Low-Dimensional Mapping of Corticostriatal Circuitry Dynamics Underlying Pair Bonding Público

Abstract

Extensive research has shed light on neurochemistry and neuroendocrinology contributing to the formation of socially monogamous relationships, known as pair bonds. However, until recently (Amadei et al., 2017), the dynamic neural circuitry underlying the formation of pair bonds has remained unstudied. By analyzing local field potential (LFP) recordings from brain regions in the “social brain network”, namely the medial prefrontal cortex (mPFC), nucleus accumbens (NAcc), and bed nucleus stria terminalis (BNST) of prairie voles, the canonical model organism of pair bonding, we can assess whether the neural dynamics exhibited during pair bonding are stereotyped between individuals or behaviors. Here, an unsupervised machine learning method, t-distributed Stochastic Neighbor Embedding (t-SNE) (van der Maaten & Hinton, 2008) was used to map the structure of LFP recordings collected from the mPFC and NAcc (hit subjects), and mPFC and within or bordering the BNST (non-hit subjects) of female prairie voles during a six-hour cohabitation period with a male partner. The primary objective was to identify behavior-specific brain-states during pair bonding. Intra-behavior variability of hit subjects’ neural dynamics was greater than the intra-animal variability, suggesting there may not be behavior-specific structure in the hit subjects’ brain-state mappings. On the other hand, the intra-animal variability of non-hit subjects’ neural dynamics was greater than the intra-behavior variability, suggesting there may be behavior-specific structure in the non-hit subjects’ brain-state mappings. Furthermore, 36 stereotyped brain-states (i.e. specific pairings of peak oscillatory frequencies) were identified, which may be used for decoding neural signal. Overall, these results provide the basis for further analyses of stereotyped neural dynamics across individuals and behaviors, as well as the temporal emergence of stereotyped neural dynamics over the course of pair bonding.

Table of Contents

INTRODUCTION 1

METHODS 5

           Experiments 5

           LFP Data Collection During Cohabitation 6

           Overview of Analyses 8

           Spectrogram Generation 9

           Spatial Embedding 10

       Jensen-Shannon Divergence 12

RESULTS 13

           Structure and Dynamics of the Low-Dimensional Embedded Space 13

           Inter-Animal and Inter-Behavior Comparisons 17

           Identified Brain-States Corresponding to Regions of the Embedded Space 28

DISCUSSION 34

           Future Directions 36       

REFERENCES 39

APPENDIX 44

A.  Experiments 44

B.   Morlet Wavelet Decomposition 44

C.   t-distributed Stochastic Neighbor Embedding Implementation 45

D.  Identified Spectral Features (PSD) For Each Region 49

FIGURES

1.    Neurologger recording device 6

2.    Ethogram definitions of scored behaviors 8

3.    Overview of the data analysis pipeline 9

4.    Low-dimensional embedding of wavelet-transformed LFP signal 14

5.    Comparison between two- and three-dimensional embedding 15

6.    Histogram of velocities within the embedded behavioral space 17

7.    Jensen-Shannon divergence for all hit and non-hit subjects 20-21

8.    Summary of Jensen-Shannon divergence statistics for hit and non-hit subjects 22

9.    Jensen-Shannon divergence for all subjects for the first hour 23-24

10. Jensen-Shannon divergence for all subjects for the last hour 25-26

11. Summary of Jensen-Shannon divergence statistics for all subjects over time 27

12. Segmentation into regions via a watershed transform 30

13. Labelled segmentation of the PDF 31

14. Examples of PSDs for signal from each brain region 32

TABLES

1.    Summary of identified region features 33

About this Honors Thesis

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Emory College
Department	Neuroscience and Behavioral Biology
Degree	B.S.
Submission	Honors Thesis
Language	English
Research Field	Biology, Neuroscience
Palabra Clave	in vivo Electrophysiology Computational neuroscience Pair bonding Dynamic neural circuitry Affiliative social behavior Unsupervised machine learning
Committee Chair / Thesis Advisor	Berman, Gordon J, Emory University
Committee Members	Sober, Samuel J, Emory University Liu, Robert C, Emory University Manns, Joseph R, Emory University

Última modificación

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Low-Dimensional Mapping of Corticostriatal Circuitry Dynamics Underlying Pair Bonding ()	2018-04-10 11:07:11 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions