• Kubska Z.R., Kamiński J. (2021). How Human Single-Neuron Recordings Can Help Us Understand Cognition: Insights from Memory Studies. Brain Sci., 11(4):443. doi: 10.3390/brainsci11040443.
• Kamiński, J., & Rutishauser, U. (2020). Between persistently active and activity-silent frameworks: novel vistas on the cellular basis of working memory. In Annals of the New York Academy of Sciences (Vol. 1464, Issue 1, pp. 64–75).
• Kamiński, J., Brzezicka, A., Mamelak, A. N., & Rutishauser, U. (2020). Combined Phase-Rate Coding by Persistently Active Neurons as a Mechanism for Maintaining Multiple Items in Working Memory in Humans. Neuron, 106(2), 256-264.
• Kamiński, J., Mamelak, A. N., Birch, K., Mosher, C. P., Tagliati, M., & Rutishauser, U. (2018). Novelty-Sensitive Dopaminergic Neurons in the Human Substantia Nigra Predict Success of Declarative Memory Formation. Current Biology, 28(9), 1333-1343.
• Kamiński, J. (2017). Intermediate-Term Memory as a Bridge between Working and Long-Term Memory. The Journal of Neuroscience, 37(20), 5045–5047.
• Kamiński, J., Sullivan, S., Chung, J. M., Ross, I. B., Mamelak, A. N., & Rutishauser, U. (2017). Persistently active neurons in human medial frontal and medial temporal lobe support working memory. Nature Neuroscience, 20(4), 590–601.

• Kubska Z.R., Kamiński J. (2021). How Human Single-Neuron Recordings Can Help Us Understand Cognition: Insights from Memory Studies. Brain Sci., 11(4):443. doi: 10.3390/brainsci11040443.

• Kamiński, J., & Rutishauser, U. (2020). Between persistently active and activity-silent frameworks: novel vistas on the cellular basis of working memory. In Annals of the New York Academy of Sciences (Vol. 1464, Issue 1, pp. 64–75).

Working memory (WM) is an ability to hold and manipulate a small amount of information in the mind. It is a fundamental human cognitive capacity crucial for mental calculation, inference, planning or decision making. Yet our understanding of neural mechanism of this function in humans is very limited. In this project I propose to use the unique opportunity to record single-neuron activity in human Dorsolateral Prefrontal cortex and Substantia Nigra using microelectrodes and novel micro/macro electrode strips in order to test neuronal mechanism of WM in humans and answer several fundamental questions about it.

Goal 1: Single-neuron mechanisms of WM in the DLPFC

Dorsolateral Prefrontal Cortex (DLPFC) is a key structure engaged in WM. Meta-analyses of fMRI studies demonstrate that this area shows strong BOLD activity during WM. Moreover, DLPFC lesions in humans and monkeys are associated with WM deficits. Similarly, single-neuron recordings in macaques demonstrate that neurons in the DLPFC code stimulus-specific information during maintenance.

What is the mechanism used to code information in WM? One of the leading models of WM proposes that memoranda are maintained through persistent neuronal activity. This means that items held in WM are maintained by sustained firing activity of neurons. This model is supported by several macaque single-neuron recording studies with a variety of paradigms. Recently, we have also observed a stimulus-specific persistent code in human single-recordings in the Medial Temporal Lobe when subjects maintained information in WM. This model, however, is being contested by recent animal research recordings from the DLPFC, resulting in an ongoing debate about the neural mechanisms of WM. For instance, Lundqvist et al proposed that neurons that maintain information during WM are activated during discrete bursts of gamma oscillations rather than in a continuous sustained activity. Additionally, it was observed that information in WM could be maintained by employing a dynamic code, where neurons carrying information change during the course of the maintenance period. How do theories of WM translate to human brain activity? It is unknown, as there are no human single-neuron recordings from DLPFC during WM tasks. This is a significant gap in our knowledge, especially when we take into account that DLPFC is a key part of the circuit involved in performing WM tasks. By exploiting the unique opportunity to record single-neuron activity in DLPFC in this project I will, for the first time, describe the neuronal mechanisms employed during WM maintenance in humans.

