Experimental mismatch in neural circuits


The finding that acute and chronic manipulations of the same neural circuit can produce different behavioural outcomes poses new questions about how best to analyse these circuits. See Article p.358

In 1949, Walter Rudolf Hess shared the Nobel Prize in Physiology or Medicine for his work using acute electrical stimulation to study neural circuits. Modern neuroscience is dominated by a newer, more sophisticated technique for acute circuit manipulation: optogenetics, in which light-sensitive ion-channel proteins are engineered to activate or inhibit select neurons1. However, a nagging doubt pervades the field — do the behavioural effects of acutely activating or silencing specific neurons reflect the normal functions of these cells? On page 358 of this issue, Otchy et al.2 systematically address this question. Their findings are bound to excite lively discussion.

If acute inactivation of a particular neural circuit alters an animal's behaviour, the seemingly logical conclusion is that the circuit controls the behaviour. But the brain's circuits are densely interconnected, so how can we be sure that these behavioural effects are not caused by changes to other, connected, circuits that normally do not participate in the targeted behaviour but are affected by the manipulation? Otchy et al. used a brilliant study design to test this idea. They reasoned that, if the effects of acute manipulation are directly caused by the manipulated neurons, then chronically manipulating those neurons, for example by permanently impairing (lesioning) them, should have the same effect. The authors compared the effects of chronic and acute neural manipulations in rats and in zebra finches. They examined behavioural tasks that were learnt before the manipulations, but that were not repeatedly practised afterwards, avoiding the confounding effect of relearning a task after an experimental manipulation.

First, Otchy et al. demonstrated that, in rats that had learnt a complex lever-pressing task, acute silencing of neurons in the brain's motor cortex using the drug muscimol profoundly impaired task performance. Acute optogenetic activation of motor-cortex neurons produced a similar effect. The same research group had shown previously3 that surgical ablation of the motor cortex blocked the initial learning of the lever-pressing task, but had no significant effect on the ability of rats to perform the task if it had been learnt before surgery. Thus, acute and chronic manipulations produce discrepant results in this circuit (Fig. 1a).

Figure 1: Mixed messages from neural manipulations.

Otchy et al.2 compared the effects of acute and chronic manipulations of neural circuits on a specific behaviour. a, The authors taught rats to perform a complex lever-pressing task. Chronic inhibition of neurons in the brain's motor cortex did not affect task performance, whereas acute perturbations strongly impaired performance. b, Likewise, chronic ablations of neurons in the sensorimotor nucleus interface (Nif) of the brains of zebra finches did not affect their songs, whereas songs became variable and unstructured after acute inhibition.

In a second set of experiments, Otchy and colleagues used muscimol to inactivate song neurons in a brain region called the sensorimotor nucleus interface (Nif) in zebra finches. This acute manipulation massively impaired birdsong, whereas chronic lesioning of Nif had no effect two days after the lesion (Fig. 1b). Investigating this apparent paradox, the authors showed that the Nif lesions did initially cause a change in the downstream neural circuitry controlling birdsong, but that this change spontaneously recovered without training after 3.4 hours. The researchers propose that homeostatic plasticity, which adjusts the overall activity level of neurons in a circuit, might be involved in this recovery. However, other processes that change the strength of the synaptic connections between these neurons are equally likely to be responsible.

How should we interpret these experiments? Two opposing hypotheses come to mind. First, that acute manipulations are unreliable and should be discarded in favour of chronic manipulations. Second, that acute manipulations elicit results that truly reflect normal circuit functions, and the lack of changes after chronic manipulations is caused by compensatory plasticity.

Before choosing between these stark alternatives, several facts should be taken into account. Many chronic manipulations of neural circuits (both permanent genetic changes and physical lesions) do actually produce major behavioural changes. For example, in rodent and human brains, lesions in the amygdala region impair fear memories4, and hippocampal lesions interfere with spatial memory5. Chronic deletion of the synaptic cell-adhesion molecule neuroligin-3 in striatal neurons alters learning of a repetitive motor task6. Thus, the finding that a chronic manipulation does not cause a behavioural change cannot simply be attributed to plasticity and compensation.

