This Week in The Journal

Dynein Moves Late Endosomes in Dendrites

Chan Choo Yap, Laura Digilio, Lloyd P. McMahon, Tuanlao Wang, Bettina Winckler

(see pages 4415–4434)

Cells continually make new proteins and degrade old ones that have become damaged. For membrane proteins, the degradation process begins with endocytosis and delivery to early endosomes, where to-be-degraded proteins are separated from those destined to return to the plasma membrane. To-be-degraded proteins then transit from early endosomes to late endosomes, and ultimately to lysosomes, where they are degraded. In neurons, early endosomes are distributed throughout dendrites and are relatively immobile. In contrast, late endosomes move in both directions along dendrites, with a slight bias for retrograde movement toward the soma and proximal dendrites, where acidified lysosomes are concentrated. Which molecular motors transport late endosomes in dendrites has been unclear, because unlike in axons, microtubules in dendrites are of mixed polarity, with ∼55% having their plus ends pointed away from the cell body and the remainder having their plus ends toward the cell body. Therefore, both plus-end-directed kinesins and minus-end-directed dynein can move cargoes in both directions in dendrites. Yap et al. asked which of these motors carry late endosomes.

The tethering and movement of endosomes are mediated by Rab GTPases and their effectors. Late endosomes contain Rab7. Notably, one Rab7 effector—Rab-interacting lysosomal protein (RILP)—binds to dynein, whereas another effector, FYCO1, binds to kinesin. Yap et al. found that overexpression of FYCO1 did not notably affect late endosome distribution in cultured hippocampal neurons. In contrast, overexpression of RILP increased retrograde movement of late endosomes, resulting in the loss of late endosomes from dendrites and the accumulation of these compartments in the perinuclear region of the soma. This effect required RILP interactions with Rab7 and dynein, both of which were recruited to late endosomes in greater numbers when RILP was overexpressed. Importantly, preventing RILP–Rab7 interaction or inhibiting dynein reduced both anterograde and retrograde movement of late endosomes, slowed late endosome maturation, and led to dendritic accumulation of proteins that otherwise would be degraded.

These results indicate that dynein moves late endosomes in both directions in dendrites and thus contributes to protein degradation. The slight bias toward retrograde movement likely stems from the slight preponderance of dendritic microtubules with minus ends toward the soma. How dynein influences endosome maturation remains to be determined.

• Download figure
• Open in new tab
• Download powerpoint

Overexpressing FYCO1 (top) does not alter the distribution of Rab7-expressing compartments in cultured hippocampal neurons, whereas overexpressing RILP (bottom) causes these compartments to cluster around the nucleus. See Yap et al. for details.

BLA Projections to BNST May Have Male-Specific Roles

Jaime E. Vantrease, Brittany Avonts, Mallika Padival, M. Regina DeJoseph, Janice H. Urban, et al.

(see pages 4488–4504)

Fear and anxiety are related emotions that promote defensive avoidance behaviors in the face of threat. Considerably less is known about the neural circuits underlying anxiety than those involved in fear, however. Anxiety typically involves prolonged states of avoidance in response to ambiguous threats or threats that occur sporadically over an extended time. In rodents, context-induced fear, prolonged responding to long-duration cues, and social hesitancy are considered anxiety-like behaviors. The basolateral amygdala (BLA) and the bed nucleus of the stria terminalis (BNST) have been implicated in anxiety and anxiety-like behaviors in both humans and rodents. One might predict, therefore, that BLA neurons that project to the BNST contribute to these behaviors. Vantrease et al. provide evidence that this is indeed the case, but perhaps only in males.

Rats underwent fear conditioning in which five footshocks were delivered at various times during an 8 min tone. Four days later, freezing behavior during the tone was quantified. Although males and females showed similar amounts of freezing during training, males froze more than females during testing, suggesting that fear responses were extinguished more quickly in females. Notably, selectively inhibiting BNST-projecting BLA neurons reduced freezing in males, but increased freezing in females. In addition, inhibiting BNST-projecting BLA neurons reduced the latency to interact with an unfamiliar rat and increased total interaction time in males, but had no effect in females.

The sex-specific effects of inhibiting BLA–BNST neurons likely stemmed from sex differences in the neurons' baseline activity. BLA neurons in females are generally more active than those in males, partly because female neurons have smaller spike afterhyperpolarization (AHP) and thus are more excitable. But BLA–BNST neurons in females had larger AHPs and were less excitable than adjacent BLA neurons. Conversely, BLA–BNST neurons in males had smaller AHPs and were more excitable than surrounding neurons. Consequently, BLA–BNST neurons were more excitable and more active in males than in females.

These data suggest that elevated excitability of BLA–BNST neurons helps to sustain fear responses to prolonged cues and inhibit social interaction in male rats. Future work should determine whether increasing the excitability of BLA–BNST neurons promotes anxiety-like behaviors in females.