Agents and Treatments
This chapter deals with a wide assortment of experimental treatments: drugs and toxicants; exposure to heat, light, or sound; modification of nutrients; induced exercise; and sleep deprivation. Although seemingly dissimilar, those treatments all have the potential for inadvertent injury to experimental animals if the animals are not carefully monitored, especially if stimuli are introduced with a mechanical device, such as a treadmill or hotplate. The first section updates material in the NIH report Methods and Welfare Considerations in Behavioral Research with Animals (NIH, 2002, pp. 57–66).
PHARMACOLOGICAL AND TOXICOLOGICAL AGENTS
Drugs and toxicants are administered for various purposes (Goldberg and Stolerman, 1986; van Haaren, 1993; Weiss and O’Donoghue, 1994). A drug or toxicant may be administered to:
observe neurobehavioral effects to determine whether a drug can alleviate health problems (such as pharmacotherapy for behavioral and neurologic disorders),
determine how a chemical causes toxicity, to characterize the abuse liability of a new pharmaceutical,
determine whether an organism’s response to a drug changes with chronic exposure and whether chronic exposure may lead to abuse or physical dependence,
examine a chemical that is known or hypothesized to have specific
neurobehavioral effects that the investigator wishes to understand in more detail (for example, drugs that block a particular neurotransmitter receptor system can help to determine the neurotransmitter’s role in modifying specific behaviors),
produce a specific neurological state (such as anxiety)
help researchers to understand the biologic and behavioral consequences and possibilities for therapy. (Weiss and O’Donoghue, 1994.)
Behavioral and Environmental Considerations
Some neurobehavioral experiments involving drug administration use animals that are trained to perform a response that can be measured objectively. The motivation for the response may be delivery of food or water, or a drug, as in drug self-administration studies (see next section). Trained responses usually involve operating a lever or switch. Other dependent variables may also be measured, such as feeding, drinking, locomotion, or exploratory activity (Iversen and Lattal, 1991; van Haaren, 1993; Wellman and Hoebel, 1997). The research methods reviewed here involve a known substantial risk to humans or animals from exposure to drugs and other chemicals. Additional information about behavioral tests that can be used to screen unknown drugs or genetic mutants is provided in Chapter 9.
Situations requiring special housing or feeding arrangements were summarized in the earlier NIH report (NIH, 2002, p. 58):
Exposure to drugs usually necessitates individual housing in order to permit repeated access to each animal for dosing and testing. Individual housing also may be preferred because, in a group situation, drug-altered behaviors may increase an animal’s risk of abuse by cage mates, as well as impair its ability to compete for food. For animals in studies of intravenous drug self-administration or of constant intragastric infusion, the animal may be fitted with a vest and tether apparatus to protect the chronically indwelling cannula. Behavior may be measured in the animal’s living cage, to which devices for presenting stimuli and recording responses have been attached (Ator, 1991; Evans, 1994). Such arrangements may preclude conventional group housing. Experiments in neuropharmacology often employ restricted access to food or water for two purposes: (1) to maintain a consistent motivation of behavioral performance (Ator, 1991) and (2) to standardize content of the digestive tract for uniform absorption and uptake of orally administered drugs. This involves scheduling the availability of food and water but not necessarily deprivation. In addition, for experiments that take place over many weeks, it may be important to keep the total amount of drug delivered relatively constant, even when drug doses are calculated on a per weight basis.
To determine dose-effect relationships, a range of doses is selected—from one that produces little or no effect to one at which significant or even toxic effects are seen. Dose-effect relationships may be determined by studying single doses given to separate groups of animals (between-subject designs) or by determining a full dose-effect relationship for each animal (within-subject, or repeated-treatment designs). Baseline performance usually is reestablished between sessions during which a drug is given. In drug-interaction studies, two doses of different drugs, are given at appropriate intervals before the experimental endpoints are recorded. Cumulative dosing procedures permit increasing doses of a drug to be administered within a relatively short period, and a brief experimental session is conducted after each dose. The effects of the drug are assumed to accumulate in an additive manner so that within a period of 2– 3 hours the effects of a range of doses can be determined (Lau et al., 2000; Wenger, 1980).
