Growth Factors Sculpt the Synapse

See allHide authors and affiliations

Science  28 Feb 1997:
Vol. 275, Issue 5304, pp. 1277-1278
DOI: 10.1126/science.275.5304.1277

The brain is a wonderfully adaptive organ, and much of this adaptive ability—the basis of learning and memory—resides in the plasticity of the cell-to-cell connections, the synapses. The strength of synapses can be changed by the pattern and nature of the stimuli they carry. A new class of modifiers has recently entered the picture with the realization that the adult nervous system reuses developmental growth molecules to promote synaptic changes in the mature brain. A report by Zhang et al. in this issue on page 1318 is an exciting example of this principle (1). These authors showed that transforming growth factor-β (TGF-β), a developmental signaling molecule associated with proliferation and dorsal-ventral patterning, can enhance synaptic communication between sensory and motoneurons in the marine mollusk Aplysia californica. Indeed, molecules active only in the embryo or only in the adult are becoming less and less frequent, suggesting that ontogeny is a continuous maturational process.

Aplysia can learn to modify several of its natural behaviors, including a defensive reflex in which the animal withdraws its gill and siphon in response to a harsh stimulus. If the animal is treated with, for example, a shock to the tail, the normally brief withdrawal response becomes sensitized and can persist for hours. Underlying this sensitization is facilitated synaptic transmission between the sensory and motoneurons that mediate the reflex; these are the same synapses studied by Zhang et al. The primary change, which is triggered by the release of the neurotransmitter serotonin from interneurons during the sensitizing stimulus, occurs in the sensory neurons; serotonin activates a cyclic AMP cascade that has immediate effects on ionic currents and later effects on gene transcription and translation. In a related study (2), the same group of investigators identified, by differential display techniques and ribonuclease protection assays, an mRNA that increases in sensory neurons after exposure to serotonin; the mRNA is from an Aplysia homolog of the Drosophilatolloid and human bone morphogenetic protein (BMP-1) genes, which encode secreted metalloproteases that act on members of the TGF-cytokine superfamily. Taking their cue from the developmental association of tolloid/BMP-1 with TGF-β and its homologs, Zhang et al. tested whether TGF-β could alter synaptic transmission at Aplysia synapses (1). They found that the application of TGF-β to sensory motoneuron synapses caused an increase in synaptic strength, measured 24 or 48 hours later. Is TGF-β signaling actually part of the serotonin-stimulated biochemical cascade that underlies sensitization? This question is hard to address at the behavioral level, but Zhang et al. do show that a TGF-β receptor antagonist can block some of the actions of serotonin on sensory neurons.

An important question to answer now is how TGF-β strengthens the synapse. Zhang and colleagues propose that activated TGF-β stimulates a second round of protein translation (the first being initiated by serotonin), but the possibility remains that TGF-β acts directly at the synapse through the activity of its receptor serine-threonine kinase and local downstream signals (see figure). This question highlights a general issue for consideration: To what extent do growth factors that modulate the mature brain use the same signal transduction cascades as those used during development?

The action of growth factors at the adult synapse.

5-HT, serotonin.

The results of Zhang et al. add breadth to a burgeoning area of research—the role that growth factors play in sculpting synaptic transmission in the developing and mature brain. In the developing visual cortex, for example, neurotrophins may participate in the activity-dependent strengthening of synapses conveying eye-specific information (3). Addition of an excess of a TrkB ligand can either block the formation of ocular dominance columns (4) or prevent the shrinkage of cell soma caused by monocular deprivation (5). It is as yet unclear whether the neurotrophins directly promote the segregation of eye-specific axons or set a threshold upon which other signals act to form the ocular dominance columns. As is true for many other synaptic signals, not much is known about how the neurotrophins, interacting with different levels of synaptic activity, might exert strengthening, weakening, or null effects on synaptic transmission. Moreover, it is not clear how the kinetics of neurotrophin release and “clearance” could subserve the function of a molecular coincidence detector that likely acts on a time scale of milliseconds.

Neurotrophins can also modulate synapses in adult animals. In many brain areas, including the hippocampus, the expression of the neurotrophins and their receptors persists well into adulthood. Long-term potentiation, a form of activity-dependent plasticity exhibited in the hippocampus and other areas, is blunted in mice that lack the gene for brain-derived neurotrophic factor (BDNF) (6, 7). In addition, in adult hippocampal slices and cultured neurons, the addition of either BDNF or neurotrophin-3 (NT-3) can cause a dramatic and long-lasting increase in synaptic transmission (8, 9). The enhancement in hippocampal slices displays a very early dependence on protein synthesis (10), which is not somatic in origin, raising the possibility that neurotrophins may stimulate the synthesis of proteins in dendrites and promote site-specific modifications of synaptic function.

Is the potentiation of synaptic transmission produced by growth factors due to, or accompanied by, structural changes, such as the addition of new synapses or alterations in the shape of existing synapses? There are now several instances in developing nervous systems where growth factors can cause changes in the morphology of pre- and postsynaptic elements (11, 12). Except for a few examples (1114), however, clear demonstrations of learning-related changes in mature synaptic structures have been much harder to come by. Indeed, some recent attempts to observe directly structural changes associated with synaptic plasticity in the mature hippocampus have yielded somewhat sobering results-little or no change in synaptic morphology was observed after long-term potentiation (15, 16). And yet, just imagining the potential for remodeling synaptic structures during learning (and in combating neurodegenerative diseases) is sufficient motivation to continue to search for structural changes in mature synaptic structures.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
View Abstract

Navigate This Article