Exploring the Cephalopod Brain: Towards Mapping the First Whole Brain Connectome and Whole Animal Projectome of a Pygmy Squid

Abstract

Cephalopods have arguably the most complex central nervous system of all studied invertebrates. They possess large, highly centralised brains for processing sensory information, higher cognitive functioning, motor control, and feeding. Cephalopods are known for their ability to solve problems in unique ways, displaying idiosyncrasies in prey/predator behaviours, courtship and conspecific signalling via dynamic colourful patterns in their skin, which can be used for their famous camouflage abilities as well. Additionally, these animals demonstrate higher-order abilities, such as learning from observation and tool manipulation. Comparisons with vertebrates are common, not only due to their remarkable intelligence but also because of their anatomically and functionally similar eyes, whose overall morphological configuration is considered evolutionarily convergent with the vertebrate eye. Despite interest in cephalopod brains dating back to the 19th century, very little is known about how they process visual information. As structure tightly constrains function, this thesis explores the first steps in obtaining multi-scale connectomic atlases for the cephalopod optic lobes and brain. To obtain synaptic-resolution volumes of neural tissue, the primary technique used was Volume Electron Microscopy (vEM), specifically Enhanced Focused Ion Beam Scanning Electron Microscopy (eFIBSEM) and Multibeam Scanning Electron Microscopy (Multibeam SEM). This study focused on Idiosepius hallami hatchling, which belongs to the family of pygmy squids, the smallest known cephalopods, of dimensions suitable for whole-brain volume EM. Using eFIBSEM, small volumes of the eye and optic lobe were acquired to test the feasibility of mapping brain wiring, dependent on sample preparation optimisation. The eye and optic lobe were chosen for analysis given abundant literature on visual systems across phyla , with vision being the most studied sense, as well as the ease with which the optic lobes can be localised, and their presumed organisation in a mosaic repetition of circuit units in two dimensions that facilitate the identification of cell types. From the first volume obtained, photoreceptors and interneurons were manually traced, demonstrating how different cell types are distributed in the first layers of the optic lobe. This study presented, for the first time, direct connections from photoreceptors to outer granule layer neurons and centrifugal inner granule layer neurons, revealing that photoreceptors can be classified into seven different types depending on these direct connections. By digitally segmenting this volume by hand and then rendering neuropils and neurons in 3D, I was able to unveil, for the first time, that these neurons pierce through the axonic boutons of photoreceptors to collect synaptic inputs. Sample fixation and embedding were iteratively optimised, enabling supervised automatic segmentation algorithms, which are essential for a whole-brain connectome of 0.25 mm3. At the time of writing this thesis, a whole-brain volume at a resolution of 4440 nm (XYZ) is being acquired using a Multibeam SEM. This volume will be the first high-resolution electron microscopy volume of a whole brain of a cephalopod, serving as an invaluable reference volume for the cephalopod neuroscience community, and a starting point for comparisons across cephalopods and beyond, particularly for neural circuits for vision in other phyla. To frame the brain and its nerves in the context of the whole body, a high-resolution volume (125 nm isotropic) of the whole pygmy squid was acquired at the ESRF synchrotron beamline ID16A. This was the largest volume acquired at this resolution at the synchrotron beamline to date. At this resolution, although individual neuron cells and synapses cannot be distinguished, it is possible to follow the nerves housing the axonal projections between brain lobes and between the brain and body organs. Brain lobes in the entire animal volume were manually segmented, showing afferent and efferent nerve tracts to the body, such as the arms and chromatophores. This is the first high-resolution image dataset of a whole cephalopod body, to complement the whole brain connectome. Future work will focus on analysing the upcoming dataset from the Multibeam SEM and the existing beamline volume, correlating both to study how the cephalopod brain integrates inputs from all its body and controls its limbs, fins, syphon and chromatophores.