Behavioral Intoxication following Voluntary Oral Ingestion of Tetramethylenedisulfotetramine: Dose-Dependent Onset, Severity, Survival, and Recovery

Abstract

Tetramethylenedisulfotetramine (tetramine, or TETS) is a highly toxic rodenticide that has been responsible for over 14,000 accidental and intentional poisonings worldwide. Although the vast majority of TETS poisonings involved tainted food or drink, the laboratory in vivo studies of TETS intoxication used intraperitoneal injection or gavage for TETS exposure. Seeking to develop and characterize a more realistic model of TETS intoxication in the present study, rats were trained to rapidly and voluntarily consume a poisoned food morsel. Initially, the overt toxic effects of TETS consumption across a large range of doses were characterized, then a focused range of doses was selected for more intensive behavioral evaluation (in operant test chambers providing a variable-interval schedule of food reinforcement). The onset of intoxication following voluntary oral consumption of TETS was rapid, and clear dose-dependent response-rate suppression was observed across multiple performance measures within the operant-chamber environment. At most doses, recovery of operant performance did not occur within 30 hours. Food consumption and body weight changes were also dose dependent and corroborated the behavioral measures of intoxication. This voluntary oral-poisoning method with concomitant operant-behavioral assessment shows promise for future studies of TETS (and other toxic chemicals of interest) and may be extremely valuable in characterizing treatment outcomes.

1. Introduction

Tetramethylenedisulfotetramine (tetramine, or TETS) is a highly potent convulsant commonly used as a black market rodenticide in Asia and is one of the most dangerous food and water contaminants that could be used in an intentional poisoning scenario [1]. TETS is easily synthesized [2], tasteless, odorless, remarkably stable [3], environmentally persistent [2], and particularly toxic to mammalian species [1, 3, 4], and it can remain in the body for up to six months [5]. TETS was once marketed as a rodenticide, and despite its use being banned worldwide since 1984 [6], it remains widely available in some regions. It has been responsible for large numbers of accidental and intentional mass poisonings and is perhaps responsible for more intentional fatal poisonings than any other chemical in recent history [7]. These incidents are particularly common in China, where over 14,000 poisonings and 900 fatalities attributable to TETS occurred between 1991 and 2010 [8]. TETS, as an agent of terror, has gained increased attention from Western countries [9] and has been used in over 50 poisonings outside of China [10]. Although an intentional poisoning involving TETS has not yet been reported in the U.S., a child was accidentally poisoned at her home in New York City by TETS brought back from China. Within 15 minutes of exposure the child experienced seizures that persisted for 4 hours despite aggressive therapy with lorazepam, phenobarbital, and pyridoxine. The child remained hospitalized for several days and at discharge appeared to have multiple neurological deficits, including absence seizures and possible cortical blindness [11]. There is currently no standard treatment for TETS poisoning in the United States, but a common Chinese protocol for TETS poisoning involves gastric lavage and high-volume or charcoal-filter hemoperfusion to remove as much TETS from the body as possible [5, 12]. Several potential drug treatments have also been identified for TETS poisoning, including diazepam [13–16], ketamine [13], MK-801 [13, 15], high-dosage γ-aminobutyric acid (GABA) [17], allopregnanolone [14, 17], and sodium dimercaptopropane sulfonate (Na-DMPS) [17]. However, no drug treatment (single or combinatorial) has been completely effective at preventing behavioral intoxication and delayed lethality following TETS poisoning, possibly due to the mechanism by which TETS toxicity occurs.

TETS is a neurotoxin which reversibly binds non-competitively to the α1 and γ2 subunits of the γ-aminobutyric acid_A (GABA_A) receptor ionophore complex [18, 19], preventing chloride ion influx and acting as a potent antagonist [18, 20]. The acute toxicity of TETS antagonism of the GABA system results from an over-excitement of central nervous system neurons. The disruption of the central nervous system caused by acute TETS poisoning leads to severe and widespread neurological and physical symptoms including convulsions, arrhythmias, hematological changes, coma, respiratory failure, refractory status epilepticus, and death [8, 10, 21]. Sub-lethal or chronic doses of TETS can lead to long-term multiple organ failure and permanent neurological impairment [10, 22]. Because of the inherent difficulty in identifying the source and nature of seizures in a clinical setting, treatment of TETS toxicity with typical regimens of anticonvulsant and antiepileptic drugs can prove inadequate. Due to the nature of convulsive status epilepticus, a time-dependent loss of efficacy and potency for many first-line drugs, such as benzodiazepines, quickly develops, and seizures can become self-sustaining [23]. Studies suggest that this time-dependent pharmacoresistance is due in part to gradual endocytosis of synaptic GABA_A receptors along with the concurrent movement of excitatory α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and N-methyl-D-aspartate (NMDA) receptors to the synaptic membrane. Further, there is evidence of depletion of some inhibitory neuropeptides during prolonged status epilepticus, which contributes to the self-sustaining nature of the seizure activity [24].

Previous studies examining TETS intoxication and potential treatments [3, 13–16, 25, 26] have relied upon TETS dosing methods that do not mimic the many historical real-world exposures. By far the most common route of poisoning from TETS in humans is via voluntary consumption of TETS-adulterated foods or liquids [1, 7, 22]. Despite this fact, previous studies of TETS intoxication used gastric gavage or intraperitoneal (IP) injection as a route of administration. Obviously, IP injection is not the same as voluntary eating or drinking. Additionally, results from gavage studies may be misleading because a disproportionately large volume of liquid is forced into the animal’s stomach at a rapid rate, causing faster absorption of the toxic chemical [27]. Likewise, changes in the volume and concentration of IP administrations can affect the pharmacokinetics of the compound being administered [28, 29]. TETS is also rapidly absorbed through the saliva and mucosa of the mouth and pharynx [30], further changing the potential pharmacokinetics for different routes of TETS exposure. To better characterize the toxic profile of ingested TETS in a manner consistent with real-world scenarios, we developed a voluntary-ingestion model in adult male Sprague-Dawley rats. In this model, rats were trained to reliably consume a known quantity of food (a single piece of Froot Loops® cereal) in a short period of time. After rats were trained to consume the food quickly and reliably, a quantity of TETS in an acetone solution was applied to the food, and the acetone vehicle was allowed to completely evaporate, leaving only the TETS on the food morsel. Rats were then allowed to eat the poisoned food, which they did promptly and reliably.

The current study sought to characterize the toxicity of TETS following voluntary consumption, with an emphasis on behavioral intoxication across a range of doses. Such behavioral measures are intended to inform the onset, type, and degree of intoxication to allow for signs to be rapidly recognized and treatment administered (a so-called “trigger to treat”). Additionally, one can determine the extent and duration over which intoxication persists when left untreated, establishing a baseline for future studies searching for effective treatments against TETS.