Goal 2: The role of a dopaminergic modulation in WM

Dopamine has become the focus of WM research since the 1970s when Brozoski et al. demonstrated that macaques with regional dopamine (DA) deprivation have similar deficits in WM as macaques with lesions of DLPFC. Currently there are dozens of studies on humans and animals showing a crucial role of DA in WM, however, we still do not know how DA modulates WM function. One of the proposed theories suggests that DA could be involved in the active gating mechanism of WM. This mechanism allows information to enter WM storage and protects this information from distractors. Support for this theory comes from a recent study in macaques that demonstrated that DA modulation is crucial for the recovery of task relevant information after the presentation of distractors in a WM task. Similarly, a study in humans demonstrated that a DA agonist modulates distractor resistance. I propose to test these gating hypotheses by using single-neuron recordings of dopaminergic neurons in substantia nigra (SN), a key site in dopamine production projecting to DLPFC.

Previous single-neuron studies in the human SN from myself and others demonstrate the feasibility of dopaminergic cells recording. The project will provide data that will characterize dopaminergic activity in humans during a WM task. Critically, I will electrically stimulate SN to casually test the contribution of dopaminergic modulation to WM processes.

Goal 3: Network-level mechanisms of WM in humans

To understand how the human brain functions we need to exploit a variety of techniques that provide data on both the micro and macro scale. One of the goals of this project is to push the boundaries of our understanding of the neuronal mechanism of WM on the micro level. Additionally, I plan to employ a series of advanced computational techniques to gain a better network-level understating of the recorded neural responses.

One approach for embedding single-neuron activity into network-level brain dynamics is to analyze spike- field interactions. Undeniably, this approach is very fruitful in pinpointing the oscillatory activity relevant to neuronal discharges. In my previous studies with neural recordings in humans, I observed that neuronal spiking in substantia nigra coincided with theta oscillatory field activity recorded from frontal cortex. Moreover, theta oscillations are often associated with working memory tasks. Indeed, oscillatory models of WM suggest that theta plays a crucial role in WM. In the proposed project simultaneous recordings of electrocorticogram (ECOG) signals and single-neuron recordings in the DLPFC give an unprecedented opportunity to test these models in human subjects. Moreover, using simulations recording form DLPFC and Substantia Nigra I will able to probe interactions between those structures during WM.

In addition to spike-field interactions, state-space analysis is a powerful method used to understand the global dynamics of recorded neuronal populations, because these methods can capture dynamic activity not visible in the analysis at a single-neuron level. State-space analysis focuses on the multidimensional analysis of firing activity of the whole recorded population. Using this technique we observed that in Medial Temporal Lobe neuronal activity form attractors during WM maintenance. These attractors were behaviorally relevant: the distance in a given trial form the attractor centre predicted accuracy and reaction time. Here, I plan to use this technique to explore activity of DLPFC and Substania Nigra – key areas responsible for WM.

SIGNIFICANCE OF THE PROJECT

The field of cognitive neuroscience is dominated by non-invasive electrophysiology and neuroimaging studies. They provide important knowledge about brain mechanisms underlying cognition; nevertheless, single- neuron correlates of our behavior remain underexplored. This is especially visible in the domain of working memory (WM). To date, only two publications have explored the cellular basis of WM in humans. Because neural recordings in humans are limited to clinically relevant target regions, key brain structures for WM functions have not yet been investigated with single-neuron recordings. Without a precise description of WM single-neuron correlates in the human brain we will not be able to fully understand the mechanisms that govern WM. This project, for the first time, will probe single neuron activity in DLPFC and in Substantia Nigra during WM task in humans. We will use clinical microelectrodes and novel micro/macro strips to record neuronal activity from large area of DLPFC. By exploiting this unique opportunity of single-neuron recordings in the DLPFC and in Substantia Nigra this project will provide unique insights to the knowledge about the neuronal mechanisms of WM. Through a combination of carefully selected cognitive tasks and comprehensive data analysis a new framework for understanding WM in humans will be built. This will inspire new ideas for treatments of many neurological and psychiatric diseases like ADHD, Schizophrenia or Depression as they are characterized by WM disturbance and abnormalities in the DLPFC and the dopaminergic system.