Drug self-administration experiments determine the drug’s reinforcing efficacy, which may indicate the drug’s potential for abuse. The animal controls the number and frequency of delivery of the test drug. That is, a quantity of a particular drug is available intravenously, orally, or via inhalation, and the subject of interest is the amount of behavior maintained by this drug at the self-administered dose. In such studies, the dose available is varied across experimental conditions, and the rate of responding to obtain the dose, the number of drug deliveries obtained, and/or the amount of drug taken are the primary dependent variables of interest. In such studies, the likelihood that the animal will produce a fatal overdose is carefully considered in the design and choice of drug. Drugs vary across classes in how likely it is that high drug doses will produce adverse effects. Overdose may be minimized by placing an upper limit on the number of doses per session or on the minimum time-lapse between doses, or by setting the magnitude of each dose available to the animal (NIH, 2002, p. 59).
Drugs for animal research are often in solid form and must be dissolved or suspended in a liquid vehicle to be administered. Sterility of the vehicle is crucial, especially when it is administered intravenously. Aqueous vehicles, such as sterile water and saline solution, have no pharmacologic action of their own in appropriate volumes; however, many drugs require more complex vehicles, for example, such an organic solvent as propylene glycol or an alcohol. Testing with the vehicle, without a drug, will provide a control for the vehicle’s influence on the function being studied and for any effects of the drug-administration procedure itself. A vehicle or vehicle-drug combination may irritate tissue. Irritation can be minimized by using less concentrated solutions or alternating injection
sites. If less concentrated solutions require volumes that are too large for a single injection site, delivery may be made by small-volume injections at different sites. In some cases, one can adjust the pH to something more similar to physiologic pH by adding another chemical after the drug is dissolved.
The route of administration may be dictated by the need to use methods comparable with those of previous neuroscience studies, by constraints on the solubility of the drug, or by a desire to match the route used in humans. The routes of drug administration include oral, subcutaneous, intramuscular, intraperitoneal, intragastric, intravenous, inhalation, or intracranial (for example, into the ventricles or a specific brain region).
Injection by hypodermic needle is the most common way of administering drugs (van Haaren, 1993). The site of injection may be determined by the characteristics of a particular drug’s absorption or by the vehicle in which it is given. A common problem is the incorrect site of intraperitoneal injection into rodents. Research staff should be trained to avoid injection into the liver, intestines, or bladder instead of the peritoneal cavity. Success of injections also can be improved by prior adaptation of animals to the handling and restraint that normally accompany injection.
Insertion of a cannula into a blood vessel, a body cavity, or the nervous system is another method of administering drugs. A permanently implanted cannula ensures that repeated injections can be given at precisely the same site and permits the study of drug effects without peripheral effects, such as pain at the injection site (Waszczak et al., 2002). Self-administration studies often use the intravenous route with a chronically indwelling venous cannula (Lukas et al., 1982). The cannula generally is guided subdermally from the intravenous implantation site to exit in the midscapular region. The animal may wear a vest that covers and protects the cannula system. There are also methods for intraventricular drug self-administration through cannulae implanted directly into the brain (Goeders and Smith, 1987).
Implantable pumps for slow drug delivery are also used for chronic delivery, as in studies of drug tolerance or physical dependence (Tyle, 1988). Aseptic technique is important in the implantation of cannulae or pumps and in assessing the system (for example, to reattach tubing or add drug solution), to reduce morbidity and prolong the useful life of the cannulae or pumps.
Inhalation is a common route of exposure for such agents as cocaine, anesthetics, and smoke (e.g., Carroll et al., 1990). Some compounds are easily administered in nasal sprays, but inhalation exposures usually require specialized equipment to measure the amount of drug exposure and to prevent leakage of the airborne chemical (Liu and Weiss, 2002; Paule et al., 1992; Taylor and Evans,
1985). The risk of hypoxia requires attention when drugs are administered by inhalation for long durations.
Oral administration can be used for drug self-administration research (Meisch and Lemaire, 1993). A specialized drinking spout (often termed a drinkometer) regulates the volume of each drink to control drug dose. That permits study of the drug’s reinforcing efficacy. Acquisition of taste aversion is studied with oral administration of a toxic substance that serves as an unconditioned stimulus to produce illness. Oral administration of the toxicant can be controlled with a surgically implanted intragastric cannula (Touzani and Sclafani, 2002).
