Vagus nerve stimulation promotes generalization of conditioned fear extinction and reduces anxiety in rats

Abstract

Background: Exposure-based therapies are used to treat a variety of trauma- and anxiety-related disorders by generating successful extinction following cue exposure during treatment. The development of adjuvant strategies that accelerate extinction learning, improve tolerability, and increase efficiency of treatment could increase the efficacy of exposure-based therapies. Vagus nerve stimulation (VNS) paired with exposure can enhance fear extinction, in rat models of psychiatric disorders, and chronic administration of VNS reduces anxiety in rats and humans.

Objective: We tested whether VNS, like other cognitive enhancers, could produce generalization of extinction for stimuli that are not presented during the extinction sessions, but are associated with the fear event.

Methods: Male Sprague Dawley rats underwent auditory fear conditioning with two easily discriminable auditory stimuli. Following fear conditioning, extinction training consisted of exposure to only one of the conditioned sounds. Half of the rats received VNS and half received sham stimulation during with sound presentations. VNS effects on anxiety were examined in a separate study where VNS was administered prior to testing on the elevated plus maze.

Results: Sham stimulated rats given 20 presentations of a conditioned stimulus (CS) during the extinction session showed performance that was matched to VNS-treated rats given only 4 presentations of the CS. Despite comparable levels of freezing to the presented CS, only the VNS-treated rats showed a significant decrease in freezing to the CS that was not presented. VNS-induced generalization of extinction was observed only when the two sounds were paired with footshock within the same fear conditioning session; VNS did not promote generalization of extinction when the two sounds were conditioned on different days or in different contexts. On the anxiety test, VNS administration significantly increased time spent in the open arms of the elevated plus maze.

Conclusion: These results provide evidence that VNS can promote generalization of extinction to other stimuli associated with a specific fear experience. Furthermore, non-contingent VNS appears to reduce anxiety. The ability to generalize extinction and reduce anxiety makes VNS a potential candidate for use as an adjunctive strategy to improve the efficacy and tolerability of exposure-based therapies.

Keywords: Anxiety; Extinction; Fear; Generalization; VNS; Vagus.

Conflict of interest statement

Conflict of interest:

This work has not been published and has not been submitted for publication elsewhere while under consideration. Authors: Noble, Meruva, and Hays declare no potential conflicts of interest. Dr. Kilgard is a paid consultant for and shareholder of MicroTransponder. Drs. Kilgard and McIntyre are authors of a patent entitled “Enhancing Fear Extinction using Vagus Nerve Stimulation”. Dr. Rennaker is an owner of Vulintis Inc. and Optokinetics and is a paid consultant for Konan Medical USA; none of these financial interests are related to this work.

Copyright © 2018 Elsevier Inc. All rights reserved.

Figures

Figure 1.. Spectrograms of each conditioned stimulus used during AFC. A. Spectrogram for CS1 (machine gun fire). B. Spectrogram for CS2 (marmoset vocalization).

Spectrograms for CS1 and CS2 were used in generalization experiments, as they are easy to discriminate given they have distinct differences in frequency bandwidth, frequency modulation, rise time, and decay time. Duration of stimuli were also different, CS1 lasted 50 seconds while CS2 lasted 39 seconds.

Figure 2.. VNS+Extinction leads to generalization of extinction for both the CS that was presented and the CS that was not presented during extinction training. A. Timeline for AFC, conditioned fear response testing, and extinction treatment.

AFC was administered across two days where both CS1 and CS2 were presented and paired with footshocks on each day. The order of CS presentation was random and interleaved such that each rat was administered 4 presentations of each CS on each day. Following AFC, rats underwent a pre-extinction conditioned fear response test (CFRT) where they were presented with four presentations of each CS to measure conditioned fear. Twenty-four hours later, rats underwent treatment consisting of either: 20 extinction trials of only CS1 or only CS2, paired with sham stimulation (Extended Extinction), 4 extinction trials of only CS1 or only CS2, paired with VNS (VNS+Extinction), or no exposures to either CS but equivalent amounts of VNS in the home cage (VNS Alone). A day later, rats underwent another CFRT to assess levels of conditioned fear to both stimuli. B. Extended Extinction and VNS+Extinction rats show equal reduction in conditioned fear to the Presented CS. Following 20 unreinforced presentations of a CS, Extended Extinction rats showed reduced freezing in response to the CS versus freezing during Pre-Extinction (p=2.3×10−4). VNS+Extinction rats also showed reduced freezing in response to the Presented CS versus Pre-Extinction (p=3.2×10−4). C. Only VNS+Extinction leads to generalization of extinction for the Non-Presented CS. Extended Extinction did not show reduced fear in response to the Non-Presented CS versus the Pre-Extinction freezing response (p=0.15). In contrast, VNS+Extinction rats showed a reduction in freezing response to presentation of the Non-Presented CS versus the Pre-Extinction freezing response, even though it was never presented during extinction (p=2.1×10−4). D. VNS Alone does not lead to extinction of either CS. Following VNS in the home cage in lieu of extinction, VNS Alone rats showed no reduction in conditioned fear for either CS versus the Pre-Extinction CFRT (p=0.16). E. Context extinction cannot explain VNS+Extinction generalization. There is no difference in baseline fear to the context between Extended Extinction and VNS+Extinction groups, rats spend equal amounts of time freezing to the context during the Post-Extinction CFRT (p=0.68). VNS Alone rats show elevated freezing to the context versus Extended Extinction (p=2.1×10−3) and VNS+Extinction rats (p=3.8×10−3).

