Astrocytes Mediate IGF-1 Effects on Synaptic Plasticity
José Antonio Noriega-Prieto, Laura Eva Maglio, Jonathan A. Zegarra-Valdivia, Jaime Pignatelli, Ana M. Fernandez, et al.
(see pages 4768–4781)
Insulin-like growth factor 1 (IGF-1) is a peptide with diverse roles in the CNS, including promotion of neurogenesis, dendrite growth, and synaptic plasticity. It acts by binding to IGF1 receptors (IGF1Rs), which are present on both neurons and glia. Downstream signaling pathways have various effects, which can differ from one brain area to another. Therefore, how IGF-1 regulates synaptic plasticity remains incompletely understood.
Noriega-Prieto et al. have investigated how IGF-1 modulates synaptic strength in the barrel cortex of young mice. Injection of IGF-1 into layer II/III increased the amplitude of local field potentials evoked by repeated whisker deflection in vivo. Similarly, IGF-1 induced long-term potentiation (LTP) of synapses between layer IV axons and layer II/III pyramidal neurons in slices. Furthermore, IGF-1 lowered the threshold for spike timing-dependent LTP and induced short-term potentiation (STP) of EPSCs in pyramidal neurons. In addition, IGF-1 induced long-term depression (LTD) of IPSCs in pyramidal cells by reducing the probability of presynaptic GABA release.
Importantly, IGF-1 also induced calcium transients in astrocytes and triggered ATP release from these cells. Moreover, STP and LTD induction by IGF-1 were blocked by chelating astrocytic calcium, knocking out astrocytic IGF1Rs (and thus blocking ATP release from astrocytes), or inhibiting adenosine A2A receptors. Finally, knocking out astrocytic IGF1Rs prevented the effects of IGF-1 on LTP threshold in vivo, and impaired the performance of mice on a whisker-mediated texture-discrimination task.
These results suggest that IGF-1 can regulate synaptic plasticity indirectly, by triggering release of ATP from astrocytes. ATP is metabolized to adenosine extracellularly, and adenosine activates A2A receptors on GABAergic terminals. The resulting decrease in GABA release may free pyramidal cells from tonic inhibition, thus increasing their spiking, which enables potentiation of excitatory input. This process appears to be essential for learning in the whisker system.
IGF-I induces LTD of IPSCs in layer II/III pyramidal neurons by stimulating ATP release from astrocytes, leading to activation of adenosine receptors that reduce GABA release from GABAergic interneurons. See Noriega-Prieto, et al. for details.
Some Neurons in Crow Brain Encode Numerosity Zero
Maximilian E. Kirschhock, Helen M. Ditz, and Andreas Nieder
(see pages 4889–4896)
The concept of zero as a number seems simple and even intuitive to most people, but it appeared relatively late in human history. Indeed, zero was first described as a number subject to mathematical rules by an Indian mathematician in AD 628, and when the idea reached Europe in the 13th century, it was controversial. This late conceptualization of zero becomes less surprising when one considers that the nervous system evolved to process sensory stimuli, not the absence of stimuli (Nieder, 2016, Trends Cogn Sci 20:830). Creating neural representations of no stimulus likely requires training.
Many animals, including insects, fish, birds, and monkeys have been trained to respond to the absence of stimulation. Whether these responses reflect the encoding of zero as a number or simply “no stimulus” is usually unclear. To assess this, researchers must determine whether errors on trials with zero items are similar to those on trials with more items. One such error is the distance effect seen in delayed-match-to-numerosity tests. In these tests, animals are shown a stimulus, followed by a delay, then a test stimulus, and they must report whether the test stimulus has the same number of items as the sample. Animals are more likely to report a match erroneously if the sample and test stimuli have similar numbers of items. This distance effect has been shown for numerosity zero in humans, macaques, and honeybees. Furthermore, neurons tuned to numerosity zero have been identified in the prefrontal cortex of macaques. Kirschhock et al. report similar encoding of numerosity zero in carrion crows.
Two crows were trained on the numerosity-matching test. One showed a numerical distance effect for numerosity zero (making more mistakes when the test stimulus had one item than when it had two items), but the second crow made too few mistakes to discern the effect. Consistent with previous work, neurons in the nidopallium caudolaterale—considered analogous to mammalian prefrontal cortex—were tuned to specific numerosities, including numerosity zero. Notably, a distance effect was apparent in the tuning curves of these neurons: neurons that spiked most in response to zero items responded progressively less to one and two items. Finally, a classifier trained on neuronal population responses could decipher the numerosity of stimuli, including numerosity zero.
These results suggest that carrion crows represent zero as a number in a continuum with higher numbers. This adds to accumulating evidence that, despite the lack of a layered cerebral cortex, corvids have cognitive abilities that rival those of primates.
Footnotes
This Week in The Journal was written by Teresa Esch, Ph.D.
In this issue
