per neuron versus 1,900 per neuron in rats housed in “traditional” settings).
But what about the injured brain? Many researchers question how exercise or mental stimulation might help those who may already face cognitive loss. Briones has addressed this question using a rat model of ischemia. She found that after injury, the number of neurons and the number of synapses per neuron increases, even in animals housed in normal cages. However, environmental enrichment fails to elicit any further increases (see Briones et al., 2005). The data suggest that the power of environmental enrichment may be limited, an important consideration in the context of AD. Recent work has shown, for example, that environmental enrichment can either reduce amyloid-β (Aβ) levels and amyloid deposition in a mouse model of AD (see ARF related news story), or increase it (see Jankowsky et al., 2003). This discrepancy needs clarification.
Assuming one could spur neurogenesis in the adult brain, would it have any functional significance? This question occupies Jack Parent of the University of Michigan Medical Center in Ann Arbor. There is no evidence to support the idea that neurogenesis in the dentate gyrus of the hippocampus can contribute to repair. In fact, the opposite might be true, Parent argued.
His studies on epilepsy have shown that the numbers of immature neurons in the dentate gyrus increase in epileptics, but this might not necessarily be a good thing because these cells are hyperexcitable and may contribute to seizure problems. In humans, these cells occur in the hilus of the hippocampus, where they are just as likely to be problematic. Even so, the overriding problem with gaining a functional benefit from adult neurogenesis remains that insufficient numbers of new neurons survive. Therefore, finding ways to increase that percentage would be an important therapeutic avenue, Parent suggested.
Further thoughts on the AD environment came from David Greenberg, Buck Institute for Age Research in Novato, California. First, he reminded the audience that human anatomy is quite different from that in small rodents. For example, the rostral migratory stream, a pathway of neuronal migration from the subventricular zone to the olfactory bulb in rodents, has not been demonstrated in humans. Greenberg speculated that such migration might not be feasible in the human brain because the distances involved are much greater than in the rodent brain.
And yet, neurogenesis does occur in the human brain. Markers of neuronal precursors, such as doublecortin, tend to increase with the severity of AD, Greenberg said. But interpreting data from human studies is complicated because patients may have been on medication or had particular dietary or lifestyle habits.
A specific question Greenberg and colleagues have focused on is whether it is cell death that stimulates neurogenesis, and judging from an AD mouse model (PDAPP Sw/Ind Tg mice), their answer is no (see ARF related news story)—despite enhanced levels of neurogenesis in these animals, they exhibit no neuronal loss. Instead, more subtle abnormalities, for example, reduced synaptic transmission, might stimulate neurogenesis in AD models, Greenberg suggested. His lab now has data suggesting that C31, a protein fragment released from the C-terminal of AβPP, might play a role in increasing neurogenesis. Work from Eddie Koo’s lab at University of California, San Diego, has shown that this peptide is cytotoxic (see ARF related news story and ARF news story), and now Greenberg and colleagues have found that in PDAPP mice that cannot make C31 because they have an additional APP mutation, neurogenesis proceeds on a par with that in wild-type animals.
Mark Noble, University of Rochester, New York, introduced a different concept of environment, namely the redox environment to which stem cells or progenitors, particularly oligodendrocyte type-2 astrocyte progenitors (O-2A), might be exposed.
O-2A progenitors are important in the context of myelination disorders. That these cells strike the right balance between self-renewal and differentiation into oligodendrocytes is important for development, but what’s more, it might also control demyelination, which is prominent in neurologic disorders as diverse as multiple sclerosis and multiple system atrophy (see ARF related news story). In fact, there are even hints that demyelination might be a factor in AD pathology (see ARF related news story; Bartzokis et al., 2004).
Noble and colleagues have found that the redox state of O-2A precursors shifts this balance between self-renewal and differentiation. The more reduced the cells are when isolated, the more likely they are to renew themselves in vitro. Noble has found that signaling molecules that promote self-renewal, such as basic fibroblast growth factor (bFGF) and neurotrophin-3 (NT-3), also lower the cellular redox potential, while factors that stimulate differentiation have the opposite effect. The findings raise the possibility that endogenous systems may be controlling neurogenesis/differentiation pathways by altering cellular redox states. This might enable scientists to predict how drugs or therapies affect cells in vivo, Noble suggested.
Several of the day’s speakers addressed the potential of small molecules to promote neurogenesis. Frank Longo, University of North Carolina, Chapel Hill, said that in considering how small molecules may promote neurogenesis, one has to think beyond stem cells. Proliferation, migration, differentiation, integration, and survival of the new neurons are key problems to be addressed. Small molecules could be designed to target any or all of these steps, he suggested. BDNF, for example, has been shown to spur neurogenesis, yet some neuronal progenitors have no BDNF receptors, implying that some neurotrophic factors may not act directly on progenitors.
Longo noted that BDNF might be a particularly exciting protein because neurotrophin mimetics may simulate its actions, for example, on the protein kinase TrkB. There is also evidence that BDNF is lacking in the brains of AD patients (see ARF related news story). Other trophic activities that could be similarly mimicked include those of NGF, a factor that has shown some promise in pilot clinical trials for AD (see ARF related news story).
Barbara Hempstead, Weill Medical College, Cornell University, New York, echoed the importance of both the mature and proforms of these two trophins, which can elicit opposing effects. Alzforum follows developments in this area closely (see, for example, ARF related news story). Hempstead also summed up her latest work with collaborator Bai Lu at the National Institute of Child Health and Human Development, Bethesda, Maryland, showing that pro- and mature BDNF can stimulate long-term depression and long-term potentiation in neurons, respectively (see ARF related news story).
