Excitotoxic cell death, the end for many cells after ischemic brain injury and in some neurodegenerative diseases, is triggered by glutamate-induced calcium influx. Downstream, activation of the p38α map kinase leads to apoptosis, but just how calcium and p38 connect has been an open question. New results from Michael Courtney’s lab at the University of Kuopio in Finland show that the small GTPase Rho is the missing link. Their work, published in the March 18 online edition of Nature Neuroscience, establishes Rho activation by calcium influx as a necessary and sufficient event to turn on p38 and cause cell death in primary neurons.

Their results open up a potential new target for drugs to block excitotoxicity, and may have some additional relevance to Alzheimer disease. Inhibitors of Rho and its downstream kinase Rock can modulate amyloid-β (Aβ) peptide production (Zhou et al., 2003; Leuchtenberger et al., 2006). In addition, Rho is farnesylated, and changes in Rho/Rock pathway activity have been implicated in alterations of amyloid precursor protein processing in response to statins (see ARF related news story).

Finding Rho in the excitotoxicity pathway was unexpected, the authors write, since many previous attempts have failed to detect Rho-mediated activation of p38α. This is because the pathway only functions in primary neurons, and not in cultured cell lines, they found. To look at Rho’s signaling role in response to excitotoxic stimuli, first authors Maria Semenova, Anu Maki-Hokkonen, and Jiong Cao used a strategy to inactivate Rho by expressing the Clostridium botulinum C3 exoenzyme, a toxin which ADP ribosylates and inactivates Rho proteins specifically. When the toxin gene was expressed in primary cerebellar neurons, it blocked activation of p38 by glutamate without affecting calcium influx. They subsequently found that glutamate stimulation activated Rho, but not the related protein Rac, in the same neurons. Rho activation was also seen in mouse brain after ischemia.

Blocking Rho not only inhibited p38α activation, but it also decreased cell death. The same effect was elicited with a dominant negative Rho mutant, suggesting that the protein was necessary for cell death. Finally, they showed that expression of constitutively active Rho was sufficient to cause cell death. They found a similar role for Rho in p38 activation and cell death after glutamate stimulation in cortical and hippocampal neurons in culture.

Using a sensitive florescence resonance energy transfer (FRET) assay for Rho activation in primary cell culture, the investigators showed that glutamate activated Rho in an NMDA receptor-dependent manner. They found a rapid, glutamate concentration-dependent increase in calcium levels, Rho activation, p38 activation, and cell death, further supporting the idea that these events are all linked. However, they did not find out how calcium stimulated Rho, although they ruled out two candidate pathways, the Rho115 guanine exchange factor and the calmodulin-dependent kinase II. Likewise, the downstream targets of Rho are also unknown, but the investigators showed the pathway did not require either RockI or II kinase activity.

The results fill in one more piece of the puzzle of excitotoxic cell death, and open up the possibility of a new target for inhibiting the pathway. “Inhibition of Rho has proven to be beneficial in models of axonal regeneration, and regulation of the release of amyloidogenic Aβ42 peptide from neuroblastoma cells," the authors write. “In light of the present data, Rho inhibition might also help to increase survival subsequent to excitotoxic challenge." The map kinases including p38 and its cousins are involved in many aspects of neuron life and death (see ARF related news story), suggesting that more roles for Rho remain to be discovered.—Pat McCaffrey


  1. Semenova et al. claim that a "necessary and sufficient" death program was activated by just a few seconds in 30 mM KCl. But when the trail of references is traced backwards to find their culture conditions, it seems that the cerebellar granule cells they used were routinely cultured in 25 mM KCl. Can a 5 mM elevation of KCl for just a few seconds really kill cerebellar granule cells? These authors and others have shown previously that high KCl is necessary for cerebellar granule cell survival. Other reports indicate that glutamate itself is a survival factor for these cells.

    It would also have been nice to see citation of Bossy-Wetzel et al. (2004). These authors showed evidence for a different mechanism to activate p38 by NMDA receptors—one that could explain a caspase-independent programmed cell death that is sensitive to Bcl-2.


    . Crosstalk between nitric oxide and zinc pathways to neuronal cell death involving mitochondrial dysfunction and p38-activated K+ channels. Neuron. 2004 Feb 5;41(3):351-65. PubMed.