2. Methods

2.1 Chemicals

Acetone (≥99.5%) was purchased from Sigma-Aldrich and stored at room temperature. The TETS (anhydrous) was obtained from the Edgewood Chemical Biological Center (Aberdeen Proving Ground, MD) at ~78% purity and stored at 4 degrees Celsius. The primary impurity (approximately 16% of the sample) within the powder was hexamethylenetrisulfohexamine (HEXS), a common TETS contaminant [26, 31] believed to be approximately 50-fold less toxic than TETS [4]. The remaining impurities were unable to be characterized at the time of this writing, but appear to be consistent with those noted by the Edgewood Chemical Biological Center in their review of TETS synthesis [32].

The TETS powder was dissolved into acetone solution by institutional chemical surety specialists to a concentration of 2 mg/mL. These multiple aliquots (1–2 mL) of TETS were stored securely at room temperature in amber vials in a dark, locked cabinet. An aliquot was typically discarded after a single use, but the initial cohorts used aliquots across multiple cohorts and therefore the solution was kept for up to 6 weeks after being opened and then resealed. All handling of the TETS occurred within the confines of a certified chemical fume hood, and personnel wore a face mask, safety goggles, lab coat, and double nitrile gloves.

2.2 Subjects

One hundred five (105) male Sprague-Dawley rats (SAS SD 400) were obtained from Charles River Laboratories (Wilmington, MA, USA). Forty-five (45) rats were assigned to the overt-toxicity assessment, and 60 rats were assigned to the behavioral-toxicity assessment (n = 12 rats × 5 doses). Rats in both assessments weighed between 201–225 g at the time of shipping and were allowed five days to acclimate to our facility. All subjects were housed individually in a vivarium with free access to water under a 12 h light/dark cycle (lights on at 0600). During acclimation all rats were fed ad libitum, after which food regulation was implemented and maintained for the remainder of the study. During food regulation rats were given a measured amount of food every afternoon (at least 30 min after behavioral assessment) to ensure their weights were approximately 85% of the ad libitum growth-curve weights provided by the vendor.

The experimental protocol was approved by the Animal Care and Use Committee at the United States Army Medical Research Institute of Chemical Defense (USAMRICD), and all procedures were conducted in accordance with the principles stated in the Guide for the Care and Use of Laboratory Animals and the Animal Welfare Act of 1966 (P.L. 89–544), as amended. The USAMRICD is a research facility fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International.

2.3 Apparatus

Sixteen commercially available operant chambers for rodents were used (Model ENV-009, 31.8 cm × 41.9 cm × 34.3 cm, MED Associates, St. Albans, VT). Each chamber was housed within a fan-ventilated sound-attenuating cubicle that masked extraneous light and noise. A pellet dispenser (ENV-203M) was located in the center of the front wall and delivered 45 mg pellets (product # F0165; Bio-Serv, Flemington, NJ) to a food trough (ENV-200R2M) near the bottom of the front wall, approximately 3 cm above the grid floor. The food trough contained a photobeam (ENV-254) for detecting pellet retrieval by the subject. An audio module (ENV-223HAM) was mounted on the upper left portion of the front wall and presented a high-pitched beep (4,500 Hz at approximately 70dB) that accompanied each pellet delivery. Two response levers (ENV-110M) were located on either side of the food trough and required a force of approximately 0.1 N to operate. Above each response lever was a triple stimulus display (ENV-222M "cue light") comprised of three LEDs (red, yellow, and green) that were illuminated simultaneously. A water dipper (ENV-202M-S) was located on the far right of the front wall and, like the food trough, had a photobeam for detecting entry. General illumination was provided by a houselight (ENV-215M) mounted centrally on the upper portion of the rear wall. A PC running MED-PC IV software was used to present stimuli and record events with 0.01 s resolution.

2.4 Voluntary Consumption Training

Training for the voluntary consumption of TETS began at least two weeks prior to the day of poisoning. Each training session occurred in a disposable plastic cage that was placed in a fume hood, hereinafter referred to as the “feeding cage.” The feeding cage was divided in half to limit the area the rats could explore, had a stainless steel dish affixed to the floor for presenting food items, and had a layer of bedding (ALPHA-dri®; Shepherd Specialty Papers, Hanover, MA) that was changed whenever soiled.

Training began by feeding each rat several pieces of Froot Loops® cereal (Kellogg Company, Battle Creek, MI) with their chow over a weekend. The following Monday, training began in the feeding cage, with a single whole piece of cereal (180–220 mg) already in the dish. The rats were then free to discover and consume the cereal. Time to consumption was measured with a stopwatch that was started when all four of the rat’s feet were in contact with the bedding and stopped when the entire food morsel was no longer visible (i.e., it was consumed or in the rat’s mouth). Consumption times were initially long (~1–5 min), but with additional daily training the time to consumption dropped significantly (typically < 30 s). The cereal used during training was initially unadulterated and later was prepared with 50 uL of acetone, the TETS vehicle. Consumption training sessions occurred daily except for weekends and holidays, and each cohort (n = 15) received between 6 and 11 (M: 7.8, SD: 1.8) days of consumption training prior to poisoning. The number of days in which acetone was used as an adulterant varied between cohorts, with the restriction that in the five sessions leading up to poisoning there had to be at least one acetone and one unadulterated food morsel presented. The acetone was applied by pipette and allowed at least 30 min to evaporate prior to consumption training. On the day of poisoning, the TETS-acetone solution was used in place of the acetone alone, and the volume applied was appropriate to the target µg/kg dose for each individual rat.

2.5 Behavioral Training

All early training sessions delivered 100 food pellets. Each food pellet was also accompanied by a 0.1 s 70 dB 4500-Hz tone. The houselight provided general illumination throughout the entirety of the session and was darkened when a session ended. Water was freely available in a session via a dipper that was equipped with a 0.1 mL cup. The dipper cup was refilled by lowering the cup into a reservoir of water then raised again. Refilling occurred 3 s after the last dipper entry (i.e., a head entry into the dipper receptacle) to minimize the possibility of schedule-induced drinking. Of the two levers, each rat was randomly assigned to have either the left or right lever active, though both levers were always present in the operant chamber. The active lever remained constant across sessions, varying only between subjects (never between sessions). The active lever was signaled by illuminating the cue lights above it and never above the inactive lever.

2.5.1 Chamber acclimation and magazine training

The magazine training and acclimation session delivered a food pellet every 120 s on average. The purpose of this session was to allow the rats to acclimate to the operant chambers and to create an association between the tone and food pellet delivery into the food trough. Lever presses made during this session were recorded but had no programmed consequence.

