Methods in Behavioral Pharmacology

Visit the link below to download the full version of this book:

https://medipdf.com/product/methods-in-behavioral-pharmacology/

Click Download Now

vin
MA. Ellenberger Division of Laboratory Animal Medicine, Tufts-New England
Medical Center, Boston, MA 02111, U.S.A.
ΒΛ. Ellenbroek Psychoneuropharmacology Research Unit, University of Nij-
megen, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands
J.L. Folk Department of Psychology, Rutgers State University, Psychology Build­
ing Room 332, Busch Campus, New Brunswick, NJ 08903, U.S.A.
M. Galizio Department of Psychology, University of North Carolina, Wilmington,
NC 28403-3297, U.S.A.
GB. Glavin Department of Pharmacology and Therapeutics and Department of
Surgery, Faculty of Medicine, University of Manitoba, Winnipeg, Manitoba,
Canada R3E OW3
F. van Haaren Department of Psychology, University of Florida, Gainesville, FL
32611, U.S.A.
5.7! Higgins Human Behavioral Pharmacology Laboratory, Department of Psy­
chiatry, The University of Vermont, Burlington, VT 05401-1419, U.S.A.
J.R. Hughes Human Behavioral Pharmacology Laboratory, Department of Psychi­
atry, The University of Vermont, Burlington, VT 05401-1419, U.S.A.
A.E. Kelley Department of Psychology, Northeastern University, 125 Nightingale
Hall, Boston, MA 02115, U.S.A.
E.D. Kemble Division of Social Sciences, University of Minnesota, Morris, MN
56267, U.S.A.
/. Krauth Institute of Psychology, University of Düsseldorf, Universitätsstrasse l,
D-4000 Düsseldorf, Germany
K Larsson Department of Psychology, University of Göteborg, POB 14158, S-400
20 Göteborg, Sweden
G.A. Lemaire Minneapolis, MN 55410, U.S.A.
A. Lenègre I.T.E.M.-LABO, 93, Avenue de Fontainebleau, 94276 Le Kremlin-
Bicêtre Cedex, France
MJ. Lewis Department of Psychology, Howard University, Washington, DC 20059,
U.S.A.
G van Luijtelaar Department of Psychology, University of Nijmegen, P.O. Box
9104, 6500 HB Nijmegen, The Netherlands
R.A. McArthur I.T.E.M.-LABO, 93, Avenue de Fontainebleau, 94276 Le
Kremlin-Bicêtre Cedex, France
A.L. Markowska Department of Psychology, The Johns Hopkins University, Balti­
more, MD 21218, U.S.A.
R.A. Meisch Department of Psychiatry and Behavioral Sciences, University of
Texas Health Science Center at Houston, TX 77030-3497, U.S.A.
D.S. Olton Department of Psychology, The Johns Hopkins University, Baltimore,
MD 21218, U.S.A.
W.P. Paré Eastern Research and Development Office, Veterans Administration
Medical Center, Perry Point, MD 21902, U.S.A.
M.J. Picker Department of Psychology, University of North Carolina at Chapel
Hill, Chapel Hill, NC 27599-3270, U.S.A.
R.C. Pitts Department of Psychology, University of North Carolina at Chapel Hill,
Chapel Hill, NC 27599-3270, U.S.A.
 IX

A. Poling Department of Psychology, Western Michigan University, Kalamazoo,

MI 49008, U.S.A.
MJ. Pontecorvo Janssen Research Foundation, 1125 Trenton-Harbourton Road,
Titusville, NJ 08560-0200, U.S.A.
R.D. Porsolt I.T.E.M.-LABO, 93, Avenue de Fontainebleau, 94276 Le Kremlin-
Bicêtre Cedex, France
N.E. Rowland Department of Psychology, University of Florida, Gainesville, FL
32611, U.S.A.
C.W. Schindler Behavioral Pharmacology and Genetics Laboratory, NIDA Addic­
tion Research Center, P.O. Box 5180, Baltimore, MD 21224, U.S.A.
S. Stevens Negus Department of Pharmacology, University of Michigan, Ann
Arbor, MI 48103, U.S.A.
LP. Stolerman Department of Psychiatry, Institute of Psychiatry, De Crespigny
Park, London SE5 8AF, England
Methods in behavioral pharmacology 1
F. van Haaren (Ed.)
©1993 Elsevier Science Publishers B. V. All rights reserved.