Clearly, it is possible to dissect the functions of some types of neuron and circuit using chronic manipulations, making this a compelling overall experimental approach. But acute optogenetic manipulations are generally easier to perform, and the conclusions drawn from many such manipulations do correlate well with those from chronic manipulations (see, for example, ref. 4). Moreover, such acute manipulations often match changes in neural activity observed during the targeted behaviour in vivo7,8,9, although a caveat of acute manipulations is that natural neural activity is normally limited to only a subset of neurons in a circuit, whereas acute manipulations are mostly not.

There are multiple explanations for why acute and chronic manipulations might produce distinct results, which makes it difficult, or perhaps even impossible, to assess whether results reflect 'off-target' or 'on-target' effects, as Otchy et al. aptly call them. The authors point out that, because neural circuits are massively interconnected, acute manipulations are probably more susceptible to off-target effects than are chronic lesions. This is because acute manipulations are more likely to spread to other connected circuits that have no normal role in the targeted behaviour. Therefore, we cannot simply assume that the behavioural readouts of such manipulations always reflect the normal functions of the manipulated circuits.

Where do we go from here? Most acute manipulation studies that use optogenetics confirm, and so add valuable support to, existing hypotheses that were established in earlier studies. But for those studies that have proposed new circuit functions, it may be advisable to re-evaluate the conclusions using independent approaches.

In the future, it might be helpful always to correlate acute and chronic manipulations of specific neurons. If results from acute and chronic manipulations are discrepant, analyses of circuits that act in parallel to the manipulated circuit, or of similar neurons that are activated by different stimuli, might be more likely to provide an explanation for the discrepancy than examination of chains of hierarchically connected neurons, because off-target effects probably propagate throughout neural circuits by spilling over into adjacent, connected circuits. Moreover, studies of a broad range of behaviours might be helpful — restricting a study to a few behaviours could make it harder to detect off-target effects. Overall, more caution about the conclusions drawn from circuit manipulations, be they acute or chronic, seems advisable, because most current studies focus on only one circuit and one behaviour.

It is both an exciting and a sobering time for neuroscience. Exciting, because it is now possible to manipulate neurons and circuits with an ease that was only dreamt of a few years ago. Sobering, because the massively parallel and interconnected nature of neural circuits is becoming apparent, and the complexity imposed on such circuits by various forms of plasticity has yet to be even touched on. By using parallel approaches to study circuits, we can develop an understanding of the brain that acknowledges the limitations of this understanding, as well as its achievements. Such a strategy will drive the field forward.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Emiliani, V., Cohen, A. E., Deisseroth, K. & Häusser, M. J. Neurosci. 35, 13917–13926 (2015).

    CAS  Article  PubMed  Google Scholar 

  2. 2

    Otchy, T. M. et al. Nature 528, 358–363 (2015).

    ADS  CAS  Article  PubMed  Google Scholar 

  3. 3

    Kawai, R. et al. Neuron 86, 800–812 (2015).

    CAS  Article  PubMed  Google Scholar 

  4. 4

    Johansen, J. P., Wolff, S. B., Lüthi, A. & LeDoux, J. E. Biol. Psychiatry 71, 1053–1060 (2012).

    Article  PubMed  Google Scholar 

  5. 5

    Squire, L. R., Genzel, L., Wixted, J. T. & Morris, R. G. Cold Spring Harb. Perspect. Biol. 7, a021766 (2015).

    Article  PubMed  Google Scholar 

  6. 6

    Rothwell, P. E. et al. Cell 158, 198–212 (2014).

    CAS  Article  PubMed  Google Scholar 

  7. 7

    Buetfering, C., Allen, K. & Monyer, H. Nature Neurosci. 17, 710–718 (2014).

    CAS  Article  PubMed  Google Scholar 

  8. 8

    Chaumont, J. et al. Proc. Natl Acad. Sci. USA 110, 16223–16228 (2013).

    ADS  CAS  Article  PubMed  Google Scholar 

  9. 9

    Kravitz, A. V., Owen, S. F. & Kreitzer, A. C. Brain Res. 1511, 21–32 (2013).

    CAS  Article  PubMed  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Thomas C. Südhof.

Related links

Related links

Related links in Nature Research

Neuroscience: When lights take the circuits out

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Südhof, T. Experimental mismatch in neural circuits. Nature 528, 338–339 (2015).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.