@article{THESIS,
      recid = {3465},
      author = {Zayyad, Zaina Adel},
      title = {Visually-Evoked Normalization in Awake Mouse V1},
      publisher = {University of Chicago},
      school = {Ph.D.},
      address = {2021-08},
      number = {THESIS},
      pages = {109},
      abstract = {Normalization is a canonical nonlinear computation that  underlies multisensory integration, attention, and other  sensory processing. In this computation, the response of  one neuron to a set of stimuli is normalized by the weight  of all the presented stimuli – not just its preferred  stimulus. Normalization is ubiquitous and conserved across  species, suggesting its fundamental role in brain  processing. When central brain computations go awry, the  consequences can be severe – and normalization has been  implicated in brain disorders like autism. While  normalization has been described in multiple modalities and  species, its functional circuitry and mechanism remain  unknown. This project aims to produce an experimental model  to address that gap. Specifically, here, we investigate  divisive normalization in awake mouse visual cortex. This  mouse model will enable the use of powerful in vivo  genetic, electrophysiology, imaging, optogenetic, and  psychophysics techniques.Using transgenic mice, excitatory  neurons in mouse V1 were functionally labelled with  GCaMP6s. Two-photon imaging was used to capture the  activity of large V1 excitatory populations in awake mice  presented with cross-inhibitory stimuli designed to evoke  normalization. We recorded from hundreds of tuned and  untuned neurons in an unbiased manner and observed tuned  normalization similar to what has been observed in other  species. These findings were cross-validated using  electrophysiological recordings. Additionally, in our  electrophysiology data, normalization strength was observed  to have a depth dependence, suggesting the need for further  study of laminar differences in this computation. Our data  suggest that normalization can be visually evoked and  measured in the V1 of awake, head-fixed mice, opening the  door to further functional and mechanistic  study.
Furthermore, we examined how normalization  influences population pairwise noise correlations. We  observed that normalization increased the correlation of  similarly orientation-selective pairs with high  normalization strength and decreased the correlation of  oppositely orientation-selective pairs with high  normalization strength. We also observed that in mice, in  contrast with macaques, normalization decreased the  correlation of similarly orientation-selective pairs with  low normalization strength. Normalization mechanisms  underlie the changes in pairwise correlations that can  explain improved behavioral performance in attention, so  further understanding of normalization will help better  understand processes that modify pairwise correlations,  like attention.
We anticipate that this work will provide  groundwork for continuing to study normalization in the  mouse and may ultimately lead to circuit dissection and  behavioral assays, contributing to an understanding of the  role of normalization in normal physiology and behavior,  aberrant circuitry, and disease.
},
      url = {http://knowledge.uchicago.edu/record/3465},
      doi = {https://doi.org/10.6082/uchicago.3465},
}