27 July 2009. The idea of using stem cell therapy to mend memory circuits that are damaged in Alzheimer disease (AD) has met with justified skepticism. Asking stem cells to replace multiple cell types and patch broken connections over wide swaths of the brain seemed a stretch. But the versatile cells continue to generate surprising results. A paper out this week in PNAS online, from Frank LaFerla’s lab at the University of California at Irvine, shows that stem cells are able to rehabilitate memory function in old mice with extensive AD pathology, but not by replacing lost neurons. Instead, the cells provide nurturing support in the form of brain-derived neurotrophic factor (BDNF), which increases the number of synapses in the hippocampus. Surprisingly, the stem cells work without causing any reduction in plaques or tangles.
The results will surely add to a growing interest in BDNF, a trophin critical for neuronal survival and synaptic formation, as a therapy for AD. Margaret Fahnestock at McMaster University in Hamilton, Ontario, along with Elliot Mufson and colleagues at Rush University in Chicago, Illinois, has shown for some time that levels of BDNF are decreased in the hippocampus and cortex of people with AD (see ARF related news story), and that the decrease is likely caused by amyloid-β (Aβ). The latest paper from Fahnestock’s group, out in the July 22 Journal of Neuroscience, suggests large Aβ oligomers in particular are responsible for downregulating BDNF levels in mouse models in vivo.
Both LaFerla’s stem cell study and Fahnestock’s work strengthen the idea that downregulation of BDNF may account for at least part of Aβ’s neurotoxic effects, and support the case for BDNF replacement as an alternative to amyloid-targeted therapeutic strategies.
As LaFerla put it, “If we make the argument that the brain is the most complex organ in the human body, and memory is the most complex function that the brain undertakes, it might be a lot to be expecting a small molecule to reverse decades of cognitive loss, and I think its about time that we try cell-based therapies. And in this regard, stem cells or even a product from the stem cells may be useful.”
LaFerla admits to questioning the approach at first. Although his group previously showed that stem cells could ameliorate cognitive defects in a model of focal hippocampal neuron loss (Yamasaki et al., 2007), they believed that AD, with its diffuse pathology and widespread damage, was different. “We were quite skeptical that stem cells would be useful for treatment of Alzheimer’s because we were operating on the assumption that they would work by a replacement mechanism.”
On the other hand, bystander effects of stem cells had been seen before (e.g., see ARF related news story on Lee et al., 2007), and first author Mathew Blurton-Jones says that is part of the reason he undertook the work. “Our rationale came from the literature in Parkinson’s and Huntington’s and other models where people were seeing benefits to behavior or survival, but when they actually looked into the mechanism they were seeing similar sorts of nursing effects or neurotrophic effects,” he told ARF.
So Blurton-Jones set out to transplant neural stem cells into the hippocampi of 18-month-old triple transgenic mice, which carry mutant human APP, presenilin 1, and tau genes. The mice, created in LaFerla’s lab, develop an extensive array of Alzheimer-like symptoms, including amyloid plaques, tau tangles, neuroinflammation, and memory defects (see ARF related news story). Blurton-Jones found that one month after stem cell transplantation, mice improved their performance to the level of control mice in two memory tests, the Morris water maze and a novel object recognition test.
Then came two unexpected results. First, there was no change in AD-related pathology: Plaque levels, soluble and insoluble Aβ, Aβ oligomers, total tau, phospho-tau, and localization of tau all remained the same in the transplanted mice. LaFerla called this result “Amazing,” saying, “We have never gotten an improvement in cognition ever in our model without affecting Aβ and tau.”
Furthermore, the effects did not appear to be due to cell replacement. Postmortem analysis indicated that five weeks after transplantation, most of the stem cells had differentiated, but not into neurons. Only about 5 percent became neurons, while the majority became astrocytes (40 percent) or oligodendrocytes (26 percent).
What had increased was synaptic density. When the investigators looked at levels of the marker protein synaptophysin, they found a 67 percent increase in the transplanted mice. Given the role of BDNF in synaptic remodeling, they measured levels of the factor and found it was also elevated.
Additional experiments proved that BDNF was necessary for the behavioral effects of the cells. Transplanting neuronal stem cells treated with BDNF shRNA to block expression of the neurotrophin failed to significantly improve the animals’ memory. In addition, direct injections of recombinant BDNF into the hippocampus improved performance in the water maze and resulted in increased synaptic density in the hippocampus.
The results jibe with recent work from Mark Tuszynski’s lab at the University of California, San Diego, who showed that viral expression of BDNF in the hippocampus in another mouse model of AD also improved cognitive performance and synaptic density, without affecting amyloid plaque levels (see ARF related news story on Nagahara et al., 2009). Together, the studies raise the question of what might be the best way to deliver BDNF. Blurton-Jones said they are now collaborating with Tuszynski to do a head-to-head comparison of viral versus stem cell delivery to see which is most effective at restoring cognition.
Stem cells could have some advantages, Blurton-Jones says. “We know that in the stem cells, if we decrease the BDNF by shRNA, we lose the benefits to cognition, but if you actually look at the data its not a complete loss. There may be other factors that the cells are making, other neurotrophins.” Another potential benefit of stem cells could lie in ease of delivery, he says. “If you want to deliver enough virus to get enough widespread distribution of, say, BDNF expression, you might have to potentially change the brain into a pincushion. One of the stem cells’ characteristics is that they can migrate over huge distances, and they can even migrate to areas of injury, so potentially you could do a single injection of stem cells and get widespread delivery.” In addition, the stem cells might provide a more physiological release of BDNF. These are only suppositions, however, and Blurton-Jones stressed that only the direct comparison will reveal the best approach for delivery.
With data accumulating behind it, BDNF is emerging as a potential key player in the toxicity of Aβ. Levels of the neurotrophin are decreased in early AD, and correlate with cognitive decline (Peng et al., 2005). More recently, decreased BDNF was shown to correlate with age-related cognitive decline (see ARF related news story on Li et al., 2009). Fahnestock and colleagues reported that Aβ oligomers cause a decrease in BDNF production in cultured cells (Garzon and Fahnestock, 2007), and in the new study first author Shiyong Peng and colleagues look at the relationship between Aβ and BDNF in vivo, using three different mouse models of amyloid pathology, and one that overexpresses APP but does not accumulate Aβ (Down syndrome mice).
The results suggest that the downregulation of BDNF is specific to the aggregation state of Aβ. Peng and colleagues report that BDNF mRNA levels are significantly reduced in two of the strains (APPNLh/PS1P264L and TgCRND8, which express mutant human APP) compared to wild-type littermates, but not in the third strain (APPswe/PS1M146V; see Kurt et al., 2001) or in the APP overexpressing strain. The reduction is associated with the presence of a high-molecular-weight (~115 KDa) assembly in soluble cortical homogenates from the two strains. All strains had strong signals for other oligomers, including the Aβ* 12-mer of 56 KDa (see ARF related news story on Lesne et al., 2006). The presence of the larger oligomer, and decreased BDNF, correlated with high levels of Aβ42, and elevated Aβ42/40 ratio in the mice. The structure of the ~115 KDa oligomer was not determined, but because it dissociated in the presence of calcium, the authors suggest it may represent a protofibril species. How it modulates BDNF levels is unclear, but there are several BDNF mRNAs, and the researchers show that the downregulation occurs mainly in two transcripts, one of which is homologous to the highly expressed human transcript IV.—Pat McCaffrey.
Blurton-Jones M, Kitazawa M, Martinez-Coria H, Castello NA, Muller F, Loring JF, Yamasaki TR, Poon WW, Green KN, LaFerla FM. Neural Stem Cells improve cognition via BDNF in a triple transgenic model of Alzheimer disease. PNAS. 2009 July 20. Abstract
Peng S, Garzon DJ, Marchese M, Klein W, Ginsberg SD, Francis BM, Mount HTJ, Mufson EJ, Salehi A, Fahnestock M. Decreased brain-derived neurotrophic factor depends on amyloid aggregation state in transgenic mouse models of Alzheimer's disease. J. Neurosci. 2009 July 22; 29(29):9321-9329. Abstract