Date of Graduation


Document Type


Degree Type



School of Medicine


Not Listed

Committee Chair

Bernard Schreurs

Committee Member

Visvanathan Ramamurthy

Committee Member

Candice Brown

Committee Member

Peter Mathers

Committee Member

George Spirou


Across many parts of the nervous system, initial neural circuit assembly involves stereotyped stages of early axonal proliferation within target territories, selective reinforcement of a subset of inputs and the removal of exuberant or unneeded connections via either activity or experience-dependent mechanisms, in a process referred to as synaptic competition. The mechanisms and associated structural transformations that underlie and define these stages of neurodevelopment have been examined in a variety of brain regions including the peripheral neuromuscular junction, the central climbing fiber innervation of cerebellar Purkinje cells and retinal ganglion cell innervation of the thalamic lateral geniculate nucleus. Each of these connections are considered representative examples of sites of synaptic competition. We and others have established the globular bushy cell (GBC) innervation of principal cells (PC) of the medial nucleus of the trapezoid body (MNTB) in the auditory brainstem via the large calyx of Held (CH) as an additional representative, or model, system for the study of terminal growth and competition.

We utilized dense reconstruction of pre and postsynaptic elements from serial block-face scanning electron microscopy (SBEM) image volumes collected during key developmental timepoints from both prehearing (postnatal days (P)2, 3, 4, 6, & 9) and adult (P30) mice. In this document, I report our most recent studies of MNTB ultrastructure, which revealed additional structural elements that define the progressive growth, competition and subsequent maturation of the CH and its innervation of PCs.

We found evidence that all developing CHs extend 3 types of collateral arbors including short filopodia (5.85 µm), and growth cone tipped collaterals of sufficient length to innervate adjacent cells or cells within 60 µm, but in some cases up to 100 µm, away from the innervation territory of the CH. Moreover, as the proportion of polyinnervated PCs with competing CH inputs declines from 23% at P6 to 11% at P9, we found a significant reduction in the length of CH collaterals until few if any were found by maturity (P30).

Using the first live imaging data collected from a living brain slice with the lattice light sheet microscope, we revealed with high spatial and temporal resolution that CH collaterals exhibited some of the fastest motility recorded in developing neural systems (average growth cone rate of motility: 2.1 µm per min). We also observed interactions among long filopodial collaterals that resulted in retraction and altered growth trajectories suggesting possible repulsive contact-mediated guidance of CH growth and innervation. These may be the predominant locations where competing CHs physically interact, since they typically expand over non-overlapping territories across the cell surface of the PC soma.

Postsynaptically, we provided the first systematic description of two types of invaginating spine (type 1 and 2) that reside specifically within recesses underneath and penetrating the developing CH. Similar to the emergence of CH collaterals, we found a strong association of type 1 spine emergence with the initial expansion of the terminal. Moreover, for the first time, somatic spine mats, composed of so-called type 2 spines that intertwine to assemble a macrostructure called the somatic spine mat, were described in the vertebrate CNS. From inspection of EM images, we suggest three modes of intercellular communication between the CH and its synaptic partner, the PC: chemical synaptic transmission, signaling via cell adhesion proteins, trans-endocytosis of spine mats into the CH.

We hypothesize that the unifying event to synchronize CH growth and emergence of collaterals and postsynaptic somatic processes is a transformation of spontaneous activity to a bursting pattern. Preliminary experiments reveal that interference with synaptic transmission using viral vector induced expression of tetanus toxin leads to morphological changes in the CH and slower growth of the postsynaptic cell body. Further experiments are proposed that combine electrophysiology and dynamic imaging with molecular and physiological perturbation to further explore these linkages.

Embargo Reason

Publication Pending

Available for download on Saturday, December 04, 2021