Cell biology

Glutamate primes the pump

Article metrics

Glutamate is the most widely distributed excitatory neurotransmitter in the central nervous system. It is stored within nerve terminals, in small vesicles that are released when the nerve cell is stimulated. Once released, glutamate diffuses to neighbouring nerve cells where it binds to glutamate receptors, resulting in an influx of sodium and calcium ions (ionotropic receptors) or mobilization of intracellular calcium (metabotropic receptors). On page 685 of this issue, Maechler and Wollheim1 report that, as well as having an extracellular function in neurotransmission, glutamate acts as an intracellular signal. They show that in insulin-secreting pancreatic β-cells, glutamate is involved in preparing the insulin-filled secretory granules for release — a process commonly referred to as ‘priming’.

Glucose stimulates the pancreatic β-cells to secrete insulin after an increase in the intracellular concentration of calcium. We have known for over 30 years2 that glucose stimulation elicits a biphasic secretory response consisting of an initial transient phase and a second, sustained component (Fig. 1). The most common form of human diabetes is associated with abnormalities in this release pattern, so we need to establish the underlying cellular mechanisms.

Figure 1: Biphasic stimulation of insulin secretion in response to glucose.

The response consists of an initial transient phase in which the readily releasable pool of secretory granules is depleted, followed by a second, slower phase in which this pool is replenished by granule ‘priming’.

The two phases of insulin secretion are thought to reflect the sequential release of two functional subsets of secretory granules3. In this model, the rapid release and depletion of a readily releasable pool of granules (RRP), containing only about 50 of the cell's 13,000 or so secretory granules, accounts for the first phase. The second, slower phase reflects the time-dependent replenishment of this pool by priming of granules originally in the reserve pool. Priming proceeds at a rate of 5–10 granules per minute per β-cell, and probably involves a product of cell metabolism because this phase is evoked only by stimuli that can be metabolized (such as glucose), and not by agents that merely increase the levels of intracellular calcium3. It has not been possible to establish the identity of this factor, but Maechler and Wollheim1 now propose that glutamate, generated by amination of the Krebs-cycle intermediate α-ketoglutarate, is the elusive metabolic signal.

How might glutamate promote granule priming? The release of neurotransmitters and many hormones (such as insulin) involves fusion of synaptic vesicles or secretory granules, respectively, with the plasma membrane (exocytosis). This requires an interaction between various fusion proteins collectively referred to as SNAREs (‘SNAP receptors’, where SNAP stands for soluble NSF-attachment protein)4,5. According to the SNARE model, exocytosis is preceded by ATP-dependent priming of secretory granules by the N-ethylmaleimide-sensitive fusion protein (NSF)4. Yet although insulin secretion depends on ATP, perturbation of NSF activity has little effect on release6.

This discrepancy can now be explained by Maechler and Wollheim's findings, which indicate that ATP hydrolysis by a granular proton pump, rather than by NSF, is more prominent in granule priming. The electrochemical proton gradient that results from the activity of this proton pump can be envisaged as driving negatively charged glutamate into the secretory granules (although we still do not know how uptake of glutamate facilitates release). The identity of the glutamate-uptake mechanism in the β-cell is also not known, but pharmacological studies indicate that it is related to the mechanism documented in synaptic vesicles7. Maechler and Wollheim propose that glutamate acts by reducing the granular membrane potential, in which case it would act as a negative counter-ion, allowing a larger pH gradient to develop across the granular membrane (Fig. 2). In the absence of such a counter-ion, the large granular membrane potential that developed on proton pumping (positive inside) would quickly prevent further uptake of H+ and, thus, acidification.

Figure 2: Model to account for the effects of glutamate on granule priming, based, in part, on the results of Maechler and Wollheim1.

The granular membrane contains v-SNAREs (black), whereas the plasma membrane contains t-SNAREs (red). The two membranes come together, and granule priming requires intragranular acidification. Glutamate uptake reduces the granular membrane potential, facilitating the pumping of H + ions into the granule.

It seems wasteful of the β-cell to use glutamate in a process where it could easily be replaced by chloride or any other cytosolic anion. One possibility is that, once priming has been completed, glutamate also has an extracellular function similar to that in the central nervous system. Interestingly, ionotropic glutamate receptors have been documented in all types of endocrine cell in the pancreatic islet8. This points to the exciting possibility that glutamate coordinates the complex interplay between the different types of islet cell and their secretory products.

With the paper by Maechler and Wollheim, researchers should finally be able to get a firm grip on biphasic insulin secretion and its metabolic regulation. Secretion involves similar mechanisms in different cell types. So the functions that the authors have documented in the β-cell — namely, the role of glutamate and a low intragranular pH in the priming of secretory granules — may have counterparts in other secretory cells. Bafilomycin (an inhibitor of the vesicular H+-ATPase) has, in fact, been reported9 to abolish a late phase of glutamate release in neurons. In this context, it may be significant that vacuole acidification is required for the pairing of the SNARE proteins in yeast10. If this also applies to the SNAREs involved in the fusion between the granules and plasma membrane, then it is easy to see how granule acidification promotes secretion. The traditional view that a low intragranular pH is required for hormone processing and uptake may represent only one side of the coin. And as the coin is now tossed, we may discover that intragranular acidification is also essential for exocytosis itself.


  1. 1

    Maechler, P. & Wollheim, C. B. Nature 402, 685–689 (1999).

  2. 2

    Curry, D. L., Bennett, L. L. & Grodsky, G. M. Endocrinology 83, 572–584 (1968).

  3. 3

    Eliasson, L. et al. J. Physiol. 503, 399–412 (1997).

  4. 4

    Hanson, P. I., Heuser, J. E. & Jahn, R. Curr. Opin. Neurobiol. 7, 310–315 (1997).

  5. 5

    Lang, J. Eur. J. Biochem. 259, 3–17 (1999).

  6. 6

    Kiraly-Borri, C. et al. Biochem. J. 314, 199–203 (1996).

  7. 7

    Özkam, E. D. & Ueda, T. Jpn. J. Pharmacol. 77, 1–10 (1998).

  8. 8

    Weaver, C. D., Yao, T. L., Powers, A. C. & Verdoorn, T. A. J. Biol. Chem. 271, 12977–12984 (1996).

  9. 9

    Zoccarato, F., Cavallini, L. & Alexandre, A. J. Neurochem. 72, 625–633 (1999).

  10. 10

    Ungerman, C., Wickner, W. & Xu, Z. Proc. Natl Acad. Sci. USA 96, 11194–11199 (1999).

Download references

Author information

Correspondence to Patrik Rorsman.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Rorsman, P., Renström, E. Glutamate primes the pump. Nature 402, 595–596 (1999) doi:10.1038/45110

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.