Oral administration is advantageous for chronically administered drugs because dosing may be accomplished without daily handling and intubation if the compound is added to the animal’s food or drinking water, as in studies of alcohol self-administration (Cunningham and Niehus, 1997) and exposure to toxic contaminants in food and water (Carpenter et al., 2002; von Linstow Roloff et al., 2002; Weiss and O’Donoghue, 1994). Special feeders and water canisters (Evans et al., 1986) are available to prevent spilling. When a drug is added to food or water, ingestion should be monitored both to determine the amount of drug consumed and to identify reduction in ingestion resulting from reduced palatability. If chronic drug exposure reduces consumption of the food, a control group (for example, pair-fed or pair-watered controls, or in studies done prior to weaning, controls that have restricted access to the lactating dam) should be used to determine whether results are attributable to the drug or to the reduced caloric or fluid intake. Drugs also can be given orally by gavage needle (for example, in rats and pigeons), by nasogastric tube (in monkeys), or in a gelatinous capsule (in monkeys).
Animal Care and Use Concerns Associated with Toxicity or Long-Lasting Drug Effects
Some chronic drug experiments involve dosing that produces cumulative deleterious effects. The animal-use protocol should include a contingency plan to define the conditions under which deleterious effects will be alleviated or an animal will be removed from the experiment. Some drugs may have long-lasting effects on feeding and drinking, on activity level, and on bodily functions such as elimination. However, other causes of behavioral changes during a drug study, such as irritation at an injection site or dental problems that affect food consumption, must also be examined.
In neuroscience experiments involving chronic drug exposure—for example, to study possible deterioration of performance after repeated exposure to a neurotoxin or the development of tolerance of an initial effect of a drug—attention must be given to the duration of drug exposure and the disposition of the animals. The decision to end chronic drug exposure should be based on predetermined criteria related to a range of changes from baseline that will be considered signifi-
cant. The observation of overt signs of toxicity, however, may necessitate a decision to terminate treatment earlier than expected. Daily observation of animals by someone familiar with the experimental protocol is especially important so that timely decision-making can occur.
Many dosing regimens do not produce long-term effects or behavioral impairment. After an appropriate washout time, the neuroscientist can determine the existence of long-lasting or irreversible effects (Bushnell et al., 1991). Irreversible effects do not pose a problem if the animal use-protocol calls for the animal to be euthanized to obtain cellular data to supplement functional results. A factor in the decision to euthanize is whether drug exposure has permanently altered a physiologic or behavioral function in such a way as to make providing adequate care for the animal difficult or to compromise continued humane use of the animal. But such an animal would be a valuable resource if the aim of the research is to understand mechanisms of tolerance, postexposure recovery, or therapeutic interventions that ameliorate long-lasting drug effects.
The previous section addressed a wide array of issues related to acute and chronic effects of various chemical agents, including drugs. This section extends that discussion by focusing on issues related to the testing of drugs that are of interest because their chronic use or exposure produces neuroadaptations thought to underlie the behavior patterns (such as tolerance and sensitization, dependence, and withdrawal) that characterize addiction to alcohol, nicotine, cocaine, heroin, and other abused drugs. Neuroscientists study the brain mechanisms that establish and maintain addiction in order to identify and characterize variables that affect risk (for example, genotype, environment, and experience) and to develop methods for treating addictive behavior and preventing relapse (e.g., Koob and Le Moal, 2001). Neuroscientists are also interested in characterizing the neurobiologic consequences of chronic exposure to addictive agents (such as changes in brain structure or function) (Becker, 1996; Obernier et al., 2002) and the process of recovery from deficits induced by such exposure.
Studies of addictive agents often require attention to dose, route of administration, vehicle, and other variables discussed in the previous section (see also NIH, 2002). When drugs are to be administered with abuse potential, the possibility that an animal will receive a harmful overdose must be carefully considered in the determination of the amount of each dose, the minimum interval between doses, and the total number of doses per session. Those factors depend on drug
class, animal species, and, in the case of rodents, strain. They can also vary among individual animals as a function of history of drug exposure, such environmental variables as ambient temperature (Finn et al., 1989), and the presence of a stimulus previously paired with drug exposure (Siegel et al., 1982).
