ILAR Journal Online, Volume 41(3) 2000: Mouse Behavioral Models in Biomedical Research

Online Issues

<< All Back-issues

<< This Issue's Table of Contents

ILAR Journal V41(3) 2000
Mouse Behavioral Models in Biomedical Research

Behavioral Phenotyping of Transgenic and Knockout Mice: Experimental Design and Evaluation of General Health, Sensory Functions, Motor Abilities, and Specific Behavioral Tests
Jacqueline N. Crawley

Jacqueline N. Crawley, Ph.D., is Chief of the Section on Behavioral Neuropharmacology, Experimental Therapeutics Branch, National Institute of Mental Health, Bethesda, Maryland.

Introduction

Targeted gene mutation technology represents a powerful new tool for biomedical research. When the targeted gene is expressed in the brain, the behavioral phenotype of the mutant mice may reveal genetic mechanisms underlying normal behaviors and may increase our knowledge of genetic factors in neuropsychiatric disorders. Transgenic mice have a new gene, or an additional copy of an existing gene, added to the genome. Knockout mice have a targeted gene deletion, such that no product of the mutated gene is synthesized in the null mutants. The methods for developing the mutation, and breeding strategies to generate null mutants in the F2 and subsequent generations, are extensively described in [Brain Research Vol 835] and elsewhere (Campbell and Gold 1996; Nelson and Young 1998; Wehner and Silva 1996).

Approximately 100 different genes expressed in the central nervous system have been targeted and phenotyped in transgenic and knockout mice to date (Campbell and Gold 1996; Bedell et al. 1997; Silva et al. 1997; Murphy et al. 1998; Nelson and Young 1998). Reported behavioral phenotypes include aberrant social, reproductive, and parental behaviors, aggression, feeding disorders, learning and memory impairments, anxiety-like behaviors, and altered responses to antidepressants, antipsychotics, ethanol, and psychostimulant drugs of abuse. Experimental design is presently being optimized for thorough evaluation of behavioral phenotyping in mutant mice. This review is designed to suggest general methods that have been validated in our laboratory and others. Specific protocols for individual behavioral tasks can be found in the original publications referenced throughout the text, and in several recent reviews (Wehner et al. 1996; Lee et al. 1996; Campbell and Gold 1996; Crawley et al. 1997, 1998; Crawley and Paylor 1997; Jucker and Ingrain 1997; Nelson and Young 1998) [see also Crawley 20001].

The Mice

After the mutation has been successfully introduced, and the gene product shown to be correctly overexpressed or absent, the first chimeras are mated with wildtype mice. A Mendelian distribution of F2 offspring is predicted. The strain of mouse used for the embryonic stem cells, for the donor blastulas, and for the breeding can greatly affect the behavioral phenotype. For example, some 129 substrains (e.g., 129/J and 129/SvJ) have an incomplete or missing corpus callosum (Livy and Wahlsten 1991), and perform poorly on learning and memory tasks (Wehner and Silva 1996; Crawley et al. 1997). Background genes that are randomly contributed by each parental strain often interact with the mutated gene of interest, creating false positives (Wehner and Silva 1996; Gerlai 1996; Crawley 1996; Banbury Conference 1997; Crawley et al. 1997). For example, human Alzheimer's mutant [3-amyloid precursor protein overexpressed in C57B6 breeder mice produces amyloid plaques in the brain and poor learning and memory on the Morris water task (Hsiao et al. 1996). The same transgenic insertion in FVB/N mice did not show plaque formation and was lethal at too young an age for learning and memory testing. No single strain can solve all of the potential breeding problems. For some experiments, breeding into the 129/Sv strain is a good solution, to unify the background genes by using the same strain for the embryonic stem cells and the parental breeders (Lijam et al. 1997). Other experiments benefit from a congenic breeding strategy into C57BL/6J, a standard, commercially available inbred strain that shows intermediate values on most behavioral phenotypes, has been characterized on a large number of behavioral tasks, and is a reasonably prolific breeder (Silver 1995; Crawley et al. 1997; Banbury Conference 1997).

Numbers of animals for standardized experimental designs and appropriate statistical tests are a minimum of N = 10 -/-null mutants, N = 10 +/- heterozygotes, and N = 10 +/+ wildtype littermate controls. If a gender effect is detected, N -- 10 of each gender and each genotype is required. Ages of the animals are approximately equivalent across genotypes, in accordance with the goals of the experiment. For example, adult mice are best tested between ages 3 and 8 months; aged mice between 12 and 18 months; juvenile mice between 2 and 6 weeks.

Preliminary Observations

A series of carefully conducted preliminary observations of general health, home cage behaviors, sensory abilities, and motor functions is first conducted for each mouse to avoid spurious false positives. If an animal has a major health problem or a gross motor defect, it will be unable to perform many behavioral tasks, for reasons not necessarily specific to the mutation. If an animal is blind or deaf, specific behavioral tests can be designed around the sensory deficit, such as olfactory learning tasks for blind mice or tactile startle tasks for deaf mice.

Our laboratory developed a set of preliminary observations and neurological reflexes which we use to assess gross defects in mutant mice (Crawley and Paylor 1997). This first step is likened to "giving your mouse a physical exam." The mouse is weighed, its body temperature is taken, and the appearance of its fur and whiskers is noted. Home cage locomotion, grooming, nesting, sleeping, and fighting patterns are recorded. Neurological reflexes are tested in each mouse. These include eye blink, ear twitch, whisker twitch, and righting reflex.

