nature cell biology doi:10.1038/ncb1102-e253 november 2002 volume 4 issue 11 pp E253 - E255 Integrins as developmental switches Trent A. Watkins and Ben A. Barres Trent Watkins and Ben Barres are in the Department of Neurobiology, Stanford University School of Medicine, Fairchild Building Room D235, 299 Campus Drive, Stanford, CA 94305-5125, USA e-mail: [email protected] Neuregulin (NRG) is a crucial regulator of oligodendrocyte development, strongly promoting both the proliferation of oligodendrocyte precursor cells and the survival and maturation of oligodendrocytes. How can the same growth factor mediate such different effects? New work in this issue of Nature Cell Biology implicates an integrin-mediated switch in signalling that results in the loss of the proliferative response and the enhancement of survival and maturation. he developing nervous system faces the daunting task of coordinating the numbers of cells to provide precise numbers of functional contacts for target cells. A common approach for achieving this type of target matching in the central nervous system (CNS) is the overproduction of cells followed by their apoptosis if they fail to make the appropriate connections. A good example of this strategy is the matching of the number of myelin-forming oligodendrocytes with the length and number of axons to be myelinated. In the developing optic nerve, for instance, a large portion of newly formed oligodendrocytes die, leaving only those that have successfully contacted and ensheathed axons. This suggests that contact with target-axons somehow promotes the survival of the proper number of oligodendrocytes. Consistent with this possibility, transection of the optic nerve results in widespread oligodendrocyte death, whereas experimentally increasing the number of surviving axons in the developing optic nerve results in a corresponding increase in the number of surviving oligodendrocytes1. Thus, a competition for limiting amounts of axon-bound survival factors is the primary determinant of the fate of newly formed oligodendrocytes. The report by Colognato et al. in this issue provides new insight into the interactions of two recently proposed axon-linked survival signals, NRG and integrin ligands, and suggests an intriguing mechanism for the emergence of axon-dependent survival in newly formed oligodendrocytes2. NRG is a member of a family of soluble and membrane-bound growth factors that are related to epidermal growth factor (EGF), and is essential for the development of myelinating cells throughout the nervous system3. Interaction of NRG with the receptor tyrosine kinases erbB2, erbB3 or erbB4 can activate at least two different survival-promoting signalling cascades: the phosphatidylinositol-3-OH kinase (PI(3)K)/Akt pathway and the mitogen-activated protein kinase (MAPK) pathway4, 5. Neutralization of NRG in cocultures of neurons and oligodendrocytes negates the survival-promoting effects of contact with axons, and delivery of NRG to the developing optic nerve decreases normal oligodendrocyte death6. In addition, spinal cord explants from erbB2-knockout mice fail to develop mature oligodendrocytes7. Thus, NRG can function as an axon-derived survival factor for maturing oligodendrocytes. These survival-promoting effects of NRG on maturing oligodendrocytes have been difficult to reconcile with its effects on other stages of the oligodendrocyte lineage. NRG is a powerful mitogen for oligodendrocyte precursor cells (OPCs) and strongly inhibits their differentiation8. In fact, NRG can even induce a phenotypic reversion of oligodendrocytes; when treated with high levels of NRG in culture, oligodendrocytes reduce their expression of mature markers, retract some of their processes and can re-enter the cell cycle5. Paradoxically, these results suggest that newly formed oligodendrocytes that contact NRG-expressing axons would fail to mature any further. So how do immature oligodendrocytes, reaching out to find axons, overcome the mitogenic effects of NRG and grow to become dependent on NRG for their survival? The new work by Colognato et al. suggests a novel mechanism for this context-dependent response to NRG at distinct stages of oligodendrocyte development2. They began with their previous observation that the laminin receptor 61 integrin helps to mediate the ability of axons to promote oligodendrocyte survival9. Integrins are a large family of heterodimeric cell-surface receptors for extracellular matrix (ECM) proteins. Functioning as both cell adhesion and signalling molecules, integrins mediate a vast array of biological effects, from migration to proliferation and survival10. A role for integrins in axon-dependent oligodendrocyte survival was suggested from findings that anti-61 antibodies reduced the survival-promoting effects of neurons in coculture with oligodendrocytes9. Additionally, laminin-2, an 61 integrin ligand, potentiated the survival of newly-formed oligodendrocytes in response to low, physiological levels of platelet-derived growth factor (PDGF). This result will come as no surprise to those familiar with integrin signalling. As key adhesion molecules allowing cells to sense and interact with their environment, integrins are well known to be signal transducers and modulators of growth factor functions. Signals from growth factors and integrins can influence each other through both direct receptor interaction and intersecting signalling networks11. So what function does integrin signalling have in oligodendrocyte development in vivo? This question is addressed in the latest study through the evaluation of 6-integrin-deficient mice2. These knockout mice die at birth from severe skin blistering, before myelination begins in many regions of the CNS. Fortunately, oligodendrocyte development can nonetheless be studied by examining brain-stem tracts that are normally myelinated before birth. Although the number of newly formed oligodendrocytes was similar to those from wild-type mice, there was a significant reduction in the number of more mature, myelin basic