S C I E N C E ’ S C O M PA S S PERSPECTIVES

65 ter at postsynaptic sites. This suggested

64 P E R S P E C T I V E S : N E U RO B I O L O G Y that the ephrinB1 present in the neuronal
63 membranes of axon terminals might induce
62
61 Learning More About NMDA the maturation of glutamatergic synapses
by promoting aggregation of NMDA re-
60 ceptors in postsynaptic membranes.
59
58
Receptor Regulation Now, Takasu and colleagues (1) extend
these observations by revealing that activa-
57 Anirvan Ghosh tion of the EphB receptor and its signaling
56 pathway greatly potentiates glutamate-in-
55 f all the neurotransmitter receptors Eph receptors, EphA and EphB, which are duced calcium influx through the NMDA re-
54
53
52
O in the brain, the N-methyl-D-aspar- selectively activated by ephrinA and ephrinB ceptor. In addition, they show that NMDA
tate (NMDA) subtype of glutamate ligands, respectively. EphrinA ligands are at- receptor–induced phosphorylation of the
receptor has an unmatched hold on the tached to the membrane through a glycosyl- transcription factor CREB and expression
51 imagination of neuroscientists. The secret of phosphatidylinositol linkage, and ephrinB of its target genes are also markedly en-
50 the NMDA receptor’s enduring appeal is its ligands are transmembrane proteins. The hanced by EphB2 receptor activation and
49 crucial involvement in regulating changes in binding of ephrinB to the EphB receptor is signaling. Because NMDA receptor–in-
48 the strength of synapses, the regions where particularly interesting because it leads to ty- duced calcium influx and altered gene ex-
47 neurons communicate. Such changes in rosine phosphorylation not only of the recep- pression are crucial for inducing long-term
46 synaptic strength (synaptic plasticity) are tor, but also of the cytoplasmic domain of changes in synaptic strength, regulation of

Downloaded from www.sciencemag.org on December 13, 2009

45 believed to underlie learning the ephrinB ligand itself. NMDA receptor activity by Eph receptor
44 and memory. The NMDA re- Axon Bidirectional signaling signaling may influence synaptic plasticity.
43 ceptor is a multimeric protein terminal through ephrinB ligands Takasu and colleagues also investigated
42 complex localized in the and the EphB2 receptor is how EphB receptors might modulate NMDA
41 membranes of postsynaptic receptor activity. Stimulating cortical neurons
40 neurons. It consists of an with ephrinB2 resulted in tyrosine phospho-
39 NR1 subunit and one or more rylation of the NR2B subunit at three tyrosine
Glutamate
38 NR2 subunits, which form a residues. They report that
37 channel that is permeable to ephrinB2-induced potenti-
36 calcium ions. The defining NMDA
ation of calcium influx
35 feature of the NMDA recep- receptor through NMDA receptors,
34 tor is that it allows calcium Ca2+ as well as tyrosine phos-
33 ions to flow into the postsyn- phorylation of NR2B, re-
EphB Src
32 aptic neuron when the neuro- quires the cytoplasmic
EphrinB
31 transmitter glutamate is re- (kinase) domain of the
30 leased into the synapse. An EphB2 receptor. Tyrosine
29 increase in the calcium ion i t e Ca2+ phosphorylation of NR2B
28 concentration of the postsyn- ndr P
appears to be necessary
De
27 apatic neuron triggers a series Src for ephrinB2 to modulate
EphB
26 of biochemical changes that result in modu- NMDA receptor activity,
25 lation of synaptic strength. Despite the fact because a mutant form of
24 that the NMDA receptor is a central player NR2B in which the three
23 in synaptic plasticity, surprisingly little is tyrosine residues cannot be
22 known about the way in which NMDA re- phosphorylated prevents
21 ceptor–mediated calcium influx is regulat- Two receptors are better than one. Activation of the EphB receptor calcium influx in response
20 ed. Reporting on page 491 of this week’s and its signal transduction pathway regulates NMDA receptor activity. to EphB signaling.
19 Science (1) and in a recent issue of Neuron When activated by binding of its ligand, ephrinB2, the EphB receptor It appears that the ef-
18 (2, 3), the Greenberg, Klein, and Pawson recruits a Src family kinase, which phosphorylates three of the ty- fects of EphB2 signaling
17 groups shed light on how this NMDA re- rosines in a subunit of the NMDA receptor. This leads to increased cal- on NMDA receptor activi-
16 ceptor–mediated postsynaptic calcium in- cium permeability of the NMDA receptor, the influx of calcium ions ty are mediated by a Src
15 flux is regulated. Using developing cortical into the postsynaptic neuron, and alterations in synaptic plasticity. family tyrosine kinase. A
14 neurons, Takasu et al. (1) show that regula- Src tyrosine kinase called
13 tion depends on the Eph receptor tyrosine important for regulating interactions be- Fyn is known to phosphorylate the NR2B
12 kinase. Meanwhile, the Neuron papers (2, 3) tween axons and their cellular targets in the subunit of the NMDA receptor (7). Takasu et
11 provide in vivo evidence for the involve- mammalian embryo (4). al. show that ephrinB2 stimulation of corti-
10 ment of Eph receptors in synaptic plasticity. The possibility that ephrinB-EphB2 re- cal neurons leads to phosphorylation of Src
9 Eph receptors are a large family of recep- ceptor interactions might regulate synaptic and its association with the EphB2 receptor
8 tor tyrosine kinases that regulate various de- strength was first suggested by the obser- (see the figure). Also, a dominant negative
7 velopmental events including cell migration, vation that EphB receptors are found on Src construct (which inhibits multiple Src
6 axon guidance, and regionalization of the postsynaptic membranes (5). Greenberg’s family kinases) suppresses ephrinB2-in-
5 nervous system (4). There are two classes of group has shown that ephrinB1 induces the duced tyrosine phosphorylation of NR2B
4 association of EphB2 receptors with the and inhibits the potentiation of NMDA re-
3 NR1 subunit of NMDA receptors (6). ceptor–mediated calcium influx. Grunwald
The author is in the Department of Neuroscience,
2 Johns Hopkins University School of Medicine, Balti- These investigators have also demonstrated et al. (2) also provide evidence for Src acti-
1 more, MD 21205, USA. E-mail: aghosh@jhmi.edu that ephrinB1 causes NR1 subunits to clus- vation and tyrosine phosphorylation of a dif-

