Reactive astrocytes crowd the brains of AD patients (Griffin et al., 1989), and cozy up to amyloid plaques in AD transgenic mice (Sturchler-Pierrat et al., 1997; Chishti et al., 2001; Oddo et al., 2003; Irizarry et al., 1997). This led Prusiner and colleagues to think that a gliosis marker could track amyloid accumulation and AD-like neuropathology in live mice. This approach has been successful for monitoring neuronal death in epilepsy models (Zhu et al., 2004) and, more recently, in Prusiner’s prion disease model (Tamgüney et al., 2009).

In the current study, first author Joel Watts and colleagues brought the GFAP-luciferase system into two AD strains—CRND8 mice, which develop Aβ neuropathology by three to four months of age, and APP23 mice, with slower amyloidosis. By Western analysis and immunohistochemistry, GFAP expression tracked well with plaque burden and brain levels of soluble Aβ in both transgenic lines. And in bioluminescence imaging done after injecting the mice with the luciferase substrate D-luciferin, luminescence signals from the animals’ brains gave a detectable readout for disease progression (see image below). Glow intensity correlated with onset of plaque accumulation (determined by immunohistochemistry) and brain Aβ levels (judged by ELISA) in the mice. In addition, the GFAP-luciferase system was able to detect accelerated Aβ deposition in younger APP23 mice that were inoculated with brain extract from aged APP23 animals.

Go With the Glow
Bioluminescence signals from brains of 12-month-old CRND8 AD mice expressing GFAP-luciferase transgene (top) and age-matched GFAP-luciferase non-AD transgenic mice (bottom). Image credit: Watts et al. PNAS 2011

Like all new technology, the approach has some caveats. Bioluminescence imaging machines are expensive, and their resolution is limited at present to the whole brain. And because the system measures GFAP transcription, not Aβ deposition per se, “there may be situations in which (the bioluminescence signal) is not completely reflective of Aβ deposition status in the brain,” Watts wrote in an e-mail to ARF. He also noted that GFAP-luciferase imaging does not replace behavioral testing. Rather, it “allows you to rapidly assess ‘disease status’ in live mice and guides you as to when to perform behavioral tests and what subset of mice may give the most uniform response,” Watts explained.

To follow up on the present study, Watts and coworkers are trying to use their GFAP-luciferase system to monitor the efficacy of candidate Aβ-directed therapeutics, and evaluate the “infectiousness” of various preparations of Aβ-containing material. They are also looking to see whether the system would improve with promoters for other genes whose expression goes up earlier during disease and/or to a greater extent than GFAP. Nevertheless, the current data provide proof of concept for using bioluminescence imaging to monitor the progression of neurological dysfunction in models of AD and other neurodegenerative diseases. “Some labs may wish to generate their own reporter mice tailored to their specific needs, but the concept will be similar,” Watts wrote.

Meanwhile, Agneta Nordberg and colleagues at Karolinska Institute in Stockholm, Sweden, are trying a similar approach—using an astrogliosis marker alongside amyloid tracers—in positron emission tomography (PET) imaging of AD patients. In this case, however, two separate PET tracers directly visualize astrocytosis and amyloid deposition, respectively. She reported new data from these studies at the Human Amyloid Imaging conference held last month in Miami, Florida (see ARF related conference story).—Esther Landhuis.

Reference:
Watts JC, Giles K, Grillo SK, Lemus A, DeArmond SJ, Prusiner SB. Bioluminescence imaging of Abeta deposition in bigenic mouse models of Alzheimer’s disease. PNAS Early Edition. 24 Jan 2011. Abstract