Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Brief Communication
  • Published:

An excitatory basis for divisive normalization in visual cortex

Nature Neuroscience volume 19, pages 568–570 (2016)Cite this article

Subjects

Abstract

Neurons in visual cortex are connected not only locally, but also through networks of distal connectivity. These distal networks recruit both excitatory and inhibitory synapses and result in divisive normalization. Normalization is traditionally thought to result from increases in synaptic inhibition. By combining optogenetic stimulation and intracellular recordings in mouse visual cortex, we found that, on the contrary, normalization is a result of a decrease in synaptic excitation.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on SpringerLink
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Figure 1: Distal cortical activation causes contrast- and time-dependent summation and division.
Figure 2: Distal network activation causes context-dependent synaptic effects.
Figure 3: Roles of excitation and inhibition in divisive suppression.

Similar content being viewed by others

References

  1. Angelucci, A. & Bressloff, P.C. Prog. Brain Res. 154, 93–120 (2006).

    Article  PubMed  Google Scholar 

  2. Sato, T.K., Häusser, M. & Carandini, M. Nat. Neurosci. 17, 30–32 (2014).

    Article  CAS  PubMed  Google Scholar 

  3. Nassi, J.J., Avery, M.C., Cetin, A.H., Roe, A.W. & Reynolds, J.H. Neuron 86, 1504–1517 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. Carandini, M. & Heeger, D.J. Nat. Rev. Neurosci. 13, 51–62 (2012).

    Article  CAS  Google Scholar 

  5. Olsen, S.R., Bhandawat, V. & Wilson, R.I. Neuron 66, 287–299 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Zhu, P., Frank, T. & Friedrich, R.W. Nat. Neurosci. 16, 1678–1686 (2013).

    Article  CAS  PubMed  Google Scholar 

  7. Haider, B. et al. Neuron 65, 107–121 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Adesnik, H., Bruns, W., Taniguchi, H., Huang, Z.J. & Scanziani, M. Nature 490, 226–231 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Ozeki, H., Finn, I.M., Schaffer, E.S., Miller, K.D. & Ferster, D. Neuron 62, 578–592 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Carandini, M., Heeger, D.J. & Movshon, J.A. J. Neurosci. 17, 8621–8644 (1997).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Chance, F.S. & Abbott, L.F. Network 11, 119–129 (2000).

    Article  CAS  PubMed  Google Scholar 

  12. Somers, D.C. et al. Cereb. Cortex 8, 204–217 (1998).

    Article  CAS  PubMed  Google Scholar 

  13. Douglas, R.J., Koch, C., Mahowald, M., Martin, K.A.C. & Suarez, H.H. Science 269, 981–985 (1995).

    Article  CAS  PubMed  Google Scholar 

  14. Rubin, D.B., Van Hooser, S.D. & Miller, K.D. Neuron 85, 402–417 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Shushruth, S. et al. J. Neurosci. 32, 308–321 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Chung, S. & Ferster, D. Neuron 20, 1177–1189 (1998).

    Article  CAS  PubMed  Google Scholar 

  17. Shu, Y., Hasenstaub, A. & McCormick, D.A. Nature 423, 288–293 (2003).

    Article  CAS  PubMed  Google Scholar 

  18. Tsodyks, M.V., Skaggs, W.E., Sejnowski, T.J. & McNaughton, B.L. J. Neurosci. 17, 4382–4388 (1997).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Katzner, S., Busse, L. & Carandini, M. J. Neurosci. 31, 5931–5941 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Reynolds, J.H. & Heeger, D.J. Neuron 61, 168–185 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Petreanu, L., Huber, D., Sobczyk, A. & Svoboda, K. Nat. Neurosci. 10, 663–668 (2007).

    Article  CAS  PubMed  Google Scholar 

  22. Madisen, L. et al. Nat. Neurosci. 15, 793–802 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Hippenmeyer, S. et al. PLoS Biol. 3, e159 (2005).

