University of Southern California Dissertations and Theses

Functional circuitry underlying cross-modality integration and the development of a novel sparse labeling technique

(USC Thesis Other)

Functional circuitry underlying cross-modality integration and the development of a novel sparse labeling technique

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
UNIVERSITY OF SOUTHERN CALIFORNIA

Functional circuitry underlying cross-modality integration and
the development of a novel sparse labeling technique

Leena Ali Ibrahim

A DISSERTATION PRESENTED TO THE FACULTY OF THE
USC GRADUATE SCHOOL IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE DOCTOR OF PHILOSOPHY
(NEUROSCIENCE)

Mentor: Dr. Huizhong Whit Tao

May 2016

© Copyright by Leena Ali Ibrahim, 2016
All rights reserved.
1

To my Parents and brother
For their love, encouragement and support

2

Acknowledgement
Firstly, I would like to express my deepest gratitude to my two advisors, Dr. Huizhong Whit Tao
and Dr. Li I Zhang, for their exceptional guidance, patience and support provided throughout my
graduate study. I could not have succeeded in accomplishing anything without them.
Secondly, I would like to thank my colleagues for their help in designing and carrying out those
challenging experiments and also for their moral support when experiments did not go as I would
have hoped for. Dr. Shengzhi Wang taught me a lot about molecular biology, and without his
help and support, I would not be where I am today. I want to thank Young Joo Kim, for his advice
on a lot of problems I encountered during molecular experiments and fun times in the lab. I
especially want to thank my colleague and friend Lukas Mesik, who has always been there for
me. He taught me a lot about Matlab programing and data analysis among many other things.
Xiaolin Chow, another very good friend and colleague has been very helpful and constructively
critical which helped me grow as an individual. I would also like to thank all my other lab mates
who have helped me with experiments and made my stay in the lab very pleasant: Lingyun Li,
Yatang Li, Ray Xiong, Qi Fang, Xuying Ji, Haifu Li, and Brian Zingg.
I would also like to thank my committee members, Dr. Karen Chang and Dr. Chien-Ping Ko for
their thoughtful suggestions and guidance throughout the qualifying and dissertation process.
Words cannot express the gratitude I have for my parents. Their love, understanding, and support
have been at the forefront of my life. I am grateful to them and my brother for always being there
for me and encouraging me to continue to strive.

3

Table of Contents

Dedication 1
Acknowledgement 2
List of Figures 6
Abstract........................................................................................................................................... 8
Chapter 1: Introduction..................................................................................................................10
1.1 Overview of the Visual System ..................................................................................... 10
1.1.1 Organization of the visual system .............................................................................. 10
1.1.2 Receptive Fields ........................................................................................................ .13
1.1.3 Orientation Selectivity ............................................................................................... 15
1.2 Inhibitory Neuron types in the Cortex ........................................................................... 17
1.3 Long Range Projections to V1 ....................................................................................... 22
1.3.1 Role of long range projections and their importance in visual processing ................ 22
1.3.2 New tools to study specific projections in the brain .................................................. 23
1.4 Overview of the Auditory System ................................................................................. 25
1.4.1 Auditory pathway....................................................................................................... 25
1.4.2 Overview of the cochlea and it’s connectivity ........................................................... 27
1.5 Sparse Labeling .............................................................................................................. 28
1.5.1 Introduction to single cell labeling ............................................................................ 28
1.5.2 Introduction to STARS .............................................................................................. 30
1.6 Specific Aims ................................................................................................................. 31
1.6.1 Specific Aim 1 ........................................................................................................... 31
1.6.2 Specific Aim 2 ........................................................................................................... 32

4

Chapter 2: Cross Modality Sharpening of visual cortical processing............................................33
2.1 Introduction ................................................................................................................... 33
2.2 Materials and Methods ................................................................................................... 35
2.2.1 Viral Injection ............................................................................................................. 35
2.2.2 Retrograde Tracer Injection ........................................................................................ 36
2.2.3 Animal surgery for in vivo recording in anaesthetized mice ...................................... 37
2.2.4 Animal surgery for in vivo recording in awake mice .................................................. 37
2.2.5 In Vivo Electrophysiology........................................................................................... 38
2.2.6 Visual Stimulation ...................................................................................................... 39
2.2.7 Auditory Stimulation ................................................................................................. 40
2.2.8 In Vivo optogenetic manipulation ............................................................................... 41
2.2.9 In vitro Electrophysiology .......................................................................................... 41
2.2.10 In vivo Two-Photon Calcium Imaging ........................................................................ 43
2.2.10 Data Analysis .............................................................................................................. 44
2.2.11 Statistical Analysis ...................................................................................................... 45
2.3 Results ............................................................................................................................ 46
2.3.1 Effect of Sound stimilation on excitatory neuron responses in V1 ............................. 46
2.3.2 V1 receives direct input from A1................................................................................. 53
2.3.3 Optogenetically stimulating A1 axons sharpens OS of L2/3 excitatory neurons ...... ..53
2.3.4 Innervation pattern of the A1-V1 projection ............................................................. ..59
2.3.5 Activation of L1 neurons by sound ............................................................................ ..62
2.3.6 Optogenetic suppression of L1 neuron activity. ....................................................... ...63
2.3.7 Effect of Sound stimilation on L2/3 PV, SOM and VIP neurons ................................ 68
2.4 Discussion. .................................................................................................................... ..71
2.4.1 Modulation of V1 processing by auditory input through a top down circuit............... 72
2.4.2 Sound effect on L1 neuron responses .......................................................................... 74
2.4.3 L2/3 inhibitory neuron subtypes .................................................................................. 74

5

Chapter 3: A Genetic Strategy for Stochastic Gene Activation with Regulated Sparseness.........77
3.1 Introduction ..................................................................................................................... 77
3.1.1 Introcution to STARS ................................................................................................ 78
3.2 Materials and Methods .................................................................................................... 80
3.2.1 Construction of STARS transgene .............................................................................. 80
3.2.2 Generation of STARS transgenic mice ....................................................................... 81
3.2.3 Generating STARS::X-Cre lines................................................................................. 86
3.2.4 Histology ..................................................................................................................... 87
3.2.5 Sound deprivation ....................................................................................................... 87
3.2.6 Image analysis and Quantification .............................................................................. 88
3.3 Results ............................................................................................................................ 88
3.3.1 Design and Construction of STARS ........................................................................... 88
3.3.2 Quantification of sparsness in the brain ...................................................................... 90
3.3.3 Quantification of sparsness in the cochlea .................................................................. 94
3.3.4 Tracing type II SPG neurons in the cochlea ............................................................... 96
3.4 Discussion and Future directions ................................................................................ 100

References................................................................................................................................... 103

6

List of Figures

Figure 1: Schematic of the Visual Pathway…………………………………………………….. 12
Figure 2: Receptive fields of visual neurons ................................................................................. 14
Figure 3: Orientation Selectivity in V1. ........................................................................................ 16
Figure 4: Inhibitory Neuron types................................................................................................. 18
Figure 5: The two types of Layer 1 neurons ................................................................................. 22
Figure 6: Optogenetics .................................................................................................................. 24
Figure 7: Schematic of the Auditory Pathway .............................................................................. 26
Figure 8: Sound induced sharpening of Orientation Selectivity (OS) in V1 ................................ 48
Figure 9: Effect of Sound on Spontaneous firing rate in Layer 2/3 of V1.................................... 49
Figure 10: Sound mediated sharpening of OS in Awake Mice .................................................... 51
Figure 11: Sound induced sharpening of OS revealed by Ca
2+
imaging ...................................... 52
Figure 12: Optogenetic Activation of A1 axons in V1.. ............................................................... 56
Figure 13: Comparison of axon densities between A1-V1 vs A1-IC projections. ....................... 57
Figure 14: Comparison between changes observed at low contrast vs high contrast. .................. 57
Figure 15: Plots of OSI measurement in different cell types ........................................................ 58
Figure 16: Correlation between changes in gOSI vs initial gOSI ................................................. 59
Figure 17: Layer 1 neurons receive maximum A1 input .............................................................. 61
Figure 18: Visual responses of Layer 1 neurons are enhanced by Sound. ................................... 64
Figure 19:Response of L1 neurons to Sound alone .......................................................................65
Figure 20: Effect of Sound intensities on layer 1 neurons and contrast dependence. .................. 66
7

Figure 21: Inhibition of Layer 1 neurons reduces the sound induced sharpening………………..65
Figure 22: Expression of GCaMP6s in various Cre::tdTomato mice…………………………….69
Figure 23: Effect of Sound stimulation on PV, SOM and VIP neurons in Layer 2/3…………….70
Figure 24: In vivo loose-patch recording from fast spiking neurons in Layer 2/3………………..71
Figure 25: Proposed circuitry mechanish for sound mediated effect in V1………………………73
Figure 26: The STARS Strategy……………………………………………………………….....80
Figure 27: Schematic Illustration of TARGATT-pBT378-STARS construct………………........83
Figure 28: Identification of STARS founders…………………………………………………….86
Figure 29: Schematic representation of the STARS construct and test in ES cells………………90
Figure 30: Quantification of sparseness in the brain……………………………………………...93
Figure 31: Quantification of sparseness in the cochlea…………………………………………...95
Figure 32: Row dependent innervation of Type II Spiral ganglion (SPG) neurons………………98
Figure 33: Analysis of innervation at P14 under normal and sound deprived conditions………..99
Figure 34: STARS2.0 Strategy…………………………………………………………………..102

8

Abstract
Animals and humans alike constantly receive a multitude of sensory inputs from their
environment. All of these inputs are processed and integrated in a very precise manner in the
brain, which ultimately leads to appropriate behaviors and survival. Initially, it was thought that
all the sensory modalities operate in isolation; i.e. they are unimodal. An auditory stimulus will
produce a response in the auditory cortex and a visual stimulus will elicit a response in the visual
cortex and that their integration occurs in higher association areas in the brain. This view has been
challenged recently and a number of recent studies have suggested that sensory cortical regions
are in fact multimodal and have the ability to integrate multiple sources of sensory inputs.
In the first part my thesis, I want to explore how sensory processing in cortex is cross-modally
modulated and what the underlying neural circuits are. In mouse primary visual cortex (V1), we
found that orientation selectivity of layer 2/3 but not layer 4 excitatory neurons was sharpened in
the presence of sound or optogenetic activation of projections from primary auditory cortex (A1)
to V1. The effect was manifested by decreased average visual responses but an increase in
response at preferred orientation. It was more pronounced under lower visual contrast, and
diminished by suppressing layer1 activity. Layer1 neurons, being strongly innervated by A1-V1
axons, exhibited elevated responses to visual stimuli under sound stimulation, while layer2/3
vasoactive intestinal peptide (VIP) neurons exhibited suppressed responses, both preferentially at
preferred orientation. Our data suggest that the cross-modality modulation is achieved primarily
through layer1 neuron and layer2/3 VIP cell mediated inhibitory and disinhibitory circuits.
The second part of my thesis will describe a novel technique, termed STARS (Stochastic Gene
Activation with Regulated Sparseness). We developed a transgenic reporter mouse line that can
9

genetically label cells sparsely in a Cre dependent manner. The number of labeled neurons is
estimated to be 8-15% of those labeled by crossing with a traditional reporter line Ai14 (Cre
dependent tdTomato expression). Genetically labeling a sparse population of different neuron
types specified by the Cre driver has enormous implications. Using this genetic labeling
technique, we are able to study the morphology and connectivity of a set number of neurons. We
are also expanding its application to include functional manipulation as well. To demonstrate the
use of this transgenic mouse, we studied the connectivity of Type II spiral ganglion neurons in the
cochlea. The cochlea is a very elegant peripheral sensory organ with precise connectivity
between spiral ganglion neurons and hair cells. We found that the synaptic contacts the spiral
ganglion neurons make onto the outer hair cells are in fact different than previously thought. At
P0, there was a row dependent length of innervation by the Type II spiral ganglion neurons, with
no discernable synaptic formations onto the OHCs. At P14 however, we are able to see synaptic
contacts being made onto the hair cells. Interestingly, at that age, the row dependent innervation
disappeared; however, each spiral ganglion neuron could innervate 7-10 OHCs in multiple rows.

Taken together, understanding precise connectivity in the nervous system is essential to
understanding the function of specific neural circuits governing sensory processing.

10

CHAPTER 1
Introduction
1.1 Overview of the Visual System
1.1.1 Organization of the visual system
One of the most important senses is that of vision. Visual information is used by many
animals to seek food, avoid danger and survive. Visual perception is conveyed by the visual
system, which consists of three major components; the eye, the lateral geniculate nucleus
(LGN) and the visual cortex, which are connected and serially process information the
incoming visual information (Figure 1). The eye, which is the first segment of the system,
consists of a variety of structures. The first of which is the lens, which projects objects from
the outside world onto the retina, which is the first sensory tissue in the visual pathway. The
retina consists of cells which transform the physical energy of light into chemical and
electrical signals that are then relayed into the brain.

The retina is composed of three layers of cells: outer nuclear layer, inner nuclear layer and
ganglion cell layer. The outer nuclear layer consists of photoreceptor, rods and cones cells,
which are the main sensory cells that absorb photons by a channel protein called rhodopsin.
The activation of the channel hyperpolarizes the membrane potential of the cell through a
series of intracellular signaling cascades and reduces the amount of glutamate release. Rods
take charge of vision at low light levels (scotopic vision), do not mediate color vision, and
have low spatial acuity, while cones are active at high light levels (photopic vision),
11

responsible for color vision, and have high spatial acuity. The inner nuclear layer is made of
bipolar, horizontal and amacrine cells and mediates the information between the outer nuclear
layer and ganglion cell layer, which carry out the first steps in information processing. The
ganglion cell layer, mainly composed of retinal ganglion cells, turns the visual input into
electrical signals in the form of action potentials and relays the visual information along their
axons or optical nerves to the next stage, the lateral geniculate nucleus (LGN) in the thalamus.
These three layers of cells form a network that performs the first stage of neural computation.
They have the ability to detect the local changes in brightness of the visual stimuli but are not
able to detect edges and orientation (Tessier-Lavigne, 2000).

The LGN is the primary center for visual information processing in the thalamus. It is a
laminated structure with cell body and neuropil layers alternating. In humans and macaques
the LGN is characterized by six distinctive layers (Layers 1 to 6). Layers 1 and 2, also called
the magnocellular layers, receive information from rod cells and are necessary for the
perception of form, movement, depth, while layers 3-6 or parvocellular layers receive
information from cone cells and are responsible for the perception of color and fine details.
Cells in the LGN inherit the properties of retinal ganglion cells and respond well to local light
intensity change but not to elongated stimuli (Wurtz and Kandel, 2000). When the visual
information travels to the visual cortex, visual features such as orientation selectivity,
binocular vision and depth begin to emerge. These tasks are performed by a neuronal network
of six layers, which are connected in a way so that visual information flows in a more or less
specified order. In cats and monkey, V1 neurons are well-organized horizontally into
functional columns like orientation and ocular dominance columns. The horizontal
12

organization is also reflected in the retinotopic map that represents the outside visual world.
(Wurtz and Kandel, 2000).

Figure 1. The Visual Pathway. Light, in the form of electromagnetic waves enters the eyes, and then
through the retina, the signal is conveyed to the LGN in the thalamus. The retinal ganglion cell fibers cross
at the optic chiasm to reach the contralateral hemisphere. From the LGN, the information goes to the first
stage of cortical processing, in the primary visual cortex (V1)
13

1.1.2 Receptive Fields
A visual neuron only responds to stimuli that are presented within a certain area in space, called
the visual receptive field (RF). Receptive fields exist because each individual visual neuron
receives convergent inputs only from a limited number of photo receptor cells that get activated.
The first cells in the visual pathway whose receptive fields were examined are the retinal ganglion
cells. Extracellular studies have demonstrated that the RFs of ganglion cells are characterized by
concentric center-surround organization (KUFFLER, 1953; Kuffler, 1973). According to the
response polarity to small bright or dark stimuli presented on the center of RFs, ganglion cells
could be divided into two groups, the On-center ganglion cells and the Off-center ganglion cells.
In the case of the On-center cells, a bright spot shining in the center of their RFs increases their
firing rates while a bright spot illuminating the periphery reduces or suppresses their firing rates.
A dark stimulus has an opposite effect; stimulation in the center suppresses their firing while
stimulation in the periphery activates the cells. On the other hand, Off-center cells respond to
local intensity change in a reverse way that light on the center decrease the firing rate while dark
on the periphery elevates firing. Because the center-surround arrangement of these RFs are
symmetric, retinal ganglion neurons respond well to local light intensity change, but not to the
change of the diffuse light or the orientation of linear stimulation like a bright or dark bar. RF
properties of LGN relay neurons are inherited from retinal ganglion cells and those cells also
respond strongly to local intensity change, but not to long oriented bars, which are characteristic
of V1 neurons (Figure 2). (HUBEL and WIESEL, 1962; Levick et al., 1972)

14

Figure 2 . Receptive fields of simple cells in the V1 are different from those of the neurons in the retina and LGN.
(A) On-center or Off-center receptive fields of retinal ganglion cells or LGN relay neurons. Note that they have center-
surround concentric configuration. (B) The receptive fields of simple cells in the V1 have narrow elongated zones that are
spatially segregated with each other. ‘+’ for excitation; ‘–’ for inhibition. (Adapted from Wurtz and Kandel, 2004)
15

1.1.3 Orientation Selectivity
Hubel and Wiesel originally demonstrated the phenomenon of orientation selectivity in their
Nobel Prize-winning work in 1962. When a bar or an edge is flashed or moved across a cortical
neuron’s receptive field at different orientations moving in either direction, a particular neuron is
most sensitive and thus fires very strongly to a very specific orientation. That orientation is called
the neuron’s preferred orientation. The orientation which is orthogonal to the preferred orientation
is known as the orthogonal orientation. When the stimulus is presented within the same region of
visual space but at the orthogonal orientation, the neuron, in most cases, does not respond at all.
In cat visual cortex, a vast majority of neurons are orientation selective (Figure 3). In primate
visual cortex, all neurons outside layer 4c (the input layer from the thalamus) are orientation
selective. In mice as well, many neurons in layer 4 and layer 2/3 are orientation selective.
Different neurons have different preferred orientations, so that some subset of cortical neurons
will respond to a stimulus of any orientation and position. In mice, neurons of different
orientations are arranged in a random, salt and pepper like fashion unlike cats and primates which
have discernable orientation columns. The selectivity of each neuron is roughly Gaussian; i.e. the
responses against stimulus orientation form an approximately Gaussian-shaped curve (Figure 3).

16

Figure 3. Orientation Selectivity in V1. Upper panel: The responses (right) of a given neuron in V1 to moving
bars oriented in different orientations (left). The responses are plotted against orientations presented and fit with a
gaussian. Lower panel: The Gaussian fit of neurons showing the highest response to be the preferred orientation
of the cell, while, the lowest is the orthogonal orientation of the same neuron.
17

1.2 Inhibitory Neuron subtypes in the Cortex
Inhibitory neurons account for only about 10-20% of the neurons found in rodent neocortex.
Nevertheless, the inhibition they provide appears to play an important role in the function of the
circuits, as at the synaptic level, pyramidal cells receive a fairly even mix of excitatory and
inhibitory inputs. In fact, developmental maturation of the cortical circuitry strictly requires
inhibitory neurons, suggesting they are indispensable for normal function (Hensch, 2005). Despite
their smaller population size, inhibitory neurons are strikingly diverse: they come with different
morphologies, connectivity patterns, intrinsic excitability properties and protein expression levels
(Markram et al., 2004). In recent years there has been a lot of effort to classify them and find
molecular markers that consistently label subgroups based on well characterized properties
(Ascoli et al., 2008). Currently, it seems that inhibitory interneurons can be divided into three
more or less mutually exclusive groups (Rudy et al., 2011) (Figure 4):
1. Parvalbumin (PV) positive neurons, which are fast spiking and with basket cell or chandelier
cell morphology.
2. Somatostatin (SOM) positive neurons, mostly corresponding to Martinotti cells.
3. 5HT3aR expressing neurons. This is the most recently characterized group and includes two
major subgroups. The first one is the neurons expressing the vasoactive intestinal peptide (VIP),
corresponding to bipolar cells and small basket cells. The second one is Reelin expressing
neurons found mostly in layer 1 and including the neurogliaform cells. Developmentally, 5HT3aR
neurons originate from the Caudal Ganglionic Eminence while PV and SOM neurons originate
from the Medial Ganglionic Eminence (Lee et al., 2010; Marín, 2013).
18

Having inhibitory interneurons classified based on their molecular markers is very useful.
Construction of corresponding Cre lines enabled targeting, imaging and optogenetic manipulation
of these cells in vitro and in vivo, bringing us closer to understanding their role in cortical circuits.

Figure 4. Inhibitory neurons. Different inhibitory interneuron morphologies along with their corresponding
molecular markers. The three major types of inhibitory neurons are shown: PV,SOM and 5HT3aR+ (VIP and
Reeling). (Adapted from Markram et al 2004)
19

Parvalbumin interneurons
PV neurons constitute about 40% of all inhibitory neurons and are found in layers 2/3 through 6.
Almost all of them are fast spiking, owing to the fast Kv3 channels granting them a very short
recovery after spike (Rudy and McBain, 2001). Most of the PV neurons are basket cells. These
receive inputs from multiple sources, although how the inputs differ from layer to layer has not
been systematically investigated. In layer 4, PV neurons are known to receive direct thalamic
excitation, providing a fast feedforward inhibition to other layer 4 neurons.
Axonal arbors of basket cells extend horizontally within a layer. Since PV neurons primarily
target the soma and proximal dendrites, the axonal organization implies that they mostly inhibit
neurons originating within the same layer. In fact, it appears that the PV neurons inhibit
pyramidal neurons within their axonal arbor without any specificity. Packer and Yuste (Packer
and Yuste, 2011) showed that PV neurons made connection to 40-70% of neurons within 200
microns horizontally from their soma, with connection probability depending on the layer and
cortical area. This pattern of connectivity would seem better suited for general modulation of
activity level rather than for fine low level computations.