Despite the theoretical benefits of mimicking the actions of a particular neurotrophin, in practice it is much easier to inhibit an enzyme activity or a protein interaction than to stimulate it. Evolution has ensured that proteins work at near optimal capacity, so finding room for improvement can be challenging. For this reason, Longo suggested that protein phosphatases might make good targets for small molecules intended to stimulate neurogenesis. His team has found, for example, that the leukocyte antigen-related phosphatase (LAR) is present in neuronal progenitors and could be a target for small-molecule inhibitors. That is because knocking out LAR leads to a doubling of the number of proliferating progenitors expressing the neuronal marker NeuN.
One of the take-home messages from this meeting was that many small molecules with neurogenic potential already exist. In some cases, such as antidepressants (see below), their mechanism of action is unclear. Judy Kelleher-Andersson, for example, founder of the small biotech company Neuronascent Inc., in Clarksville, Maryland, is screening for small molecules that may spur neurogenesis. She is using commercially available human fetal stem cells to test libraries for compounds that can simultaneously increase proliferation and survival of human neuronal progenitors. One of the potentially interesting compounds she has found not only increases the neurogenic potential of the fetal cells, but also protects them against the toxic effects of Aβ1-42, she reported.
Christophe Labie from the pharmaceutical company Sanofi-Aventis, Toulouse, France, reported on neurogenic compounds that are in clinical trials for AD and/or PD. Xaliproden and SR57667B at nanomolar concentrations can increase the number of adult neural stem cells that differentiate to express neuronal progenitor markers, such as Tuj1 and MAP2, Labie said. They can also counteract the effects of leukemia inhibitor factor (LIF), which acts to keep stem cells in self-renewal mode, preventing their differentiation.
Company researchers have found that these compounds can increase the number and migratory rate of neuronal progenitors in vivo, Labie reported. In the normal dentate gyrus, the compounds increased the number of neural progenitors. Perhaps more dramatically, in a neurodegenerative model where medial septum neurons are ablated by the neurotoxin vincristine, oral delivery of SR57667B induces the migration of Tuj1-positive neural precursors into the septum within 7 days, and by 21 days mature neurons have begun to replace the lost cells. Sanofi-Aventis is committed to developing these compounds even though their mode of action is unknown, Labie said.
The same curious situation—an intriguing effect on neurogenesis by a mysterious mode of action—applies to antidepressants. Jessica Malberg from Wyeth Research in Princeton, New Jersey, reported that multiple classes of antidepressants increase cell proliferation in the adult mammalian hippocampus. The finding is interesting because major depression is linked to the risk for AD (see ARF related news story).
Malberg said that several hints at how antidepressants might stimulate neurogenesis are emerging, and some of them may have nothing to do with neurotransmission. Instead, antidepressants might counter stress by increasing the levels of BDNF and other trophic factors such as CREB. The drug rolipram, for example, can increase the levels of cAMP and CREB, and it also increases cell proliferation in the dentate gyrus of the hippocampus. The cAMP/CREB connection is of particular interest to AD researchers because Aβ appears to tamp down these pathways (see ARF related news story).
Allopregnanolone, by contrast, is a potential therapeutic whose mechanism of action is better understood. Roberta Diaz Brinton, from the University of Southern California, Los Angeles, reported that this steroid derivative can increase proliferation of both rodent and human neural stem cells by up to 30 percent within 24 hours of exposure. The mechanism of allopregnanolone is through the GABA(A) receptor in the cell membrane. It has been known for some time that in adult neurons allopregnanolone potentiates chloride influx through the GABA(A) receptor. More recent work has shown that in immature neurons, such as neural stem cells, allopregnanolone promotes an efflux of chloride that leads to excitation of the cell membrane and an influx of calcium. In work just published (see Wang et al., 2005), Brinton and colleagues show that activation of L-type calcium channels is required for allopregnanolone to stimulate neural stem cell proliferation.
Furthermore, Brinton and colleagues have found that allopregnanolone induces expression of cell cycle genes (for example, cyclins and CDKs) and decreases expression of genes that inhibit proliferation (such as CDK inhibitors). Of particular interest will be how allopregnanolone affects neurogenesis in triple transgenic mice (see ARF related news story). These experiments are presently underway, Brinton said, adding that allopregnanolone penetrates the blood-brain barrier, which is critical for its development as a neurogenic therapeutic. [Editor's note: Corrections to Brinton section posted 7 November 2005.]
In Alzheimerology, whenever a new idea takes hold, researchers test how it might relate to the predominant pathogenic players, and with neurogenesis this is no different. At the ISOA meeting, Anne Cataldo, McLean Hospital, Belmont, Massachusetts, addressed the question of whether the N-terminal as well as the C-terminal of AβPP could be relevant in neurogenesis. This last talk took the audience full circle to bone marrow-derived adult progenitor cells (MAPCs), with which the day had begun. Cataldo’s lab is testing the effect on the differentiation of these cells of sAβPPα, a peptide released from AβPP by the action of α-secretase.
The scientists have found that sAβPPα enhances neuronal differentiation of MAPCs in culture. Of particular interest is that it promotes the development of cholinergic neurons, which are the subtype most devastated in AD. Presently, Cataldo’s group is investigating the effect of intravenous injection of sAβPPα into the Ts65DN mouse, which mimics the trisomy that causes Down syndrome. The Ts65DN animals develop cholinergic degeneration in the forebrain. Results from these studies are currently being gathered. In toto, research on the role of neurogenesis in the aging and diseased brain remains in its infancy, but at least it has begun to diversify, as basic as well as some translational and clinical projects are moving ahead in parallel.—Tom Fagan.