Meet Henry Gustav Molaison, a man who saw his doctor every day for many years and introduced himself every time as if they had never met [1].Molaison suffered from anterograde amnesia, the inability to form new memories [1]. How did he come to have this strange condition? In 1953, Molaison, who was previously an epileptic, underwent experimental surgery that removed his hippocampus [1]. The surgery did cure his epilepsy, but also removed his ability to remember new information. From Molaison, we learned that the hippocampus is the region of the brain responsible for declarative memory. For years, the classic memory pathway of the hippocampus has involved a 4-neuron trisynapticcircuit: the entorhinal cortex II (EC II) à the dentate gyrus àthe CA3 region of the hippocampus à the CA1 region of the hippocampus [2].This circuit was the basis for how declarative memories are made and stored in the brain. A recent study publishedin Naturechallenges this classical view of the hippocampus by emphasizing the role of the overlooked CA2 region.
The CA2 region, located between CA3 and CA1, has been poorly characterized .The classical definition of the CA2 region includes two criteria. First, CA2 neurons lack the dendritic spines found on CA1 and CA3 neurons [2]. Spines are highly convoluted regions of the dendrite plasma membrane which create a higher surface area-to-volume ratio and thus more room for synapses. Second, CA2 neurons do not receive input from the dentate gyrus [2], which is thought to be a gatekeeper for information flowing into the hippocampus. This lack of input suggests that the CA2 region plays an insignificant role in memory.
This studynot only provides a new molecular definition of CA2 neurons, it also indicates the presence of an influential synaptic connection between the dentate gyrus and CA2 neurons whereas previously no connection was thought to exist at all.
The new definition of CA2 neurons of the hippocampus uses three molecular markers as its criteria [2]. The authors noted three genes expressed in all CA2 neurons, RGS14, PCP4, and STEP,and used immunohistochemistry to visualize expression of these genes on a slice of the mouse hippocampus. The authors found that the three genes’ rate of coexpression, that is, the degree to which their expressions overlap, was almost 100%, making them ideal markers for CA2 neurons [2].
Once these three marker genes were verified, the authors set about comparing the CA2 neurons defined by the new marker genes to the CA2 neurons defined by classical anatomical criteria – namely the absence of dendritic spines. They used diolistic labeling to visualize the dendrites of the newly defined CA2 neurons[2]. Diolistic labeling, a type of gene gun therapy, is a dye-labeling technique in which target DNA attached to fluorescent particles is propelled into cells on a brain slice[3]. The authors visualized 43RGS14-positive CA2 neurons in mice brains sand found that all 43lacked complex spines on their dendrites [2]. They then tested RGS14-negative CA3 neurons and found that 29 of 30 did indeed have spines[2].Therefore, CA2 neurons defined by the molecular definition correspond to neurons defined by classical anatomical criteria.
I would like to highlight the almost perfect correlation, highlighting the specificity of these genes markers and suggesting that these markers can be used to define CA2 neurons in the future.The nature of these markers also lends themselves to a wider variety of applications. While in the past CA2 cells could be defined only by the presence or absence of spines, CA2 neurons can now be defined by their protein, DNA, or RNA labels, which makes it possible to perform procedures such as coimmunolabeling. Discovery of these molecular markers will facilitate research on the CA2 region of the hippocampus and is an important step in further exploring the memory pathways of the brain.
The other big finding in this study is the presence of a connection between dentate gyrus mossy fibers and CA2 pyramidal cells [2], which directly refutes the classical criterion of CA2 neurons that no such connection exists.To do this, the authors used a technique known as viral tracing [2], a technique which injects a virus into a neuron to map neural pathways [4]. As the virus replicates, the entire path that the virus travels can be visualized[4]. Viral tracing was successfully employed to confirm that mossy fibers of the dentate gyrus do project to PCP4-positive CA2 neurons[2].
Further evidence of this connection was provided via immunohistochemistry. ZnT3 and VGluT1, two proteinmarkers in presynaptic vesicles of mossy fibers in the dentate gyrus, were also found in RGS14-positive dendrites of CA2 neurons[2]. The authors also found that clusters of ZnT3 overlapped with a known scaffold protein, PSD-95, on RGS-14-positive CA2 neurons[2]. These findings provide strong evidence of a previously unknown connection between the dentate gyrus and the CA2 region[2].It is important to note that without the CA2 neuron molecular markers previously discussed, this finding would not have been possible. Thus already we can see applications of the new definition of CA2 neurons.
Another significant finding was the strength of this newly discovered dentate gyrus – CA2 connection [2]. In general, neurons that synapse closer to the body of the target neuron will have a larger impact on neural activity. This is because the postsynaptic current generated at the synapse will have less distance to travel, thus allowing less leakage. Similarly, neurons that synapse farther from the cell body will have less impact on neural activity.The authors performed an experiment in which this rule held true.Optogenetic stimulation of a dentate gyrus mossy fiber, which synapses close to the CA2 neuron cell body, resulted in an excitatory postsynaptic current of -122 pA [2]. Meanwhile, stimulation of a distal CA3 synapse resulted in a current of -40 pA in the CA2 neuron [2], showing that the dentate gyrus à CA2 connection is much stronger than the CA3 à CA2 connection. Other studies confirmed the strength of this dentate gyrus à CA2connection and found that this connection was stronger than two other circuits well-known for their influence on behavior, the entorhinal cortex IIIàCA1 and dentate gyrus àCA3[5]. These results suggest this linkage could have a large impact on memory.
Then the question becomes, what is the nature and function of this connection? Prior research indicates that CA2 neurons project to CA1 neurons [6]. With the knowledge that the dentate gyrus projects to CA2 neurons, a new trisynaptic circuit emerges which runs parallel to the classic DG à CA3 à CA1 circuit [2].Thisalso raises many questions: To what degree do these two circuits work together, or do they even work together? How do they influence one another? One suspected function of the CA2 region is prevent “overlearning,” as its synapses on CA1 neurons are inhibitory[2]. It is possible that the new DG à CA2 à CA1 circuit plays a role in how memories are suppressed and eventually forgotten. Another role of the CA2 region has been proposed in relation to its connection to the dentate gyrus. The dentate gyrus has been implicated in pattern separation, the ability of the brain to distinguish objects and memories from each other [7]. This then suggests that the dentate gyrus – CA2 linkagemay play an important role in separating the new elements of a memory from those we are already familiar with [8].
Research into the poorly characterized CA2 region of the hippocampus has just begun. Much work remains to be done on discovering the relation of the CA2 region to other partsof the memory pathway. The findings of this study not only provide a direction for future research, but also facilitate this work through the discovery of gene markers.