Somatostatin interneurons
While somatostatin is occasionally found in other cell types, it is largely associated with the
Martinotti cells (Figure 1), which always express SOM and never express PV or VIP. SOM
interneurons are found in layers 2-6 and their proportion increases with depth. Their population
ranges from 25% of inhibitory neurons in layer 2/3 up to 50% of inhibitory neurons in layer 6. In
20

layer 2/3, Martinotti cells receive most of their excitatory input from layer 2/3 and layer 4,
although for those containing calretinin the layer 2/3 contribution is much bigger (Xu and
Callaway, 2009). Where Martinotti cells differ from PV cells much more strikingly is their axonal
arborization. They have a fairly elaborate vertical organization, including a projection all the way
to layer 1 where their axons spread horizontally across large distances and inhibit tuft dendrites
(Markram et al., 2004). SOM axons in general seem to preferentially target distal dendrites,
although somas and proximal dendrites are also targeted. Functionally, SOM neurons are not very
well characterized. They are generally much less active than PV neurons (Ma et al., 2010) but the
activity level seems to be modulated by behavioral state of the animal (Fanselow et al., 2008;
Gentet, 2012). Specific functional roles suggested for SOM neurons include regulation of burst
firing (Gentet, 2012) and subtractive inhibition (Wilson et al., 2012).

VIP interneurons
The 5HT3aR neurons have been identified as a distinct group only recently, but virtually all of the
neurons belonging to this group have been studied before to some extent. Most of the neurons
expressing VIP, Reelin, Neuropeptide Y and Cholecystokinin belong under the 5HT3aR umbrella
(Lee et al., 2010). Of these, reelin positive cells appear to be the biggest subgroup followed by
VIP. VIP positive cells constitute only ~12% of all inhibitory interneurons but this number is
about double in layer 2/3 where they are most numerous. They include several morphological
types (Figure 1) although the association with bipolar cells and small basket cells is most
common. Not much is known about the connectivity of VIP neurons. In an early study (Morrison
et al., 1984), most extensive dendritic branching was found in layer 1 and on the boundary
21

between layers 4 and 5. The density of axonal varicosities was highest in layers 2 to 4. According
to David et al. 2007, all VIP neurons receive inputs from PV neurons and 94% of PV neurons are
innervated by VIP neurons. Recently it has been shown that VIP neurons are at the center of
several disinhibitory circuits. For example, during locomotion, VIP neurons in the visual cortex
were activated, which in turn inhibited SOM neurons (with which they are specifically
connected). The inhibition of SOM neurons in turn relieved the inhibition onto excitatory
neurons, thereby activating them (Fu et al., 2014). Several other studies have also shown VIP
neurons being involved in disinhibition (Lee et al., 2013; Zhang et al., 2014).

Layer 1 neurons
In the most superficial layer of the cortex (layer 1), all neurons are inhibitory in nature. They are
very sparse in number and the majority of which are Reelin positive. As mentioned earlier, all
layer 1 neurons and VIP neurons in layer 1 and layer 2/3 are 5HT3aR positive. In terms of
morphology, these neurons are classified either as elongated neurogliaform cells (eNGFs) or
single bouquet cells (SBCs) (Figure 5). The eNGFs mainly have axonal and dendritic arborization
localized to layer 1 spreading horizontally across the layer. The SBCs on the other hand, have a
bipolar morphology, with their axons extending all the way to layer 5. Layer 1 inhibitory neurons
are the least understood type of neurons. This stems from the fact that there were no Cre mouse
lines specific for layer 1 neurons. Layer 1 neurons also receive direct thalamic inputs and
therefore have strong visual responses. Additionally, layer 1 is a hub of inputs coming from
various far away regions in the brain. Whether or not, layer 1 neurons are directly innervated by
22

long range projections and what their role is in integrating these inputs is a very interesting
question.

1.3 Long range projections to V1
1.3.1 Role of long range projections and their importance in visual
processing
Information processing in a primary cortical area (such as V1) is determined by many factors,
incoming sensory experience, higher cortical feedback and attention. While the incoming
sensory experience is relayed via the thalamus, feedback is usually cortico-cortical or could
involve inputs from higher order thalamic nuclei. Long range projections to V1 also include
inputs from neuromodulatory systems, such as cholinergic, from the basal forebrain. These
projections have been implicated in a variety of functions such as memory, attention as well
as modulation of sensory processing (Alitto and Dan, 2012; Hasselmo and Sarter, 2011;
Kilgard and Merzenich, 1998; Robbins et al., 1997). In addition to sub cortical inputs
modulating sensory processing, cortico-cortical projections have recently been shown to be
Figure 5. The two major types of Layer 1 neurons. Adapted from Rudy et al, 2011)
23

implicated in surround suppression and attentional mechanisms. A recent study has shown
that V1 receives inputs from a higher frontal brain region, the anterior cingulate cortex.
Activating this projection in the mouse was found to enhance the visual activity at the
activated site, while suppressing responses in the surrounding region, a mechanism similar to
selective attentional modulation (Zhang et al., 2014).

Primary cortical areas have been thought to be strictly uni-sensory; a visual stimulation will
only elicit a response in the visual cortex, for example. Recently however, some extracellular
recordings reveal interactions, thought to be inhibitory in nature, among different primary
sensory cortices, challenging the strict uni-sensory view. What the role of these interactions is
in sensory processing, where they originate and how they contribute to altering unique
properties of neurons in a specific sensory region remain largely unknown. We have observed
direct projections from the primary auditory cortex (A1) to V1. In the first part of the thesis, I
will be focusing on this projection to reveal its importance in modulating visual cortical
processing through this long range projection.

1.3.2 New tools to study role of specific projections in the brain
Recent advances in optogenetics have made it possible to activate or inhibit specific pathways
or specific cell types. Channelrhodopsin, a non-selective cation channel, when expressed in
specific neurons in a Cre dependent manner, could be activated by blue light (Boyden et al.,
2005). This activation in turn activates the neuron by causing fast depolarization. Similarly,
halorhodopsin or Arch, could be specifically expressed in neurons to inhibit them when green
24

light is shone on them. While halorhodopsin is a chloride ion channel, Arch is a proton pump.
Both cause hyperpolarization in the neurons when activated (Figure 6).

Additionally, recent mapping of connections across the brain (Oh et al., 2014; Zingg et al., 2014)
has provided a huge and important database of connectivity across the mouse brain. These
projections were mapped using anterograde and retrograde tracers to reveal input and output
targets of different neuron types in specific regions in the brain. Expression of ChR2 and NpHR
in specifically mapped projections and their activation by the corresponding wavelength of light
Figure 6. Optogenetics. Left panel shows the ion channel channelrhodopsin (Chr2). When activated by blue light
(470nm) it activates the neuron in which it is expressed. Right panel: Expression and activation of Halorhodopsin
or Arch by green light inhibits the neurons. Adapted from Nature Communications
25

has shed light on the function and the importance of specific pathways as well the role of specific
cell types in sensory processing, among many other brain functions.

1.4 Overview of the auditory system
1.4.1 Auditory Pathway
Another very important sense is that of audition. Auditory information in the form of mechanical
sound waves enters the brain via the ear. The ear consists of three parts; outer, middle and inner
ear. While the outer and middle ear are important in conducting the sound through the bony
structures, the inner ear is the main organ responsible for transforming the mechanical energy into
electrical energy which can then be propagated along the auditory nerve via the bipolar spiral
ganglion neurons (SGN) to the brain. The cochlea, a critical structure of the inner ear is where this
mechanotransduction takes place. The hair cells in the cochlea are the primary sensory cells that
accomplish this task. Once the sound has been converted to electrical energy, the sound
information is sent to the next stage of auditory processing to the cochlear nucleus in the brain
stem. After several more relays, the sound information eventually reaches the primary auditory
cortex where further processing of the sound takes place (Figure 7).

26

Figure 7. The Auditory Pathway. Sound in the form of mechanical waves enters the ears. The hair cells in the
cochlea (inner ear structure) convert the mechanical vibrations of the sound into electrical signals which are
propagated by the spiral ganglion neurons through the auditory nerve. The information is sent to the cochlear
nucleus in the brainstem, then to the inferior colliculus, to the MGB in the thalamus and eventually to the cortex
for further processing.
27

1.4.2 Overview of the cochlea and it’s connectivity
Different aspects of sound information, such frequency, intensity and timing is first decoded by
the very specialized type of epithelial cells known as hair cells that are located in the inner ear.
The information is subsequently relayed to the central nervous system for further processing via
the bipolar spiral ganglion neurons (SGN). Each spiral ganglion neuron has one peripheral
projection that innervates hair cells through a special type of synapse known as the ribbon
synapses, and one central process to the cochlear nucleus in the brainstem. This flow of auditory
information from hair cells to the spiral ganglion neurons is indispensable for our sense of
hearing. Many hearing loss disorders are in fact known to be sensorineural deafness, which in
humans occur due to the loss of either hair cells or the spiral ganglion neurons. The only available
treatment for such sensorineuronal hearing loss is cochlear implants, which basically directly
stimulate the spiral ganglion neuron fibers via electric stimuli. Therefore it is pertinent to
understand the precise connectivity between hair cells and spiral ganglion neurons. Mammals are
capable of hearing sounds of frequencies ranging from 20 Hz to 20 kHz and detect small
differences in intensities. They can respond to sound induced vibration at atomic scales and are
able to amplify the signal over 100 fold (Gillespie et al., 2005; Müller, 2008; Vollrath et al.,
2007). How is such sensitivity and robustness achieved by the auditory system? Remarkably, the
peripheral auditory system has evolved into an elegant structure that is tailored to achieve this
daunting task. The cochlea epithelium extends throughout the cochlear duct along its baso-apical
axis. Depending on location along this axis, individual hair cells respond to a narrow range of
sound frequencies. Hair cells that are located in the basal part of the cochlea with wider open
space will respond to high frequency and those located in apical region respond to low frequency
sounds. This arrangement gives rise to tonotopy along the longitudinal (base to apex) axis. There
28

are two types of hair cells that make up the sensory epithelium: one row of inner hair cells (IHCs)
and three rows of outer hair cells (OHCs). IHCs are the believed to be the primary receptors for
sound perception, with the OHCs on the other hand to be amplifiers that increase the sound
detection sensitivity (Dallos et al., 2008; Liberman et al., 2002). Spiral ganglion neurons (SPG)
are also divided into two types. The majority (more than 90% percent) are the type I SPG
neurons, which extend unbranched fiber to make contact with one inner hair cell only; the
remaining 10% are the type II SPG neuron. These extend their processes bypassing the inner hair
cell layer and turn towards the base, making contacts with multiple outer hair cells along their
way (Appler and Goodrich, 2011; Meyer et al., 2009; Rubel and Fritzsch, 2002). A single inner
hair cell is connected with 10 to 20 type I spiral ganglion neurons (Meyer et al., 2009; Rubel and
Fritzsch, 2002), possibly to increase the probability to detect soft sound (Meyer et al., 2009).
Inversely, a single type I SPG neuron connects with only one IHC. In contrast, a single type II
SPG neuron has the ability to innervate multiple OHCs.

1.5 Sparse Labeling
1.5.1 Introduction to Single Cell Labeling
The Cre/lox recombination system has provided a powerful tool for gene expression control in
specific cell types and regions. Cre recombinase, which is derived from the bacteriophage P1 is a
member of λ integrase superfamily. It has the ability to specifically recognize identical pairs of
lox site, which are nothing but two 14 base pair inverted repeats separated by an 8 base pair
spacer which is assymetric, the latter dictating the direction of the lox site (Sauer, 1993). The
29

catalytic actions of Cre recombinase include excision, inversion, translocation and integration
events depending on the location and orientation of the lox site. Because of its ability to transform
the DNA in such diverse ways it has been widely applied in genetic analysis (Lewandoski, 2001;
Luo, 2007; Luo et al., 2008; Schnütgen et al., 2003). An essential part of a gene can be flanked by
a pair of lox sites in order to achieve conditional control of gene. And essentially the function of
the gene would not be affected when no Cre is present; i.e. without recombination to delete the
flanked part of the gene.
Transgenic mice can be generated in which lox sites flank a gene of interest. When the mice are
crossed to a tissue or cell type specific Cre mouse line, Cre recombinase activity will excise the
‘floxed’ sequence and lead to inactivation of the allele with spatial and temporal accuracy that is
dictated by the promoter driving Cre expression (Lewandoski, 2001). Alternatively, floxed
“STOP” elements (usually a multi-polyA sequence) between the promoter and the coding
sequence of a gene could be added so that the gene expression is inactive unless it is crossed to a
specific Cre line to kick out the STOP signal and thus activate the gene that is present after the
STOP (Dragatsis et al., 2000).
The nervous system consists of many different types of neurons with various morphologies,
molecular profiles and electrophysiological properties. In order to understand how individual
neurons connect to other neurons, both locally and over long distances, it is important to be able
to visualize individual neurons and perturb their activity in a sparse and specific manner. Early in
the century, Ramon y Cajal used Golgi staining to reveal specific axonal and dendritic
morphologies of many cell types in the brain. This was very helpful in revealing the intricate
diversity of neurons in the brain. However, the method lacked specificity for the cell type being
analyzed, and it was random in that any region could be labeled by chance. Many approaches
30

(Introduced in Chapter 3) have recently been developed to circumvent these problems but they
have certain limitations. One limitation is their ability to achieve genetic “Golgi” staining as well
as genetic manipulation in a small population of cells in nervous tissue in an efficient and precise
manner.

1.5.2 Introduction to STARS
STARS, or Stochastic gene Activation with Regulated Sparseness, can provide a straightforward
way to achieve genetic “Golgi” staining as well as genetic manipulations in the same cells
labeled. Using STARS, a fluorescent protein marker such as EGFP, or any gene of interest, could
be expressed sparsely in a set number of to achieve genetic “Golgi” staining when these STARS
transgenic mice are crossed to tissue or cell type specific Cre lines. Previously we showed that it
is possible to regulate the Cre-lox reaction kinetics by varying the spacer distance between
identical pairs of lox sites (Wang et al., 2009). The larger the distance, the more time it takes for
the sites to be brought together and thus we can control the probability of a certain gene of
interest being expressed. A variety of questions could be addressed using such a genetic system.
Firstly, the morphology of molecularly specified neurons could be revealed in great detail.
Secondly, axonal spans of individual projection neurons could be traced in order to anatomically
visualize where and to what extent single neurons innervate their targets. Additionally,
perturbations and manipulations could be applied to a sparse number of neurons, so as to not
disrupt the whole circuit. Additionally, the molecular and cellular mechanisms underlying neuron
circuit development, tracing cell lineage and analyzing the cell autonomous function of a gene
31

could be also be studied. The strategy which STARS employs to achieve genetically controlled
sparse labeling is addressed in detail in chapter 3.

1.6 Specific Aims
1.6.1 Specific Aim 1
In the first part of the thesis, I want to explore how a long range projection, such as from the
auditory cortex, influences properties of neurons in the primary visual cortex V1. Primary
cortical areas have been thought to be strictly uni-sensory; a visual stimulation will only elicit a
response in the visual cortex, for example. Recently however, some extracellular recordings
reveal interactions, thought to be inhibitory in nature, among different primary sensory cortices,
challenging the strict uni-sensory view. What the role of these interactions are in sensory
processing, where they originate and how they contribute to altering unique properties of neurons
in a specific sensory region remain largely unknown. We want to address this question by
activating specific connections originating from layer 5 of the primary auditory cortex that project
to the superficial layers of the primary visual cortex. Using layer 5 specific Cre mouse lines and
optogenetics to activate auditory fibers, we want to test the hypothesis that layer 5 auditory
projections alter the orientation tuning of cells in the primary visual cortex. Additionally, we want
to use sound and see whether the properties of V1 neurons, such as orientation selectivity get
altered. And if so, what circuitry mechanism contributes to such an effect.

32

1.6.2 Specific Aim 2
In the second part of my thesis, I want to discuss our efforts in generating the STARS transgenic
mouse line. A genetic system that can precisely label a set number of neurons and achieve a
genetic ‘Golgi’ like staining in a conditional Cre dependent manner is needed. We want to test
whether we can indeed use this strategy to generate a transgenic reporter mouse line in which we
can express a particular gene of interest in a molecularly defined population of neurons. Using
this mouse line we can then trace the projections of individual type II spiral ganglion neurons in
the cochlea and reveal the precise connectivity pattern between the SPGs and outer hair cells.

33

CHAPTER 2
Cross-modality Sharpening of Visual Cortical Processing

2.1 Introduction
Animals constantly receive a multitude of sensory inputs, such as visual, auditory and
somatosensory information, from the outside world, which are eventually processed in the cortex.
The precise processing of sensory input from each of these modalities is essential for an animal’s
perception, behavior and survival. Previously, it has been thought that information of different
sensory modalities is initially processed in anatomically isolated primary sensory areas, and that
merging of different sensory information takes place only in higher association areas (Felleman
and Van Essen, 1991; Jones and Powell, 1970). This view mainly stems from lesion studies
showing that deficits in a specific primary sensory cortex only affect the perception of the
corresponding modality (Dewson et al., 1970; Winans, 1967). Earlier anatomical studies have not
been particularly sensitive in revealing cross-modality projections (KUYPERS et al., 1965).
The view has been challenged more recently, as a number of functional magnetic resonance
imaging (fMRI) studies have shown that cross-modality interactions may occur at the level of
primary sensory cortices (Clavagnier et al., 2004; Kayser et al., 2005; Kok et al., 2012). For
example, it has been found that the spatial pattern of activation of a primary sensory cortical area
(for example V1) is different when stimuli from a different modality (auditory or tactile) are
presented (Liang et al., 2013). This finding suggests that there may be discrete locations within
34

V1 that respond to a specific cross-modal input. Furthermore, an increasing number of animal
studies using viral tracing in both anterograde and retrograde manners have begun to reveal
patterns of potential cross-modality interactions (Campi et al., 2010; Cappe and Barone, 2005;
Falchier et al., 2002; Frostig et al., 2008; Sieben et al., 2013; Stehberg et al., 2014; Zingg et al.,
2014). In humans, psychophysics studies have shown that combining of sensory information of
different modalities is important for speeding reaction times (Gielen et al., 1983; HERSHENSON,
1962) and for detecting indistinct stimuli (Driver and Spence, 1998; Frens et al., 1995; Jaekl and
Harris, 2009; McDonald et al., 2000; Vroomen and de Gelder, 2000). A recent behavioral study
in rats has also shown that performance success rate is enhanced and reaction time is faster when
visual targets are accompanied by a simultaneous sound (Gleiss and Kayser, 2012).
While evidence is now emerging to increasingly support occurrence of multisensory integration in
primary sensory cortices (Iurilli et al., 2012), the role of cross-modality inputs in modulating
sensory processing properties of cortical neurons and the underlying neural circuits remain not
well understood. In this study, we found that in the presence of sound or optogenetic activation of
projections from the primary auditory cortex (A1) to V1, the orientation tuning of layer (L) 2/3
pyramidal neurons was sharpened. This resulted from an increase of firing rate at the preferred
orientation in accompany with a general reduction of average visual responses. Recordings from
brain slices revealed that L1 inhibitory neurons received the maximum amount of direct input
from A1 as compared with other cell types across all the cortical layers, including pyramidal
neurons, parvalbumin (PV), somatostatin (SOM), and vasoactive intestine peptide (VIP)
inhibitory neurons. Further studies on the visual response properties of different inhibitory
neurons in superficial layers of V1 in the presence of sound suggested complex local inhibitory
35

and disinhibitory circuits that could underlie the sound induced cross-modality modulation in the
visual cortex.

2.2 Materials and Methods
All experimental procedures used in this study were approved by the Animal Care and Use
Committee of USC. In all experiments, adult (45–60 days) female mice were used.

2.2.1 Viral injection
For L5-specific expression, Rbp4-Cre mice were crossed with tdTomato reporter (Ai14) mice
(The Jackson Laboratory, C57BL/6J background). For injections, we anaesthetized mice with 2%
isofluorane, thinned the skull over A1 (or V1) and performed ∼0.2 mm
2
craniotomy. We
delivered the virus using a beveled glass micropipette (tip opening: 40–50 µm in diameter)
attached to a microsyringe pump (World Precision Instruments). Adeno-associated viruses
(AAVs) to deliver hChR2 were acquired from the UPenn Viral Vector Core:
AAV2/9.EF1α.DIO.hChR2(H134R)-EYFP.WPRE.hGH (Addgene 20298). We made one
injection of virus in each mouse at a volume of 50 nl and at a rate of 15 nl/min in A1 at a depth of
600 µm (n = 15 mice). We then sutured the scalp, and administered an analgesic (0.1 mg/kg
Buprenex). We made in vivo recordings 2–3 weeks after viral injections. We examined the
expression pattern of hChR2(H134R)-EYFP in each injected mouse to make sure that the
expression level was sufficiently strong before the experiment. For slice recording, non-floxed
36

AAV9.hSyn.hChR2(H134R)-eYFP.WPRE.hGH (Addgene26973P, also from UPenn Vector
Core) was injected into A1 (one injection at a depth of 600 µm) of PV-, SOM-, VIP- or GAD2-
Cre::tdTomato mice (n = 3 mice each). For calcium imaging of pyramidal neurons, non-floxed
AAV2/1.Syn.GCaMP6s.WPRE.SV40 (UPenn Vector Core) was injected into V1 (at two depths
of 200 and 400 µm and at two sites) at a of wild-type C57BL/6J mice (n = 4 mice). For calcium
imaging of inhibitory neurons, AAV2/1.Syn.Flex.GCaMP6s.WPRE.SV40 (UPenn Vector Core)
was injected into V1 (at two depths of 200 and 400 µm at two sites) of GAD2-, PV-, SOM- or
VIP-Cre::tdTomato mice (n = 4 mice each). For inhibiting L1 neurons, AAV1-CAG-FLEX-
ArchT-GFP (Addgene 28307, from UNC vector core) was injected into GAD2-Cre mice via
iontophoresis using 3µA currents for 5 sec, at a depth of 50 µm and at 4 different sites within V1.
Cells included in the analysis were obtained only from mice that had ArchT-GFP expression
largely limited to L1 only (n = 4 mice).

2.2.2 Retrograde Tracer Injection
For retrograde tracer injections into V1, 40 nl of fluorescently conjugated Cholera Toxin subunit
B (CTb Alexa 488, 0.25%; Invitrogen) was pressure injected into V1 through a pulled glass
micropipette at a rate of 15 nl/min and at a depth of 400 µm at one site (n = 3 mice in total). After
5–7 days, the animal was deeply anesthetized and transcardially perfused with 4%
paraformaldehyde. Brain tissue was sliced into 150 μm sections using a vibratome (Leica), and
sections were mounted onto glass slides and imaged under a confocal microscope (Olympus).

