General Animal-Care Concerns
TRAINING AND SUPERVISION
Oversight and training of all individuals associated with animal care and use (PI, research personnel, students, animal-care staff, veterinary staff, and IACUC members) is critical for the success of research. Gaining consensus on the importance of training is easy; implementation and participation present challenges. Neuroscience research often involves situations in which the research team and the animal-care staff must work in close cooperation to optimize both animal welfare and research outcomes. The diversity in education and experience of these multi-disciplinary teams adds to the training challenge.
Proper training is fundamental in ensuring animal welfare, and is recognized by regulatory agencies. For example, both the AWRs and PHS Policy require institutions to ensure that every person who works with animals is appropriately qualified (AWR 2.32(a) and PHS Policy IV.C.1.f.). There are several good references that provide guidance on training, including Essentials for Animal Research: A Primer for Research Personnel (Bennett, Brown, & Schofield, 1994) and Education and Training in the Care and Use of Laboratory Animals: A Guide for Developing Institutional Programs (NRC, 1991).
Although the PI is ultimately responsible for ensuring that appropriate training has been provided to the research staff, it is an institutional responsibility to make available training in animal anesthesia, surgery, experimental manipulations, and occupational health and safety. The ultimate responsibility for overseeing training rests with the IACUC, which must consider the qualifications of personnel involved in conducting research as part of its protocol review and approval process (AWR 2.31 (d)(1)(viii); NRC, 1996).
Training of research personnel should include procedure-specific training in neuroscience-research techniques, which the PI or senior research staff are usually best suited to teach, and more general training in such subjects as regulation, aseptic technique, anesthesiology, euthanasia techniques, and animal handling, which members of the veterinary or animal-care staff are generally most qualified to teach. The extent of training can depend on the duties and responsibilities of the staff involved. If the procedures to be used have the potential to cause pain and/or distress, mechanisms must be in place to ensure that the research staff can perform them competently. The selection of a trainer should be flexible and adaptive because it will depend on who is best qualified and prepared to provide training. A consortium of individuals from various disciplines may be necessary for complex projects (Kreger, 1995).
Training should be a continuing process. Open communication and cooperation between the veterinary staff and the investigative staff regarding innovations in technique are essential to ensure the most up-to-date and refined use of animals.
Evaluation of outcomes and results is critical in assessing technical experience and the need for training. The IACUC must be prepared to re-review training and experience whenever problems occur in projects.
MONITORING EXPECTED AND UNEXPECTED CONSEQUENCES
Assessing the nature and context of the clinical problems that an animal may experience during neuroscience experiments can be challenging for both researchers and veterinarians. For example, some strains of genetically modified mice spontaneously develop severe and debilitating disease unrelated to experimental manipulation. In some models, animals may develop substantial or exacerbated neurologic abnormalities because of drug treatment or experimental lesions. The assessment of postprocedure pain, distress, and general health is a matter of subjective clinical judgment that depends on evaluating a variety of measures, including behavioral factors, and recognizing that the interpretation of these measures differs greatly among species; for example, some species mask pain or distress from the observer. However, how a trained animal performs a behavioral task can be a sensitive index of its general condition. Changes in baseline experimental measures can also be informative. In many cases, a change in a specific behavioral measure, rather than changes in a general repertoire of behaviors, is particularly informative. Accordingly, thorough record keeping is essential in any behavioral monitoring program, and the frequency and method of record keeping should be described in detail in the animal-use protocol.
Review of proposed experiments that involve the care and use of animals with induced neurologic deficits poses special concerns for IACUCs. Depending on the nature and extent of the deficits, animals with induced neurologic disease may be limited in their ability to ambulate, obtain food and water, groom, urinate, or defecate, or they may experience pain, behavioral depression, or fear. The
experimental induction of debilitating neurologic deficits must be well grounded in scientific need, the animals must receive appropriate specialized care as needed, the number of animals exposed to a debilitating deficit must be minimized, the experimental end point must be well defined (for example, as to the length of time that an animal may be debilitated or the degree of debility), and the experimental protocol should be refined to reduce or eliminate pain, distress, discomfort, and mortality to the greatest extent consistent with valid experimental and statistical design.
Studies of neural injury and disease necessitate stringent requirements for the assessment and alleviation of animal pain and distress. Prolongation or repetition of many treatments, chronic alteration of neural activity, or the destruction of a population of neurons can cause substantial or permanent neurologic deficits. Neuroactive agents and even treatments themselves can cause adverse side effects or toxicity. Evaluating the likelihood of such adverse outcomes and designing strategies for avoiding or alleviating them without compromising the scientific goal of an experiment can be challenging for investigators, veterinarians, and IACUCs. For example, in some studies, the repeated application of an agent or a treatment might require multiple major survival surgeries. In such cases, the stress of undergoing general anesthesia repeatedly, the level of necessary asepsis, and the need to perform the procedures in a surgical setting may be special considerations.
The personnel in a research laboratory usually have some knowledge or expectation about the likely effects of a specific neuroscience procedure on animal health and well-being. Such information is typically solicited as part of the protocol evaluation. A structured approach to developing a profile of anticipated pain, distress, or disease should consider whether any major body systems are likely to be substantially affected during a study. Such an assessment can also guide the development of a systematic approach to animal monitoring and record keeping. The plan should incorporate a list of variables to be assessed and a timetable of observations. Three general considerations apply to research projects that require animal monitoring and maintenance to promote animal well-being (NIH, 1991):
Consultation. Consultation between neuroscientists and veterinarians is essential for the design and implementation of monitoring and maintenance procedures. Achieving appropriate solutions to problems that arise in neuroscience experiments requires continuing discussion and collaboration between the PI and veterinarians. Interaction should begin before experiments are initiated. The interaction between the research team and the veterinary staff should provide an opportunity for mutual education and support.
Responsibility. Periodic or regular veterinary assessment of both the animal and the experimental records is important in ensuring adequate veterinary care. Animal monitoring and maintenance are conducted and should be documented by neuroscientists as a routine part of their experiments. Documentation should in-
clude objective data to identify clinical trends. The records should be readily available to the attending veterinarian acting on behalf of the IACUC. Veterinary oversight is essential to the process for two reasons. First, laboratory-animal veterinarians are trained specialists in the recognition and management of animal health problems, whether spontaneous or iatrogenic. Second, regulatory responsibility for providing appropriate veterinary care rests with the veterinarian.
Record keeping. Good record keeping is essential. Records should be written as soon as practical after the animal observation is conducted and should not be phrased with excessive jargon or abbreviations. They should be dated and signed by the record keeper. Observations should be clearly understood by all persons who may have reason to read the records. This documentation serves at least four purposes:
It facilitates detection of gradual changes in health that might not be obvious during a single observation period. A change in condition (such as weight loss) can sometimes be more informative than the condition at any given time.
It requires an advance decision regarding the characteristics that will be assessed and the frequency of monitoring. Completing a form or checklist designed for a particular study promotes diligence and consistency.
It becomes an archive that can be used to improve future study design and animal management.
It documents that appropriate monitoring and maintenance activities were conducted.
All personnel who use animals should be trained to recognize health problems in their animals. That requires knowledge of the appearance and behavior of normal and abnormal animals and a solid understanding of what conditions are acceptable and unacceptable. Animals should be observed initially in an undisturbed state in their home cage. Making such observations can be difficult in some modern high-density caging systems, but attempts should nonetheless be made to evaluate the animals for general activity levels, posture, the condition of the hair coat, signs of self-induced trauma, pattern of respiration, and the general condition of the cage.
Next, the animals should be examined, especially if it is suspected that they have problems. The frequency of individual examination depends on the nature of the debility or disease and the expected rate of progression. For example, for general rodent-colony health surveillance, evaluations should probably be done only on a scheduled basis, such as during a cage change. This limits the number of times that the cages are opened as opening rodent cages properly can be time-consuming and is not risk-free, particularly in light of the possibility of subclinical infectious disease. Characteristics that can be assessed through manipulation are the response to handling; tremors, seizures, vocalization; ulceration; masses; injury; abnormalities of the eyes, ears, nose, or mouth; hyperthermia or hypothermia; and general body condition. Body-condition assessment in rodents requires
that personnel learn to palpate the vertebral column to look for emaciation (Ullman-Cullere and Foltz, 1999). Body-condition scoring can be superior to simply weighing animals (especially rodents) because it minimizes the potential for the spread of disease through a shared scale, a reference weight is not needed to calculate a percentage of weight loss for assessment of health status, and body condition can be evaluated more rapidly than body weight.
Neuroscience preparations can cause various degrees of debility that may be predictable in both severity and duration. Sedating the animals at critical post-procedural intervals may prevent discomfort and even inadvertent injury. If debility is unexpectedly severe or prolonged, the PI and the attending veterinarian must intervene to ensure the animal’s welfare. For example, a necessary intervention for animals that are not drinking is fluid replacement to prevent dehydration. Similarly, an anorectic animal may be encouraged to eat by being provided easy access to soft, rather than hard, food or a highly palatable food rather than the standard diet.
Appropriate scheduling of procedures that are potentially debilitating, painful, or stressful is important. It may be challenging to provide adequate veterinary care at night, on the weekend, or over holidays because of a shortage of trained staff, closure of diagnostic laboratories, or an inability to obtain specific drugs or equipment. For this reason, it is recommended that researchers schedule experimental procedures that may necessitate supportive or interventional care so that the time during which the animals would be expected to experience distress falls during normal operating hours.