CHAPTER 1

The use of animal models in behavioral

pharmacology

MARY A. ELLENBERGER

Division of Laboratory Animal Medicine, Tufts-New England Medical Center, Boston, MA 02111, U.S.A.

1. Selection of an animal model

1.1. General considerations

Animal models are used throughout the biological sciences to provide complex
systems for the study of complex problems. While animal studies are often con­
ducted in lieu of or prior to human studies, in other instances an animal species is
itself the definitive subject of basic research. The use of animals in scientific
study should be accompanied by careful planning and a high level of conscien­
tiousness. The ethical and legal criteria that must be met prior to conducting
research on animals, surpass those placed upon any other use of animals in our
society.
The process of animal model selection is one of evaluation and compromise. For
example, mice are inexpensive and easy to maintain, but their usefulness can be
limited due to their small size and poor trainability. Dogs are much larger and more
trainable, but substantial expenses will likely be incurred for dog purchase and care.
The ability of an animal model to meet the scientific criteria needed for successful
studies is a crucial factor. Review of the scientific literature in the selected field of
study is essential to understanding why and how particular animal models have been
used to meet experimental needs. Characteristics of the species' genetics, behavior,
anatomy, physiology, or other traits that have been the subject of scrutiny will be
presented for evaluation by the reader. Biological characteristics of the animal model
that are omitted from discussion in the literature likewise bear due consideration.
A paper which conveys the suitability of the mouse as a model for feeding behavior
is not likely to point out that the animal's small size precludes certain procedures.
Collection of frequent blood samples from individual animals, for example, would
be impossible due to the small total blood volume of a mouse. Any proposed altera-
2

tion of procedures described in a published animal model merits careful consider­

ation. The scientific literature focuses on how well a particular animal model fits the
technical demands of a study and rarely provides any indication of difficulties or
failures with the model. Despite this short-coming, the historical successes of par­
ticular animal models as recounted in the literature are extremely important
guidelines to animal model selection. The scientific compatibility of the animal
model with the study to be performed is of chief importance in designing a successful
experiment.
By their nature, the practical considerations of animal model selection are rarely
topics of scientific discourse. Difficulties in procuring, maintaining, and working
with a particular animal species are not conveyed in publications which emphasize
experimental results. Laboratory animal veterinarians, animal care technicians, com­
mercial animal vendors, and researchers with experience using animal models can
convey logistical concerns that are not covered in scientific reports. To a researcher
lacking practical information, the Virginia opossum (Didelphis virginiana) may seem
to be an ideal animal model for studies which require mammals with primitive
placentas. These marsupials are large enough for the proposed surgical implant of
a drug delivery device and for the frequent blood withdrawals needed to complete
a study of placentation. Consultation with a laboratory animal veterinarian would
quickly reveal that, because there are no commercial breeders of this species,
availability of pregnant opossums is limited by their abundance and their breeding
season in the wild (Jurgelski, 1987). The diet and habits of these animals result in
a significant rate of parasitism and infection with pathogenic organisms such as Sal-
monella (Runkel et al., 1991). Recently captured opossums are difficult to handle
safely and must be housed in sturdy, escape-proof enclosures. Lacking urgent scien­
tific necessity, most investigators would be well advised to avoid studies of an animal
that is so difficult to procure, risky to handle, and potentially hazardous to human
health. The scientific advantages to be gained through the use of this species would
require particular merit in order to overcome the practical disadvantages.
A researcher's consultation with persons who have already invested the time and
effort to become knowledgeable about research animals pays off not only in the
return of practical information but also in the acquisition of important scientific
information about the animal model. For example, a laboratory animal veterin­
arian's knowledge that the hamster is used as a model for antibiotic-associated
enterocolitis triggers a warning to proceed cautiously should postoperative antibiotic
administration become necessary in this species (Small, 1987). A behavioral endo-
crinologist who plans to perform surgery on hamsters to implant hormone pellets
may not be familiar with this characteristic of the species and so may propose to
routinely administer a prophylactic dose of antibiotics to each hamster at the time
of surgery. Persons involved with animal care and use programs are likely sources
for information on the biological nature of a research animal species as well as any
new developments in anesthetic agents, drug delivery systems, housing units, infec­
tious diseases, and other factors that may have direct effects on the results obtained
from experiments involving animals. Laboratory animal vendors can also help in­
vestigators fulfill scientific needs by supplying appropriate animal models. Perfor-
 3

mance of surgical procedures, such as rodent olfactory bulb ablation, by an animal

vendor's specially trained in-house technical staff is efficient and cost-effective. In­
vestigators can purchase surgically-modified animals at reasonable cost and avoid
the expenses of surgical instruments, anesthetic agents, and technician time.

