Prenatal methylmercury exposure increases responding under clocked and unclocked fixed interval schedules of reinforcement

Abstract

Recent experiments have suggested that developmental methylmercury exposure produces perseverative behavior in adulthood. In the present experiment, interactions between developmental low-level methylmercury (MeHg) and nutritionally relevant dietary selenium (Se) on operant behavior and its persistence were examined in aged animals. Female rats were exposed, in utero, to 0, 0.5, or 5 ppm mercury as MeHg via drinking water, approximating mercury exposures of 0, 40, and 400 μg/kg/day. They also received both pre- and chronic post-natal exposure to a diet that was marginal (0.06 ppm) or rich (0.6 ppm) in Se, a nutrient believed to protect against MeHg's toxicity. This created a 2 (chronic Se) × 3 (gestational MeHg) full factorial design, with 6–8 female rats per cell. At eleven months of age, a multiple schedule consisting of alternating fixed interval (FI) and clocked FI (CFI) components was arranged. The CFI component was divided into 5, 24-second bins, each associated with a different auditory stimulus, providing a “clock.” Low and high response rates were evaluated using the initial 40% (bins 1 and 2) and last 20% (bin 5) of the FI and CFI components, respectively. Rats exposed to 5 ppm Hg made more responses than the other two groups during the last 20% of the intervals, regardless of selenium exposure or presence of the clock stimuli. They did not differ from the other groups during the initial 40% of the FI and CFI components. Following reinforcement omission for half of the intervals at 21 months of age, the 5 ppm Hg group continued to respond at higher rates than the other groups in both components.

Introduction

Methylmercury (MeHg) is a known developmental neurotoxicant found in fish and marine mammals. Fish, however, are also an important source of nutrients such as selenium and long-chain polyunsaturated fatty acids. In rodent studies examining adult-onset MeHg exposure, selenium (Se) ameliorated some of the effects of chronic, high-level MeHg exposure [9], [12], [21]. For developmental MeHg exposure, however, evidence that Se confers protection is less pronounced. With developmental MeHg exposure, a diet severely deficient in Se enhanced MeHg's fetolethality [30], as well as MeHg's detrimental effects on the development of gait, thermal preference, and open-field activity, but these effects did not persist into adulthood [46]. In contrast, a Se-sufficient diet has not attenuated MeHg's neurobehavioral toxicity [4], [36] with the exception of one study [46]. In that study, Se excess attenuated MeHg-induced hypoactivity in animals exposed to 6 mg/kg MeHg by gavage on days 6–9 of gestation and examined at two months of age [8]. Thus, while Se ameliorates some of MeHg's effects following adult-onset exposure, the picture is less clear for developmental exposure.

A recent study focused on the potential interactions between low-level developmental MeHg exposure and nutritionally relevant dietary Se on spatial discrimination reversals in adulthood [36]. Although all rats acquired the original discrimination similarly, MeHg-exposed rats, regardless of Se exposure, made more errors than controls on the first and third reversals, which were away from the lever that was reinforced in the original discrimination. MeHg-exposed rats also had shorter choice latencies than controls, implying an impulsive or perseverative response pattern. Rats consuming a low-Se diet, regardless of their MeHg exposure, made more omissions (trials without a response) during the first reversal and required more sessions to complete this reversal than rats exposed to a high-Se diet. Thus, while there were main effects of both MeHg and Se, on no measure was there an interaction between Se and MeHg exposure. However, behavioral procedures that permit greater variation in response rate, such as the fixed-interval schedule of reinforcement, might be sensitive to such an interaction.

The behavioral effects produced by MeHg exposure could be viewed as reflecting a reduced sensitivity to changing reinforcement contingencies or an increased reinforcer efficacy [23]. Disentangling these two ideas is difficult, since each would result in persistent, or even perseverative, responding when reinforcement contingencies are altered. Both interpretations are consistent with findings that gestational MeHg exposure slowed transitions during a choice in-transition procedure [27], [28], resulted in more rapid acquisition of lever-pressing and lack of ratio strain under large fixed ratio schedules of reinforcement [32], and a tolerance for higher ratios under a progressive ratio procedure [32], [35].

The behavioral patterns seen in MeHg-treated animals in previous studies allow us to make predictions about the behavior of exposed animals under other reinforcement schedules. For example, response rates under fixed interval (FI) schedules are positively related to the reinforcement magnitude of food pellets [19], [20], sucrose solution [42], and cocaine [2]. If a reinforcer's efficacy is increased for animals exposed to MeHg, then increased response rates under the FI schedule, especially in the last portion of the interval, would be expected.

Previous studies of developmental MeHg exposure have not identified deficits in discrimination [e.g. [5], [39], [41]], an observation that supports a second prediction. If exteroceptive stimuli are correlated with the passage of time in an FI schedule, a “clocked” FI (CFI) [16], [31], then we would expect animals to make fewer responses, particularly in the first portion of the interval, as compared with the typical FI. Since there is little evidence that developmental MeHg exposure affects discrimination processes [5], [24], [39], [41], then we might expect no differential effect of MeHg on clocked performance. However, MeHg-exposed animals might be expected to have greater response rates during the latter portion of the interval in both components if reinforcer efficacy is altered by MeHg exposure. Finally, if the reinforcer is omitted at the end of the FI and CFI components, but responses continue to be recorded, then we would expect MeHg-exposed animals to make more responses than controls.

The present study was designed to examine the role of exteroceptive stimuli and reinforcement omission in rats exposed developmentally to MeHg and chronically to a diet either marginal or rich in Se, a nutrient hypothesized to protect against MeHg's neurotoxic effects [18], [34], [44], [45], [48]. The experiments were conducted using a 2 (chronic Se) × 3 (gestational MeHg) full factorial design, which allows for the direct examination of the interactions between MeHg and Se, as well as the main effects of either element. The MeHg concentrations chosen produce levels spanning the low to moderate range [6], [26], as determined by brain mercury [25]. Likewise, the Se diets were at the low and high end of recommended intakes. The 0.06 ppm Se concentration is lowest possible with a casein-based diet and is still a nutritionally adequate level for rodents [22], [38]. The higher, 0.6 ppm, concentration is at the high end of adequate and represents an excess over the AIN-93 formulation, which contains 0.15 ppm of Se [37], [38], but is below that thought to be toxic [1].

Upon reaching adulthood, female offspring were trained to respond under a Mult FI 120″, Clock FI (CFI) 120″ schedule of reinforcement. When the FI schedule was in effect, the first lever-press after 120″ produced sucrose. When the CFI was in effect, five distinct auditory stimuli were presented sequentially for 24″ each, resulting in a 120″ interval, and the first lever-press after 120″ produced sucrose. Rats experienced twenty-two sessions of the Mult FI CFI schedule with auditory stimuli before group comparisons of baseline responding were made at 13 months of age. After this first comparison, drug challenges began (to be described elsewhere) with multiple doses of cocaine, desipramine, SKF-38393, quinpirole, SCH-23390 and sulpiride. At 20 months of age, thirty days after completing the last dose-effect determination, responding under the Mult FI CFI schedule was reassessed and compared with their performance at 13 months of age. Finally, reinforcement omission trials were instated for 10 sessions at 21 months of age: responses at the end of the interval were not followed by sucrose for half of the FI and CFI components, but responding continued to be monitored for an additional 240″. The responses at the end of the interval in the remaining FI and CFI components were followed by sucrose as in previous sessions.