    PubMed  PubMed Central  Google Scholar 

  24. Margrie, T.W., Brecht, M. & Sakmann, B. Pflugers Arch. 444, 491–498 (2002).

    Article  CAS  PubMed  Google Scholar 

  25. Haider, B., Häusser, M. & Carandini, M. Nature 493, 97–100 (2013).

    Article  PubMed  Google Scholar 

  26. Schmidt-Hieber, C. & Häusser, M. Nat. Neurosci. 16, 325–331 (2013).

    Article  CAS  PubMed  Google Scholar 

  27. Barry, P.H. J. Neurosci. Methods 51, 107–116 (1994).

    Article  CAS  PubMed  Google Scholar 

  28. Wehr, M. & Zador, A.M. Nature 426, 442–446 (2003).

    Article  CAS  PubMed  Google Scholar 

  29. Li, Y.T., Ibrahim, L.A., Liu, B.H., Zhang, L.I. & Tao, H.W. Nat. Neurosci. 16, 1324–1330 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Lien, A.D. & Scanziani, M. Nat. Neurosci. 16, 1315–1323 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We thank C. Reddy, C. Schmidt-Hieber, M. Basche and M. Rizzi for technical help, K. Svoboda for the gift of pCAGGS-ChR2-Venus, H. Kawasaki for the gift of pCAG-mCherry, and M. Dipoppa, N. Steinmetz, K.D. Harris and others in our laboratories for helpful discussions. This work was supported by the Wellcome Trust, the Simons Foundation, the Gatsby Charitable Foundation and the European Research Council. M.C. holds the GlaxoSmithKline/Fight for Sight chair in Visual Neuroscience.

Author information

Author notes

  1. Tatsuo K Sato & Bilal Haider

    Present address: Present addresses: Institute of Neuroscience, Technische Universität München, Munich, Germany (T.K.S.), Coulter Department of Biomedical Engineering, Georgia Institute of Technology & Emory University School of Medicine, Atlanta, Georgia, USA (B.H.).,

Authors and Affiliations

  1. UCL Institute of Ophthalmology, University College London, London, UK

    Tatsuo K Sato, Bilal Haider & Matteo Carandini

  2. Wolfson Institute for Biomedical Research, University College London, London, UK

    Michael Häusser

Authors
  1. Tatsuo K Sato
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Bilal Haider
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Michael Häusser
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Matteo Carandini
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

T.K.S., B.H., M.H. and M.C. designed the study. T.K.S. and B.H. performed the experiments. T.K.S., B.H. and M.C. analyzed the data. T.K.S., B.H., M.H. and M.C. wrote the paper.

Corresponding authors

Correspondence to Tatsuo K Sato or Matteo Carandini.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 Reliability of Vm depolarization caused by distal network activation in the absence of visual stimulation.

(a) Vm depolarization caused by distal network activation in a representative neuron (the neuron in Fig. 2a). The horizontal line indicates Vbottom (see Online Methods). The thick and thin curves indicate mean ± s.d. (n = 20 trials). Note smaller s.d. in the earlier period (< 150 ms). (b) Reliability in Vm response, measured as mean/s.d., i.e. the reciprocal of coefficient of variation. Response was measured from Vbottom. (c) Similar to b, but for n = 14 neurons (mean ± s.e.m.). (d) Trials in a were separated into less quiescent (n = 11 trials, thick black trace) or more quiescent trials (n = 9 trials, thick gray trace) based on Vm level for the 100 ms just before the activation. Note how black and gray lines overlap. The horizontal line indicates Vbottom. (e) Similar to d, without distal network activation. (f) The height of Vm depolarization measured from Vbottom. We considered neurons with mean/s.d. > 2 that had a transient depolarization in the earlier period (<150 ms, n = 9 neurons, see Online Methods) or slow depolarization in the later period (>150 ms, n = 7). We saw no significant difference between less quiescent and more quiescent trials (12.4 ± 1.9 vs 11.6 ± 2.2 mV for transient, p>0.05, 12.2 ± 1.6 vs 12.2 ± 1.9 mV for slow depolarization, P > 0.05, Wilcoxon signed rank test). See also Supplementary Figure 4, which analyzes EPSCs and IPSCs rather than Vm.