Figure 3.. Separating CS1 and CS2 conditioning by twenty-four hours blocks the VNS effect on generalization. A. Timeline for AFC, CFRT, and extinction treatment.

AFC was administered across two days where CS1 was presented on 1 day and CS2 was presented twenty-four hours later. A CFRT to CS1 was administered on day 3 where CS1 was presented four times without reinforcement to measure conditioned fear to CS1. The next day, a CFRT was administered to CS2 where CS2 was presented four times without reinforcement to measure conditioned fear to CS2. Twenty-four hours later, rats underwent extinction consisting of either: 20 extinction trials of CS1 or CS2 paired with sham stimulation (Extended Extinction), or 4 extinction trials of CS1 or CS2 paired with VNS (VNS+Extinction). Following extinction, a Post-Extinction CFRT was administered where four presentations of each CS were presented and randomly interleaved to assess conditioned fear to each CS following extinction. B. Extended Extinction and VNS+Extinction rats show equal reduction in conditioned fear to the Presented CS. Following 20 presentations of the Presented CS, both Extended Extinction and VNS+Extinction rats showed reduced freezing to the Presented CS (p=7.1×10−3 vs PreExtinction) and (p=4.4×10−3 vs. Pre-Extinction), respectively. C. Neither Extended Extinction rats or VNS+Extinction rats show reduced freezing to the Non-Presented CS. After extinction training, freezing to the Non-Presented CS was not reduced for Extra Extinction rats (p=0.12). In contrast to Figure 1, freezing to the Non-Presented CS was not reduced in (p=0.29). D. Baseline freezing to the context is not different between groups. Extended Extinction and VNS+Extinction rats spent equivalent amounts of time freezing to the extinction context during the five minutes prior to CS presentation at the Post-Extinction CFRT (p=0.55).

Figure 4.. Separating conditioning of CS1 and CS2 by context blocks the VNS+Extinction effect on generalization. A. Timeline for AFC, CFRT, and extinction treatment.

AFC was administered across two days, where CS1 was presented only in Context A, and CS2 was presented only in Context C, but both CS1 and CS2 were presented during the same session. Twenty-four hours after AFC, a Pre-Extinction CFRT was administered for CS1 in Context A, where CS1 was presented four times without reinforcement. On the same day, a Pre-Extinction CFRT was also administered to CS2 in Context C, where CS2 was presented four times without reinforcement. Following the Pre-Extinction CFRT day, rats were subjected to extinction treatment consisting of either: 20 presentations of CS1 in Context A paired with sham stimulation (Extended Extinction), or 4 presentations of CS1 in Context A paired with VNS (VNS+Extinction). Twenty-four hours following extinction, a Post-Extinction CFRT was given to CS1 in Context A, and then on the same day a Post-Extinction CFRT was given for CS2 in Context C. B. Extended Extinction and VNS+Extinction rats show equal reduction in conditioned fear to the Presented CS. Following extinction of CS1 in Context A, Extended Extinction and VNS+Extinction rats showed reduced conditioned fear to the Presented CS versus Pre-Extinction (p=2.3×10−3) and (p=1.2×10−3), respectively. C. Neither Extended Extinction rats or VNS+Extinction rats show reduced freezing to the Non-Presented CS. Freezing to the Non-Presented CS was not reduced after Extinction training for Extended Extinction rats (p=0.38). Similar to Figure 2, freezing to the Non-Presented CS was also not reduced for VNS+Extinction rats versus Pre-Extinction (p=0.40). D. Baseline freezing to the extinction context is not different between groups. Extended Extinction and VNS+Extinction rats spent an equal amount of time freezing to the extinction context during the five minutes prior to presentation of CS1 during the Post-Extinction CFRT (p=0.53). E. Baseline freezing to the non-extinction context is not different between groups. Extended Extinction and VNS+Extinction rats spent an equal amount of time freezing to the non-extinction context during the five minutes prior to CS2 presentation during the Post-Extinction CFRT (p=0.62).

Figure 5.. VNS reduces anxiety.

Rats were given one non-contingent train of VNS or sham stimulation in their home cage. Ten minutes later, they were removed from their home cage and placed on the elevated plus maze (EPM). VNS-treated rats spent significantly more time in the open arms of the EPM when compared to sham-treated rats (p=2.1×10−3), indicating a VNS-induced reduction in anxiety. The percent of time spent moving during EPM testing was similar between groups, indicating no general locomotor effect.

Figure 6.. Dual benefits of VNS.

Cognitive enhancers show promise as adjuncts to exposurebased therapies; however, most pharmaceutical enhancers of cognition do not improve tolerability of the therapy by reducing anxiety. Anxiolytic drugs may improve tolerability but they are not effective, possibly because they interfere with the consolidation of new extinction memories. However, VNS enhances memory consolidation and accelerates extinction, and it also reduces anxiety. This unique combination of effects suggests that VNS may offer a desirable alternative to drugs as an adjunct therapy.