1 February 2010. Bigger, smarter, and more amenable than mice to the imaging techniques that are rapidly becoming indispensable in Alzheimer’s research, rats could be a valuable model for studying AD. The major downside of the sagacious creatures is that they are about five times more expensive to maintain than mice, but then again, maybe you get what you pay for. For all the research into mouse models, many still come up short, failing to recapitulate some of the most basic AD pathologies, such as neurofibrillary tangles and neuronal loss. What if you could have it all in one animal?
That could soon be possible, according to Terrence Town, Cedars-Sinai Medical Center, Los Angeles, California. At the recent Keystone Symposium, Alzheimer’s Disease Beyond Aβ, held 10-15 January 2010 at Copper Mountain, Colorado, Town debuted a new rat model at the end of a talk focusing on the role of the innate immune system in AD. The model is the result of collaboration with Robert M. Cohen and Robert Pechnick at the same institution. If the model characteristics Town presented turn out to be true, researchers may be salivating over more than their ratatouille.
The rats express both human APP with the Swedish mutation and human PS1 with the exon 9 deletion, a la David Borchelt’s APP/PS1 mouse (see Savonenko et al., 2005). The transgenes are driven by the hamster prion promoter, as in Karen Hsiao Ashe’s Tg2576 mice (see Hsiao et al., 1996). Town reported that the animals show reduced NeuN staining compared to controls (around 25 percent lower in the hippocampus and a slightly greater loss in the cingulate cortex), suggesting neuronal loss with age. The rats develop plaques that can be detected by FDDNP imaging; importantly, they also develop nearby tangles as seen by immunohistochemistry (using Cp13 and PHF1 antibodies to tau) and ultrastructural electron microscopy. FDDNP imaging discriminates transgenic animals from controls, which opens up the possibility of following pathology longitudinally in individual animals, Town reported. (FDDNP is thought to bind to both plaques and tangles). Caspase 3, a marker of cell death, is also elevated in the APP/PS1 rats compared to controls. Levels of the caspase increase with age, and the protein appears in the vicinity of plaques. Tunel staining of 16- and 27-month-old rat brain tissue suggests progressive cell loss in the cingulate and hippocampus, Town said.
Town believes his may be the first AD rat to have a chance of becoming widely used. He noted that he hopes to make it freely available to academic labs, though companies may have to deal with some red tape and pay a fee. He suggested these rats better mimic human AD pathology than do similar mouse models because the rat tau proteome is more akin to that of humans. For example, humans express six different tau isoforms that differ by the number (three or four) and type of repeat units and by the extent of inserts in the N-terminal of the protein (see Gustke et al., 1994). Whereas mice express a four-repeat tau exclusively in the brain, a recent study suggests that the rat brain boasts the full complement of six isoforms (see Hanes et al., 2009).
The transgenic rats also exhibit gliosis, another hallmark of AD, and interestingly, Town showed confocal microscopy data suggesting that activated microglia (as judged by IBA1 staining) seem to take up both Aβ (seen by ThioS or 4G8 staining) and are filled with tau (Cp13 staining). “This could be a unique form of microgliosis,” suggested Town.
This data all seems fairly hot of the press. Town showed no behavioral results, but in response to questions, he did say that the animals show a significant decline in hippocampal-based learning and memory that kicks in around 15 months of age when plaque deposition is evident. He concluded by suggesting that these animals may present a better platform for preclinical testing than the current crop of transgenic mice.
Other rat models that express human APP, PS1, or both have been produced in the past (see Vercauteren et al., 2004; Folkesson et al., 2007; Agca et al., 2008; Liu et al., 2008; Flood et al., 2009). It is not clear why these models have not been more widely used, but Town told ARF that some of them appear to be short-lived, making them less suitable for AD research, while others lack the extensive pathology. “One of the key features of our AD rat model is that it produces high levels of total Aβ with age (over 100 microgam/wet gram of brain tissue), and it has an almost 1:2 ratio of Aβ1-42:Aβ1-40,” Town told ARF via e-mail. “Perhaps other rat models have not attained the requisite levels/type of Aβ in order to precipitate the full amyloid cascade hypothesis,” he suggested.—Tom Fagan.