2.5.2 Autoshaping

Following magazine training, all subjects were given an autoshaping session, wherein food pellets were again delivered based on the passage of time (120 s average). The cue lights above the active response lever were illuminated for 8 s prior to pellet delivery and were darkened concurrent with pellet delivery. If a press occurred during the 8-s interval in which the cue lights were illuminated, a food pellet was delivered immediately, and the cue lights were darkened. The purpose of these sessions was to engender lever pressing reinforced by delivery of a food pellet.

2.5.3 Progressive-ratio schedule

The progressive-ratio (PR) schedule session followed autoshaping and occurred for a single session. During the PR session, food pellets were delivered only after a requisite number of lever presses. The cue lights were illuminated for the entirety of the session, signaling the constant availability of food pellets. The initial number of lever presses per food pellet started at one and was increased by one for every 20 food pellets delivered. The session lasted for 100 reinforcers, so at the end of the session each food pellet was delivered after five lever presses.

2.5.4 Variable-ratio schedule

A variable-ratio (VR) schedule session followed the progressive-ratio session. During the VR session, food pellets were delivered after a variable number of responses. The response requirement was drawn without replacement from the following list: 1, 2, 3, 4, 5, 5, 6, 7, 8, 9 (i.e., a VR-5 schedule). No changes to the response requirement were made during the session.

2.5.5 Variable-interval schedule

The variable-interval (VI) schedule sessions made up the remainder of the behavioral training and started at a short interval and were gradually increased. In a VI schedule, a certain interval of time must elapse before the next response is reinforced. These intervals are variable and range from very short to very long intervals, but are defined by the mean interval. Each interval was selected without replacement from a geometric distribution of 12 items so that the passage of time did not reliably increase the probability of food-pellet availability [33]. The mean intervals used here were 20, 30, 45, 60, 75, and 90 s. A new interval from this progression was used for each session until the 90-s schedule was implemented. After two sessions of the VI 90-s schedule, the limit of 100 reinforcers per session was removed, and the session time was lengthened to five hours. During these lengthened sessions the rats earned approximately 170 food pellets. These 5-h sessions embodied the terminal parameters and occurred for six sessions prior to TETS poisoning. This constituted the pre-exposure baseline against which post-exposure performance was compared for each individual rat.

2.6 TETS Dosing

Rats were assigned to different dose groups either by body weight (for the overt-toxicity assessment) or by baseline lever-press response rates (for the behavioral-toxicity assessment). Group assignments were made by rank ordering rats according to weight or response, then splitting them into high, middle, and low thirds. Group assignment was then done randomly on each of these thirds, such that each dose group had equal representation of the high, middle, and low ranks.

Poisoned cereal morsels were prepared the afternoon prior to poisoning and left to dry overnight in the chemical fume hood. The process was identical to the procedure described for the voluntary-consumption training except the TETS-acetone solution was used in place of acetone alone and the preparation occurred at an earlier time point. The volume of TETS-acetone solution varied according to the dose and volume needed based on each rat’s body weight. At the time of exposure, rats weighed between 269.9 and 313.7g (M = 291.7g, SD = 9.1g). The cereal morsels were selected to be nearly uniform in size and their mass ranged from 180 to 220 mg.

The nominal doses of TETS used for the overt-toxicity portion of the experiment were 25, 50, 100, 200, 400, 600, 800, 1600, and 2400 µg/kg. However, these doses were based upon the assumption that the TETS was 100% pure. TETS aliquots were tested at key points throughout the study and immediately after the experiment, via use of both gas chromatography mass spectrometry (GC/MS) and nuclear magnetic resonance (NMR) spectroscopy, revealing that the TETS was ~78% pure and stable throughout the study. Therefore, measured doses used herein were: 19.5, 39, 78, 156, 312, 468, 624, 1248, and 1872 µg/kg. All doses from here onward will be referred to by the actual mean measured TETS dose.

The 24-h LD₅₀ of TETS was determined using the stage wise, sequential groups design [34–36]. Briefly, 1 to 3 rats were allocated randomly to each of 3–6 TETS challenge levels per stage (a dosing day). TETS doses were assigned randomly to the animals. In the first stage, a range of TETS doses were selected to span the predicted range of lethality from 0–100 percent, although the highest dose used did not produce complete lethality. The 24-h lethality results of the first stage were used to select TETS doses for the next stage. In the second stage, TETS doses were selected to refine the data of the previous stage, including increasing the upper dose range to include a dose that produced lethality in all rats. This approach allowed animals in subsequent stages to be placed at doses which would allow better estimation of the LD₅₀. After the last stage, probit dose-response models using maximum-likelihood estimation were fitted to the combined data for all stages [37]. Estimated parameters were used to calculate the LD₅₀ at 1, 5, 24, and 30 h. The delta method was used to compute a 95% CI for the LD₅₀ [37, 38].

Based on the overt-toxicity results, we selected a range of doses for the operant-behavior assessment that were expected to produce a range of behavioral intoxication as well an acceptable range of survival rates. The doses used for the behavioral-toxicity assessment were 19.5, 39, 78, 156, and 312 µg/kg.

2.7 Post-Exposure Assessment

2.7.1 Overt toxicity

Rats assigned to the overt-toxicity assessment were placed individually into dedicated transport cages for observation immediately after poisoning. Continuous observation occurred for one hour and then hourly checks occurred until five hours post-poisoning before the animals were returned to their home cages and given a measured amount of food. A 24-h observation recorded food wastage and body weights and then an additional observation occurred 30 h post-exposure, after which the rats were humanely euthanized by overdose of a pentobarbital-based solution. Continuous observations were recorded as a running narrative for each individual, and hourly observations were recorded similarly along with associated toxic signs scoring using a toxic signs scoresheet.

2.7.2 Behavioral toxicity

Following TETS poisoning, rats in the behavioral-toxicity assessment were immediately placed into the operant chamber and the session commenced (the same 5-h VI-90 s schedule used in the baseline was again used here). A second identical behavioral session was then conducted 24 h post-poisoning. Surviving rats were humanely euthanized 30 h post-poisoning.

2.8 Statistical Analyses

All statistical analyses were conducted using IBM SPSS Statistics 22. Baseline session data were collected from the five sessions preceding TETS exposure. Aggregated data are expressed as means ± standard error of the means (SEM). Median lethal dose estimates and associated 95% confidence intervals were obtained by fitting a probit function. Behavioral measures (i.e., lever presses, food pellets earned, food-trough entries, and water-trough entries) are shown as rates, calculated as the total number of events over the session length (300 min). All rates are expressed as events/min. Group differences were compared using one-way and repeated-measures ANOVAs followed by multiple comparison tests of pairs of groups using a Bonferroni correction. Statistical significance was set at p < .05.

