More evidence that dendritic structures matter for brain preservation
How cutting off parts of neurons using lasers helps tells us how neural information is stored
Park 2019 is an excellent paper on how pyramidal cells integrate dendritic inputs, which Ken Hayworth recently summarized in an Aspirational Neuroscience Prize journal club video.
I wanted to highlight it here because I think it’s helpful for grounding discussions about brain preservation in actual results from neurobiology about how information is encoded.
What did this study find?
In this paper, the authors studied pyramidal neurons in layer 2/3 of the primary visual cortex of the mouse. They looked at how the removal of different dendritic branches affected the information-processing properties of these neurons.
A key distinction in neurobiology is between the apical and basal dendrites. Apical dendrites emerge from the apex of the soma of a pyramidal cell, while basal dendrites emerge from its base:
In this study, the authors developed a system for dissecting away dendritic branches from the soma using a laser:
They found that removing the apical dendrite of a layer 2/3 pyramidal neuron in the mouse's primary visual cortex had little effect on the cell's overall visual orientation tuning.
“Tuning” in neurobiology refers to a cell’s electrophysiologic selectively for a particular type of stimuli.
So in other words, the neurons they recorded from pre- and post-apical dendrite ablation were still selective for the same orientations of visual stimuli. Here’s an example:
It’s remarkable to me that any neurons even survive the ablation of apical dendrites, which provide around 40% of the neuron’s synaptic inputs — only about half did — let alone survive with the same type of information processing properties.
While there is only one apical dendrite branch per neuron, there is a variable neuron of basal dendrite branches. On average, the cells they studied have around six main basal dendrite branches.
Removing two, but not just one, of the cell’s basal dendrites resulted in a small but significant change in the cell's overall visual orientation preference:
They also tried to remove three basal dendrites, but the all neurons died in this case, so they couldn’t test the effects of that intervention.
Based on modeling results, the authors conclude that there is a degree of heterogeneity of tuning across basal dendrites.
This heterogeneity may arise from dendrite-specific forms of plasticity, potentially mediated by spatially restricted biochemical signaling and/or dendritic spikes.
What are the implications for brain preservation?
In brain preservation, our goal is to preserve enough structural information that mediates the functional properties of cells. These results show that even the removal of a large percentage of a neuron's inputs does not necessarily result in a loss of function.
It’s hard to argue that preservation quality has to be perfect when in vivo results like this show that information encoding in neural circuitry is in fact quite redundant and robust.
These results also underline the well-known phenomenon that dendrites play an important role in computation. This makes dendritic structures and morphology essential to preserve. It’s an example of why it’s I think it’s effectively not possible that preserving each cell’s DNA or epigenetics alone — without morphological information — would be sufficient to preserve memories.