Striatal Axons Guide Corticofugal Axons through Subpallium
Jacqueline M. Ehrman, Paloma Merchan-Sala, Lisa A. Ehrman, Bin Chen, Hee-Woong Lim, et al.
(see pages 3344–3364)
Axons that extend long distances often make multiple turns as they grow along stereotyped trajectories toward their targets. Some axons seem to cheat at this task by simply following trajectories established by earlier-growing pioneer axons. Corticothalamic and thalamocortical axons rely partly on this strategy, following each other's earlier trajectories after meeting in the internal capsule. Cortical axons that project to the hindbrain and spinal cord veer off from corticothalamic axons in the posterior internal capsule and instead extend through the cerebral peduncle to the midbrain. Ehrman et al. provide evidence that these axons follow pioneer axons projecting from the striatum to the substantia nigra pars reticula (SNr).
Examination of striatal, corticofugal, and thalamocortical axons at different embryonic ages revealed that striatal axons are the first to traverse the region of the internal capsule, with many axons reaching as far as the SNr by embryonic day 14.5 (E14.5). Corticofugal and thalamocortical axons grow through the internal capsule over the next 2 d. As expected, corticofugal axons were closely associated with thalamocortical axons in the rostral part of the internal capsule at E16.5. Toward the posterior internal capsule, however, descending corticofugal axons were more closely associated with striatonigral axons, and they continued to grow alongside these axons into the cerebral peduncle. Importantly, knocking out Isl1—a transcription factor expressed by striatonigral neurons but not by corticofugal or thalamocortical neurons—disrupted growth of both striatonigral and corticofugal axons. Instead of growing together in a tight bundle passing underneath the subthalamic nucleus, both sets of axons grew through this nucleus in small bundles. Corticofugal axons remained defasciculated as they grew through, rather than under, the SNr. Consequently, no cerebral peduncle was present in mutant mice. RNA-sequencing analyses comparing wild-type and Isl1-deficient striatal tissue identified several molecules whose increased or decreased expression might contribute to these navigational errors.
Altogether, the data suggest that striatonigral axons act as pioneers for corticofugal axons as they grow through the internal capsule and form the cerebral peduncle. It remains possible, however, that other Isl1-expressing cells that line the pathway contribute to guidance of these axons. Future work should address this possibility and pinpoint the molecules responsible for axon growth along these pathways.
At E14.5, striatonigral axons (green) have begun to reach the SNr, while thalamocortical axons (blue) have just reached the globus pallidus, and corticofugal axons (red) remain in the lateral striatum. See Ehrman et al. for details.
Somatic Depolarization Boosts Effects of Dendritic Spikes
Tobias Bock, Adrian Negrean, and Steven A. Siegelbaum
(see pages 3406–3425)
Pyramidal neurons in hippocampal area CA1 receive information from entorhinal cortex at their distal dendrites, and receive information from CA3 at their proximal dendrites. EPSPs evoked by entorhinal inputs typically dissipate before reaching the soma, but, with sufficient local depolarization, they can generate dendritic calcium spikes that propagate to the soma. When paired with proximal stimulation, these dendritic spikes can trigger bursts of somatic action potentials that are thought to contribute to learning. Recordings in slices indicate that dendritic spikes rarely produce suprathreshold depolarization at the soma on their own, but Bock et al. suggest that is not the case in vivo, where the resting membrane potential of pyramidal cells is more depolarized than in slices.
In paired patch recordings from the distal apical dendrite and soma of CA1 pyramidal cells in mouse hippocampal slices, dendritic spikes were greatly attenuated if the soma was held near the typical in vitro resting membrane potential (–70 mV). Attenuation was greatly reduced, however, when the soma was held closer to in vivo resting potentials (–55 mV). Somatic depolarization to –55 mV also increased the number of dendritic spikes generated during trains of paired distal and proximal stimulation and increased the probability that these spikes would trigger a burst of somatic spikes. Consequently, somatic depolarization facilitated a form of synaptic plasticity induced by such stimulation. Notably, somatic depolarization did not alter dendritic spike threshold, amplitude, or duration or enhance propagation of distal postsynaptic potentials that did not generate a dendritic spike.
Because A-type potassium channels are present in apical dendrites, attenuate backpropagating action potentials, and become inactivated with mild depolarization, the authors asked whether inactivation of these channels underlies the enhancement of dendritic spike propagation resulting from somatic depolarization. Indeed, blocking the channels replicated and occluded the effects of somatic depolarization on dendritic spike generation and propagation. Computational modeling further demonstrated that somatic depolarization can enhance the propagation of dendritic spikes by increasing inactivation of A-type channels in proximal dendrites. Thus, subthreshold modulation of somatic membrane voltage, which may result from changes in synaptic input or neuromodulation, can alter the ability of entorhinal inputs to influence CA1 pyramidal cell output and plasticity.
Footnotes
This Week in The Journal was written by Teresa Esch, Ph.D.