Detection, assessment, and alleviation of pain and distress are additional critical aspects of animal monitoring. Both pharmacologic and nonpharmacologic interventions can be used to alleviate pain and distress. The research and animal-care staff must ensure that instances of animal pain and distress are reported promptly to a veterinarian. Research personnel and animal-care staff must be trained to recognize signs of pain and distress in the species they care for or use.
In summary, appropriate monitoring of animals and maintenance of clinical and experimental records are essential for maximizing the well-being of experimental animals in neuroscience research. Training of personnel and good communication among the research personnel, animal-care staff, and the veterinarian are key components of success.
ANIMAL HUSBANDRY AND NURSING CARE
Many animal models used in neuroscience research demand exceptional attention to daily care. Induction of neural injury and disease may compromise animals’ basic coping and survival mechanisms, as well as their ability to eat, drink, and defecate. Communication, coordination, and creativity in implementation of basic nursing support by the research team and veterinarian are necessary for successful outcomes in these challenging animal models.
The research team should describe the model to the IACUC in the animal-use protocol and should review the approved protocol with the animal-care and veterinary team before beginning the study. The process should include an overview of the scientific benefits that could be achieved from the study and a frank discussion of the challenges involved in maintaining the comfort of the animals after development of a deficit. This front-end investment will go a long way toward creating a team approach to maintaining what are, in effect, intensive-care patients.
The clear delineation of responsibility for monitoring animals is fundamental in ensuring adequate postprocedure care. The Guide’s general recommendation for daily observation may be inadequate in many cases. Ideally, frequent observation and the opportunity for intervention constitute a team effort involving both the research group and the animal-care and veterinary staffs. Clearly defined and well-understood scientific goals allow informed intervention (as opposed to inaction) by the caregivers to manage the animals optimally without compromising research goals. A planned strategy for undertaking defined nursing interventions benefits both the animals and the research.
The basics of animal husbandry that are so readily provided in modern housing systems—bedding, food and water, waste-handling—may require extensive modifications or personnel intervention for animals with impaired nervous system function. Enlisting the animal-care group early to consider strategies that will meet basic needs and maximize well-being presents an opportunity to build a team approach.
Generally, recovery from neurosurgery involves the same considerations as recovery from other surgical procedures. Cranial surgery is typically well tolerated by laboratory species. Postoperative analgesia should be used whenever it would not compromise scientific goals. Moistening of chow or providing a diet of softer or more palatable foods for several days postoperatively may make eating more comfortable for the animal and promote food intake, but nutritional modifications are often unnecessary. It may be necessary to consider the use of specialized or modified caging for animals with implanted devices, for example, it may be necessary to remove hanging food bins from rodent cages and place the food on the floor of the cage as hanging food bins could potentially damage a cranial implant.
Special considerations with respect to social housing may be warranted for animals that have had devices implanted for neuroscience research. Animals recovering from such surgery should generally be housed individually during recovery. If damage to implanted devices by a cagemate is unlikely, most animals can then be gradually reintroduced to social housing after their behavior returns to normal.
Some neuroscientists study animal models of human disease. Thus, some surgery is intended to alter the normal physiologic functions of the animal systematically and can affect the psychologic or behavioral state of the animal during postsurgical recovery.
Such procedures include those that reduce a subject’s ability to interact socially or with the environment. Examples are procedures that result in impairment of sensory perception, limit an animal’s movement capacities, and impair cognitive abilities. After those procedures, appropriate accommodations should be made in an animal’s housing environment or access to enrichment devices to maximize the extent to which it can interact socially and with the environment. Such accommodations can include housing the animal in a social group where it will not be subject to aggressive attacks, giving it manipulanda that can be used with a particular sensory or motor deficit, and giving increased personal attention to an animal that can no longer be put in social housing (NRC, 1998).
On occasion, changes in standard husbandry practices are warranted by the scientific goals of an experiment. For example, cats may be reared in total darkness to determine the influence of visual experience on the development of the visual system (Lein and Shatz, 2000; Mower and Christen, 1985) or animals with lesions of the labyrinth may be housed in the dark to prevent visual compensation for altered vestibular cues (Fetter et al., 1988; Zennou-Azogui et al., 1996). In each of those types of neuroscience research, the animal protocol must ensure appropriate care and monitoring of the animal while maintaining the environmental requirements of the experiment; for example, food and water might be provided in the same locations before and after the lesion is produced.
Care of animals used in neuroscience or behavioral research often requires creativity and exceptions to an institution’s normal husbandry procedures. For example, the research team often provides all or much of the daily care of animals used in behavioral studies because of protocol-specific issues or special housing situations. If husbandry responsibilities (including cleaning and sanitization) are to be shared by the animal-care staff and the research staff, the role of each group must be clearly delineated and the care must be documented and freely available to both parties. Integrated husbandry responsibility can work well but only when all members of the team know and accept their roles. The IACUC is authorized to approve exceptions to standard husbandry practices that deviate from the Guide’s recommendations when the exceptions have a sound justification and appropriate performance standards are met.
SPECIAL ENVIRONMENTS AND ENCLOSURES AND HOUSING OF MULTIPLE SPECIES
Experimental designs for neuroscience or behavioral studies may involve the use of special environments, including periodic or chronic housing of animals in unusual, nontraditional settings; for example, animals may be reared in total darkness or exposed to omnidirectional sound, microgravity or hypergravity, hyperbaric, or magnetism-free environments. The need to use a special environment may require housing multiple species in close proximity. The Guide recommends “physical separation of animals by species to prevent interspecies disease transmission
and to eliminate anxiety and possible physiologic and behavioral changes due to interspecies conflict” (p. 58). However, the well-defined health status of most research animals allows the risk of interspecies disease transmission to be reasonably assessed. The possibility of interspecies physiologic and behavioral stressors must also be evaluated. Occasionally, those stressors are an integral part of an experimental design. The veterinarian and IACUC should carefully evaluate such factors and work with the investigator to develop reasonable compromises that allow a balance between animal welfare and research objectives.
Neuroscience or behavioral research may also require the use of nontraditional primary enclosures or caging. Special configurations may allow less space than the standard minimal recommendations in the Guide. The Guide (p. 25) encourages the use of professional judgment and performance outcomes in assessing space needs for animals with special research needs. It is important that deviations from the Guide’s space recommendations be evaluated continuously, not just approved prospectively.
SURGERY AND PROCEDURES
Frequently, surgical procedures are required to meet the scientific needs of neuroscience research, and it is the responsibility of PIs, veterinarians, and IACUCs to ensure that the procedures are designed and conducted in a manner that complies with applicable animal-welfare guidelines and regulations. Interpreting the guidelines and regulations and applying them to a specific neuroscience procedure can be complicated, and it is important for all concerned to be cognizant of the relevant guidelines and regulations.
The Guide states that:
In general, surgical procedures are categorized as major or minor and in the laboratory setting can be further divided into survival and nonsurvival. Major survival surgery penetrates and exposes a body cavity or produces substantial impairment of physical or physiologic functions (such as laparotomy, thoracotomy, craniotomy, joint replacement, and limb amputation). Minor survival surgery does not expose a body cavity and causes little or no physical impairment (such as wound suturing; peripheral-vessel cannulation; such routine farm-animal procedures as castration, dehorning, and repair of prolapses; and most procedures routinely done on an “outpatient” basis in veterinary clinical practice) [pp. 61–62].
Minor procedures are often performed under less-stringent conditions than major procedures but still require aseptic technique and instruments and appropriate anesthesia. Although laparoscopic procedures are often performed on an “outpatient” basis, appropriate aseptic technique is necessary if a body cavity is penetrated [p. 62].
The definition of a major operative procedure in the AWRs is almost identical with that in the Guide except that it refers to permanent, rather than substan-
tial, impairment of functions (AWR 1.1). Both the Guide and the AWRs offer additional language related specifically to the conduct of survival surgical procedures—those in which the animal is allowed to awaken from surgical anesthesia. The Guide provides detailed recommendations regarding facility requirements for survival surgery (pp. 62–63, 78–79) and also states:
In general, unless an exception is specifically justified as an essential component of the research protocol and approved by the IACUC, nonrodent aseptic surgery should be conducted only in facilities intended for that purpose [p. 62].
The relative susceptibility of rodents to surgical infection has been debated; available data suggest that subclinical infections can cause adverse physiologic and behavioral responses (Beamer, 1972; Bradfield et al., 1992; Cunliffe-Beamer, 1990; Waynforth, 1980, 1987) that can affect both surgical success and research results. Some characteristics of common laboratory-rodent surgery—such as smaller incision sites, fewer personnel in the surgical team, manipulation of multiple animals at one sitting, and briefer procedures—as opposed to surgery in larger species, can make modifications in standard aseptic techniques necessary or desirable (Brown, 1994; Cunliffe-Beamer, 1993). Useful suggestions for dealing with some of the unique challenges of rodent surgery have been published (Cunliffe-Beamer, 1983, 1993) [p. 63].
The AWRs stipulate:
Activities that involve surgery include appropriate provision for pre-operative and post-operative care of the animals in accordance with established veterinary medical and nursing practices. All survival surgery will be performed using aseptic procedures, including surgical gloves, masks, sterile instruments, and aseptic techniques. Major operative procedures on non-rodents will be conducted only in facilities intended for that purpose which shall be operated and maintained under aseptic conditions. Non-major operative procedures and all surgery on rodents do not require a dedicated facility, but must be performed using aseptic procedures. Operative procedures conducted at field sites need not be performed in dedicated facilities, but must be performed using aseptic procedures [AWR 2.31 (d)(1)(ix)].