1.2. Animal species

More than any other feature, the species of an animal will determine its feasibility
and desirability as a research model. Scientific attributes such as behavior, anatomy,
and physiology are inexorable characteristics of the animal species. The untrained
observer of a rat, for example, may see merely a large-scale mouse. As anyone who
has searched in vain for the non-existent rat gallbladder can attest, the two rodent
species are by no means interchangeable. The solitary, aggressive Syrian hamster
(Mesocricetus auratus) and the much smaller, docile Siberian hamster (Phodopus
sungorus) are anatomically similar but have vastly different social structures and
behaviors (Nowak and Paradiso, 1983; Gibber et al., 1984). Species differences or
similarities that may seem apparent to the casual observer do not necessarily with­
stand the more rigorous analysis that must be applied to prospective animal models.
Additional detective work is necessary for evaluation of the practicality of using a
particular species. Consideration of animal availability, costs, health status, en­
vironmental requirements, and manageability may have profound impact on the
feasibility of a proposed study. If a model employing either rats or mice fits the scien­
tific criteria of a study, the choice of species might be influenced by the knowledge
that there is a critical shortage of mouse housing space at the research institute but
plenty of room for rats. Conversely, a research team member's severe allergy to rats
could sway the decision towards the selection of mice (Platts-Mills et al., 1987). Eval­
uation of both the technical and operational facets of using an animal species is fun­
damental to testing this first defining characteristic of the animal model.

1.3. Animal genetics

The genetic makeup of the animal selected for study is the next consideration once
the appropriateness of an animal species has been determined. Anatomical features,
the likelihood of congenital diseases, and certain aspects of behavior among
members of Canis familiaris can be predicted based on knowledge of the dog's gene­
tic background. The genetics of an animal can determine not only the outcome of
an experiment but also the number of animals needed to achieve reliable results
(Mann et al., 1991). One investigator has estimated that a given level of statistical
precision obtained with outbred mice can be achieved using less than half as many
inbred mice (Festing, 1987). Outbred animals are genetically variable, whereas
members of an inbred strain are virtually identical due to isogenicity at more than
99% of their genetic loci. The substrain designations for inbred strains maintained
at different breeding facilities are necessary due to genetic differences which may
arise from mutation, residual heterozygosity, or contamination (Greenhouse, 1984).
Most large suppliers of laboratory rodents have genetic monitoring programs for
4

their inbred animals. Inbred strains are now candidates for indefinite stability thanks
to frozen embryo technology (Bleby, 1987). Outbred stocks are likely to change
significantly over time due to selection, random drift, inbreeding, mutation, and
genetic contamination (Festing, 1987). Inbred strains of most rodent species are
available commercially, and some larger inbred species may be obtained from
research colonies (Greenhouse and Cohen 1981; Research Resources Information
Center, 1988). Although the cost of inbred animals exceeds that of outbreds, the
reduction in animal numbers needed and improved precision of data obtained from
inbred strains are important research considerations.

1.4. Health status

The health of animals used in research affects the outcome and reproducibility of
experimental results. Animals may be born with congenital defects or may acquire
infectious, nutritional, or degenerative diseases. The physical and social environment
of an animal likewise affects its overall condition. In a laboratory setting, an
animal's condition is largely dependent upon human knowledge and activities. Infor­
mation on the infectious diseases and dietary requirements of rats and mice is abun­
dant, but this information is not always used to greatest benefit. There have been
numerous instances in which rodents purchased in specific-pathogen-free condition
from an animal vendor are exposed to endemic diseases upon arrival at the research
facility. The initiation of experiments would often overlap with the onset of disease
in the naive animals. Optimal health for some unusual animal species cannot be en­
sured due to a lack of knowledge about its health requirements. Suboptimal animal
health can negate the validity of research results despite the most exacting experi­
mental procedures and rigorous data collection.

2. Conducting controlled animal experiments

Several avenues of entry for experimental error must be considered if animal

research is to be accurate and reproducible. A recognized source of experimental
error, such as diagnosis of a viral infection that suppressed animal performance for
2 weeks, may be dealt with by re-analyzing the affected data. An undetected source
of animal data 'noise', such as the stress induced by a weekend spate of jackhammer-
ing above the animal room, may forever perplex the investigator (Nayfield and
Besch, 1981; Gamble, 1982). The data from an uncontrolled situation such as this
would best be disregarded. In the absence of clear warnings, however, many resear­
chers have pursued the enticing but never-to-be duplicated results of experiments
gone awry.