37

2.2.3 Animal surgery for in vivo recording in anaesthetized mice
We sedated the mouse with the correct expression with an intramuscular injection of
chlorprothixene hydrochloride (10 mg/kg in 4 mg/ml water solution) and then anesthetized it with
urethane (1.2 g/kg, i.p., at 20% w/v in saline), as previously described (Li et al., 2013; Liu et al.,
2009; Niell and Stryker, 2008). We maintained the animal’s body temperature at ∼37.5° by a
heating pad (Havard Apparatus, MA). We performed tracheotomy, and inserted a small glass
capillary tube to maintain a free airway. We performed cerebrospinal fluid draining, removed the
skull and dura mater ( ∼ 1 × 1 mm) over the V1, and applied artificial cerebrospinal fluid solution
(ACSF, containing [in mM] 140 NaCl, 2.5 KCl, 2.5 CaCl
2
, 1.3 MgSO
4
, 1.0 NaH
2
PO4, 20
HEPES, 11 glucose, pH 7.4) to the exposed cortical surface when necessary. We trimmed
eyelashes contralateral to the recording side, and covered the eyes with ophthalmic lubricant
ointment until recording, at which time we rinsed the eyes with saline and applied a thin layer of
silicone oil (30,000 centistokes) to prevent drying while allowing clear optical transmission. Eye
movements and the receptive field drift were negligible within the time window of our recording
(Mangini and Pearlman, 1980). For optogenetic activation applied to the surface of A1, the
ipsilateral eye was enucleated to eliminate potential stimulation of this eye by the blue LED light.

2.2.4 Animal surgery for in vivo recording in awake mice
One week before electrophysiological recordings, mice were anesthetized with isoflurane (1.5%
by volume) and a screw for head fixation was mounted on top of the skull with dental cement, as
previously described (Xiong et al., 2013; Zhou et al., 2014). Afterward 0.1 mg/kg buprenorphine
was injected subcutaneously before they were returned to home cages. During the recovery
38

period, the mice were trained to get accustomed to the head fixation on the recording plate, which
was flat and could rotate smoothly around its center. To fix the head, the head screw was tightly
fit into a metal post. One day before electrophysiological recordings the mouse was anesthetized
with isoflurane and craniotomy was made over the V1 (2.6 mm, lateral to midline; 3.9 mm
posterior to bregma). We then covered the opening with Vaseline and then agarose, and injected
the animal with buprenorphine before returning it to the home cage. On the day of the recording,
the animal was fixed to a metal post through the screw, and agarose and Vaseline were cleared.
Due to potential complex effects of locomotion on visual responses (Fu et al., 2014; Niell and
Stryker, 2010), we have only selected recording trials in which animals remained stationary.

2.2.5 In vivo electrophysiology
We pre-penetrated the pia with a broken pipette under visual guidance before in vivo recordings,
and then performed cell-attached recordings with an Axopatch 200B amplifier (Molecular
Devices). To record from excitatory neurons, the patch pipette had a tip opening of ∼2 µm (4 – 5
MΩ impedance). The intrapipette solution contained (in mM): 140 NaCl, 2.5 KCl, 2.5 CaCl
2
, 1.3
MgSO
4
, 1.0 NaH
2
PO4, 20 HEPES, 11 glucose, pH = 7.4. A 100–250 M Ω seal was formed on the
targeted neuron. We recorded spikes in the voltage-clamp mode with a small commend potential
applied to achieve a zero baseline current. The spike signal was filtered at 10 kHz and sampled at
20 kHz. All the neurons recorded under this condition showed regular-spikes (the spike waveform
had a trough-to-peak interval of 0.85 ± 0.10 ms, n = 72 cells), consistent with the sampling bias
towards excitatory neurons as shown previously with cell morphology reconstructions (Liu et al.,
2010; Wu et al., 2008). The recordings were done in layer 2/3 (200-350 µm from the pia) and
39

layer 4 (375-510 µm). The layer assignment of the blindly recorded neurons was made mostly
according to the vertical travel distance of the electrode. The assignment was reasonably precise
because our use of a high-magnification objective (40×) on the microscope allowed a precise
identification of the cortical surface and our application of pre-penetration minimized the
dimpling of the cortical surface (Li et al., 2013; Li et al., 2012a). To record from fast-spiking
(FS) inhibitory neurons, smaller pipettes with a higher impedance (10 MΩ) were used and
neurons with fast-spikes (trough-to-peak interval of the spike waveform < 0.5 ms) were actively
searched (Wu et al., 2008).

2.2.6 Visual stimulation
The visual stimuli were generated using Matlab with Psychophysics Toolbox and were displayed
with a gamma-corrected LCD monitor (refresh rate 75 Hz, maximum luminance 280 cd/m
2
)
placed 0.25 m away from the right eye. We placed the center of the monitor at 45° Azimuth, 25°
Elevation, and it covered ±35° horizontally and ±27° vertically of the mouse visual field. We
made recordings in the monocular zone of the V1. We recorded spontaneous activity when
presenting a uniform grey background. To measure orientation tuning, we applied drifting
sinusoidal gratings (spatial frequency of 0.04 cycles per degree and temporal frequency of 2Hz)
of 12 directions (30° steps) in a random sequence. The visual stimulation with and without sound
or LED illumination were alternated, but the stimulus sequence was randomized independently
for sound/LED off and sound/LED on trials. We set the inter-stimulus interval at 10 s to allow a
full recovery of ChR2 function from desensitization (Li et al., 2013). Each cell was recorded
under high contrast (95%) and low contrast (25%) conditions. We applied five to ten sets of
40

stimuli to each cell, with the sequence different between sets. The long recording time in our
experimental conditions prevented us from applying different combinations of spatial and
temporal frequency. To better drive as large fraction as possible of V1 neurons, we have carefully
chosen our visual stimuli (0.04 cpd, 2Hz) based on a previous study of a large number of neurons
(Niell and Stryker, 2008), which shows that the largest fraction of mouse V1 cells prefers the
spatial frequency of 0.04 cpd, and the average preferred spatial frequency in L2/3 is ~0.04 cpd,
and that most of units prefer a temporal frequency of ~2Hz. Since the level of change in
selectivity was not correlated with the initial selectivity level or the overall response level
(Supplementary Figure 4D-4F), this suggests that both optimally driven and non-optimally driven
neurons could increase their selectivity under sound/A1 stimulation.

2.2.7 Auditory Stimulation
The auditory stimulation consisted of white noise pulses at 70 dB sound pressure level (SPL)
presented at 10Hz (50 ms on, 50 ms off) throughout the duration of the visual stimulus. The onset
and offset of auditory stimulation were the same as visual stimulation. The sound was delivered
by a single speaker located contralateral to the recording side. The visual stimuli alone and those
coupled with sound, were alternated, but the stimulus sequence was randomized independently
for sound off and sound on trials.

41

2.2.8 In vivo optogenetic manipulation
To photoactivate hChR2, we used a blue (470 nm) fiber-coupled LED (0.8 mm diameter, Doric
Lenses) placed on top of the exposed cortical surface. LED light spanned the entire area of V1.
We applied black pigment stained agar to cover the tip of the optic fiber, as to prevent LED light
leakage reaching the contralateral eye. We had verified that LED light did not directly stimulate
the contralateral eye in wild-type mice (data not shown). The LED was driven by the analog
output from a NIDAQ board (National Instruments). The intensity of LED was around 5 mW
(measured at the tip of the fiber). To photoactivate ArchT, we used a green light (530 nm) fiber-
coupled LED (0.8 mm diameter, Doric Lenses) and followed the same procedure as with hChR2.

2.2.9 In vitro electrophysiology
Slice preparation. Viral injected mice were anesthetized with urethane. After decapitation, the
brain was rapidly removed into an ice-cold oxygenated dissection buffer (60 mM NaCl, 3 mM
KCl, 1.25 mM NaH
2
PO
4
, 25 mM NaHCO
3
, 115 mM sucrose, 10 mM glucose, 7 mM MgCl
2
, 0.5
mM CaCl
2
; bubbled with 95% O
2
and 5% CO
2
; pH = 7.4). Coronal cortical slices of 350 µm
thickness were cut from the infected brain hemisphere by a vibrating microtome (Leica
VT1000s). After being incubated in a warmed (at 34 °C) artificial cerebral spinal fluid (ACSF;
126 mM NaCl, 2.5 mM KCl, 1.25 mM Na
2
PO
4
, 26 mM NaHCO
3
, 1 mM MgCl
2
, 2 mM CaCl
2
, 0.5
mM ascorbic acid, 2 mM sodium pyrurate, and 10 mM glucose, bubbled with 95% O
2
and 5%
CO
2
) for >30 min, the slice was transferred to the recording chamber at room temperature.
Electrophysiological recording. Recording was made under an upright fluorescence microscope
(Olympus BX51WI) equipped with an infrared light source. Slices were examined under a 4X
42

objective before recording to determine whether hChR2-EYFP was expressed in A1. In slices
with good expression sites, whole-cell voltage-clamp recordings were selectively performed on
fluorescence-labeled inhibitory neurons in PV-Cre, SOM-Cre, VIP-Cre or GAD2-Cre::tdTomato
slices or non-fluorescent excitatory cells in GAD2-Cre::tdTomato slices under epifluorescence
imaging in V1. The extracellular solution contained: 60 mMNaCl, 3 mMKCl, 1.25 mM
NaH
2
PO
4
, 25 mM NaHCO
3
, 115 mM Glucose, 10 mM Sucrose, 7 mM MgCl
2
, 0.5 mM CaCl
2
, PH
7.2, and bubbled with 95% O
2
and 5% CO
2
. For examining monosynaptic excitatory responses
only, recordings were made with TTX (a sodium channel blocker, 1µM) and 4-aminopyridine (a
potassium channel blocker, 1mM) present in the external solution. Glass pipette (4-7 MΩ
impedance) was filled with a cesium-based internal solution (125 mM Cs-gluconate, 5 mM TEA-
Cl, 4 mM MgATP, 0.3 mM GTP, 10 mM phosphocreatine, 10 mM HEPES, 1 mM EGTA, 2 mM
CsCl, 1% biocytin, pH = 7.2). The pipette and whole-cell capacitances were completely
compensated and the initial series resistance was compensated for 50% at 100 μS lag. Recordings
were made with an Axopatch 200B amplifier (Molecular Devices). Excitatory synaptic currents
were recorded by clamping the cell’s membrane potential at -70 mV. Signals were filtered at 2
kHz and sampled at 10 kHz. In each slice, multiple neurons were recorded. The cortical depth of
each recorded cell was based on the vertical distance of the cell body from the pial surface of the
cortex, which was set as 0 µm. The distance was measured with a micromanipulator coupled with
a digital reader (SD Instrument DR1000). L1 was defined as 0 −150 µm.
Photostimulation. The hChR2 was activated by blue light pulses from a mercury Arc lamp gated
by an electronic shutter (Li et al., 2014). The excitation light was passed through a blue light
filter and the objective. A calibrated aperture placed at the conjugate plane of the slice was used
to control the size of the illumination area. The aperture was adjusted so that the entire V1 area
43

was illuminated. The power of light stimulation was 3 mW measured at the focal plane. Brief
light pulses (3 ms) were applied individually (0.033 Hz). For each condition, 10 – 30 trials were
given and responses were averaged.

2.2.10 In vivo two-photon calcium imaging
Imaging was performed after at least 2 weeks of viral expression. We used a custom built Mai Tai
(Spectra-physics) based 2-photon system and recorded data using a custom-modified version of
the Scanimage software (Pologruto et al. 2003). Calcium imaging from labeled cells was
performed in multiple subregions each spanning ~200 x 200 up to ~400 x 400 µm. Scan lines
were drawn across each of the clearly visible cell bodies (typically 5-15 cells) and then imaged
continuously (rapidly alternating between all the scan lines) while presenting moving sinusoidal
grating stimuli (1.5 sec) at 10 second intervals. Depending on the number of scan lines the scan
rate ranged between 25-70 Hz. Within each scan line, we manually defined regions of interests
(ROIs) based on the presence of fluorescence transients. The selected ROIs were compared with
the 2 dimensional snapshot of the region to make sure that we were imaging the labeled cell
bodies rather than neuropils. The signal within a ROI was processed using standard methods to
derive fractional change over baseline, i.e. ΔF/F0 (Jia et al. 2011). Neurons located within 150
µm below the pia surface were considered as L1 cells.

44

2.2.11 Data analysis
We performed data analysis with custom-developed software (LabVIEW, National Instruments;
and MATLAB, Mathworks). We counted the spikes evoked by drifting sinusoidal gratings within
a time window covering the visual stimulation duration with a 70 ms delay, and subtracted the
average spontaneous firing rate from the stimulus-evoked spike rate.
We quantified the strength of OS with a global orientation selectivity index (gOSI), which
considers responses at all test orientations (Mariño et al., 2005):
gOSI= ∥∑R(θ)×e
2 iθ
∥/∑R(θ)
i is √−1. θ is the angle of the moving direction. R(θ) is the response level at angle θ. The
preferred orientation was determined from this vector sum of all the responses. We then averaged
responses to the gratings of opposite directions, and obtained the orientation tuning curve between
0 – 180 degrees, and fitted it with a Gaussian function: R(θ) = A × exp(−0.5 × (θ – ϕ)
2
/ σ
2
) + B. ϕ
is the determined preferred orientation and σ is the tuning width. We also computed an
orientation selectivity index (OSI) defined as (R
pref
– R
orth
)/(R
pref
+ R
orth
), where R
pref
is the
response level at the angle of ϕ, and R
orth
is that at the angle of ϕ + 90. The gOSI and OSI values
vary between 0 and 1, with 0 being the value for an untuned neuron and 1 for a perfectly tuned
neuron (Carandini and Ferster, 2000; Mariño et al., 2005; Ringach et al., 2002). To compare
tuning between conditions, the preferred orientation was the peak of the Gaussian fit for the data
in the sound/LED off condition (e.g. in Figure 2E, right panel), or was the test orientation closest
to the peak of the Gaussian fit in the sound/LED off condition (e.g. in Figure 1B, 1C, 2E left
panel, and 2G). Similarly, the orthogonal orientation (e.g. in Figure 1D, 2H) was also determined
based on the data in the sound/LED off condition of each cell. Tuning curves were aligned
45

according the preferred orientation (set as zero degree) before averaging. The global direction
selectivity index (gDSI) was quantified according to:
gDSI= ∥∑R(θ)×e
iθ
∥/∑R(θ)
For slice recording data, peak response (EPSC) amplitude was measured for each cell. In order to
take into account the varying levels of channelrhodopsin expression among different slices, we
quantified a relative response amplitude. Specifically, in each slice, at least three L2/3 pyramidal
cells were recorded, and their EPSC amplitudes were averaged to obtain a mean pyramidal cell
response for that slice. The mean pyramidal cell responses of different slices were then averaged
to obtain a global average pyramidal cell response. In each slice, a scaling factor was determined
by the ratio between the mean pyramidal cell response for that slice and the global average
pyramidal cell response. EPSC amplitudes of all of cells recorded in the same slice were then
scaled according to the scaling factor for that slice.

2.2.12 Statistical analysis
We first performed Lilliefors test to check whether the data were normally distributed. In the case
of a normal distribution, we performed paired t-test. Otherwise, we performed a non-parametric
test (Wilcoxon signed-rank test in this study). For the L1 inhibition experiment, one-way
ANOVA with repeated measures with Bonferroni post-hoc test was performed. For the slice
recording experiment, the relative input strengths of different groups were compared using one-
way ANOVA with Bonferroni post-hoc test. No statistical method was used to pre-determine
sample sizes, but our sample sizes were similar to those reported in previous publications in the
field.
46

2.3 Results
2.3.1 Effects of sound stimulation on excitatory neuron responses in V1
Using in vivo cell-attached loose-patch recording, we examined spike responses of L2/3 neurons
in V1 to drifting sinusoidal gratings at 12 different directions (6 orientations) and at 25% contrast
in anaesthetized mice. Our recording method was highly biased towards sampling excitatory
neurons (Wu et al., 2008; Liu et al., 2010). Visual stimuli in the absence and presence of white
noise pulses (50 ms pulse duration, at 10 Hz, for 1.5 sec, 70 dB sound pressure level (SPL)) were
interleaved, and stimuli of different directions were presented in a random fashion. As shown by
an example cell in Figure 1A, sound apparently increased the evoked spike number at the cell’s
preferred orientation, while decreased the spike number at the orientation orthogonal to the
preferred. We used a global orientation selectivity index (gOSI), which is equivalent to 1 minus
circular variance (Mariño et al., 2005), to quantify the degree of orientation selectivity (OS) (see
Experimental Procedures). The cell exhibited an increased gOSI in the presence of sound (Figure
8A, right panel), indicating that the cell’s OS had been sharpened.
In a population of 26 similarly recorded orientation selective excitatory neurons, we normalized
the evoked firing rates to the value at the preferred orientation for each cell, and then calculated
the average normalized firing rates across different orientations relative to the preferred (Figure
8B). We found that the increase of evoked firing rate at the preferred orientation in the presence
of sound was statistically significant (p < 0.001, Wilcoxon signed rank test), and that there was a
decrease of firing rate at orientations away from the preferred (Figure 8B). On average, there was
a 22% increase in evoked firing rate at the preferred orientation (Figure 8C), and 32% decrease at
the orthogonal orientation (Figure 8D), and 9% decrease in the overall firing rate averaged across
47

orientations (p < 0.01 paired t-test). The trend of these changes was evident in individual neurons
(Figure 8C, 8D). Together, these changes led to a significant increase in gOSI (Figure 8E).
Additionally, the spontaneous firing rate was also decreased in the presence of sound stimulation
(Figure 9 A-B), indicating that sound responses of layer 2/3 pyramidal neurons. There was no
significant effect on the activity and orientation selectivity of layer 4 neurons (Figure 9D-E). We
also confirmed this phenomenon in the stationary awake state. As shown in the example neuron
(Figure 10A), we observed an increase of firing rate at the preferred orientation, decrease of firing
rate at the orthogonal orientation, and a corresponding increase in the gOSI of the neurons. On
average, there was a 24% increase of firing rate at the preferred orientation (Figure 10C) and 48%
decrease of firing rate at the orthogonal orientation (Figure 10D). These together led to a
significant increase in gOSI (p < 0.01 paired t-test) (Figure 10E).

48

A B
C
D E
0.0
0.5
1.0
FR at Pref Ori (-S)
FR at Pref Ori (+S)
0 4 8 12 16
0
4
8
12
16
Norm. FR
***
0 90 180 270
0
2
4
6
8
(8)
90
o
270
o
180
o
0
o
gOSI (-S)= 0.532
gOSI (+S) = 0.771
Orientation Angle θ (
o
)
Firing Rate (Hz)
0 0.2 0.4 0.6 0.8 1.0
0
0.2
0.4
0.6
0.8
1.0
gOSI (-S)
gOSI (+S)
***
0.0
0.2
0.4
0.6
***
0.0
0.5
1.0
FR at Orth Ori (-S)
FR at Orth Ori (+S)
0 2 4 6 8
0
2
4
6
8
Norm. FR
-S
+S
Norm. Evoked FR
-90 0 90
0.2
0.6
1
1.4
θ (
o
)
+S
-S
***
***
***
***
***
gOSI
Figure 8. Sound induced sharpening of orientation selectivity in V1
(A) In vivo loose-patch recordings from layer 2/3 pyramidal neurons. Left panel, post-stimulus spike time
histograms (PSTHs) of spike responses of an example neuron to drifting sinusoidal gratings without (left) and
with (right) coupling with sound stimulation. The preferred orientation of the cell was marked by arrows. Scale:
60 Hz, 500 ms. Right panel, firing rates at the different stimulus directions for the same cell. Red: visual
stimulation only. Blue: visual plus sound stimulation. Dashed grey line marks the level of spontaneous firing
rate. Top inset, polar graph plotting of orientation tuning of the cell. The axial value (number of spikes evoked)
is marked within the parentheses. The gOSI values without (-S) and with (+S) sound are labeled.
(B) Average normalized firing rates for 26 pyramidal cells without (red) and with (blue) sound, fitted with a
Gaussian function. The tuning curves were aligned according to the preferred orientation (defined as 0
0
)

without
sound stimulation. Responses to the two directions of the same orientation were averaged. Error bar = SEM.
***, p < 0.001, Wilcoxon signed rank test.
(C-E) Plots of evoked firing rate at preferred orientation (C) and orthogonal orientation (D), as well as of gOSI
values (E), in the presence of sound (+S) versus that without sound (-S) for individual cells. Insets, mean
relative change in firing rate of all cells (C, D), or mean change in gOSI value (E). ***, p < 0.001, Wilcoxon
signed rank test (C) or paired t-test (D, E). Error bar = SEM (C, D) or SD (E).
49

To further confirm the observed phenomenon with a larger number of cells, we performed two-
photon Ca
2+
imaging in anaesthetized mice using a genetically encoded Ca
2+
indicator. AAV1-
Syn-GCaMP6s virus was injected into V1 and experiments were performed after 2-3 weeks of
viral expression. Although the non-floxed AAV virus in principle had labeled both excitatory and
0
1
2
3
0
2
4
6
*
Spike Rate (Hz)
- Sound + Sound
B C
*
- LED + LED
A
0 0.5 1 1.5
0
0.12
0.24
0.36
Time (Sec)
Spike Rate (Hz)
Sound onset
Spike Rate (Hz)
gOSI (-S)
gOSI (+S)
FR at Pref Ori (-S)
FR at Pref Ori (+S)
FR at Orth Ori (-S)
FR at Orth Ori (+S)
0.0
0.5
1.0
0.0
0.5
1.0
1.5
0.0
0.2
0.4
0.6
Norm. FR
Norm. FR
n.s
n.s
E F
D
n.s
- S
+S
6 12 18
6
12
18
2 6 10
2
6
10
0.2 0.4 0.6 0.8
0.2
0.4
0.6
0.8
.gOSI
L4 L4 L4
Figure 9. Effect of optogenetic activation/sound on the spontaneous firing rate in layer 2/3 neurons in V1.
(Related to Figure 1) and effect of sound on L4 neuron responses
(A) PSTH responses of an example layer 2/3 neuron to sound alone (0.5-1.5 sec) in a total of 50 trials.
(B) Line plotting of the spontaneous firing rate without and with sound stimulation. The average FR for no sound
condition was 1.175 Hz and for sound condition was 0.85 Hz. Bar=SEM. *, p < 0.05, paired t-test.
(C) Line plotting of the spontaneous firing rate without and with optogenetic stimulation of the A1 fibers. The
average FR for no LED condition was 0.83 Hz and for the LED condition was 0.6 Hz. Bar=SEM. *, p < 0.05,
paired t-test.
(D) Plot of the evoked firing rate at the orthogonal orientation in the presence (+S) versus that in the absence (-S)
of sound stimulation for L4 cells. Inset, mean evoked firing rate normalized to the value in the sound off condition
of all the cells. ‘ns’, p > 0.5, paired t-test. Bar = SEM.
(E) Plot of gOSI in the presence (+S) versus absence (-S) of sound stimulation for L4 cells. Inset, mean gOSI
value. ‘ns’, p > 0.5, paired t-test. Bar = SD.