When preparing animal-use protocols for neuroscience experiments that require surgical procedures, PIs must take care to describe all aspects of their proposed perioperative procedures accurately and completely. In reviewing the protocols, veterinarians and IACUCs must ensure that proposed surgical procedures are properly classified as major or minor, survival or nonsurvival, and rodent or nonrodent. Furthermore, veterinarians and IACUCs must ensure that the surgical procedures are performed in a manner that complies with the Guide and the AWRs; for example, major nonrodent survival surgery should be conducted in a dedicated surgical suite.
Although that sounds relatively straightforward, the complexities of contemporary neuroscience research demand that professional judgment, guided by outcome or performance-based considerations (NRC, 1996, p. 3) be used at each
step of the process. Both the PHS Policy and the AWRs permit a great deal of flexibility in their application to research by allowing the IACUC to grant exceptions to their recommendations when acceptable justification is provided. Thus, the PHS Policy states:
The IACUC shall confirm that the research project will be conducted in accordance with the Animal Welfare Act insofar as it applies to the research project, and that the research project is consistent with the Guide unless acceptable justification for a departure is presented [Policy IV.C.1].
The AWRs state:
In order to approve proposed activities or proposed significant changes in ongoing activities, the IACUC shall conduct a review of those components of the activities related to the care and use of animals and determine that the proposed activities are in accordance with this subchapter unless acceptable justification for a departure is presented in writing; [AWR 2.31 (d)(1)].
A common exception to the AWRs and to the PHS Policy surgical requirements that IACUCs allow is to permit major surgery to be performed in a modified laboratory setting when necessary equipment (such as electrophysiologic recording equipment) cannot be moved to a dedicated surgical suite (see section on “Asepsis and Physical Environment,” below).
One area of confusion for IACUCs, veterinarians, and researchers alike is the definition of what actually constitutes a major surgery. Neuroscience research often involves procedures that do not meet the strict definitions of major survival surgery given in the Guide and AWRs. Some procedures do not involve both penetration and exposure of a body cavity (for example, endoscopic surgery), or they do not penetrate or expose a body cavity at all (for example intravenous infusion or injection of neuroactive or neurotoxic substances, closed-head trauma, or peripheral neurectomy). Determining whether such procedures meet the definitions of major survival surgery hinges on whether they seem likely to produce “substantial impairment of physical or physiological functions” (NRC, 1996, pp. 11-12, 61) or “permanent impairment of physical or physiological functions” (AWR 1.1).
The IACUC must assess whether a proposed minimally invasive procedure seems likely to result in an impairment of physical or physiologic functions that is substantial or permanent. If so, the procedure must be categorized as a major surgical procedure and reviewed as such by the IACUC to ensure compliance with the provisions of the Guide and the AWRs. However, both the Guide and the AWRs expect the IACUC to exercise professional judgment in applying their criteria to the review of surgical protocols. For example, the Guide does not define what constitutes a “substantial impairment of physical or physiologic functions,” and does not require that an induced impairment be permanent to be considered major surgery. The AWRs stipulate that a noninvasive procedure should result in a “permanent” impairment to be classified as major surgery, but
they do not require the impairment to be substantial. Because minimally invasive procedures like those mentioned above can result in impairments whose severity ranges from no apparent loss of function to obvious major functional deficits or impairments whose severity changes markedly (either lessening or worsening) over time, IACUC review of these kinds of protocols can be challenging.
Rather than debate the extent to which a particular neuroscience procedure meets the various regulatory criteria for a major surgical procedure, the authors of this report strongly recommend that PIs, veterinarians, and IACUCs collaborate on the development of animal-use protocols that are designed to safeguard animal welfare and address the scientific needs of the research. The Guide and the AWRs provide for sufficient flexibility in the application of their standards for major surgical procedures to allow those involved to craft such protocols. Careful attention should be given to the outcomes of earlier neuroscience studies that used the same or similar procedures. In the absence of precedents for a particular minimally invasive procedure, consideration should be given to obtaining preliminary data from a pilot study performed under direct veterinary supervision and with appropriate reporting to the IACUC.
Multiple Major Survival Surgeries
In general, multiple major survival surgeries are discouraged, but they may be conducted if they are scientifically justified, related components of a research project approved by the IACUC (NRC, 1996). For example, animals that receive a unilateral visual cortex lesion neglect visual stimuli presented to them on the side opposite the lesion (contralateral visual neglect). Subsequent lesioning of the tectum can ameliorate this neglect (Sprague, 1966). A physical or chemical lesion of the basal ganglia causes a Parkinson’s-like tremor in animals that can be reduced or eliminated by a second lesion (Bergman et al., 1990; Wichmann and DeLong, 1996) or by stereotaxic implantation of a stimulating device (Benazzouz et al., 1993, 1996; Boraud et al., 1996). The use of cranial implants for experimental restraint, recording chambers, or implanted monitoring devices is another situation where multiple survival surgeries may be justified.
Some neuroscience-research designs involving multiple surgeries and procedures may have special requirements for asepsis and facilities that will be discussed later. Careful monitoring of the animal, in consultation with the veterinary staff, is necessary. Techniques should be continually refined to minimize pain and/or distress and the monitoring program should be appropriately matched to the anticipated level of pain and/or distress.
Multiple major survival surgeries may also be used to conserve scarce animal resources or if the multiple major surgeries are needed for clinical reasons (NRC, 1996). When a research project involves species covered by the AWA, a waiver must be obtained in writing from USDA for multiple major surgeries that are not related components of a research project (APHIS/AC Policy 14). The
balance between the welfare of the individual animal and minimization of the numbers of animals used must be carefully weighed by the IACUC. Cost alone is not an acceptable reason for performing multiple major surgeries (NIH, 1991).
Planning for Survival Procedures
In survival-surgery experiments, monitoring and maintenance issues arise before, during, and after surgery, and in relation to long-term survival and animal health. Responsibility for and details of monitoring of individual animals during and after surgery must be clearly defined in the animal-use protocol (APHIS/AC Policy 3) and presurgical planning should identify the personnel who will perform these duties (NRC, 1996). Medical records should be maintained on each animal throughout the course of an experiment; in fact, this is a requirement for all AWA covered species (APHIS/AC Policy 3). Before any surgery, the weight, general health, and distinctive characteristics of the animal should be noted.
Careful consideration should be given to the anesthetic agents used so that adverse effects of the agents on data collection can be minimized (Cherry and Gambhir, 2001) and the need for post-operative analgesics reduced. Provisions should be made to monitor vital signs and depth of anesthesia frequently, and methods to monitor the animal before and after the surgical procedures should be established.
Removal of food is a standard veterinary practice before any major or minor recovery surgical procedure (Flecknell, 1996), although water should not be restricted. However, as rodents and rabbits cannot vomit, it is unnecessary to fast them prior to a surgical procedure (Waynforth et al., 2003). To maintain proper hydration throughout the surgical procedure, an intravenous line may be established through which supplemental doses of anesthetic or emergency drugs may also be delivered when appropriate. This is especially encouraged for larger animals such as nonhuman primates. Administration of fluids is especially important for smaller mammals during lengthy surgical procedures because their ratios of surface area to body weight and higher metabolic rates necessitate nearly double the fluid supplementation necessary for larger mammals (Balaban and Hampshire, 2001). Fluids should be warmed before infusion to prevent their contributing to hypothermia (Balaban and Hampshire, 2001). Maintaining body temperature during the surgical procedure and post-operative recovery is critical, as a side effect of sedation is hypothermia. Rodents and other small mammals are particularly susceptible to irreversible hypothermia leading to death (Hedenqvist and Hellebrekers, 2003).
If surgery facilities are not in the vivarium, the method and route of transportation to and from the surgery facilities should be considered when preparing animal-use protocols. When these routes take the animal through public areas, such as hospital corridors, care must be taken to minimize any potential contact with the public, who are not expecting nor are prepared to encounter the animal.
Special attention must be given to the public health concerns, such as B virus exposure, that may arise from transporting an animal through public areas. Procedures should be established for dealing with emergencies that may arise during transport, such as bites, scratches, or splashes to a member of the public or the research staff. Consideration must also be given to the potential for animal tissues or fluids to contaminate public corridors, elevators, or patient areas during the transport. Furthermore, the personnel and methods used to monitor the animals and to administer appropriate care to ensure their well-being during transport should be identified. Planning for the transfer of an animal from a vivarium to a surgery facility should include all personnel that will be involved in the transport.
Anesthesia and Analgesia
The goal of this section is to provide investigators, veterinarians, and IACUCs with a general understanding of the differences between anesthetics and analgesics and the concepts underlying preemptive analgesia and balanced anesthetic regimens. The purpose underlying the use of any of these drugs or regimens is to relieve unintended pain and/or distress (experiments involving unrelieved pain and/or distress are discussed later in this chapter). As noted in the US Government Principles (IRAC, 1985), “proper use of animals, including the avoidance or minimization of discomfort, distress, and pain when consistent with sound scientific practices, is imperative”.
As noted in the Guide (p. 64),
The selection of the most appropriate analgesic or anesthetic should reflect professional judgment as to which best meets clinical and humane requirements without compromising the scientific aspects of the research protocol.