3. Results

3.1 Voluntary Consumption

Consumption latencies, or the amount of time to consume the cereal morsel, were recorded when the cereal was unadulterated as well as adulterated with acetone (vehicle) or TETS. Figure 1 shows the distribution of consumption latencies from the session most proximal to TETS poisoning under each of these different conditions of adulteration for all subjects (the overt-toxicity and behavioral-toxicity groups; n=105). The majority of the rats consumed the cereal morsel in less than 30 s, and outliers above 60 s occurred only five times out of a possible 315 instances (1.6%). The latencies for the unadulterated (Median: 25.4s), acetone (Median: 25.0s), and TETS (Median: 25.4s) food types appeared equivalent. To verify this, the latencies were compared in a repeated-measures ANOVA, using session as a within-subject variable and consumption latency as the outcome variable. The ANOVA showed no difference in consumption latency as a function of the adulterant, F(2,208) = 0.14, p = .87. Thus, these rats could not discern the difference between cereal morsels that were unadulterated, subjected to acetone, or adulterated with the TETS-acetone solution, apparently confirming that TETS is tasteless and odorless to rats and, presumably, to humans as well.

Figure 1.

Time to consume the cereal morsels as a function of adulterant for all rats (both the overt-toxicity and behavioral-toxicity assessments). Each box indicates the interquartile range, the middle line indicates the median, and the whiskers extend to the 5^th and 95^th percentiles. There were no significant differences in consumption latency across adulterants.

3.2 Overt Toxicity

Following TETS exposure, the rats assigned to the overt-toxicity assessment (n = 45) were continuously monitored for an hour and then checked hourly thereafter for toxic signs. The typical progression of intoxication started with lethargy, which progressed into ataxia, piloerection, and the rat lying prone or in a hunched posture, followed by tremor, convulsions, and then death (at the highest doses). Lethargy typically presented as immobility and a lack of the explorative behaviors typically shown by the animals. Our description, described as lethargy, has also been referred to as quiescence [16, 26] and somnolence [25] by others. The ataxia observed was often characterized by poor motor control of the limbs while walking and sometimes animals dragged themselves along the floor instead of walking. Convulsions were typically of the tonic-clonic variety, characterized by a few seconds of muscle rigidity followed by a longer period of rapidly alternating muscle relaxation and contraction. These convulsive episodes were often preceded by tremors of escalating severity. Blood was sometimes noted on the animals’ snouts after convulsive episodes, presumably from traumatic contact with the cage walls during convulsions. Ejaculate was also sometimes found dried and caked near the rats’ penises after death or extended convulsions.

Rats that survived to 24 h following a 156 µg/kg dose or higher were often hypersensitive and would react disproportionately to normal stimuli. Sudden noise or being touched would cause the rats to exhibit an exaggerated startle response. However, once these subjects were picked up they would typically be very docile and still, exhibiting none of the typical responses that healthy rats do when picked up (sniffing, moving their heads, moving their bodies and tails to maintain balance, etc.). Figure 2 depicts the percentage of rats in the various dose groups that displayed an important subset of the observed toxic signs: lethargy, tremor, convulsions, and death. These toxic signs exhibited a clear dose dependency, with higher doses of TETS producing an increased prevalence and severity of intoxication. Convulsions and death sometimes occurred at the 156 µg/kg dose, and these same outcomes became more prevalent as the dose increased. In all of the fatal cases at the 312 µg/kg dose, convulsions preceded death, but, importantly, not all rats exhibiting convulsions died. Thus, the incidence of convulsions shortly after poisoning only imperfectly predicted fatality within 24 h at this moderate dose. At the 624 µg/kg dose all rats exhibited convulsions and died within 5 h.

Figure 2.

Percentage of subjects showing various overt toxic signs as a function of dose in the overt-toxicity assessment. Lethargy is shown by the white bars, tremor by the gray striped bars, convulsions by the dotted bars, and death by the black bars. Toxic signs prevalence and severity increased as a function of dose. Data shown here were collected up to 30 h post-exposure, though signs typically occurred within an hour (with the exception of death).

Lethality was documented at 1, 5, 24, and 30 h post-exposure in the overt-toxicity group to determine median lethal doses (LD₅₀). The probit functions for the 1-, 5-, and 24-h LD₅₀ estimates are displayed in figure 3. No deaths occurred between 24 and 30 h, so the 30-h estimate equaled the 24-h estimate. Several subjects that died approximately 70–80 min after exposure were not captured in the 1-h estimate but were captured in the 5-h estimate. The hashed line indicating 50% lethality intersects with each curve at the LD₅₀ value. The median lethal doses (and their 95% confidence intervals) for 1, 5, and 24 h were 505 (354 – 912), 387 (280 – 539), and 292 (192 – 387) µg/kg, respectively.

Figure 3.

Probit functions fitted to lethality data collected at 1, 5, and 24 h post-exposure in the overt-toxicity assessment. The hashed line at 50% lethality indicates the median lethal dose (LD₅₀). The 1-h curve is shown as the dashed line, the 5-h curve is shown as the dotted line, and the 24-h curve is shown as the solid line.

Of the subjects that survived to 24 h post-exposure, percent food wasted and percent body weight change were calculated at that time point. Figure 4 shows both of these measures as a function of TETS dose. The group sizes are denoted in the upper panel and vary as a function of survival rate. An ANOVA using dose group as a between-subjects factor revealed a main effect of TETS dose for both overnight percent food wasted, F(4,59) = 19.16, p < .001, η_p² = .57, and percent body weight change, F(4,59) = 12.12, p < .001, η_p² = .45. In general, the amount of food left uneaten and body weight loss increased as the TETS dose increased. Significant differences are denoted by letters in figure 4, wherein groups that share a letter are not significantly different and groups that do not share a letter are significantly different. Very little or no food was wasted by the 19.5 (M = 0.0%, SD = 0.0%) and 39 µg/kg (M = 2.8%, SD = 10.9%) dose groups, whereas significantly higher wastage was evident at higher doses (p < .01). No difference in food wastage was observed between the 156 (M = 60.2%, SD = 44.5%) and 312 µg/kg (M = 53.4%, SD = 30.6%) dose groups (p = .99), but please note that these data are only from the 24-h survivors. The 78 µg/kg dose group showed an intermediate percentage of food wasted (M = 26.8%, SD = 33.1%) and was only significantly different from the 156 µg/kg group (p = .018).

