Antibiotics have been touted as potentially beneficial in many neurodegenerative diseases including Alzheimer and Parkinson diseases (see recent caveat about clinical trials of doxycyclin and rifampin for AD). But these benefits have not necessarily been linked to antibacterial action. In fact, in addition to preventing aggregation of amyloid-β (Aβ), rifampicin was recently shown to inhibit the aggregation of α-synuclein, a major protein player in the pathology of Parkinson disease (PD) (see ARF related news story).

But by increasing protein levels, the β-lactam antibiotics seem to claim yet a different mode of action. Fisher, at Columbia University Medical Center, New York, together with first author Jeffrey Rothstein at Johns Hopkins University, Baltimore, Maryland, and their colleagues at those two institutions and The ALS Association, Florida, stumbled upon this surprising property when they screened a diverse library of drugs and nutritional compounds for the ability to increase expression of the glutamate transporter (GLT1). Of the 20 or so compounds that increased expression of rodent GLT1 by over twofold, 15 turned out to be β-lactam antibiotics, and because the screening process was “blinded,” it was completely unbiased.

So what might the antibiotics be doing? Rothstein and colleagues found that they can increase expression of luciferase if it is hooked up to the GLT1 promoter, suggesting that the antibiotics are transcriptional activators. Ampicillin, capable of stimulating luciferase expression by about threefold, was the most potent activator tested. Significantly, the lactams could activate luciferase at concentrations known to be attainable in the brain, and indeed the authors found that the drugs could induce protein expression in vivo. When Rothstein tested transgenic mice expressing green fluorescent protein driven by the GLT1 promoter, GFP levels in hippocampus and spinal cord increased in response to ceftriaxone. This lactam seemed to have the most specific effects, as synthesis of other proteins was unaffected.

To test if the antibiotics may be useful in models of neurotoxicity, the authors turned their attention to oxygen glucose deprivation (OGD), which leads to glutamate-mediated toxicity in cultured neurons. When they added cefuroxime to the cultures 48 hours before an hour of OGD, only 20 percent of the neurons died as opposed to 50 percent of neurons in control cultures. In spinal cord models, ceftriaxone could completely prevent motor neuron loss induced by the transporter blockers, such as threo-β-hydroxyaspartate.

But the most exciting results came from studies on G93A SOD1 mice. These animals, having a glycine-to-alanine mutation at position 93 of the superoxide dismutase gene, suffer from a motor neuron disease akin to ALS, caused by the same mutation in humans. The authors found that ceftriaxone treatment, started at 12 weeks of age (about the time of disease onset) increased lifespan by about 10 days and improved body strength and weight as compared to control littermates. Postmortem analysis also showed that the drug prevented neuron loss in the spinal cord, as about 20 percent fewer neurons were found in age-matched control animals.

This data fits well with previous experiments showing that glutamate transporter overexpression can protect mice with motor neuron disease (see Guo et al., 2003), and it “provides a new pathway for drug discovery and manipulation of glutamate transmission in disease,” according to the authors. In fact, given that the pathway for the promoter activation is not yet known, there could be some interesting β-lactam targets waiting to be discovered.—Tom Fagan.

Reference:
Rothstein, JD, Patel S, Regan MR, Haenggeli C, Huang YH, Bergles SE, Jin L, Dykes Hoberg M, Vidensky S, Chung DS, Vang Toan S, Bruijn LI, Su Z-Z, Gupta P, Fisher PB. β-lactam antibiotics offer neuroprotection by increasing glutamate transporter expression. Nature. 2005 Jan 6;433:73-77. Abstract