Animal Care and Use Concerns Associated with Chronic Exposure to Addictive Agents
Studies of the effects of chronic exposure to addictive agents may involve prolonged or repeated exposure to high drug doses over a period of several days, weeks, months, or years. Such studies raise several issues that require consideration. One basic concern is whether extended periods of intoxication interfere substantially with normal feeding, drinking, and other activities (such as grooming) that are important for maintaining the health and well-being of animals. When that concern arises, consideration should be given to alternative methods of providing adequate nutrients and fluids, and of avoiding unsanitary cage conditions.
An additional concern in chronic studies is the possibility that long-term drug exposure will produce long-lasting tissue or functional changes that have adverse effects. In some cases, producing such changes is important to the scientific goals of the study, for example, a study designed to model neurologic deficits associated with Wernicke-Korsakoff syndrome.
In protocols involving prolonged or repeated drug exposure, criteria should be established for determining the duration of exposure and, if necessary, for terminating drug treatment earlier than planned. Daily observation of animals by someone familiar with the experimental protocol is important in such studies to ensure that decision-making is timely.
Physical Dependence and Withdrawal
In some studies of addictive agents, repeated or chronic drug exposure may produce physical dependence. Physical dependence is revealed by a characteristic withdrawal syndrome on termination of the drug regimen. The salient features and course of the withdrawal syndrome depend on the drug class, the animal species, and, in rodents, the strain (Metten and Crabbe, 1996; Way, 1993; Yutrzenka and Patrick, 1992). And the severity of withdrawal typically depends on the dosing regimen. Withdrawal signs may include irritability, activity changes, body-temperature changes, weight loss, tremor, and convulsive seizures. Drug withdrawal typically produces dysphoria and distress in humans (Jaffe, 1992), and investigators should consider the possibility that withdrawal may produce discomfort and distress in animals.
Whether or how withdrawal is treated in the laboratory will depend on the purpose of the experiment and the nature and extent of the withdrawal syndrome.
In some cases, induction of withdrawal is part of the experimental design, and treatment of the syndrome (for example, with a pharmacologic agent) would interfere with achieving the scientific goals of the study. Nevertheless, even when the schedule of exposure to an addictive agent is designed to allow the expression of a withdrawal syndrome, consideration should be given to establishing contingencies in the event of life-threatening signs, such as excessive weight loss or protracted seizure. Such contingencies might involve supplementary administration of food or fluids through a feeding tube or treatment with an appropriate anticonvulsant drug. When withdrawal is not the subject of the study and the withdrawal syndrome is expected to be severe, dose titration or other drugs may be used to alleviate withdrawal symptoms.
Occupational Health and Safety Considerations
Use of drugs that are restricted by the US Drug Enforcement Agency (DEA) requires supervision and inventorying by an institutional staff member who is licensed by DEA. The Public Health Security and Bioterrorism Preparedness Act of 2002 and the USA Patriot Act of 2001 require that research institutions collect information regarding hazardous substances classified as “select agents” and register their presence with the federal government.
Staff working with drugs and toxicants must be trained in the use of gloves, gowns, goggles, and eyewash and the appropriate disposal of “sharps.” Animals exposed to hazardous materials, including carcinogens and radioactive agents, must be handled and disposed of separately from other animals. Care must also be taken when cleaning the cages or enclosures of these animals to avoid contacting hazardous materials that may have been excreted in the urine or feces. Some hazardous materials may be administered to animals in drinking water from special spillproof canisters to avoid spilling of hazardous materials and exposure of staff to hazardous materials (Evans et al., 1986). If an animal is a possible source of contamination, behavior and physiologic measures can be monitored while the animal remains in its home cage without requiring staff to touch it. Home-cage observations may use a rating scale for cage-side observation, photobeam equipment for detecting locomotion, or telemetry devices implanted in the animal at the start of the experiment, which can be monitored with remote equipment.
