Pretty exciting (although not directly related to consciousness). Some very interesting nuggets in here.
The Demise of the Synapse As the Locus of Memory: A Looming Paradigm Shift?
Synaptic plasticity is widely considered to be the neurobiological basis of learning and memory by neuroscientists and researchers in adjacent fields, though diverging opinions are increasingly being recognized. From the perspective of what we might call “classical cognitive science” it has always been understood that the mind/brain is to be considered a computational-representational system. Proponents of the information-processing approach to cognitive science have long been critical of connectionist or network approaches to (neuro-)cognitive architecture, pointing to the shortcomings of the associative psychology that underlies Hebbian learning as well as to the fact that synapses are practically unfit to implement symbols. Recent work on memory has been adding fuel to the fire and current findings in neuroscience now provide first tentative neurobiological evidence for the cognitive scientists' doubts about the synapse as the (sole) locus of memory in the brain. This paper briefly considers the history and appeal of synaptic plasticity as a memory mechanism, followed by a summary of the cognitive scientists' objections regarding these assertions. Next, a variety of tentative neuroscientific evidence that appears to substantiate questioning the idea of the synapse as the locus of memory is presented. On this basis, a novel way of thinking about the role of synaptic plasticity in learning and memory is proposed. ...
Now, when we are looking for a mechanism that implements a read/write memory in the nervous system, looking at synaptic strength and connectivity patterns might be misleading for many reasons. Most pressingly, as Gallistel and King point out, synapses might already be too complex in terms of implementing such a very basic function:
In the final analysis, however, our skepticism rests most strongly on the fact that the synapse is a circuit-level structure, a structure that it takes two different neurons and a great many molecules to realize. It seems to us likely for a variety of reasons that the elementary unit in the memory mechanism will prove to be a molecular or sub-molecular structural unit. (
Gallistel and King, 2009, p. 282)
Hence, they suggest turning to DNA and RNA, which already implement the functionality of a read/write memory at the sub-molecular level. Interestingly, in discussing recent work on memory, Poo et al. reach a similar conclusion when they remark that “[…] some other mechanisms, potentially involving epigenomic modifications in engram neurons, appear to be necessary for memory trace storage” (
Poo et al., 2016, p. 8).
A mechanism as essential as memory has to be efficient in all respects, be it implementational complexity or energy efficiency. Another part of Gallistel and collaborators' argument for the point of view they put forward is the observation that neural computation is demonstrably incredibly fast, therefore making it much more likely that the memory mechanism is (sub-)molecular in nature so that computational machinery and memory can be located in close physical proximity in order to minimize the distance over which a signal has to be transmitted (a process which evidently is “slow” in the nervous system in comparison to, for example, conventional computers). ...
To this day, tentative evidence for the (classical) cognitive scientists' reservations toward the synapse as the locus of memory in the brain has accumulated. A lot of groundbreaking work concerning the way in which the brain carries forward information in time was actually performed on comparatively simple model organisms such as
Aplysia and has then been extrapolated to speculate about what might be going on in human mind/brains (e.g.,
Kandel and Siegelbaum, 2013). Interestingly, it is recent work in this exact domain which has indicated that the idea of synaptic conductance as the basic memory mechanism is insufficient and incomplete at best.
In Kandel and collaborators' by now classic work with
Aplysia, changes in synaptic conductivity were shown to alter how the animal reflexively responds to its environment. But not even in
Aplysia all synapses are equally susceptible to change, many appeared not to be very plastic (
Kandel and Siegelbaum, 2013). Recent work with cultured
Aplysia motor and sensory neurons by
Chen et al. (2014) has revealed that long-term memories appear to persist covertly in cell bodies and can be restored after synapses have been eliminated. Long-term memory persisted after pharmacological elimination of synapses that had been produced only after learning had occurred, calling the role of synapses as the presumed locus of memory into serious doubt.
Similarly and possibly even more convincing, in a groundbreaking study bulding on earlier work (
Hesslow et al., 2013) that already pointed to the mismatch between LTD in Purkinje cells and cerebellar learning,
Johansson et al. (2014) investigated how the response of Purkinje cells changes during learning. Studying eyeblink conditioning, they showed that the cells could learn the temporal relationship between paired stimuli during conditioning. Strikingly and in stark contrast to widespread belief, the timing of responses exhibited by conditioned Purkinje cells after conditioning did not depend on a temporally patterned input. Consequently, Johansson et al. conclude that both, timing mechanism and memory trace, are located within the Purkinje cell itself. As they put it, “[…] the data strongly suggest that the main timing mechanism is within the Purkinje cell and that its nature is cellular rather than a network property” (
Johansson et al., 2014, p. 14933).
Lastly, in a recent study supposed to demonstrate the increase in synaptic strength and density of dendritic spines during memory consoldiation,
Ryan et al. (2015), to their own surprise, showed that changes in synaptic strength are not directly related to storage of new information in memory. In accordance with the literature on memory consolidation, Ryan et al. found that injection of protein synthesis inhibitors induced retrograde amnesia, meaning that the memory could not be retrieved. However, when optogenetically activating the neurons previously tagged during the conditioning process, memories could nevertheless be retrieved despite chemical blocking, indicating that the formation of synapses or strengthening of synaptic weights is not critical to memory formation as such. ...