2.1. Animal stress

Even before birth, an animal is under the influence of a multitude of factors that will
affect its performance in a research project. Stresses caused by an animal's social,
 5

nutritional, environmental, or health condition may be fleeting or could have a

lasting impact. Transportation from the vendor to the research facility is a predic­
table source of significant stress for animals (Landi et al., 1982; Aguila et al., 1988).
The environment, caging, food, and water at their destination are unfamiliar.
Animals experience conflict during establishment of a new social hierarchy within
their housing group (Taylor and Moore, 1975). Circadian rhythms may be disrupted
during adaptation to a new light-dark cycle. Exposure to infectious agents may result
in subclinical or overt disease. Insidious disruptions of homeostasis are common at
the time of an animal's arrival in the research facility, but these perturbations are
by no means limited to a brief post-shipment period. To minimize or eliminate
animal stress during experiments, the following recommendations should be applied:

1. Be familiar with the characteristics and needs of the animals that will be
studied. Convey special needs to the animal care staff before animals arrive.
2. Acquire healthy animals from a reputable source. Ascertain the infectious
disease status of the existing populations of the species at the site where
animals will be housed. Take precautions to safeguard the health of new
arrivals.
3. Allow animals at least 2-3 days for recovery from shipment prior to use in
experiments.
4. Note the types of caging, bedding, food, and water that are provided to
animals at the research facility. Be aware of the procedures and schedule of
the animal maintenance services provided by animal care personnel.
5. Avoid unnecessary disruptions of the animals' social setting by maintaining
stable housing groups.
6. Understand the environmental parameters (temperature, humidity, light,
noise) of the animal housing room. Request notification if parameters exceed
desired ranges.
7. Respect the animal health assessments and technical advice provided by the
animal care professional staff.

2.2. Environment

Standards for animal caging, space requirements, room ventilation, lighting,

temperature, sanitation, and many other aspects of the animal environment are well-
established (National Research Council, 1985). Most research institutions support
centralized animal facilities to fulfill the requirements for appropriate animal hous­
ing and care. Inattention to these aspects of animal care is a poor reflection on the
institution and on its research programs.
Seemingly insignificant materials and conditions in the animal's environment will
affect experimental results. Volatile components contained in some softwood bed­
dings commonly used for laboratory animals are known to induce increased levels
of drug-metabolizing enzymes in the liver (Weichbrod et al., 1988). Dietary formula­
tions and impurities affect animal physiology, as is noted when a diet containing
estrogenic plant residues alters the estrous cycles of mice (Thigpen et al., 1987).
6

Water supplied to laboratory animals is likewise subject to microbial and chemical

contamination. Faulty operation or improper location of animal facility air supply
intake ducts may result in circulation of air laden with diesel truck exhaust fumes.
Ambient temperature, humidity, light, and noise have obvious impacts on animal
physiology and behavior (Besch, 1990). Cessation of breeding is often observed when
these parameters are disrupted in rodent housing rooms. Environmental distur­
bances are equally disruptive to studies of non-breeding research animals.

2.3. Animal identification

The ability to identify individual animals is essential to accurate experimentation in

most areas of research. Factors to be considered in selecting an animal identification
system are the uniqueness and permanence of the individual label, its appropriate­
ness for the animal species, the skill level required to apply and decode the identify­
ing mark, and the cost of using the system. Reliance upon identifying cards placed
on an animal's cage is a traditional but risky method. Data on a cage card can easily
be separated from an animal. Rodents, rabbits, and non-human primates will con­
sume the evidence of their identity if a misplaced card comes within their reach.
Codes written on skin or fur with a waterproof marker typically last for only a day
or two. An intradermal ink tattoo gives a unique, and if properly applied, permanent
identity to an animal. Ear tags are useful in some species of animals. Many varieties
of plastic and metal ear tags in an array of sizes are sold commercially. Some animal
vendors will identify animals by tag or tattoo at a nominal cost prior to shipment.
Subcutaneous implantation of a microchip provides a unique and permanent means
of identification with the capacity for computer integration. Due to its small size,
the microchip can be implanted in a matter of seconds using a modified hypodermic
needle. Descriptive or photographic recording of an animal's natural markings may
be useful for field studies and in some non-albino laboratory species. With the excep­
tion of the ear notch system used for mice, identification methods which involve
removal of animal tissue are generally disfavored. Animal identification should
always be confirmed by the best available method before any procedures are per­
formed.

