Embedded Files
Computational models of neural system development
Computational models of neural system development
Publications:
Khakhalin 2019 - I use high-speed calcium imaging to reconstruct connectivity within small subnetworks in the tectum. I then analyze this connectivity graphs using methods from network science, and verify these analyses in a transparent developmental model. I show that if STDP (spike-time-dependent plasticity), intrinsic plasticity, and synaptic competition are allowed to roam free, and the network is exposed to realistic visual stimuli, it spontaneously develops looming selectivity, mediated by a set of resonant synfire chains (neurons, connected in the same order, in which they are activated in response to stimuli of interest).
Github repo for this project (data files, analysis programs, model files)
Jang, ... Khakhalin 2016 - We show that a realistically tuned spiking model of the optic tectum is naturally selective to looming stimuli, but in the absence of inhibition this selectivity can be easily disrupted by neuronal noise.
Multidimensional tuning in biological networks
Multidimensional tuning in biological networks
Neurons in the brain can be classified into several different subtypes. Within each subtype, cells are relatively alike to each other: they have similar shapes, similar global connections, release same neurotransmitters, and clearly work together to process information. Yet while they are similar, they are of course not exactly identical. In a way, they have to be different, to contribute something unique to the overall working of the network. By looking at how individual neurons differ from each other, we can hope to understand more about how they work together.
My research shows that neurons are more similar to each other in younger brains, but become "fancier" and diverse later in development, as their tuning gets nuanced and unique. Moreover, different sensory experiences uniquely shape neuronal properties, making cells either more, or less diverse, and tuning them to different types of sensory stimuli.
In my 2019 paper, we describe a novel way in which different neurons in the same brain area are uniquely tuned to their inputs. It turns out that even neurons that don't usually engage in oscillations, and don't support periodical firing, are often temporally tuned, and prefer either fast (synchronous) or slow (asynchronous) synaptic inputs. Moreover, these preferences are plastic, as they depend on the environment the animal is exposed to. This new finding shows that the tuning of neurons within brain networks is even more multifaceted than we used to think!
Publications:
Busch Khakhalin 2019 (bioRxiv preprint); published in J Physiol
Github repo for this project (data files, analysis programs)
Ciarleglio and Khakhalin, .. Aizenman 2015 (eLife; co-first author)
Resources:
Full data set for the 2015 eLife paper: available at Dryad
Neural mechanisms of escape and startle behaviors
Neural mechanisms of escape and startle behaviors
In this series of papers I try to understand how behaviorally relevant visual stimuli (such as a predator approaching you from the depth, or a snowball flying into your face) are encoded, processed, and identified by the brain, and how the brain generates useful motor outputs (movements that help you to avoid both the predator, or the snowball) in response to these stimuli. In both humans and tadpoles, collision detection mostly happens in the midbrain structure called the optic tectum (known as the superior colliculus in mammals). Motor outputs in the tadpole are generated by the hindbrain networks, and are then supported by the central pattern generators in the spinal cord.
To figure out how the collision avoidance system works in a tadpole I rely on a series of sequentially reduced preparations: from intact animals swimming in a virtual environment, through in vivo recordings in partially immobilized animals, and to in vitro whole brain preparations.
Publications:
Press:
Resources (videos) for the 2014 paper:
Experiments in partially immobilized animals ("fictive avoidance" or "tailflicks")
A set of sample visual stimuli used in the study
Multisensory integration in tectal and hindbrain networks
Multisensory integration in tectal and hindbrain networks
In this collaborative project with the group of Carlos Aizenman at Brown we look into how information from the visual and the auditory systems is integrated in the tadpole brain.
Publications:
Truszkowski ... Khakhalin, Aizenman 2017 - Behavior, Ca imaging, whole-cell ephys
Felch, Khakhalin, Aizenman 2016 - cell-attached and whole-cell ephys
Press:
For the 2016 study: Brown, Science Daily, Eureka Alert
Xenopus tadpole as a model for the study of neurodevelopmental disorders
Xenopus tadpole as a model for the study of neurodevelopmental disorders
As tadpoles are developing externally from an egg to an independent organism in a matter of just about two weeks, this model is particularly inviting for any kind of neurodevelopmental research. Xenopus is also closer to humans evolutionary than some the "competitive" small animal models, such as invertebrates (e.g. fruitflies) or Zebrafish. It makes it relatively easy to use it as a model for human neurodevelopmental disorders, as most of the molecular mechanisms underlying these disorders in humans seem to be conserved between Homo and Xenopus genera (between humans and frogs).
Publications:
Khakhalin et al. 2020- a pre-print collection of detailed protocols for behavioral assays in Xenopus. Some parts of it are now published as peer-reviewed pieces, but some will remain there as a citable protocols collection
James, ... Khakhalin, Aizenman 2015 (pdf; co-corresponding author)
Press:
Resources:
The figure "Xenopus tadpole as a model" (the one on the right) is available in vector Adobe Illustrator format, and rasterized PNG format (2000x1457) under CC-BY (Creative Commons Attribution) license. Please cite Pratt Khakhalin 2013 if you use this figure in your work.
Maturation and development of inhibitory networks
Maturation and development of inhibitory networks
GABA is the most important inhibitory neurotransmitter in the vertebrate brain. Yet very early in development GABA doesn't act quite as inhibitory as it does in adult brains. Before birth, and, in most mammals, for some time after birth GABA may depolarize cells, and even cause them to spike. The current assumption is that this early depolarizing action of GABA is important for neural system development.
One of the beautiful features of the Xenopus tadpole as an experimental model is that its brain continues to produce and incorporate new neural cells even as the animal freely swims around, feeding, and avoiding predators. At these stages new cells are born at the caudal and dorsal edges of the optic tectum.
It turnes out that GABA action in tectal cells follows same temporal maturation pattern that is known from studies in mice, rats and humans. In tadpoles, as in mice, GABA is more depolarizing in younger animals, and more hyperpolarizing in older ones. In Xenopus tectum, however, this temporal profile is superimposed onto a spatial one, as GABA is more depolarizing in caudal and dorsal parts of the tectum (where younger cells are located), but is more hyperpolarizing in older cells, located more rostrally and ventrally. It means that different parts of the tadpole's brain do indeed have different functional "ages"!
Publications:
Protocols:
Google Sites
Report abuse