Figure 4.

The percent of food wasted overnight (top panel) and percent body weight change (bottom panel) as a function of dose for all rats that survived to 24 h post-exposure (the time when these measures were recorded) in both the overt-toxicity and behavioral-toxicity assessments. The number of subjects in each group is noted at the top. Error bars represent the standard error of the mean. Significant differences between dose groups are denoted by the letters, where groups that are not significantly different share a letter and groups that do not share a letter are significantly different. As dose increased, the amount of food wasted overnight and amount of bodyweight lost both increased. Because only subjects that survived to 24 h post-exposure were included in this analysis, the number of subjects represented in each group varied.

Presumably as a direct result of the increased food wastage, rats in the higher dose groups typically lost weight overnight. The 19.5 (M = 0.8%, SD = 1.7%) and 39 µg/kg (M = 0.6%, SD = 1.7%) groups gained weight normally and were not significantly different (p = .999), whereas the 156 (M = −7.9%, SD = 5.1%) and 312 (M = −7.4%, SD = 2.8%) µg/kg dose groups lost significantly more weight (p < .001). Rats in the 78 µg/kg dose group lost an intermediate amount of weight (M = −2.8%, SD = 3.8%) and were only significantly different from the 19.5 (p = .031) and 156 µg/kg (p = .001) dose groups. Thus, for both food wastage and weight loss, a clear dose-dependent effect was observed, but significant effects were only evident at doses of 78 µg/kg and higher.

3.3 Behavioral Toxicity

Rats in the behavioral-toxicity groups were immediately placed into the operant chamber after consuming the cereal morsel, and four separate measures of behavior were recorded in each 5-h session: lever presses, food pellets earned, food-trough entries, and water-dipper entries. Rates for each of these measures were calculated by taking the total number of events divided by the session duration and are expressed as events/min. The rates for each measure, proportioned to each rat’s individual baseline rates, are shown in figure 5. The top panel depicts data from the day of exposure (TETS), and the bottom panel depicts data from the session that began 24 h post-exposure (POST). A repeated-measures ANOVA was conducted using session (baseline, TETS exposure, 24 h post-exposure) and measure (lever presses, food pellets earned, food-trough entries, water-dipper entries) as within-subject variables and dose (19.5, 39, 78, 156, 312 µg/kg) as the between-subject variable. The results of the repeated-measures ANOVA are shown in table 1. There were main effects of session, measure, and dose, and all interactions (2-way and 3-way) were significant. Multiple comparison tests using a Bonferroni correction revealed several significant pairwise comparisons. The group means and standard error of the means for each of the behavioral measures are shown in table 2. The italicized values indicate where rates were significantly different from baseline.

Figure 5.

Percent of baseline rates for four separate behavioral measures as a function of dose in the behavioral-intoxication assessment. Each session shown here was 5 h in duration. The top panel depicts data from the session immediately following TETS exposure, and the bottom panel depicts data from the session that occurred 24 h post-exposure. Lever-pressing rates are shown as black bars, reinforcers (food) earned are shown as dotted bars, food-trough entries are shown as the gray striped bars, and water-dipper entries are shown as the white bars. Error bars represent the standard error of the mean. Rates were suppressed more severely as dose increased, and lever-pressing rates were the most sensitive measure. For most doses and measures, no recovery was observed at 24 h post-exposure.

Table 1.

F-ratios for a Repeated-Measures ANOVA Comparing Rates of Four Dependent Behavioral Measures (Lever Presses, Reinforcers Earned, Food-Trough Entries, and Water-Dipper Entries) as a Function of Session, Measure Type, and Dose in the Behavioral-Toxicity Assessment

Effect	F	df	P	ηp2
Session	368.43	2,110	< .001	0.87
Session × Dose	42.53	8,110	< .001	0.76
Measure	2123.66	3,165	< .001	0.97
Measure × Dose	92.52	12,165	< .001	0.87
Dose	75.50	4,55	< .001	0.85
Session × Measure	325.87	6,330	< .001	0.86
Session × Measure × Dose	40.16	24,330	< .001	0.74

Note: partial eta squared represented by η_p²

Table 2.

Group Means and Standard Error of the Means of Rates (expressed as events/min) for Four Dependent Behavioral Measures in the Behavioral-Toxicity Assessment

		Lever Presses		Reinforcers Earned		Food-Trough Entries		Water-Dipper Entries
Dose (µg/kg)	Session	Mean	SEM	Mean	SEM	Mean	SEM	Mean	SEM
	BL	15.09	2.85	166.73	2.25	5.99	1.05	1.64	0.32
19.5	TETS	11.06	1.42	163.75	10.15	5.31	0.70	1.65	0.32
	POST	14.47	2.10	162.42	6.61	5.78	0.83	1.55	0.28
	BL	12.99	2.85	166.38	2.25	5.61	1.05	2.06	0.32
39	TETS	7.40	1.42	140.25	10.15	5.76	0.70	1.52	0.32
	POST	7.46	2.10	126.17	6.61	4.97	0.83	1.81	0.28
	BL	10.48	2.85	168.92	2.25	4.74	1.05	1.92	0.32
78	TETS	2.60	1.42	56.92	10.15	3.15	0.70	0.98	0.32
	POST	1.78	2.10	62.92	6.61	3.18	0.83	1.09	0.28
	BL	14.13	2.85	164.77	2.25	5.88	1.05	1.07	0.32
156	TETS	1.50	1.42	14.83	10.15	1.38	0.70	1.14	0.32
	POST	0.21	2.10	5.25	6.61	1.31	0.83	0.84	0.28
	BL	12.37	2.85	166.58	2.25	6.08	1.05	1.28	0.32
312	TETS	1.16	1.42	1.58	10.15	1.31	0.70	1.92	0.32
	POST	-	-	-	-	-	-	-	-

Note: values significantly different (p < .05) from baseline (BL) are italicized and shaded

Behavioral intoxication was evidenced as response suppression across multiple dependent measures (p < .001). This suppression increased directly as a function of dose, and was most evident in the higher dose groups. Despite mean rates of less than 80% of baseline lever presses, the 19.5 µg/kg group showed no statistically significant intoxication (i.e., response suppression) at the level of session averages (p = .163). The 39 µg/kg group showed a modest level of behavioral intoxication (p = .026), and lever presses was the measure most affected. Behavioral suppression was severe in the 78 µg/kg group (p = .001), wherein rates of all four measures were less than half that observed during baseline, and among the four measures, lever presses was again the most sensitive measure of intoxication. The 156 µg/kg group exhibited response rates ranging from near zero (lever presses) to a high of 23% (dipper entries) of baseline responding (p < .001). This suppression was even more pronounced in the 312 µg/kg group (p < .001), which had rates of less than 1% of baseline for lever presses, food pellets earned, and trough entries, and only 7% of the baseline rate for dipper entries. The 312 µg/kg group has no visible data in the bottom panel of figure 5 because of the fatalities that occurred, leaving no survivors for the 24 h post-exposure behavioral assessment. Across all of the dose groups, lever presses and reinforcers earned appeared to be the most sensitive measures for detecting intoxication, affecting the 39 µg/kg and higher doses. Food-trough entries were only significantly decreased in the 156 and 312 µg/kg groups, and the water dipper entries were only significantly decreased for the 78 µg/kg group. Across all dose groups and behavioral measures, no significant differences were observed between day of exposure and 24 h post-exposure (p = .911). This indicates that no behavioral recovery occurred in the 24 h between sessions and that behavioral intoxication persisted into the following day.