  2. Many an Alzheimer’s researcher has pondered the question of whether there is some way for a neuron to release excess Aβ as a sign of ill health and imminent death. Andrea LeBlanc among others provided evidence many years ago that cultured neurons undergoing apoptosis released excess Aβ. Giuseppina Tesco and Rudy Tanzi followed up on this in 2003 in JBC and again in the Hot Topics session at ICAD in Madrid (Madrid BACE News Roundup: Part 3). There, Tesco and Tanzi reported progress in elucidating a molecular basis for apoptosis-induced Aβ generation in their demonstration that the sorting protein GGA3 was cleaved during apoptosis, thereby elevating BACE levels and/or access to APP, and so favoring Aβ generation. So, apoptosis clearly fulfills the criteria for being an amyloidogenic cell death pathway.

    New data from Semenova and colleagues converge with data from our own lab and that of Giulio Pasinetti to implicate Rho signaling as another way to cause release of excess Aβ. The Semenova paper focuses on excitotoxic cell death, a death pathway long associated with excess calcium influx and excess activation of stress-associated protein kinases (SAPKs) such as p38a. Through a series of elegant and convincing experiments, Semenova demonstrates that Rho is required for glutamate-induced activation of SAPKs, and that Rho is activated when calcium levels rise. Rho toxicity was reversed by Bcl-2 and appeared to be an essential component of excitotoxic cell killing. Curiously, Rho actions were not blocked by Y-27632, prompting the conclusion that Rho was acting independently of Rho kinase (ROCK). This last conclusion should be drawn cautiously, however, since we found examples of ROCK involvement that were resistant to Y-27632. In our hands, ROCK activation is “bad” because ROCK tonically suppressed basal α-secretase-type, non-amyloidogenic APP ectodomain shedding.

    In parallel work, Pasinetti and colleagues discovered that caloric restriction (CR) lowered brain amyloid burden in plaque-forming transgenic mice in a pathway that appeared to involve sirtuins. Their report coincided exactly with our report that isoprenoids modulate non-amyloidogenic α-secretase-type ectodomain shedding via ROCK. Recognizing a possible link, Pasinetti and colleagues worked in collaboration with our group to implicate ROCK as a modulator of APP metabolism downstream of CR. The linkage appears to be due to sirtuin control of ROCK expression.

    In both situations, Rho is the common link that associates Aβ metabolism to CNS insults in parallel or in sequences of events that extend beyond the traditional “amyloid hypothesis” in which all neurotoxicity is caused by Aβ. Rather, it would appear that while Aβ is frequently at “the scene of the crime,” causing neurotoxicity, there are circumstances in which more indirect and complex relationships underlie its involvement.


    . Increased production of 4 kDa amyloid beta peptide in serum deprived human primary neuron cultures: possible involvement of apoptosis. J Neurosci. 1995 Dec;15(12):7837-46. PubMed.

    . Caspase activation increases beta-amyloid generation independently of caspase cleavage of the beta-amyloid precursor protein (APP). J Biol Chem. 2003 Nov 14;278(46):46074-80. PubMed.

    . Modulation of statin-activated shedding of Alzheimer APP ectodomain by ROCK. PLoS Med. 2005 Jan;2(1):e18. PubMed.

    . Neuronal SIRT1 activation as a novel mechanism underlying the prevention of Alzheimer disease amyloid neuropathology by calorie restriction. J Biol Chem. 2006 Aug 4;281(31):21745-54. PubMed.

    View all comments by Samuel Gandy

Make a Comment

To make a comment you must login or register.


News Citations

  1. Statins Boost α-Secretase, but Not Through Cholesterol
  2. Signal Transduction and Memory—SCOP Scuppers MAPK Pathway

Paper Citations

  1. . Nonsteroidal anti-inflammatory drugs can lower amyloidogenic Abeta42 by inhibiting Rho. Science. 2003 Nov 14;302(5648):1215-7. PubMed.
  2. . Inhibitors of Rho-kinase modulate amyloid-beta (Abeta) secretion but lack selectivity for Abeta42. J Neurochem. 2006 Jan;96(2):355-65. PubMed.

Further Reading


  1. . Novel modulators of amyloid-beta precursor protein processing. J Neurochem. 2007 Jan;100(2):314-23. PubMed.
  2. . Inhibitors of Rho-kinase modulate amyloid-beta (Abeta) secretion but lack selectivity for Abeta42. J Neurochem. 2006 Jan;96(2):355-65. PubMed.
  3. . Rho kinase, a promising drug target for neurological disorders. Nat Rev Drug Discov. 2005 May;4(5):387-98. PubMed.

Primary Papers

  1. . Rho mediates calcium-dependent activation of p38alpha and subsequent excitotoxic cell death. Nat Neurosci. 2007 Apr;10(4):436-43. PubMed.