A variety of physical agents influence neural function, including heat, light, and sound. This section surveys exposing animals to heat, light, and sound to study environmentally induced stress or neural dysfunction.
Some studies of temperature regulation allow animals to indicate a preferred ambient temperature by changing their location. For example, an animal may be
given access to a thermal gradient. Animals can also learn to control their ambient temperature with conditioned responses that increase or decrease the ambient temperature of their immediate environment (Carlisle and Stock, 1993; Gordon et al., 1998; Zhong et al., 1996). In other studies, animals may be exposed to inescapably cold or warm environmental conditions (Mechan et al., 2001); in these cases, consideration should be given to providing a period of adaptation to the new temperature, for example, by exposing the animals for increasing periods, or by gradually increasing or decreasing the ambient temperature. In the case of cold exposure, increased availability of food is important.
Changes in luminance and in daily light cycles are used to alter circadian rhythms (Cheng et al., 2002). Such studies may be performed to investigate health problems caused by disturbances in circadian rhythms resulting from jet lag or by shift work.
Auditory stimuli may be studied for their aversive or damaging properties. An important problem in contemporary society is the risk that work-related or environmental noise may damage auditory organs or interfere with auditory perception (Fechter, 1995). Studies of the neurobiology of sensory function or learning may use auditory-reflex methods. With rodents, a brief auditory stimulus is often used to induce a startle reflex (Le Pen and Moreau, 2002). The startle response provides a basis on which to evaluate variables that influence auditory learning and perception. Auditory startle-reflex techniques also are used to evaluate effects of drugs and toxicants that may alter sensory function or response to unexpected stimuli (Crofton, 1992). Other studies involve the use of noise exposure as a general stressor or to cause hearing loss (Fredelius, 1988; Hamernik and Qiu, 2001; Rao et al., 2001). The period and magnitude of noise exposure should be minimized but kept consistent with experimental goals. If noise is not used as an experimental manipulation to produce stress, researchers should recognize that noise may cause stress or induce seizures (Neumann and Collins, 1991). Consideration should also be given to avoiding inadvertent exposure of personnel and other people to excessive noise.
MODIFICATION OF DIETARY NUTRIENTS
A large body of research focuses on the effects of specific nutrients on neurologic function and dysfunction. For instance, folic acid deficiency in pregnant women leads to neural-tube defects in their children (Werler et al., 1993), and vitamin A deficiency can cause blindness (Anonymous, 1966). When neuroscience and behavioral research involves selective nutrient deficiency or toxicity, the research and veterinary staff must be prepared to deal effectively with the pain and/or distress that may result.
Numerous references identify the clinical signs and physiologic effects of specific nutrient deficiency or toxicity in rodents (NRC, 1995), nonhuman pri-
mates (NRC, 2003b), cats and dogs (Aiello, 1998a; NRC, 1985, 1986), and rabbits (NRC, 1977).
An animal use-protocol that involves purposeful deficiency or toxicity of a nutrient should include a comprehensive plan for monitoring the expected physiologic effects. The plan should outline the clinical signs or testing regimen for identifying animals in pain and/or distress, animals at risk of reduced feeding, and animals with physiologic impairments due to the nutritional modification. The animal-care staff should be made aware of the plan, because they may be the first to notice expected or unexpected adverse effects of nutritional modification. Steps should be established in the animal-use protocol and approved by the IACUC in advance to manage the animals adequately without compromising the goals of the experiment and to define clear endpoints for removal of animals from the study.
Many diets used in nutritional studies are ordered, stored, and dispensed outside the normal husbandry operation, so quality control must be ensured. The diets are often administered by the research staff. Record keeping that can be accessed by the husbandry staff and by the IACUC during its semiannual inspection of facilities and animal-study areas is necessary to ensure that animals are being fed in the manner described in the approved animal-use protocol.
The running wheel has been a fundamental tool in neurobehavioral research in rodents since the pioneering studies of Richter (1967, 1971). Voluntary wheel-running is studied to understand neurologic mechanisms controlling circadian rhythms, metabolism, and energy expenditure (Cotman and Berchtold, 2002). If an experiment requires that animals live in the running wheel or in a cage that has been specifically modified to include a running wheel, the cage should comply with the space recommendations of the Guide. Forced running, in which a rodent is placed briefly on a moving treadmill or on a rotating bar, is used to measure deleterious effects of drugs on coordination and stamina (see “Behavioral Screening Tests” in Chapter 9).