The use of professional judgment, open discussion, and the flexibility of all involved parties are particularly encouraged when tackling this complex issue.
In developing a pain-relieving regimen, it is important to understand the difference between anesthesia and analgesia. General anesthesia produces a loss of awareness or consciousness and is used for surgical procedures or experiments that cannot be conducted in awake animals (NRC, 1992). Examples of general anesthetics are inhalation anesthetics, such as isoflurane; opioids, such as fentanyl; and dissociatives, such as ketamine. Inhalation anesthetics produce unconsciousness and muscle relaxation sufficient for surgical intervention (NRC, 1992). However, many injectable anesthetics do not provide enough sedation, muscle relaxation, or analgesia to be used alone. For example, fentanyl provides sedation and analgesia, but muscle relaxation is poor (Hedenqvist and Hellebrekers, 2003); ketamine does not produce visceral analgesia (NIH, 1991). For that reason, they are seldom used as the sole anesthetic in major surgery but instead are combined with other agents in a balanced anesthesia regimen (NIH, 1991). In these cases, drugs with different pharmacological effects are used in combination to produce
a surgical anesthesia; for example, fentanyl and midazolam may be used in concert or ketamine and medetomidine may be combined (Hedenqvist and Hellebrekers, 2003). Using local anesthetics in combination with a general anesthetic is another example of a balanced anesthesia regimen, the benefit being that local anesthetics can reduce the need for general anesthesia and side effects associated with higher doses of general anesthesia (Gordon et al., 2002).
Analgesia is the inability to feel pain; an analgesic drug relieves pain but does not cause a loss of awareness. Analgesics include opioid drugs, such as morphine, and NSAIDs, such as aspirin and ibuprofen (NRC, 1992). There is evidence that surgery (or tissue injury) induces sensitization of central neural function, causing nociceptive inputs from the surgical wound to be perceived as more painful (hyperalgesia) than they would otherwise have been, and causes innocuous inputs to give rise to pain (allodynia). Studies have shown that preemptive analgesia (such as opiates, local anesthetics, or NSAIDs) prevents this sensitizing, reducing postoperative pain intensity and decreasing postoperative analgesic requirements for periods much longer than the duration of action of the preemptively administered analgesic (Coderre et al., 1993). Researchers should be encouraged to preemptively use analgesics.
Sedatives and anxiolytics may be used for the relief of non-pain-induced distress. They are often combined with analgesics to produce a state free of pain and distress—for example, in the management of postsurgical pain or pain associated with disease—and are also useful for restraint during minor procedures (NIH, 1991; NRC, 1992).
Systemic paralysis is commonly used in neuroscience experiments. These experiments require that the animal be paralyzed with a neuromuscular blocking agent to prevent movement, such as movement of the ocular muscles during visual experiments. Neuromuscular blocking agents are used only in fully anesthetized animals (NRC, 1996). They do not interact substantially with anesthetics and analgesics, but they leave an animal unable to respond behaviorally to pain or distress. That can make it difficult to evaluate the depth of anesthesia and the adequacy of analgesia, so other signs of pain or distress must be used, such as lacrimation, salivation, reactivity of heart rate and arterial blood pressure to noxious stimuli, or electroencephalographic recordings (NIH, 1991). Such signs are not adequate singly, but in combination they can provide valuable information about an animal’s physiologic status (NIH, 1991). In addition, care should be taken to ensure that the animal has recovered control of respiration and locomotion before it is returned to the home cage. A detailed discussion of monitoring paralyzed animals can be found in Chapter 5.
It is important to confer with a laboratory-animal veterinarian to develop an adequate anesthetic and analgesic regimen. In fact, the AWRs states that a veterinarian be consulted during the planning of any procedure that could cause pain in animals (AWR 2.31(d)(1)(iv)(B)). Many resources are available to help the investigator and laboratory-animal veterinarian to develop a balanced anesthetic
and analgesic regimen (Deyo, 1991; Flecknell, 1996, 1997; Hillyer and Quesenberry, 1997; Kohn et al., 1997; NRC, 1992; Rosenberg, 1991; Smith and Swindle, 1994; Stoelting, 1999).
Asepsis and Physical Environment
Among the more problematic Guide recommendations for reviewers of proposed neuroscience protocols are those pertaining to physical environment and asepsis during surgery. The Guide states:
In general, unless an exception is specifically justified as an essential component of the research protocol and approved by the IACUC, nonrodent aseptic surgery should be conducted only in facilities intended for that purpose [p. 62].
For most rodent surgery, a facility may be small and simple, such as a dedicated space in a laboratory appropriately managed to minimize contamination from other activities in the room during surgery [p. 78].
The AWRs state:
All survival surgery will be performed using aseptic procedures, including surgical gloves, masks, sterile instruments, and aseptic techniques. Major operative procedures on non-rodents will be conducted only in facilities intended for that purpose which shall be operated and maintained under aseptic conditions. Non-major operative procedures and all surgery on rodents do not require a dedicated facility, but must be performed using aseptic procedures. Operative procedures conducted at field sites need not be performed in dedicated facilities, but must be performed using aseptic procedures [AWR 2.31(d)(1)(ix)].
The Guide further states:
The species of animal influences the components and intensity of the surgical program. The relative susceptibility of rodents to surgical infection has been debated; available data suggest that subclinical infections can cause adverse physiologic and behavioral responses (Beamer, 1972; Bradfield et al., 1992; Cunliffe-Beamer, 1993; Waynforth, 1980, 1987) that can affect both surgical success and research results [p. 63].
Many neuroscience procedures can be performed in full compliance with the AWRs and Guide recommendations for asepsis and the physical environment. However, if a survival surgical procedure requires the use of specialized equipment, facilities, or substances, performing it in a manner that complies fully with all recommendations can be impractical or impossible (NIH, 1991). Even when full compliance is not possible, most aspects of the recommendations can be met, such as the use of sterile surgical gloves, gowns, caps, and face masks; the use of sterile instruments; aseptic preparation of the surgical field; and appropriate postsurgical care.
IACUCs may receive requests to conduct survival neuroscience procedures in a modified laboratory setting to meet the scientific needs of an experiment, for
example, when the experiment requires specialized equipment that cannot be sterilized or moved into a dedicated surgical facility. Both the AWRs and the Guide allow for such exceptions when “an acceptable justification for a departure is presented in writing” (AWR 2.31 (d)(1)) or “an exception is specifically justified as an essential component of the research protocol” (Guide, p. 62). Before granting such an exception to the AWRs and Guide recommendations, an IACUC should consider the extent to which animals will be susceptible to increased risk of infection. In particular, the IACUC should carefully review the various safeguards that will be used to minimize the risk. Examples of the safeguards are aseptic preparation of a separate area of the laboratory in which the surgery will be conducted; the use of aseptic surgical attire, instruments, and supplies; and aseptic preparation and maintenance of the surgical field during the procedure (NIH, 1991). Maintaining equipment that cannot be moved or sterilized under plastic or equivalent cover when not in use is encouraged to decrease any potential contamination of the equipment.
Several things can be done to make a laboratory setting more suitable for major survival surgery. The room should be free of unnecessary equipment. In some situations, a large, general-purpose laboratory can be partitioned to isolate a smaller surgical area. The room in which surgery is to be performed must be sanitized immediately before each procedure. The walls, ceiling, and floor should have smooth surfaces that are impervious to moisture and easily cleaned (NIH, 1991).
A decision to allow major survival surgery to be performed in a modified laboratory setting should be contingent on the development of a set of stringent postsurgical monitoring and reporting procedures. For example, the IACUC may approve an exception to the Guide’s recommendations subject to receiving a status report from veterinary staff on the health and welfare of animals during the postsurgical survival period, to ensure the efficacy of the various procedures proposed to mitigate the risk of postsurgical infection.
Postsurgical Recovery Period
As noted in the Guide:
The investigator and veterinarian share responsibility for ensuring that postsurgical care is appropriate. An important component of postsurgical care is observation of the animal and intervention as required during recovery from anesthesia and surgery. The intensity of monitoring necessary will vary with the species and the procedure and might be greater during the immediate anesthetic-recovery period than later in postoperative recovery [p. 63].
Temperature and hydration should be monitored, maintained, and recorded until recovery from anesthesia. Monitoring heart rate and respiratory rate may also prove useful. Animals should be monitored until it is determined not only that the animal is normothermic, but also that it can maintain a normal body
temperature in the absence of supplemental heat (Waynforth et al., 2003). Small mammals, such as rodents, are particularly susceptible to hypothermia and must be monitored closely (Hedenqvist and Hellebrekers, 2003). Because the period of recovery from anesthesia is often a time of substantial physiologic change, policies and procedures that ensure adequate observation of the animal to facilitate prompt correction of problems should be implemented. According to the Guide:
During the anesthetic-recovery period, the animal should be in a clean, dry area where trained personnel can observe it often. Particular attention should be given to thermoregulation, cardiovascular and respiratory function, and postoperative pain or discomfort during recovery from anesthesia. Additional care might be warranted, including administration of parenteral fluids for maintenance of water and electrolyte balance (FBR, 1987), analgesics, and other drugs; care for surgical incisions; and maintenance of appropriate medical records [pp. 63–64].
After anesthetic recovery, monitoring is often less intense but should include attention to basic biologic functions of intake and elimination and behavioral signs of postoperative pain, monitoring for postsurgical infections, monitoring of the surgical incision, bandaging as appropriate, and timely removal of skin sutures, clips, or staples (UFAW, 1989, p. 64).