We next evaluate motor functions. Each mouse is tested for normal exploratory locomotion on the Digiscan open field (Norflus et al. 1998). The mouse is placed in the photocell-equipped automated open field box for a 5-min test session, during which the Accuscan software calculates total distance traversed, number of movements, horizontal activity, vertical activity, and center/perimeter time. Each mouse is next tested for motor coordination on the Basile automated accelerating rotarod (Jones and Roberts 1968; Sango et al. 1995). The mouse is placed on the cylinder and the speed of the cylinder rotation is gradually accelerated from 4 to 40 revolutions per minute over a 5-min period. Latency to fall from the rotarod is recorded. The fall is approximately 6 inches, a height that mice can easily fall and land on their feet without injury. Neuromuscular strength is tested by the wire hang test (Paylor et al. 1998). The mouse is placed on a wire cage lid and the lid is gently waved in the air so the mouse grips the wire. The lid is then turned upside down, approximately 6 inches above the surface of soft bedding material. Latency to fall onto the bedding is recorded, with a 60-sec cut-off time. Gait is measured with the footprint test (Barlow et al. 1996). The two hindpaws are dipped into black ink. The mouse is then immediately placed onto white paper in a dark, narrow tunnel, approximately 12 inches long. The ink is then wiped off the feet, and the pattern and pathway of the black footprints on the white paper are calculated.

Sensory functions are then assessed. Visual ability is measured in the visual cliff test conducted in a box with a ledge. The inner surface of the box and ledge are covered with black and white checkerboard contact paper, which emphasizes the ledge drop-off. A piece of clear plexiglas spans the ledge so that there is no actual drop-off, just the visual appearance of a cliff. Normal mice will stop at the "edge" and explore the Plexiglas floor before walking forward. Blind mice will not see the appearance of the edge and will walk forward across the plexiglas immediately. Hearing is assessed with the acoustic startle test (Paylor and Crawley 1997). The reflexive flinch and eyeblink to a sudden loud noise comprise the standard acoustic startle response. An automated startle system is used to deliver the startle stimulus and measure the response of the mouse. The mouse is placed in a small cylindrical restraining tube within a sound-attenuating chamber. Stimulus tones are varied randomly from 70 dB to 120 dB, 40 msec duration, in the presence of background white noise at 70 dB. A standard battery of tones begins at threshold sound levels of 72 and 75 dB, includes moderately loud tones of 80, 90, and 100 dB, and ends at the loudest sound levels, 110, 115, and 120. Whole body flinch amplitude is automatically recorded by the pressure transducer. Tactile startle is measured by whole body flinch amplitude to a brief puff of air delivered by the system to an area near the face. A typical session lasts for about 30 min, after which the mouse is returned to the home cage. Sense of touch can be measured with Von Frey hairs, made of thin wire, which are used to gently touch the footpads, to measure the reflexive twitch to touch. A simple olfactory test can be used, e.g., latency to locate an odiferous piece of food, e.g., a piece of cheese or smear of peanut butter, buried under the litter in a clean test cage, or time spent sniffing a novel odor such as vanilla extract painted onto the test cage wall.

Specific Behavioral Tasks To Test Discrete Hypotheses

Targeted gene mutation is designed to address specific hypotheses about the behavioral role of a gene. The set of specific behavioral tasks to be used for behavioral phenotyping is designed around the hypothesis. Genes for neuropeptides in the hypothalamus might be analyzed in tests of feeding behaviors, sexual behaviors, and stress-related behaviors. Signalling genes expressed in the hippocampus might be analyzed in tests of spatial learning and memory. Genes for serotonin receptor subtypes might be analyzed in tests of aggression, feeding, and mouse behaviors relevant to symptoms of human depression. A thorough knowledge of the behavioral literature is required to choose the optimal constellation of behavioral tests to address the critical hypotheses.

In some cases, tests designed for rats can be directly used for mice. For example, the Digiscan automated open field works equally well for rats and mice, and the resident/intruder procedure for aggressive behaviors is applicable to males of both species. In some cases, the equipment can be modified for mice often simply by building a smaller version of the rat test apparatus. For example, a mouse rotarod and a mouse elevated plus maze are downsized versions of the rat rotarod and rat elevated plus maze. In some cases, the equipment must be changed to accommodate the behavior of the mouse. For example, lighter weight levers allow a mouse to conduct a lever press task in an operant chamber; alternatively, a nose-poke insert replaces the lever in an operant chamber to accelerate the autoshaping process, because mice tend to explore dark holes more readily than they press levers. In some cases, mouse and rat behaviors are too dissimilar. For example, social interaction is easily measured between two rats placed together in a new environment, but mice will primarily explore the new environment rather than interacting with each other. Rats can perform complex operant tasks which appear to be far beyond the abilities of mice.

This article provides an overview of several behavioral tasks that our laboratory and others have validated for mice. The tasks described in this review represent only a small subset of the applicable behavioral tests available in the wider behavioral neuroscience literature. The reader is referred to the many excellent chapters in [Brain Research Vol 835] which describe multiple tasks in specific behavioral domains in much greater detail.

Because each task has its own limitations, it is best to employ multiple tasks for each behavioral domain of interest to avoid false negatives. For example, a mutation in a gene expressed mainly in the amygdala may produce performance deficits on fear-conditioned learning tasks, which require an intact amygdala, but not on spatial navigation learning tasks, which require an intact hippocampus. When sensory or motor deficits are detected during the preliminary screen, the choice of the multiple tasks is limited by the physical disability of the mutants. Knowledgeable behavioral neuroscientists can generally identify two or three good tasks for each type of behavior relevant to each hypothesized function of the gene product.

Feeding

Twenty-four hr consumption of normal rat chow in the home cage is measured by daily weighing of the food (Erickson et al. 1996; Huszar et al. 1997). Standard amounts of food are provided by the investigator. Another approach is a challenge test, measuring consumption of a palatable food source or specific macronutrients during a test session (Crawley and Corwin 1994). A high carbohydrate/high fat source, such as sweetened condensed milk, cookie mash, or a commercially available powdered chow diet, is provided in a clean test cage. The mouse is placed in the cage for the fixed time period. Alternatively, a two-choice paradigm involves presentation of the special diet in a second food hopper next to the standard chow in an identical food hopper, or presentation of test liquid solutions in separate water bottles, attached to the home cage (Belknap et al. 1993; Berrettini et al. 1994). Food or water restriction may be performed in special circumstances, e.g., to study a gene that regulates consumption only after an overnight fast, or as a necessary component of food or water reinforcement in learning and memory tasks.