Rodents are introduced into a pool of water for tests of endurance (Cryan et al., 2002) and learning and memory (Reed et al., 2002). The Morris water maze is one of the most commonly used methods of studying learning and memory in rodents (Barnes et al., 1994; Harker and Whishaw, 2002). Its wide use is based on the rodent’s ability to swim without training, and the task requires a shorter training period than do such responses as lever-pressing. Exposure to water at
ambient temperature requires less adaptation than exposure to such motivating procedures as food or water restriction. Forced swimming is used to create a standardized stress experience (Griebel et al., 2002; Porsolt et al., 1977). Rodents commonly become immobile after several minutes of swimming if there is no possibility of escape from the water. The dependent variable often is the time until the first episode of immobility or the percentage of the test session spent immobile. Animal-welfare issues include maintenance of an appropriate water temperature, and provision of proper care of the wet rodent after it is removed from the water, and the establishment of unambiguous humane endpoints for testing in the animal-use protocol (see “Mood-Disorder Models” in Chapter 9).
Animal Care and Use Concerns
Neuroscience studies involving physical conditioning and exercise require appropriate attention to adaptation of the animal to the testing situation, its gradual conditioning to develop stamina, and close animal observation and record keeping during the exercise period. Swim tanks and automated treadmills and running wheels are the most common equipment used to force or promote exercise. Animals should initially be trained on automated devices at low speed, incline, and duration, which should increase gradually as the animals gain stamina. Similarly, the duration of swimming periods should be increased gradually as the animals’ condition improves. Weekly or even daily increases may be possible. However, animals should be closely observed by knowledgeable personnel during training and exercise sessions—particularly during the early phases of a conditioning program, near the end of individual training sessions, and during sessions in which performance requirements are increased—and detailed records of the animal’s performance and general health should be kept and made available to veterinary staff and the IACUC. In some systems, a rodent’s toes or tails may be at risk of becoming entrapped in the treadmill device. The continuous presence of an observer is essential to prevent injury in such situations. The use of remote monitoring systems, such as closed-circuit cameras, is sometimes warranted. As part of its review of the animal-use protocol, the IACUC may consider evaluation of equipment and a preliminary assessment of animal performance in a device.
Many automated treadmill systems apply a mild electric shock to animals whenever they fail to keep up with the programmed pace and drift back on the device. Although the number of shocks experienced by well-trained and conditioned animals is typically low, monitoring and recording shocks that animals experience and the pattern of shock administration during a training session can provide information about the adequacy of the training or exercise in light of the animals’ physical condition. A humane endpoint for removal of animals from the testing situation should be specified in the animal-use protocol and approved by the IACUC.
Studies that require an animal to exercise to exhaustion require special consideration. The need for such extreme effort by an animal should be carefully defined and justified, and endpoints should be clearly established and well defined in the animal-use protocol. Specific behaviors, circumstances, or physiological markers should be established to alert the observer that a trial must be terminated. Continuous animal observation is essential near the time of the expected development of animal exhaustion. In all cases, accurate records of test conditions and of performance should be maintained for each animal, and they should be available to veterinary staff and the IACUC. Such records will allow day-to-day adjustment of testing, if warranted by an animal’s condition or ability.
A final consideration is the need to maintain sanitation of devices used for exercise or learning. Devices should be constructed of an impervious material to the greatest extent possible. A regular sanitation procedure and schedule should be established, maintained, and clearly documented. A thorough description of the sanitation process should be included in the animal-use protocol.
Short-term sleep loss in humans typically has no adverse physiologic consequences other than increasing sleepiness and impaired performance in some tasks (Horne, 1985; Naitoh et al., 1990). Because sleep is a homeostatic process, adverse effects associated with short-term sleep loss are probably alleviated simply by providing the opportunity to “catch up” on sleep (Everson, 1997; Everson et al., 1989), much as thirst is immediately relieved by taking a drink of water. In rats, biologically significant adverse effects of sleep deprivation have been reported only after sleep deprivation of more than 5 days (Everson and Toth, 2000).
