Synaptic plasticity is a contest between long-term potentiation (LTP) and long-term depression (LTD), with the former strengthening synaptic connections and the latter weakening them. Since activation of NMDA-type glutamate receptors induces both LTP and LTD, one might think this tug-of-war is destined to end in a tie. So what determines which has more pull? A paper in the March 1 Neuron shows that, surprisingly, the enzyme glycogen synthase kinase-3β (GSK3β) appears to sway the balance. Graham Collingridge and colleagues suggest that the kinase plays anchor man on the LTD side, but apparently not very effectively. After induction of LTP, the kinase becomes temporarily inactivated, support for LTD is lost, and LTP comes out on top. The finding is relevant for Alzheimer disease (AD) drug discovery. It implies that inhibiting GSK3β might boost LTP and depress LTD—good, in a first approximation, for learning and memory. That would be a new benefit on top of the better-known one of reducing phosphorylation of the GSK3β substrates tau and presenilin-1, which would presumably prevent toxicities associated with neurofibrillary tangles and amyloid-β. Add to that a report in the February 21 Journal of Neuroscience from Eliezer Masliah and colleagues showing that inactivating GSK3β reduces phosphorylation and processing of amyloid precursor protein (APP), as well, and the kinase begins to look like an even better therapeutic target.

To test the role of GSK3β in synaptic plasticity, Collingridge, from the University of Bristol, U.K., and colleagues there and at Canada’s University of British Columbia, Vancouver, and the NIH in Bethesda, Maryland, inhibited the kinase in rat hippocampal slices. Joint first authors Stéphane Peineau and Changiz Taghibiglou found that GSK3β inhibition had no effect on LTP induced by high-frequency stimulation, but completely abolished LTD induced by low-frequency stimulation. The researchers obtained the same result with several GSK3β inhibitors, including the highly selective SB415286, lithium, and kenpaullone, but not with roscovitine, which inhibits the related cyclin-dependent protein kinases. The data indicate that LTD is specifically sensitive to GSK3β inhibition. In support of this, the authors found that during LTD, the kinase is activated by dephosphorylation at serine number 9 (ser9), a post-translational modification. Biochemical assays of GSK3β activity also showed a post-LTD increase of about 30 percent.

It was when the authors tested the duration of GSK3β activation that they found LTP has the opposite effect on the kinase. While ser9 GSK3β phosphorylation declined by about 30 percent 20 minutes after inducing LTD, 20 minutes after induction of LTP, phosphorylation of that residue increased by about 50-60 percent. This observation suggested that LTP might actually inhibit LTD, and when the scientists tested this idea that’s exactly what they found. LTP completely prevented subsequent induction of LTD for as long as 1 hour. One possible role for such regulation might be to “stabilize a synaptic modification over the short term by protecting synapses from the effects of additional NMDA receptor-dependent plasticity until the information can be either consolidated or erased by NMDA receptor-independent mechanisms,” write the authors.

GSK3β has many substrates, and it is unclear how its action on LTD and LTP is regulated. To approach this question, Collingridge and colleagues tested if this mostly cytosolic kinase might be present in synapses. Using immunohistochemistry, they found that it not only occurs on dendritic spines, but also associates with the AMPA receptor subunits GluR1 and GluR2 there. Moreover, NMDA receptor-triggered insertion of AMPA receptors into the plasma membrane appears to be accompanied by a decrease in AMPA-linked GSK3β activity. All told, the work points to GSK3β having some role in the installation or maintenance of AMPA receptors, a process in which prior studies have implicated Aβ and that is drawing intense research attention.

