Churchland lab

 By S. TANABE, A. ZANDVAKILI, A. KOHN;  Neurosci., Albert Einstein Col. of Med., Bronx, NY

This group tackled a long-standing question in the field of decision-making: how do you tell the degree to which a single neuron “weighs in” on a behavioral choice? The question is important, but hard to get at via traditional single cell recording methods, especially for a fine discrimination task as the authors used here. 

To get around this problem, this group recorded from 15-30 neurons in V4 while well-trained subjects made decisions about the orientation of a stimulus. As predicted, they didn’t find a strong relationship between the firing rates of single neurons and the subjects’ choices, but things changed dramatically when they looked at the population level and used a linear classifier. What might account for this? The authors argue that in high dimensional state spaces, the decision axis and variability axis are aligned. This means that even if a given neuron is a key player in a decision, the relationship between its firing and the animal’s choice might be weak.  

At the end of the talk, the authors suggested a “trick” for evaluating the degree to which decision and variability axes are aligned: they shuffled the trials and found that sometimes the classifier did better! The effect of trial shuffling on the classifier, they argue, offers insight into the weighting profile of the neurons. I haven’t heard of anyone taking this approach before- it will be interesting to test on other datasets, especially those on coarse discrimination tasks where the prediction differs.

I am pleased to announce that my blog was selected as one of the official blogs for the Society for Neuroscience Meeting this year. I will focus on two themes: “Sensory systems and behavior” and “Cognition and Behavior”. You can also follow me on Twitter at @anne_churchland.

This is exciting news as there will be a wealth of fantastic science to talk about at the meeting. But in the meantime, I’ve been arguing with students at Penn about the importance of  an orthogonal representation of task parameters in the parietal cortex. They are a tough crowd and had many great questions after my talk today. In this photo, we are discussing the science at the edge of a pretty pond behind the biology building. Marino Pagan from Nicole Rust’s lab had some interesting ideas about new classification analyses we should try inspired by analyses they undertook in their recent paper.

Last week I attended the Champalimaud Neuroscience in Lisbon, Portugal. I heard many fantastic talks, including those from Susana Lima, Dora Angelaki, Michale Fee and Matteo Carandini. I also got updates from many investigators with labs at the Champalimaud, a number of whom are thinking deeply about body movements: how to track them and what they tell us about underlying neural processes. I spoke with Megan Carey and some of her team who are investigating how sensory inputs are processed differently in the brains of moving versus stationary animals.

I also spoke in detail with Joe Paton whose lab has been tackling questions about how future decisions affect current movements. This approach builds on an existing body of work suggesting that a signature of developing decisions is sometimes evident in premotor areas, and even in the movements themselves. The animal’s in Joe’s lab are freely moving so getting handle on their complex full-body movements is a challenge. The standard approach is to track one or two parameters that might turn out to be important- head angle, for instance. Thiago Gouvêa along with Asma Motiwala, graduate students in Joe’s Lab, came up with an approach that is fundamentally different: rather than trying to guess what the right body parameter might be, they image the whole animal and then reduce the
dimensionality of the large collection of images that results (see below, and also this video).

This analysis will inform them which dimensions are the right ones. Because the approach doesn’t commit the investigator to a particular movement parameter, it allows for the fact that animals might be different from each other, or that a single animal might change over time: for instance, early in training showing anticipatory movements that are suppressed when the animal is an expert at the task. This is a new project, but has the potential to lead to a novel method of evaluating movements during decisions. Down the line, the movements could be related to neurons in different parts of the brain, and could help to interpret how those neurons contribute to a developing choice.

Cold Spring Harbor’s historic undergraduate research program continued this summer as 25 students from around the world joined us for 10 weeks. As co-directors of this program, Mike Schatz and I have the pleasure of seeing these students tackle challenging scientific problems and report their findings to their peers. Like last year, we were surprised both by the progress that the students made on their projects, and also by the way their scientific thinking changes over the course of the summer.

One project that particularly interested me this year was undertaken by Gregory Fuller, a student from Johns Hopkins who worked in the laboratory of Dr. Z. Josh Huang. Josh is one of the world’s experts on inhibitory neurons, and he has a particular interest in a class of such neurons called Chandelier Cells, named for their beautiful morphology. Like many neurons, numerous Chandelier Cells die over the course of development, mostly by a programmed cell death known as apoptosis. But Chandelier Cells are unique in that they tend to die later in development than most inhibitory neurons. Greg wondered, why do Chandelier Cells die so much later than other cells? And what will happen if their programmed cell death is delayed- will that have consequences for the circuit?