protein-expressing oligodendrocytes that resulted, at least in part, from increased apoptosis of the maturing cells. When Colognato et al. grew isolated wild-type oligodendrocytes in culture on laminin-2, the survival-promoting effects of NRG were greatly enhanced. Taken together, these new findings show that integrin-mediated signalling normally regulates oligodendrocyte survival and that it may do so by enhancing NRG signalling. These observations raise the question of how integrin signalling enhances the effects of NRG. A clue came when the authors found that the effects of NRG and PDGF treatment on the survival of oligodendrocytes grown on laminin-2 were additive2. This was unexpected because the promotion of survival by both PDGF and NRG had previously been shown to be dependent on PI(3)K signalling4. This result raised the possibility that after 61 ligation, NRG and PDGF may function through separate signalling pathways. Colognato et al. discovered that when oligodendrocytes were cultured on laminin-2, PI(3)K inhibitors blocked the survival-promoting effects of PDGF, but not those of NRG2. Instead, inhibitors of the MAPK pathway blocked the survival response of oligodendrocytes to NRG. These findings show that integrin ligation switches the survival signalling cascade activated by NRG from a PI(3)K-dependent to a MAPK-dependent pathway. This switch has an important functional consequence. Whereas activation of PI(3)K by NRG in the absence of laminin-2 inhibits differentiation and myelin membrane formation, activation of MAPK by NRG in the presence of laminin-2 enhances survival and maturation. In this way, integrin-mediated signalling offers a resolution to the seemingly paradoxical actions of NRG on developing oligodendrocytes. These new data also provide an elegant model for how a single growth factor can induce a variety of responses at distinct stages of development within a lineage (Fig. 1). OPCs proliferating in the context of axons receive mitogenic signals from axons and other sources. As OPCs differentiate and extend processes to contact axons, only those newly formed oligodendrocytes that are signalled by both integrin ligands and NRG on axons will be allowed to fully differentiate and survive. Those that only receive signals from NRG (perhaps including soluble isoforms from axons and other sources) will either be prevented from differentiating by activation of the PI(3)K pathway or will fail to survive for lack of the integrin-mediated enhancement of the survival signal provided by axonal contact. Excess oligodendrocytes that are unable to contact target axons will die from lack of trophic support, in this way precisely matching oligodendrocytes to axons. Thus, the activation of 61 integrins provides a means for a context-appropriate response to NRG. This model raises some interesting questions. In developing axon tracts, how are the levels of NRG and integrin ligands regulated? It is possible that the onset of expression of one or both of these may help to trigger the start of myelination. Immature oligodendrocyte processes that contact NRG-expressing axons prematurely in the absence of integrin-mediated signals may be blocked from further maturation. Then, when myelination is set to proceed, the requirement for at least two axon-derived signals may help to ensure that only those oligodendrocytes that have made sufficient contact with a number of axons will survive. Is the integrin-mediated switch in signalling reversible? The answer to this question has important implications for demyelinating conditions, such as multiple sclerosis (MS), in which oligodendrocytes may lose contact with axons. Does exogenous NRG, which is known to be beneficial in animal models of MS12, support the survival of mature oligodendrocytes or promote their reversion to a more immature state in preparation for remyelination? Although these questions relate specifically to oligodendrocyte development and remyelination, many researchers will be intrigued by the broader implications of these new findings. Integrins, already appreciated for their ability to modulate signals from growth factors and function as signalling molecules in their own right, now provide a mechanism for a complete switch in growth-factor response within a cell lineage, from blocking to promoting differentiation. Such a mechanism may help migrating precursors in a variety of contexts know when they have found their destinations; encountering the appropriate integrin ligands may switch progenitor responses to mitogenic signals, leading them to settle down and differentiate. Similarly, could integrin-mediated signalling be involved in converting the responses of growth cones to neurotrophins from elongation to maturation when a presynaptic axon contacts its target neuron? With so many functions for integrins, from adhesion to signalling, we can't help but wonder what they will be up to next. References 1. Barres, B. A. & Raff, M. C. J. Cell Biol. 147, 1123-1128 (1999). | Article | PubMed | ISI | 2. Colognato, H. et al. Nature Cell Biol. 4, 833-841 (2002). | Article | 3. Adlkofer, K. & Lai, C. Glia 29, 104-111 (2000). | Article | PubMed | ISI | 4. Flores, A. I. et al. J. Neurosci. 20, 7622-7630 (2000). | PubMed | ISI | 5. Canoll, P. D., Kraemer, R., Teng, K. K., Marchionni, M. A. & Salzer, J. L. Mol. Cell. Neurosci. 13, 79-94 (1999). | Article | PubMed | ISI | 6. Fernandez, P. A. et al. Neuron 28, 81-90 (2000). | PubMed | ISI | 7. Park, S. K., Miller, R., Krane, I. & Vartanian, T. J. Cell Biol. 154, 1245-1258 (2001). | Article | PubMed | ISI | 8. Canoll, P. D. et al. Neuron 17, 229-243 (1996). | PubMed | ISI | 9. Frost, E. E., Buttery, P. C., Milner, R. & ffrench-Constant, C. Curr. Biol. 9, 1251-1254 (1999). | Article | PubMed | ISI | 10. Miranti, C. K. & Brugge, J. S. Nature Cell Biol. 4, E83-E90 (2002). | Article | PubMed | ISI | 11. Schwartz, M. A. & Ginsberg, M. H. Nature Cell Biol. 4, E65-E68 (2002). | Article | PubMed | ISI | 12. Marchionni, M. A. et al. Adv. Exp. Med. Biol. 468, 283-295 (1999). | PubMed | ISI |