In another segment of our Neurovox series, Vania Cao, PhD, interviews Amar Sahay, PhD, an Assistant Professor at Massachusetts General Hospital, Center of Regenerative Medicine, Harvard Medical School. He is also a Principal Faculty at the Harvard Stem Cell Institute, and an Associate member of the Broad Institute of Harvard and MIT. He tells us about his research into the neural circuits underlying fear responses, why it’s important to study neural circuits, some of the challenges neuroscience faces, and what excites him about working in neuroscience today.
VC: Please tell us a little bit about your research interests.
AS: My laboratory’s main focus is understanding how adult hippocampal neurogenesis, which is a unique form of plasticity in the brain, contributes to hippocampal functions, in both cognition as well as the regulation of mood. And towards this goal, we’re particularly interested in the circuit mechanisms by which neural stem cells are physiologically regulated by environmental stimuli. We’re interested in the circuit mechanisms by which adult born neurons integrate into the hippocampal circuit. And finally, and importantly, we’re interested in the circuit mechanisms by which these adult born neurons contribute to encoding functions of the hippocampus.
We use a range of tools, from molecular approaches to in vivo optogenetics, optical imaging, viral tracing, and molecular genetics, as well as behavioral analysis. And we also empower our experimental approaches with theoretical interrogations trying to understand how these circuits work in the hippocampus using theoretical models.
VC: Is there a particular research topic or interest that your lab is pursuing for DECODE?
AS: For the purposes of the DECODE award, which we’re very excited to receive two years ago, we wanted to ask and test the question: “How is information that’s computed in the hippocampus, in particular as it relates to the evaluation of environmental threats, relayed to subcortical targets?” A large amount of work has gone into understanding how information processing in the hippocampus is relayed to brain regions like the amygdala and the hypothalamus, whose activation ultimately allows an organism to calibrate its fear responses. We were particularly interested in whether a specific set of neural pathways that had been previously identified contributed to this relay mechanism. Or are there other neural pathways that have been previously understudied or even identified that may be important for relaying hippocampal-dependent computations to brain regions that subserve fear.
To this end we spent a fair bit of time thinking about subcortical regions that receive inputs from the hippocampus, and based on an extensive brain-wide analysis we performed, we’ve identified a small subset of subcortical targets that receive inputs from the hippocampus, which we have also identified as exquisitely sensitive to the environment.
With the DECODE project, we wanted to use the mini-endoscopes for deep brain calcium imaging of distinct populations of inhibitory interneurons in these subcortical targets to determine the extent to which these inhibitory interneurons respond to computations made in the hippocampus as the mouse is evaluating threats in its environment. The basic idea is that if we identify the patterns of firing of calcium transients in these inhibitory interneurons as the animal is computing threat in its environment, it may edify strategies to rely on these populations of neurons as sensors and potentially predictors of the animal’s behavior…in this case, adaptive fear responses.
VC: What level of knowledge do you feel that nVista calcium imaging data provides to help you understand this whole process?
AS: In the subcortical target, the brain region that we are most interested in, it’s comprised primarily of GABAergic inhibitory interneurons that come in many different flavors or subtypes. Therefore, there is a critical need to identify how distinct populations behave as the animal is navigating its environment. And so the combination of Cre-dependent viruses encoding GCaMP with genetically modified mouse lines in which distinct populations of inhibitory interneurons express the recombinase Cre, we can selectively image different subtypes of inhibitory interneurons in these animals. It’s the combination of viral genetics, optical imaging, and the miniaturized endoscopes that allow us to look at the performance of these distinct cell types in awake behaving animals.
VC: What do you feel is important about studying neural circuits in general?
AS: Studying neural circuits to understand how the brain works in health and disease I think, as do many others, is going to be a major challenge, and a transformative step in basic science in both understanding how the brain works but also being able to relate the functions of the brain to human disease. There are several reasons for this. One of them, especially with psychiatric diseases, is that most of these psychiatric diseases have at a molecular level hundreds if not thousands of genes that are essentially culpable agencies. It’s going to be very difficult to target each of these genes, even if you figured out what they were doing, at the level of the circuit. However, we do know that genes exert their effects in the brain through their actions on circuits. We also know that many psychiatric diseases, outside of schizophrenia for example where there is a very strong heritable genetic component, have both an integral contribution of the environment and the genetic culpability that comes from mutations in the DNA.
And so it’s the interactions between genes and environment that is thought to produce vulnerability, or in some cases resilience. However, both again genes and environment converge at the level of neural circuits. If anything, it underscores the need to identify neural circuits and the functions of neural circuits, as well as neural circuit-based signatures to ultimately inform our understanding of behavior in health and disease.