Although the session averages of lever-pressing rates changed dose dependently, these averages are unable to provide a clear depiction of within-session patterns of lever presses. To investigate the temporal patterning of operant lever presses within the session, we examined pausing (i.e., the absence of lever presses). All pauses exceeding 120 s in length were extracted and the durations were summed. These sums were then converted to a percent of the total session and are depicted in figure 6. The baseline session is shown in white, the session on the day of TETS exposure is shown in gray, and the 24 h post-exposure session is shown in black. Across all groups, the baseline duration of pausing within the session was very low, accounting for only 0.7% to 4.7% of the total session. To determine how pausing changed as a function of TETS exposure, pause durations were analyzed in a repeated-measures ANOVA as a function of dose and session. There were main effects of both session, F(2, 82) = 81.69, p < .001, η_p² = .67, and dose, F(3, 41) = 295.42, p < .001, η_p² = .88, and the session × dose interaction was also significant, F(6,82) = 15.42, p < .001, η_p² = .53. Post-hoc comparisons were then conducted and significant differences from baseline are denoted as asterisks in figure 6. On the day of TETS exposure and 24 h post-exposure, within-session pausing significantly increased for the 39, 78, 156, and 312 µg/kg groups. In addition, pause durations were equivalent between the day of exposure and 24 h post-exposure, indicating that no recovery occurred overnight for any of the groups that showed increased pausing (i.e., all groups except 19.5 µg/kg). The percent of the session spent pausing increased as a function of dose. The 19.5 µg/kg group paused for 3.0% of the session on the day of exposure, whereas the 39 µg/kg group paused for 16.6% of the session. The 78 µg/kg group paused for 68.0% of the session, the 156 µg/kg group paused for 91.0% of the session, and the 312 µg/kg group paused for 99.3% of the session on the day of exposure. All groups were significantly different from one another (all p < .03) except for the 156 and 312 µg/kg groups (p = .14).

Figure 6.

Percent of the session spent pausing (interresponse times ≥ 2 min) as a function of dose in the behavioral-intoxication assessment. Baseline sessions are shown as white bars, the session immediately following TETS exposure is shown as gray bars, and the session that occurred 24 h post-exposure is shown as black bars. Error bars represent the standard error of the mean. Asterisks indicate doses where pausing significantly increased from baseline in the TETS and post-exposure sessions. No significant recovery occurred the day after exposure for any dose group.

The pause analysis showed that low, consistent rates of lever presses were not typical for TETS-exposed rats, but instead the predominant pattern was lever-pressing punctuated by lengthy pauses. To further analyze the response rates across the entirety of the session, we binned lever-pressing rates into 5-min blocks and averaged within each dose group. These data are shown in figure 7: the solid line represents the baseline sessions, the hollow line represents the sessions immediately following TETS exposure, and the dotted line represents the 24-h post-exposure session. The shaded gray area is the 95% confidence interval for the baseline data. The 312 µg/kg group was not included in this analysis because of the high level of mortality and corresponding lack of data prior to the end of the 5-h behavioral assessment. Across all doses, a clear pattern emerged with intoxication occurring rapidly (within 5 to 10 min), and decreased response rates were evident for all groups. However, the degree and duration of response-rate suppression increased with larger doses of TETS. Response-rate suppression was observed at the start of each session for all doses, as lever-pressing rates were below baseline levels. The 39 µg/kg dose decreased initial response rates to 50% of baseline levels. The decrease was even greater in the 78 and 156 µg/kg dose groups, with response rates approximating only 25% of baseline levels. Thus, the severity of TETS intoxication increased with dose, with the most profound response suppression occurring in the 78 and 156 µg/kg groups. It should be noted, however, that even the 39 µg/kg group showed a prolonged period of response suppression, with response rates approximating 50 to 75% of the baseline rate throughout the entirety of the session. Lastly, the rate of recovery was affected by dose. The 19.5 and 39 µg/kg groups recovered the fastest, whereas the 78 and 156 µg/kg groups did not exhibit any significant recovery throughout the entire 5-h assessment or the following day.

Figure 7.

Lever-pressing rate analyzed in 5-min blocks across multiple sessions as a function of dose. The baseline sessions are shown as the black line, the session immediately following TETS exposure is shown as the hollow line, and the session that occurred 24 h post-exposure is shown as the dotted line. The gray shaded area indicates the 95% confidence interval for the baseline session. Behavioral intoxication increased as dose increased, and the suppression of lever presses was more severe and protracted. Onset of intoxication was often rapid (within 5 to 10 min), and recovery did not occur in the session that occurred 24 h post-exposure.

