Stay the Executioner's Hand

See allHide authors and affiliations

Science  14 Apr 2000:
Vol. 288, Issue 5464, pp. 283-284
DOI: 10.1126/science.288.5464.283

The French physician Charcot provided the original clinical description of the motor neuron disease amyotrophic lateral sclerosis (ALS) back in 1869. The disease causes astonishingly rapid loss of motor neurons in the cortex, brainstem, and spinal cord over 1 to 5 years with consequent paralysis and death. Patients with ALS (commonly referred to as Lou Gehrig's disease in the United States after the baseball star who died of the illness) are typically 40 to 50 years of age at diagnosis. Why motor neurons are selectively vulnerable to disease, and what triggers their destruction, still remains a mystery more than a century later.

The great leap forward in our understanding of ALS began in 1993 with the demonstration (1) that a familial form of the illness could be attributed to mutations in the SOD1 gene, which encodes the cytosolic antioxidant enzyme, copper/zinc superoxide dismutase (Cu,Zn SOD). At first it seemed that the mutations might cause disease by reducing the amount or activity of Cu,Zn SOD, thereby decreasing the protection from oxidative stress of motor neurons in ALS patients. But mice lacking Cu,Zn SOD do not develop motor neuron disease, whereas transgenic mice overexpressing a mutant human SOD1 gene do (2, 3). This suggested that the mutant protein itself is in some way selectively toxic for motor neurons, perhaps through altered oxidative chemistry, protein misfolding, or protein aggregation (4).

On page 335 of this issue, Li et al. (5) provide evidence that a mutant SOD1 transgene causes motor neuron death in mice through caspase-mediated programmed cell death (apoptosis). Initiator caspases—enzymes activated from their dormant precursor forms in response to a variety of cellular insults—act on the precursors of downstream caspases such as caspase-3, which are the executioners in the breakdown of essential cellular proteins (see the figure). Apoptosis is characterized by a complex series of cellular changes leading to the noninflammatory demise of the cell. It is a normal, highly regulated process that is crucial for proper cell growth and development. In pathological states, however, it can be abrogated (cancer) or exacerbated (neurodegeneration), either condition leading to some of the most devastating diseases known.

Players in a deadly game.

Pathways for caspase activation in the mutant SOD1G93A transgenic mouse model of ALS. Cellular insults that result from expression of mutant Cu,Zn SOD lead to activation of endogenous proforms of caspase-1 and caspase-3. Caspase-3 initiates destruction of cellular proteins and cell death by apoptosis. Caspase-1 promotes production of the proinflammatory cytokine IL-1β, which increases transcription of both caspases and exacerbates cell death.

The idea that motor neuron death in SOD1 transgenic mice is through an apoptotic pathway is bolstered by experiments in which overexpression of bcl-2, a well known mitochondrial inhibitor of apoptosis, protects against neuronal loss (6). Indeed, mitochondrial involvement in the apoptotic pathway leads to release of cytochrome c, an activator of the initiator caspase-9 that, in turn, activates the executioner caspase-3 (7). The Li et al. study builds on previous observations that caspase-1 and caspase-3 are activated in SOD1G93A mice (which overexpress an SOD1 transgene carrying an ALS mutation that results in glycine being substituted for alanine at position 93). When these mice are crossed with transgenic mice expressing a dominant negative mutant form of caspase-1 that is inactive, they have a slight increase in life-span (8). Li and co-workers (5) show that a small peptide caspase inhibitor (zVAD-fmk) prolongs the survival of SOD1G93A transgenic mice, and they provide evidence for the activation of both caspase-1 and caspase-3 in neurons within the ventral horn of the mouse spinal cord.

