This Week in The Journal

Cellular/Molecular

PSD-95 Extends Perpendicularly from the Postsynaptic Membrane

Xiaobing Chen, Christopher D. Nelson, Xiang Li, Christine A. Winters, Rita Azzam, et al.

Postsynaptic densities (PSDs) comprise orthogonal horizontal and vertical filaments that serve as scaffolds for clustering glutamate receptors and signaling molecules. Scaffold proteins include membrane-associated guanylate kinases (MAGUKs), the most abundant of which is PSD-95. By labeling the ends of PSD-95 molecules and measuring their distance from the postsynaptic membrane, Chen et al. found that the protein is nearly fully extended in the PSD, forming vertical filaments with N termini toward the membrane. Knockdown of PSD-95 resulted in loss of vertical filaments, along with associated horizontal filaments and putative AMPA receptors, especially at the PSD periphery. Nonetheless, numerous vertical filaments remained in the center of the PSD. Likewise, putative NMDA receptors, which are primarily found in the PSD center, were minimally affected by PSD-95 knockdown. These data suggest that PSD-95 is more stable at the center of the postsynaptic density and/or that other proteins—possibly other MAGUKs, which remained after PSD-95 knockdown—form central vertical filaments.

Rendering of EM tomogram of a portion of dendritic spine after PSD-95 knockdown, showing patchy loss of vertical filaments (red) and cytoplasmic sides of putative NMDA receptors (cyan) and AMPA receptors (dark blue). See the article by Chen et al. for details.

Development/Plasticity/Repair

Reprogrammed Neural Crest Cells Produce Myelin In Vivo

Ellen Binder, Marion Rukavina, Hessameh Hassani, Marlen Weber, Hiroko Nakatani, et al.

(see pages 6379–6391)

As development proceeds, cells become progressively restricted to specific fates. Even pluripotent stem cells, which can self-renew and generate multiple cell types, typically produce limited cell types in situ. For example, stem cells derived from the neural crest (NCSCs) produce melanocytes, PNS neurons, and glia, but NCSCs residing in different areas differ in their ability to produce each cell type, and they normally do not produce CNS cells. Nonetheless, because a progenitor's potential is determined by its molecular environment, maintaining self-renewing progenitors in culture for many generations can reprogram them to produce different cell types. Binder et al. show that when NCSCs isolated from mouse dorsal root ganglia were maintained in media containing certain growth factors, they stopped expressing neural crest and PNS genes and began to express CNS markers. When these reprogrammed progenitors were grafted into mouse brain, they survived for months and primarily differentiated into oligodendrocytes that could myelinate axons.

Behavioral/Systems/Cognitive

Training Reduces Selectivity of PFC Neurons

Travis Meyer, Xue-Lian Qi, Terrence R. Stanford, and Christos Constantinidis

(see pages 6266–6276)

The prefrontal cortex (PFC) maintains representations of transient stimuli for several seconds, which is thought to contribute to working memory. Dorsal PFC receives visual information largely from areas involved in motion and spatial processing, whereas ventral PFC receives information from areas that encode color and shape, suggesting that these regions represent spatial and feature information, respectively. Although some electrophysiological studies support this hypothesis, others suggest that PFC neurons respond to different types of information depending on task requirements. Meyer et al. recorded PFC neurons before and after monkeys learned spatial and shape memory tasks. Before training, most visually responsive dorsal neurons were differentially activated by stimuli presented in different spatial locations, but some had differential responses to shapes. Similarly, ventral neurons showed differential responses to locations and/or shapes. After training, the proportion of visually responsive neurons increased in both areas, but both shape and location selectivity decreased. These results counter predictions of both hypotheses, indicating neither fully explains PFC activity.

Neurobiology of Disease

Myelin Lipid Sulfatide Inhibits Axon Growth

Alissa M. Winzeler, Wim J. Mandemakers, Matthew Z. Sun, Melissa Stafford, Carolyn T. Phillips, et al.

(see pages 6481–6492)

Multiple mechanisms have evolved to limit axon growth in the mature mammalian CNS, including both a loss of axons' ability to grow and the presence of multiple inhibitory molecules in myelin. Although this suggests continued growth is detrimental in mature animals, it confounds attempts to promote regeneration after injury. Even attempts to target multiple inhibitory proteins while simultaneously providing growth-promoting molecules have had limited success in restoring functional connections. Winzeler et al. have now identified yet another class of myelin-associated molecule that inhibits axonal growth: the prominent membrane lipid sulfatide. Purified sulfatide inhibited growth of cultured rodent retinal ganglion cell (RGC) axons, whereas myelin treated with anti-sulfatide antibodies or taken from sulfatide-deficient animals was less inhibitory than wild-type myelin in vitro. Like knocking out individual inhibitory proteins, however, knocking out sulfatide or its precursor did not by itself improve RGC regeneration in vivo.