Learning and Memory Tasks

Morris swim tasks (Morris 1981) have been standardized to measure spatial navigation learning and memory in mice (Wehner and Silva 1996; Crawley et al. 1997; Sterneck et al. 1998). Each mouse learns to swim in a circular pool of water to locate a submerged hidden platform. The water level is 30 cm deep, maintained at 24-26°C with aquarium heaters, and changed daily. Non-toxic Crayola white paint is added to make the water opaque and the platform invisible to the mouse. An automated video tracking software system is used to quantitate swim speed, swim pathway, latency to reach the platform, and time spent in each of the four quadrants of the pool. Mice are excellent swimmers and readily dry themselves off and groom when removed from the pool. While additional drying or heat is usually not necessary, a soft cloth is available to dry the mice and a hot plate is available to place under the cage, if needed. During training on the visible platform task, the platform is raised above the surface of the water, or a prominent cue is attached to the platform to extend above the surface of the water. The mouse is gently guided to the platform by the experimenter and allowed to remain on the platform for 20 sec. The mouse is then placed along the edge of the pool, facing the wall, in one of four randomized start locations equidistant around the perimeter. The mouse is given 60 sec to swim to the visible platform and allowed to remain on the platform for 20 sec. Latency to find the visible platform over four successive trials is recorded. Mice that do not swim to the visible platform are noted, and given further guided training by the experimenter. Any mouse that does not swim, and any mouse that only floats, is eliminated from the experiment. Mice that learn to navigate to the visible platform are then trained on the hidden platform task. The platform is submerged below the surface of the water, so that the mouse cannot see the location of the platform. Each trial begins with the mouse being placed in the pool, facing the wall, in a different quadrant on each trial. Most acquisition training on the hidden platform task [the series of training trials over which the animals learn the location of the hidden platform, based on distal environmental cues] consists of two or three blocks a day of four trials per block = 8 to 12 trials per day, for three to five consecutive days. Fewer trials per day may be used for mice that learn rapidly. More days may be needed for mice that learn slowly. After the acquisition curve reaches the performance criteria (usually 10- to 20-sec latency to find the hidden platform), the probe trial is administered. [The probe trial, which is the final test, proves that the task was learned using distal environmental cues.] The platform is removed from the pool. The mouse is placed in the pool for a single 60-sec trial. Time spent in the trained quadrant, where the platform was previously located, is compared to time spent in the other three quadrants, to confirm acquisition. If time spent in the trained quadrant is significantly greater than time spent in each of the other quadrants, the mouse has learned the location of the hidden platform, rather then having learned an effective general search strategy. The probe trial is generally administered at the end of the last training day. Additional probe trials may be administered on subsequent days, as a measure of retention. Additional training may precede subsequent probe trials. Another approach is to test reversal learning, in which the mouse is retrained with the hidden platform in another quadrant. Probe trial testing is similarly conducted after reversal training.

Cued and contextual conditioning is a fear-conditioning task that measures memory of an aversive experience and the stimuli present during the aversive experience (Fanselow 1990; LeDoux 1994; Crawley et al. 1997; Wehner et al. 1997; Caldarone et al. 1997; Sterneck et al. 1998). A standard footshock shuttle box, or a specialized automated freeze monitor and software package, is used to control footshock delivery and to measure duration of freezing behavior. Freezing, a standard response to a sudden aversive stimulus, is defined as complete behavioral immobility except for respiration. On day 1, the mouse is placed in a chamber and allowed to explore for 2 min. The chamber is square and illuminated by normal house lighting. An auditory stimulus (80 dB white noise) is then presented for 30 sec. One footshock, 0.5 mA, 2-sec duration, is then delivered. The animal is then left in the cage for 30 sec, then returned to its home cage. On day 2, the mouse is returned to the same testing chamber. The time spent in freezing behavior is measured over a 5-min period. Total number of seconds spent in complete behavioral immobility are tallied. One hour later, the mouse is placed into the same testing chamber, but the context has been changed. The grid floor is now covered with Plexiglas, the square is divided by a piece of Plexiglas to form a triangle-shaped chamber, a red chamber light is on, and an odor is painted onto the chamber walls, e.g., vanilla or lemon extract. Freezing is quantitated for 3 min. Then the auditory cue is presented. Freezing is quantitated over the next 3 min, in the presence of the 80 dB white noise cue. The mouse is then returned to its home cage. Thus, memory is assessed by measuring freezing under three conditions for each mouse: context, altered context, and auditory cue. Chambers are thoroughly cleaned after each test session.

Maze learning and avoidance tasks are among the oldest learning and memory tasks used in rodents. Passive and active avoidance tasks measure memory of an aversive experience, through simple avoidance of a location in which the aversive experience occurred (Oliverio et al. 1973; Brioni and McGaugh 1988; Mathis et al. 1994; Sango et al. 1995). A commercially available automated apparatus consists of two connected chambers, one lighted, one dark. Mice tend to prefer a dark environment and will immediately enter the darkened chamber. Passive avoidance is a two day task. On day 1, the training session, the mouse is placed in the lighted chamber for ten seconds. The door to the dark chamber is then opened, and latency to enter the dark chamber is measured as a control for visual ability and motor activity. Immediately after the mouse enters the dark chamber, a 0.3 mA, 1-sec footshock is delivered. The mouse remains in the dark compartment for 10 sec after the shock to allow formation of the association between the dark compartment and the footshock. The mouse is then returned to the home cage. On day 2, the retention test session, the mouse is placed in the lighted compartment and the door is opened. Latency to enter the dark compartment is measured, with a 300-sec cut-off time. The mouse is then returned to its home cage. An additional exposure to the single footshock on day 2, followed by latency to enter the dark on day 3, can be used to obtain additional information on acquisition of passive avoidance in mice with apparent poor acquisition. For active avoidance, using the same chamber, with the same parameters, the mouse must move into the opposite chamber to avoid receiving a footshock. Latency to enter the non-shocked chamber is the measure of learning.

Spatial maze tasks include the T-maze, Y-maze, radial mazes, and Barnes maze (Crusio et al. 1987; Paylor et al. 1993; Heinrichs et al. 1996). Acquisition of location of a reinforcer over repeated trials provides the measure of learning and memory. The Barnes maze is a circular platform, 1.3 meters in diameter, in which 36 holes, each 8 cm in diameter, are equally spaced around the perimeter of the circle, 10 cm from the edge. Mice tend to poke their noses into holes, especially to explore dark holes, which may lead to escape routes away from a lighted, open field. One of the holes leads to a dark, enclosed box, located just below the circular platform. The mouse is placed in the center of the platform under a start box. The start box is lifted, and the mouse is guided to the escape tunnel, where it remains for I min. Then the mouse is placed back in the start box, and allowed to explore the maze. When the mouse finds the correct hole, it escapes into the dark box. The mouse is allowed to remain in the escape box for 1 min, then removed from the box and given repeated trials in which it is placed in the center of the platform and allowed to explore for 5 min per trial, 1-4 trials per day, for up to 20 training days. Time to locate the tunnel hole, distance traveled, and number of errors on each training day (noses pokes into the incorrect hole) provide the measures of acquisition.

The T-maze is shaped like the letter T, with either the left or right arm containing concealed cups. One Noyes food pellet or 0.05 ml tap water is used for food or water reinforcement, respectively, in food or water restricted mice, respectively. In the continuous reinforcement procedure, the mouse is placed in the start end of the T, allowed to explore the maze and obtain the reinforcer. The mouse must then return to the start box and choose the opposite arm on the next trial to obtain the reinforcer. Twelve trials per day are conducted for up to 10 consecutive days. Acquisition of the procedure and choice accuracy in this alternation task are the measures of procedural and learning abilities.

Operant tasks are more difficult for mice than for rats, but mice can learn some lever press acquisition tasks (McDonald et al. 1998; Ripple and Stephens 1999). Food or water restricted mice are habituated to the operant chamber and shaped to press the lever for the food (small food pellet) or water (0.05 ml tap water) reinforcer, respectively. Number of trials to acquire the lever press task is the simplest measure of learning in mice. The more complex non-matching to position task requires the mouse to receive reinforcement for pressing the opposite lever to the lever illuminated on the previous trial by the cue lamp above the lever. A discriminative go-no go task for mice requires the mouse to press the lever only when the cue light is illuminated to obtain the reinforcement, or to press the lever only during the period in which a tone is sounding. Modifications of the rat operant chamber for mice include levers that are smaller and more sensitive to the touch of the mouse paw, and a module which substitutes the lever for a nose-poke into a dark hole, to obtain the reinforcement.

Motor learning is evaluated by repeated daily testing on the rotarod (Paylor et al. 1998). Reduction in latency to fall, over repeated testing sessions, provides a measure of cerebellar learning.

Pain Sensitivity Tests

The two methods to assess analgesia that have been well validated and standardized for mice are the tail flick and hot plate tasks (Dewey et al. 1969; Pick et al. 1991; Rubenstein et al. 1996). Tail flick nociception measures a simple spinal reflex to a sudden, painful thermal stimulus. A photobeam is used to apply a heat stimulus to the tail. Latency to flick the tail out of the path of the light beam is measured. The photobeam is turned off if the tail is not flicked away within 15 sec, to avoid tissue damage.

Hot plate nociception is a similar reflex, which requires higher brain centers. The mouse is placed on the surface of a hotplate, which is maintained at 50-55°C. A plastic frame encloses the surface, so that the mouse cannot jump out. Latency for the mouse to raise and lick its forepaw, or to jump up, is recorded. The mouse is then immediately removed from the hotplate. If the mouse has not responded within 30 sec, it is removed from the hotplate, to prevent tissue damage.

Pain sensitivity to footshock is assessed by threshold analysis, as a control for procedures requiring a footshock, e.g., some of the learning and memory tasks described above. A sequence of single footshocks is delivered in the avoidance chamber, and the mouse is observed for flinching, jumping, running, and vocalization. When these behaviors are observed, the sequence is terminated. The sequence begins with 0.075 mA and proceeds through 0.1 mA, 0.15 mA, 0.25 mA, and 0.35 mA, and 0.45 mA single footshocks of 1 second duration each. Most mice show behavioral responses to the lower footshock levels, and therefore do not receive the higher footshocks.

Anxiety-related Behaviors

Light <---> dark exploration is an ethologically-based approach-avoidance conflict test which is sensitive to anxiolytic drug treatments (Crawley 1981; Crawley and Davis 1982; Mathis et al. 1995; Crawley et al. 1997). Mice prefer a dark, enclosed, small space over a brightly lit, open, large space. However, mice are also highly exploratory. The light e--> dark exploration task represents a naturalistic conflict between the tendency of mice to explore a novel environment versus the tendency of mice to avoid a brightly lit open field. The mouse is placed into the large, lighted compartment of a two-compartment chamber. The mouse repeatedly enters the adjoining small, dark, enclosed chamber, then emerges back into the larger, lit, open chamber. Number of transitions between the two chambers and time spent in each chamber are automatically recorded through a photocell array across the border between the two chambers, over a single 10-min test session.

The elevated plus maze task similarly measures the conflict between exploration of a novel environment and avoidance of brightly lit open areas (Lister 1987; Trullas and Skolnick 1993; Stenzel-Poore et al. 1994; Flint et al 1995). An additional factor is that the surface is raised 1 m from the floor. The elevated plus maze is in the shape of a +. Two alternate arms are dark and enclosed, while two alternate arms are open, lit, and without edges. The mouse is placed in the center of the + and allowed to explore the maze for 5 min. Numbers of entries into the open arms versus number of total arm entries, and time spent on the open arms versus the closed arms, provide the measures of anxiety-related behavior.

The Digiscan open field measure of center time versus perimeter time gives a less specific measure of anxiety-related behaviors. In a brightly lit open area, mice will tend to stay near the walls of the open field, rather than enter the center region. Habituation to the novelty of the open field, i.e., reduction of perimeter time and increase in the center time, is measured by recording activity in 5-min consecutive intervals over a 1-hr period. Repeated 1-hr tests over several days gives a further measure of habituation to the aversive properties of the novel open field. While the open field test is not as specific for anxiety-related behaviors, the center time parameter provides some indications that can be used as a starting point for further testing on the more specific anxiety-related tests.

Depression-related Behavior

The Porsolt swim test is used to evaluate "behavioral despair," a measure of failure to seek escape from an aversive stimulus (Porsolt et al. 1977). The mouse is placed in a cylinder of room temperature tap water and swimming behavior is observed over a 1-min test session. The water is approximately 12 in deep, such that the animal cannot balance on its feet or tail. The water surface is approximately 6 in from the top of the cylinder, such that the animal cannot jump out. Rodents will generally swim. Animals treated with certain drugs or lesions will stop swimming and will float. Floating time is considered the measure of depression-like behaviors, in that the animal has stopped swimming and "given up" on finding an escape route. Floating time is decreased by treatment with antidepressant drags. Any mouse that does not swim or float is immediately removed from the water. At the end of the 10-min swim test, the mouse is dried with a towel and returned to its home cage.

Schizophrenia-related Behavior

Deficits in prepulse inhibition are common in schizophrenic patients and may measure attentional dysfunctions that contribute to auditory hallucinations (Geyer et al. 1990). Prepulse inhibition is a sensorimotor gating reflex, similarly quantitated in mice, rats, and humans (Geyer et al. 1990; Paylor and Crawley 1997). When the startle stimulus is immediately preceded by a milder stimulus, delivered immediately before the startle stimulus, the mouse will flinch less to the startle stimulus. Prepulse tones of 74, 82, or 90 dB are randomly presented 100 msec before each 100 dB or 120 dB startle tone. Whole body flinch amplitude is again automatically recorded. To evaluate a separate sensory modality, a puff of air is used as the prepulse, before the acoustic or tactile startle stimulus.

Social Behaviors

A variety of social behaviors can be quantitated by videotaping the mice over 24-hr periods in the home cage (Lijam et al. 1997). Observations during the night are conducted under red light, which does not disturb the mice. Group huddling while sleeping is one characteristic of normal mouse social behavior. Nest building can be observed over a 1-hr period after placing a Nestlet cotton square or other nesting material in the floor of the home cage. Depth of nest provides a quantitative measure of nest building. Fighting can be quantitated from the videotapes by scoring number and duration of attacks.

The tube test for social dominance can be used to evaluate aggressive behaviors (Lindzey et al. 1961; Lijam et al. 1997). Two mice are placed in opposite ends of a Plexiglas tube, 3 cm in diameter, 35 cm long. The mouse that advances beyond the midpoint is considered the dominant male. The mouse that backs away from the midpoint is considered the subordinate male. Mice are removed immediately after scoring. Fighting does not occur.

Isolation-induced standard opponent and resident-intruder testing for aggression in male mice is conducted according to established methods (Miczek et al. 1994; Nelson et al. 1995; Crawley et al. 1997). A standard subordinate test opponent male is placed into the cage of the test male. The test male has been singly housed for approximately 1 month. Attacks over a 5-min test session by the test male are scored for latency to first attack, number of attacks, number of bites, chasing, and tail-rattling.

Sexual and Reproductive Behavior

Mouse sexual and parental behaviors can be scored using standardized methods (Cohen-Salmon et al. 1985; Nishimori et al. 1996; Rissman et al. 1997; Wersinger et al. 1997). Male sexual behavior is quantitated by observational methods for latencies and frequencies of mounts, intromissions, and ejaculations. Female sexual behavior is quantitated by standardized observations methods for the lordosis response. Pregnancies and number of pups delivered per pregnancy are measures of reproductive success. Lactation can be scored by an observer. Presence of milk in the stomach of the pup can be ascertained in cases of early postnatal death. Parental behaviors are measured by quantitating retrieval of pups to the nest and sitting with the pups in the nest. Before weaning, isolated mouse pups emit ultrasonic vocalizations, which alert the parent that the pup is out of the nest and triggers retrieval behavior. Olfactory cues are used by the parents to identify their own pups. Quantitating parental behaviors provides a good example of the need to test sensory abilities before conducting a specific test. If sensory abilities are not tested, the parent may be deaf or anosmic due to the mutation, and deficits in parental behaviors can be overinterpreted (Brown et al. 1996).

Addictive Drugs

Good paradigms are available to assess responsivity to drugs of abuse in mice. Self-administration of drugs of abuse, including alcohol, cocaine, morphine, and nicotine, are measured in mice (Belknap et al. 1993; Berrettini et al. 1994; Grahame and Cunningham 1995; Alexander et al. 1996; Matthes et al. 1996; Grisel et al. 1997; Miner 1997; Maldonado et al. 1997; Picciotto et al. 1998; Crabbe et al. 1998). Drugs are self-administered intravenously or in the drinking water. Conditioned place preference quantitates the rewarding internal cues created by a drug treatment. Measures of tolerance, dependence, and withdrawal symptoms are quantitated by an observer, using standardized scoring methods.

Conclusions

In the opinion of this author, the present technology for the generation of mutant mice is most useful for modeling the symptoms of single gene mutation human hereditary diseases. Behavioral phenotypes identified for transgenic and knockout mice may lead to important medical applications. When the mutation is designed to mimic the genetic mutation in a human hereditary disease, the mouse behavioral phenotype can serve to evaluate the efficacy of new pharmacological and gene therapy treatments (Campbell and Gold 1996; Blomer et al. 1996; Bedell et al. 1997; Norflus et al. 1998). Behavioral phenotypes revealed in the mutant mouse model can thus serve as quantitative surrogate markers to test the efficacy of potential therapeutics.

Two major problems of the current transgenic and knockout technology limit its usefulness to the more general study of the role of genes in normal and abnormal behaviors. The first is the presence of the mutation in all cells. A gene expressed in the brain may also be expressed in many peripheral organs. It is impossible to assign a behavioral difference in the mutant mouse to a specific brain structure or pathway, or even to the nervous system. Tissue-specific "conditional" knockouts are being generated to solve this problem (Mayford et al. 1996; Tsien et al. 1996; Kistner et al. 1996; Kammescheidt et al. 1997; and see other reviews in [Brain Research Vol 835]). The second is the presence of the mutation from the earliest stages of development. Other genes may compensate during development for the absence of the deleted gene, or for the overexpression of the transgene. Lack of behavioral phenotype may be due to compensation by a redundant gene which takes over the function of the missing gene. An opposite behavioral phenotype to that predicted may be due to overcompensation by one or more redundant genes. An observed behavioral phenotype may in fact characterize the role of the compensatory gene, not the targeted mutation. Temporally-selective "inducible" transgenics are being generated to solve this problem (Mayford et al. 1996; Kistner et al. 1996; and see other reviews in [Brain Research Vol 8351).

If conditional and inducible mutations become feasible, it is likely that the transgenic and knockout technology will become the premiere research tool to further our understanding of the genetic basis of behavior. Gene mutations may replace pharmacological tools, such as drugs that act as selective antagonists at neurotransmitter receptor subtypes. Gene mutations may also replace lesions of brain structures or pathways, to study the functions of endogenous neurochemicals and neuroanatomical pathways. Behavioral phenotyping methods continue to evolve, to optimize our approaches to the growing opportunities of the mutant mouse technology.

References

Alexander RC, Wright R, Freed W. 1996. Quantitative trait loci contributing to phencyclidine-induced and amphetamine-induced locomotor behavior in inbred mice. Neuropsychopharmacology 15:484-490.

Banbury conference on genetic background in mice. 1997. Neuron 19:755-759.

Barlow C, Hirotsune S, Paylor R, Liyanage M, Eckhaus M, Collins F, Shiloh Y, Crawley JN, Ried T, Tagle D, Wynshaw-Boris T. 1996. Arm-deficient mice: A paradigm of Ataxia-Telangiectasia. Cell 86:159-171.

Bedell MA, Jenkins NA, Copeland NG. 1997. Mouse models of human disease. Part I: Techniques and resources for genetic analysis in mice. Genes Devel 11:1-43.

Belknap JK, Crabbe JC, Young ER. 1993. Voluntary consumption of ethanol in 15 inbred mouse strains. Psychopharmacology 112:503-510.

Berrettini WH, Ferraro TN, Alexander RC, Buchberg AM, Vogel WH. 1994. Quantitative trait loci mapping of three loci controlling morphine preference using inbred mouse strains. Nature Genet 7:54-58.

Blomer U, Naldini L, Verma IM, Trono D, Gage FH. 1996. Applications of gene therapy to the CNS. Hum Molec Genet 5:1397-1404.

Brioni JD, McGaugh JL. 1988. Post-training administration of GABAergic antagonists enhances retention of aversively motivated tasks. Psycho-pharmacology 96:505-510.

Brown JR, Ye H, Bronson RT, Dikkes P, Greenberg ME. 1996. A defect in nurturing in mice lacking the immediate early gene fosB. Cell 86:297-309.

Caldarone B, Saavedra C, Tartaglia K, Wehner JM, Dudek BC, Flaherty L. 1997. Quantitative trait loci analysis affecting contextual conditioning in mice. Nature Genet 17:335-337.

Campbell IL, Gold LH. 1996. Transgenic modeling of neuropsychiatric disorders. Mol Psychiatry 1:105-120.

Cohen-Salmon Ch, Carlier M, Roubertoux J, Jouhaneau C, Semal C, Paillette M. 1985. Difference in pup care V-pup ultrasonic emissions and pup care behavior. Physiol Behav 35:167-174.

Crabbe JC, Gallaher EJ, Cross S J, Belknap JK. 1998. Genetic determinants of sensitivity to diazepam in inbred mice. Behav Neurosci 112:668-677.

Crawley JN. 1981. Neuropharmacological specificity of a simple animal model for the behavioral actions of benzodiazepines. Pharmacol Biochem Behav 15:695-699.

Crawley JN. 1996. Unusual behavioral phenotypes of inbred mouse strains. Trends Neurosci 19:181- 182.

Crawley JN, Belknap JK, Collins A, Crabbe JC, Frankel W, Henderson N, Hitzemann R J, Maxson SC, Miner LL, Silva A J, Wehner JM, Wynshaw-Boris A, Paylor R. 1997. Behavioral phenotypes of inbred strains of mice. Psychopharmacology 132:107-124.

Crawley JN, Corwin RL. 1994. Biological actions of cholecystokinin. Pep-tides 4:731-755.

Crawley JN, Davis LG. 1982. Baseline exploratory activity predicts anxiolytic responsiveness to diazepam in five mouse strains. Brain Res Bull 8:609-612.

Crawley JN, Gerfen CR, McKay R, Rogawski MA, Sibley DR, Skolnick P. 1998. Current Protocols in Neuroscience. New York: John Wiley and Sons, Inc.

Crawley JN, Paylor R. 1997. A proposed test battery and constellations of specific behavioral paradigms to investigate the behavioral phenotypes of transgenic and knockout mice. Horm Behav 31:197-211.

Crusio WE, Schwegler H, Lipp H-P. 1987. Radial-maze performance and structural variation of the hippocampus in mice: A correlation with mossy fibre distribution. Brain Res 425:182-185.

Dewey WL, Snyder JW, Harris LS, Howes JF. 1969. The effect of narcotics and narcotic antagonists on the tail-flick response in spinal mice. J Pharm Pharmacol 21:548-550.

Erickson JC, Hollopeter G, Palmiter RD. 1996. Attenuation of the obesity syndrome of ob/ob mice by the loss of neuropeptide Y. Science 274:1704-1707.

Fanselow MS. 1990. Factors governing one-trial contextual conditioning. Anim Learn Behav 18:264-270.

Flint J, Corley R, DeFries JC, Fulker DW, Gray JA, Miller S, Collins A. 1995. A simple genetic basis for a complex psychological trait in laboratory mice. Science 269:1432-1435.

Gerlai R. 1996. Gene-targeting studies of mammalian behavior: Is it the mutation of the background genotype? Trends Neurosci 19:177-181.

Geyer MA, Swerdlow NR, Mansbach RS, Braff DL. 1990. Startle response models of sensorimotor gating and habituation deficits in schizophrenia. Brain Res Bull 25:485-498.

Grahame NJ, Cunningham CL. 1995. Genetic differences in intravenous cocaine self-administration between C57BL/6J and DBA/2J mice. Psychopharmacology 122:281-291.

Grisel JE, Belknap JK, O'Toole LA, Helms ML, Wenger CD, Crabbe JC. 1997. Quantitative trait loci affecting methamphetamine responses in BXD recombinant inbred mouse strains. J Neurosci 17:745-754.

Heinrichs SC, Stenzel-Poore MP, Gold LH, Battenberg E, Bloom FE, Koob GF, Vale WW, Merlo Pich E. 1996. Learning impairment in transgenic mice with central overexpression of corticotropin-releasing factor. Neuroscience 74:303-311.

Hsiao K, Chapman P, Nilsen S, Eckman C, Harigaya Y, Younkin S, Yang F, Cole G. 1996. Correlative memory deficits, A_Belevation, and amyloid plaques in transgenic mice. Science 274:99-102.

Huszar D, Lynch CA, Fairchild-Huntress V, Dunmore JH, Fang Q, Berkemeier LR, Gu W, Kesterson RA, Boston BA, Cone RD, Smith FJ, Campfield LA, Burn P, Lee F. 1997. Targeted disruption of the melanocortin-4 receptor results in obesity in mice. Cell 88:131-141.

Jones BJ, Roberts DJ. 1968. A rotarod suitable for quantitative measurements of motor incoordination in naive mice. Naunyn-Schmiedebergs Arch Pharmacol 259:211.

Jucker M, Ingram DK. 1997. Murine models of brain aging and age-related neurodegenerative diseases. Behav Brain Res 85:1-25.

Kammescheidt A, Karo K, Ito KI, Sumikawa K. 1997. Adenovirus-mediated NMDA receptor knockouts in the rat hippocampal CA1 region. NeuroReport 8:635-638.

Kistner A, Gossen M, Zimmermann F, Jerecic J, Ullmer C, Libbert H, Bujard H. 1996. Doxycycline-mediated quantitative and tissue-specific control of gene expression in transgenic mice. Proc Natl Acad Sci U S A 93:10933-10938. LeDoux JE. 1994. Emotion, memory and the brain. Sci Am 270:50-57.

Lee MK, Borchelt DR, Wong PC, Sisodia SS, Price DL. 1996. Transgenic models of neurodegenerative diseases. Current Opin Neurobiol 6:651-660.

Lijam N, Paylor R, McDonald MP, Crawley JN, Deng C-X, Herrup K, Stevens KE, Maccaferri G, McBain CJ, Sussman DJ, Wynshaw-Boris A. 1997. Social interaction and sensorimotor gating abnormalities in mice lacking Dv11. Cell 90:895-90.

Lindzey G, Winston H, Manosevitz M. 1961. Social dominance in inbred mouse strains. Nature 191:474-476.

Lister RG. 1987. The use of a plus-maze to measure anxiety in the mouse. Psychopharmacology 92:180-185.

Livy DJ, Wahlsten D. 199l. Tests of genetic allelism between four inbred

mouse strains with absent corpus callosum. J Hered 82:459-464.

Maldonado R, Salardi A, Valverde O, Samad TA, Roques BP, Borrelli E. 1997. Absence of opiate rewarding effects in mice lacking dopamine D2 receptors. Nature 388:586-589.

Mathis C, Paul SM, Crawley JN. 1994. Characterization of benzodiazepine-sensitive behaviors in the A/J and C57BL/6J inbred strains of mice. Behav Genet 24:171-180.

Mathis C, Neumann PE, Gershenfeld H, Paul SM, Crawley JN. 1995. Genetic analysis of anxiety-related behaviors and responses to benzodiazepine-related drugs in AXB and BXA recombinant inbred mouse strains. Behav Genet 25:557-568.

Matthes H, Maldonado R, Simonin F, Valverde O, Slowe S, Kitchen I, Befort K, Dierich A, Le Meur M, Dolle P, Tzavara E, Hanoune J, Roques BP, Kieffer BL. 1996. Loss of morphine-induced analgesia, reward effect and withdrawal symptoms in mice lacking the g-opioid-receptor gene. Nature 383:819-823.

Mayford M, Bach ME, Huang YY, Wang L, Hawkins RD, Kandel ER. 1996. Control of memory formation through regulated expression of a CaMKII transgene. Science 274:1678-1683.

McDonald MP, Wong R, Goldstein G, Weintraub B, Cheng SY, Crawley JN. 1998. Hyperactivity and learning deficits in transgenic mice beating the human mutant thyroid hormone ~1 receptor gene. Learn Mem 5:289-301.

Miczek KA, Weerts E, Hanley M, Tidey J. 1994. Neurobiological mechanisms controlling aggression: Preclinical developments for pharmacotherapeutic interventions. Neurosci Biobehav Rev 18:97-110.

Miner LL. 1997. Cocaine reward and locomotor activity in C57BL/6J and 129/SvJ inbred mice and their Fl cross. Pharmacol Biochem Behav 58:25-30.

Morris RGM. 1981. Spatial localisation does not depend on the presence of local cues. Learn Motivat 12:239-260.

Murphy DL, Wichems C, Li Q, Heils A. 1998. Molecular manipulations as tools for enhancing our understanding of 5-HT neurotransmission. Trends Pharmacol Sci 20:246-252.

Nelson RJ, Demas GE, Huang PL, Fishman MC, Dawson VL, Dawson TM, Synder SH. 1995. Behavioural abnormalities in male mice lacking neuronal nitric oxide synthase. Nature 378:383-386.

Nelson RJ, Young KA. 1998. Behavior in mice with targeted disruption of single genes. Neurosci Biobehav Rev 22:453-462.

Nishimori K, Young LJ, Guo Q, Wang Z, Insel TR, Matzuk MM. 1996. Oxytocin is required for nursing but is not essential for parturition or reproductive behavior. Proc Natl Acad Sci U S A 93:11699-11704.

Norflus F, Tifft C J, McDonald MP, Goldstein G, Crawley JN, Hoffmann A, Sandhoff K, Suzuki K, Proia RL. 1998. Bone marrow transplantation prolongs life span and ameliorates neurologic manifestations in Sandhoff disease mice. J Clin Invest 101:1881-1888.

Oliverio A, Eleftheriou BE, Bailey DW. 1973. A gene influencing active avoidance performance in mice. Physiol Behav 11:497-501.

Paylor R, Baskall L, Wehner JM. 1993. Behavioral dissociations between C57BL/6 and DBA/2 mice on learning and memory tasks: A hippocampal-dysfunction hypothesis. Psychobiology 21:11-26.

Paylor R, Crawley JN. 1997. Inbred strain differences in prepulse inhibition of the mouse startle response. Psychopharmacology 132:169-180.

Paylor R, Nguyen M, Crawley JN, Patrick J, Beaudet A, Orr-Urtreger A. 1998. ~7 nicotinic receptor subunits are not necessary for hippocampal-dependent learning or sensorimotor gating: A behavioral characterization of Acra7-deficient mice. Learn Mem 5:302-316.

Picciotto MR, Zoli M, Rimondini R, Lena C, Marubio LM, Merlo Pich E, Fuxe K, Changeaux JP. 1998. Acetylcholine receptors containing the b2 subunit are involved in the reinforcing properties of nicotine. Nature 391:173-177.

Pick CG, Cheng J, Paul D, Pasternak GW. 1991. Genetic influences in opioid analgesic sensitivity in mice. Brain Res 566:295-298.

Porsolt RD, Anion G, Blavet N, Jalfre M. 1977. Behavioral despair in rats: A new model sensitive to antidepressant treatments. Eur J Pharmacol 47:379-391.

Ripley TL, Rocha BA, Oglesby MW, Stephens DW. 1999. Increased sensitivity to cocaine, and over-responding during cocaine self-administration in tPA knockout mice. Brain Res 826:117-127.

Rissman EF, Early AH, Taylor JA, Korach KS, Lubahn DB. 1997. Estrogen receptors are essential for female sexual receptivity. Endocrinology 138:507-510.

Rubinstein M, Mogil JS, Japon M, Chan EC, Allen RG, Low MJ. 1996. Absence of opioid stress-induced analgesia in mice lacking ~-endorphin by site-directed mutagenesis. Proc Nail Acad Sci U S A 93:3995-4000.

Sango K, Yamanaka S. Hoffmann A, Okuda Y, Grinberg A, Westphal H, McDonald MP, Crawley JN, Sandhoff K, Suzuki K, Proia RL. 1995. Mouse models of Tay-Sachs and Sandhoff diseases differ on neurologic phenotype and ganglioside metabolism. Nature Genet 11:170-176.

Silva AJ, Smith AM, Giese KP. 1997. Gene targeting and the biology of learning and memory. Ann Rev Genet 31:527-546.

Silver LM. 1995. Mouse genetics: Concepts and applications. New York: Oxford University Press.

Stenzel-Poore MP, Heinrichs SC, Rivest S, Koob GF, Vale WW. 1994. Overproduction of corticotropin-releasing factor in transgenic mice: A genetic model of anxiogenic behavior. J Neurosci 14:2579-2584.

Sterneck E, Paylor R, Jackson-Lewis V, Libbey M, Przedborski S, Tessarollo L, Crawley JN, Johnson PF. 1998. Selectively enhanced contextual fear conditioning in mice lacking the transcriptional regulator CCAAT/ enhancer binding protein fi. Proc Natl Acad Sci U S A 95:10908-10913.

Trullas R, Skolnick P. 1993. Differences in fear motivated behaviors among inbred mouse strains. Psychopharmacology 111:323-331.

Tsien JZ, Chen DF, Gerber D, Tom C, Mercer EH, Anderson DJ, Mayford M, Kandel ER, Tonegawa S. 1996. Subregion- and cell type-restricted gene knockout in mouse brain. Cell 87:1317-1326.

Wehner JM, Bowers B J, Paylor R. 1996. The use of null mutant mice to study complex learning and memory processes. Behav Genet 26:301-312.

Wehner JM, Radcliffe RA, Rosmann ST, Christensen SC, Rasmussen DL, Fulker DW, Wiles M. 1997. Quantitative trait loci analysis of contextual fear conditioning in mice. Nature Genet 17:331-334.

Wehner JM, Silva A. 1996. Importance of strain differences in evaluations of learning and memory processes in null mutants. Ment Retard Dev Disabil Res Rev 2:243-248.

Wersinger SR, Sannen K, Villalba C, Lubahn DB, Rissman EF, De Vries GJ. 1997. Masculine sexual behavior is disrupted in male and female mice lacking a functional estrogen receptor a gene. Horm Behav 32:176-183.
¹Crawley JN. 2000. What's Wrong With My Mouse? Behavioral Phenotyping of Transgenic and Knockout Mice. New York: Wiley.