A Transporter That Removes D-Serine from Synapses
Oded Bodner, Inna Radzishevsky, Veronika N. Foltyn, Ayelet Touitou, Alec C. Valenta, et al.
(see pages 6489–6502)
NMDA receptors (NMDARs) are essential for many types of synaptic plasticity, and thus for learning and cognition. Activation of NMDARs requires not only glutamate binding and postsynaptic depolarization, but also binding of a coagonist. This coagonist can be glycine or D-serine, but one or the other plays a more prominent role depending on the brain area, NMDAR subunit composition, and other factors. D-serine appears to play a predominant role in hippocampal areas CA1 and prefrontal cortex.
D-serine is produced in the brain primarily by neurons. Astrocytes produce and secrete L-serine, which is imported by neurons and converted to D-serine by serine racemase. D-serine is then released by neurons via the amino acid antiporter Asc-1. How D-serine is removed from the synapse has been unclear, however. Notably, blocking or knocking out Asc-1 or sodium-dependent alanine-serine-cysteine transporters does not increase extracellular D-serine levels, suggesting they are not essential for uptake. Bodner et al. therefore hypothesized that a different amino acid transporter is required.
Amino acid transporters typically have several substrates that compete for transport. Therefore, the authors applied different amino acids to mouse synaptosomes and asked whether any reduced D-serine uptake. Glutamine reduced uptake of both L-serine and D-serine, not only by synaptosomes, but also in cortical slices and in the striatum in vivo. Moreover, glutamine increased the NMDAR component of field EPSPs and enhanced LTP in the hippocampus. Importantly, these effects were not seen in mice lacking serine racemase.
These experiments suggested that a glutamine transporter is responsible for serine uptake in the brain. Indeed, the so-called system A transporters Slc38a1 and Slc38a2, which mediate most glutamate uptake in the brain, also mediated uptake of D-serine and L-serine when expressed in HEK cells. This uptake was reduced in the presence of excess glutamine or MeAIB, another substrate of Slc38a1 and Slc38a2.
These and additional control experiments suggest that Slc38a1 and Slc38a2 mediate the uptake of glutamine, L-serine, and D-serine in the brain. Elevating extracellular glutamine decreases D-serine uptake, increasing extracellular D-serine levels and thus increasing its binding to NMDARs. Future work should determine whether similar modulation of D-serine uptake, for example by increased glutamine release from astrocytes, regulates NMDAR activity under physiological conditions.
At P21, basal dendrites are somewhat shorter in visual cortical pyramidal cells from dark-reared mice (green) than in those from mice raised under normal light conditions (blue). See Richards et al. for details.
Postnatal Growth of Basal Dendrites in Visual Cortex
Sarah E. V. Richards, Anna R. Moore, Alice Y. Nam, Shikhar Saxena, Suzanne Paradis, et al.
(see pages 6536–6556)
The size and shape of dendritic arbors determine the type of input neurons receive and how that input is integrated. Although gross dendritic morphology depends largely on cell type-specific genetic programs, morphological details such as branch location, and segment length are influenced by extracellular molecules and synaptic input, which regulate cytoskeletal dynamics via intracellular signaling cascades.
In Xenopus tadpoles, visual experience shapes dendritic arbors in the optic tectum. Richards et al. asked whether experience similarly shapes dendritic trees in visual cortex. To answer this, they measured basal dendrites of layer 2/3 pyramidal cells in mouse visual cortex at several points, starting before eye opening [postnatal day 7 (P7)] and continuing until the peak of the critical period (P30). They compared growth in mice housed under normal lighting conditions and mice reared in darkness, and compared wild-type mice with mice lacking Rem2, a small GTPase implicated in dendrite growth.
In normally reared wild-type mice, basal dendrites grew little between P7 and the day of eye opening (P12). Arbor length increased after that, especially in the days before the critical period opened (P21), primarily through increases in the length of individual segments, rather than branch formation. Somewhat surprisingly, dendritic arbors grew little after the onset of the critical period. Even more surprisingly, dark rearing had relatively mild effects on dendritic growth. Dendritic segments still grew between eye opening and critical period onset in dark-reared mice, but the arbors were ∼15% smaller than in wild type.
Constitutive knockout of Rem2 reduced the number of dendritic branches at P7, but this difference disappeared by eye opening, and Rem2-deficient dendrites had significantly more branches than wild type at P21. Notably, this increase in branching did not occur in dark-reared Rem2-deficient mice. Finally, sparse knockout of Rem2 starting at P21 caused transient changes in branch number, segment length, and arbor orientation.
These data suggest that visual experience has a relatively small effect on basal dendritic arbors of pyramidal cells in mouse visual cortex. This may be because cortical neurons receive more input from other cortical cells than from the periphery. The results also indicate that mature dendritic arbors must be maintained by factors like Rem3. Future work should explore the role of activity in this maintenance.
Footnotes
This Week in The Journal was written by Teresa Esch, Ph.D.