Determining the Appropriate Restraint Procedure
Restraint has been characterized as a physiologic and psychologic stressor (Norman et al., 1994; Norman and Smith, 1992). The method of animal restraint used to achieve a particular objective in an experimental protocol should, according to the US Government Principles (IRAC, 1985), be chosen so as to minimize distress to the animal. And the Guide (p. 11) states that prolonged restraint should be avoided unless it is scientifically justified and approved by the IACUC. The Guide recommends that less restrictive methods be chosen when possible. The objectives of restraint should be clearly defined in the animal-use protocol. Specifically, the degree of restraint needed (head only, arms and head, whole body, and so on) will guide the method of restraint and the type of equipment used.
Personnel safety may also necessitate restraint of an animal. For example, neuroscience studies that require physical proximity of macaques and the experimenter should involve sufficient restraint of the primates to minimize the risk of handler exposure to B virus by scratch, bite, or splash. Similarly, a rabbit-restraint device has been described that secures the animal and immobilizes its feet to prevent scratching of the experimenter (Abell et al., 1995). Animals that have been habituated to a suitable restraint method will probably be less stressed and agitated during a procedure, thereby reducing risk of injury to the handler (Sauceda and Schmidt, 2000).
Newer restraint devices and techniques reduce distress and enhance the health and comfort of animals by taking into account their behavior and typical postural adjustments. For example, Binder (1996) describes a short-term mouse-restraint device that is based on the mouse’s preference to seek shelter; the mouse reliably and voluntarily crawls forward so that its head is in an opening of a box while the experimenter gently restrains it by holding on to its tail. Such a restraint procedure is presumably less stressful for the animal, because the animal can express a normal coping behavior.
Observation of an animal during restraint is critical and should be thoroughly described in the animal-use protocol. Observation may be direct, such as through a viewing window (e.g., Ator, 1991), or indirect, such as by remote video. The frequency of animal observation may depend on the specific experimental procedure, but it may also be determined by the species of animal, the degree and duration of restraint, the type of restraint device or technique, the stage of training of the animal to the restraint, and the animal’s degree of habituation to the restraint. Observation may include a physical examination, an evaluation of the animal’s behavior, and/or an assessment of various physiologic measures, such as concentrations of cortisol/corticosterone, leuteinizing hormone (LH), testosterone, and blood glucose) (Flecknell and Silverman, 2000; Norman and Smith, 1992; Rogers et al., 2002; Wade and Ortiz, 1997). Monitoring should be frequent because restraint-device failures and unanticipated actions by the animal can sometimes place the animal in jeopardy. Detailed records should indicate the date, the time, the observer’s name, and the observations made.
Occasionally, an animal will not adapt well to restraint. Criteria for the temporary or permanent removal of an animal from a study that requires restraint must be developed in advance of the study and be reviewed and approved by the IACUC. The development of physical or behavioral abnormalities should constitute a basis of a decision to temporarily or permanently remove an animal from a study.
Methods of Physical Restraint
The two principal methods of physical restraint are manual and device-facilitated. In general, manual restraint is used for short-term procedures. Personal protective equipment (such as gloves) is often used to enhance worker safety by preventing bites, scratches, or contact allergy that can occur with some species (Egglestone and Wood, 1992).
Device-facilitated restraints can be used safely in some situations. Depending on the goals of the study, the equipment can facilitate short-term restraint (e.g., Abell et al., 1995) or be incorporated into the animal housing (e.g., Coelho and Carey, 1990). Innovative restraint equipment, such as slings and restraining boxes, has been used successfully without increasing stress (Flecknell and
Silverman, 2000). Sauceda and Schmidt (2000) have reviewed common restraint devices used with macaques.
Conditioning an Animal to the Restraint Technique
An experimental procedure that requires performance or measurements of restrained animals can be successful only if the animal is not unduly stressed and is sufficiently habituated to the restraint so that it will attend to the task rather than focusing on the restraint itself. Thus, an initial investment of time to train the animal to accept the restraint is highly recommended, particularly for chronic or repeated restraint. In accordance with the AWRs (AWR 2.32), personnel working with restrained animals should be trained in using the equipment properly and in handling the animals safely while causing them minimal distress. Well-qualified personnel will have a sound understanding of when restraint should be suspended or stopped if it compromises animal welfare.
The period required for habituation of an animal to a restraint technique varies. For a single brief period of restraint, habituation may not be critical for obtaining valid data. In some studies, providing social animals with companionship during the restraint period may reduce restraint-related stress. For example, Fleischman and Chez (1974) chair-restrained baboons as pairs to reduce anxiety. Restraint can result in various physiologic changes in animals (e.g., Bush et al., 1977; Gartner et al., 1980), so a substantial period of habituation may be required to obtain valid data. The period of training depends on the species and the animal’s experience and behavior; animals will habituate to the procedure better if the equipment and handling procedure are species-appropriate, sized correctly for the animals, and otherwise adjusted to maximize the animals’ comfort. The habituation period is especially critical for studies that require more restrictive restraint (NIH, 2002). Maintenance of physiologic and behavioral measures within normal limits during restraint suggests that an animal is well adapted to the restraint, as do voluntary movement of the animal into the restraint equipment and performance of the requisite task (NIH, 2002). For example, Wade and Ortiz (1997) demonstrated that well-habituated monkeys had no rise in urinary cortisol associated with restraint.
Potential Consequences of Restraint
The correct use of restraint can facilitate the collection of accurate research data. However, the inappropriate application of restraint can adversely affect health, physiologic measurements, and behavior. An animal that has had an adverse experience during restraint may be more difficult to use in the future because of increased anxiety resulting from memory of the experience. The development of ulceration on the ischial callosities of some primates as a result of chronic restraint has been reported (e.g., Wade and Ortiz, 1997). Restraint has
also been shown to inhibit LH and testosterone secretion in male macaques and LH secretion during the follicular phase of the menstrual cycle of macaques, resulting in reduced fertility (Norman et al., 1994; Norman and Smith, 1992). Restraint that causes stress activates the hypothalamic-pituitary-adrenal axis. Thus, the potential scope of adverse effects on an animal bears careful consideration. Under all circumstances, the minimal restraint feasible should be used— such as a tether in lieu of chair restraint, or a shorter period of restraint—and the availability of adequate alternatives to the restraint should be assessed (Flecknell and Silverman, 2000).
Prolonged Physical Restraint
The Guide states, “prolonged restraint, including chairing of nonhuman primates, should be avoided unless it is essential for achieving research objectives and is approved by the IACUC” (p. 11). Although prolonged restraint has not been defined, as the duration of restraint increases, a concomitant increase in attention should be given to alternatives to restraint, the health and well-being of the animal, and endpoint criteria for the restraint. The AWRs direct the IACUC to review procedures to avoid or minimize animal discomfort, distress, and pain and direct the investigator to consider alternatives to procedures that may invoke more than momentary or slight pain and/or distress (AWR 2.31 (d)(1)(i,ii,iii,iv) and APHIS/AC Policy 11). The AWRs go on to state that in instances where long-term (greater than 12 hours) restraint is required, a nonhuman primate must be provided the opportunity daily for unrestrained activity for at least one continuous hour during the period of restraint, unless continuous restraint is justified for scientific reasons and approved by the IACUC (AWR 3.81 (d)).
FOOD AND FLUID REGULATION
Neuroscience-related protocols occasionally require the regulation of animals’ food or fluid intake to achieve a specific experimental goal. The regulation process may entail scheduling of access to food or fluid sources so an animal consumes as much as desired at regular intervals, or restriction, in which the total volume of food or fluid consumed is strictly monitored and controlled. As stated in the Guide, “the least restriction that will achieve the scientific objective should be used” (p. 12). Research protocols that use food or fluid regulation can be divided into at least three main categories: studies of homeostatic regulation of energy metabolism or fluid balance, studies of the motivated behaviors and physiologic mediators of hunger or thirst, and studies that regulate food or fluid consumption to motivate animals to perform novel or learned tasks (Toth and Gardiner, 2000).
In studies of homeostatic regulation, the manipulation of food or fluid availability would be predicted to directly influence a dependent variable that is being
measured as a specific aim of the experiment, for example, food restriction leads to neurally mediated hormone release.
In contrast, regulation of food or fluid is commonly used as motivation in experiments that require animals to perform a behavioral task with a high degree of repeatability (Toth and Gardiner, 2000), but the food or fluid consumption is not the experimental variable. In those studies, food and fluid regulation is used to motivate the animals to perform a specific behavioral task for a food or fluid reward; regulation of food or fluid outside the experimental session ensures response reliability to the food and fluid reward in each session (NIH, 2002). That allows the investigator to elicit and monitor the same movement repeatedly, to present the sensory stimuli under highly controlled conditions, and to obtain physiologic discriminations from the animal. For example, water-regulated monkeys may be trained to press a button for a juice reward, while the investigator measures the effect on neuronal firing rates. In conditioned-response experiments, (for example, a monkey may be conditioned to associate a light with a fluid reward), consideration should be given to whether the use of highly preferred food or fluid as positive reinforcement can be used instead of restriction.