Greg’s project took advantage of techniques that make it possible to block cell death: specifically, he studied brains from knockout mice that lack the proteins required for apoptosis. He closely examined tissue from different parts of the cortex, with a focus on layers 2 and 6 where the Chandelier Cells proliferate. One surprising observation was that perturbing the programmed cell death in Chandelier Cells not only caused them to proliferate, but also caused the appearance of some small cells that didn’t look much like Chandelier cells at all. Greg’s project is still at an early stage, but his observations are intriguing. Might the small cells be Chandelier cells that didn’t develop properly? Or are they a different class of neuron that was only made apparent through Greg’s manipulation? More work is necessary to get to the bottom, but the implications could be profound. Chandelier Cells likely play a key role in shaping cortical activity, and understanding how they are integrated into circuits during development may provide major insights into how they accomplish this.

I gave a talk this week at a European Science Foundation workshop about noise in decision-making. It took place at Sant Fruitos des Bages in Catalonia, Spain. The workshop was notable in a number of respects. First, for someone like me who thinks a lot about neural variability, it was great to be among so many like-minded folks (I wore my VarCE t-shirt with pride). Second, the organizers of the meeting did an unusually good job bringing together some great women who work on this topic: Hendrikje Nienborg , Laura Busse, Sophie Deneve, Tania Pasternak, Catherine Tallon-Baudry, Satu Palva, Petra Ritter, Georgia Gregoiou, and a postdoc from Marlene Cohen‘s lab. One of the most interesting talks from my point of view was presented by Encarni Marcos from Paul Verschure’s lab at the UPF in Barcelona. She described work from a recent publication of hers in which they calculated a quantity called the Variance of the Conditional Expectation (also known as the VarCE). Mike Shadlen and I (along with some co-authors) defined this quantity in our 2011 paper about variance in parietal cortex. Encarni used the VarCE in the service of a different goal: she tested whether a signature of a known trial history effect might be evident in the VarCE. To this end, she calculated the VarCE of dorsal premotor neurons recorded during a countermanding task where subjects sometimes have to cancel a planned movement. Her main finding was that trials that were just after a “stop” trial were highly variable: some had a much higher-than-average firing rate, and some a much lower-than-average firing rate. Encarni’s dataset is particularly intriguing because the two conditions she compared had nearly identical firing rate means. By examining VarCE, she was able to uncover a neural mechanism that would have been invisible using traditional analyses.

This week, my lab members and I attended a mini symposium at our own Cold Spring Harbor Lab. Organized by colleagues Tony Zador and Adam Kepecs, the focus of the day was understanding how the cortex and striatum interact to guide behavior. The speakers were Rui Costa, Linda Wilbrecht, Bernardo Sabatini, Anatol Kreitzer, and Tony Zador.

Linda Wilbrecht presented data aiming to understand the “currency” that the brain uses to make decisions. The talk included data from her recent Nature Neuroscience paper about behavioral effects of stimulating each of two classes of striatal dopamine neurons (D1-expressing and D2-expressing neurons). The first finding is that the two classes exerted opposite effects: stimulating D1 neurons caused more choices in one direction, while stimulating D2 neurons caused more choices in the opposite direction. The really interesting part, though, is that she found the effect of stimulation interacted with the animals’ reward history: for example, when stimulating neurons that caused more leftwards choices, she found the effect was enhanced if the animals had just experienced a few failed trials on the left side. WIthout stimulation, they would rarely go left after so many failures. But with the stimulation, they overcame this reluctance, and went left anyway. Understanding how recent reward history and current sensory information interact might one day give us insight into why the cycle of addiction is so hard to break.

Bernardo Sabatini described his experiments that aim to understand role of the direct versus indirect pathways. He stimulates each pathway optogenetically and looks at the effects on both behavior and firing rates of neurons in the motor cortex. He finds that the two pathways are not equal in the degree to which they affect firing rates and behavior. To visualize firing rates of motor cortex neurons, which are notoriously dynamic and complex, he reduced the dimensionality of the data using PCA. This analysis makes it possible to tell whether stimulation simply modifies an ongoing trajectory in high-dimensional space, or whether the stimulation drives the network to a novel state. The ability to perturb motor cortex and watch the resulting neural activity affords insight not just into striatal projections, but into developing motor commands as well.