50

inhibitory neurons, the great majority of fluorescent cells we imaged ought to be pyramidal cells
due to their large abundance in the cortex. We made line scanning across cell bodies of individual
neurons so that Ca
2+
responses of 10-20 neurons could be imaged simultaneously with a
reasonably high temporal resolution (Figure 11A, left panel, see Experimental Procedures).
Visual stimuli without and with coupling of sound were interleaved. The time-dependent
fractional change of fluorescence intensity ( ΔF/F0) was determined for each orientation over an
imaging period of 10 sec. Figure 11A (right panel) shows plotting of fluorescence change in an
example L2/3 neuron without and with sound. An increase in fluorescence intensity was
observed at the preferred orientation and a simultaneous decrease was seen at some non-preferred
orientations (Figure 11B). Overall, there was a slight decrease in Ca
2+
response averaged across
orientations in the sound on condition for this cell (Figure 11C).
Tuning curves generated from ΔF/F0s (normalized to the value at the preferred orientation) in the
absence and presence of sound were plotted for a total of 75 L2/3 neurons (from 4 mice) (Figure
11D). For summarizing results, we averaged data in two ways. In the first, we averaged
normalized ΔF/F0s across cells for each orientation (relative to the preferred orientation in the
absence of sound), and fit the average results with Gaussian functions (Figure 11E, left panel). In
the second, for each individual cell, we first fit data with Gaussian curves, and then averaged
Gaussian curves of all the cells (Figure 11E, right panel). By either way of analysis, we observed
an increase of Ca
2+
signal at the preferred orientation, and a decrease of the signal at orientations
away from the preferred in the population of L2/3 neurons. In general, the preferred orientation
was not significantly different between sound off and sound on conditions (p > 0.05, paired t- test,
Figure 11F). There was a 10% increase in Ca
2+
response at the preferred orientation (Figure
11G), and 15% decrease in the response at the orthogonal orientation (Figure 11H), and a 5%
51

decrease in overall averaged responses (p < 0.001, paired t-test). These changes resulted in a
significant increase in gOSI (Figure 11I). Together, these Ca
2+
imaging data were consistent with
the loose-patch recording results, confirming the sharpening effect of sound input on visual
processing.

Orientation Angle θ (
o
)
Firing Rate (Hz)
Norm. FR at Pref. Ori
*** ** **
0.0
0.5
1.0
0.0
0.5
1.0
0.0
0.2
0.4
0.6
0.8
A
B
C D E
θ (
o
)
-90 0 90
0.2
0.6
1.0
1.4
Norm. Evoked FR
**
*** ***
**
Norm. FR at Orth. Ori
gOSI
-S
+S
-S +S -S +S -S +S
0 90 180 270
2
4
6
8
10
(15)
90
270
180 0
gOSI (-S) = 0.560
gOSI (+S)= 0.661
Figure 10. Sound mediated sharpening of OS in awake mice. (Related to Figure 1)
(A) In vivo loose-patch recordings from layer 2/3 pyramidal neurons. Left panel, PSTHs of spike responses of an
example neuron to drifting sinusoidal gratings without (left) and with (right) coupling with sound stimulation. The
preferred orientation of the cell was marked by arrows. Scale: 60 Hz, 500 ms. Right panel, firing rates at the
different stimulus directions for the same cell. Red: visual stimulation only. Blue: visual plus sound stimulation.
Top inset, polar graph plotting of orientation tuning of the cell. The axial value (number of spikes evoked) is
marked within the parentheses. The gOSI values without (-S) and with (+S) sound are labeled.
(B) Average normalized firing rates for 10 pyramidal cells without (red) and with (blue) sound, fitted with a
Gaussian function. The tuning curves were aligned according to the preferred orientation (defined as 0
0
)

without
sound stimulation. Responses to the two directions of the same orientation were averaged. Error bar = SEM.
(C-E) Mean relative change in firing rate of all cells at preferred orientation (C), orthogonal orientation (D), or
mean change in gOSI value (E). **, p < 0.01, ***, p < 0.001, paired t-test. Error bar = SEM (C, D) or SD (E).
52

Norm. ΔF/F0
A
D
F
0.2
0.6
1.0
1.4
-90 0 90 -90 0 90
B C
E
G H I
Orientation Angle θ (
o
)
0 90 180 270 360
2
3
4
ΔF/F0
θ (
o
)
+S
-S
0
1
2
3
Time (s)
10
+S
-S
-20 -10 0 10 20
0
4
8
12
16
Δ Ori
No. of Cells
0 2 4 6 8 10
0
2
4
6
8
10
ΔF/F0 Pref Ori (-S)
ΔF/F0 Pref Ori (+S)
Norm. FR
***
0.0
0.5
1.0
-90 0 90
0.6
0.8
1
1.2
Mean Norm. ΔF/F0
+S
-S
***
0 0.15 0.30 0.45
0
0.15
0.30
0.45
gOSI (+S)
gOSI (-S)
0.0
0.1
0.2
0.3
ΔF/F0 Orth Ori (+S)
ΔF/F0 Orth Ori (-S)
0 5 10 15
0
5
10
15
0.0
0.5
1.0
Norm. FR
***
-S
+S
0.6
0.8
1
1.2
***
-90 0 90
***
** ** ***
Orientation Angle θ (
o
)
0
4.5
Time (s)
10
Orientation Angle θ (
o
)
ΔF/F0
0
gOSI
Figure 11. Sound induced sharpening of OS revealed by two-photon Ca
2+
imaging
(A) Left panel, an example imaging plane in layer 2/3 (250 µm below the pial surface). Red lines were used for
line scanning across labeled cell bodies. Scale bar: 100 µm. Right panel, color map of average fractional change in
fluorescence (ΔF/F0) over an imaging period of 10 sec (5 repetitions) for 12 different stimulus directions, plotted
for an example cell marked with a dotted white circle on the left. Time zero represents the onset of visual stimuli.
Stimulus orientation is marked by a black line below the color map. First block shows visual responses alone;
second block shows the responses to visual plus sound stimulation. Black arrows point to the preferred orientation
of the cell.
(B) Tuning curves of the example cell shown in (A).
(C) Average Ca
2+
responses across all the orientations for the same cell.
(D) Normalized Ca
2+
responses (peak amplitude) plotted for all the imaged layer 2/3 cells (n =75) in the absence
(left) and presence (right) of sound stimulation. Tuning curves were aligned according to the preferred orientation
of each cell in the sound off condition.
(E) Left panel, average normalized Ca
2+
responses of all the cells. Bright red and blue lines represent the Gaussian
fits. Right panel, average of Gaussian curves fitted individually for each cell. **, P < 0.01; ***, p < 0.001,
Wilcoxon signed rank test.
(F) Histogram of differences in preferred orientation between sound off and sound on conditions for cells analyzed
in this study. Preferred orientations were determined from Gaussian fits.
(G-I) Plots of ΔF/F0 at the preferred orientation (G) and orthogonal orientation (H), as well as of gOSI value (I) in
the sound on versus sound off condition, for all individual cells (n=75). Insets, mean relative change in Ca
2+

response of all cells (G, H), or mean change in gOSI value (I). ***, p < 0.001, Wilcoxon signed rank test. Error
bar = SEM (G, H) or SD (I).
53

2.3.2 V1 receives direct inputs from A1
For sound to modulate visual cortical responses, auditory information must have entered the
visual cortex by a certain route. A previous study has suggested that A1 might make direct
connections with V1 through its layer 5 neurons (Iurilli et al., 2012). Here, we traced the direct
(CTb-Alexa 488) into V1 (Figure 12A, left panel). The retrogradely labeled cells in A1 were
observed in different layers, but most densely in layer 5 (Figure 12A, middle and right panel). To
further demonstrate direct projections from A1 to V1, we injected AAV encoding floxed
humanized channelrhodopsin 2 (hChR2)-EYFP into A1 of Rbp4-Cre mice, a L5-specific Cre line
(Figure 12B), since layer 5 appeared to contain the largest population of V1 projecting neurons.
The EYFP expression was largely limited to layer 5 as expected, with labeled neurons extending
apical dendrites into layer 1 (Figure 12B, left panel). In V1, we found that fluorescence-labeled
A1 axons mainly terminated in superficial layers of V1, with the highest axonal density being
observed in layer 1 (Figure 12B, right panel). The strength of this A1-V1 projection is compared
with the projection to the inferior colliculus (IC), a well-known corticofugal projection from the
same mouse in order to qualify the intensity of innervation. There was no significant difference
between the intensity of projections to dorsal cortex of IC (DCIC) and L1 of V1 (Figure 13).

2.3.3 Optogenetically stimulating A1 axons sharpens OS of L2/3 excitatory
neurons
After 2-3 weeks of viral expression of hChR2 in A1, we optogenetically activated A1 axons and
recorded spike responses from excitatory neurons in V1. Blue LED light (470 nm) illumination
was delivered through an optic fiber placed either on the surface of V1 or A1 (see Experimental
54

Procedures). To activate ChR2-expressing axons, trains of LED pulses (duration = 10 ms) were
applied at 20 Hz for 1.5 sec to cover the duration of visual stimulation. We examined spike
responses of single cells to moving visual stimuli with and without LED stimulation at two
different contrasts (95% and 25%) of the gratings, using loose-patch recordings. In layer 2/3, we
observed a similar phenomenon as with sound stimulation: sharpening of orientation tuning, as
manifested by an increase and a decrease in firing rate at the preferred and orthogonal orientation,
respectively (Figure 12C-J). Interestingly, the effect was more pronounced under the lower visual
contrast (compare Figure 12C-F with Figure 12G-J). The increase in firing rate at the preferred
orientation was 26% at 25% contrast (p < 0.001, Wilcoxon signed rank test), and 12% at 95%
contrast (p < 0.01, Wilcoxon signed rank test). The decrease in firing rate at the orthogonal
orientation was 40% at 25% contrast (p < 0.01, paired t-test), and 24% at 95% contrast (p < 0.05,
paired t-test). The increase in gOSI was 25% at 25% contrast (p < 0.001, Wilcoxon signed rank
test) and 15% at 95% contrast (p < 0.001, Wilcoxon signed rank test). The differences in the
results between the low and high contrast stimulation were significant (Figure 14). With the
visual stimulation alone, orientation tuning was sharper at the higher contrast (gOSI = 0.37 ± 0.12
at 25% contrast, 0.46 ± 0.13 at 95% contrast, p < 0.01, paired t-test, n = 10), consistent with our
previous report (Li et al., 2012a). Interestingly, the percentage change in the gOSI values
observed was negatively correlated with the initial gOSI value. This suggests that sound/LED has
a better effect of increasing the tuning of neurons with low-medium gOSI values, and may not be
able to increase the tuning of very sharply tuned neurons even further (Figure 16). Additionally,
the changes in the OSI values were computed and we observed a similar increase in the OSI
values with sound/LED stimulation (Figure 15). Together, these data suggest that particularly
under low stimulus strength conditions (as with the low contrast gratings), sound has a great
55

benefit to the vision of mice in that it renders principal neurons in the visual cortex to be better
tuned. On the other hand, in layer 4, stimulating A1 axons did not have any significant effects on
orientation tuning or evoked firing rates (Figure 12K-R), consistent with the notion of a top down
rather than a thalamocortical modulation of visual processing.

Figure 12. Optogenetic activation of A1 axons in V1 (next page)
(A) Retrograde labeling. Left panel, CTb-Alexa 488 (green) injection site in V1. Red shows Rbp4-Cre::tdTomato
expression. Middle panel, retrogradely labeled neurons (green) in A1. Right panel, expression pattern of Rbp4-
Cre::tdTomato in A1. Scale bar, 200 μm.
(B) Anterograde labeling. Left panel, Multiple 10X images stitched together to show AAV2/9-hChR2-DIO-eYFP
injection site in A1 of an Rbp4-Cre mouse. Scale bar, 250 μm. Right panel, image of A1 axons in V1 of an Rbp4 -Cre
mouse. Scale bar, 200 μm.
(C) Upper panel, PSTHs for responses of an example layer 2/3 pyramidal neuron to gratings at high (95%) contrast,
without (left) and with (right) LED illumination (Light source). Black arrows mark the preferred orientation. Scale: 70
Hz, 500 ms. Bottom panel, normalized evoked firing rate (FR) at different orientations (relative to the preferred) and
polar graph plotting of orientation tuning (top) for the same cell. Red, visual stimulation only; blue, visual plus LED
stimulation. The gOSI values under different conditions are labeled. Inset, local field current recorded in superficial
layers of V1 without (upper) and with (lower) LED illumination (duration: 50 ms). Arrow marks the timing of the
trigger signal. Scale, 20 pA, 100 ms.
(D-F) Plots of evoked firing rate at preferred (D) and orthogonal (E) orientations, as well as of gOSI (F) in the LED off
versus LED on conditions for all individual cells (n = 18). Insets, average relative change in evoked firing rate (D, E)
and average change in gOSI for all the cells (F) *, p < 0.05; **, p < 0.01; ***, p < 0.001, paired t-test (D, E) or
Wilcoxon signed rank test (F). Error bar = SEM (D, E) or SD (F).
(G) The same cell as in (C), except that visual stimuli were at low (25%) contrast. Scale: 60 Hz, 500 ms.
(H-J) Responses under low contrast stimuli. Data were from the same cells in (D-F) and are presented in similar
manners (n = 18). **, p < 0.01; ***, p < 0.001, Wilcoxon signed rank test (H) or paired t-test (I, J).
(K-N) Responses of layer 4 neurons to high contrast stimuli. Data are presented in similar manners. Scale: 80 Hz, 500
ms. “n.s”, nonsignificant, p > 0.05, Wilcoxon signed rank test (L, M) or paired t-test (N), n = 20 cells.
(O-R) Responses of layer 4 neurons to low contrast visual stimuli. Data are presented in a similar manner as before.
Scale: 80 Hz, 500 ms. “n.s”, nonsignificant, p > 0.05, Wilcoxon signed rank test (L, M) or paired t-test (N), n = 20
cells.
56

0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.2
0.4
0.6
0.8
1.0
0 5 10 15 20
0
5
10
15
20
AAV-hChR2-eYFP
A1
95%
(15)
90
o
270
o
180
o
0
o
gOSI (-L) = 0.667
gOSI (+L) = 0.690
-90 0 90
0.2
0.6
1.0
Norm. FR
θ (
o
)
Layer 2/3
-90 0 90
0.2
0.6
1.0
1.4
θ (
o
)
A
C
B
E
F
G
H
J
K
L
0 4 8 12 16
0
4
8
12
16
**
FR at Pref Ori (+L)
Norm. FR
n.s
Norm. FR
n.s
Norm. FR
LED off
-90 0 90
0
0.2
0.6
1.0
θ (
o
)
0.0 0.2 0.4 0.6
0.0
0.2
0.4
0.6
***
gOSI (+L)
0.0
0.2
0.4
0.6
*
0 1 2 3 4
0
1
2
3
4
FR at Orth Ori (+L)
Norm. FR
0.0 0.2 0.4 0.6 0.8
0.0
0.2
0.4
0.6
0.8
***
0.0
0.2
0.4
0.6
***
0 4 8 12
0
4
8
12
Norm. FR
- S
+S
25%
**
0 1 2 3
0
1
2
3
Norm. FR
L2/3 L2/3 L4
Rbp4-Cre-tdtom
CTb-Alexa 488
L5
V1 A1 A1 V1
L2/3
L1
L5
Layer 4
(8) 90
o
270
o
180
o
0
o
gOSI (-L) = 0.597
gOSI (+L)= 0.818
D
I
M
N
0.0
0.5
1.0
1.5
0.0
0.5
1.0
0.0
0.5
1.0
0.0
0.5
1.0
1.5
-90 0 90
0.2
0.4
0.6
0.8
1.0
θ (
o
)
gOSI (-L) =0.742
gOSI (+L)=0.751
(25) 90
270
180 0
o
o
o
o
gOSI (-L) = 0.723
gOSI (+L) = 0.707
(10) 90
270
180 0
o
o
o
o
95%
0.0 0.2 0.4 0.6 0.8
0.0
0.2
0.4
0.6
0.8
n.s
0.0
0.2
0.4
0.6
0 6 12 18
0
6
12
18
0 10 20 30
0
10
20
30
0 10 20 30 40
0
10
20
30
40
0.0
0.5
1.0
Norm. FR
Norm. FR
n.s
n.s
0.0
0.5
1.0
0.0
0.2
0.4
0.6
0.8
O
P
Q
n.s
R
0.0
0.5
1.0
0.0
0.5
1.0
***
** **
**
* *
LED off
L4
gOSI
gOSI
gOSI
gOSI
25%
57

A
B
V1 SC A1
V1
L2/3
L1
A1
ICC
IC
Norm. intensity
0.0
0.5
1.0
A1-IC L1 L2/3
***
ns
***
A1-V1
0.0
0.5
1.0
Norm. intensity
V1-SC V1-A1
***
midline
wm
L1
midline
** *
-60
-40
-20
0
20
40
60
Change (%)
High Contrast
Low Contrast
gOSI Pref Ori Orth Ori
ns
Figure 13. Comparison of axon densities between A1-IC vs A1-V1 projections.
A The projections to IC from an injection in A1 and the projections to V1 from the same
injection (B). Both images are in binary mode and the pixel value was measured across DCIC, V1 L1 and L2/3.
(C) The pixel value was normalized to that of the A1-DCIC projection and compared with those of V1. Bar = SD.
‘ns’, p = 0.1; *** p < 0.001, two-sample t-test. Scale Bar=200μm.
Figure 14. Comparison between changes observed at low contrast versus high contrast. (Related to Figure 3)
(A) The percentage changes in gOSI value, firing rate at the preferred orientation, firing rate at the orthogonal
orientation were compared between high contrast and low contrast visual stimulation conditions. Bar = SD.
‘ns’, p = 0.07; * p < 0.05 ; ** p < 0.01, paired t-test.
58

0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.2
0.4
0.6
0.8
1.0
OSI (Sound off)
OSI (Sound on)
A B
D
n.s
***
0.0
0.5
1.0
L2/3
C
L4
0.0
0.4
0.8
0.0
0.2
0.4
***
0.2 0.4 0.6 0.8
0.2
0.4
0.6
0.8
L2/3 Ca
2+
imaging
OSI (Sound off)
0.2 0.4 0.6 0.8 1.0
0.2
0.4
0.6
0.8
1.0
gDSI (LED off)
gDSI (LED on)
0.0
0.2
0.4
0.0
0.2
0.4
0.2 0.4 0.6 0.8
0.2
0.4
0.6
0.8
0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.2
0.4
0.6
0.8
1.0
High Contrast Low Contrast
G F
n.s n.s
gDSI (LED off)
OSI (Sound off)
0.2 0.4 0.6 0.8 1.0
0.2
0.4
0.6
0.8
1.0
OSI (LED off)
OSI (LED on)
0.2 0.4 0.6 0.8 1.0
0.2
0.4
0.6
0.8
1.0
0.0
0.5
1.0
***
0.0
0.5
1.0
E
n.s
L2/3 L4
OSI (LED off)
- Sound
+Sound
- LED
+LED
OSI
gDSI
Figure 15. Plots of the OSI and DSI measurement.
(A) Plot of the OSI value in the sound off versus sound on condition for L2/3 pyramidal neurons (from loose-patch recording
data). Inset, mean OSI value of all cells. ***, p < 0.001, paired t-test.
(B) Plot of the OSI values in the sound off versus sound on condition for L4 neurons (from loose-patch recording data). ‘ns’, p
> 0.05, paired t-test.
(C) Plot of the OSI value in the sound off versus sound on condition for Ca
2+
responses of L2/3 neurons.
(D) Plot of the OSI value in the LED off versus LED on condition for L2/3 neurons.
(E) Plot of the OSI value in the LED off versus LED on condition for L4 neurons.
(F,G) Plots of the gDSI value in the LED off versus LED on condition for L2/3 neurons under high contrast (F) and low
contrast (G) visual stimulation. Inset, mean gDSI value of all the cells.

59

2.3.4 Innervation pattern of the A1-V1 projection
In order to understand the mechanism by which the sound effect on visual responses is produced
and in particular how inhibitory neurons are involved in this process, in slice preparations we
examined the responses to activation of A1 axons in known major types of inhibitory neurons in
V1. Utilizing available inhibitory neuron-specific Cre lines (PV, SOM, VIP and GAD2-Cre)
(Taniguchi et al., 2011), we crossed each of these lines with the Ai14 tdTomato reporter line to
label individual inhibitory neuron types. We then injected non-floxed hChR2 virus in A1, and
performed whole-cell voltage clamp recordings from desired inhibitory neurons (Figure 17A, left
panel) as well as pyramidal neurons (morphologically identified, or non-fluorescence-labeled
0 1 2 3 4 5
0.0
0.1
0.2
0.2 0.4 0.6 0.8
0.0
0.2
0.4
2
= 0.04 R
p > 0.3
OSI (Sound off)
ΔOSI
OSI (LED off)
2
= 0.03 R
p > 0.2
Avg. FR (LED off)
= 0.01 R
p > 0.3
2 A B C
0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.2
0.4
0.6
0.8
1.0
Norm. ΔOSI
0.0 0.2 0.4 0.6 0.8 1.0
0.0
0.2
0.4
0.6
0.8
1.0
2
= 0.02 R
p > 0.5
2
= 0.06 R
p > 0.5
0 1 2 3 4 5
0.0
0.2
0.4
0.6
0.8
1.0
2
= 0.04 R
p > 0.2
0.2 0.4 0.6 0.8
0.0
0.1
0.2
Figure 16. Correlation between change in gOSI vs initial gOSI
(A,B) change in OSI plotted against the original OSI value (in the sound/LED off condition) (upper panel),
or the normalized change in OSI versus the original OSI value (bottom panel). Data in low visual contrast
condition were used. Normalized change was calculated as ΔOSI/(1-OSI
off
). R is the correlation coefficient.
P value is indicated.
(C) The change (upper) or normalized change (bottom) in OSI plotted against the average firing rate across
orientations in the LED off condition.