Supplementary Figure 2 Vm depolarization and hyperpolarization caused by distal network activation during current injection.

(a, b) Vm measurements from a representative neuron in the absence (a) or in the presence (b) of visual stimulation. Either positive (+50 pA), no or negative (-250 pA) current was injected from -200 ms to +475 ms as indicated. (c) Vm depolarization with current injection in the absence of visual stimulation (n = 14). Difference was calculated between Vm with and without distal network activation for the same current injection. (d) Same as c, in the presence of visual stimulation (n = 14). (e) Vm change as a function of baseline Vm in the absence of visual stimulation (n = 14 neurons) in 0-150 ms (top), 150-300 ms (middle) and 300-450 ms (bottom) intervals. No significant correlation (P > 0.05, Spearman's rank correlation). (f) Similar to e, in the presence of visual stimulation (n = 14 neurons). In the 150-300 ms interval, Vm change correlated negatively with baseline Vm (P = 0.047, Spearman's rank correlation, n = 42 points).

Supplementary Figure 3 Vm hyperpolarization caused by distal network activation with a potassium chloride-based internal solution.

(a, b) Vm measurements from an example neuron in the presence of visual stimulation. Average Vm traces without (a) and with (b) distal network activation. (c) Difference in Vm between two conditions a and b. (d) Same as c, for all 4 neurons.

Supplementary Figure 4 Reliability of EPSCs and IPSCs caused by distal network activation in the absence of visual stimulation.

(a) EPSC response caused by distal network activation in a representative neuron. Thick and thin red lines indicate mean ± s.d. (n = 15 trials). Note smaller s.d. in the earlier period (< 150 ms). (b) Reliability in the EPSC. The reciprocal of coefficient of variation (1/C.V.) was calculated from a. (c) Trials were grouped into less or more quiescent trials (8, 7 trials respectively) base on the PSC level prior to distal network activation (-100 to 0 ms). Note the transient PSC and the slow PSC overlap between two groups. (d) The peak of the transient PSC (1/C.V. > 2, n = 8 neurons) for less or more quiescent trials. No significant difference (-112 ± 23 vs -105 ± 26 pA, P = 0.38) (e) Same as d for the slow PSC (the average PSC between 300 and 450 ms, n = 7 neurons, 1/C.V. > 2) No significant difference (-97 ± 22 vs -96 ± 23 pA, P = 0.94) (f-j) Same as a-e for the IPSC. f-h were from the same neuron in a-c. No significant difference for i (610 ± 157 vs 550 ± 146 pA, P = 0.37, n = 7). Sample size for j was too small to check for significance (314 ± 39 vs 308 ± 60 pA, n = 3).

Supplementary Figure 5 Distal network activation at 100% visual contrast reduces synaptic conductances in all 10 cells recorded in voltage clamp.

Traces from 10 cells showing conductance responses to distal network activation during responses to 100% contrast visual stimulation. Each conductance was normalized to the average measured during visual responses in the 100 ms before distal network activation. Red: excitatory conductance, Cyan: inhibitory conductance. The average across cells of these data appears in Fig. 3a.

Supplementary Figure 6 Vm and EPSCs measurements from the same neurons with a potassium gluconate-based internal solution.

(a, b) Vm measurements from a representative neuron in the absence (a) or in the presence (b) of visual stimulation. Thin and thick traces represent with and without distal network activation. (c, d) Current measurements in the absence (c) or in the presence (d) of visual stimulation. (e) Time course of Vm (magenta) and current (red) in the absence of visual stimulation. To compare response in different units (mV and pA), the response was scaled to the size of visual response for the 100 ms before distal network activation. (f) Same as e in the presence of visual stimulation (blue: Vm, red: current). (g) Same as e, averaged over 5 neurons. Thick and thin lines indicate mean ± s.e.m. (h) Same as g in the presence of visual stimulation. (i) Relationship between Vm and currents (0 - 600 ms after distal network activation) for the single neuron shown in a-f, in the absence (gray) or presence (black) of visual stimulation. Dots indicate values measured in the 100 ms before distal network activation. (j) Same as i, averaged over 5 neurons.

