Changes in calcium homeostasis have been reported in AD in many contexts, but a fundamental question of whether amyloid plaques affect calcium has not been addressed. To look at this, first author Kishore Kuchibhotla and colleagues developed a multiphoton imaging method to measure calcium and look at morphology simultaneously in neurons in living mice. They used a chameleon calcium sensor, a fusion of two fluorescent proteins linked by a calcium-binding domain that makes it possible to visualize cell morphology at the same time as it provides a fluorescence resonance energy transfer-based measure of calcium concentration. When expressed by adeno-associated virus in pyramidal neurons of adult AD transgenic mice, the protein filled soma, axons, dendrites, and spines, and reliably reported calcium transients in dendrites after pipette application of puffs of glutamate.

Using the probe, the researchers measured calcium concentration in multiple neurites from normal mice and found a narrow range of values between 65-91 nM. A few percent of neurites showed calcium overload, which they defined as sustained levels higher than 147 nM. They then measured calcium in four different transgenic AD mouse lines, two of which (APP/PS1 and Tg2576) develop plaques and two of which (PS1-ΔE9 and PS1M146V) do not. In the plaque-bearing APP/PS1 mice, about 20 percent of neurites displayed elevated calcium (mean of 499 nM). The rest had normal calcium. In Tg2576 mice the results were similar, while the PS mutant mice, which lacked plaques, showed no changes in neuronal calcium.

The results suggested that calcium overload was induced by amyloid plaques, leading the researchers to investigate if plaques had a localized effect on calcium. Visualization of plaques with an amyloid binding dye along with imaging of calcium showed that a greater proportion of neurites with calcium overload lay near plaques. Within 10 μm of plaques, 33 percent of dendrites had overload, but beyond 30 μm, the number dropped to 17 percent, and remained steady even with increasing distance.

In addition to higher calcium, the dendrites showed an abnormal compartmentalization of calcium. In neurons with normal calcium levels, there was no relationship between the calcium concentrations in the spine versus its dendritic base—that is, the compartments seem to be independently regulated. In contrast, in the plaque-bearing mice, neurons with elevated calcium show that the spine and dendrite concentrations were highly correlated, suggesting that the neurons had lost their ability to regulate spine calcium fluxes independently, an important part of dendritic signal integration.

Turning to morphology, the investigators found that a higher percentage of the calcium overloaded neurites showed abnormal morphology, appearing beaded, blebbed, or lacking spines at all. Calcium levels correlated with the extent of morphological changes. Spine beading has been shown to result in some cases from activation of the calcium-dependent phosphatase calcineurin (CaN), and the researchers showed that treatment of APP/PS1 mice with the CaN inhibitor FK-506 resulted in both reduced calcium overload and the appearance of abnormal neurites.

“We propose a multistage degenerative process in which amyloid-β aggregates induce calcium influx as an initial acute trigger,” the authors write. “The resulting moderate calcium overload initiates a second pathological stage that activates the phosphatase, CaN. Increased CaN activity then triggers the final stage of neurite structural degeneration and even more severe calcium overload.” This hypothesis is consistent with the ability of FK-506 to partially restore learning and memory in Tg2576 mice (Dineley et al., 2007), perhaps by preventing the morphological breakdown of plaque-proximal neurites.

The study, which shows no calcium effects in presenilin-1 mutant mice that do not harbor plaques, suggests that the presence of mutant PS1 by itself, though important in acceleration of plaque deposition by mutant APP, does not seem to play a role in the calcium overload measured in this study. PS1 was recently shown by Frank LaFerla and colleagues to regulate uptake of calcium into the ER by the SERCA pump, and by Kevin Foskett’s lab to modify release from the same stores via the IP3 receptor (see ARF related stories cited above). In the paper, Bacskai and colleagues allow that their results do not rule out that presenilins could have a more subtle effect on calcium handling, which could work alongside the calcium overload they describe.

As Kim Green and LaFerla of the University of California at Irvine put it in a mini-review that accompanies the paper, “These results highlight for the first time that proximity to Aβ plaques induces calcium dysregulation, and in discrete subcellular compartments, which are relevant to learning and memory, LTP, and synaptic loss. This seminal study provides data linking AD pathology and calcium dysregulation with subsequent morphological changes and opens the door to a number of compelling future studies.”

Among those, Green and LaFerla suggest, is whether elevation of dendritic calcium occurs in young Tg2576 mice that show cognitive impairments before the onset of plaque deposition. This would help to sort out if the proximity of calcium disruption to plaque is due to the plaques themselves, or if diffusible Aβ species, released from plaque, cause calcium imbalances. In addition, the question of where the calcium comes from remains to be explored.—Pat McCaffrey.

References:
Kuchibhotla KV, Goldman ST, Lattarulo CR, Wu H-Y, Hyman BT, Bacskai BJ. Abeta plaques lead to aberrant regulation of calcium homeostasis in vivo resulting in structural and functional disruption of neuronal networks. Neuron. 2008 July 31;59:214-225. Abstract

Green KN, LaFerla FM. Linking calcium to Abeta and Alzheimer’s disease. Neuron. 2008 July 31;59:190-193. Abstract