How to Enter  Winning Entries  Past Prize Coverage  Contact Information  Eppendorf & Science Prize Press Releases

Establishing Synaptic Independence: How Neurons Create Diffusional Barriers

By Bernardo Sabatini

The computational power of the human brain arises from the enormous number of neurons (~10¹⁰) and synapses (~10¹⁴) that it contains. In order to fit so many signaling elements within a relatively small volume, a remarkable degree of miniaturization has been achieved. For example, synapses, the contact points between two neurons at which information is exchanged, are often found at densities greater than 3/m³ and are associated with cellular structures that can be far smaller than 1 m in diameter (1). Despite this tight packing, each synapse is thought to operate independently. This is exemplified by the ability of specific patterns of activity to trigger the regulation of a single synapse without affecting its neighbors (2). This degree of precision is surprising because synaptic plasticity is typically induced by the intracellular accumulation of diffusible molecules, such as calcium (3-5). If these molecules diffused freely within the neuron, they would reach inactive synapses within a few milliseconds and the activity-dependent regulation of individual synapses would not be possible.

What biophysical specializations have neurons developed to prevent the spread of regulatory signals from active synapses to inactive neighbors? Answering this question is difficult because the biophysical properties of the tiny cellular compartments associated with individual synapses cannot be probed by conventional biochemical and electrophysiological methods. For this reason, we and others make extensive use of two-photon laser microscopy and photoactivation to manipulate and monitor the activity of an individual synapse within a living brain slice (6-15). Below I summarize the results of recent studies in which we examined the mechanisms that maintain the biochemical isolation of synapses in two morphologically distinct types of neurons (10, 12).

In many classes of neurons, excitatory synapses are sequestered onto dendritic spines, a specialized cellular compartment consisting of a tiny, bulbous spine head (<1 femtoliter in volume) that is separated from the dendrite by a thin neck (~100 nm in diameter) [reviewed in (16)]. The spine head contains the machinery necessary to read out and regulate the activity of the associated synapse, and its biochemical isolation may be the key mechanism that allows the independent regulation of each synapse. We measured the diffusional isolation of excitatory synapses made onto spines of hippocampal pyramidal neurons by monitoring the movement of photoactivatable green fluorescent protein (PAGFP) across the spine neck (10). This genetically engineered protein has the useful property that its fluorescence can be turned on by a pulse of light (17). Using a custom-built microscope, PAGFP was activated in individual spines and the movement of the activated protein was tracked. We found that the strength of diffusional coupling across the spine neck is highly variable, spanning nearly three orders of magnitude. Furthermore, we uncovered a class of highly diffusionally isolated spines that retained activated PAGFP within the spine head for many seconds. PAGFP is of similar size (28 kD) to proteins involved in the regulation of synapses such as calmodulin, small GTPases, and some kinases and phosphatases (10 to 40 kD). Thus, the retention of these proteins in the spine head following activation by a synaptic stimulus is likely to exhibit similar heterogeneity.

Diffusional isolation of synapses in spiny and aspiny neurons: (A) Image of a pyramidal neuron in a hippocampal organotypic slice culture (left) and a higher magnification showing an apical dendrite with numerous spines (right). A thin axon with bulbous boutons is also seen on the right. Synapses are made at the points of contact between dendrites and axons. This neuron has been transfected with GFP and green fluorescence was used to form the images. (B) Image of a dendrite and spine of a hippocampal pyramidal neuron transfected with PAGFP and dsRed, a genetically encoded red fluorophore. Scale bar, 1 m. (C) Red (dsRed) and green (PAGFP) fluorescence acquired in the line scan show in (B). At the time indicated by the arrowhead, a laser pulse was used to activate PAGFP in the spine head. The depicted spine was initially strongly diffusionally isolated from the parent dendrite (left) but spontaneously became more coupled to the dendrite (right) a few minutes later. The time course of decay of green fluorescence in the spine head was used to calculate the time constant of diffusional equilibration (_equ) across the spine neck. Scale bar, 200 ms. (D) Image of a stellate cell (red fluorescence) overlaid on a laser-scanning differential interference image of the cerebellar slice. The arrowhead indicates the position of the stimulating electrode in the molecular layer. On the right is seen the recording electrode used to fill the cell with fluorophores and to monitor synaptic currents. (E) Top: Image of a stellate cell dendrite depicting three regions that displayed independent, synaptically evoked calcium transients. Bottom: The green fluorescence from a calcium-sensitive indicator was collected in a line scan along the dendrite during three independent trials of synaptic stimulation (left, middle, and right). Scale bars, 2 m and 150 ms. (F) The relative increases in green fluorescence (G/R_syn), indicative of calcium influx, are shown for each of the highlighted regions (red, blue, and green lines) and stimulation trials (left, middle, and right) shown in (E).