2.4. Procedures

Consistency in the timing and performance of experimental procedures involving

animals is crucial to the collection of valid data. Blood collection by three laboratory
technicians using different procedures may create three subsets of blood analysis
data (Bickhardt et al., 1983). The circadian rhythms of animals may influence experi­
mental results unless an appropriate research schedule is maintained (Hastings and
Menaker, 1976). The route used for drug administration can markedly affect the
onset, intensity, and duration of the dose (Borchard et al., 1990). Improved accuracy
and ease of drug administration and sample collection can be achieved through the
use of indwelling devices. The need for frequent drug re-dosing, for example, can be
eliminated by surgically implanting a sustained drug delivery device (Theeuwes and
 7

Yum, 1976). A similar benefit can be derived through vascular access catheters that
permit blood samples to be collected without stress to the animal (Wixson et al.,
1987; MacLeod and Shapiro, 1988). Although the implantation of these devices
requires proficiency in animal anesthesia and surgical techniques, the investment is
amply repaid in terms of more accurate and reproducible research data. Procedures
that demand a high level of technical skill warrant practice sessions conducted with
animal cadavers or anesthetized animals just prior to euthanasia.

3. Selected animal research procedures

3.1. Handling and restraint

The use of appropriate handling techniques for laboratory animals minimizes stress
and risk of injury for both the animal and its human handler. In addition to the
mechanical trauma sustained in an animal bite or scratch, the handler faces the risk
of developing a zoonotic infection (NRC, 1985). Non-human primates may harbor
the greatest array of agents capable of causing human infection, but even a bite from
a mouse poses some risk for zoonotic infection in the handler. Any human injury
caused by a laboratory animal requires medical attention from an occupational
health professional who is aware of zoonotic disease risks.
The principles of animal handling are addressed in reference books pertaining to
each animal species. Videotapes designed to provide instruction in animal handling
and procedures are readily available. Direct observation of procedures on videotape
or in person is an important prelude to animal handling. Demonstrations and hands-
on training are likely to be available from the veterinary or technical staff of the
animal facility.
The restraint of animals for injection or sample collection generally lasts only a
few seconds or minutes. A proposal involving long-term restraint of unanesthetized
animals will be closely scrutinized at the time of animal protocol review. Equipment
and techniques to avoid or minimize long-term restraint are favorable to the animals
and may improve the conduct of the study as well. These include sustained drug
delivery implants, tether systems that employ animal jackets or implanted ports,
radio telemetry, and animal training (Ruiz de Elvira and Abbott, 1986; Garner et
al., 1988; Parker and Martin, 1989; Schmidt et al., 1989; Varosi et al., 1990).

3.2. Drug administration

Some test substances are well suited to oral administration. Accurate doses can be
administered directly into the stomach via gavage in rodents. Oral (also known as
per os, abbreviated p.o.) dosing by adding drug to food or water suffers from the
disadvantage of imprecision, particularly if the drug is unpalatable to the animal.
Drugs that are irritating to tissue can often be safely administered by slow intra­
venous (i.v.) infusion. This route results in rapid distribution and onset of action of
the drug. Subcutaneous (s.c. or s.q.) administration is a simple procedure useful for
8

the injection of non-irritating drugs. A favored site for s.c. injection is under the
loose skin of the suprascapular region. Absorption from the subcutaneous space can
be slow, particularly if the drug has vasoconstrictive properties. Sites for intra­
muscular (i.m.) injection in small rodent species are limited in number and capacity.
Fairly rapid absorption results from the ample blood supply, and mildly irritant
drugs can be administered i.m. Care must be taken to avoid inadvertant placement
of the drug i.v. or in close proximity to sensitive sites. Intraperitoneal (i.p.) injections
are fairly easy to accomplish and carry low risk of visceral puncture if performed
properly. Onset of drug action administered by i.p. injection is considerably slower
than that from i.v. injection. Intracerebroventricular (i.c.v.) and intrathecal (i.t.)
routes of administration may be necessary for drugs that do not cross the blood-
brain barrier. Injection via these routes requires general anesthesia of the animal. If
multiple doses are needed, the surgical placement of a cannula is advisable (Bor-
chard et al., 1990; McCully et al., 1990; Paredes et al., 1990). All non-enteric dosing
methods, such as i.v., s.c, i.m., and i.p. injections, are defined as parenteral routes
of administration.
The effects of drug concentration and volume merit special concern in small ani­
mals. Use of either too large or too small a volume for drug administration can be
problematic. Physiologic changes due to administration of excessive fluid volume
can alter or obliterate drug effects. When a very small volume of highly concentrated
drug is injected, dosing errors become more likely due to dead space in the delivery
system and imprecise measurement.

