Behind every successful neuron, there are dozens of good glia. Far from being an inert scaffold, astrocytes and microglia are increasingly recognized to be active participants in the life and death of neurons (see related ARF Live Discussion). Now, work from the lab of developmental neurobiologist R. Douglas Fields reveals a new and unexpected role for astrocytes in the myelination of axons in the central nervous system. In a report in the March 16 Neuron, Fields and colleagues at the National Institute of Child Health and Human Development in Bethesda and the National Cancer Institute in Frederick, Maryland, show that the activity-dependent myelination of neurons in culture by mature oligodendrocytes requires astrocytes. Specifically, the astrocytes respond to ATP produced by electrically stimulated axons by themselves producing the cytokine leukemia inhibitory factor (LIF), which in turn stimulates oligodendrocytes to increase myelination.

As pointed out in an accompanying commentary by Ivo Spiegel and Elior Peles of the Weizmann Institute, Rehovot, Israel, the study raises the intriguing possibility that glial cells can participate in activity-dependent plasticity of myelination in mature neurons. Such a scenario would present new opportunities to intervene in demyelinating diseases like multiple sclerosis. Perhaps it could also help explain the demyelination that occurs in Alzheimer disease, possibly in response to changes in neural activity.

Myelination of axons by oligodendrocytes occurs mainly during nervous system development, but it continues into adulthood for some nerves. Early on, electrical activity regulates myelination via adenosine-mediated effects on oligodendrocyte proliferation and differentiation. However, the regulation of myelination later, by mature oligodendrocytes, has not been as well understood.

Using an in-vitro system where dorsal root ganglion (DRG) explants and mature oligodendrocytes were cultured together in a special chamber, first author Tomoko Ishibashi was able to show a threefold increase in myelination when the DRG neurons were electrically stimulated. Adding a non-hydrolyzable ATP analog to cultures mimicked this result, and further experiments showed that increased myelination by either the ATP analog or electrical stimulation required the cytokine LIF, which was being produced in the cultures.

That production was traced to astrocytes, and Ishibashi et al. showed that the ATP released from neurons as a result of electrical activity increased LIF mRNA in, and release from, astrocytes. To prove the astrocytes were indispensable, they repeated the experiments in cultures lacking these cells, and showed myelination could not be stimulated. Adding back astrocytes restored the effect, but not if they came from LIF knockout mice.

The results could explain defects in myelination found in either GFAP knockout mice, which lack astrocytes, and also in LIF knockout mice. In each case, the phenotypes support a role of astrocytes in modulating myelination, rather than as primary regulators. “Taken together, these results reveal a new mechanism by which electrical activity promotes myelination of CNS axons at a later developmental stage and possibly into postnatal life,” the authors conclude.—Pat McCaffrey


Make a Comment

To make a comment you must login or register.

Comments on News and Primary Papers

  1. The paper by Ishibashi, Fields, and colleagues describes a systematic and thorough series of experiments that does much to advance our understanding of the interactions between axonal electrical activity and induction of myelination programs in oligodendrocytes. The studies elucidate an indirect pathway that is initiated by axonal ATP release during action potential activity and mediated by astrocyte release of the cytokine, leukemia inhibitory factor (LIF), which acts in a paracrine mode to activate oligodendrocyte myelination processes.

    As noted in the article, several of the steps in this sequence have been described previously, including the links of LIF and astrocyte functions to oligodendrocyte maturation and myelination (e.g., Bugga et al., 1998; Meyer-Franke et al., 1999). However, the present work identifies a novel integrated pathway, providing the first evidence that activity-dependent promotion of myelination is mediated by ATP release from axons followed by ATP-induced LIF release from astrocytes. Thus, it adds importantly to our growing but still nascent understanding of the complexity of molecular interactions among the multiple cell types of the CNS.

    In the context of Alzheimer disease (AD), the Ishibashi et al. work, and related studies, appear particularly relevant to an “oligodendrocyte growth factor” model of AD that we proposed in a recent article (Fig. 3 in Blalock et al., 2004). That model arose from our statistical analyses of gene microarray data from hippocampal tissue of control and incipient AD subjects. Those analyses showed that a substantial number of oligodendrocytic growth factor genes, as well as many tumor suppressor genes, were up-regulated and correlated with cognitive and pathological markers of AD, even in the early stages of the disease. To account for these findings, we suggested that oligodendrocyte activation, associated with up-regulation of oligodendrocyte growth factors, is triggered as an early event in AD pathogenesis, possibly in response to myelin damage or endogenous dysregulation. This up-regulation results in elevated release of the growth factors and excessive paracrine stimulation of surrounding cells. To control the effects of this excessive growth stimulation, adjacent neurons and glial cells up-regulate cell type-specific compensatory tumor suppressor and differentiation responses, resulting in the axonopathy, tangles, β amyloid generation, and synaptic loss underlying the cognitive impairment of AD. We also suggested that the sequential overactivation of adjacent oligodendrocytes might help account for why AD appears to advance along myelinated fiber pathways from entorhinal cortex to hippocampus and neocortex.