The optimal diffusional isolation of each spine likely depends on the state of the associated synapse. For example, at certain times the synapse may need to exchange components with the dendrite, requiring a relatively open neck. At other times, it may be beneficial to constrict the neck and favor the accumulation of plasticity-inducing molecules in the spine head. We found that the strength of diffusional coupling across the neck is a dynamically regulated parameter that is adjusted in response to changes in activity levels. Furthermore, the diffusional isolation of each synapse is controlled by a local signaling loop, such that when a synapse and the postsynaptic cell are consistently coactive, the neck of the spine that houses the active synapse is rapidly constricted. Since these patterns of activity are known to induce strengthening of hippocampal synapses [reviewed in (18)], we propose that adjusting the properties of the neck may titrate the threshold for plasticity induction by determining the lifetimes of plasticity-inducing molecules near active synapses.

The mechanisms described above for the biochemical isolation of synapses are not universal because not all neurons have dendritic spines. For example, the majority of interneurons lack spines and receive excitatory input directly onto the dendritic shaft (19, 20). Does this absence of obvious physical barriers to diffusion mean that aspiny neurons are unable to independently regulate each synapse?

To address this question, we examined the diffusional isolation and plasticity of excitatory synapses made onto cerebellar stellate cells (SCs), a compact aspiny cell class that receives excitatory input from parallel fibers (PFs) and makes inhibitory synapses onto local Purkinje cells. We uncovered a novel form of long-term synaptic depression (LTD) at the PF-to-SC synapse that specifically regulates active synapses without affecting neighboring inactive synapses (12). LTD was induced by prolonged trains of synaptic activity and required signaling through at least two diffusible messengers, calcium and endocannabinoids. The ineluctable deduction from these findings is that, despite the absence of spines, SC dendrites possess specializations that prevent the spread of plasticity-inducing molecules from active to inactive synapses. Indeed, the calcium signals responsible for LTD induction are highly compartmentalized and the accumulation of calcium is confined to a ~1 m segment of SC dendrite surrounding the active synapse. This compartmentalization results from two mechanisms. First, buffering of calcium by the calcium-binding protein parvalbumin specifically restricts the mobility of this signaling ion. Second, the movements of all molecules in SC dendrites are slowed by roughly a factor of 10, reflecting a general retardation of small-molecule mobility that may result from microscopic and nonspecific diffusion barriers such as high tortuosity or molecular crowding within the dendrite (21, 22).

In summary, our results revealed a diverse and rich set of mechanisms by which neurons maintain the diffusional isolation of individual synapses. The regulation of these mechanisms by activity may provide a local mechanism by which individual synapses can dynamically adjust the biochemical and electrical consequences of their stimulation (23, 24).

References

1. K. M. Harris, J. K. Stevens, J. Neurosci. 9, 2982 (1989).
2. M. Matsuzaki, N. Honkura, G. C. Ellis-Davies, H. Kasai, Nature 429, 761 (2004).
3. S. N. Yang, Y. G. Tang, R. S. Zucker, J. Neurophysiol. 81, 781 (1999).
4. R. C. Malenka, M. F. Bear, Neuron 44, 5 (2004).
5. R. Malinow, H. Schulman, R. W. Tsien, Science 245, 862 (1989).
6. A. Sobczyk, V. Scheuss, K. Svoboda, J. Neurosci. 25, 6037 (2005).
7. B. L. Sabatini, T. G. Oertner, K. Svoboda, Neuron 33, 439 (2002).
8. K. Svoboda, D. W. Tank, W. Denk, Science 272, 716 (1996).
9. A. G. Carter, B. L. Sabatini, Neuron 44, 483 (2004).
10. B. L. Bloodgood, B. L. Sabatini, Science 310, 866 (2005).
11. T. J. Ngo-Anh et al., Nat. Neurosci. 8, 642 (2005).
12. G. J. Soler-Llavina, B. L. Sabatini, Nat. Neurosci. 9, 798 (2006).
13. R. Yuste, W. Denk, Nature 375, 682 (1995).
14. W. Denk, R. Yuste, K. Svoboda, D. W. Tank, Curr. Opin. Neurobiol. 6, 372 (1996).
15. M. Matsuzaki et al., Nat. Neurosci. 4, 1086 (2001).
16. E. A. Nimchinsky, B. L. Sabatini, K. Svoboda, Annu. Rev. Physiol. 64, 313 (2002).
17. G. H. Patterson, J. Lippincott-Schwartz, Science 297, 1873 (2002).
18. G. Q. Bi, J. Rubin, Trends Neurosci. 28, 222 (2005).
19. S. L. Palay, V. Chan-Palay, Cerebellar Cortex: Cytology and Organization (Springer-Verlag, New York, 1974), pp. 216-233.
20. T. F. Freund, G. Buzsaki, Hippocampus 6, 347 (1996).
21. M. J. Kushmerick, R. J. Podolsky, Science 166, 1297 (1969).
22. R. J. Ellis, Trends Biochem. Sci. 26, 597 (2001).
23. C. Koch, A. Zador, J. Neurosci. 13, 413 (1993).
24. I. Segev, W. Rall, J. Neurophysiol. 60, 499 (1988).

To Advertise   |   Find Products