Another reason we think it’s important to emphasize neural circuits is that many of these psychiatric diseases are based on the DSM, and are ostensibly extremely heterogeneous in structure. However, if one were to zoom into the neural circuit underpinnings of these diseases and specifically how different neural circuits contribute to distinct components of these different diseases, we may find ourselves identifying common shared endophenotypes at the circuit level that are conserved across different diseases. For example, in my laboratory one of the driving forces of our work and our approach is that identifying basic computational mechanisms by which the hippocampus encodes information, and keeps for example similar memories separate through neural mechanisms such as pattern separation. We think this will ultimately edify how brain dysfunction contributes to memory impairments in aging and Alzheimer’s, but also as we proposed overgeneralization of fear and PTSD. Because if you step back and think about it, it’s very intuitive the way an animal encodes its environment is the way in which the animal is likely to respond to it. And so by identifying core signatures, in this case the neural mechanisms that underlie pattern separation and pattern completion in the hippocampus, we might be able to essentially shed light on how pattern separation impairments, or pattern completion impairments, contribute to episodic memory impairments in aging and Alzheimer’s, but also as we proposed the overgeneralization of fear and PTSD. Therefore, we think studying neural circuits allows us to tackle this barrier of understanding the complexity of gene and environment interactions, but also identifying biomarkers, or signatures, that underlie endophenotypes of different psychiatric diseases.
A final point on why we think studying neural circuits to advance our understanding of diseases is important is because circuits, again unlike the function of individual genes, are more likely to be conserved when you move from mice to men. This is particularly true of the hippocampus. It is true that the prefrontal cortex and higher order cortices may not be as conserved. But a large number of limbic circuits are conserved to a high extent between rodents and humans. The hippocampus is certainly paradigmatic of this notion, and so if you identified a signature at the level of a neural circuit in terms of say for example again pattern separation and completion mechanisms in the hippocampus, we might be able to identify a biomarker in humans which is noninvasively imaged, and this may allow us to then stratify individuals for risk to a disease. In the case of PTSD it would be a vulnerability to stress. And it will allows us to objectively monitor therapeutic approach’s efficacy non invasively. And so these are the three reasons that we think studying neural circuits is going to, or is already catalyzing both our understanding of disease, but also is going to illuminate therapeutic strategies for targeting complex psychiatric diseases.
VC: Given all these reasons why studying neural circuits is important, what challenges do you see for the field to really make some inroads?
AS: There are several challenges that I see as necessary to tackle to advance our understanding of psychiatric diseases using this neural circuit framework. One of them is reproducibility. I think it’s important for the biomedical science field to be able to essentially reproduce important findings so that there also is an interest from industry to co-invest with academia to further disease research. And by one way this could be accomplished is incentivizing reproducibility. Another challenge I see is, at least in using rodent models, to understand human disease is the contribution of genetic backgrounds. And again I think science researchers should be incentivized to pursue observations that we made in one strain, for example, in other strains, and including rats. And it’s true for genetic mutations, but I think it’s likely to be true for requirements for circuit dissection studies where the dissection of a specific circuit in one strain may not apply in another strain or in a rat.
The third challenge I see is going to require a lot more effort, which is scalability. For example, we know that even though most circuits are homologous between rodents and humans, there is tremendous heterogeneity. This is very evident in rodents. This heterogeneity is likely to be highly magnified in humans, and so the question then becomes: “how are we going to be able to resolve at the cellular level, as we can do now in rodents, in humans?” Where we know the genetic cell type, so as to then resolve the heterogeneity, to then infer how the circuit functions as the human is performing a specific task. I think the scalability question may require non-invasive imaging tools in humans, and also a way by which we can genetically identify cell types. Together, this combination, as is happening with rodents, may allow us to generate more tractable approaches to dissect circuit function in humans.
Actually, a final point that I’ll add that is relatively recent but I think is going to gain a lot of traction is the idea of humanizing mouse circuits. Here the premise is that we can now generate a range of neuronal subtypes from a specific individual’s own skin or through the generation of induced pluripotent stem cells, or transdifferentiation of fibroblasts. This allows us then to generate human neurons from a specific individual. One of the challenges is going to be how do these neurons contribute to human circuits, and a way to do that would be to humanize mouse or marmoset circuits. In the case of marmosets, there may be some ethical challenges that we’re going to have to address also upfront. Together, all of these approaches I think will facilitate the bridging of a potential gap in our insights from rodent studies and human studies.
VC: What do you find to be exciting about being in neuroscience and studying neural circuits today?
AS: I’m very excited and fortunate, I think, to be a neuroscientist today, in this era, because I do think that an understanding of the brain is going to change the future of our world as we know it. Our understanding of how the brain forms new memories, how memories can change, how we can update memories, how we can change their valence–a lot of this has already been realized. We still have a ways to go but I think the way technology is maturing, and the innovation that we see in neuroscience, it’s very likely that it’s going to have a profound effect on the quality of human life. But also in the way even healthy humans experience the world that we live in. I see this all the time when I talk about the work that we do to nonscientists, and the ease with which non scientists can easily relate to the significance of our understanding of the brain. I see this as having a profound effect on not just the health space, but literature, poetry, art, all of which is dependent on the way we think about and experience our world.