Due to the rapid onset of intoxication, an additional analysis with greater granularity than 5-min blocks was performed to further characterize the onset of intoxication. Lever presses in the first five minutes of the session were analyzed to determine when each rat emitted its first pause exceeding 300 s. If a rat did not press the lever within the first 300 s of the session, we counted that as a pause occurring at the beginning of the assessment (t = 0). We conducted a Kaplan-Meier analysis using these times to the first pause as our event of interest. Figure 8 shows these data, and each line represents a different dose group. The analysis clearly showed that behavioral intoxication was rapid and typically occurred within 5 min, although a trend of dose-dependent onset was evident. That is, higher doses produced a more rapid onset of pausing. Nearly all rats paused within 180 s of poisoning in the three highest dose groups. All rats in the 39, 156, and 312 µg/kg groups, and all but one rat in the 78 µg/kg group, paused during the first 5 min of the assessment. The higher dose groups also showed signs of immediate intoxication, as rats did not press the lever in the first 300 s, a finding not evident in the 19.5 and 39 µg/kg groups. In the 19.5 µg/kg group, 48% of the subjects stopped responding during this 5-min period. The median time to the first pause for each dose is shown in table 3, as are individual Wilcoxon pairwise comparisons of the latency to pause onset. The median time to the first pause was not determined for the 19.5 µg/kg group because 7 out of 12 subjects did not have a ≥ 5 min pause. However, for those subjects that did pause, their median time to the first pause was 206 s. As the TETS dose increased, the median time to the first pause decreased and equaled 72 s for the 312 µg/kg group, underscoring the rapid onset of TETS intoxication following voluntary oral ingestion. The individual pairwise comparisons also revealed that doses were significantly different from one another, with the exception of the 78 and 156 µg/kg comparison and the 156 and 312 µg/kg comparison. Even though this analysis was only for the first 300 s of the session, clear dose-dependent differences were evident, and the response suppression was profound at the higher doses. It should be noted that this analysis does not determine why the subject paused for > 300 s, just that a protracted pause did occur. It is a certainty that some subjects were experiencing tremor and even convulsions while others were simply lethargic or immobile. Nevertheless, this measure clearly shows the interruption of lever presses, which in this model represents a high-probability functional class of operant responding. Regardless of the reason for the pause, it functions as a good behavioral measure for determining intoxication and it is reasonable to assume that such doses would functionally disrupt other operant responses in other contexts (e.g., driving, the ability to assist others, and self-care behaviors).

Figure 8.

Kaplan-Meier analysis using time to the first pause (≥ 5 min) as the event of interest as a function of dose in the behavioral-intoxication assessment. Time to the first pause decreased as a function of dose and higher doses showed immediate intoxication (t = 0) more frequently.

Table 3.

Wilcoxon Pairwise Comparisons for Time to First Pause and Median Time to First Pause in the Behavioral-Toxicity Assessment

Dose	Pairwise Comparison					Median Time to First Pause (s)
	19.5	39	78	156	312
19.5	-					n.d.
39	0.00	-				162
78	0.00	0.05	-			124
156	0.00	0.04	0.52	-		116
312	0.00	0.00	0.03	0.38	-	72

Note: median time to first pause was not determined (n.d.) for the 19.5 µg/kg dose group because the majority of subjects (7/12) did not have a ≥ 5 min pause; significant p-values (< .05) are italicized and shaded

4. Discussion

The current experiment was successful in establishing a voluntary food-consumption model in rats to evaluate TETS poisoning. Rats consumed the TETS promptly and reliably and could not apparently discern either the TETS or the acetone vehicle as consumption latencies were comparably short (< 30 s) across all adulterants (including none). The voluntary consumption of an acute poison is both a novel and realistic model for mass-poisoning scenarios. Much of the literature that uses voluntary consumption falls into one of several types: chronic low-level exposures (e.g., [39]), descriptions and analyses of events that previously occurred (e.g., [40, 41]), masking/delivering medicinal doses (e.g., [42–44]), or delivering a substance that has reinforcing properties on its own, such as ethanol or morphine [45–48]. There is a severe lack of information regarding the oral consumption of acute poisons, despite the fact that many chemicals are oral hazards. Gavage or IP injections are often selected as substitute routes of exposure for the oral route, likely because they are simpler and dosing is easily controlled. The voluntary consumption model presented here still allows for accurate dosing, but does require more time investment for training subjects to promptly and reliably consume the food item. However, one advantage which should not be overlooked is the fact that gavage produces a stress response that can persist for up to an hour [49] and could alter experimental findings. The voluntary consumption model presented here appears to produce no stress and the delivery of the poison through the mouth and esophagus to the stomach is essentially guaranteed.

In this work, our voluntary-consumption model was first used to quantify overt toxic signs and determine the median lethal dose following TETS consumption. The progression of TETS toxic signs generally proceeded from lethargy, ataxia and lying prone, to tremor, convulsions, and death at the higher doses. The LD₅₀ values were measured at 1, 5, 24, and 30 h after poisoning, and these latter two estimates equaled 292 µg/kg. From the initial study, a broad range of doses (approximating 0.07 to 1.07 times the 24-h LD₅₀) was selected for more intensive behavioral assessment. Even at the lowest dose (19.5 µg/kg), transient behavioral intoxication was observed followed by complete recovery of operant performance. TETS intoxication typically developed in less than 5 min under all doses. As the TETS dose increased, the onset latency of operant behavioral suppression decreased (figure 8), and the duration (figure 6) and severity (figures 5 and 7) increased, with no recovery observed at the three highest doses (78, 156, and 312 µg/kg). Overnight food wastage and body weight loss were also evident, and the severity of these two overt toxicity measures, like behavioral intoxication, was dose dependent.

The LD₅₀ estimates for TETS established in the present study are higher than those previously reported using different exposure methods or are higher than the lethal doses reported by others [13–16, 26, 30]. This may be due to differences in the species, vehicle, volume, greater level of stress produced by gastric gavage/IP injection relative to voluntary food consumption, or some combination of these factors. Acetone served as the vehicle for the current study and evaporated completely before TETS poisoning, whereas previous studies have used DMSO and saline solutions. Because we used an acetone vehicle to place TETS onto the food morsels, we did not have a means to compare our vehicle volumes to those used in previously published IP and gavage studies; however, the mass of the cereal morsels used in our study approximated only 0.1% of the rats’ body weights. Based on the information available from the published literature, the factor(s) responsible for differences in LD₅₀ estimates cannot be clearly deduced at this time. Presumably, studies could be performed to clarify the relative role of these factors in TETS toxicity. It is our hope that other laboratories will adopt the more real-world analogous voluntary-ingestion model developed herein.

The toxic signs observed following TETS exposure that have been discussed thus far have been primarily lethargy, tremor, and convulsions, but other symptoms were also observed and are worth mentioning here. At moderate doses, piloerection and hunched posture were observed in nearly all conscious, non-convulsing animals. We also observed hypersensitivity in many of the rats, especially at higher doses. Noises or being touched, which would cause minimal reaction pre-exposure, led to an exaggerated startle response. The severity of the hypersensitivity was surprising, as many of these animals would jump with enough force to knock a cage lid ajar. Precisely quantifying the degree of hypersensitivity and latency to its onset would be difficult, though the presence or absence of such hypersensitivity was often obvious.

At the 78 µg/kg dose, toxic signs were variable between subjects, and this group exhibited the broadest range of intoxication levels, from very little overt toxicity to moderate/severe toxicity. Also at higher doses (156 and 312 µg/kg), some animals died within the operant chamber environment, and a few of these animals were found with their teeth tightly clenched over the stainless steel grid bars comprising the floor. One further note is that rigor mortis seemed to be accelerated in the rats that died after TETS exposure, likely due to the extreme physical exertion associated with prolonged convulsions which precede death caused by convulsant compounds [50, 51].