60

cells in GAD2-Cre::tdTomato V1, see Figure 17A right panel for an example image). To record
monosynaptic excitatory responses only, TTX (1 µM) and 4-AP (1 mM) were present in the bath
solution (Cruikshank et al., 2010; Ji et al., 2015; Petreanu et al., 2009). LED illumination was
applied to the entire visual cortical area to activate A1 axon fibers in this region (Figure 17A, left
panel). As shown by a number of example neurons (Figure 17B) and summary plots of relative
response peak amplitude (Figure 17C, 17D), layer 1 inhibitory neurons, except for VIP positive
cells which only account for a small subset of layer 1 population (about 10%, see (Rudy et al.,
2011), received in general strong excitatory input from A1, much stronger as compared with other
cell types examined, including pyramidal neurons as well as PV, SOM and VIP inhibitory
neurons (Figure 17D). In layer 2/3, the A1 input to PV neurons was comparable to that to
pyramidal cells (p = 0.97, two sample t-test), both of which tended to be stronger than the inputs
to SOM and VIP neurons (p < 0.01, t-test). Beyond 400 µm depth, none of the cell types received
direct excitatory input from A1 (Figure 17C). This is consistent with the imaging result that most
of A1-V1 axonal fibers are located in superficial layers of V1 in particular layer 1, and with the in
vivo electrophysiological result that layer 4 neuron responses were not modulated by auditory
input. Thus, the strongest excitatory inputs from A1 are received by layer 1 inhibitory neurons
(those negative for VIP), followed by excitatory neurons and PV cells in layer 2/3.

61

0 200 400 600 800
0
500
1000
1500
PV
SOM
L1
Pyr
VIP
Relative Peak Amp. (pA)
Depth ( μm)
L2/3
L2/3 L1
A
B
C D
VIP-
A1
inputs
L2/3
L1
+ TTX & 4AP
L 1 L2/3 L 4
GAD2+/
VIP-
PV
SOM
VIP+
Pyr.
***
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Pyr PV SOM VIP+
Relative input strength
4.0
3.0
2.0
1.0
0.0
* * *
***
*
L1
L2/3
Figure 17. Layer 1 neurons receive maximum A1 input
(A) Left panel, schematic diagram of whole-cell recording in a slice preparation. The thickness of green arrows
represents density of A1 axons in each layer. Red, inhibitory neuron; dark blue, pyramidal neuron; light blue,
LED illumination. Right panel, fluorescence labeled inhibitory neurons in V1 of a GAD2-Cre::tdTomato mouse.
Scale bar, 200 µm.
(B) Average EPSC traces from example neurons of different types in different layers. Note that in layer 1, VIP+
cells did not show EPSCs. The responding layer 1 cells were then most likely VIP- cells (all layer 1 cells are
5HT3aR+/GAD2+). Scale: 50 pA, 50 ms.
(C) Relative peak EPSC amplitudes of all recorded cells at different depths. Blue marks VIP-negative layer 1
neurons.
(D) Relative strength of A1 input to different types of neurons (n = 12 for pyramidal, 9 for VIP- L1, 19 for PV,
12 for SOM, 10 for VIP in layer 2/3). Small box, mean value; large box, SD; whisker, full range. *, p < 0.05;
***, p < 0.001, one-way ANOVA and post hoc test.

62

2.3.5 Activation of L1 neurons by sound
To examine whether visual responses of L1 neurons were affected by sound stimulation, we
performed Ca
2+
imaging in GAD2-Cre::tdTomato mice injected with AAV encoding floxed
GCaMP6s in V1. We found that most of imaged L1 neurons were robustly activated in the
presence of sound as opposed to visual stimulation alone. As shown by the plots of ∆F/F0 for an
example L1 neuron (Figure 18A), there was a marked increase in Ca
2+
response level at almost
every stimulus orientation (Figure 18B), resulting in a pronounced enhancement of the overall
response averaged across orientations (Figure 18C). Tuning curves of normalized peak response
level are plotted for 40 L1 cells (Figure 18D). In the presence of sound, we observed an increase
of visually evoked Ca
2+
response at nearly all orientations, either by averaging normalized
response levels across cells (Figure 18E, left panel) or by averaging individually fitted Gaussian
curves (Figure 18E, right panel). On average, there was a 20% increase in Ca
2+
response at the
preferred orientation (Figure 18F), 10% increase at the orthogonal orientation (Figure 18G), and
13% increase in the overall response level (Figure 18H). The latter increase was relatively larger
at the lower visual contrast (Figure 20A), and enhanced with increasing sound intensities (Figure
20B). L1 neurons overall were tuned, although the tuning was relatively broad (Figure 18E).
Since the increase in Ca
2+
response was relatively larger at the preferred than orthogonal
orientation (Figure 18F, 18G), the OS of L1 neurons was enhanced in the presence of sound
(Figure 18I).
Additionally, sound stimulation alone could excite L1 neurons in V1 (Figure 19A-B). In one
experiment, we imaged Ca
2+
responses of a group of L1 neurons and L2/3 pyramidal cells in the
same animal and with the same temporal resolution, and found that visual responses in L2/3
appeared delayed relative to sound-evoked responses in L1 (Figure 19C). To be able to more
63

precisely compare the onset timings of these two types of responses, we performed loose-patch
recordings from L1 neurons, which were easily identified by their spike responses to sound
(Figure 19D, 19E). Indeed, relative to the visual responses of L2/3 pyramidal cells (Figure 19F),
the onset of sound-evoked spike responses of L1 neurons was much earlier (Figure 19G). Given
that sound and visual stimulation were applied simultaneously in our experiments of cross-
modality interaction, this result suggests that L1 neurons may be able to modulate L2/3 responses
as early as the latter start.

2.3.6 Optogenetic suppression of L1 neuron activity
To examine whether the sound-induced modulation of OS could be mediated by L1 neurons, we
sought to selectively suppress their activity during sound presentation. Since there is no L1-
specific Cre line currently available, we employed an iontophoresis method to limit the spread of
viral particles within layer 1 when injecting floxed ArchT-GFP AAV in GAD2-Cre mice (see
Experimental Procedures) (Figure 6A, lower panel). Green LED light (530 nm) illumination was
applied onto the surface of V1 for inhibiting ArchT-expressing L1 neurons (Figure 21A, upper
panel). Using loose-patch recordings, we examined spike responses of L2/3 excitatory cells to
interleaved trials of visual stimulation only (moving gratings at 25% contrast), visual plus sound,
and visual plus sound plus ArchT-mediated L1 inhibition. As shown by the tuning curves of two
example cells (Figure 21B) and summary of all the cells (Figure 21C), OS was enhanced in the
presence of sound as compared with visual stimulation alone (compare red and blue curves),

64

Orientation Angle θ (
o
)
Norm. ΔF/F0
ΔF/F0 Orth Ori (+S)
ΔF/F0 Orth Ori (-S)
0 2 4 6 8 10 12
0
2
4
6
8
10
12
0.0
0.5
1.0
*
Orientation Angle θ (
o
)
0 90 180 270 360
0
1
2
3
ΔF/F0
θ (
o
)
A B
D E
F G H
0 10
0
1
2
Time (s)
+S
-S
0 5 10 15 20
0
5
10
15
20
ΔF/F0 Pref Ori (-S)
ΔF/F0 Pref Ori (+S)
0.0
0.5
1.0
***
Norm. FR
Norm. FR
gOSI (+S)
gOSI (-S)
0.00 0.05 0.10 0.15 0.20 0.25
0.00
0.05
0.10
0.15
0.20
0.25
***
0.00
0.05
0.10
0.15
***
Average ΔF/F0 (-S)
Average ΔF/F0 (+S)
0.0
0.5
1.0
Norm. FR
0 3 6 9 12
0
3
6
9
12
+S
-S
0.6
0.8
1
1.2
1.4
-90 0 90
***
*** ***
** ** ** **
Time (s)
10
0
3.5
ΔF/F0
0.2
0.6
1.0
-90 0 90
0.2
0.6
1.0
-90 0 90
0.8
1
1.2
1.4
0.6
Avg. Norm. ΔF/F0
-90 0 90
Orientation Angle θ (
o
)
0
C
I
-S
+S
gOSI
Figure 18. Visual responses of Layer 1 neurons are enhanced by sound
(A) Ca
2+
responses of an example layer 1 neuron to gratings at 12 different directions (average of 5 repetitions) in the
absence or presence of sound stimulation.
(B) Left, tuning curves of peak Ca
2+
responses in the absence (red) and presence (blue) of sound plotted for the same
cell in (A). Right, time-dependent ΔF/F0 averaged across orientations for the same cell.
(C) Normalized tuning curves in the absence (left) and presence (right) of sound for all the imaged layer 1 neurons (n
= 40).
(D) Left, average normalized Ca
2+
response amplitudes across imaged layer 1 cells. Bright red and blue curves are
Gaussian fits. Right, average of all Gaussian fits for individual layer 1 cells.
(E-H) Plots of peak Ca
2+
response amplitude at the preferred orientation (E) and orthogonal orientation (F), of average
response amplitude across all orientations (G), as well as of gOSI value (H) in the sound on versus sound off
conditions. Each data point represents one layer 1 cell. Insets, average relative change across all cells (E-G), or
average change in gOSI value (H). *, p < 0.05; ***, p < 0.001, Wilcoxon signed rank test (E, F) or paired t-test (G,
H). Error bar = SEM (E-G) or SD (H).
65

2
6
10
0 5
0
1
0 1 2 3
1.2
2.4
3.6
4.8
6.0
A B
0 10
0.0
0.1
0.2
0.3
Sound Only
No Sound/Visual
ΔF/F0
Time (sec)
0 0.2 0.4 0.6
0
0.5
1.0
ΔF/F0
Cumulative Fraction
Onset
Time (sec)
Norm. ΔF/F0
Visual Only L2/3
Sound Only L1
C
FR (Hz)
0
100
200
300
400
D F
Onset Latency (ms)
G
0 1 2 3
0.8
1.6
2.4
3.2
4.8
E
Sound
onset
Peak FR (Hz)
Time (sec)
Visual
Onset
Sound
onset
0 1 2 3
3
6
9
Stim.
onset
***
Sound Only
No Sound/Visual
Figure 19. Response of layer 1 neurons to sound alone.
(A) Average Ca
2+
response trace of an example L1 neuron to sound stimulation alone (blue). Black, trigger only with zero
sound output.
(B) Cumulative distribution of L1 neurons’ peak Ca
2+
response amplitudes to sound (n = 27 cells).
(C) Population Ca
2+
response of 5 L1 neurons to sound stimulation alone (blue) and that of 5 L2/3 neurons to visual
stimulation alone (red) imaged in the same mouse, both normalized to their respective peak amplitude.
(D) PSTH of spike responses of an example L1 neuron to sound stimulation alone. Solid black line marks the sound
duration.
(E) Average PSTH for 6 L1 neurons. Inset, peak firing rates of individual neurons and their mean. Bar = SD.
(F) PSTH for responses of an example L2/3 neuron to visual stimulation alone. Black bar marks the duration of moving
gratings.
(G) Average onset latencies of L1 neuron spike responses to sound stimulation and L2/3 neuron spike responses to
gratings. Onset latency was determined from the PSTH of an individual cell by the time when spike rate exceeds the
average baseline firing rate plus two standard deviations of baseline fluctuations. Bar = SD. **p < 0.01, t-test, n = 6 for
L1 and 14 for L2/3.

66

as the firing rate at the preferred orientation increased and that at the orthogonal orientation
decreased, similarly as observed earlier. When L1 neurons were suppressed, the sound-induced
increase of response level at the preferred orientation was reduced by approximated half, and the
decrease of response level at the orthogonal orientation was nearly abolished (Figure 21C). These
changes were summarized in Figure 21D and 21E. Together, the reduction of overall response
level was prevented (Figure 21F). As a result, the OS of pyramidal cells in the condition of L1
suppression was significantly weaker as compared with the visual-plus-sound condition (Figure
21G). Together, our data indicate that suppressing L1 neurons could block, at least partially, the
sharpening of OS of L2/3 pyramidal cells by sound, suggesting that L1 neurons are involved in
mediating the sound-induced modulation effect.

A
n.s
0.0
0.5
1.0
1.5
+ Sound
Low High
Ctrst Ctrst
- Sound
Average Norm. ΔF/F0
** **
*
0.0
0.5
1.0
1.5
0 40 60 80
Sound Intensity (dB SPL)
Average Norm. ΔF/F0
*
**
B
Figure 20. Effect of sound on visual responses of layer 1 neurons.
(A) Mean normalized average Ca
2+
response amplitude (to the value under no sound condition) of all the neurons (n = 40)
under low and high contrast visual stimulation.
* p < 0.05; ** p < 0.01, one-way ANOVA with repeated measures.
(B) Mean normalized average Ca
2+
response amplitude at four sound intensities tested (0,40,60 and 80 dB SPL) for all the L1
neurons. ‘ns’, p > 0.05; * p < 0.05 ; ** p < 0.01, one- way ANOVA with repeated measures.
67

0.00
0.25
0.50
0.75
1.00
0.00
0.25
0.50
0.75
1.00
0.00
0.50
1.00
1.50
0.00
0.50
1.00
1.50
0 90 180 270 360
0
2
4
6
8
Orientation Angle
Visual Only gOSI = 0.132
Visual + Sound gOSI= 0.342
Visual + Sound + gOSI= 0.211
Pref Ori. Orth. Ori. gOSI
A B
D E F
θ (
o
)
0 90 180 270 360
0
2
4
6
8
Firing Rate (Hz)
Visual Only gOSI = 0.367
Visual + Sound gOSI= 0.662
Visual + Sound + gOSI= 0.430
Cell 1 Cell 2
** *
Norm. Firing Rate
* *
* *
G
Norm. gOSI
** *
Avg. FR
Depth (μm)
Fluor. Intensity (a.u)
L1 L2/3
0 50 100 150 200 250 300
300
250
200
150
100
50
0
C
V1
LED
-90 0 90
0.0
0.2
0.4
0.6
0.8
1
1.2
1.4
θ (
o
)
Norm. Firing Rate
+S
-S
+S+
L1
150 μm
Figure 21. Inhibition of layer1 neurons reduces the sound induced sharpening effect
(A) Left, schematic diagram of the experimental setup. Green LED illumination was applied to V1 surface. Middle,
expression pattern of ArchT-GFP in a GAD2-Cre mouse. Right, average fluorescence intensity of ArchT-GFP at
different depths across animals examined (n = 4). Shade = 95% confidence level.
(B) Tuning curves of evoked firing rate for two example L2/3 pyramidal neurons. Red, blue and green lines represent
visual stimulation only, visual plus sound stimulation, visual plus sound plus inhibiting L1, respectively.
(C) Average tuning curves in three different conditions across all the cells recorded (n = 11).
(D-G) Normalized evoked firing rates at the preferred (D) and orthogonal orientation (E), average evoked firing rates
across all orientations (F), as well gOSI values (G) under the three different conditions. Data represent mean ± SEM. *,
p < 0.05; **, p < 0.01, one-way ANOVA with repeated measures.

68

2.3.7 Effects of sound stimulation on L2/3 PV, SOM and VIP neurons
Taking advantage of the inhibitory neuron specific Cre lines, we further imaged Ca
2+
responses
from PV, SOM or VIP neurons in layer 2/3 of V1 that had been injected with AAV encoding
floxed GCaMP6s (Figure 23). Although all the inhibitory subtypes received some degree of
direct excitatory input from A1 (Figure 17D), the effects of sound on gOSI and overall response
level were variable. For PV neurons, we did not find significant changes in the visually evoked
Ca
2+
response at the preferred orientation or in the average response level when sound was
applied (Figure 24). Similar results were obtained when we identified PV neurons based on their
fast spiking properties (Atallah et al., 2012; Li et al., 2015; Ma et al., 2010) in our blind loose-
patch recordings (Figure 25A-C). For SOM neurons, we found that visually evoked Ca
2+

responses could not be detected in the majority (~90%) of SOM neurons imaged under the current
anesthetized condition, suggesting that the firing rate of SOM neurons might be in general low
(see also (Adesnik et al., 2012). In the small subset of SOM neurons which exhibited significant
visually evoked Ca
2+
responses, we did not find a significant change in their average response
level (p > 0.05, paired t-test) or in the response level at the preferred orientation (p > 0.5, paired t-
test) (Figure 24C-D). Therefore, neither PV nor SOM neurons appeared to contribute to the
sound-induced modulation of excitatory neuron responses. For VIP neurons however, we
observed a significant decrease of Ca
2+
response at the preferred orientation (p < 0.001, Wilcoxon
signed rank test) (Figure 24E), and a smaller decrease in the average response level (p < 0.001,
paired t-test) (Figure 24F). Ca
2+
responses of an example VIP neuron are shown in Figure 24G-I,
and summary tuning curves for all VIP neurons are shown in Figure 24I. Interestingly, there was
a more pronounced decrease of response level at the preferred than other orientations (Figure
69

24I), suggesting that VIP neurons are in a suitable position to preferentially disinhibit pyramidal
cells at their preferred orientation in the presence of sound.

tdTomato
GCaMP6s Merged
A
B
C
D
PV-Cre
SOM-Cre
GAD2-Cre
VIP-Cre
Figure 22. Expression of GCaMP6s in various Cre::tdTomato mice.
Left panel, GCaMP6s expression (green) in V1in GAD2-Cre (A), VIP-Cre (B), PV-Cre (C) and SOM-Cre
(D)::tdTomato mice. Middle panel, tdTomato expression (red) in V1. Right panel, superimposed image of
GCaMP6s and tdTomato expression. Note that all green cells are also red. Scale = 250 μm.
70

0.4 0.6 0.8 1.0
0.4
0.6
0.8
1.0
0.4 0.6 0.8 1.0
0.4
0.6
0.8
1.0
PV VIP
ΔF/F0 Pref Ori
A
D B
C
H G
Sound on
0 4 8 12 16
0
4
8
12
16
0.0
0.5
1.0
***
Norm. FR
0.0
0.5
1.0
***
0 2 4 6 8 10
0
2
4
6
8
10
Norm. FR
0.0
0.5
1.0
n.s
0 4 8 12 16
0
4
8
12
16
Norm. FR
0.0
0.5
1.0
n.s
0 3 6 9 12
0
3
6
9
12
Norm. FR
Avg. ΔF/F0
Sound on
0 90 180 270
2.0
3.0
4.0
θ (
o
)
0 10
0.2
1.0
1.8
Time (s)
-90 0 90
0.5
0.6
0.7
0.8
0.9
1
θ (
o
)
Avg. Norm. ΔF/F0
***
* * * *
Time (s) 10
Sound off
ΔF/F0
0.0
0.5
1.0
Norm. FR
SOM
E
F
I
0.0
0.5
1.0
Norm. FR
n.s
n.s
-S
+S
0
Sound off Sound off
Sound off Sound off Sound off
Sound on
Sound on
Sound on
Sound on
- Sound
+Sound
0
3.75
ΔF/F0
Figure 23. Effects of sound stimulation on PV, SOM and VIP neurons in layer 2/3
(A-) Plots of peak Ca
2+
response amplitude at the preferred orientation (A) and of average response amplitude across
orientations (B) in the sound on versus sound off conditions, for PV neurons (n = 30). Insets, average relative change in
amplitude across all cells “n.s.”, p > 0.05, Wilcoxon signed rank test
(C-D) Plots of responses for SOM neurons (n = 8). Data are presented in the same manner as in (A-B). “n.s.”, p > 0.05,
paired t-test.
(E-F) Plots of responses for VIP neurons (n=40). ***, p <0.001, Wilcoxon signed rank test (E) or paired t-test (F).
(G) Fractional change in fluorescence (ΔF/F0) at 12 different directions of gratings plotted for an example VIP neuron in
layer 2/3. First block shows visual responses alone, next block shows the responses to visual stimulation coupled with
sound stimulation.
(H) Upper panel, tuning curves for the same cell in (G). Lower panel, time-dependent fractional change in fluorescence
averaged across all orientations for the same cell in (G). Red, visual stimulation alone; blue, visual stimulation coupled
with sound.
(I) Average of Gaussian fits of tuning curves of individual VIP neurons. Red curve, visual stimulation only; blue curve,
visual stimulation coupled with sound. * p < 0.05; *** p < 0.001, Wilcoxon signed rank test.
71

2.4 Discussion
In this study, we examined how input of one sensory modality influences specific response
properties of cortical neurons in another modality. Using loose-patch recording and Ca
2+

imaging, we found that activation of auditory cortex sharpens OS of layer 2/3 excitatory neurons
in V1, more so when the visual stimuli were at a lower contrast. The sharpening effect was
attributed to an increase of response level at the preferred orientation and a decrease at the
orthogonal orientation, with an overall decreased average response level. These effects were
achieved, at least partially, through intracortical connectivity from A1 to V1: cortico-cortical
axons from A1 layer 5 neurons mainly terminated in superficial layers of V1 and activated L1
inhibitory neurons. The latter likely caused the overall suppression of pyramidal cells observed in
the underlying layer 2/3. In addition, L1 neurons might inhibit other inhibitory neurons in layer
0
20
40
60
FR at Pref Ori.
- S +S
A
B
C
0
20
40
60
Average FR
- S +S
0
0.1
0.2
0.3
gOSI
- S +S
Figure 24. In vivo loose-patch recordings from fast spiking neurons in layer 2/3.
(A) Evoked firing rate at the preferred orientation without (-S) and with (+S) sound, plotted for all recorded fast-spiking cells
(n = 12). Data points for the same cell are connected by a line.
(B) Average evoked firing rate across all orientations, plotted for all the cells.
(C) gOSI values under sound off and sound on conditions (p > 0.05, paired t-test).
72

2/3, generating a complex disinhibition effect which contributed to the increased firing rate at the
preferred orientation of the pyramidal neurons.