www.sciencemag.org SCIENCE VOL 295 18 JANUARY 2002 449

 S C I E N C E ’ S C O M PA S S
65 ferent NMDA receptor subunit (NR2A) by though this is an important issue that needs One of the most interesting implications
64 EphB2 signaling (2). These observations to be resolved, the studies agree on the prin- of the Greenberg study (1) is that EphB2
63 suggest that ephrin ligands activate EphB2 cipal conclusion that EphB2 receptors con- receptor signaling can acutely influence
62 receptors on the postsynaptic membrane, tribute to hippocampal synaptic plasticity. NMDA receptor activity. It will be impor-
61 which leads to recruitment of a Src family Together the studies from the Green- tant to further examine this possibility in
60 kinase to the EphB2 receptor. The Src family berg, Klein, and Pawson laboratories (1–3) vivo, both in the context of developmental
59 kinase, in turn, phosphorylates one or more strongly suggest that EphB2 receptors can maturation and adult synaptic plasticity. It
58 NR2 subunits, and this posttranslational modulate NMDA receptor–mediated should be determined whether NMDA re-
57 modification leads to increased calcium per- synaptic plasticity, but several mechanistic ceptor–mediated calcium influx is reduced
56 meability of the NMDA receptor (see the details are not yet resolved. Perhaps the in ephB2-deficient mice, and whether the
55 figure). In this way, ephrinB signaling could most surprising finding is that of the Klein effects of EphB2 signaling on NMDA re-
54 exert a marked influence on synaptic plastic- and Pawson groups (2, 3), who report that ceptor clustering and calcium permeability
53 ity at glutamatergic synapses. the in vivo deficits in ephB2-deficient mice can be mechanistically separated by genet-
52 The Grunwald et al. (2) and Henderson can be rescued by overexpressing an EphB2 ic manipulation. It would also be interest-
51 et al. (3) papers provide in vivo evidence receptor that lacks the cytoplasmic domain. ing to know whether the effects of EphB
50 that the EphB2 receptors are involved in At first glance, this appears to be at odds signaling on the NMDA receptor are partly
49 synaptic plasticity. Both groups report that with the observation from the Greenberg regulated by insertion of NMDA receptors
48 synaptic plasticity is compromised in mice group that EphB receptor–induced potenti- into the postsynaptic membrane, which is
47 that lack the ephB2 gene, although the stud- ation of NMDA receptor activity requires one way in which glutamate receptor activ-
46 ies differ in some important respects. Hen- the cytoplasmic domain of the EphB recep- ity is modulated. Addressing these ques-