Fluid reward is preferable to food reward in some types of experiments. For example, studies that monitor neuronal activity in the brain may require the minimization of jaw or head movement to avoid displacing a microelectrode from its position. Because fluid rewards can be delivered through a tube positioned near the animal’s mouth and tongue, they offer a particular advantage: licking and swallowing a fluid reward are much less disruptive to the neuronal recordings than chewing or crunching movements of the teeth or jaws that accompany the consumption of food rewards (NIH, 2002).
Fluids offer additional experimental advantages. They can be easily delivered in small quantities, maximizing the number of trials that can be executed before satiation of the animal. In contrast with food rewards that require chewing before swallowing, fluids are quickly consumed, reducing the intertrial interval—an important advantage when an animal must perform a behavior hundreds or even thousands of times in an experimental session to allow for statistical analysis.
In other studies, there may be disadvantages to using fluid rewards. For example, milk and juice require more extensive cleaning than water or solid food if spilled on the experimental apparatus. Milk and juice are also more susceptible to rapid spoilage and require frequent assessment or replacement.
In designing and evaluating an animal-use protocol that proposes to regulate access to food or fluid to facilitate operant training, the following questions should be considered:
What type food or fluid regulation is most appropriate for meeting the specific goals of the experiment?
Do alternative procedures exist that would facilitate the generation of the desired behavior without food or fluid regulation, or is food or fluid regulation the best option?
What is the proposed schedule of food and fluid access, and does it allow periodic ad lib access to food and fluid?
What is the proposed schedule for monitoring, so adverse effects will be recognized quickly.
Is laboratory chow or fluid the only item to be offered, or will other foods or fluids be considered?
What are the endpoints for intervention with supplemental feeding or hydration?
The development of animal protocols that involve the use of food or fluid regulation requires the determination of three fundamental details: the necessary level of regulation, the potentially adverse consequences of regulation, and methods for assessing the health and well-being of the animals. Consideration of each of those details facilitates the establishment of interventional endpoints to maintain the animals’ health and well-being.
The Guide states that when the experimental situations require food or fluid regulation, at least minimal quantities of food and fluid should be available to provide for development of young animals and to maintain long-term well-being of all animals. Regulation for research purposes should be scientifically justified and approved by the IACUC. A program should be established to monitor physiologic or behavioral indexes, including criteria (such as weight loss or state of hydration) for temporary or permanent removal of an animal from the experimental protocol (NRC, 1996, p. 12). Reducing an animal’s body weight by 15–20% (compared with cage-matched controls) is commonly the goal of food regulation (NIH, 2002).
In general, the total caloric intake of a food-regulated animal is 50–70% of that associated with ad libitum feeding (Bucci, 1992). In some cases, however, the attending veterinarian may determine that an animal needs to be removed from a study for health or behavioral reasons even if it has not reached that weight loss. In addition, species, strain, and individual differences may influence what is considered an acceptable amount of weight loss.
Special attention should be given to ensuring that the diet meets the animal’s nutritional needs (New York Academy of Sciences and Ad Hoc Committee on Animal Research, 1988) unless the scientific needs of the research protocol necessitate otherwise. Caloric restriction must not produce unintended nutritional imbalances. In some cases, it might be necessary to increase the concentration of selected nutrients to provide the same nutrition as provided to animals fed ad
libitum (NRC, 1995). Typically, the restricted diet contains proportionate decreases in energy sources (protein, fat, and carbohydrates) rather than a reduction in only one source of energy (Bucci, 1992).
The AWRs state that procedures involving more than momentary or slight aversive stimulation that is not relieved with medication or other acceptable methods should be undertaken only when the objectives of the research cannot be achieved otherwise. APHIS/AC’s Policy 11, “Painful Procedures,” lists “food or water deprivation beyond that necessary for normal presurgical preparation” as a procedure that may cause pain or distress. The IACUC should closely evaluate the pain-distress categorization of animals that are food-restricted in accordance with APHIS/AC Policy 11.
In some rats, a decrease of 20% in baseline body weight within 1 week is associated with increased serum corticosterone, which may reflect the physiologic response to caloric restriction, and with substantially greater freezing behavior in the open field test, which may be a stress response (Heiderstadt et al., 2000). Generally, it is recommended that animals be gradually reduced to a target weight and acclimated to the feeding schedule over some period, such as several weeks, to mitigate the stress response. The stress response associated with a rapid reduction in body weight can also be relieved by following the reduction with a diet designed to maintain the rat at 80% normal body weight, compared with a rat fed ad libitum. (Heiderstadt et al., 2000).
Mechanisms for Mediation of Food Consumption on an Ad Libitum Diet
There is individual variability in food intake and adult body weight (Toth and Gardiner, 2000). Food intake may be regulated more by satiety than by hunger (Stricker, 1984). Physiologic signs of satiety include gastric distention and increases in insulin secretion and metabolic processes in the liver (Toth and Gardiner, 2000). Food consumption is influenced by palatability and accessibility (Collier et al., 1972; Peck, 1978; Rolls et al., 1983). The amount of work an animal must do for food also influences the amount consumed; increases in workload reduce consumption to about 50% of control consumption (Collier, 1989; Collier et al., 1972; Nicolaidis and Rowland, 1977). Rowland et al. (1996) have summarized the various endogenous and exogenous regulators of food intake.
Determination of Minimum Caloric Consumption
A sound approach to developing a food-regulation protocol requires information about the minimum caloric requirements of animals. A general method of assessing caloric consumption is to require an animal to work for all its food under different reinforcement schedules (Collier, 1989; Nicolaidis and Rowland, 1977). Caloric needs and consumption vary with the life stage of the animal, such as growth, maintenance, gestation, and lactation; activity level; environmental
conditions, including social housing and thermoregulatory demands; and circadian rhythm. However, some animals consume more than is necessary to meet their metabolic needs (Toth and Gardiner, 2000). In some species, preventing coprophagy (ingestion of feces) results in an increase in nutritional requirements. Investigators should consider those factors in determining nutritional requirements related to different circumstances.
Two common methods are used for regulating the food intake of animals to motivate them to perform tasks. The first method restricts the amount of time available to the animals to eat, and the second restricts the amount of food available. When the former method, referred to as meal feeding, is used, the growth curve of animals generally remains below that of animals fed ad libitum; meal feeding also results in a different pattern of drinking, in which most of the volume is consumed during the meal period. When the latter method, referred to as restricted feeding, is used, a lower body weight also occurs. Both types of regulation cause a temporary increase in food efficiency in young rats, and the reduced intake is correlated with a reduction in the animal’s relative fat content and an increase in relative water content (Brownlow et al., 1993). However, rats on a restricted diet that was provided as two meals, rather than one, had substantially more adipose tissue. Food-restricted rats normalize their rate of weight gain (relative to control animals), although they will not achieve the mean body weight of satiated animals. The choice of regulation method should be based on the goals of the study and the behavior of the animal.
Species- and Strain-Specific Considerations
As there are species- and strain-related differences in feeding behaviors, there can also be variation in responses to food scheduling, so the amount and pattern of food regulation necessary to induce animals to perform tasks varies. Rats and mice are meal-eaters rather than nibblers (Classen, 1994a), and rats have a circadian rhythm of feeding (Classen, 1994b) and are more likely to eat during the dark cycle (Lima et al., 1981); these factors may affect an animal’s response to a particular food schedule (Bellinger and Mendel, 1975). Rats adjust to reasonable food-restriction regimens; however, several studies indicate that mice are less resilient and that food restriction can compromise their well-being (Nelson et al., 1973). Intolerance of food restriction is aggravated by single housing and by feeding during the light cycle (Hotz et al., 1987; Van Leeuwen et al., 1997). Thus, the species and the sensory environment can affect the physiologic response to food scheduling.
Depending on the species, food regulation can have secondary effects on research. For example, in rats, scheduled access to food can result in an increase
in the self-administration of drugs that are perceived by the subject as reinforcing agents (Laties, 1987). Diet-restricted rats may also be more agitated during restraint (Albee et al., 1987). Rats can exhibit hemoconcentration that is directly related to the degree of food restriction (Levin et al., 1993). Other physiologic alterations observed in meal-fed rats include reduced white-cell counts, reduced platelet counts, reduced serum protein concentration, increased serum bilirubin, decreased cholesterol (females only), imbalances in electrolytes, reduced hematopoietic tissue in sternal bone marrow, and, in severely restricted rats, bone marrow necrosis, thymic atrophy, and mild testicular degeneration (Levin et al., 1993). Nonhuman primates that are fed a calorie-restricted diet have a reduced bone mass, slightly lower body temperature, and increased glucose tolerance and insulin sensitivity (Roth et al., 2000).
Influence of Circadian Rhythm
The circadian rhythms of several physiologic and behavioral variables are affected when access to food is limited to particular times of the day. Rats that are meal-fed shift their activity pattern relative to the timing of meal presentation in such a way that periods of quiescence and activity are anchored to the time of day of feeding (Classen, 1994b). Factors that are affected in the mouse include the circadian rhythm of plasma corticosterone concentrations and core body temperature (Classen, 1994b).