2.4.1 Modulation of V1 processing by auditory input through a top down
circuit

Our results indicate that in V1 the axons of A1 L5 neurons project mostly to the superficial layers,
with the highest axonal density in layer 1, which is consistent with recent mouse brain
connectome studies (Oh et al., 2014; Zingg et al., 2014). This projection pattern is also supported
by slice recording results: when A1 axons were optogenetically stimulated, L1 neurons (except
those positive for VIP) received the maximum direct input among all the major cell types. In
layer 2/3, pyramidal neurons as well as PV, SOM and VIP inhibitory neurons also received direct
A1 inputs but to a varying degree, with the inputs to pyramidal and PV neurons stronger than
those to SOM and VIP neurons. Neither type of neuron below 400 µm received direct A1 inputs.
These results do not support the notion of a previously suggested layer 5 to layer 2/3 inhibitory
route (Iurilli et al., 2012) underlying the sound induced suppressive effect. Instead, our data
indicate that L1 VIP-negative neurons are the primary targets of A1 input. L1 neurons are known
to connect with all the cell types (including all inhibitory types) in the underlying layer 2/3 (Jiang
et al., 2013; Lee et al., 2015; Wozny and Williams, 2011; Xu and Callaway, 2009; Zagha et al.,
2013), and thereby are in a suitable position to provide general inhibition to layer 2/3 pyramidal
cells and to produce disinhibitory effects by inhibiting certain layer 2/3 inhibitory neurons. Such
L1-mediated top down circuit is a likely candidate for mediating the A1 modulation of visual
processing, since suppressing L1 neuron activity reduced the sharpening of OS by sound. It is
73

thus tempting to postulate that layer 1 of V1 can serve as a hub of top down connectivity,
receiving inputs from different sensory modalities or even non-sensory higher cortical regions
(Zhang et al., 2014). Interestingly, while A1 profoundly innervates V1, viral injections in V1 did
not reveal any connectivity from V1 to A1 (data not shown, also see (Oh et al., 2014). This
indicates that the direct connection between A1 and V1 is unidirectional, at least in some species.

L5
L1
L2/3
in
X
in
VIP
Figure 25. Proposed circuitry mechanism for the sound mediated sharpening of orientation selectivity
Sound or Optogenetic activation of A1 fibers activates Layer 1 neurons in V1. The increase of L1 activity in turn inhibits
neurons in Layer 2/3. The inhibition of excitatory neurons by layer 1 activation gives rise to the overall decrease of firing
observed in these neurons. The inhibition of VIP neurons by Layer 1 neuron activation specifically at their preferred
orientation disinhibits excitatory neurons. The differential effect of overall inhibition and activation specifically at the
preferred orientation is responsible for the sharpening of the orientation selectivity of layer 2/3 excitatory neurons to
sound/optogenetic stimulation.
74

2.4.2 Sound effect on L1 neuron responses
Consistent with an involvement of L1, sound stimulation resulted in an elevation of visual
responses in L1 neurons. The effect was dependent on the sound intensity. The minimum testing
intensity at which there was a consistent increase in L1 responses was 60 dB SPL, while the
intensity threshold for A1 neurons was lower than 40 dB SPL (Li et al., 2014; Liu et al., 2007;
Sun et al., 2013). The largest increase occurred when the maximum testing intensity (80 dB SPL)
was applied. These results suggest that there is a certain intensity threshold for the cross-modal
interaction coming into play. Interestingly, the fractional increase of response level was larger at
the preferred than orthogonal orientation of L1 neurons, enabling them to generate relatively
larger disinhibition at their preferred orientation. Such “supralinear” effect is possible and could
be mediated by certain voltage-dependent conductances such as NMDA receptors and dendritic
Ca
2+
channels (Lavzin et al., 2012; Xu et al., 2012).

2.4.3 L2/3 Inhibitory neuron subtypes
L1 neurons are likely the source of the general inhibition observed in pyramidal neurons in layer
2/3, given that L1 neurons directly innervate L2/3 pyramidal cells. In layer 2/3, PV, SOM and
VIP neurons together account for nearly 100% of the inhibitory cell population, and VIP neurons
comprise the largest inhibitory subgroup in layer 2/3 (Rudy et al., 2011). All of these inhibitory
subtypes receive direct A1 inputs, but to varying degrees. They are also innervated by L1
inhibitory neurons (Jiang et al., 2013). Additionally, within the L2/3 inhibitory neuron
population, there is complex inhibition between inhibitory cells (Pfeffer et al., 2013). All
together, these interactions determine that the effects of sound on layer 2/3 inhibitory neuron
75

responses could be rather complex, which depend on the delicate balance between the excitation
from A1, inhibition from L1 of V1, and intra-laminar or local inhibition. Indeed, for PV neurons,
we did not observe any significant changes of response levels in the presence of sound. PV
neurons are thus unlikely to play a significant role in mediating the sound-induced sharpening of
OS of pyramidal cells. The number of visually driven SOM neurons was so low that SOM
neurons were also unlikely a major player.
What types of inhibitory neuron in layer 2/3 are more likely to contribute to the sharpening of OS,
in particular the disinhibition effect seen at the preferred orientation of pyramidal cells? Several
previous studies have suggested that VIP neurons, the most abundant L2/3 inhibitory neurons, are
the main driver of a disinhibitory circuit: their activation by various long-range projections
inhibits SOM neurons, which in turn disinhibits pyramidal cells (Fu et al., 2014; Lee et al., 2013;
Pfeffer et al., 2013; Pi et al., 2013). Contrary to the expectation based on previous studies, VIP
neurons were in fact suppressed in the presence of sound. There was a decrease in their overall
Ca
2+
response, and an even greater decrease at the preferred orientation (Figure 24G-I). The latter
effect may overcome the increased layer 1 inhibition in pyramidal cells, resulting in an overall
disinhibition effect at this stimulus orientation. As suggested by our previous reports that
excitation and inhibition in an excitatory cell display a similar preferred orientation (Li et al.,
2012b; Liu et al., 2011), it is certainly possible that the connected VIP and pyramidal cells share
similar orientation preferences. Although a slice recording study shows that a VIP cell’s
connection to a pyramidal neuron is relatively weak (Pfeffer et al., 2013), optogenetic activation
of VIP neurons causes a significant decrease of firing rate at least in a subset of pyramidal cells
(Lee et al., 2012), indicating that VIP neurons as a population are able to affect pyramidal cell
firing rates. Together, the behavior of VIP neurons in the presence of sound, i.e. preferentially
76

decreasing their firing rate at the preferred orientation, suggests that these inhibitory neurons are
in a suitable position to generate a disinhibition effect at the preferred orientation of the connected
pyramidal cells.
Altogether, our results suggest that auditory inputs can modulate visual processing in V1 neurons
through a projection from A1 layer 5 neurons to superficial layers of V1, in particular layer 1.
Such cross-modality top-down modulation may be abundant in sensory systems. It is likely
advantageous for animals’ survival in that other senses (audition in this case, and probably
somatosensation also) can aid the visual system by improving the performance and
discriminability of V1 principal neurons, especially under low visual contrast conditions. For
example, in a dim environment, barely detectable images of predators may be better perceived by
rodents when there are sounds coincidentally associated with the threats. Several intriguing
questions arise from the current results: how this top-down circuit is formed during development
and whether this circuit is hardwired or if it can be shaped by visual or auditory experience
(Petrus et al., 2014; Petrus et al., 2015) . Studying the A1-V1 circuit in the developing brain
of both naïve and experientially manipulated animals will be an exciting future research direction.

77

CHAPTER 3
A genetic strategy for stochastic gene activation with
regulated sparseness (STARS)

3.1 Introduction
A central feature of the nervous system is the presence of an enormous diversity of neurons with
different morphologies, molecular profiles and physiological properties (Ascoli et al., 2008; Lee
et al., 2010; Rudy et al., 2011; Sun et al., 2013; Swanson, 2007). Accurately determining the
morphology of molecularly specified neurons and their corresponding physiological properties is
essential for the understanding of the functional organization of the nervous system. In the past,
Golgi staining was widely used to visualize the morphology of neurons. Though the method
revealed detailed dendritic and axonal morphologies of various neurons, it was nevertheless
random without molecular information on the cells being analyzed. There have also been many
efforts to label neurons intracellularly during electrophysiological recordings by filling the cell
with tracers such as biocytin or fluorescence conjugated dextrans (Smith and Armstrong, 1990;
Wilson et al., 1990). Though these methods provided considerable information regarding the
morphology and electrophysiological properties of neurons, their molecular identity often
remained unclear. Combining intracellular recording and single-cell RNA extraction and analysis
allows the assessment of the cell’s molecular identity (Belinsky et al., 2014; Fuzik et al., 2016;
Okaty et al., 2011)but the process is challenging and laborious.
78

The Cre-lox system and its variants have been revolutionary in providing tools to access neurons
with a specific molecular profile. Strategies to label and manipulate neurons involve the usage of
transgenic mouse lines that have a set of genetically identified cell types expressing Cre
recombinase (Madisen et al., 2010; Taniguchi et al., 2011). While these strategies have been
extensively employed and proved very useful in addressing a multitude of questions, it has
become increasingly important to have a genetic system in which expression of a certain
transgene or disruption of a certain gene of interest occurs in a small population of a genetically
identified cell type. The application of sparse genetic manipulation is tremendous, including but
not limited to, tracing projections of single neurons, determining their morphology as well as
perturbing activity in a sparse manner that does not affect the whole system. There have been
several efforts to address the issue of sparsely labeling neurons. One approach is to screen
transgenic lines with variegated gene expression, taking position effects into account, and use the
line with the desired expression pattern (Feng et al., 2000; Young et al., 2008). Secondly, the
CreER-lox mediated recombination is utilized to achieve sparse labeling in a tissue of interest,
which requires careful administration of an appropriate dose of tamoxifen (Hayashi and
McMahon, 2002). The appropriate dose however is only determined empirically. Thirdly, low
dose viral injections into cell-type specific Cre driver lines can sometimes achieve sparse
labeling. But it is not a reliable method to be able to consistently achieve a desired sparseness.
Fourthly, Cre-lox mediated recombination during mitosis (such as Mosaic Analysis with Double
Markers, MADM) is useful for tracing cell lineage (Zong et al., 2005), but is not applicable to
stages past cell differentiation. Finally, brainbow transgenic strategy was developed in which
Cre/lox recombination was utilized to create a stochastic choice of expression among several
different fluorescence proteins (Livet et al., 2007). Cells can be labeled with as many as tens of
79

different colors depending on the copy number variation during integration of transgenes. This
method enables imaging a large number of individual neurons in the same circuit (Lichtman et al.,
2008). All of the above methods have proved useful to some degree, but a precise genetic control
to regulate the percentage of cells expressing or inactivating a certain gene of interest is not
readily achievable.
Previously in cell lines, we showed that it was possible to regulate the Cre-lox reaction
kinetics by varying the base-pair distance between identical lox sites (Wang et al., 2009). The
larger the distance, the more time it takes for the pair of lox sites to be brought together and less
efficient is the recombination reaction occurring on those lox sites (Fig. 1a). Following this
proofed principle, we designed and generated a transgenic reporter mouse line STARS (STochastic
gene Activation with Regulated Sparseness), in which Cre-dependent eYFP expression is limited
to only ~10% of the Cre expressing cells. Since the expression of eYFP was dependent on Cre,
we could have a good control over the identity of the cells being labeled. Here, we characterized
the STARS mouse and quantified the sparseness level in different regions of the brain and
cochlea. To demonstrate potential applications of STARS, we assayed the postnatal development
of connectivity of individual type II spiral ganglion fibers in the cochlea and examined show how
it was affected by various sound experiences.

80

3.2 Materials and Methods
3.2.1 Generation of STARS transgene

All constructs were constructed using standard molecular cloning methods. CMV β- actin
enhancer (CAGGS) promoter (Niwa, Yamamura et al. 1991; Okada, Lansford et al. 1999) was cut
out from the pCAGGS-ES plasmid (gift of Dr. Le Ma) using SpeI and EcoRI restriction enzymes.
This fragment was sub-cloned into the pBluescript SK+/- plasmid digested with SpeI and SmaI.
Figure 26. STARS strategy. A, The structure of STARS transgene: gene X flanked by lox2272 pair (Unit A)
and gene Y flanked by loxP pair (Unit B) are crossly linked and subjected to Cre action. B, Two Cre
molecules bind to each lox site and two identical lox sites are brought together for the recombination to occur.
Cre stochastically excises recombination unit A or B (the reaction kinetics reflected by the rate constant k1
and k2, respectively) and leads to mutually exclusive expression of gene Y and gene X. C, We hypothesize
that the length of recombination unit may affect the reaction kinetics and leads to differential outcomes after
Cre action.
81

To obtain lox -flanked eYFP, lox3172 and loxP pairs were introduced in the primers amplifying
the membrane bound eYFP (Köster and Fraser, 2001) using the following primer pairs:
5’primer:agatatcctgcagataacttcgtataggatactttatacgaagttataccatgggatgtattaaatca; and
3’primer:ccatgggaagcttataacttcgtataatgtatgctatacgaagttatgaattcggcgcgccggccggccgaattcgtcgaggccg
cgaattaaaaaacc). mCherry (Shaner et al., 2004) was amplified using 5’ primer:
ccatgggaagcttataacttcgtataggatactttatacgaagttatcacgtgccaccatggtgagcaagggcgagga, and 3’primer:
ggaattcctcgagataacttcgtataatgtatgctatacgaagttatagatctgtcgaggccgcgaattaaaaaacc. The fragments
which were sequentially inserted downstream of CAGGS promoter using PCR. We then
generated multiples of tetra-poly(A) fragment (6X-pA) flanked by AscI and FseI in pCRScript
plasmid and inserted these fragments or a single poly(A)(1kb) fragment flanked into the spacer of
STARS by AscI and FseI. The entire STARS construct cassette was subcloned using XhoI plus
NotI sites into pcDNA5/FRT expression vector (Invitrogen) previously modified to remove its
CMV promoter. The entire cassette of 1kb STARS was eventually cloned into the TARGATT
vector (Applied StemCell Inc.) using NotI and Xho I. The spacer fragment (6X-pA=7.2kb) was
then inserted using AscI and FseI enzymes (Figure 28).

3.2.2 Generation of the STARS transgenic mice

A knock-in mouse was generated using the TARGATT technology from Applied StemCell Inc.
The generation of the STARS transgenic mouse model involved four steps. The first step was to
generate a genetically modified founder mouse line, designated Rosa26-3attP (R26P3), by
knocking in three tandem attP sites (3attP) into the mouse Rosa26 locus. The second step
integrated the STARS gene into the R26P3 sites. This step was accomplished by microinjecting
82

an integration cocktail into the pronucleus of zygotes from heterozygous R26P3 FVB mice. The
integration cocktail consisted of the plasmid pBT378-STARS DNA and in vitro transcribed ϕC31
mRNA. Since the STARS transgene is flanked by two attB sites, the ϕC31 enzyme will catalyze
site-specific DNA integration at the attP sites within the modified R26P3 loci. In the third step,
zygotes injected with the integration cocktail were implanted into CD1 foster mice. The fourth
step was to identify STARS transgenic/founder mice by PCR-based genotyping using a panel of
primers. Sixty-one embryos injected with the integration cocktail into pronucleus were implanted
in three foster mice. One female mouse was identified as STARS founder. The F0 founder was
bred with a wildtype FVB male. Out of seventeen pups, five were identified as F1 STARS
founders (Figure 29). Advantages of generating a mouse in such a way include, but not limited to:
Single Copy insertion, no disruption of internal genes and elimination of position effects.

83

Figure 27. (A) Schematic illustration of TARGATT pBT378-STARS construct. Lower panel: Restriction
enzyme digest confirmation for pBT378-STARS constructs. (B) pBT378-STARS-1kb spacer, digested with
XhoI+NotI (Left); pBT378-STARS-7kb spacer, digested with FseI+AscI (Right). GeneRuler 1kb DNA marker.
B
A
84

Identification of STARS mice
Tail tissues from F0 and F1 mice were collected and DNA extraction was performed individually.
A panel of primer pairs was designed to identify and confirm 1) attP site-specific insertion by the
amplification of novel junction sequences at the STARS-targeted R26P3 allele; and 2) the
insertion of STARS gene. The following PCR amplifications were performed.
List of primers:
R10N: 5’- AGTTCTCTGCTGCCTCCTGGCTTCT -3’
PR425N: 5’- GGTGATAGGTGGCAAGTGGTATTCCGTAAG -3’
PR436N: 5’- CCACCTCGACCCGTTCATCATGATG -3’
CAG-R: 5’- CATATATGGGCTATGAACTAATGACCCCGT -3’
CAG-F: 5’- GCCTCTGCTAACCATGTTCATGCCTTCTTC -3'
mCherryR3: 5’- CTTGAAGCGCATGAACTCCTTGATGATG -3’
WPRE-F: 5’- CTCCTCCTTGTATAAATCCTGGTTGCTG -3’
EYFP-F: 5’- GAGTACAACTACAACAGCCACAACGTCT -3’
WPRE-R: 5’- CAGCAACCAGGATTTATACAAGGAGGAG -3’
PR522N: 5’- GACGATGTAGGTCACGGTCTCGAAG -3’
R13: 5’- CATAAACCCCAGATGACTCCTATCCTC -3’

PCR1: R10N/PR436N. R10N (forward primer) recognizes a unique sequence on Rosa26 allele,
while PR436N (reverse primer) recognizes the attL sequence (Figure 1). The attL sequence is
produced only after ϕC31 integrase-mediated insertion at attP site of the 5’arm of R26P3. The
PCR amplification using R10N/PR436N can detect site-specific insertion at 5’ insertion site. The
expected size of the amplicon is 346bp for 2nd and 3rd attP insertion (2, 3- insertion), or 276bp
for 1, 2- or 1, 3-insertion.
PCR2: PR425N/WPRE-F. WPRE-F (reverse primer) recognizes WPRE sequence. The PCR
amplification using PR425N/WPRE-F can detect the insertion of STARS transgene at the 5’-
85

junction. The expected fragment sizes are 1,067 bp for 1,2- or 1,3- insertions, or 1,137bp for 2, 3-
insertion.
PCR3: EYFP-F/WPRE-R. PCR using primer pair EYFP-F and WPRE-R amplifies a 464bp
fragment within the EYFP-WPRE region.
PCR4: CAG-F/mCherry-R3. PCR using CAG-F and mCherry-R amplifies a fragment of 219bp
from CAG promoter region, to mCherry sequence.
PCR5: CAG-R/R13: A PCR amplification using CAG-R/R13 primer pair can detect transgene
insertion at 3’ junction of the Rosa26 locus. The expected size of the PCR amplicons is 439bp for
the 2, 3- or the 1, 3- insertion or 509bp for the 1, 2- insertion.
PCR5: CAG-R/R13: A PCR amplification using CAG-R/R13 primer pair can detect transgene
insertion at 3’ junction of the Rosa26 locus. The expected size of the PCR amplicons is 439bp for
the 2, 3- or the 1, 3- insertion or 509bp for the 1, 2- insertion.
PCR6: PR522N/R13: PR522N (forward primer) binds to the attR sequence (Figure 1). The attR
sequence is produced only after ϕC31 integrase-mediated insertion at attP site of the 3’arm of
R26P3. The reverse primer R13 binds to the R26P3 genomic sequence. A PCR amplification
using the PR522N/R13 primer pair can detect site-specific insertion at 3’ insertion site of the
Rosa26 locus. The expected size using the PR522N/R13 primer pair is 369bp for either the 2, 3-
or the 1, 3- insertion or 439bp for the 1, 2- insertion.

86

3.2.3 Generating STARS::X-Cre lines

Since the STARS mouse is like a sparse version of any traditional reporter line, we need to cross
it with a Cre driver mouse line in order for eYFP to express in specific cell types driven by the
specific promoter driving Cre. To analyze the sparseness level in the brain, the STARS reporter
mouse was crossed with Rbp4-Cre; PV-IRES-Cre; and VIP-IRES-Cre. For analyzing STARS
expression in the cochlea, to label hair cells and spiral ganglion neurons, we crossed the STARS
mouse with PV-IRES-Cre mouse. The resulting progeny were identified by genotyping for Cre
and STARS transgene. Additionally, brains from Ai14 (tdtomato) reporter crossed with the
various Cre driver mice lines were used for comparison.

Figure 28 . Idenification of STARS founder.
PCRs using primer set: Lanes 1: R10N/PR436N; Lane 2: PR425N/WPRE-F; Lane 3: EYFP-
F/WPRE-R; Lane 4: CAG-R/R13; Lane 5: PR522N/R13. GeneRuler 100bp DNA marker.
* indicates expected band sizes
87

3.2.4 Histology

Analyzing sparseness in the brain: Adult mice (progeny of STARS X various Cre driver mice and
Ai14 X Cre mice) were transcaridually perfused with 1X PBS and 4% paraformaldehyde. The
brains were removed and brain tissue was sliced into 150 μm sections using a vibratome (Leica),
and sections were mounted onto glass slides and imaged using a confocal microscope (Olympus).
Analyzing sparseness in the cochlea: Cochlea from PV-Cre::STARS and PV-Cre::Ai14 pups
were dissected out at different ages (P0, P1, P5 and P14). Whole cochleas were mounted and
imaged using a confocal microscope.
Immunohistochemistry: Anti-GFP antibody was used to confirm the sparse expression of our
transgene of interest (eYFP). Fixed samples were permeablized with 0.8% TrionX-100 followed
by incubation in 10% serum blocking buffer for at least 1 hour at room temperature. Primary
antibody incubation overnight at 4°C was followed by secondary antibody incubation for 2 hours
at room temperature. Primary antibodies used were: rabbit anti-GFP (1:1000; Abcam); mouse
anti-CTBP2 (1:200); Rhodamine Phalloidin (1:400; Invitrogen). Secondary antibodies: goat anti-
rabbit Alexa 647; goat anit-rabbit 488; goat anti-mouse 405 (1:500; Invitrogen).

3.2.5 Sound Deprivation

Newborn mice (P0) were raised with their mothers in a sound booth on a regular 12h day/night
cycle. The mice were then perfused at P14, their cochleas dissected, and imaged.

88

3.2.6 Image analysis and quantification

After z-stacks were taken using the confocal microscope, the analysis was done using the ImageJ
software. Multiple 100X100X100 areas were drawn and number of cell bodies was counted
within a particular brain region. Same was done for slices from the tdtomato reporter and the two
were compared and quantified as percentage sparseness (compared with tdtomato as 100%). In
case of analysis of sparseness in the cochlea, the hair cells were simply counted and compared
with the total number of hair cells within that region. To trace spiral ganglion neuron connectivity
in the cochlea, individual spiral ganglion neurons were traced across the z-stack image using
ImageJ and fibers and their processed from individual Type II spiral ganglion neurons were
pseudo-colored to reflect their path and connectivity.