    Of course, much additional testing will be required to assess this hypothesis. In particular, the possible mechanisms that initiate the putative oligodendrocyte activation are clouded in the model. However, the studies by Ishibashi et al. and others showing that astrocytic factors can modulate activation of myelination programs in oligodendrocytes appear to raise the interesting additional possibility that reactive astrocytes, which are prominent around amyloid plaques in AD and even in normal brain aging, might exhibit up-regulated LIF and other cytokines, and thereby play a role in triggering an oligodendrocytic growth/myelination response. If further studies continue to support the oligodendrocyte model of pathogenic progression in AD, the question of initiating factors will become increasingly important, and studies such as the present work may prove highly relevant to AD.

    View all comments by Philip Landfield
  2. This paper has significant potential implications for multiple sclerosis, and one can reasonably speculate about potential relevance to Alzheimer disease, as well. The general principle the authors are describing, that is, that electrical activity in neurons releases a signal—in this case ATP—that tells the astrocytes to release other signals that then feed back on neighboring cells such as oligodendrocytes is quite interesting.

    One possible connection to neurodegenerative disease is that there is much evidence that astrocytes are releasing signals that are crucial for the promotion of CNS neuron survival, though no one yet knows what these astrocyte-derived trophic signals are. Could electrical activity in neurons induce astrocytes to release more of these neurotrophic signals? If so, decreased activity with aging or in neurodegenerative disease certainly might lead to less release of trophic signals (which in turn could lead to failure of myelin maintenance).

    In fact, I showed previously that LIF and CNTF are co-mitogens for oligodendrocyte precursor cells (Barres et al., 1993). If activity is inducing astrocytes to release LIF, this could mean that when activity is blocked, there is also less proliferation of oligodendrocyte precursor cells and therefore less new oligodendrocyte generation. Over time, this might lead to loss of myelin in humans. Such a loss of myelin has recently been found in the temporal lobes of patients with major depressive disorder (Aston et al., 2005). I wonder if decreased activity in a temporal lobe pathway, perhaps because of stress, might eventually lead to depression by causing myelin loss. If so, then it’s fair to speculate that major depressive disorder could be a type of neurodegenerative disease.


    . Multiple extracellular signals are required for long-term oligodendrocyte survival. Development. 1993 May;118(1):283-95. PubMed.

    . Protein kinase CK2 phosphorylates the cell cycle regulatory protein Geminin. Biochem Biophys Res Commun. 2004 Mar 19;315(4):1011-7. PubMed.

  3. Myelin, brain aging, and Alzheimer disease
    The protracted myelination of the human brain throughout life results in a roughly quadratic (inverted U) trajectory of myelin content, reaching a maximum in mid-life and then declining in older age. The extensive scope of myelination is arguably the most uniquely human aspect of our brain. It results in the high processing speeds underlying our cognitive functions, and is extremely vulnerable during both brain development and degeneration. In this "myelin model" of the human brain, the breakdown of myelin integrity in old age is hypothesized to also be the first step in the development of uniquely human age-related diseases such as Alzheimer disease (AD) (for review, see Bartzokis, 2004, 2004a).

    The model posits that many of the risk factors associated with AD, such as brain cholesterol and iron levels, head trauma, and apolipoprotein E (ApoE) alleles, may affect age-related myelin breakdown and thus contribute to the ultimate manifestations of age-related cognitive decline and degenerative brain disorders. For example, ApoE genotype shifts the age at onset of AD, and is the most influential AD risk factor after age itself. We recently demonstrated that in regions that myelinate later than others, for example, the frontal lobes, the trajectory of age-related myelin breakdown in healthy older individuals is altered by ApoE alleles (Bartzokis et al., 2006). Postmortem and in vivo data indicate that myelin breakdown progresses at different rates (in the absence of gross axonal damage) and eventually culminates in AD, which is characterized by more severe myelin breakdown than matched healthy controls (Bartzokis, et al., 2003).

    By pointing out yet another pathway by which brain activity may increase signals instructing the brain to myelinate, the article by Ishibashi et al. and commentary by Spiegel and Peles suggest once again that the old saying "use it or lose it" also applies to the brain. Thus, this interesting article has indirect relevance to AD, as well as to finding ways by which we could eventually change our trajectory of age-related cognitive decline and thus possibly postpone AD. In the meantime, keeping your mind active and occupied with fun things should be just that: fun and possibly good for you.

    View all comments by George Bartzokis


Webinar Citations

  1. Are Glia Active Participants in Neurodegenerative Disease?

Further Reading

No Available Further Reading

Primary Papers

  1. . Astrocytes promote myelination in response to electrical impulses. Neuron. 2006 Mar 16;49(6):823-32. PubMed.
  2. . A new player in CNS myelination. Neuron. 2006 Mar 16;49(6):777-8. PubMed.