The role of mPFC and MTL neurons in human choice under goal-conflict

ARTICLE https://doi.org/10.1038/s41467-020-16908-z OPEN The role of mPFC and MTL neurons in human choice under goal-conflict 1234567890():,; Tomer Gazit1,2,10, Tal Gonen1,3,10, Guy Gurevitch 1,4,10, Noa Cohen1,2, Ido Strauss2,5, Yoav Zeevi6,7, Hagar Yamin1, Firas Fahoum 2,8, Talma Hendler 1,2,4,6,10 ✉ & Itzhak Fried1,2,5,9,10 Resolving approach-avoidance conflicts relies on encoding motivation outcomes and learning from past experiences. Accumulating evidence points to the role of the Medial Temporal Lobe (MTL) and Medial Prefrontal Cortex (mPFC) in these processes, but their differential contributions have not been convincingly deciphered in humans. We detect 310 neurons from mPFC and MTL from patients with epilepsy undergoing intracranial recordings and participating in a goal-conflict task where rewards and punishments could be controlled or not. mPFC neurons are more selective to punishments than rewards when controlled. However, only MTL firing following punishment is linked to a lower probability for subsequent approach behavior. mPFC response to punishment precedes a similar MTL response and affects subsequent behavior via an interaction with MTL firing. We thus propose a model where approach-avoidance conflict resolution in humans depends on outcome value tagging in mPFC neurons influencing encoding of such value in MTL to affect subsequent choice. 1 Sagol Brain Institute Tel Aviv, Wohl Institute for Advanced Imaging, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel. 2 Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel. 3 Department of Neurosurgery, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel. 4 School of Psychological Sciences, Faculty of Social Sciences, Tel Aviv University, Tel Aviv, Israel. 5 Functional Neurosurgery Unit, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel. 6 Sagol School of Neuroscience, Tel Aviv University, Tel Aviv, Israel. 7 Department of Statistics and Operation Research, Tel Aviv University, Tel Aviv, Israel. 8 Epilepsy Unit, Department of Neurology, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel. 9 Department of Neurosurgery, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, CA, USA. 10These authors contributed equally: Tomer Gazit, Tal Gonen, Guy Gurevitch, Talma Hendler, Itzhak Fried. ✉email: NATURE COMMUNICATIONS | (2020)11:3192 | https://doi.org/10.1038/s41467-020-16908-z | www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-020-16908-z H umans often find themselves facing a choice involving conflicting emotions. Spinoza defined such conflicting emotions as those which draw a man in different directions (Part IV of the Ethics, on Human Bondage). Indeed approach-avoidance behavioral choices are resolved by the human capacity to adapt goal-directed behaviors to the emotional value of prospective outcomes. Rewarding outcome serves to strengthen or reinforce context-behavior associations, thereby increasing the likelihood of future approach behavior1. Aversive outcomes, on the other hand, are encoded so as to avoid similar future punishment, thus encouraging avoidance behavior2. Animal studies have investigated the neural mechanism responsible for encoding the effects of various outcomes on subsequent behavior, mostly in the context of reinforcement learning. Accumulating evidence points to the striatum as an important region involved in signaling prediction errors (PEs)3 and to the medial prefrontal cortex (mPFC)4, and medial temporal lobe (MTL) as processing outcome values and valence5,6. Of particular importance is the known role of the hippocampus and the amygdala in forming, respectively, contextual and emotional associations7 that guide future behavior in reinforcement learning procedures. However, less is known regarding the effects of outcome valence on the probability of subsequent behavior in situations of goal conflict. Goal conflicts arise when we encounter potential gains and losses simultaneously within the same context8. Such conflicts are thought to be central to the generation of anxiety; a state of high arousal and negative outcome bias that often leads to disadvantageous dominance of choosing avoidance behavior9,10. Classical animal studies using goal-conflict paradigms such as the elevated plus maze (EPM)11,12 have implicated the amygdala13, hippocampus9,14, and mPFC15,16 as being crucial in triggering avoidance behavior in goal conflict situations. For example, in Kimura et al.12, rats were punished with a delivery of an electrical shock as they consumed food (avoidance training). Over time, control animals increased their latency to enter the target box, while rats with hippocampal lesions presented impaired acquisition of such passive avoidance behavior. However, classical animal studies have not clearly differentiated the neural substrates involved in using information regarding the valence of outcomes (reward vs. punishment) for subsequent adaptation of approach behavior, from those that mediate the actual resolution of the goal conflict17. Schumacher et al.17 showed that the ventral hippocampus (vHPC) is involved in the resolution of approach-avoidance conflict at the moment of decision making rather than in learning about the value of outcomes for future decisions. On the other hand, further studies showed that the hippocampus, as well as the amygdala, seems to support learning from outcomes and thus affect future behavior. For example, Davidow et al.5 showed that adolescents were better than adults at learning from outcomes to adapt subsequent decisions, and that this was related to heightened PE-related BOLD activity in the hippocampus. Using lesions to macaque amygdalae, Costa et al.18 present evidence that the amygdala plays an important role in learning from outcomes to influence subsequent choice behavior. With relation to psychopathology, it has been suggested that patients suffering from depression are unable to exploit affective information to guide behavior19. For example, Kumar et al.20 found reduced reward learning signals in the hippocampus and anterior cingulate in patients suffering from major depression. Disruption of prediction-outcome associations in the bilateral amygdala–hippocampal complex was found in patients with schizophrenia21. Yet, it remains to be seen whether these results, pointing to the significance of the MTL in the processing of outcomes and adapting behavior, are relevant to outcomes that appear in the context of an approach-avoidance conflict. To investigate these processes, we use a rare opportunity to perform intracranial recordings from multiple sites in the MTL and mPFC of 14 patients with epilepsy (Table 1). We apply a previously validated game-like computerized task22 that enables the measurement of goal-directed behavior (the tendency to approach) under high or low goal conflict and evaluate the neural response to the outcome of this behavior (reward or punishme (...truncated)