3.3 Results
3.3.1 Design and Construction of STARS
In our previous study (Wang et al., 2009), a STARS transgene was designed such that two
different sets of lox sites were interleaved. Therefore when one recombination occurs between a
pair of identical lox sites, the recombination between the other pair is excluded. As the distance
between a pair of lox sites is increased, the combination efficiency at these sites is lowered,
resulting in a change in the chance of expressing gene X versus gene Y (Figure 26 and 29A). To
achieve maximum sparseness, besides varying the distance between a pair of lox sites, we
screened for readily available lox variants that exhibited the lowest efficiency relative to lox p,
based on the percentage of cells expressing mCherry instead of eYFP (Figure 29B). We found
89

that lox 3172 had the lowest efficiency among tested lox variants. Thus, we have selected lox
3172 and lox p as two lox pairs in our design of the transgene for making STARS mouse. For the
transgene, the fluorescence reporter mCherry plus a spacer cassette composed of multiple poly-A
sequences (6 x pA = 7.2kb) was flanked by two lox 3172 sites. Membrane bound eYFP (meYFP-
WPRE-pA) was flanked by two lox p sites. Because the size of the meYFP-WPRE-pA sequence
was considerably smaller than the mCherry-pA-spacer sequence, the recombination at the lox p
sites is more favorable than that at the lox 3172 sites. Therefore, in the majority of cells mCherry
would be expressed, and in only a small subset of cells eYFP would be expressed (Figure 29C).
Transfection of the STARS construct into mouse embryonic stem (ES) cells revealed reliable
sparse labeling of cells with eYFP when Cre recombinase was also transfected (Figure 29D), The
eYFP expressing cells did not overlap with cells expressing mCherry (Figure 29D), confirming
the exclusiveness between recombination reactions at lox p and lox 3172 sites. The STARS
transgene, controlled by the CAGSS promoter, was cloned between the attB sites in the pBT378
vector (Figure 29C). Using TARGATT technology (Applied StemCell, Inc), a knock-in mouse
was created by inserting the STARS transgene at the Rosa26 locus in a TARGATT mouse that
harbors attP sites, allowing recombination to occur between attB and attP sites in the presence of
the PhiC31 integrase (see Methods).

The STARS mouse was confirmed by genotyping (see Method). A very prominent feature of the
mouse was the widespread expression of mCherry in most of the cells in the body. In the brain,
the mouse typically had very strong expression in blood vessels as well as neurons. The cells
labeled in blood vessels were most likely pericytes and smooth muscle cells.

90

3.3.2 Quantification of sparseness in the brain
The STARS mouse was crossed with different available driver lines that have Cre expression in
excitatory layer 5 neurons (Rbp4-Cre) or subtypes of inhibitory neuron (PV-IRES-Cre and VIP-
IRES-Cre). Coronal sections were made through primary visual (A1) and auditory (V1) cortical
regions. The number of eYFP labeled neurons was quantified in the cortex and hippocampus, and
compared with that in Ai14 (Cre-dependent tdTomato reporter) crossed with the same Cre lines.
Due to the membrane bound version of the eYFP, we were able to visualize all the processes of a
neuron, including dendritic spines and axonal boutons. In the V1 region of STARS mouse
crossed with Rbp4-Cre (Rpb-Cre::STARS), a much smaller number of pyramidal neurons were
labeled by eYFP compared with Rbp4-Cre::Ai14, and the sparsely labeled region matched the
+Cre -Cre
mCherry meYFP merge
ES Cells
C
D
A
B
mCherry
meYFP
merge
lox4271 - loxP
combination
Figure 29. A. Schematic representation of the STARS construct showing the probability of mCherry
and eYFP expression in the presence of Cre.
B. Left Panel: Test of STARS construct in ES cells. Right panel, expression of mCherry in STARS
mouse brain

91

tdTomato labeled region in the Rbp4-Cre::Ai14 mouse (Figure 30A). The eYFP labeled neurons
had clearly visible apical dendrites extending into layer 1, consistent with the known layer 5
pyramidal cell morphologies (Sun et al., 2013; Winer and Prieto, 2001; Wong and Kelly, 1981).
Even though the number of labeled cell bodies was small (7-10% of that in Rpb4-Cre::Ai14),
there was dense labeling of neuronal processes in layer 5 (and layer 1), suggesting that either local
layer 5 neurons have dense arborization or there are some long-range axons from other regions
that terminate and ramify profusely in layer 5 and layer 1.
We also examined the sparseness level in the inhibitory neuron population, utilizing PV- and VIP-
Cre lines that have Cre expression in the parvalbumin (PV) and vasoactive intestinal polypeptide
(VIP) subtype of inhibitory neurons respectively. In V1 of the PV-Cre::STARS mouse, the
number of eYFP labeled neurons was only 9-11% of that labeled by tdTomato in the PV-
Cre::Ai14 mouse (Figure 30B). The laminar distribution of eYFP-labeled PV neurons also
matched that of tdTomato labeled neurons in the PV-Cre::Ai14 mouse, with the highest density
occurring in layer 4 (Figure 30B) (Desgent et al., 2005; Hendry and Jones, 1991; Van Brederode
et al., 1990) . This suggests that the sparseness level was relatively uniformly achieved across
layers. In V1 of the VIP-Cre::STARS mouse, the number of eYFP labeled VIP neurons was 10-
12% of that of tdTomato labeled neurons in the VIP-Cre::Ai14 mouse (Figure 30C). Layer 2/3
contained the highest density of eYFP labeled neurons, consistent with the general laminar
distribution pattern of VIP neurons (Lee et al., 2010; Mesik et al., 2015). Bipolar morphology
was observed for the majority of labeled VIP neurons. Non-bipolar morphologies (e.g. Small
basket cells) were also observed for a small population of labeled VIP neurons (yellow arrow in
Figure 30C), suggesting that VIP neurons may be further divided into subtypes (Markram et al.,
2004; Wang et al., 2002).
92

Similar comparisons were made in A1 (data not shown) and hippocampus (Figure 30D-E).
Across imaged sensory cortical and hippocampal regions as well as across different cell types, we
observed a relatively stable sparseness level, indicating a rather uniform regulation of sparseness
in the brain with STARS. On average, 10±1.88 % of sparseness level was achieved in the cortex
and hippocampus. Together, our data suggest that STARS can be used for sparsely labeling
genetically identified neurons in different parts of the brain and that the morphology of single
labeled neurons can be identified.

93

STARS Ai14
A B C
D
% Cells labeled
(Sensory Cortex)
Rbp4-Cre PV-Cre VIP-Cre
STARS Ai14
DG DG
% Cells labeled
(Hippocampus)
V1 V1 V1
0
2
4
6
8
10
12
14
A1 V1 A1 V1 A1 V1
PV Rbp4 VIP
0
2
4
6
8
10
12
14
DG CA1 DG
PV Rbp4
E
Rbp4-Cre PV-Cre
GCL
GCL
DG
CA1
Figure 30. Quantification of STARS expression in the brain. Comparison of expression in the cortex
and hippocampus in the Rbp4-Cre and PV-Cre lines crossed with either Ai14 or STARS. Bottom
panel: Sparseness percentage comparison between different brain regions among specific Cre driver
lines
94

3.3.3 Quantification of sparseness in the cochlea
Next, we wanted to examine whether we can use STARS to sparsely label cells in peripheral
sensory organs. To that end, we imaged cochleae from PV-Cre mice crossed with either Ai14 or
STARS. PV is expressed in both hair cells (HCs) and spiral ganglion neurons (SGNs) in the
cochlea (Kim et al., 2016; Wang et al., 2013). Since the HCs in the organ of Corti are arranged in
a precise manner (i.e. three rows of outer HCs and one row of inner HCs) (Ruben, 1967; Yang et
al., 2011), the quantification of HC number is relatively straightforward. Cochleae of P0-P1 PV-
Cre::Ai14 or PV-Cre::STARS pups were dissected, imaged and compared between the two
reporter lines. In addition, anti GFP antibody was used to label eYFP and confirm its expression
in the STARS cochlea. As shown in Figure 31A, in the PV-Cre::Ai14 cochlea, tdTomato was
expressed in the majority of HCs (98%, Figure 31C and SGNs. In the PV-Cre::STARS cochlea,
however, the eYFP expression was only observed in about 2-4% of HCs (Figure 31B-C).
We also looked at the cochlea at different ages (P0-P1, P5, P14 and Adult) and quantified the
hair cells labeled and the corresponding spiral ganglion innervation pattern. Interestingly and to
our surprise, we actually found that there is more number of cells labeled consecutively with age
until P14 (The percentage of hair cells labeled at P0 was 1.75±0.55%, at P5 was 3.96±1.56% and
at P14, 11.8±3.2% There was no significant difference between the numbers of cells labeled at
P14 and thereafter. This suggests to us that until P14, which is around the time of hearing onset,
there is actually some regenerative capability in the organ of corti and that there are new hair cells
being generated either by direct trans-differentiation of supporting cells or actual division. Since
only one recombination event can occur in one cell, i.e. either the expression of mCherry or eYFP
can take place; the increase in the number of cells labeled with eYFP suggests an increase in the
overall hair cell number in the cochlea.
95

native GFP anti-GFP merge
P0
PV-CreXAi14 PV-CreXSTARS
OHCs
IHCs
SpG
0
20
40
60
80
100
% Cells labeled
Hair Cells
P0 P5 P14 Adult
PV-Cre::Ai14
PV-Cre::STARS
Figure 31. Quantification of STARS expression in the cochlea. Comparison of expression in the hair
cells of the PV-Cre line crossed with either Ai14 or STARS. Bottom panel: Quantification of percentage
of hair cells labeled in different ages.

96

3.3.4 Tracing type II spiral ganglion fibers (SPG) in the cochlea
Type II spiral ganglion (SG) neurons constitute about 5-10% of all SGNs, and they innervate
outer hair cells (OHCs) (Appler and Goodrich, 2011; Echteler, 1992). In the developing PV-
Cre::STARS cochlea, we were able to visualize individual type II SG fibers in the organ of Corti,
which was not possible in the PV-Cre::Ai14 mouse (Figure 32A-B). After reaching OHC layers,
these fibers make a turn, project tangentially along the HC rows, and make putative synaptic
contacts with OHCs (Figure 32B). We traced these fibers at two developmental time points, P0
and P14, and quantified their lengths in the OHC layers and putative synaptic contacts they made.
Interestingly, at P0, the length of type II SG fiber was different at different rows of OHCs. Fibers
in the first row of OHCs (which is closest to IHCs) were the shortest, and the length progressively
increased in rows 2 and 3 (Figure 32C). On average, the length of a tangential type II fiber in
OHC layers is 162 ± 16 µm in row 1, 307± 31 µm in row 2, and 543 ± 66 µm in row 3 (Figure
32E). There were 9 ± 0.89 OHCs traversed by a type II fiber in row 1, 17.2 ± 1.72 OHCs
traversed in row 2 and 29.2 ± 3.76 OHCs traversed in row 3 (Figure 32F). Though there was no
apparent formation of synaptic contacts by these premature fibers at this age yet, a thickening of
the fiber was often observed near the end (Figure 32B, arrows), suggesting a possible site on the
fiber to form future synaptic contacts. At P14, the length of tangential type II fibers became
similar at different OHC rows, which was also similar as that in Row 3 at P0 (Figure 32D). This
result suggests that type II fibers develop in an outside-in fashion, arriving in Row 3 first and in
Row 1 last. In addition, formation of synaptic structures was obvious, as manifested by filopodia-
like protrusions extending from the tangential fiber near its end (Figure 32B). The number of
synaptic contacts per tangential fiber was not different between OHC rows (8.8 ± 0.74 protrusions
in row 1, 9.2 ± 0.76 in row 2, and 8.4 ± 0.49 in row 3, p > 0.05, one-way ANOVA test) (Figure
97

32F). Furthermore, we observed in some type II fibers that besides the principal OHC row
innervated; they extend a secondary branch to traverse one or even two rows of OHCs (Figure
32B. Figure 33A). Therefore, these fibers could potentially make synapses with OHCs in
multiple rows in a narrow column across the organ of Corti. Finally, high-magnification images
with OHCs stained with Ctbp2 revealed that the protrusions from type II fibers made contacts
with the base of OHCs, the known site for synaptic contacts between type II fibers and OHCs
(Bulankina and Moser, 2012; Huang et al., 2007). Occasionally, we observed that Ctbp2 puncta
co-localized with the end of the protrusions. Together, these observations suggest that the
protrusions from type II fibers do make synaptic contacts with OHCs.

Sound deprivation increases the innervation by type II SG fibers
Since we were able to reveal the synaptic contacts made by individual type II SG fibers
morphologically, we intended to determine the effects of activity on their innervation pattern. As
to understand whether normal acoustic experience is required for the proper formation of type II
fiber innervation, we raised pups in a sound deprived environment from birth (~P0) until hearing
onset (~P14). We then traced individual fibers in a similar fashion (Figure 33C) and compared
with those in normally reared mice. The sound deprivation had no effects on the morphologies of
IHCs and OHCs, as shown by phalloidin staining of hair bundles (Figure 33C). To our surprise,
in sound-deprived animals, we found twice as many synapses formed by type II fibers as in
normally raised (control) animals (Figure 33E). And those synapses were only present in a single
row of OHCs. In other words, there was no secondary branches extending from the tangential
fiber (Figure 33F). For the tangential fiber, it was longer in sound-deprived than control animals,
as also shown by the number of OHCs traversed (Figure 33G). The difference between control
98

and sound-deprived animals is schematically illustrated in Figure 33I. Our data suggest
enhanced type II fiber innervation in the condition of sound deprivation.

PV-Cre::Ai14
No. of OHC traversed
PV-Cre::STARS
B
Row 1 Row 2 Row 3
A
D
C
G
Innervation
portion
0
10
20
30
40
***
ns
***
P0
P14
0
200
400
600
Total length ( μm)
***
ns
***
No. of synaptic contacts
at P14
Row 1
Row 2
Row 3
0
5
10
P0 P14
E F
Figure 32. Row dependent innervation of Type II spiral ganglion neurons onto OHCs at P0 and P14. Upper panel:
Tracing of individual type II fibers innervating the different rows of OHCs at P0. Bottom panel: Tracing of individual
type II fibers innervating the different rows of OHCs at P14. Quantification of the number of OHCs contacted by
individual type II fibers in different rows.

99

0
200
400
600
800
20
30
40
50
P14
A
Control
B
C
Control
Sound Deprived
No. of Synaptic
contacts
H
D
Ctbp2
GFP
Phall
Control
Control
Sound
Deprived
**
Sound Deprived
No. of secondary
branches/fiber
Control
Sound
Deprived
P14
apex
base
OHC
IHC
eYFP
Phall
E F
G
Sound deprived
merged
0
3
6
9
12
15
18
21
Length ( μm)
OHCs traversed
0
2
4
6
8
10
12
Length of
Protrusions ( μm)
I
**
**
IHC OHC
**
0
1
2
Figure 33. Innervation pattern of type II SPGs at P14 under normal and sound deprived conditions. A) Tracing of
individual type II fibers innervating the different rows of OHCs in normal sound environment. B) Juxtaposition of
protrusions and the base of the outer hair cell indicative of synapse formation. C) Tracing of type II fibers in sound
deprived environment. D) Phalloidin staining to show hair cell bundles in the two conditions. E-H) Quantification of
length, synapses, branching and length of protrusions in the two acoustic conditions. I) Schematic representation of
the innervation pattern.

100

3.4 Discussion and Future Directions

In this study, we characterize the STARS transgenic mouse line and show the precise innervation
pattern of typeII spiral ganglion neurons in the cochlea. We show that we can indeed achieve a
genetic version of the “Golgi” staining, specifying exactly which molecularly identified cells we
are looking at. By increasing the length of the spacer region between two identical lox sites, we
are able to change the kinetics of Cre action and thus control the probability of expression of a
gene of interest (eYFP in this case). Utilizing this mouse line, we were able to sparsely label a
few type II spiral ganglion neurons and analyze their innervation pattern at different ages in the
mouse cochlea.
At P0, there was a row dependent innervation onto the OHCs with OHCs in row 3 having the
most number of hair cells contacted, while in row 1 having the least number of cells innervated.
At P14 however, there was no such row dependence, and clear synaptic contacts onto OHCs were
visible. Additionally, we were able to trace individual SPG fibers and observed a multi row
innervation pattern in the OHC layer.
In the case of acoustic deprivation conditions, we in fact observed an increased number of
synapses made by the type II fibers. This was an unexpected observation initially since it is
believed that activity is important for synapse formation. Since the mice were raised in a sound
deprived environment devoid of external auditory inputs, the type II spg neurons tended to make a
lot more connections in order to transmit as much sound information as possible and hence the
increased number of synapses that we observed under those conditions. Thus it can be thought
101

that sound input is required for synaptic pruning as well as branch formation confined to a narrow
frequency range for each type II fiber.
Though we are able to trace individual neuron projections, it is not free of certain limitations. For
example, mCherry is expressed in ~90% of all the cells since it is driven the CAG promoter. This
causes a potential background problem, with high background that can be problematic when
imaging fine axons. This problem can however be circumvented with the generation of a second
generation of STARS mice (STARS2.0) (Figure 35) in which mCherry can be mutated. Secondly,
the sparseness level can be further reduced. Instead of a 6X-pA repeat, the spacer region could be
lengthened and have 8X-pA instead, which can further increase the sparseness. In addition to
these, one important addition to be made to the STARS mouse is being able to manipulate the
sparse number of neurons in a variety of ways. In order to achieve this capability, it is pertinent to
have a gene of interest that can be used to control the cells. For example, if we express Flpe in a
sparse, Cre dependent manner, we have the capability of doing various types of manipulations to
the cells. Firstly, we could inject an Flpe dependent virus expressing either channelrhodopsin or
halorhodopsin. This way, we can activate or inhibit a very small number of neurons in a defined
brain region and defined cell types (driven by Cre). Secondly, we can cross the mouse to a ‘Frted’
mouse that has a critical gene of interest flanked by FRT sites. This way, the gene will be deleted
only in a small percentage of neurons defined by Cre. This is beneficial in that the whole system
will be preserved, but the actions of a few neurons will be disrupted and that could provide more
insight into specific gene function in a defined population of neurons without worrying about
compensatory effects or even perturbation at a very large scale.
This new STARS2.0 mouse is currently under development. The construct is shown in Figure 35.
102

‘Frted’ virus (frted ChR2/GCaMP)
Or Cross with ‘Frted’ mouse
Manipulate activity of few neurons
CAGGS
Lox P
TARGATT Vector
attB
attB
+ Cre
Lox3172 Lox P
Lox3172
mCherry-pA
Flpe-IRES
eCFP
WPRE-pA
9.6 kb spacer
8xpA
X
<10% Cre+ cells  Flpe-IRES-eCFP-WPREpA
>90% cells  No Color
Figure 34. Schematic representations of the STARS2.0 construct showing the probability of Flpe-IRES-eCFP-WPRE-
pA expression in the presence of Cre.

103

References

Adesnik, H., Bruns, W., Taniguchi, H., Huang, Z.J., and Scanziani, M. (2012). A neural circuit
for spatial summation in visual cortex. Nature 490, 226-231.
Alitto, H.J., and Dan, Y. (2012). Cell-type-specific modulation of neocortical activity by basal
forebrain input. Front Syst Neurosci 6, 79.
Appler, J.M., and Goodrich, L.V. (2011). Connecting the ear to the brain: Molecular mechanisms
of auditory circuit assembly. Prog Neurobiol 93, 488-508.
Ascoli, G.A., Alonso-Nanclares, L., Anderson, S.A., Barrionuevo, G., Benavides-Piccione, R.,
Burkhalter, A., Buzsáki, G., Cauli, B., Defelipe, J., Fairén, A., et al. (2008). Petilla terminology:
nomenclature of features of GABAergic interneurons of the cerebral cortex. Nat Rev Neurosci 9,
557-568.
Atallah, B.V., Bruns, W., Carandini, M., and Scanziani, M. (2012). Parvalbumin-expressing
interneurons linearly transform cortical responses to visual stimuli. Neuron 73, 159-170.
Beaulieu, C., Kisvarday, Z., Somogyi, P., Cynader, M., and Cowey, A. (1992). Quantitative
distribution of GABA-immunopositive and -immunonegative neurons and synapses in the
monkey striate cortex (area 17). Cereb Cortex 2, 295-309.
Boyden, E.S., Zhang, F., Bamberg, E., Nagel, G., and Deisseroth, K. (2005). Millisecond-
timescale, genetically targeted optical control of neural activity. Nat Neurosci 8, 1263-1268.
Campi, K.L., Bales, K.L., Grunewald, R., and Krubitzer, L. (2010). Connections of auditory and
visual cortex in the prairie vole (Microtus ochrogaster): evidence for multisensory processing in
primary sensory areas. Cereb Cortex 20, 89-108.
Cappe, C., and Barone, P. (2005). Heteromodal connections supporting multisensory integration
at low levels of cortical processing in the monkey. Eur J Neurosci 22, 2886-2902.
Carandini, M., and Ferster, D. (2000). Membrane potential and firing rate in cat primary visual
cortex. J Neurosci 20, 470-484.
Chu, Z., Galarreta, M., and Hestrin, S. (2003). Synaptic interactions of late-spiking neocortical
neurons in layer 1. J Neurosci 23, 96-102.
Clavagnier, S., Falchier, A., and Kennedy, H. (2004). Long-distance feedback projections to area
V1: implications for multisensory integration, spatial awareness, and visual consciousness. Cogn
Affect Behav Neurosci 4, 117-126.
Cruikshank, S.J., Urabe, H., Nurmikko, A.V., and Connors, B.W. (2010). Pathway-specific
feedforward circuits between thalamus and neocortex revealed by selective optical stimulation of
axons. Neuron 65, 230-245.
Dallos, P., Wu, X., Cheatham, M.A., Gao, J., Zheng, J., Anderson, C.T., Jia, S., Wang, X., Cheng,
W.H., Sengupta, S., et al. (2008). Prestin-based outer hair cell motility is necessary for
mammalian cochlear amplification. Neuron 58, 333-339.
104