Several approaches are used to produce sleep deprivation in laboratory animals. The method most widely used is probably the so-called “gentle-handling” technique. This method has been applied to rodents, rabbits, and cats and is usually used to cause loss of both rapid-eye-movement sleep (REMS) and non-rapid-eye-movement sleep (NREMS). The animals are under continuous observation by the experimenter and are physically roused by the experimenter whenever they either enter a state of electroencephalographically defined sleep or assume a sleeplike posture. Animals are generally aroused by such actions as tapping on the cage, providing novel objects, or prodding gently. As the duration of the deprivation period increases, particularly beyond a few hours during the species’ normal “rest” period, the experimenter must gradually increase the intensity or frequency of handling or of environmental stimulation to maintain arousal (of both the animal and the experimenter!). Because of its labor-intensive nature, the gentle-handling method of achieving sleep loss is rarely extended beyond a 24-hour period. Such short-term sleep loss does not appear to have marked adverse effects in humans or animals other than the progressive development of moderate to severe sleepiness, cognitive and performance impairment,
and perhaps irritability or aggression (Everson, 1997; Horne, 1985; Naitoh et al., 1990).
Another relatively common approach to inducing sleep loss in animals is the so-called “flowerpot” technique. This approach produces REMS deprivation by taking advantage of the muscle atony that develops during REMS (Cohen and Dement, 1965; Jouvet et al., 1964). The animals (typically rats) are placed on a small platform (historically an inverted flowerpot) in a tank of water. The platform is large enough to allow the rat to engage in slow-wave sleep, in which residual muscle tone allows it to retain a stable sleeping posture. However, as the animal enters REMS and develops skeletal-muscle atony, it slips from the platform into the water and awakens.
A third approach to causing sleep loss in animals is called the “disk-over-water” technique; it can be used to deprive animals of REMS alone or of both NREMS and REMS (Bergmann et al., 1989; Rechtschaffen et al., 1983). The animals are housed on a rotating platform, or disk, that is positioned over a pan of water. When the electroencephalogram indicates that an animal is entering a state of sleep, a computer algorithm causes the disk to rotate at a low speed. The animal then generally awakens and walks to avoid contacting the water.
A fourth approach is forced locomotion, usually in a slowly revolving drum (Frank et al., 1998; Mistlberger et al., 1987; Rechtschaffen et al., 1999). Animal-welfare considerations relevant to this method are similar to those mentioned previously for exercise models. Interpretation of data collected with this method is confounded by the effect of continuous locomotion or exercise as opposed to the effects of sleep loss itself (Rechtschaffen et al., 1999).
In contrast with the gentle-handling method, the flowerpot and disk-over-water techniques can be easily imposed for long periods, and these approaches create some animal-use concerns. The flowerpot method of REMS deprivation causes alterations in several biochemical indexes of stress (Suchecki et al., 2002). In a refinement of the flowerpot and disk-over-water methods, multiple platforms are used in one large pool so that animals can engage in locomotor activity (Suchecki et al., 2002). Several animals can be housed together in these conditions. Social interactions may reduce some of the nonspecific stress associated with the environment and the physiologic challenge (Suchecki et al., 2002).
Sleep deprivation of over 7 days with the disk-over-water system results in the development of ulcerative skin lesions, hyperphagia, loss of body mass, hypothermia, and eventually septicemia and death in rats (Everson, 1995; Rechtschaffen et al., 1983). The duration of sleep deprivation must be well justified scientifically, particularly if it will be continuous for more than a few days. However, relatively few studies have imposed sleep deprivation long enough to cause those signs. In general, animals that are maintained on chronic sleep-deprivation schedules should be closely monitored for injury and general well-being, and observations should be recorded. The task is simplified by the fact that research teams typically monitor such animals very closely to ensure that they are
experiencing the targeted amount of sleep loss. The use of interventions, particularly in chronic studies, must be compatible with the scientific goals of the experiment.
The use of automated sleep-deprivation devices, like the use of exercise devices, requires regular sanitation, good animal observation, and accurate record keeping.