3.3. Sample collection

Total blood volume for most laboratory animal species is within the range of 5-7%
(50-70 ml/kg) of total body weight. Removal of excessive amounts of blood from
an animal changes its physiologic state and may endanger its life (Wiggers, 1950).
Both the volume and the frequency of blood withdrawal must be considered in plan­
ning sample collections. When blood is removed, replacement of its fluid component
occurs fairly rapidly, while cellular replacement is much slower. A healthy animal
can tolerate removal of one fourth of its total blood volume as a single or cumulative
collection over 2 weeks. Blood or fluid replacement therapy may be indicated if
volumes in excess of 15 ml/kg must be collected. In small rodents, each time point
in a study may necessitate termination of one animal in order to obtain a sufficient
volume of blood. General anesthesia is required for exsanguination.
Collection of urine and feces from an animal is easily accomplished with the use
of a metabolic cage. The animal is housed on a slatted floor suspended over a waste
collection pan. Most commercially available models of metabolic cages for rodents
collect urine and solid waste separately so that drug excretion via each route may
be analyzed independently. Food consumption, water intake, and even expired gases
can be monitored in more elaborate models of metabolic cages.
The limitations and technical difficulties of cerebrospinal fluid (CSF) collection
in small animals resemble those of blood collection amplified by several orders of
magnitude. The volume of CSF that can be safely withdrawn is small, access is
 9

generally limited to one site, considerable skill is needed to perform the tap, and
anesthesia is required for the animal. Repeated sampling of CSF, like repeated dos­
ing at this site, is best accomplished by surgical placement of an intraventricular
cannula.
Some aspects of animal behavior can be measured by automated devices. Activity
monitors can measure a mouse's nocturnal running on an exercise wheel or a dog's
location within a cage (McClearn, 1982; Hughes et al., 1988). Video monitoring
coupled with computerized image analysis can be used to quantitate social interac­
tions between animals. Monitoring systems are also used extensively in the measure­
ment of learned behaviors (Deyo et al., 1989). Collection of behavioral data from
animals without artifacts caused by the presence or intervention of humans can
readily be accomplished.

3.4. Anesthesia and surgery

Classification of the two major categories of anesthetic agents used in laboratory

animals is based on the route of entry into the body. Inhalant anesthetics are volatile
compounds that enter the body through the respiratory tract with inhaled air or
oxygen mixtures. Parenteral anesthetics require injection into the body. The
difference in route of exposure between inhalant and injectable anesthetic agents has
direct implications for an animal's recovery from anesthesia. The speed with which
inhalant anesthetics take effect is nearly matched by the rate of anesthetic recovery
when exposure to the agent is curtailed. Injectable anesthetic agents continue to exert
an effect until metabolized and/or excreted.
Modern inhalant agents commonly used for general anesthesia in laboratory
animals include methoxyflurane, halothane, and isoflurane. All three form non-
explosive gases. Inhalant anesthetics are best used with a precision vaporizer that
controls the concentration of anesthetic in oxygen. The principles and equipment in­
volved in the use of a vaporizer to deliver gas anesthesia are discussed in detail in
anesthesiology textbooks (Short, 1987; Flecknell, 1988). The open-drop method for
induction of inhalant anesthesia without the use of a vaporizer is appropriate only
for rodents. Cotton balls soaked in anesthetic liquid are placed in the bottom of a
container. The rodent is placed on a raised platform above the chemical, and the
chamber is closed. As the liquid volatilizes in the ambient air of the chamber, an
anesthetic gas mixture is produced. The rodent is watched carefully and is removed
when fully anesthetized. If sustained anesthesia is needed, a cotton ball soaked in
anesthetic is pressed into a plastic holder. This nose cone can be positioned over the
rodent's face as needed to maintain anesthesia. Concentrations of oxygen and
anesthetic gas cannot be regulated in the open-drop system. Care must be taken to
prevent death due to anesthetic overdose or anoxia. Open-drop systems are
restricted to use within a chemical fume exhaust hood to minimize exposure of
personnel to anesthetic agents. Methoxyflurane is well suited to the open drop
method because its low vapor pressure precludes development of a lethal concentra­
tion of vapor (Stimpfel and Gershey, 1991). The use of ether and chloroform as
inhalant anesthetics is discouraged in light of the much improved agents described
10

above. Ether is explosive, causes irritation and secretion in mucous membranes, and
rapidly reaches lethal concentrations at room temperature. Chloroform is a potent
toxin and suspected carcinogen that is also easily overdosed.
A number of injectable anesthetic agents are used with success in laboratory
animals. Basic information on these agents and their effects on various body systems
can be obtained from anesthesiology and pharmacology texts (Booth and
McDonald, 1988; Goodman Gilman et al., 1990). Safe and effective anesthetic
regimens employing injectable drugs are also discussed in texts pertaining to the
commonly used species of laboratory animals. Barbituric acid derivatives such as
sodium pentobarbital, thiamylal sodium, and thiopental sodium are widely used in
some species despite their depressant effects on the respiratory and cardiovascular
systems. Other disadvantages of barbiturate anesthetics include relatively narrow
margins of safety and record-keeping requirements due to their classification as
controlled substances. The dissociative anesthetics ketamine and tiletamine produce
sedation, immobility, amnesia, and analgesia in most species. They have wide
margins of safety, produce minimal cardiopulmonary effects, and are not con­
trolled substances. Lack of muscle relaxation, the major deficiency of dissociative
anesthesia, is quite effectively overcome by addition of other pharmacologie agents
such as xylazine or zolazepam. There are other parenteral agents that are useful in
various species. Most of these agents are safer, longer-lasting, and/or more effective
when an animal has been pre-treated with a suitable sedative, tranquilizer, or
analgesic drug. Like all aspects of animal experimentation, the investigator's selec­
tion of an anesthetic regimen must be approved by an institutional review com­
mittee.
Surgical procedures involving animals can be categorized according to their
degree of invasiveness as either major or minor surgeries. Laparotomy, craniotomy,
thoracotomy, and orthopedic surgery are examples of major procedures. Minor
surgery generally involves only the skin or subcutaneous tissues. Further categoriza­
tion of surgery in laboratory animals is based on surgical outcome. Many research
procedures involving surgery are terminal events in which the animal is euthanized
while under anesthesia. Performance of survival surgery carries with it respon­
sibilities for postoperative monitoring and care of animals. All major survival
surgical procedures involving rabbits or other non-rodent mammalian species must
be performed using aseptic technique in a facility intended for that purpose.
Adherence to aseptic procedures is highly recommended for any survival surgery,
although special surgical facilities are not required for performance of surgery on
rodents (NRC, 1985). Postoperative care records must be maintained for rabbits and
other non-rodent mammals. Proposed surgical procedures in laboratory animals
undergo institutional review along with other components of the research protocol.

3.5. Euthanasia

The criteria used for evaluating euthanasia methods incorporate both animal and
human perceptions (AVMA, 1986). Termination of an anesthetized animal by
exsanguination or other vital tissue removal is a humane method of euthanasia. Pen-
 11

tobarbital overdose by i.p. injection in rodents or i.v. injection in other animals is

equally appropriate. Although cervical dislocation is acceptable and quite widely
used in mice, physical methods of killing animals are generally viewed as aesthetical­
ly displeasing. The proposed method of euthanasia for animals in a research project
must be specified in the protocol submitted for institutional review.

