Neurotrophins: key regulators of cell fate and cell shape in the vertebrate nervous system

Neurons are specialized cells with a complex morphology that represent the functional unit of the nervous system. They are generated in remarkable numbers, particularly in higher vertebrates. In the human brain, for example, there may be ∼85 billion neurons (Williams and Herrup 1988). There is little cell division in the adult nervous system of vertebrates, and in most areas, the final number of neurons is determined early in development, at about the time when neurons extend axons (Oppenheim 1991). Neuronal numbers are controlled both by cell-intrinsic and cell-extrinsic programs. Cell-intrinsic programs govern basic aspects of neuronal differentiation in vertebrates, and a number of transcription factors have been shown to be expressed in well-defined areas of the nervous system (for review, see Rubenstein et al. 1998). Cell-extrinsic mechanisms play a prominent role in vertebrates. They involve the secretion of diffusible molecules controlling the survival of neurons produced in excess early in development, a process thought to help match the size of neuronal populations with the territory they innervate (Purves 1988; Oppenheim 1991). However, much of the developmental growth of the animal must still take place by the time these numerical adjustments are completed (Purves 1988). The neurons that have escaped elimination grow proportionally with the organism, enlarging their size by adding dendrites that grow out from the cell bodies. Secreted proteins play a crucial role in the control of neuronal numbers and of dendritic growth. The best studied group is a family of structurally related molecules termed neurotrophins (Barde 1990). The first neurotrophin identified was originally designated “the” nerve growth factor (NGF; Levi-Montalcini 1966). However, only very few neurons were found to be NGF responsive in the central nervous system (CNS), and the isolation of brain-derived neurotrophic factor (BDNF) from the brain helped establish the concept that the fate and the shape of most vertebrate neurons can be regulated by diffusible growth factors (Hofer and Barde 1988). In the context of the regulation of neuronal shape, a particularly attractive and important feature of the neurotrophins is that they are synthesized and released by neurons and that both their biosynthesis and secretion depend on neuronal activity (Thoenen 1995). In addition to ngf and bdnf, two other neurotrophin genes have been identified in mammals, neurotrophin-3 (nt3) and neurotrophin-4/5 (nt4/5). These four genes encode pre-pro-neurotrophins. The processed proteins have a size of ∼13,000 D, and they exist in solution as noncovalently linked homodimers (for review, see Barde 1990; Ibanez 1998). They all have very basic isoelectric points, a somewhat unusual property for secreted proteins, which may serve the purpose of limiting their range of action. The structural hallmark of the protomer is a characteristic arrangement of the disulfide bridges known as the cystine knot (McDonald et al. 1991), later identified in other secreted proteins such as the plateletderived growth factors and the transforming growth factor– s (TGFs; McDonald and Hendrickson 1993). With the exception of NT4/5, neurotrophin sequences are highly conserved in mammals. In bony fishes, more neurotrophin and receptor genes have been isolated than in mammals (for review, see Hallbook 1999). Based on sequence comparisons and on the isolation of neurotrophin genes in various vertebrates, it is thought that ngf/nt3 and bdnf/nt4/5 evolved from separate duplication events (Hallbook 1999). The most primitive neurotrophin genes have been isolated from jawless fishes, a river lamprey and the Atlantic hagfish. The jawless fish lineage diverged about 460 million years ago in vertebrate history, and the neurotrophin receptors of the trk family (see below) seem to have coevolved with the neurotrophin genes (Hallbook, 1999). So far, no neurotrophin-like sequences have been detected in invertebrates typically used by geneticists, and unlike other growth factors such as members of the Wnt, fibroblast growth factor or TGFfamilies, genes coding for neurotrophins and their receptors have not been identified in the genome of the nematode Caenorhabditis elegans (Bargmann 1998). Clearly then, a nervous system can be put together in the absence of neurotrophins, including precise wiring, chemical neurotransmission, and the 1Corresponding author. E-MAIL yves.barde@neuro.mpg.de; FAX 49-89-8578-3749. Article and publication are at www.genesdev.org/cgi/doi/10.1101/ gad.841400.