Flush from a decade of unprecedented economic growth, Celtic coffers are now funding basic science at an appropriate level. Government is expected to back research and development to the tune of €2.5 billion ($3.4 billion) per year by 2006—not bad for a country with a population of fewer than four million people. (As a yardstick, the NIH budget in the U.S. was around $27 billion in 2003, in a country of 280 million). Part of the Irish funding goes to the relatively new Science Foundation Ireland. With some of its five-year grants approaching $5 million, its easy to see why Irish scientists are staying put and ex-pats like meeting organizer Dominic Walsh, as well as new faces such as Jochen Prehn, formerly of the Johann Wolfgang Goethe University, Frankfurt, Germany, have been lured to the Emerald Isle.

The story below is the first of a series of conference news we will post in the next few days.

Dublin: Chaperoned in Celtic Capital
It is great to have a chaperone when you visit a new city. At the meeting on Molecular Mechanisms of Neurodegeneration, held March 14-16 in Dublin, Ireland, chaperones were out in force.

Isabella Graef, Stanford University, opened the meeting by describing her efforts to bust plaques with small molecules that recruit larger chaperones. Preventing protein-protein interactions is a tall order for small molecules because they cannot hope to cover the vast number of contacts that can hold two proteins together. In the case of fibrils, the situation is exacerbated by the lack of “hotspots,” regions where the binding energy is concentrated. Breaking fibrils apart with a small molecule is a bit like asking a four-year-old to separate two boxers.

Graef’s solution is to have the pipsqueak bring along her heavyweight brother. Last October, Graef reported how a compound made by fusing the dye Congo red to a ligand for FK506 binding protein (FKBP) can prevent fibril formation. The compound, called SLF-CR for synthetic ligand for FKBP-Congo red, has enough affinity to bind to the fibrils and enough bulk, because it recruits FKBP, to interfere with fibril growth. In the presence of FKBP, the compound prevents fibril formation in cell-free systems and Aβ toxicity in cultured hippocampal neurons, though it has not yet been tested in vivo (see ARF related news story).

Graef hinted at how she might improve the plaque buster. Her strategy is to play with the three different modules of the compound: The targeting element, which binds the compound to Aβ fibrils; the recruiting element, which ropes in FKBP; and the linker that connects the two. In Dublin, Graef showed how other fibril-binding chemicals, such as the recently developed imidazole pyrimidine compound TZDM (see Kung et al., 2003), and the natural antioxidant curcumin, which may have plaque-busting talents of its own (see ARF related news story), may serve as better targeting elements. These two compounds cross the cell membrane, and bifunctional variants of TZDM, carrying either SLF or FK506, are even more cell-permeable, Graef reported. As for the recruiting element, she is currently working on substituting chaperones for FKBP, or cutting out the middle man altogether and recruiting a protease such as insulin-degrading enzyme or neprilysin, both of which degrade Aβ fibrils. Such a therapeutic strategy could prove beneficial not only for preventing plaque formation, but for clearing aggregates already formed, Graef suggested. As for the linker, Graef had already shown in the Science paper how lengthening it improved the potency several-fold to an IC50 of 50 nM.

The chaperone angle also marked the presentation of Linda Greensmith from University College London. Greensmith has been studying data on the use of small-molecule “co-inducers” of heat shock proteins. These co-inducers, such as arimoclomol, a hydroxylamine derivative, can amplify the heat shock response and so protect neurons against the toxic effects of mutant superoxide dismutase (SOD), which is responsible for about 20 percent of familial cases of the motor neuron disease amyotrophic lateral sclerosis (ALS).

Last spring, Greensmith’s group demonstrated that these co-inducers improve motor neuron survival and muscle strength in a mouse model of ALS and increase lifespan in these animals by about 25 percent (see Kieran et al., 2004). These effects are not just protective. Even when arimoclomol is given after the first symptoms have appeared, muscle strength improves and the mice survive about 18 percent longer than do littermates on placebo.

How, exactly, do these co-inducers work? In Dublin, Greensmith said that they may affect expression of key heat shock proteins. Her previous work had shown how BRX220, a derivative of the hydroxylamine bimoclomol, leads to increased expression of Hsp70 in astroglia when mice are subjected to acute injury to the sciatic nerve (Kalmar et al., 2002). The treatment also doubles the number of ventral horn neurons that survive injury. These neurons do more than merely hang on by the skin of their teeth, said Greensmith, as electrophysiological measurements showed that the number of working motor neuron contacts doubled in response to the drug.

Greensmith next posed the question of what causes the elevated Hsp70 expression? Could it relate to the finding that bimoclomol causes hyperphosphorylation and prolonged action of the major heat shock protein transcription factor, heat shock factor 1 (HSF-1) (Hargitai et al., 2003)? Possibly so, because when Greensmith examined mouse ventral horns from SOD mutant mice treated with arimoclomol, she found that HSF-1 was indeed hyperphosphorylated. What’s more, she found elevated expression of Hsp70 and Hsp90 in ventral horn neurons of these animals, but no changes in expression of Hsp27.

The pattern of heat shock protein expression in response to arimoclomol is interesting for two reasons. First, in an acute injury model, the effect on Hsp70 appeared in glia, not neurons. It is both plausible that the damage occurs faster than the ALS model, or that a totally different mechanism is at work, suggested Greensmith. Second, while Hsp70 and Hsp90 are known to be under the control of the HSF-1 promoter, Hsp27 is not. Therefore, the data fit nicely with a scenario whereby the hydroxylamine co-inducers act primarily on HSF-1, which in turn upregulates expression of Hsp70 and Hsp90. The question now is what causes increased phosphorylation of HSF-1. Meanwhile, it has become clear that the co-inducers are neuroprotective, said Greensmith.—Tom Fagan.