4. Species considerations

4.1. The mouse

Mice are members of the family Muridae in the order Rodentia. Numerous outbred
stocks and inbred strains of the laboratory mouse (Mus domesticus) may be purchas­
ed from research animal vendors. Coat colors and other characteristics vary con­
siderably among mouse strains. Several other species and genera of mice are also
available commercially. Mice that are free from murine infectious diseases and
parasites are produced by many vendors. The acquisition costs for healthy and
infected mice approach parity, while differences in experimental outcome and
reliability of data between the two groups are likely to be significant. With use of
appropriate housing procedures and equipment, disease-free mice may maintain
their health status despite the presence of endemic rodent infectious diseases within
the animal facility. The most commonly used type of mouse housing unit is the
polycarbonate shoebox cage. The cage bottom is covered with 1-2 cm of appro­
priate contact bedding. A wire bar lid is placed on top of the shoebox to hold a water
bottle and commercial rodent chow. A perforated plastic top with a filter that serves
as a microbiologie barrier may be placed over the assembled shoebox unit.
Automatic water delivery and active cage ventilation are available as options for the
mouse shoebox cage. An average adult laboratory mouse weighs around 25 g and
requires 97 cm 2 of cage floorspace. The average lifespan for most stocks and strains
of the laboratory mouse is under 2 years.
Mice can be picked up by mid-tail and held either manually or within a restraining
device. Manual restraint is achieved by grasping the loose skin over the mouse's neck
and shoulders. Intraperitoneal or subcutaneous injections are easily accomplished in
manually restrained mice. A restraint device that permits access to the tail veins is
recommended for intravenous injections. Intramuscular injection in the mouse is
restricted to the small volume of material that can be placed within the caudal thigh
muscles. Small amounts of urine and fecal material are usually expelled when a
mouse is manually restrained. More extensive collection of these samples can be ob­
tained by housing the mouse in a metabolic cage. Blood sampling may be performed
as a survival or non-survival procedure. The easiest method for obtaining a blood
sample employs a scalpel blade. The ventral tail artery can be nicked in a restrained
mouse to permit blood collection. This procedure does not require anesthesia and
is most effective when the mouse has been briefly exposed to a heat lamp to promote
vasodilation of the tail vessels. The site must be digitally compressed immediately
after sampling is completed to prevent hemorrhage. Terminal blood collection
12

requires that the mouse be under general anesthesia. For cardiac puncture, a 1-3-ml
syringe and 21- or 23-gauge needle are used to draw blood directly from the heart
or large vessels. Other sites of vascular access may be used to collect blood from an
anesthetized mouse during a terminal surgical procedure. Methoxyflurane is a safe,
easily administered inhalant anesthetic agent for mice. The injectable anesthetic
agents most often used in mice include pentobarbital (40-60 mg/kg i.p. after dilution
in saline) and 2% tribromoethanol solution (0.2 ml/10 g i.p.). Parenteral agents
usually provide 20-30 min of general anesthesia in the mouse and can be re-dosed
to prolong their effects. Recommended methods of euthanasia for mice include car­
bon dioxide inhalation, cervical dislocation, and i.p. overdose with pentobarbital.

4.2. The rat

The laboratory rat (Rattus norvégiens) is also a member of the family Muridae in the
order Rodentia. Rats are available for purchase in a number of inbred strains and
outbred stocks, most of which are albino. All of the commonly used rat stocks and
strains can be obtained free of specific pathogens. Like mice, rats are most comfor­
tably housed in polycarbonate shoebox cages that contain contact bedding. Water
and commercial rodent chow are supplied through a wire bar lid that, unlike the
mouse shoebox lid, has built-in clips to prevent dislodgment by the inhabitant. Cage
options like those described for the mouse shoebox are available. Adult rats range
from 300 to 500 g or more and may require up to 452 cm2 of cage floorspace. The
average lifespan of the laboratory rat is under 3 years.
A rat may be picked up near the base of its tail and can be held by hand or in
a restrainer. In order to be impenetrable to a rat bite, gloves must be so heavy as
to preclude gentle handling. Rats object to being manually restrained by the method
used for mice. Many handlers prefer to use a circumferential grip around the rat's
torso. The handler's fingers can be protected by bunching up the rat's forelegs and
loose skin into a collar around its neck. Movement of the rat's head may also be con­
trolled by placement of a finger along each side of the jawline. There are con­
siderable variations in temperament and response to handling among the stocks and
strains of rats used in research. Injection sites and sample collection procedures for
the rat are comparable to those for the mouse. Rat's skin is considerably thicker than
mouse skin, causing difficulties with i.v. injection of the tail veins. The rat's size
makes chronic catheterization for i.v. injection and serial blood collection a
reasonable option. Methoxyflurane can be administered to rats in an open-drop
system. Pentobarbital (40-50 mg/kg i.p. or 20-40 mg/kg i.v.) and xylazine/ketamine
(5 mg/kg xylazine s.c, followed in 10 min by 20-40 mg/kg ketamine i.m.) are useful
parenteral anesthetic agents. Euthanasia for rats can be accomplished with carbon
dioxide or pentobarbital overdose.

4.3. The rabbit

Rabbits and hares are members of the family Leporidae in the order Lagomorpha.
The only lagomorph commonly used in research is the European rabbit, Oryctolagus