Downloaded from www.sciencemag.org on December 13, 2009

45 derson et al. (3) report that synaptic plastici- tor. But the Greenberg lab had previously tions should provide important insights in-
44 ty in the hippocampus is reduced at both shown that EphB2 receptors induce NMDA to the link between EphB receptor signal-
43 CA1 and dentate gyrus synapses in ephB2- receptor clustering in neurons, and that this ing and NMDA receptor activity, simulta-
42 deficient mice. They also found that NMDA effect does not require the EphB receptor’s neously enhancing our understanding of
41 receptor–mediated synaptic currents in hip- cytoplasmic domain (6). Therefore, one the molecular basis of synaptic plasticity.
40 pocampal granule cells are reduced in the possible explanation is that the in vivo
39 mutant mice. Grunwald et al. (2) also re- deficits in synaptic plasticity are principally References
38 port a reduction in synaptic plasticity at due to lack of proper clustering of NMDA 1. M. A. Takasu et al., Science 295, 491 (2002); published
online 20 December 2001 (10.1126/science.1065983).
37 CA1 synapses, but they did not observe dif- receptors at the synapse. This further under- 2. I. C. Grunwald et al., Neuron 32, 1027 (2001).
36 ferences in NMDA receptor–mediated cur- scores the importance of determining 3. J. T. Henderson et al., Neuron 32, 1041 (2001).
35 rents in CA1 neurons. Thus, it is not yet whether the localization and function of 4. J. G. Flanagan, P. Vanderhaegen, Annu. Rev. Neurosci.
34 21, 309 (1998).
clear whether altered plasticity in ephB2- NMDA receptors are affected in ephB2-de- 5. R. Torres, Neuron 21, 1453 (1998).
33 deficient mice can be entirely explained by ficient mice, and whether that can explain 6. M. B. Dalva et al., Cell 103, 945 (2000).
32 changes in NMDA receptor activity. Al- the observed effects on synaptic plasticity. 7. T. Nakazawa et al., J. Biol. Chem. 276, 693 (2001).
31
30 P E R S P E C T I V E S : C L I M AT E C H A N G E
29 surface and bed velocities are similar, al-
28 lowing measurements of surface velocities
27
26
On Thickening Ice? and ice thicknesses to constrain ice outflow.
Early, often heroic efforts to measure
25 Richard B. Alley mass balance produced important baseline
24 data but left considerable uncertainties be-
23 he big ice sheets in Greenland and stability. The West Antarctic Ice Sheet has cause sampling was too sparse to capture the
22
21
20
T Antarctica are key elements of the
global climate system. By storing
large volumes of water as ice or ice-con-
changed greatly since it first formed a few
million years ago (3) and has been far from
static since humans began observing it a
spatial variability. Improvements in ice-core
analyses, airborne geophysical surveying,
and satellite remote sensing are rapidly re-
19 tact lakes and sometimes releasing that few decades ago (1). Yet in the modern ducing these uncertainties and form the basis
water abruptly, they warm period (interglacial), it has long out- of the new work by Joughin and Tulaczyk
Enhanced online at can affect sea level, lasted the melting of most ice-age ice, and (1). Focusing on the West Antarctic drainage
www.sciencemag.org/cgi/ global ocean circu- circumstantial evidence indicates that the into the Ross Sea, the authors show that on
content/full/295/5554/451 lation, and hence ice sheet persisted through the previous in- average the ice sheet is thickening slowly.
Earth’s climate, as terglacial (4) and probably the two inter- This new result differs from the best
18 highlighted on page 476 of this issue by glacials before that (3). older estimates, which indicated a net thin-
17 Joughin and Tulaczyk (1). Predicting the future of the West Antarc- ning for this region (5). Improved data
16 Modern attention is especially focused tic Ice Sheet bears many challenges. Even from interferometric synthetic-aperture
15 on the West Antarctic Ice Sheet (2). Its bed just measuring the mass balance—whether radar and other techniques contributed to
14 is well below sea level and deepens toward the ice sheet is growing or shrinking—has the difference. However, the discharge
13 the center. In some models and in recon- proved difficult. One approach is to com- from this region also has decreased sub-
12 structions of the behavior of some past ice pare the snow input with the flow output. stantially over the last decades as Whillans
11 sheets, these characteristics are linked to in- This requires enough ice-core or other data Ice Stream (formerly called Ice Stream B)
10 to determine accumulation rates and slowed near the Ross Ice Shelf (6). Consid-
9 enough velocity measurements to capture ering the century-old near stoppage of ad-
The author is in the EMS Environment Institute and
8 the Department of Geosciences, Pennsylvania State
the ice outflow. Much West Antarctic ice jacent Ice Stream C (see the figure) (5), it
7 University, University Park, PA 16802, USA. E-mail: discharges through ice streams with slip- is tempting to identify a trend. Perhaps af-
6 ralley@essc.psu.edu pery beds. This simplifies the problem as ter 10,000 years of retreat from the ice-age

www.sciencemag.org SCIENCE VOL 295 18 JANUARY 2002 451