Supplementary Figure 7 Roles of excitation and inhibition in additive effects. Same as Figure 3b,c, but in the absence of visual stimulation.

(a) Relationship between inhibition and excitation in the absence of visual stimulation. Data are from Figure 2h,l, normalized in each cell so that 0% and 100% are the average values measured in the 100 ms before distal network activation at 0% contrast (thick gray) and 100% contrast (thick black). When distal network activation arrives in the presence of 0% contrast, it increases both inputs proportionally (thin gray, 0-700 ms). The crosses show mean ± s.e.m. (n = 10) for excitation and inhibition 100 ms and 420 ms after distal network activation. (b) Predictions of Vm based on synaptic conductances measured at 0% contrast. The measured Vm averaged over 14 neurons following distal network activation (dashed) is poorly predicted by inhibitory conductance alone (cyan), as inhibition would predict a hyperpolarization. It is better predicted by synaptic excitation (pink) especially in combination with inhibition (black). Shaded areas indicate mean ± s.e.m. (n = 10 neurons).

Supplementary Figure 8 Lack of transient effects of distal network activation at high contrast.

(a) On rare occasions, distal network activation caused transient depolarizations, visible only during current injection. Bars show Vm measurements for each neuron, averaged 80–120 ms after distal network activation, during hyperpolarizing current injection (for example, Supplementary Figure 2b). Vm was transiently depolarized in 3/14 neurons (three thick lines, P = 0.009, 0.006, 0.0009, 2.1 ± 0.8, 0.7 ± 0.2, 4.1 ± 1.0 mV, n = 50, 30, 36 trials, respectively). No significant effect was found in the population (n = 14 cells, -0.3 ± 0.5 mV, P = 0.4, Wilcoxon signed rank test). No neurons reach significance without current injection. (b) Distal network activation caused a transient increase in EPSCs in only one out of 10 cells (thick line, n = 15 trials, -19.7 ± 5.4 pA, P = 0.002). No significant effects were seen in the population (n = 10 cells, -2.2 ± 4.1 pA, P = 0.70). (c) Distal network activation caused no transient increases in IPSCs. Significance is reached neither across the population (n = 10 cells, -17.4 ± 34.8 pA, P = 0.77) nor in any individual cell.

Supplementary Figure 9 Current clamp and voltage clamp measurements of GABAA input elicited optogenetically.

(a) Design of current clamp experiments. Channelrhodopsin-2 (ChR2) was expressed in PV interneurons in Pvalb-IRES-Cre;Ai32 transgenic mice. Recordings were made in excitatory neurons with a potassium-gluconate internal solution. (b) PV activation evoked depolarizing or hyperpolarizing PSPs depending on baseline Vm. Baseline Vm was controlled through current injection (62.5, 0 and -125 pA, light gray to black). (c) The sign of PSPs reversed at -71.7 mV for the neuron shown in b. The average was -71.4 ± 1.5 mV (n = 4 neurons), close to the potential (-74 mV) predicted by the Nernst equation based on the chloride concentration in aCSF. The linear regression line was drawn ignoring points for Vm < -90 mV, to avoid strong contributions of nonlinear membrane conductances. (d) Design of voltage clamp experiments. Recordings were made with a cesium methanesulphonate internal solution. (e) PV activation evoked inward or outward PSCs, depending on command voltage (Vcmd, -51, -61 and -72 mV, light gray to black). (f) The sign of PSCs reverses at ‑62.4 mV for a neuron shown in e (-63.2 ± 1.3 mV, n = 4 neurons). This is close to the potential (-62 mV) predicted by the Nernst equation based on the chloride concentration in aCSF.

Supplementary information

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Sato, T., Haider, B., Häusser, M. et al. An excitatory basis for divisive normalization in visual cortex. Nat Neurosci 19, 568–570 (2016). https://doi.org/10.1038/nn.4249

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nn.4249

This article is cited by

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing