Verifying, Challenging, and Discovering New Synapses Among Fully EM-Reconstructed Neurons in the Leech Ganglion

Abstract

Neural circuits underpin the production of animal behavior, largely based upon the precise pattern of synaptic connectivity among the neurons involved. For large numbers of neurons, determining such "connectomes" by direct physiological means is difficult, as physiological accessibility is ultimately required to verify and characterize the function of synapses. We collected a volume of images spanning an entire ganglion of the juvenile leech nervous system via serial blockface electron microscopy (SBEM). We validated this approach by reconstructing a well-characterized circuit of motor neurons involved in the swimming behavior of the leech by locating the synapses among them. We confirm that there are multiple synaptic contacts between connected pairs of neurons in the leech, and that these synapses are widely distributed across the region of neuropil in which the neurons' arbors overlap. We verified the anatomical existence of connections that had been described physiologically among longitudinal muscle motor neurons. We also found that some physiological connections were not present anatomically. We then drew upon the SBEM dataset to design additional physiological experiments. We reconstructed an uncharacterized neuron and one of its presynaptic partners identified from the SBEM dataset. We subsequently interrogated this cell pair via intracellular electrophysiology in an adult ganglion and found that the anatomically-discovered synapse was also functional physiologically. Our findings demonstrate the value of combining a connectomics approach with electrophysiology in the leech nervous system.

Keywords: EM reconstruction; invertebrate neurobiology; leech; serial block face scanning electron microscopy; synapse.

Figure 1

Most, but not all, of the predicted physiological connections were found anatomically after reconstructing the arbors of six pairs of dorsal motor neurons. (A) Predicted circuitry based on dual electrophysiological recordings, adapted from Ort et al. (1974). Lines ending in circles represent inhibitory connections; lines ending in a T-junction indicate excitatory connections; resistors indicate non-rectifying gap junctions; diodes represent rectifying gap junctions. (B) Updated circuitry based on what was directly observed after anatomical reconstruction. Electrical connections are grayed out as these are not directly observable with serial blockface electron microscopy (SBEM). All predicted connections were found except those onto the L cell. A few unexpected synapses were found (e.g., from cell 1 to cell 102); these typically involved far fewer overall synapses (Table 1). The total number of synaptic contacts made by both the right and left pairs of neurons are shown next to each line (see also Table 1). (C) Examples of synapses between the right DI-1 and the right DE-3 (upper panel), the right DI-102 and the right DE-3 (middle panel), and the left VI-2 and the right VE-4 (lower panel). In these examples, the cells are fully segmented to display the relative scale of the participating processes; the remainder of their arbors were traced via skeletonization. Scale bars 300 nm.

Figure 2

The right DE-3 receives numerous widely-distributed synaptic inputs from the right DI-1 and none from the left DI-1. (A) The right DE-3 (blue skeleton) receives synaptic input from the right DI-1 (green skeleton) at 18 sites (red dots) widely distributed throughout the contralateral half of its arbor. Inset displays the previously-known connectivity among these three cells. (B) The left DI-1 arbor (pink skeleton) overlaps with the right DE-3 arbor. Even where the left DI-1 forms presynaptic boutons and the right DE-3 receives synaptic inputs, no synapses are found (region within black box). Scale bars 10 μm. Arbors are presented as viewed from above, with anterior to the top. Cell bodies are omitted for clarity, as their position above the arbors would partially obscure them.

Figure 3

The close apposition of cell pairs known to be electrically coupled could harbor gap junctions. Though the arbors shown were all traced by skeletonization, we fully segmented them in each particular section shown here to highlight their membrane appositions. (A) The confluence of the S cell (blue) and both coupling interneurons (pink and green). (B) Close apposition between two processes of the left DI-102 (red) and left DI-1 (yellow). (C) A small branch of the left DE-3 (orange) invades the main branch of the right DE-3 (purple). Scale bars 500 nm.

Figure 4

A synapse discovered anatomically makes an electrophysiological connection. (A) Skeleton arbors of the presynaptic S cell (blue) and postsynaptic cells 116 (green and orange) with pink dots representing sites of synaptic contact. Scale bar 10 μm . Arbors are presented as viewed from above, with anterior to the top. Cell bodies are omitted for clarity, as their position above the arbors would partially obscure them. Inset displays the connections between the S cell and cells 116 that we tested physiologically. (B) Examples of synapses from S onto the left 116 (top) and right 116 (bottom). scale bars 300 nm. Cells are fully segmented in these example sections to display the relative scale of the individual processes; the remainder of the arbors were reconstructed via skeletonization as shown in (A). (C) Example recordings from one adult nervous system preparation of the S-116 connection. Spikes were induced in the S cell in one ganglion (bottom trace) whereupon they traveled across the S cell network down the nerve cord, eliciting a reliable depolarization in cell 116 (middle trace). The S cell spike was visible in an extracellular recording of the connective nerves posterior to the ganglion containing the recorded 116, indicating that the spike successfully passed through (top trace). A single spike in the S cell is presented for clarity in the bottom trace while the middle and top represent recordings following 15 separate S-cell spikes from the same preparation (gray) and their average (black).