טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
M.Sc Thesis
M.Sc StudentYuval Omer
SubjectNeuronalizer: A Neuron Tracing and Analysis Tool And
A Computational Analysis of the PVD Neuron in
Caenorhabditis elegans
DepartmentDepartment of Biology
Supervisors Assistant Professor Tom Shemesh
Professor Benjamin Podbilewicz
Full Thesis textFull thesis text - English Version


Abstract

Individual neurons and neural networks possess complex morphologies thought to be tightly connected to their function and to the behavior of the animal. Despite more than 100 years of observations and tremendous advances in microscopy and image processing techniques, the methods used to study neurons did not advance much.


Due to the enormous complexity of nervous systems, there is an interest in studying the neural networks of relatively simple organisms in order to shed light on fundamental principles of their workings. One of the simplest and highly investigated organisms that possesses a nervous system is the roundworm Caenorhabditis elegans. This, along with being transparent and having a relatively short life cycle, makes C. elegans an attractive model organism for studying neuronal morphology. The PVD neuron is a mechano-receptor, a thermo-sensor and a proprioceptor neuron in C. elegans. It exhibits a complex and dynamic morphology, which is thought to be closely related to its function and to the behaviour of the worm.


However, no rigorous mathematical model, protocol or procedure have been developed in order to accurately and consistently investigate the shape of the PVD neuron. The analysis done today is mostly or completely manual, a process that is elementary, time consuming and prone to inconsistencies and human error. Here I present a new method, and a tool - Neuronalizer, for the characterization of the morphology of neurons. I demonstrate the use of the Neuronalizer through investigation of the PVD neuron. The tool provides a fast and reproducible method for morphological characterization based on mathematical quantifications, and enables the extraction of key morphological features that cannot be extracted manually.


Integrating the computational analysis with experimental results enables the detection of subtle morphological differences between animals with different genetic, developmental and environmental backgrounds. Altogether, it provides new insights about the interconnection between neuronal structure, function and the behaviour of the animal.


Using the Neuronalizer we were able to detect and quantify novel morphological features in the PVD neuron. First, by examining the orientation of the outsets of segments in each junction, we found two types of membranal 3-way junctions' morphologies - symmetrical junctions and linear junctions. Second, we examined the PVD of mechanosensory-deprived worms and found that it exhibits activity-dependent structural plasticity and that this plasticity is partially eliminated in mutants that lack an ion channel that opens up in response to touch. Finally, we quantified the morphological changes that the PVD undergoes during development.


Altogether, the Neuronalizer provides a robust method and a user-friendly interface for tracing, visualization and analysis of neuronal images. It contains advanced analytical tools and allows the integration of new neuron-specific analysis modules.