The caspase inhibitor zVAD-fmk blocks all known caspases and acts at several points in the activation cascade (see the figure). The VAD backbone positions the inhibitor within the caspase active site with the aspartate carboxylic acid facing downward into a deep, positively charged pocket, so that the fluoromethyl ketone warhead is positioned to react covalently with the cysteine in the caspase active site. Because of low oral bioavailability and limited brain penetrance, zVAD-fmk was delivered by infusion into the cerebral ventricle, which communicates with the spinal cord central canal through the fourth ventricle. In humans, intrathecal perfusion within the vertebral column would be preferred, but the small size of mice prevents access to that site for chronic infusion.

Infusion of zVAD-fmk into the cerebral ventricle extended survival of the SOD1G93A mice by nearly 4 weeks. At an intermediate time point after zVAD-fmk administration, there were a greater number of motor neurons in cervical sections of the spinal cord—although fewer in lumbar segments (perhaps because of poor penetrance of the inhibitor down the length of the spinal column)—in treated compared to control animals. There was also improved preservation of the axons of thoracic phrenic neurons that innervate the diaphragm. Treatment with zVAD-fmk also decreased interleukin-1β (IL-1β), an indication that caspase-1 activity was inhibited. The particular SOD1G93A transgenic mouse line used by Li et al. (5) normally shows the first outward signs of clinical disease at 3 months of age, as evidenced by tremor in two or more limbs. The duration of disease averages 30 to 40 days in the many laboratories that use this model. Thus, zVAD-fmk probably extends survival after the onset of clinical disease by ∼70%. In comparison, riluzole, which is thought to decrease the neurotoxic effects of the excitatory neurotransmitter glutamate, extends survival by ∼30% in the same SOD1G93A transgenic line and by about 3 months in ALS patients. Antioxidants and copper chelators also show neuroprotection in the mouse model, although their effect is primarily on disease onset rather than on disease duration. Clearly, an open issue is the extent to which treatment with multiple neuroprotective agents might produce an even greater clinical response, much as cocktails of protease and reverse transcriptase inhibitors produce greater clinical benefit in AIDS patients.

Caspase-3 is an obvious target for inhibiting the apoptotic cascade; its modes of activation have been well characterized, and its activity is largely responsible for the destruction of structural and maintenance proteins in the cell (9). Caspase-1 is another matter. Originally referred to as ICE (interleukin converting enzyme) because it promotes production of IL-1β from the proform, caspase-1 is considered to be a proinflammatory rather than an apoptotic caspase. An interesting aspect of the Li et al. report is the observation of increased levels of active caspase-1, both in the SOD1G93A mice and in the spinal cord of ALS patients, as well as increased levels of caspase-1 and caspase-3 mRNA in the transgenic mice. Up-regulation of caspase-1 correlates with increased levels of IL-1β, an activator of transcription, and this could explain the increased production of mRNAs for these caspases. Accordingly, the authors propose a non-cell-autonomous pathway of apoptosis that is subject to paracrine control through IL-1β (see the figure). These findings strongly suggest that apoptosis is a process that can be influenced by proinflammatory agents such as the cytokine IL-1β.

A number of hypotheses have been proposed to explain how mutant Cu,Zn SOD triggers the destruction of motor neurons. These include altered chemical reactivity of the mutant enzyme leading to the generation of reactive radicals, its aggregation (seen in SOD1G93A transgenic mice and ALS patients), and alterations in glutamate excitotoxicity. Fundamental questions remain with respect to how one or more of these mechanisms ultimately might trigger caspase activation and consequent motor neuron death. However, the links between earlier work (10, 11), the Li et al. findings, and these possible triggers of motor neuron death provide a compelling argument for the participation of apoptotic pathways, and for the value of caspase inhibitors as potential therapeutic agents in the treatment of ALS and other neurodegenerative diseases.

It is more than 60 years since Lou Gehrig, the champion of the New York Yankees, in his last press conference before retiring as a consequence of ALS, claimed to be “the luckiest man alive.” Although we have yet to hit a home run in the quest for a treatment or cure for ALS, work with transgenic mice builds hope that we are at bat in the final inning against this deadly disease.


View Abstract

Stay Connected to Science

Navigate This Article