Dissociation of learned helplessness and fear conditioning in mice: a mouse model of depression

Abstract

The state of being helpless is regarded as a central aspect of depression, and therefore the learned helplessness paradigm in rodents is commonly used as an animal model of depression. The term 'learned helplessness' refers to a deficit in escaping from an aversive situation after an animal is exposed to uncontrollable stress specifically, with a control/comparison group having been exposed to an equivalent amount of controllable stress. A key feature of learned helplessness is the transferability of helplessness to different situations, a phenomenon called 'trans-situationality'. However, most studies in mice use learned helplessness protocols in which training and testing occur in the same environment and with the same type of stressor. Consequently, failures to escape may reflect conditioned fear of a particular environment, not a general change of the helpless state of an animal. For mice, there is no established learned helplessness protocol that includes the trans-situationality feature. Here we describe a simple and reliable learned helplessness protocol for mice, in which training and testing are carried out in different environments and with different types of stressors. We show that with our protocol approximately 50% of mice develop learned helplessness that is not attributable to fear conditioning.

Fig 1. Time course of learned helplessness protocol and subsequent additional behavioral tests.

Five days before training, mice were transferred to individual cages. Training on days 1 and 2 was done at ZT9 (3 pm). On day 3 mice were tested three hours earlier at ZT6 (12 noon) in order to prevent time of day related anticipation. At ZT5 (11 am) on the following day, additional behavioral tests (tail suspension test or open field test) were conducted. Acclimation to water bottles used in sucrose testing started 4 days prior to LH testing (for details see “Methods” section). Two sucrose preference tests were conducted for ~1 day each. The first test started immediately after the second training session, and the second test started immediately after the testing session.

Fig 2. Setup of trans-situational LH.

(A) Shuttle boxes are comprised of two compartments separated by a movable gate. Both compartments have a metal grid floor through which the electric foot shocks were delivered. Current intensity can be adjusted for each individual box. Infrared detectors determined the position of the mice at all times during the experiment. (B) During training, mice were kept in a restrainer, and conductive paste was administered to their tails. Conductive metal rings with screws (taken from luster terminals) were gently attached to the tails (about 1 cm apart). Cables that deliver the electric shocks from the shuttle boxes to the mice were connected to the metal rings with alligator clips (red: positive and black: negative). (C) Overview of the whole trans-situational LH setup. Each shuttle box is connected to two cables (positive and negative) so that the electric shocks from the grid floor inside the shuttle box compartments can be delivered to the tails of the mice outside the boxes. Each mouse was connected to one box. The mice were kept in restrainers that were placed on a cart with high borders, which blocked the shuttle boxes from view. In addition, black plastic roofs separated the mice from each other to avoid visual contact. The shuttle boxes were run by a computer (not shown).

Fig 3. Inescapable electric shocks lead to escape failures in a different environment but not to fear-related behavior in mice.

(A) Before testing, mice had a 60 sec acclimation time to explore the shuttle boxes with open gates, and the number of gate crossings was measured. The exploratory behavior of the mice did not correlate with the grade of helplessness subsequently detected. Black line: linear regression line, dashed lines: 95% confidence band, n = 63. (B) After two days of tail-shock training, mice were tested in shuttle boxes, and average escape latency time and number of escape failures were measured. Resilient animals (open circles) were defined as those showing escape latencies within 2 standard deviations of those of naive mice (shorter than 9.5 sec) and numbers of escape failures within 2 standard deviations of those of naive mice (less than 2 failures). Animals with greater escape latencies and larger numbers of escape failures were defined as helpless (black circles). Thresholds are shown as dashed lines. n = 63. (C) Escape latencies and numbers of escape failures of resilient animals were in the range of naive mice that never received training. Data are shown as average ± SEM; ***p ≤ 0.001, (1-way ANOVA with Bonferroni post-test comparing all data sets); n-values are shown in brackets.

Fig 4. Shock intensity during testing affects escape behavior.

Latency times (top panel) and number of escape failures (bottom panel) are shown for PER2::LUC mice tested with 0.10 mA current intensity and WT mice tested with different current intensities ranging from 0.10 mA to 0.30 mA. WT mice required higher current intensities to reach escape latency times and failure frequencies comparable to those of PER2::LUC mice. Data are shown as box-and-whisker plots; **p ≤ 0.01, (1-way ANOVA with Dunnett post test comparing all wild-type data with PER2::LUC data); n-values are shown in brackets.

Fig 5. Helplessness is partly correlated with other behavioral phenotypes, but not with total locomotor activity.

(A) Sucrose preference test was performed immediately after the second LH training (1. sucrose preference) and a second time after LH testing (2. sucrose preference). Top panel: sucrose preference comparing naive mice that did not receive LH training/testing (spotted bar), resilient (white bars), and helpless mice (black bars). Bottom panel: total consumption of water and 1% sucrose solution of the same mice. Data are shown as average ± SEM; *p ≤ 0.05, (1-way ANOVA with Bonferroni post test comparing all data sets); n-values are shown in brackets. (B) Immobility time (top panel) and latency until the first 2 sec of immobility (bottom panel) in the tail suspension test are not dependent on helplessness. Data are shown as average ± SEM; n-values are shown in brackets. (C) In the open field test, learned helplessness does not change general locomotor activity, on the basis of total distance (left panel) and immobility time (second panel from the left). Helpless mice spend slightly more time in the center of the open area (second panel from the right), but latency until the center is entered for the first time is not different among the groups (right panel). Data are shown as average ± SEM; *p ≤ 0.05, (1-way ANOVA with Bonferroni post test comparing all data sets with each other); n-values are shown in brackets. All tests were carried out in PER2::LUC mice.

Fig 6. Trans-situationality results in less fear-related behavior in mice.

(A) Before mice were tested, they had 60 sec acclimation time to explore the shuttle boxes. The numbers of gate crossings were measured as a marker of contextual fear and exploratory behavior. Data are shown as average ± SEM; *p ≤ 0.05, ***p ≤ 0.001 (Student’s t-test); n = 8. (B) The top panel shows average escape latencies of mice that received training in restrainers with tail shocks, and the bottom panel shows results for mice that received training in one side of the same shuttle boxes where they were subsequently tested. Mice were tested either 1 day after the second training (open circles) or 8 days after the second training (gray circles). A: FR-1 trials 1+2, B: FR-1 trials 3–5, 1–5: blocks of 5 FR-2 trials. Data are shown as average ± SEM; n = 8 (the same mice as in (A); 2-way ANOVAs with Bonferroni post tests did not reveal significant differences between mice tested 1 day or 8 days after training. All tests were carried out in PER2::LUC mice.