Dewson, J.H., Cowey, A., and Weiskrantz, L. (1970). Disruptions of auditory sequence
discrimination by unilateral and bilateral cortical ablations of superior temporal gyrus in the
monkey. Exp Neurol 28, 529-548.
Dragatsis, I., Levine, M.S., and Zeitlin, S. (2000). Inactivation of Hdh in the brain and testis
results in progressive neurodegeneration and sterility in mice. Nat Genet 26, 300-306.
Driver, J., and Spence, C. (1998). Cross-modal links in spatial attention. Philos Trans R Soc Lond
B Biol Sci 353, 1319-1331.
Falchier, A., Clavagnier, S., Barone, P., and Kennedy, H. (2002). Anatomical evidence of
multimodal integration in primate striate cortex. J Neurosci 22, 5749-5759.
Fanselow, E.E., Richardson, K.A., and Connors, B.W. (2008). Selective, state-dependent
activation of somatostatin-expressing inhibitory interneurons in mouse neocortex. J Neurophysiol
100, 2640-2652.
Felleman, D.J., and Van Essen, D.C. (1991). Distributed hierarchical processing in the primate
cerebral cortex. Cereb Cortex 1, 1-47.
Feng, G., Mellor, R.H., Bernstein, M., Keller-Peck, C., Nguyen, Q.T., Wallace, M., Nerbonne,
J.M., Lichtman, J.W., and Sanes, J.R. (2000). Imaging neuronal subsets in transgenic mice
expressing multiple spectral variants of GFP. Neuron 28, 41-51.
Frens, M.A., Van Opstal, A.J., and Van der Willigen, R.F. (1995). Spatial and temporal factors
determine auditory-visual interactions in human saccadic eye movements. Percept Psychophys
57, 802-816.
Frostig, R.D., Xiong, Y., Chen-Bee, C.H., Kvasnák, E., and Stehberg, J. (2008). Large-scale
organization of rat sensorimotor cortex based on a motif of large activation spreads. J Neurosci
28, 13274-13284.
Fu, Y., Tucciarone, J.M., Espinosa, J.S., Sheng, N., Darcy, D.P., Nicoll, R.A., Huang, Z.J., and
Stryker, M.P. (2014). A cortical circuit for gain control by behavioral state. Cell 156, 1139-1152.
Gentet, L.J. (2012). Functional diversity of supragranular GABAergic neurons in the barrel
cortex. Front Neural Circuits 6, 52.
Gielen, S.C., Schmidt, R.A., and Van den Heuvel, P.J. (1983). On the nature of intersensory
facilitation of reaction time. Percept Psychophys 34, 161-168.
Gillespie, L.N., Marzella, P.L., Clark, G.M., and Crook, J.M. (2005). Netrin-1 as a guidance
molecule in the postnatal rat cochlea. Hear Res 199, 117-123.
Gleiss, S., and Kayser, C. (2012). Audio-visual detection benefits in the rat. PLoS One 7, e45677.
Hasselmo, M.E., and Sarter, M. (2011). Modes and models of forebrain cholinergic
neuromodulation of cognition. Neuropsychopharmacology 36, 52-73.
Hayashi, S., and McMahon, A.P. (2002). Efficient recombination in diverse tissues by a
tamoxifen-inducible form of Cre: a tool for temporally regulated gene activation/inactivation in
the mouse. Dev Biol 244, 305-318.
Hensch, T.K. (2005). Critical period plasticity in local cortical circuits. Nat Rev Neurosci 6, 877-
888.
105

HERSHENSON, M. (1962). Reaction time as a measure of intersensory facilitation. J Exp
Psychol 63, 289-293.
HUBEL, D.H., and WIESEL, T.N. (1962). Receptive fields, binocular interaction and functional
architecture in the cat's visual cortex. J Physiol 160, 106-154.
Iurilli, G., Ghezzi, D., Olcese, U., Lassi, G., Nazzaro, C., Tonini, R., Tucci, V., Benfenati, F., and
Medini, P. (2012). Sound-driven synaptic inhibition in primary visual cortex. Neuron 73, 814-
828.
Jaekl, P.M., and Harris, L.R. (2009). Sounds can affect visual perception mediated primarily by
the parvocellular pathway. Vis Neurosci 26, 477-486.
Ji, X.Y., Zingg, B., Mesik, L., Xiao, Z., Zhang, L.I., and Tao, H.W. (2015). Thalamocortical
Innervation Pattern in Mouse Auditory and Visual Cortex: Laminar and Cell-Type Specificity.
Cereb Cortex.
Jiang, X., Wang, G., Lee, A.J., Stornetta, R.L., and Zhu, J.J. (2013). The organization of two new
cortical interneuronal circuits. Nat Neurosci 16, 210-218.
Jones, E.G., and Powell, T.P. (1970). An anatomical study of converging sensory pathways within
the cerebral cortex of the monkey. Brain 93, 793-820.
Kayser, C., Petkov, C.I., Augath, M., and Logothetis, N.K. (2005). Integration of touch and sound
in auditory cortex. Neuron 48, 373-384.
Kilgard, M.P., and Merzenich, M.M. (1998). Cortical map reorganization enabled by nucleus
basalis activity. Science 279, 1714-1718.
Kok, P., Jehee, J.F., and de Lange, F.P. (2012). Less is more: expectation sharpens representations
in the primary visual cortex. Neuron 75, 265-270.
KUFFLER, S.W. (1953). Discharge patterns and functional organization of mammalian retina. J
Neurophysiol 16, 37-68.
Kuffler, S.W. (1973). The single-cell approach in the visual system and the study of receptive
fields. Invest Ophthalmol 12, 794-813.
KUYPERS, H.G., SZWARCBART, M.K., MISHKIN, M., and ROSVOLD, H.E. (1965).
OCCIPITOTEMPORAL CORTICOCORTICAL CONNECTIONS IN THE RHESUS
MONKEY. Exp Neurol 11, 245-262.
Köster, R.W., and Fraser, S.E. (2001). Tracing transgene expression in living zebrafish embryos.
Dev Biol 233, 329-346.
Lavzin, M., Rapoport, S., Polsky, A., Garion, L., and Schiller, J. (2012). Nonlinear dendritic
processing determines angular tuning of barrel cortex neurons in vivo. Nature 490, 397-401.
Lee, A.J., Wang, G., Jiang, X., Johnson, S.M., Hoang, E.T., Lanté, F., Stornetta, R.L.,
Beenhakker, M.P., Shen, Y., and Julius Zhu, J. (2015). Canonical Organization of Layer 1
Neuron-Led Cortical Inhibitory and Disinhibitory Interneuronal Circuits. Cereb Cortex 25, 2114-
2126.
Lee, S., Hjerling-Leffler, J., Zagha, E., Fishell, G., and Rudy, B. (2010). The largest group of
superficial neocortical GABAergic interneurons expresses ionotropic serotonin receptors. J
Neurosci 30, 16796-16808.
106

Lee, S., Kruglikov, I., Huang, Z.J., Fishell, G., and Rudy, B. (2013). A disinhibitory circuit
mediates motor integration in the somatosensory cortex. Nat Neurosci 16, 1662-1670.
Lee, S.H., Kwan, A.C., Zhang, S., Phoumthipphavong, V., Flannery, J.G., Masmanidis, S.C.,
Taniguchi, H., Huang, Z.J., Zhang, F., Boyden, E.S., et al. (2012). Activation of specific
interneurons improves V1 feature selectivity and visual perception. Nature 488, 379-383.
Levick, W.R., Cleland, B.G., and Dubin, M.W. (1972). Lateral geniculate neurons of cat: retinal
inputs and physiology. Invest Ophthalmol 11, 302-311.
Lewandoski, M. (2001). Conditional control of gene expression in the mouse. Nat Rev Genet 2,
743-755.
Li, J., and Schwark, H.D. (1994). Distribution and proportions of GABA-immunoreactive
neurons in cat primary somatosensory cortex. J Comp Neurol 343, 353-361.
Li, L.Y., Ji, X.Y., Liang, F., Li, Y.T., Xiao, Z., Tao, H.W., and Zhang, L.I. (2014). A feedforward
inhibitory circuit mediates lateral refinement of sensory representation in upper layer 2/3 of
mouse primary auditory cortex. J Neurosci 34, 13670-13683.
Li, L.Y., Xiong, X.R., Ibrahim, L.A., Yuan, W., Tao, H.W., and Zhang, L.I. (2015). Differential
Receptive Field Properties of Parvalbumin and Somatostatin Inhibitory Neurons in Mouse
Auditory Cortex. Cereb Cortex 25, 1782-1791.
Li, Y.T., Ibrahim, L.A., Liu, B.H., Zhang, L.I., and Tao, H.W. (2013). Linear transformation of
thalamocortical input by intracortical excitation. Nat Neurosci 16, 1324-1330.
Li, Y.T., Ma, W.P., Li, L.Y., Ibrahim, L.A., Wang, S.Z., and Tao, H.W. (2012a). Broadening of
inhibitory tuning underlies contrast-dependent sharpening of orientation selectivity in mouse
visual cortex. J Neurosci 32, 16466-16477.
Li, Y.T., Ma, W.P., Pan, C.J., Zhang, L.I., and Tao, H.W. (2012b). Broadening of cortical
inhibition mediates developmental sharpening of orientation selectivity. J Neurosci 32, 3981-
3991.
Liang, M., Mouraux, A., Hu, L., and Iannetti, G.D. (2013). Primary sensory cortices contain
distinguishable spatial patterns of activity for each sense. Nat Commun 4, 1979.
Liberman, M.C., Gao, J., He, D.Z., Wu, X., Jia, S., and Zuo, J. (2002). Prestin is required for
electromotility of the outer hair cell and for the cochlear amplifier. Nature 419, 300-304.
Liu, B.H., Li, P., Li, Y.T., Sun, Y.J., Yanagawa, Y., Obata, K., Zhang, L.I., and Tao, H.W.
(2009). Visual receptive field structure of cortical inhibitory neurons revealed by two-photon
imaging guided recording. J Neurosci 29, 10520-10532.
Liu, B.H., Li, P., Sun, Y.J., Li, Y.T., Zhang, L.I., and Tao, H.W. (2010). Intervening inhibition
underlies simple-cell receptive field structure in visual cortex. Nat Neurosci 13, 89-96.
Liu, B.H., Li, Y.T., Ma, W.P., Pan, C.J., Zhang, L.I., and Tao, H.W. (2011). Broad inhibition
sharpens orientation selectivity by expanding input dynamic range in mouse simple cells. Neuron
71, 542-554.
Liu, B.H., Wu, G.K., Arbuckle, R., Tao, H.W., and Zhang, L.I. (2007). Defining cortical
frequency tuning with recurrent excitatory circuitry. Nat Neurosci 10, 1594-1600.
107

Livet, J., Weissman, T.A., Kang, H., Draft, R.W., Lu, J., Bennis, R.A., Sanes, J.R., and Lichtman,
J.W. (2007). Transgenic strategies for combinatorial expression of fluorescent proteins in the
nervous system. Nature 450, 56-62.
Luo, L. (2007). Fly MARCM and mouse MADM: genetic methods of labeling and manipulating
single neurons. Brain Res Rev 55, 220-227.
Luo, L., Callaway, E.M., and Svoboda, K. (2008). Genetic dissection of neural circuits. Neuron
57, 634-660.
Ma, W.P., Liu, B.H., Li, Y.T., Huang, Z.J., Zhang, L.I., and Tao, H.W. (2010). Visual
representations by cortical somatostatin inhibitory neurons--selective but with weak and delayed
responses. J Neurosci 30, 14371-14379.
Mangini, N.J., and Pearlman, A.L. (1980). Laminar distribution of receptive field properties in the
primary visual cortex of the mouse. J Comp Neurol 193, 203-222.
Markram, H., Toledo-Rodriguez, M., Wang, Y., Gupta, A., Silberberg, G., and Wu, C. (2004).
Interneurons of the neocortical inhibitory system. Nat Rev Neurosci 5, 793-807.
Marín, O. (2013). Cellular and molecular mechanisms controlling the migration of neocortical
interneurons. Eur J Neurosci 38, 2019-2029.
McDonald, J.J., Teder-Sälejärvi, W.A., and Hillyard, S.A. (2000). Involuntary orienting to sound
improves visual perception. Nature 407, 906-908.
Meyer, A.C., Frank, T., Khimich, D., Hoch, G., Riedel, D., Chapochnikov, N.M., Yarin, Y.M.,
Harke, B., Hell, S.W., Egner, A., and Moser, T. (2009). Tuning of synapse number, structure and
function in the cochlea. Nat Neurosci 12, 444-453.
Morrison, J.H., Magistretti, P.J., Benoit, R., and Bloom, F.E. (1984). The distribution and
morphological characteristics of the intracortical VIP-positive cell: an immunohistochemical
analysis. Brain Res 292, 269-282.
Müller, U. (2008). Cadherins and mechanotransduction by hair cells. Curr Opin Cell Biol 20, 557-
566.
Niell, C.M., and Stryker, M.P. (2008). Highly selective receptive fields in mouse visual cortex. J
Neurosci 28, 7520-7536.
Oh, S.W., Harris, J.A., Ng, L., Winslow, B., Cain, N., Mihalas, S., Wang, Q., Lau, C., Kuan, L.,
Henry, A.M., et al. (2014). A mesoscale connectome of the mouse brain. Nature 508, 207-214.
Packer, A.M., and Yuste, R. (2011). Dense, unspecific connectivity of neocortical parvalbumin-
positive interneurons: a canonical microcircuit for inhibition? J Neurosci 31, 13260-13271.
Petreanu, L., Mao, T., Sternson, S.M., and Svoboda, K. (2009). The subcellular organization of
neocortical excitatory connections. Nature 457, 1142-1145.
Petrus, E., Isaiah, A., Jones, A.P., Li, D., Wang, H., Lee, H.K., and Kanold, P.O. (2014).
Crossmodal induction of thalamocortical potentiation leads to enhanced information processing in
the auditory cortex. Neuron 81, 664-673.
Petrus, E., Rodriguez, G., Patterson, R., Connor, B., Kanold, P.O., and Lee, H.K. (2015). Vision
Loss Shifts the Balance of Feedforward and Intracortical Circuits in Opposite Directions in
Mouse Primary Auditory and Visual Cortices. J Neurosci 35, 8790-8801.
108

Pfeffer, C.K., Xue, M., He, M., Huang, Z.J., and Scanziani, M. (2013). Inhibition of inhibition in
visual cortex: the logic of connections between molecularly distinct interneurons. Nat Neurosci
16, 1068-1076.
Pi, H.J., Hangya, B., Kvitsiani, D., Sanders, J.I., Huang, Z.J., and Kepecs, A. (2013). Cortical
interneurons that specialize in disinhibitory control. Nature 503, 521-524.
Ringach, D.L., Shapley, R.M., and Hawken, M.J. (2002). Orientation selectivity in macaque V1:
diversity and laminar dependence. J Neurosci 22, 5639-5651.
Robbins, T.W., McAlonan, G., Muir, J.L., and Everitt, B.J. (1997). Cognitive enhancers in theory
and practice: studies of the cholinergic hypothesis of cognitive deficits in Alzheimer's disease.
Behav Brain Res 83, 15-23.
Rubel, E.W., and Fritzsch, B. (2002). Auditory system development: primary auditory neurons
and their targets. Annu Rev Neurosci 25, 51-101.
Rudy, B., Fishell, G., Lee, S., and Hjerling-Leffler, J. (2011). Three groups of interneurons
account for nearly 100% of neocortical GABAergic neurons. Dev Neurobiol 71, 45-61.
Rudy, B., and McBain, C.J. (2001). Kv3 channels: voltage-gated K+ channels designed for high-
frequency repetitive firing. Trends Neurosci 24, 517-526.
Sauer, B. (1993). Manipulation of transgenes by site-specific recombination: use of Cre
recombinase. Methods Enzymol 225, 890-900.
Schnütgen, F., Doerflinger, N., Calléja, C., Wendling, O., Chambon, P., and Ghyselinck, N.B.
(2003). A directional strategy for monitoring Cre-mediated recombination at the cellular level in
the mouse. Nat Biotechnol 21, 562-565.
Shaner, N.C., Campbell, R.E., Steinbach, P.A., Giepmans, B.N., Palmer, A.E., and Tsien, R.Y.
(2004). Improved monomeric red, orange and yellow fluorescent proteins derived from
Discosoma sp. red fluorescent protein. Nat Biotechnol 22, 1567-1572.
Sieben, K., Röder, B., and Hanganu-Opatz, I.L. (2013). Oscillatory entrainment of primary
somatosensory cortex encodes visual control of tactile processing. J Neurosci 33, 5736-5749.
Stehberg, J., Dang, P.T., and Frostig, R.D. (2014). Unimodal primary sensory cortices are directly
connected by long-range horizontal projections in the rat sensory cortex. Front Neuroanat 8, 93.
Sun, Y.J., Kim, Y.J., Ibrahim, L.A., Tao, H.W., and Zhang, L.I. (2013). Synaptic mechanisms
underlying functional dichotomy between intrinsic-bursting and regular-spiking neurons in
auditory cortical layer 5. J Neurosci 33, 5326-5339.
Taniguchi, H., He, M., Wu, P., Kim, S., Paik, R., Sugino, K., Kvitsiani, D., Kvitsani, D., Fu, Y.,
Lu, J., et al. (2011). A resource of Cre driver lines for genetic targeting of GABAergic neurons in
cerebral cortex. Neuron 71, 995-1013.
Vollrath, M.A., Kwan, K.Y., and Corey, D.P. (2007). The micromachinery of
mechanotransduction in hair cells. Annu Rev Neurosci 30, 339-365.
Vroomen, J., and de Gelder, B. (2000). Sound enhances visual perception: cross-modal effects of
auditory organization on vision. J Exp Psychol Hum Percept Perform 26, 1583-1590.
Wang, S.Z., Liu, B.H., Tao, H.W., Xia, K., and Zhang, L.I. (2009). A genetic strategy for
stochastic gene activation with regulated sparseness (STARS). PLoS One 4, e4200.
109

Wilson, N.R., Runyan, C.A., Wang, F.L., and Sur, M. (2012). Division and subtraction by distinct
cortical inhibitory networks in vivo. Nature 488, 343-348.
Winans, S.S. (1967). Visual form discrimination after removal of the visual cortex in cats.
Science 158, 944-946.
Winer, J.A., and Larue, D.T. (1989). Populations of GABAergic neurons and axons in layer I of
rat auditory cortex. Neuroscience 33, 499-515.
Wozny, C., and Williams, S.R. (2011). Specificity of synaptic connectivity between layer 1
inhibitory interneurons and layer 2/3 pyramidal neurons in the rat neocortex. Cereb Cortex 21,
1818-1826.
Wu, G.K., Arbuckle, R., Liu, B.H., Tao, H.W., and Zhang, L.I. (2008). Lateral sharpening of
cortical frequency tuning by approximately balanced inhibition. Neuron 58, 132-143.
Xu, N.L., Harnett, M.T., Williams, S.R., Huber, D., O'Connor, D.H., Svoboda, K., and Magee,
J.C. (2012). Nonlinear dendritic integration of sensory and motor input during an active sensing
task. Nature 492, 247-251.
Xu, X., and Callaway, E.M. (2009). Laminar specificity of functional input to distinct types of
inhibitory cortical neurons. J Neurosci 29, 70-85.
Young, P., Qiu, L., Wang, D., Zhao, S., Gross, J., and Feng, G. (2008). Single-neuron labeling
with inducible Cre-mediated knockout in transgenic mice. Nat Neurosci 11, 721-728.
Zagha, E., Casale, A.E., Sachdev, R.N., McGinley, M.J., and McCormick, D.A. (2013). Motor
cortex feedback influences sensory processing by modulating network state. Neuron 79, 567-578.
Zhang, S., Xu, M., Kamigaki, T., Hoang Do, J.P., Chang, W.C., Jenvay, S., Miyamichi, K., Luo,
L., and Dan, Y. (2014). Selective attention. Long-range and local circuits for top-down
modulation of visual cortex processing. Science 345, 660-665.
Zhou, F.M., and Hablitz, J.J. (1996). Morphological properties of intracellularly labeled layer I
neurons in rat neocortex. J Comp Neurol 376, 198-213.
Zingg, B., Hintiryan, H., Gou, L., Song, M.Y., Bay, M., Bienkowski, M.S., Foster, N.N.,
Yamashita, S., Bowman, I., Toga, A.W., and Dong, H.W. (2014). Neural networks of the mouse
neocortex. Cell 156, 1096-1111.
Zong, H., Espinosa, J.S., Su, H.H., Muzumdar, M.D., and Luo, L. (2005). Mosaic analysis with
double markers in mice. Cell 121, 479-492.

Abstract (if available)

Abstract Animals and humans alike constantly receive a multitude of sensory inputs from their environment. All of these inputs are processed and integrated in a very precise manner in the brain, which ultimately leads to appropriate behaviors and survival. Initially, it was thought that all the sensory modalities operate in isolation

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Functional circuits underlying sensory representation in mouse primary auditory cortex

PDF

Contextual modulation of sensory processing via the pulvinar nucleus

PDF

Axon guidance cues in development of the mammalian auditory circuit

PDF

Synaptic mechanism underlying development and function of neural circuits in rat primary auditory cortex

PDF

Functional properties of the superficial cortical interneurons

PDF

Characterization of the STARS (STochastic gene Activation with Regulated Sparseness) mouse line

PDF

Neural circuits control and modulate innate defensive behaviors

PDF

The molecular mechanism regulating the timing of cell cycle exit during development of the mouse organ of Corti

PDF

The behavioral and neural bases of tactile object localization

PDF

Synapse maintenance and function at the mouse neuromuscular junction: implications in diseases

PDF

Neural circuits underlying the modulation and impact of defensive behaviors

PDF

Slit/Robo signaling underlies the spatial patterning of spiral ganglion neurons to shape the peripheral auditory circuitry assembly

PDF

The role of glial cells in spinal muscular atrophy

PDF

Synaptic circuits for information processing along the central auditory pathway

PDF

The roles of TRPM8 in cold sensation: the six sides of TRPM8

PDF

Characterizing the hippocampal synaptic and sleep abnormalities of a mouse model of human chromosome 16p11.2 microdeletion

PDF

Differential regulation of anatomical and functional visual plasticity by NgR1

PDF

Dual genetic screens for mutants in synaptic homeostatic plasticity and a characterization of insomniac as a regulator for retrograde homeostatic signaling

PDF

Exploiting novel transport properties of adeno-associated virus for circuit mapping and manipulation

PDF

Functional study of C9ORF72 and its implication in the pathogenesis of amyotrophic lateral sclerosis

Asset Metadata

Creator Ibrahim, Leena Ali (author)

Core Title Functional circuitry underlying cross-modality integration and the development of a novel sparse labeling technique

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Neuroscience

Publication Date 04/20/2016

Defense Date 12/03/2015

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag auditory cortex,cochlea,cross modality interactions,experience dependent morphological development,OAI-PMH Harvest,optogenetics,sparse labeling,type II spiral ganglion fibers,visual cortex

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Tao, Huizhong W. (committee chair), Zhang, Li I. (committee chair), Chang, Karen (committee member), Ko, Chien-Ping (committee member)

Creator Email ibrahimm@usc.edu,leena.ali10@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-234659

Unique identifier UC11277332

Identifier etd-IbrahimLee-4308.pdf (filename),usctheses-c40-234659 (legacy record id)

Legacy Identifier etd-IbrahimLee-4308.pdf

Dmrecord 234659

Document Type Dissertation

Format application/pdf (imt)

Rights Ibrahim, Leena Ali

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA