NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Buccafusco JJ, editor. Methods of Behavior Analysis in Neuroscience. 2nd edition. Boca Raton (FL): CRC Press/Taylor & Francis; 2009.

Cover of Methods of Behavior Analysis in Neuroscience

Methods of Behavior Analysis in Neuroscience. 2nd edition.

Show details

Chapter 8The Behavioral Assessment of Sensorimotor Processes in the Mouse: Acoustic Startle, Sensory Gating, Locomotor Activity, Rotarod, and Beam Walking

, , , and .


Assessment of sensorimotor competence is an important part of the evaluation of animal behavior. Measurement of sensorimotor performance is of obvious importance in investigations of sensory or motor processes; however, the effects of experimental manipulations on sensorimotor performance have broader implications for behavioral neuroscience because behavioral experiments typically measure motor responses to sensory information. Thus, the results of behavioral experiments designed to assess other neurobiological processes often cannot be properly interpreted without considering concomitant effects on sensorimotor function. For example, if a lesion or genetic manipulation impairs performance on a spatial memory test, such as the radial arm maze, this impairment cannot be interpreted as evidence of cognitive dysfunction unless it is first established that it is not the result of sensorimotor deficits. Moreover, sensorimotor effects of manipulations can often be used in animal models as surrogates for effects that are more difficult to measure, and relatively simple variations of sensorimotor measures can be used as indices of performance in other behavioral domains, including cognition and emotion. A number of behavioral tasks have been designed to assess sensorimotor performance in rodents, and this chapter focuses on five general procedures—acoustic startle, sensory gating, open field exploration, rotarod, and beam walking.

The startle reflex is a stereotyped motor response to a sudden, intense stimulus that has been assessed experimentally in a variety of species, including rats, mice, cats, monkeys, and humans [1,2]. In rodents, the startle response is typically evoked using either acoustic or tactile stimuli and is characterized by contractions of the major muscles of the body, generally leading to extension of the forepaws and hind paws followed by muscle flexion into a hunched position. Mapping studies have demonstrated that the acoustic startle reflex is mediated by a specific neural pathway with acoustic information entering the CNS through auditory nerve input to the cochlear nucleus, which projects to the reticular pontine nucleus via the lateral lemnisus. Motor outputs are generated in the reticular pontine nucleus, which projects to the ventral spinal horn through the reticulospinal tract. Although the basic reflex pathway appears to be relatively simple, the reflex is subject to modulatory influences from higher brain structures [3].

Measurement of acoustic startle responses can provide general information regarding sensorimotor processing, but measurement of the reflex under conditions that engage the influence of higher brain centers provides an even richer source of data. For example, presentation of lower intensity acoustic stimuli immediately prior to the acoustic startle stimulus attenuates the response to the startle stimulus. This phenomenon, called prepulse inhibition (PPI) of startle reflex, is regulated by forebrain neural circuits and is considered an operational measure of sensorimotor gating, a filtering mechanism to prevent information flooding in the brain so that attentional recourses can be selectively allocated to salient stimuli [4,5]. The normally functioning brain has endogenous mechanisms that filter out the multitude of irrelevant sensory stimuli from those of importance.

Sensory gating is a term given to this filtering mechanism and can be measured. A sensory gating deficit, or the lack of an ability to inhibit responding to the test stimulus, is used as a clinical measure of schizophrenia [6–9]. The two methods of sensory gating described in this chapter, PPI and N-40 gating, both measure the transmission of auditory sensory information to the nervous system. For PPI the observable and measured response is based on the motor output (startle) following a loud acoustic stimulus. Inhibition of the response to this stimulus is observed if the stimulus is preceded very shortly (10–500 msec) by a prepulse stimulus to which the organism normally does not respond. PPI is defined as reduction of a response to a stimulus. In contrast, the N-40 response is a measure of electrical brain activity to repeated auditory stimuli. Although more involved, there are several advantages to measuring sensory gating using electrophysiological recordings. First, afferent neuronal activity largely contributes to sensory gating evoked potential (EP) response, while afferent, efferent, and neuromuscular activity contribute to the PPI response. Therefore, recording EPs allows for the study of sensory information processing without the influence of descending motor responding as in PPI. Secondly, although the hippocampus and cortex are most frequently recorded, electrophysiological techniques allow the study of the inhibitory processes in different brain regions [10].

Pharmacological manipulations and strains of mice yield sensory deficits reminiscent of those seen in schizophrenic patients. For example, the pro-psychotic amphetamine produces N-40 sensory gating deficits in rodents and P-50 deficits in humans [11,12]. As another example, the α7 nicotinic receptor subtype, is thought to be deficient in schizophrenia, and decreased hippocampal α7 receptor levels in DBA/2 mice are thought to account for the N-40 sensory gating deficits expressed in this strain [13]. Also, as in schizophrenic patients and PPI procedures in rodents, atypical antipsychotics reverse sensory gating deficits in both PPI and N-40 in DBA/2 mice [14,15]. Thus, sensory gating measures show considerable parallels across species and have good rodent–human translation. Other sensory gating deficit models have been reported, but the use of the DBA/2 mouse is described herein as representative of an approach to study the nicotinic-acetylcholine system and sensory gating.


8.2.1. Equipment

The equipment used to measure the startle response has varied from simple lab-made devices [16,17] to more sophisticated units available from various commercial suppliers. We have used the SR-LAB Startle Response System (San Diego Instruments, San Diego, California, USA) and Kinder Scientific Startle Monitor (SM 100 version 4.1, Poway, California, USA). In addition, acoustic startle equipment can also be obtained from Coulbourn Instruments. The ability to deliver stimuli for 5–1000 msec with consistent intensity is important. Each device is typically enclosed in a larger soundproof cubicle that isolates the animal in the presence of background noise. This also serves to protect the animals in the immediate vicinity from being exposed to the acoustic startle stimulus. A simpler, cheaper SR-LAB screening system is also available from San Diego Instruments, but it has no enclosure and the animals must obviously be isolated by location from other test animals. The magnitude of the response of an animal depends on the size of the animal, which means that the assessment of the acoustic startle reflex in the mouse requires more sensitive equipment than the assessment of the reflex in the rat.

The SR-LAB system and Kinder Scientific Startle Monitor include a separate isolation chamber for each individual startle unit. The outer sound-attenuating chamber is illuminated and ventilated with a small fan that also provides some level of background noise. An acoustic sound source is located in the upper part of this chamber and consists of a loudspeaker that delivers a full spectrum white noise that is computer controlled for duration and decibel level. The startle unit is available in assorted sizes to accommodate mice or rats. With the SR-LAB system, the cylindrical animal enclosure and the enclosure base are installed within the SR-LAB chamber. Each animal enclosure has an attached piezoelectric motion sensor that detects the movement of the animal. The signal from the motion sensor is sent to a computer for digital transformation. The motion sensor supplied by San Diego Instruments has an adjustable potentiometer on the underside. Compared to the SR-LAB system, the Kinder Scientific Startle Monitor has some unique features that we liked. The Kinder Scientific equipment animal enclosure is separable from the sensing plate that serves as a motion sensor, allowing for easy cleaning of the animal enclosure without damaging the motion sensor. Also, the SR-LAB system requires calibration to standardize test chambers before each test, whereas calibration is not necessary for each test with the Kinder Scientific Startle Monitor. Furthermore, Kinder Scientific uses direct readings for the magnitude of acoustic stimuli rather than analog levels used by the SR-LAB system that require the use of a sound meter to calculate decibel levels.

8.2.2. Setup and Decibel Confirmation

The initial setup of the equipment is fairly simple and the experimental programming is easily accomplished by following the manual provided. Measuring the decibel levels in each enclosure prior to commencing experiments is recommended. A RadioShack (Allied Electronics) Sound Level Meter #33-2050 set on slow response (“A” weighting) is a simple and inexpensive device for measuring the intensity of the acoustic stimuli. Kinder Scientific sells a device with a sensor connected via a cord to the decibel meter, making it convenient to read the sound level outside the closed chamber. The startle unit sound duration must be set long enough (e.g., 6000 msec) to accurately measure sound level, and the sound meter should be placed in the position normally occupied by the animal holder. If isolation chambers are used, calibration should be done with all of the chamber doors closed to reduce the likelihood that sound from the other chambers will influence the reading made in the chamber being calibrated.

8.2.3. Stimulus Parameters

Startle responses can be measured over a period of time up to 1 sec after the presentation of the startle stimulus. However, since the startle response is typically over within 100 msec of stimulus presentation, the window should be set to include only movements generated within the first 200 msec following the startle stimulus. When analyzing data, use the maximal voltage generated during this 200-msec period; however, it is also possible to use average voltage across the entire response window if desired. For mice, stimuli with intensities of 90 dB or higher will typically produce startle responses, although there is some strain-dependent variability [18]. The magnitude of the startle response varies as a function of the intensity of the startle stimulus, so more reliable responses are often obtained at higher intensities. Acoustic stimuli intensities should not be set higher than 120 dB to avoid producing damage to the ear and the loudspeakers.

8.2.4. Testing Location

Animals are brought to a convenient holding area near the room containing the startle chambers for acclimation. Thus, the startle equipment is best located within an inner room with a heavy door. This provides additional sound attenuation and keeps animals held in the vicinity of the testing room from being exposed to the startle stimulus.

8.2.5. Subjects

Among the mouse strains used, the DBA/2J (Jackson Laboratories, USA) mice, 11–17 g, exhibit a naturally occurring low PPI and thus provide a window to detect a PPI enhancing effect (See Figure 8.1) [19,20]. One caveat concerning the DBA/2J mice is mice older than 8 wk have hearing loss and thus only young DBA/2J mice are used. Also used are CD1 and C57BL/6 mice, 28–40 g, for PPI or startle habituation studies. The animals are housed eight per cage (reflecting the number of test chambers employed) with water and food available ad libitum. Aggressive dominant males should be removed from holding cages before commencement of any experiment. Best results are often obtained from animals that have been protected from stressors and have been habituated to the laboratory/animal quarters for at least 7 days.

FIGURE 8.1. The effect of 70, 75, and 80 dB prepulse on prepulse inhibition (PPI) displayed by CD1, C57BL/6, and DBA/2J mice.


The effect of 70, 75, and 80 dB prepulse on prepulse inhibition (PPI) displayed by CD1, C57BL/6, and DBA/2J mice. Shown are mean ± SEM. Note that increasing the intensity of the prepulse levels increases PPI. Kinder Scientific Startle Monitor (more...)

8.2.6. Acoustic Startle Protocol

After placing the animal in the test chambers allow a 5-min adaptation period before the start of the session. Background white noise (65 dB) is present during this adaptation period and throughout the session. The session starts with four 120 dB, 40 msec sound bursts. These are not included in the analysis because the responses to the first few startle bursts generally differ in magnitude from the rest of the trials. Thus, exposure to these initial bursts allows for the establishment of a stable baseline. Following this, acoustic startle trials are initiated. For simple assessment of acoustic startle responses, use two or three stimulus intensities (90 and 105 dB or 90, 105, and 120 dB). The stimuli are 40 msec in duration and are presented in a quasi random order so that an equal number of presentations of each stimulus intensity is included in each half of the session, and no single intensity is presented more than two times in succession. The time between stimuli averages 15 sec but this interval should be varied within a range of 5–30 sec so that the animals do not anticipate the stimulus. At least 10 trials at each stimulus intensity should be used to obtain reliable results. An alternative to evaluating the magnitude of the startle response is to measure the startle threshold. Here, a wider range of stimulus intensities, e.g., from 70 to 120 dB, is used. Stimuli are presented in quasi-random order, and the lowest intensity producing a reliable response is determined. Since some habituation of the response can occur both within a session and between sessions (see below), trials at each intensity should always be evenly distributed within a session (but they should not, of course, be presented in a predictable sequence).

8.2.7. Startle Habituation

This measure is of interest in the area of schizophrenia research, since schizophrenics show impaired startle habituation, an impairment that may be related to the hypervigilance characteristic of this condition [21,22]. For startle habituation, a single stimulus intensity is repeatedly presented using either a fixed or variable interval. Responses normally decline (habituate) over trials. An example of the data generated in such an experiment is shown in Figure 8.2. In this case, each session is initiated with a 10-min acclimation period followed by 121 successive 120 dB, 40 msec trials separated by a fixed inter-trial interval (ITI) of 10 sec. The first trial is excluded from analysis due to variability. Each block has 20 trials. Two groups of CD1 mice received ip injection of water and MK-801 at 0.3 mg/kg, respectively, 15 min before the test. As shown in the figure, startle habituation is displayed by the water-treated group but not by the MK-801-treated group. Long-term habituation of the startle response can also be used as a memory index by conducting a second session at a later time (e.g., 24 hr later) and assessing long-term retention of habituation.

FIGURE 8.2. An example of startle habituation in two groups of CD1 mice treated with water and MK-801 at 0.


An example of startle habituation in two groups of CD1 mice treated with water and MK-801 at 0.3 mg/kg, respectively. The study was carried out in Kinder Scientific Startle Monitor equipment. Shown are mean ± SEM. The study showed that MK-801 (more...)


8.3.1. Prepulse Inhibition

In this version of the test, the attenuation produced by a low intensity stimulus presented just before the startle stimulus is assessed. We sometimes assess PPI on the day following standard startle testing. This serves to habituate the animals to the basic handling procedures and, in the case of pharmacological studies of PPI, allows groups to be matched for baseline startle, as well as elimination of animals that either startle excessively or do not respond. The animals are given a 5-min acclimation period in the startle chambers during which a 65-dB background noise is presented. This background noise remains throughout the entire test. Following the 5-min acclimation period, four successive trials of 40-msec noise bursts at 120 dB are presented. These trials are not included in data analysis. Subjects are then exposed to five different types of acoustic stimuli in a randomized order: pulse alone (120-dB noise for 40 msec), no stimulus (no stimulus is presented), and three separate prepulse + pulse combinations, with prepulse set at three sound levels of 70, 75, and 80 dB for 20 msec followed by a 40-msec pulse at 120 dB. There are 100 msec between the prepulse and the pulse. A total number of 12 trials under each acoustic stimulus condition are presented with average 20-sec variable intervals ranging from 5 sec to 25 sec. The inclusion of four pulse-alone trials in the beginning of the experiments is to help normalize the response of the mice, as there is rapid habituation to the startle responses seen within the first few trials.

More than one level of prepulse intensity is used, whereas a “no stimulus” trial is used to assess the influence of background movement on startle measures. To reduce variability, at least 12 trials of each type should be conducted within a single session. As with standard startle testing, trial types and ITIs should be presented in a quasi-random, balanced manner with equal representations of trial types and intervals in each half of the session. This allows the session to be analyzed in two blocks to assess any changes over time. Statistical Analysis

The results can either be printed out or, more practically, assembled in a computer text file. Microsoft Excel or similar spreadsheet software easily assemble the data into the various trial types. Percent PPI is calculated as follows:

[1-(startle responseto prepulse+pulse)÷(startle responseto pulse alone)]×100.

If there is a drug treatment involved using naïve rodents, data are typically analyzed with two-way analysis of variance (ANOVA), with the drug treatment as a between-subjects variable and prepulse level as a repeated measure. If a significant treatment effect and interaction of treatment and prepulse are identified, percent PPI is then analyzed by post-hoc comparisons to compare group means at each prepulse level. If a significant treatment effect is identified without the presence of significant interaction of treatment and prepulse, percent PPI can be collapsed across prepulse levels and analyzed by post-hoc comparisons for treatment difference. Use of a post-hoc test such as Fisher’s post-hoc PLSD and Dunnett’s test in order to compare the groups treated with vehicle and the groups treated with drug should be decided a priori. To evaluate drug effects on the response to the startle (s120) alone, a separate one-way ANOVA, followed by post-hoc comparisons if there is a significant treatment effect, is calculated. Programs such as JMP statistics software package (SAS Institute, Cary, North Carolina, USA) or Graph Pad Prism (Graph Pad Software, San Diego, California, USA) are useful for statistical analysis. Association of PPI and Startle Responses

Substantial effects of an experimental treatment, resulting in an increase or decrease of the startle response (s120), may confound the effects on PPI. However, there is no agreement upon how PPI and startle responses are correlated. A reduced startle response could be accompanied by a PPI reduction, PPI enhancement, or no change of PPI dependent upon the drug treatment. Our previous studies show that antipsychotics enhance PPI responding while substantially decreasing startle responses (Figure 8.3) [20]. To address the question whether the PPI-enhancing effects observed with antipsychotics is secondary to their effects on startle response [20], we pooled data of vehicle and 1 mg/kg risperidone-treated mice from studies using risperidone as positive control. We then sorted the data by startle responses in ascending values. In order to obtain equal startle magnitude in the vehicle and risperidone groups, the pairs of individuals with risperidone and vehicle treatment listed next to each other were kept for further analysis. In cases where there were two possible pairs (e.g., if there was a sequence of vehicle-risperidone-vehicle, risperidone could be paired with the first vehicle and the second vehicle), the pair that had the smaller startle difference between vehicle- and risperidone-treated mice was chosen. This selection resulted in n = 19 in water- and risperidone-treated groups, respectively. The remaining mice were excluded. One-way ANOVA revealed (Figure 8.4) that there was a significant difference (p < 0.01) in percent PPI between risperidone- and water-treated mice when their mean values of startle magnitude were equal, suggesting that the effects on PPI responding can be independent of the effects on startle magnitude. A study investigating basal acoustic startle responses and PPI among several inbred mouse strains showed there was no correlation between the magnitude of basal acoustic startle responses and PPI, [23], suggesting that different physiological processes are involved in basal acoustic startle and PPI of startle reflex. These observations support the notion that different neurobiological processes may underlie gating processes and the startle reflex.

FIGURE 8.3. Effects of the antipsychotics, haloperidol and risperidone on prepulse inhibition (PPI) in DBA/2J mice.


Effects of the antipsychotics, haloperidol and risperidone on prepulse inhibition (PPI) in DBA/2J mice. The effects on PPI and startle responses are presented on the left and the right side of the panels, respectively. Haloperidol (a) and respiridone (more...)

FIGURE 8.4. All of the data of vehicle- and 1-mg/kg-risperidone-treated DBA mice from several mouse studies were pooled and then the startle magnitude of vehicle- and risperi-done-treated individuals were matched (for details, see section 8.


All of the data of vehicle- and 1-mg/kg-risperidone-treated DBA mice from several mouse studies were pooled and then the startle magnitude of vehicle- and risperi-done-treated individuals were matched (for details, see section 8.3.1, “Prepulse (more...) Sample Prepulse Inhibition Experiment

An experiment was conducted to evaluate the effects of BP 897, a preferential dopamine D3 receptor antagonist, on PPI [20]. In pharmacological experiments, it is often important to conduct the maximum number of trials in the shortest possible time because of pharmacokinetic considerations with many test compounds. Session duration will obviously increase as more trial types/numbers are added, and this was taken into account when we set up our standard rodent testing paradigm. Please see section 8.3.1, “Prepulse Inhibition,” for details of the paradigm. The sequence of trials is illustrated in Table 8.1. The 68-trial session used ran for approximately 25 min. In this experiment, DBA/2J mice (11–13 per group) were used. Included in the study was a positive control group in order to be certain that the experiment ran as expected. In this study, risperidone at 1 mg/kg was used as a positive control. Both BP 897 and risperidone were dissolved in 1 N HCL and then titrated to a final pH of 5 with 1 N NaOH. Both of the compounds were given ip in a volume of 10 mL/kg 30 min before the test. The results, which were analyzed using a two-way ANOVA with treatment as a between-subjects variable and prepulse as a repeated measure, revealed a significant main effect of treatment on percent PPI [F(5, 66) = 2.936, p < 0.05] in the absence of significant interaction of treatment and prepulse; the data was collapsed across the three prepulses and the average percent PPI values of the three prepulses were analyzed with Fisher’s PLSD to compare the vehicle-treated group to each of the drug-treated groups. As shown in Figure 8.5, the positive control, risperidone, significantly increased PPI (p < 0.05), suggesting the study was valid. The test compound, BP 897, also significantly enhanced PPI at 8 mg/kg (p < 0.05). Startle response to s120 was analyzed with a one-way ANOVA with treatment as a between-subject variable. A significant main effect of treatment on startle was identified [F(5, 66) = 5.575, p < 0.01]. The follow-up Fisher’s PLSD showed that risperidone, but not BP 897, significantly reduced startle response to s120.



Session Protocol for Prepulse Inhibition Experiment

FIGURE 8.5. Prepulse inhibition (PPI)-enhancing effects of BP 897 in DBA/2J mice.


Prepulse inhibition (PPI)-enhancing effects of BP 897 in DBA/2J mice. BP 897 at 8 mg/kg significantly increased percent PPI without affecting startle response to pulse alone, while the positive control, risperidone, significantly increased PPI and decreased (more...)

As previously mentioned, DBA/2J mice have a lower PPI response compared to other mouse strains and thus would provide enough of a window for seeing a PPI-enhancing effect following a drug treatment. We also use other mouse strains such as CD1 mice in studies to investigate PPI deficits. Pharmacological disruption of PPI in mice can also be obtained with compounds that influence dopaminergic (e.g., apomorphine, amphetamine), glutaminergic (e.g., phencyclidine, MK-801), muscarinic cholinergic (e.g., scopolamine), and serotoninergic (e.g., 2,5. dimethoxy-4-iodoamphetamine) neurotransmission [15].

8.3.2. N-40 Sensory Gating Introduction

Evoked potentials are synchronous discharges of neuronal circuits or populations that are time locked to a sensory stimulus. Sensorimotor gating can be measured by recording electroencephalographic (EEG) EP responses to pairs of identical auditory stimuli [9]. Each stimulus of a pair in a sensory gating paradigm is given a specific name; the first stimulus of a pair is typically called the conditioning stimulus, or S1, while the second is referred to as the test stimulus, or S2. The terminology used for stimuli in this section will be “condition” and “test.” Ordinarily, the EP elicited by the test stimulus, typically an auditory click or tone, is smaller in amplitude than that evoked by the conditioning stimulus (Figure 8.6). A lowered test response amplitude is widely considered to be indicative of sensory gating or the filtering function of the brain [6,7]. A gated EP response to visual stimuli, such as to paired strobe flashes, is not observed in humans; however, cross-modal visual to auditory sensory gating phenomenon have been reported [24–26]. The majority of animal sensory gating studies are reported to use auditory stimuli.

FIGURE 8.6. Example of an auditory evoked potential sensory (inhibitory) gating response from a C3H strain mouse.


Example of an auditory evoked potential sensory (inhibitory) gating response from a C3H strain mouse. These electrographic potentials were recorded from the CA3 region of the hippocampus. The condition and test stimuli are separated by a delay of 500 (more...)

The component of the human auditory EP that shows this gating response is the P-50 wave, a major positive deflection in the ongoing EEG with a latency of about 50 msec after the stimulus. It is thought that some form of active inhibition of neuronal activity is initiated by the first stimulus, and suppresses neuronal activity thereafter [27,28]. Although somewhat controversial, in rodents the major EP component analogous to the human P-50 is thought by some to be the N-40 (negative polarity, 40 msec latency), and is typically suppressed in many strains after presentation of the test stimulus [29,30]. Method

The techniques for recording sensory EPs in anesthetized and unanesthetized rodents are well established. Recording freely moving unanesthetized versus anesthetized rats and mice is somewhat more difficult because of movement artifact; however, this approach avoids any possible confounds of the anesthetic in pharmacological studies. The recording techniques described herein are of auditory EPs recorded in the hippocampus from unanesthetized DBA/2 mice, but many of these methods are applicable to recordings in rats and recordings in other areas of the brain. Subjects and Surgery

For implantation of EEG recording electrodes, DBA/2 mice (16–20 g, 6–7 wk, Harlan) are anesthetized with a solution of 2.8% ketamine, 0.28% xylazine, 0.05% acepromazine (Sigma Chem. Co.) at 140 mg/kg of ketamine. Other anesthetics produce less satisfactory results in mice in terms of survival. After achieving a stable plane of anesthesia, scalp hair is removed and the skin is cleaned with a standard veterinary disinfectant solution (e.g., povidone iodine). The mouse is placed in a Kopf student stereotaxic frame and a sagittal incision approximately 6 mm long is made along the centerline to expose the bone between lambda and bregma. The bone is dried with a 30% hydrogen peroxide solution, which makes suture landmarks easier to see. Three drill holes (#68 drill bit) are made at medial lateral (ML) 1.0, 1.8, and 2.6 mm from the central suture. All three are located at anterior posterior (AP) −1.8 mm from bregma, thus, they are in a plane perpendicular to the central suture. The hole at ML 2.6 and AP −1.8 mm is for the electrode directed at the CA3 region of the hippocampus. The holes at ML 1.0, AP −1.8 mm, and ML 1.8, AP −1.8 mm, are for two electrodes that lie on the surface of the cortex. The depth of the hippocampal electrode tip is dorsal ventral (DV) 1.65–1.70 mm below the surface of the cortex. The depth of the cortical electrodes are dorsal ventral (DV) 0.5 mm from the surface of the skull; a distance that results in the electrode tip being in contact with, but not penetrating, the cortical tissue. The electrodes we use for recording mouse EPs are from Plastics One, Inc., Roanoke, Virginia, USA. They are tripolar stainless steel wires that have an integral mounting pedestal for connecting a tether when recording the EEG. These electrodes are cut to a length that is appropriate for the cortical or subcortical target. After experiments are completed, the accuracy of the electrode placement should be verified histologically (e.g., crystal violet staining) and, if inaccurate, coordinates and electrode length should be adjusted accordingly. Two additional holes are drilled in the contralateral skull for placement of anchoring screws (#00-90, 1/16 in). After the screws are driven into the skull, the tripolar electrode is lowered into the brain with a stereotaxic electrode holder. Before completely inserting the electrodes, a drop of cyanoacrylic glue is placed on the skull underneath the electrode pedestal. The electrodes and pedestal are then completely lowered and the glue is allowed to dry for several minutes. The pedestal is permanently affixed to the skull with dental acrylic. The mice should be allowed to recover for at least four days before conducting experiments. Recording of Paired Stimulus Sensory Gating Evoked Potentials

Routine use of this technique requires a dedicated lab space. To minimize the influence of ambient noise, the animals are recorded in plastic shoebox cages within acoustically isolated chambers (Med Associates) that are internally lined with sound-absorbent foam. A mouse previously implanted with indwelling electrodes is connected to a flexible electrical tether and swivel device, also called a commutator, which is mounted to the ceiling of the isolation chamber. The commutator allows the animal to rotate freely without twisting the tether. Commercially available tether and commutator systems from Plastics One, Inc. have at least two channels, and some models have up to 16 channels or more to record multiple brain areas simultaneously. For recording EEGs from one location, two channels will be required; one is the active channel, in our example the hippocampal electrode, and the other is a reference channel, in our example one of the cortical electrodes. We do not use the third electrode of the tripolar configuration. The small microvolt EEG biosignals must be amplified and filtered, thus the commutator is electrically cabled to differential AC amplifiers (Grass Instrument Division, Astro-Med, Inc., West Warwick, Rhode Island, USA). Cortical and subcortical EEGs are typically amplified by a factor of 1000, and band pass filters are set at 1 and 300 Hz. A standard PC with large storage capacity is used in conjunction with acquisition software (e.g., Datawave, Inc., Berthoud, Colorado, USA) that digitizes the EEG signals at 1000 Hz, a sample rate that captures the rapid rise and fall of auditory EP waveforms. Currently, our lab has eight recording chambers that can record two channels of EEG per mouse, thus EEGs from eight mice can be recorded simultaneously. This shortens the amount of time required to record animals in a given day without reducing N size. Auditory Stimuli

Sensory gating auditory EPs are generated by presentation of paired white noise bursts (5 msec duration) from a wide-dynamic-range speaker mounted within the recording chamber at a distance of approximately 15–20 cm from the mouse. Tone bursts of 2–3 KHz have also been successfully used. The auditory stimulus can be generated by a computer, or by an external audio generator (e.g., from Med Associates). The first conditioning auditory stimulus is followed 0.5 sec later by an identical auditory test stimulus. The latency between the two stimuli of the pair, or inter-stimulus interval (ISI), is critical. Generally, ISIs of less than 1 sec result in lowered test stimulus EP amplitudes. While an ISI of 0.5 sec produces a consistent suppression of the test EP response in humans and animals, ISIs of greater than 2–3 sec yield similar condition and test stimulus EP amplitudes (Figure 8.7). Thus, gating function, initiated by sensory input, seems to gradually extinguish over time to a point where the brain is reset to a ready-state level of maximal sensitivity to sensory input. The length of time between stimulus pairs, or trials, is 15 sec. ITIs are generally reported to be 10–15 sec, a range that minimizes a potential influence from the previous stimulus pair. Clear and measurable auditory EP responses cannot be obtained from a single stimulus because of variability of the background EEG. We present 120 paired auditory stimuli to subjects and the data acquisition computer averages each EEG trace starting 100 msec before and 900 msec after every conditioning stimulus. This averaging produces prototypical-evoked waveforms (potentials) with peaks at relatively stable latencies after the auditory stimulus. Thirty minutes are required to obtain an average EP using an ITI of 15 sec. Other labs using anesthetized mice report using fewer stimulus repetitions to obtain auditory EPs, probably because of a more stable background EEG under anesthesia. The volume of auditory stimuli is 65 dB, which is about 5 dB above the constant 60 dB background noise of the recording chambers. The use of louder intensities has been reported, but this relatively low level yields clear averaged EPs for us without producing a startle response. The acquisition software is set to acquire 1 sec of data starting 100 msec before and ending 900 msec after the initial conditioning stimulus. In addition to recording the EEG EPs, the software triggers the audio generator to synchronize the EP recording with the stimulus.

FIGURE 8.7. Condition and test stimulus evoked potential (EP) responses are superimposed.


Condition and test stimulus evoked potential (EP) responses are superimposed. The recording is from the hippocampus of a CD-1 mouse. In A, the 500 msec inter-stimulus interval (ISI) results in an attenuated test stimulus EP amplitude. In B, no attenuation (more...) Evoked Potential Analysis

The hippocampal auditory EP response is identified as a peak in the ongoing EEG at a latency of 15–25 msec after the stimulus, followed by the peak of opposite polarity at 30–50 msec after the stimulus. The difference in amplitude between these peaks is defined as the N-40 amplitude in microvolts (μV). N-40 amplitude is determined for both the averaged conditioning (CAMP) and test (TAMP) EPs. A sensory gating ratio is calculated by dividing the test amplitude by the conditioning amplitude. This calculation, termed the T:C ratio, is the index by which sensory gating is assessed (Figure 8.8). T:C ratios in schizophrenic patients are often well above 0.4; in other words, TAMP is greater than 40% of CAMP. In anesthetized and unanesthetized DBA/2 mice, T:C ratios usually exceed 0.5, while in other strains such as C3H, T: C ratios often are below 0.4 (Figure 8.8). These findings have led to a considerable use of the DBA/2 mouse strain as a model of sensory gating deficits. On average, we have found that about 10% of DBA/2 mice have control T:C ratios below 0.4. Mice with low T:C ratios are considered to be clinically normal and show little or no response to drugs that improve gating. Thus, animals with control T:C ratios below 0.4 are retrospectively dropped from studies and data analysis.

FIGURE 8.8. Examples of sensory evoked potentials from either DBA/2 or C3H mice.


Examples of sensory evoked potentials from either DBA/2 or C3H mice. The N-40 amplitude is measured from the positive peak about 20 msec after the stimulus (P20) to the trough of the N-40 wave. In (a), the DBA/2 mouse has a calculated test amplitude:conditioning (more...) Example Drug Studies with DBA/2 Mice

In a typical study, drugs are administered immediately before mice are placed into the isolation chambers and initiation of auditory EP recording. The duration of the recordings can vary, but they last for 30 min in many of our studies. Multiple treatments (e.g., doses), including vehicle control, are administered to each mouse on separate days with at least 48 hr between treatments. So, for example, a three-point dose response with vehicle would require 4 recording days. With a 48-hr washout between days, this study would take 7 days. This within-subjects design allows each mouse to serve as its own control. For drugs with robust effects on sensory gating, such as nicotine, an N size of eight is sufficient to demonstrate a statistically significant effect. However, we routinely use an N size of 12 or more to detect low dose effects, or the effects of weaker drugs. Paired t-tests for two–group or repeated measures ANOVA for multiple-group statistical evaluation are used. For a drug study where only the dose is varied, a one-way ANOVA is used, while for time course or multiple drug studies, a two-way ANOVA is employed. Post-hoc analyses for between-group comparisons are Newman-Keuls for the one-way ANOVA and Boferroni for the two-way ANOVA.

As mentioned before, the atypical antipsychotics clozapine and olanzapine improve sensory gating in DBA/2 mice, that is, they lower T:C ratios (Figure 8.9) [15]. Sensory gating deficits in schizophrenia are hypothesized to be, in part, mediated by α-7 receptor hypofunction [31]. The selective α-7 agonist GTS-21 and the nonse-lective nicotinic agonist nicotine improve sensory gating in both DBA/2 mice and schizophrenic patients [32–34]. Thus, one of the uses of the DBA/2 mouse is to study compounds that may have potential therapeutic effects in schizophrenia. Figure 8.10 shows the effects of the selective α-7 agonist A-582941 on sensory gating T:C ratios in DBA/2 mice [35]. As with GTS-21, A-582941 lowers T:C ratios, an effect consistent with improved sensory gating. Substantial literature exists on the pharmacology of sensory gating using in vivo electrophysiological recordings of auditory EPs. It has been particularly useful in advancing the understanding of the role of brain cholinergic systems in information processing, and the role of cholinergic deficiencies in disease states such as schizophrenia.

FIGURE 8.9. Clozapine significantly lowers T:C ratios in DBA/2 mice.


Clozapine significantly lowers T:C ratios in DBA/2 mice. **p = 0.0056, t9 = 3.622, paired, two-tailed t-test. Source: Author’s unpublished data.

FIGURE 8.10. Example of the effects of a selective α7 agonist.


Example of the effects of a selective α7 agonist. In panel (a), A-582941 (3.0 μmol/kg ip) significantly lowers test amplitude:conditioning amplitude (T:C) ratios in DBA/2 mice, one-way repeated measures ANOVA p = 0.0158 F(2,38) = 4.955, (more...)


Although many experimenters view locomotor activity as an overly simplistic measure that provides only limited information, alterations in this behavior can reveal important information on potential mechanisms of drug action. Moreover, locomotor activity may influence functional outcome in animal models of CNS injury or disease. For example, many different psychoactive drugs can act at neuronal receptor sites and directly affect motor function. Similarly, brain injury models employed by many researchers can produce subtle or sometimes pronounced alterations in motor behavior. Furthermore, genetically altered animals have become popular in an attempt at unmasking the molecular and cellular correlates of such behaviors as learning and memory in addition to numerous disease states. These animal models, however, are not without their drawbacks, including profound changes in motor function that can confound the interpretation of behavioral results. Therefore, it is important for the neuroscientist to be aware of, and to characterize, these changes carefully. The following examines several methods for assessing motor and exploratory activity in the adult mouse.

Locomotor and exploratory behavior may also be influenced by several other factors such as time of day (rodents are nocturnal animals and are therefore significantly more active during periods of darkness); anxiety (animals may be more or less active depending on the situation to which they are exposed); state of wakefulness or arousal (stimulants will tend to increase activity, whereas sedatives will tend to decrease activity, although the magnitude of these effects can often depend upon the strain of animal used); environmental novelty (mice tend to exhibit increased exploratory behavior when 0exposed to a novel environment and decreased activity upon reexposure to the same environment, i.e., “habituation”); motivation (food-deprived mice may show increased activity); age (younger rodents are more active than aged animals); general health; and genetic strain (C57BL/6 mice are more active than 129-derived animals). All of these factors necessitate careful experimental design and it is therefore prudent to control for these and maintain consistency from the outset. However, natural variations in activity and stress levels often exist between mice of the same strain despite controlling for all of the factors described above, hence the need for adequate group sizes that will accommodate appropriate statistical analyses.

8.4.1. Spontaneous Activity Open Field (Non-automated)

Perhaps the simplest and most economical method for assessing both exploratory and locomotor activity is the open-field apparatus. As the name suggests, this generally consists of a square (or circular, depending on personal preference) arena of adequate size (e.g., 50 × 50 cm or 50 cm diameter for mice) surrounded by walls to prevent the animal from escaping. The box itself may be composed of either wood or plastic, although the latter is preferred to reduce olfactory issues and for ease in cleaning. In its simplest form, the floor is divided into equally spaced regions by marker pen which has been allowed sufficient time to dry so as not to produce unwanted olfactory effects. Alternatively, the box can be video monitored and lines can be drawn on the screen to divide the arena. Typical Protocol

  1. The open field should be located in a quiet room with controlled temperature and ventilation. A low-level illumination is preferred to reduce anxiety and thus lessen freezing behaviors, unless this is a component of the task that you wish to study. The observer should be seated comfortably at a distance from the apparatus, or ideally watching a monitor fed by a video camera positioned above the open field. If visible to the behaving rodent, the investigator should be consistent with seating position, clothing, and potential olfactory cues.
  2. If stimulant activity of a drug is to be examined, the rodent should first be habituated to the apparatus for three or more 5-min sessions to reduce baseline activity. Habituation should be omitted if anxiety or response to novelty is being studied. For brain lesion or injury studies, you may wish to examine performance at discreet times before and after surgery. Bear in mind, however, that in general, activity and/or exploration following repeated exposure to the open field will decrease with habituation and cognition.
  3. On the test day, administer the test drug, if required, at an appropriate time prior to placing the rodent into the center of the open field. The investigator records the following specific behaviors using prepared data sheets and appropriate counters over a specified period of time, usually 5–10 min.

Parameters to record: locomotion (number of square crossings within the specified time); rearing; grooming; and stereotypical behaviors such as licking, biting, and head weaving. These activities may be recorded separately for peripheral regions or in the center of the arena, the latter being thought to reflect the degree of anxiety experienced by the rodent (i.e., animals with higher levels of activity in the center of the arena are less anxious). Defecation frequency may also be recorded as a measure of fear, but this tends to be more variable due to the relatively small numbers involved. Open Field (Automated)

For large studies it is impractical to directly observe each animal individually. This can be resolved by making use of a set of automated activity boxes consisting of arenas similar to those described above, but with regions demarcated by infrared beams instead of marker pen. Each box is connected individually to a computer that collates all data from up to 30 or more boxes at a time. Equipment available from AccuScan Instruments (formerly Omnitech, Columbus, Ohio, USA; offers flexibility in experimental design and data analysis. Although the cost can be somewhat prohibitive for smaller laboratories (upwards of $80,000 for a set of 16 boxes), the major advantage of such systems is that they allow the collection of both vertical (rearing) and horizontal activity from the periphery and center of the apparatus over time periods that could not be accurately completed by a manual observer. For example, the computer can be programmed to accept data every 15 sec and to calculate a mean for each 1 min time bin for up 2 hr or more. Clearly the amount of data collected rises considerably with increased time. In our laboratories, a system of 16 boxes is employed (AccuScan Instruments) in a dedicated quiet room with dimmed lighting. Each arena is 40 × 40 cm in size with removable clear Plexiglas chambers for ease of cleaning. Two sets of infrared photocells (one for detecting rearing, the other for locomotion), are fixed to a rack that surrounds the Plexiglas and that can be adjusted in the vertical plane to allow measurements from rats or mice. Typical Protocol

  1. If a stimulant drug-induced increase in activity is expected, habituate the mice to the apparatus for several 1-hr sessions. Do not habituate for drugs expected to decrease activity.
  2. Program the computer to record activity as desired. An example would be to bin data every 5 min for up to 1 hr and to distinguish horizontal activity from vertical in both the peripheral and central regions of each arena.
  3. Administer test drug at the appropriate time. If decreased activity is expected, inject before placing mice into the center of the arena. Conversely, if increased activity is expected, place mice into the center of the arenas and allow for habituation to the novel environment for at least 30 min before drug administration. If animals are subjected to brain injury or other surgery, allow sufficient time for recovery (at least 24 hr, if not more, depending on severity) before placement into the arenas.
  4. Record data for a predetermined period of time, usually 30–120 min. Print out all raw data as a hard copy backup and convert the data file produced into a form suitable for analysis using software programs such as Microsoft Excel (Microsoft Corporation, Seattle, Washington, USA; and JUMP (SAS Institute, Cary, North Carolina, USA;

Arenas employed by more advanced systems (AccuScan Instruments; can be subdivided into zones both physically by using a Plexiglas insert and virtually by specifying different software parameters. This allows many additional animals to be assessed simultaneously. In addition, this system can allow measurements of distance traveled, movement time, rearing duration, etc. If the additional cost of such systems is prohibitive, but something more sophisticated than the manual version of the open field apparatus is desired, other equipment options are available. Software for video tracking systems that identify images in contrast to the background and track a center point for movement, and which have been used for tracking in water mazes, are available for multiple chambers for analysis. These can produce measures of distance and orientation over time, although rearing activity may have to be recorded separately. In addition, there are more sophisticated video systems that track the directionality of the animal and other behaviors such as weaving and stereotypy. Therefore, if cost is an important factor, the investigator should determine whether existing equipment could be easily modified to measure motor activity. In addition, some photocell systems have been designed to measure activity in the home cage, where response to a novel environment is undesirable, or for monitoring over the light-dark cycle. Such systems tend to come in specially designed cages, and racks and can be expensive. However, there is now a very competitive market for analyzing locomotor activity, and this has led to a reduction in the price of systems, allowing a small lab to tailor equipment according to need. A list of vendors is included at the end of the chapter. Variations

Locomotor activity can also provide indices of learning and memory and anxiety. Habituation of locomotor activity in a novel environment can be used to assess memory in mice [36]. For this procedure, the mouse is briefly exposed (e.g., 5 min) to a novel open field and locomotor activity is assessed. Memory for the novel experience is then tested at a later time by reexposing the mouse to the same open field. Activity during the second exposure is used as an index for assessing memory, with lower activity being indicative of better memory for the open field. Of course, it is important that the treatments evaluated with this method do not have direct effects on locomotor behavior. One way to minimize the effects of the drug is to treat immediately following the first activity session.

The pattern of exploration can also be an important index of anxiety. Informal assessment of anxiety can be derived by comparing time spent in the periphery of the arena relative to time spent in the center. Anxious animals tend to spend more time in the periphery. In addition, initial freezing in an open field is an index of anxiety, so the latency to move a given distance (or to move through a given number of squares) can also be used to assess fear and anxiety. A more formal assessment of anxiety can be made using a modified open-field apparatus. The open field is separated into a well-lit area and a dark area and the relative time and activity in these two zones is compared. Anxiolytics increase the time spent in the well-lit zone in this light–dark test [37]. Sample Experiment

Data from a typical automated experiment are presented in Figure 8.11. In this study, vertical (or rearing) (Figure 8.11A) and horizontal (Figure 8.11B) activity was assessed for three mouse strains, with data collected in 5-min bins for 30 min [38]. Note that BALB/c mice appear significantly less active than animals from the other strains. Note also the habituation response indicated by decreased activity over time for most groups. It is important that appropriate statistical methods be used for analysis of behavioral data. Locomotor data are generally normally distributed so parametric ANOVAs are used routinely. A repeated measures ANOVA should be considered in most cases when a time course is employed. Post-hoc tests that examine the mean square error relative to the overall analysis (e.g., Tukey’s) can then be used for multiple comparisons between groups. Individual t-tests should not be used unless they are corrected for multiple comparisons. For the data presented in Figure 8.11B, a repeated measures ANOVA yielded a significant group effect [F(2,26) = 4.382, p < 0.0229], indicating overall differences between the strains in the study; time effect [F(5,130) = 126.103, p < 0.0001], reflecting the decreased activity with time (habituation) overall; and group × time interaction [F(10,130) = 3.049, p < 0.0017], indicating significant differences between groups over time. Post-hoc analysis with Tukey’s pairwise comparisons detected significant differences for the 5-, 10-, 15-, and 30-min time points between C57BL/6 and BALB/c mice (p < 0.01).

FIGURE 8.11. Automated measure of vertical (a) and horizontal (b) activity in the same animals from three different mouse strains.


Automated measure of vertical (a) and horizontal (b) activity in the same animals from three different mouse strains. Data were collected over 5-min intervals for a total of 30 min. Most mice habituated to the test environment, as evidenced by the decline (more...) Motor Function

Motor function can be differentially affected depending on experimental parameters. For example, unilateral brain injury models often produce hemiparesis-like effects, which may be reflected by deficits in grip strength, balance, and turning behavior, or may induce forepaw flexion. Many drugs can have either sedative or stimulant properties. Consequently, several models have been developed to examine specific motor deficits such as these. Two commonly used procedures are thus described.

8.4.2. Rotarod

The ability of a rodent to maintain balance and keep pace with a rotating rod has been used with varying degrees of success over the years to assess motor function. Several versions of this test (commonly referred to as the rotarod test) have been described over the years. Most require the mouse to walk on a rotating rod of fixed diameter (3.5 cm for the apparatus we use) that increases in speed over a predetermined period of time until the animal can no longer maintain its position. The rotarod apparatus employed in our laboratories consists of a central drive rod connected to a stepper motor (AccuScan Instruments) that is divided into four separate testing stations. The speed at which the rod rotates can be accelerated up from 0 rpm to over 100 rpm over a set time period. Other rotarod models are available and can be found in the vendor list at the end of the chapter. Typical Protocol

  1. Administer drug, as appropriate. For lesioned or injured animals, wait at least 24 hr following the surgical procedure.
  2. Set the apparatus to accelerate from 0–40 rpm over 60 sec. This is a good standard for young adult mice, although it should be noted that juvenile and older animals perform poorly at this task.
  3. Place four mice on the rotarod, one per testing station, and then start the stepper motor and timer. Many models come equipped with a timer that begins automatically when the motor is switched on and stops when the animal falls down to the floor of the apparatus, as detected by interruption of an infrared beam.
  4. As the speed increases, the mouse is required to walk faster to remain in a stationary position. The latency to fall from the rotating rod is determined and taken as a measure of motor function. It is generally a good idea to take the mean of at least two to three measures from each animal. Variation

Some investigators [39,40] modify the rod itself by enclosing the core of the rod with a series of stainless steel bars of a specific diameter (Figure 8.12A). In this instance the time either to fall (Figure 8.12B) or to cling and make two full rotations is recorded as the outcome measure. This design may offer some advantages over the more traditional, relatively smooth rod in that data, particularly in brain injury studies, may be more consistent within groups. With rodent strains that exhibit a poor baseline performance in this task, it is usually beneficial to pre-train these animals at least two to three times before commencing the study proper.

FIGURE 8.12. The effect of moderate controlled cortical impact (CCI) brain injury on rotarod performance.


The effect of moderate controlled cortical impact (CCI) brain injury on rotarod performance. As the device gradually ramped up to speed (35 rpm), the mouse was required to walk faster to maintain a stationary position on the rod, which has been modified (more...) Sample Experiment

Data from a typical experiment are presented in Figure 8.12C, where the effect of sham surgery and controlled cortical impact (CCI) brain injury on time spent on the rotating rod is shown for three mouse strains. Sham operated controls exhibited a stable performance over the 4 wk of testing, whereas a decrease in time spent on the rotarod device was observed in injured mice from all strains for up to 7 days following injury. Because of these pronounced deficits, it would be unwise to conduct cognitive experiments with a significant motor component (e.g., Morris water maze) during this time period. Once again a repeated measures ANOVA is appropriate for comparing groups over time as rotarod data tends to be normally distributed and this test was conducted repeatedly over a 4-wk period in this study. A significant group effect [F(5,57) = 16.601, p < 0.0001], indicating overall differences between the strains in the study; time effect [F(7,399) = 47.183, p < 0.0001], reflecting the attenuation of the deficits with time in the injured groups; and group × time interaction [F(35,399) = 6.480, p < 0.0001], indicating significant differences between groups over time, were observed. Using a post-hoc test (Tukey’s pairwise comparison), a significant impairment was detected among CCI injured mice from all three strains when compared with their respective surgery controls on days 1, 2, and 3 (p < 0.05) following surgery. A one-way ANOVA would be suitable for comparing these groups if no time component was involved. A t-test may also be appropriate in such instances. No significant difference was observed between strains for either treatment group.

8.4.3. Beam Balance/Walking

While the rotarod is useful for determining gross motor deficits in the rodent, the detection of more subtle motor effects requires a different approach. Fine motor coordination, for example, can be assessed using a beam walking or balance task. This test essentially examines the ability of the animal to remain upright and to walk on an elevated and relatively narrow beam (Figure 8.13A) without falling to the cushioned pads below or slipping to one side of the beam. Again, unilateral brain injury models tend to induce a hemiparesis-like effect, which can cause the rodent to slip to one side, usually contralateral to the injury site (Figure 8.13B).

FIGURE 8.13. The effect of moderate controlled cortical impact (CCI) brain injury on beam walking performance for the C57BL/6 mouse.


The effect of moderate controlled cortical impact (CCI) brain injury on beam walking performance for the C57BL/6 mouse. Surgery-naive or sham-operated mice perform well on this task, traversing the beam several times, gripping its horizontal edge with (more...) Typical Protocol

  1. For mice, set up a beam approximately 0.6 cm wide and 120 cm in length, suspended about 60 cm above some foam pads. (A larger beam, approximately 1.8 cm wide and 240 cm in length, in addition to a flat platform at one end to rest between trials is required for rats.)
  2. Place the animal on one end of the beam (for the rat this would be farthest from the platform). Animals from active strains such as the C57BL/6 mouse or the Long-Evans rat will instinctively walk along the beam to reach the opposite end. Once at this point they will generally turn 180° and continue to walk on to the opposite end. Establish a basal level of performance before surgery or treatment, and allow sufficient time for recovery (at least 24 hr) before retesting.
  3. Count the number of foot faults, defined as the number of times the fore-paws and/or hindpaws slip from the horizontal surface of the beam over a predetermined number of steps (50 is usually adequate). Allow the performing animal sufficient time (approximately 5 min) to complete this task. (It is useful to use a mirror on the side of the beam opposite the observer and to videotape the performance for scoring.)
  4. Remove to home cage and retest as appropriate. It should be noted that rodents, especially rats, tire and are reluctant to move if exposed to this test repeatedly over a short period on the same day. Variation

This task works well for active rodent strains and may not be suitable for less active animals. Another variation partly designed to address this issue in the rat involves training animals to walk across the beam to a “safe” dark box; the cognitive requirements for this version, however, may influence motor outcome to some degree so care should be taken here. A simpler approach measures the time taken to fall down onto the foam pads. In this instance, the investigator should vary the beam width until an acceptable latency is found for the particular strain to be used. Attention should also be paid to the body weight of the animal, as the suitable width of the beam may change according to the mouse’s ability to grip the edge of the beam, for example, mice heavier than 35 g generally require a beam approximately 0.9-cm thick. Example Experiment

Data from a typical experiment are reproduced in Figure 8.13C. In this experiment, adult C57BL/6 mice were subjected to mild (4.5 m/s) or moderate (6.0 m/s) unilateral CCI brain injury, and the number of contralateral hind limb foot faults were recorded over a 4-wk period. An obvious deficit, dependent on injury severity, was observed when compared with sham-operated controls. For statistical analysis, beam-walking data are generally normally distributed so parametric ANOVAs are advised. A repeated measures ANOVA should be considered in most cases when a time course such as that presented above is employed. Post-hoc tests that examine the mean square error relative to the overall analysis (e.g., Tukey’s) can then be used for multiple comparisons between groups. Individual t-tests should not be used. For the data presented in Figure 8.13C, a repeated measures ANOVA yielded a significant group effect [F(2,33) = 94.265, p < 0.0001], indicating overall differences between the different treatment groups in the study; time effect [F(7,231) = 89.383, p < 0.0001], indicating significant overall changes in performance over the duration of the study; and group × day interaction [F(14,231) = 20.995, p < 0.0001], indicating significant performance differences between groups over time. Post-hoc analysis with Tukey’s pairwise comparisons detected significant differences for days 1–28 between sham controls and CCI-injured mice from both groups (p < 0.001). There were no significant differences between groups before injury (day 0; p > 0.05).


Davis M. Neurochemical modulation of sensory-motor reactivity: Acoustic and tactile startle reflexes. Neurosci. Biobehav. Rev. 1980;4(2):241–263. [PubMed: 6106916]
Davis M. The mamalian startle response. In: Eaton RC, editor. Neural mechanisms of startle behavior. New York: Plenum Press; 1984. pp. 287–351.
Koch M, Schnitzler HU. The acoustic startle response in rats—circuits mediating evocation, inhibition and potentiation. Behav Brain Res. 1997;89:1–2. 35–49. [PubMed: 9475613]
Braff DL, et al. Impact of prepulse characteristics on the detection of sensorimotor gating deficits in schizophrenia. Schizophr Res. 2001;49:1–2. 71–78. [PubMed: 11343875]
Swerdlow NR. Cortico-striatal substrates of cognitive, motor and sensory gating: Speculations and implications for psychological function and dysfunction. In: Panksepp J, editor. Advances in biological psychiatry. 2. Greenwich, CT: JAI Press; 1996.
Alder LE, et al. Neurophysiological evidence for a defect in neuronal mechanisms involved in sensory gating in schizophrenia. Biological Psychiatry. 1982;17(6):639–654. [PubMed: 7104417]
Boutros NN, Zouridakis G, Overall J. Replication and extension of P50 findings in schizophrenia. Clinical EEG (electroencephalography) 1991;22(1):40–45. [PubMed: 1991411]
Freedman R, Alder LE, Myles-Worsley M, et al. Inhibitory gating of an evoked response to repeated auditory stimuli in schizophrenic and normal subjects: Human recordings, computer simulation, and an animal model. Archives of General Psychiatry. 1996;53(12):1114–1121. [PubMed: 8956677]
Freedman R, Alder LE, Waldo MC, Pachtman E, Franks RD. Neurophysiological evidence for a defect in inhibitory pathways in schizophrenia: Comparison of medicated and drug-free patients. Biological Psychiatry. 1983;18(5):537–551. [PubMed: 6134559]
Bickford PC, Luntz LV, Freedman R. Auditory sensory gating in the rat hippocampus: Modulation by brainstem activity. Brain Research. 1993;607(1–2):33–38. [PubMed: 8481808]
Adler LE, Rose G, Freedman R. Neurophysiological studies of sensory gating in rats: Effects of amphetamine, phencyclidine, and haloperidol. Biological Psychiatry. 1986;21(8–9):787–798. [PubMed: 3730461]
Light GA, Malaspina D, Geyer MA, et al. Amphetamine disrupts P50 suppression in normal subjects. Biological Psychiatry. 1999;46(7):990–996. [PubMed: 10509182]
Stevens KE, Freedman R, Collins AC, et al. Genetic correlation of inhibitory gating of hippocampal auditory evoked response and alpha-bungarotoxin-binding nicotinic cholinergic receptors in inbred mouse strains. Neuropsychopharmacology. 1996;15(2):152–162. [PubMed: 8840351]
Light GA, Geyer MA, Clementz BA, Cadenhead KS, Bratt DL. Normal P50 suppression in schizophrenia patients treated with atypical antipsychotic medications. The American Journal of Psychiatry. 2000;157(5):767–771. [PubMed: 10784470]
Simosky JK, Stevens KE, Freedman R. Nicotinic agonists and psychosis: Current drug targets. CNS and Neurological Disorders. 2002;1(2):149–162. [PubMed: 12769624]
Hunter KP, Willott JF. Effects of bilateral lesions of auditory cortex in mice on the acoustic startle response. Physiology & Behavior. 1993;54(6):1133–1139. [PubMed: 8295953]
Weiss GT, Davis M. Automated system for acquisition and reduction of startle response data. Pharmacol. Biochem. Behav. 1976;4(6):713–720. [PubMed: 981288]
Crawley JN, et al. Behavioral phenotypes of inbred mouse strains: Implications and recommendations for molecular studies. Psychopharmacology (Berl.) 1997;132(2):107–124. [PubMed: 9266608]
Olivier B, et al. The DBA/2J strain and prepulse inhibition of startle: A model system to test antipsychotics? Psychopharmacology (Berl.) 2001;156(2–3):284–290. [PubMed: 11549230]
Zhang M, et al. Effect of dopamine D3 antagonists on PPI in DBA/2J mice or PPI deficit induced by neonatal ventral hippocampal lesions in rats. Neuropsychopharmacology. 2006;31(7):1382–1392. [PubMed: 16395300]
Geyer MA, Braff DL. Startle habituation and sensorimotor gating in schizophrenia and related animal models. Schizophr. Bull. 1987;13(4):643–668. [PubMed: 3438708]
Geyer MA, et al. Startle response models of sensorimotor gating and habituation deficits in schizophrenia. Brain Res. Bull. 1990;25(3):48598. [PubMed: 2292046]
Paylor R, Crawley JN. Inbred strain differences in prepulse inhibition of the mouse startle response. Psychopharmacology (Berl.) 1997;132(2):169–180. [PubMed: 9266614]
Adler LE, Waldo MC, Freedman R. Neurophysiologic studies of sensory gating in schizophrenia: Comparison of auditory and visual responses. Biological Psychiatry. 1985;20(12):1284–1296. [PubMed: 4063418]
Jin Y, Potkin SG. P50 changes with visual interference in normal subjects: A sensory distraction model for schizophrenia. Clinical EEG (electroencephalography) 1996;27(3):151–154. [PubMed: 8828978]
Lebib R, Papo D, de Bodes S, Bandonnière PM. Evidence of a visual-to-auditory cross-modal sensory gating phenomenon as reflected by the human P50 event-related brain potential modulation. Neuroscience Letters. 2003;341(3):185–188. [PubMed: 12697279]
Moxon KA, Gerhart GA, Brickford PC, et al. Multiple single units and population responses during inhibitory gating of hippocampal auditory response in freely-moving rats. Brain Research. 1999;825(1–2):75–85. [PubMed: 10216175]
Moxon KA, Gerhart GA, Gwinetto M, Alder LE. Inhibitory control of sensory gating in a computer model of the CA3 region of the hippocampus. Biological Cybernetics. 2003;88(4):247. [PMC free article: PMC4170679] [PubMed: 12690484]
Ellenbroek BA. Pre-attentive processing and schizophrenia: Animal studies. Psychopharmacology. 2004;174(1):65–74. [PubMed: 15205880]
Miyazato H, et al. A middle-latency auditory-evoked potential in the rat. Brain Research Bulletin. 1995;37(3):247–255. [PubMed: 7627567]
Martin LF, Freedman R, Anissa AbiDargham G, Olivier International review of neurobiology. Vol. 78. 2007. Schizophrenia and the 7 Nicotinic Acetylcholine Receptor; pp. 225–246. [PubMed: 17349863]
Alder LE, et al. Normalization of auditory physiology by cigarette smoking in schizophrenic patients. American Journal of Psychiatry. 1993;150:185–161. [PubMed: 8238642]
Olincy A, Harris JG, Johnson LL, et al. Proof-of-concept trial of and 7 nicotinic agonist in schizophrenia. Arch. Gen. Psychiatry. 2006;63(6):630–638. [PubMed: 16754836]
Stevens KE, Wear KD. Normalizing effects of nicotine and a novel nicotinic agonist on hippocampal auditory gating in two animal models. Pharmacology Biochemistry and Behavior. 1997;57(4):869–874. [PubMed: 9259018]
Bitner RS, Bunnelle WH, Anderson DJ, et al. Broad-spectrum efficacy across cognitive domains by alpha-7 nicotinic acetylcholine receptor agonism correlates with activation of ERK1/2 and CREB phosphorylation pathways. Journal of Neuroscience. 2007;27(39):10578–10587. [PubMed: 17898229]
Platel A, Porsolt RD. Habituation of exploratory activity in mice: A screening test for memory enhancing drugs. Psychopharmacology (Berl.) 1982;78(4):346–352. [PubMed: 6818597]
Costall B, et al. Actions of buspirone in a putative model of anxiety in the mouse. J. Pharm. Pharmacol. 1988;40(7):494–500. [PubMed: 2904990]
Fox GB, LeVasseur RA, Faden AI. Behavioral responses of C57BL/6, FVB/N, and 129/SvEMS mouse strains to traumatic brain injury: Implications for gene targeting approaches to neurotrauma. J. Neurotrauma. 1999;16(5):377–389. [PubMed: 10369558]
Fox GB, et al. Sustained sensory/motor and cognitive deficits with neuronal apoptosis following controlled cortical impact brain injury in the mouse. J. Neurotrauma. 1998;15(8):599–614. [PubMed: 9726259]
Hamm RJ, et al. The rotarod test: An evaluation of its effectiveness in assessing motor deficits following traumatic brain injury. J. Neurotrauma. 1994;11(2):187–196. [PubMed: 7932797]


AccuScan Instruments, Inc., 5098 Trabue Road, Columbus, OH 43228 USA, Tel: 614-878-6644; 800-822-1344, (USA/Canada), Fax: 866-650-8265, E-mail: moc.asu-nacsucca@selas

Coulbourn Instruments, 7462 Penn Drive, Allentown, PA 18106 USA, Tel: 610-395-3771, E-mail: moc.nruobluoc@selas

Clever Sys., Inc., 11425 Isaac Newton Square, Suite 202, Reston, VA 20190 USA, Tel: 703-787-6946, Fax: 703-757-7467,

Columbus Instruments, 950 N. Hague Avenue, Columbus, OH 43204 USA, Tel: 614-276-0861; 800-669-5011, Fax: 614-276-0529, Email: moc.tsniloc@selas

Kinder Scientific, Michael Kinder, President, 12655 Danielson Court, Suite 308, Poway, CA 92064 USA, Tel: 858-679-1515, Fax: 858-679-4811, E-mail: moc.redniknotlimah@rednikm

Noldus Information Technology, Inc., 1503 Edwards Ferry Road, Suite 201, Leesburg, VA 20176 USA, Tel: 703-771-0440; 800-355-9541, Fax: 703-771-0441, E-mail: moc.sudlon@ofni

Source: Data reproduced with permission from Fox, G. B., LeVasseur, R. A., and Faden, A. I. 1999. Behavioral responses of C57BL/6, FVB/N, and 129/SvEMS mouse strains to traumatic brain injury: Implications for gene targeting approaches to neurotrauma. J. Neurotrauma 16(5): 377–89.

Copyright © 2009, Taylor & Francis Group, LLC.
Bookshelf ID: NBK5236PMID: 21204341


  • PubReader
  • Print View
  • Cite this Page

Other titles in this collection

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to PubMed

Similar articles in PubMed

See reviews...See all...

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...