Physiologic Mechanisms of Fluid Consumption
Three main physiologic stimuli mediate thirst, fluid consumption, and hydration in normal animals maintained on an ad libitum fluid-consumption schedule (Rolls and Rolls, 1981; Rolls et al., 1980; Stricker and Verbalis, 1988; Toth and Gardiner, 2000; Wood et al., 1982). The first is cellular dehydration, which may result from inadequate water consumption, excessive renal or evaporative water loss, or ingestion of excessive quantities of solutes, such as sodium. Fluid consumption maintains osmotic balance by reducing the extracellular solute concentration and thus allowing fluid to move back into the cells to restore intracellular fluid volume. The second is hypovolemic thirst, which occurs when fluid is lost from the blood, as occurs during dehydration. The hypovolemic status of an animal can be ascertained by measuring hematocrit or plasma protein, which increase when fluid is lost from the plasma (Toth and Gardiner, 2000). The restoration of normal plasma volumes requires solute ingestion concurrent with fluid consumption because fluid consumed in the absence of solutes moves into the intracellular compartment. The third stimulus is the hormone angiotensin, which stimulates drinking in some physiologic states.
Mechanisms for fluid conservation are activated in the kidneys during cellular dehydration or hypovolemia. The progression of dehydration over consecutive 12-hour periods is nonlinear because mechanisms for fluid conservation are invoked progressively to retard water loss (Toth and Gardiner, 2000). There are species-specific differences in the degree of fluid loss from cellular and plasma compartments (Rolls et al., 1980), which affect the degree of cellular dehydration or hypovolemia that occurs because of fluid deprivation. Especially with small animals (NIH, 2002), there is a potential for dehydration as a result of fluid loss from both intracellular and plasma compartments (Toth and Gardiner, 2000). In rats, dogs, and monkeys, cellular dehydration and hypovolemia are the primary physiologic variables that mediate fluid consumption (Fitzsimons, 1998), and species-specific differences in drinking after fluid deprivation are apparent.
Determination of Minimum Fluid Consumption
It is difficult to specify minimum fluid requirements for the various animal species, because there is a dearth of evidence in the scientific literature. That contrasts with the growing literature on the health consequences of caloric restriction.
Physiologic needs for water are influenced by many factors, including the water and electrolyte content of the diet, the ambient temperature and humidity, and exercise. Fluid consumption is also influenced by nonhydration variables, such as habit, social factors, palatability, and ease of access to fluids. Those variables tend to increase the average daily consumption of fluids to more than is necessary to maintain homeostasis (Nicolaidis and Rowland, 1975; Rowland and Flamm, 1977). Attempts to estimate a socially housed animal’s daily fluid needs on the basis of “everyday experience” are likely to lead to inflated estimates because much fluid consumption is motivated by social or other variables rather than by hydration needs (Mountcastle, 1980).
Fluid maintenance requirements vary markedly across species; fluid maintenance requirements range from 35 to 140 mL/kg of body weight (BW) per day (Aiello, 1998a; Kirk and Bistner, 1985; NRC, 1995; Wells et al., 1993; Wood et al., 1982). There can also be a wide range of ad libitum consumption levels within a species; for example, daily fluid consumption in nonhuman primates has been reported at 75 mL/kg BW (Kerr, 1972; Wayner, 1964), 90 mL/kg BW (Wayner, 1964), and 110 mL/kg BW (Evans, 1990). There can also be a wide range within a strain; for example, daily ad libitum fluid consumption in 3 different Spraque-Dawley rat colonies was reported at 80 mL/kg BW, 105 mL/kg BW, and 125 mL/kg BW (Wells et al., 1993). There can even be significant gender differences; for example, daily ad libitum fluid consumption in male golden hamsters was reported at 50 mL/kg BW, while female golden hamsters consumed 140 mL/kg BW (Fitts and St Dennis, 1981). Consequently, assessing the ad
libitum fluid consumption for each fluid-regulated experimental animal might be an important step in ensuring the health and well-being of the animal.
However, voluntary fluid-consumption levels in a laboratory setting might not be equivalent to the animal’s minimal fluid-requirement levels. Limited availability of fluid is a common determinant of consumption in natural settings, and physiologic and behavioral mechanisms have evolved to enable animals to adapt to the limitation. For example, rats and monkeys quickly learn to consume much, if not all, of their daily fluid needs in a short, restricted period (reviewed by Evans, 1990). Species that drink from watering sites only once per day invoke homeostatic mechanisms to control urine output in relation to their hydration state (Toth and Gardiner, 2000). Mammals may also use torpor to adapt to the dry season in their natural habitat (Schmid and Speakman, 2000). Thus, it is difficult to designate specific minimum fluid needs, because requirements may vary with species, strain, environment, efficiency of fluid-saving mechanisms, and so on. IACUCs, veterinarians, and researchers should take into account the possibility that laboratory animals can be adequately physiologically sustained with less fluid than they would voluntarily consume.
At the start of a new research protocol involving restricted or altered access to fluid, the amount of fluid consumed, body weight, and a hydration assessment should be recorded daily for each animal, as individual animals may manifest physiologic and behavior differences. Those data will help in refining the protocol and evaluating the adequacy of access to fluid. In evaluating the adequacy of access to fluid, each animal should be evaluated individually to determine how it is adapting to the imposition of restricted or altered access. For example, if an animal attains and then maintains a new body weight, it could suggest successful adaptation even if the new weight is below the weight recorded during ad libitum access to fluid.
When fluid regulation is selected as a behavioral motivator, access to fluid outside the experimental setting has to be regulated to motivate performance of the rewarded behavior (NIH, 2002). Generally, fluid regulation is patterned after one of two designs. In “fluid restriction,” animals are given access to a metered volume of fluid per day and may consume that volume over any length of time. In “fluid scheduling,” the experimenter determines the time of day during which the animal has access to fluid, but the duration of drinking and the volume consumed are determined by the behavior of the animal. For example, in many behavioral protocols, animals are given continual access to fluid for as long as they continue to perform a task. Because food and fluid generally are not freely available in the wild and some effort (foraging) is required to obtain them (NIH, 2002), such scheduling designs may model the effort expenditure necessary to obtain food and fluid in the wild.
For both types of fluid regulation, animals generally should be given free access to fluid for some period on days when experimental sessions are not scheduled, unless scientifically justifiable reasons preclude such fluid supplementation (NIH, 2002).
When developing a restriction design to motivate an animal to perform a task, the main consideration is determining what level of restriction is necessary to achieve the desired performance. Generally, the more complicated the task, the more stringent the restriction protocol needs to be. For example, in a study of water-restricted rats, where the rats were required to bar press to obtain their daily allotment of water (Collier and Levitsky, 1967), mild restriction (rats receive 75% of their average ad libitum intake) resulted in poor performance while a more stringent restriction (rats receive 32% of their average ad libitum intake) resulted in maximal performance. This is why fluid restriction levels used in one study may not provide adequate motivation for learning or performing other more demanding tasks (Toth and Gardiner, 2000). However, the most severe restriction is not always necessary for achieving maximal performance. In this same study (Collier and Levitsky, 1967), bar-press rates were similar when water was restricted to 32%, 42%, or 56% of average ad libitum intake.
The health implications of fluid regulation have been of concern, although even chronic restriction schedules have not been found to cause physiologic impairment of animals that are adapted to the restriction and receive enough fluid to replenish daily losses (Toth and Gardiner, 2000). For example, the consequences of using fluid scheduling to motivate lever-pressing behavior have been examined in rats deprived of fluid for 7, 14, or 21 hours/day for 3 months. The animals showed no observable adverse effects compared with ad libitum controls with respect to weight loss, organ and tissue appearance at necropsy, hematologic examination, or clinical chemical analysis (Hughes et al., 1994).
Species- and Strain-Specific Considerations
The development of fluid-regulation schedules should include some consideration of species variations in fluid consumption behaviors. Some species consume fluid intermittently, sometimes only once per day, but others consume smaller quantities more frequently (Rolls et al., 1980). Efforts should be made to match an animal’s typical watering schedules with circadian variables, because the risk to animals on fluid regulation is reduced if periods of access and total amounts available are appropriate to the species (NRC, 1995; Toth and Gardiner, 2000). In addition to the behavioral aspects of fluid consumption, relationships between fluid intake and food intake should be considered. Food ideally is provided at close to the same time as daily fluid provision (for example, after the experimental session):
The concurrent availability of water and food incurs two benefits. First, fluid intake promotes food intake, thereby reducing the likelihood of dehydration-
related anorexia. Second, the consumption of food associated with water allows animals to consume solutes that will help retain water in the circulation, correct volume deficits, and avoid excessive hemodilution that will cause urinary excretion of the ingested water (Toth and Gardiner, 2000).
Other Influences on Fluid Homeostasis
In some situations, fluid reinforcers (such as fruit juice) are used because they may maintain behavioral performance when access to fluid is restricted; for example, some monkeys prefer fruit juice when performing long behavioral sessions in which many reinforcements are delivered (NIH, 2002). Investigators, veterinary personnel, and IACUCs should consider and monitor for any potential physiologic ramifications of total substitution of solute-containing fluids for water in a fluid-restricted protocol. Sweetened milk or juices may be unfavorable choices for use in a long-term study in which an animal will participate for many months or years, because of the potential for dental caries (NIH, 2002).
Provision of treats, such as fruits or vegetables, is recommended when appropriate to provide variety and nutritional balance to an experimental animal’s diet (NRC, 1996). The water content of these dietary supplements can be difficult to estimate, so their potential contribution to hydration should not be considered in determining the minimal ration of fluids to be given to the animal (see Pennington et al., 1998, for data on water content of fruit and vegetable supplements). However, investigators, veterinary personnel, and caretakers should be aware of the potential need for restriction or substitution of supplemental food items in fluid-regulated animals.