The behavioral-toxicity assessment was successful in characterizing the time-course and severity of TETS intoxication across multiple measures, including lever presses, food pellets earned, food-trough entries, and water-dipper entries. Among these measures, changes in lever presses and food pellets earned were the most sensitive indicators of intoxication. This is perhaps not surprising given that this response may be viewed as the most effortful among those measured and that the number of food pellets delivered was correlated with lever-pressing rate. In contrast, changes in food-trough and water-dipper photobeam breaks were the least sensitive behavioral measure for characterizing dose-dependent changes in TETS intoxication. In comparison to lever presses, photobeam breaks are a much less effortful response. These responses are also more directly associated with a reinforcer location (food or water) and an immediate consummatory response, which may enhance its perseveration despite chemical intoxication.

TETS is a GABA_A antagonist that disrupted operant performance and decreased food consumption. Other GABA antagonists administered in lower, less toxic doses have increased free-feeding food consumption [52]. It is likely that the inappetence produced by TETS is not directly related to GABA antagonism but is instead resulting from general malaise that occurs post-poisoning. Rats exposed to high doses of TETS did not recover within 24 h, so decreased food consumption in the hours following poisoning is not surprising. Other GABA agonists and antagonists have been shown to modify the operant performance of ethanol self-administration (e.g. [53, 54]), reduce response rates on a fixed-interval schedule of food reinforcement in rats [55], and disrupt performance in delayed matching-to-sample tasks and serial-probe recognition in nonhuman primates [56, 57].

It is clear that GABAergic compounds affect operant performance [58, 59] and the results from the current study are similar to those found in previous studies (i.e., reduced response rates), though it is difficult to determine to what degree the operant disruptions observed here are unique to TETS and other GABA antagonists. It is likely that the changes in operant performance observed in the current experiment are a result of incompatible behaviors (e.g., ataxia, convulsions, reduced appetite, headache, nausea, etc.) occurring that would result from any number of ingested poisons. A sufficiently high dose of any poison would likely produce similar disruptions in operant behavior, though the latency to onset, severity of intoxication, and rate of recovery would vary across poisons.

Examples of this can be seen in an operant assessment using a fixed-ratio schedule to detect intoxication after orally ingested pyridostigmine bromide [60] and perchloroethylene [61]. Both studies showed clear intoxication following oral ingestion of the poison, but the latency to intoxication was faster following perchloroethylene ingestion. Importantly, the study by Warren et al. [61] was limited in part by use of the gavage technique, which caused transient perturbations in operant performance even under control conditions, further underscoring the value of the present voluntary ingestion model. Warren’s simple schedule of operant reinforcement, like ours, was nevertheless capable of discerning these effects from those of the poison, and also clearly elaborated the time-course and dose-dependency of the intoxication. Thus, these operant assessments can use any number of schedules of reinforcement to characterize the time-course and degree of intoxication, lending further credence to the strength of this general approach (i.e., within-subjects operant behavioral assessment). The intent of the current study was to use an assessment that would be capable of detecting intoxication from virtually any ingested poison, not just GABA antagonists or other convulsants. As such, the operant procedure used here was sensitive enough to detect changes at very low doses of an ingested poison (~0.07 LD₅₀) and yet presumably general enough to be used to examine very different classes of poisons. In addition, this assessment appears suitable to capture multiple dependent measures and differentiate various types of disruption (i.e., long initial pauses, changes in interresponse time distributions, changes in latencies to collect reinforcers) when doses are low enough to disrupt operant performance but not so high that responding is completely suppressed, providing another valuable tool for comparing across different doses and classes of ingested poisons.

One of the more important findings from the current study is the sensitivity of the different dependent measures. For doses at which overt toxicity was not readily apparent (39 µg/kg), behavioral intoxication was detectable and more pronounced. In addition, one of the more reliable overt signs of intoxication was the amount of food wasted overnight and subsequent weight loss. Animals that wasted more food performed worse in the behavioral assessment than those that ate all the food provided. While it is much easier to collect overt toxicity data than behavioral intoxication data, the latter provides a more comprehensive and sensitive assessment for TETS intoxication and was capable of parsing important temporal patterns of intoxication and recovery. However, one limitation of the operant assessment is that we cannot determine if TETS intoxication led to inappetence and therefore inhibited operant performance. One solution to this problem would be to use a behavioral assessment such as two-way active shuttlebox avoidance that does not use a food reinforcer, but instead relies upon aversive stimulation (e.g., electric shock). These assessments are currently underway in our laboratory. Pilot (unpublished) data suggest that TETS produces behavioral deficits even when an aversive stimulus (and not a food reinforcer) is used.

Our analysis revealed that the onset of behavioral intoxication occurs within 5 min, but overt toxicity was not as well correlated with this finding. Lethargy was the first toxic sign typically observed, but this sign can be difficult to quantify in a holding cage environment, wherein little activity may be observed under normal conditions (i.e., in the absence of poisoning). We were able to quantify lethargy by ensuring that rats undergoing overt toxicity assessment did not remain in their transport/holding cage long before poisoning, thus preserving their exploratory behavior. Without this, determining when lethargy onset occurred would be difficult, as lethargy typically presented as immobility or a lack of exploration. Also, unlike the operant-behavioral assessment, time to intoxication using toxic signs scoring is not only labor intensive, but is also prone to variability in scoring across time and between observers. The automated behavioral assessment also allowed for each subject to serve as its own control, further enhancing the ability to detect subtle changes in behavior. Within-subject controls are also less time-consuming, easier to quantify, and provide a throughput advantage for behavioral assessments.

Taken together, the current voluntary oral-ingestion model not only mimics real-world scenarios of TETS poisoning, but is also flexible enough to be used in conjunction with other procedures. Treatment efficacy can easily be added to this model, additional types of behavioral assessment are possible, and investigation into long-term recovery is done simply by continuing to run behavioral sessions daily after exposure. Although TETS is a potential agent for food-based terrorism, much still remains to be discovered about this chemical. Realistic models of exposure, comprehensive behavioral assessments, and multifaceted treatment investigations are needed to elaborate TETS toxicity and to discover and optimize effective treatments against this potent poison.

Highlights.

1. Tetramethylenedisulfotetramine is a rodenticide implicated in over 14,000 poisonings.

2. Rats were trained to rapidly and voluntarily consume a poisoned food morsel.

3. Overt/behavioral toxic effects of TETS across a range of doses were characterized.

4. Intoxication onset was rapid (< 5 min) and observed across multiple operant measures.

5. At most doses, recovery of operant performance did not occur within 30 hours.