The work by Masliah’s group highlights another relevant substrate of GSK3β—amyloid-β precursor protein (APP). First author Edward Rockenstein and colleagues at the University of California at San Diego discovered the GSK3β/APP connection while testing if inhibition of GSK3β is neuroprotective. GSK3β is known to phosphorylate tau and presenilin-1. Inhibiting the kinase can reduce Aβ production, plaque load (see ARF related news story), and tau phosphorylation in various mouse models of AD (see Nakashima et al., 2005 and Noble et al., 2005). But Rockenstein and colleagues wanted to see if this inhibition has any effect on learning and memory or on other disease markers. The researchers inhibited GSK3β in APP transgenic mice (with London and Swedish mutations—hAPPtg, line 41) by administering the drug lithium or by crossing the animals with GSK3β dominant-negative mice. In both cases, GSK3β activity was reduced and the animals performed on par with wild-type mice in a water maze test of learning and memory. Consistent with previous findings, the treated animals had reduced Aβ and phospho-tau immunoreactivity in the brain, but more unexpectedly, the amounts of full-length phosphorylated APP and phosphorylated APP C-terminal fragments were reduced in both cases, as well.

Li-Huei Tsai’s group at Harvard Medical School, Boston, has shown that APP phosphorylation is linked to maturation of the precursor protein (see Lee et al., 2003 and also da Cruz e Silva et al., 2003), suggesting that inhibiting the APP kinase might slow the production of Aβ. This work from Rockenstein and colleagues suggests that GSK3β may be the APP kinase to target. “These results support the notion that, via regulation of phosphorylation, blockage of GSK3β might retain APP in the early secretory pathway, depleting mature APP-p before it is transported to the axon terminals, in which a large proportion of Aβ generation by the γ-secretase complex takes place,” write the authors.—Tom Fagan


  1. My initial enthusiasm, shared with commentator Tom Fagan, for this study was cooled when I tried to discover which dominant-negative mutant of GSK3β was used. The answer appears to be "none" since wild-type GSK3β was incorporated in the injected DNA construct (see materials & methods section, page 1982). Hence, overexpression of wild-type GSK3β decreases GSK3 enzyme activity in brain....

    I noticed this rather "uncommon" outcome of a transgenic mouse while attending the AD-PD meeting in Salzburg (although time to read was scarce) and discussed the matter with several "knowledgeable" colleagues. We failed to come up with an acceptable explanation and were not satisfied with the one provided in the paper. Indeed, when GSK3 activity is downregulated by activated Akt and phosphorylation of S9, that would restore (or even lower more) Akt activity…and a futile cycle would result.

    We were therefore wondering if the transgenic construct is correct—i.e., an unnoticed mutation slipped in during PCR?—or whether the enzymatic assay provided the correct answers, since it is known to be a very tricky procedure. On the other hand, if indeed the GSK3β activity is down in these mice, the cause could also be due to genetic regulation and/or integration site effects, which will be more difficult to "un-tangle" (to maintain the jargon of this field)!

    View all comments by Fred Van Leuven

Make a Comment

To make a comment you must login or register.


News Citations

  1. Lithium Hinders Aβ Generation, Buffing Up GSK as Drug Target

Paper Citations

  1. . Chronic lithium treatment decreases tau lesions by promoting ubiquitination in a mouse model of tauopathies. Acta Neuropathol. 2005 Dec;110(6):547-56. Epub 2005 Oct 14 PubMed.
  2. . Inhibition of glycogen synthase kinase-3 by lithium correlates with reduced tauopathy and degeneration in vivo. Proc Natl Acad Sci U S A. 2005 May 10;102(19):6990-5. PubMed.
  3. . APP processing is regulated by cytoplasmic phosphorylation. J Cell Biol. 2003 Oct 13;163(1):83-95. PubMed.
  4. . Protein phosphorylation and APP metabolism. Neurochem Res. 2003 Oct;28(10):1553-61. PubMed.

Further Reading

Primary Papers

  1. . LTP inhibits LTD in the hippocampus via regulation of GSK3beta. Neuron. 2007 Mar 1;53(5):703-17. PubMed.
  2. . Neuroprotective effects of regulators of the glycogen synthase kinase-3beta signaling pathway in a transgenic model of Alzheimer's disease are associated with reduced amyloid precursor protein phosphorylation. J Neurosci. 2007 Feb 21;27(8):1981-91. PubMed.