This Week in The Journal

DLK Contribution to Corticospinal Axon Regeneration

Junmi M. Saikia, Carmine L. Chavez-Martinez, Noah D. Kim, Sahar Allibhoy, et al.

(see pages 3716–3732)

Nerve damage activates signaling pathways that trigger either regeneration or degeneration of injured axons and sprouting of spared axons. The mitogen-activated protein-kinase-kinase kinase DLK is a critical component of these signaling pathways. DLK associates with elements of the cytoskeleton and is activated when cytoskeletal integrity is disrupted. Its downstream effectors activate transcriptional programs necessary for launching appropriate responses. Surprisingly, loss of DLK can have opposite effects in different neuron types. For example, loss of DLK in peripheral nerves, which typically regenerate after injury, reduces regeneration, whereas loss of DLK in retinal ganglion cells, which normally degenerate after optic nerve injury, reduces degeneration. Therefore, Saikia et al. asked how loss of DLK and its homolog LZK affects regeneration of corticospinal tract axons after spinal cord hemisection in mice.

Because corticospinal axons show little regrowth after transection in wild-type mice, the authors knocked out PTEN to increase regeneration. Whereas corticospinal axons in control mice had retracted from the lesion site and showed little regeneration 6 weeks after hemisection, most PTEN-deficient axons had regrown after an initial retraction, and many extended beyond the lesion site. Knocking out LZK did not reduce regeneration of PTEN-deficient axons, but knocking out DLK blunted regeneration, and knocking out both DLK and LZK eliminated the beneficial effect of PTEN knockout.

Although wild-type corticospinal axons do not regenerate substantially after transection, some motor function can be restored by sprouting of spared corticospinal axons. Knocking out PTEN enhances this sprouting, and, as with regeneration, knocking out both DLK and LZK prevented the effect of PTEN knockout. In addition, DLK/LZK knockout reduced baseline sprouting of corticospinal axons with normal levels of PTEN expression. Knocking out either DLK or LZK by itself did not affect sprouting of axons with or without PTEN, however.

These results suggest that DLK contributes to axon regeneration and sprouting after spinal cord injury and that LZK can partially compensate for loss of DLK. Little is known about how DLK and LZK are activated after injury or what effectors regulated by these signaling pathways promote axon growth. Answering these questions may help identify new ways to promote functional recovery after CNS injury.

• Download figure
• Open in new tab
• Download powerpoint

Unlike control corticospinal axons (top), those lacking PTEN (middle) regenerate and extend past the lesion site after spinal hemisection. Knocking out DLK and LZK (bottom) eliminates the effect of PTEN knockout. See Saikia et al. for details.

Roles of Central Amygdala Neurons in Place Preference

Marion Ponserre, Federica Fermani, Louise Gaitanos, and Rüdiger Klein

(see pages 3783–3796)

The central nucleus of the amygdala (CeA) drives appropriate autonomic and behavioral responses to stimuli with innate or acquired, positive or negative valence. This function is mediated by the activity of two major populations of inhibitory neurons: those that express somatostatin and those that express protein kinase C δ (PKCδ). Exactly how these populations contribute to the processing of inputs and the generation of outputs that select specific behaviors remains poorly understood. To elucidate the roles of these populations, Ponserre, Fermani, et al. recorded from or optogenetically inhibited each population as mice learned and performed a conditioned place preference task.

When mice were trained to associate one chamber of a two-chamber arena with food, they subsequently spent more time in the food-paired chamber than in the neutral chamber. Calcium imaging suggested that the representation of the food-paired chamber by both somatostatin-expressing and PKCδ-expressing neurons increased as mice learned to associate the chamber with food. For example, a subset of neurons of each type began to respond when the mouse entered the food-paired chamber and/or the food-delivery zone of that chamber. Furthermore, the pattern of activity in either population could be used reliably to infer whether the mouse was in the food-paired chamber. Nevertheless, only somatostatin-expressing neurons were consistently active during food consumption. Moreover, whereas silencing PKCδ-expressing neurons either throughout the training period or selectively during the test period reduced place preference, silencing somatostatin-expressing neurons did not alter place preference.

These results suggest that both somatostatin-expressing and PKCδ-expressing CeA neurons guide mice as they forage for food in a familiar environment. Although this is consistent with previous work suggesting that subsets of somatostatin-expressing neurons promote feeding, it differs from several studies suggesting that PKCδ-expressing CeA neurons suppress eating and promote aversive responses. The authors hypothesize that PKCδ-expressing neurons may instead have a more general role in representing contexts in which salient events occur. It is also likely that different subpopulations of somatostatin-expressing and PKCδ-expressing neurons have different roles. Dissecting these roles is an important direction for future work.