Variability Between Individuals
When presented with the homeostatic challenge of dehydration, animals can respond by conserving water and excreting concentrated urine (physiologic regulators) and/or by drinking more fluid and excreting dilute urine (behavioral regulators) (Kanter, 1953; Toth and Gardiner, 2000). Animals on fluid restriction or scheduling protocols may implement different compensatory mechanisms to different extents. Animals that are physiologic regulators may be problematic when used in behavioral studies, because they often tend to accommodate to the consumption of a minimal volume of fluid by excreting more concentrated urine instead of consuming more fluid (Toth and Gardiner, 2000). In contrast, the behavioral regulator tends to modify its behavior during the experimental task to obtain more fluid as a reward. In both instances, nonhuman primates often supplement fluid consumption by licking water from cages after washing. Therefore, the assessment of each animal on a fluid-regulation protocol is prudent.
The difficulty of the behavioral task that an animal must learn and the goals of the experiment often influence the degree to which the animal must be motivated to perform the task and thus the degree of fluid regulation that is necessary. When possible, palatable rewards rather than regulation should be used to motivate behavior. However, if fluid regulation is determined to be the preferred method of motivating a particular behavior, consultation with veterinary personnel and a review of recent literature regarding animal training may be appropriate (NIH, 2002).
In training of a naïve subject to perform a new task with a fluid regulation, gradual introduction to the concept that fluid availability is restricted or context-dependent (for example, earned while in the experimental apparatus) is important (Toth and Gardiner, 2000). After the animal has experienced the absence of ad libitum fluid, its motivation to learn or perform tasks to earn fluid usually increases. The degree of restriction may require periodic adjustment to generate adequate motivation to learn or perform difficult phases of a task (Toth and Gardiner, 2000). However, the restriction often may be reduced after the animal learns the task and becomes proficient at it. As noted in Methods and Welfare Considerations in Behavioral Research with Animals (NIH, 2002):
When the study begins, be prepared to consider and address a range of behavioral, environmental, or equipment-related variables that might hinder training or disrupt performance. Inexperienced personnel may presume that a source of problems in training or maintaining a food- or fluid-motivated behavior is that the restriction is not strict enough (or, in some cases, that it is too strict). The other types of variables that should be considered first, however, are equipment malfunctions, programming errors, task criteria that are raised rapidly or set too high for the animal’s level of training, illness, or nonprogrammed water restriction (in the case of food-motivated behavior).
Furthermore, experimental animals, like humans, may have deficits, such as myopia, that impair performance on tasks because of perceptual limitations. “In all circumstances, careful monitoring of animals under food or fluid control is necessary every day to avoid additional nonprogrammed restriction” (NIH, 2002).
Assessment of Animal Subjects as Individuals
The previous paragraphs emphasize that animals on food or fluid regulation schedules are individuals whose performance is likely to vary from day to day. Variations between individual animals in performance on a given task are also expected. The differences between individuals and even within an individual during different phases of an experiment may occasionally necessitate some adjustment of food or fluid scheduling to maintain homeostatic balance and achieve the desired experimental goals. The frequency of observations should therefore
be adjusted according to how fast an animal can be compromised in the experimental situation. Diligent record keeping on daily food or fluid volume consumed, hydration status, appearance, general affect, experimental performance, and routine weighing are reliable for identifying changes in behavior patterns. Those records should be reviewed regularly and kept readily accessible to the veterinary staff and others who may have a need to evaluate them, such as the IACUC during its semiannual inspections. The need for intervention or reassessment of the hydration needs of an experimental animal can thus be recognized before adverse physiologic consequences develop.
Methods of Assessing Nutrition and Hydration
A system of daily monitoring procedures is essential for animals that are food or fluid regulated. Records should be kept of the amount of food or fluid earned in the behavior task as well as any supplements given. Careful observation of the animal’s behavior and regular clinical monitoring of the animal’s health are critical to ensuring successful application of food or fluid regulation (NIH, 2002).
Clinical monitoring should include assessments of the nutritional and hydration state of experimental animals whose access to food or fluid is regulated. There are various methods for assessing a food- or water-regulated animal and reliance on a single variable is discouraged. Instead, investigators, caretakers, and veterinary staff should use several methods concurrently to ensure the health and well-being of the animals. Variables that can be monitored to assess the nutritional or hydration status of experimental subjects include the following.
Weight and food intake
Experimental animals on food or fluid regulation should be weighed several times a week, ideally before experimental sessions (NIH, 2002). Some accommodations in the frequency of weighing may be necessary if experimental animals require sedation or anesthesia to be weighed (NIH, 2002). Often, animals can be trained to cooperate with the procedures. Despite conditioning, however, the process of weighing may be very stressful to some experimental animals. In such cases, an animal can be weighed less frequently, and other reliable methods of hydration monitoring can be used.
Aside from daily fluctuations in weight due to fluid gain and loss, animals on fluid regulation may lose weight as a result of decreased food ingestion. Using a percentage weight-loss criterion during fluid deprivation as an endpoint for determining when an animal should be removed from a fluid restriction paradigm and their fluid requirements reassessed can indicate not only a proper level of motivation, but also health (Bolles, 1975). The amount of food consumed by experimental animals is a good measure of general health and of hydration status and should be monitored by caretakers or investigators, or
both. Persistent decreases in food consumption should be brought to the attention of appropriate veterinary personnel.
The texture and elasticity of skin are important indicators of an animal’s hydration status. Ordinarily, dehydration will cause a slow return of the skin to its normal position after it has been lifted. However, that characteristic is less reliable in obese animals because their skin tends to maintain its elasticity even in the presence of dehydration (Kirk and Bistner, 1985).
Solid and fluid waste output and moistness of feces
As part of the long-term adaptation to fluid restriction, healthy animals produce concentrated urine and feces that are less moist than normal (Toth and Gardiner, 2000). Regular observation of the quantity and qualities of the excrement produced by an animal on a fluid or food regulation provides information about both hydration status and physiologic compensation for fluid regulation.
General appearance and demeanor and quality of fur and skin
Investigators and veterinary personnel share the responsibility for observing behavior, general appearance, and demeanor of experimental animals, which can be valuable indicators of their health status. For example, dry mucus membranes and sunken eyes are indications of dehydration (Aiello, 1998a). If signs indicate that an animal is developing problems related to dehydration, hemoglobin content or hematocrit and blood urea nitrogen can be measured to determine its physiologic status.
A plan of action, complete with endpoints for therapeutic intervention, should be considered when the experimental animal protocol is being developed. The plan should include standard operating procedures to be used if an animal develops diarrhea or vomiting that would prompt the return to an ad libitum fluid or food schedule and application of a schedule for veterinary evaluation to prevent serious health consequences due to dehydration or malnutrition.
GENETICALLY MODIFIED ANIMALS
Genetically modified animals have induced mutations that are human-made alterations in their genetic code. The generic phrase genetically modified includes both transgenic and targeted mutations that are created to study the expression, overexpression, or underexpression of a specific gene (ARENA-OLAW, 2002). A transgenic animal has genes from another organism or species incorporated into its genome, whereas an animal with a targeted mutation has had the coding sequence of a gene in its own genome altered. For a genetic modification to be useful in research animals, the introduced or altered gene must be transmitted to
the offspring. Most induced mutations have been made in laboratory strains of mice (Mus) or rats (Rattus). Although mice are used as examples in the following discussion, the general considerations are applicable to induced mutants of any species (ARENA-OLAW, 2002).
Genetically modified animals are used to test hypotheses in several ways: the phenotype of the modified animal is evaluated to determine the pathogenesis of disease or gene influences on development, the modified animal is used to test interventions to treat its condition, or the animal is used as a tool to study the pathogenesis of other conditions.
A transgenic animal has exogenous (foreign) deoxyribonucleic acid (DNA) inserted into its cells. Typically, transgenic animals are created by the ”pronucleus method,” in which complimentary deoxyribonucleic acid (cDNA) made from specific messenger ribonucleic acid (mRNA) is inserted into cells by using microinjection, electroporation, or nonpathogenic viruses. Each of those methods has been used to insert new DNA into the pronucleus of a fertilized mouse egg to create viable transgenic mice. The manipulated fertilized eggs may be cultured in vitro for several days before they are surgically implanted into the oviducts or uterus of pseudopregnant female mice. The successful production of a transgenic animal will be affected by several events: the inserted DNA will incorporate into the chromosomes of only a percentage of the embryos developing from the microinjected eggs; the DNA will incorporate at different genetic locations; and different numbers of copies of the DNA will incorporate in different embryos. Therefore each embryo has the potential to become a unique transgenic mouse even though the same quantity and type of DNA was injected into genetically identical fertilized eggs. Not all manipulated, fertilized eggs become live-born transgenic mice. Losses occur at every step from injection through gestation and delivery (ARENA-OLAW, 2002).
Although an individual mouse may carry transgenes, it cannot transmit the transgene to its offspring, unless the cDNA incorporates into germ cells. A “founder” is a mouse that passes the transgene to its descendants. Thus, many fertilized eggs must be microinjected to obtain a few transgenic mice, and only a few of the transgenic mice will be founders of a particular transgenic line (ARENA-OLAW, 2002).
Knockout and Knockin Mutants: Animals with Targeted Mutations
Targeted mutation refers to a process whereby a specific gene is made non-functional (“knockout”) or, less frequently, made functional (“knockin”). A mouse with a targeted disruption, or knockout, of a specific gene is typically created through the embryonic-stem-cell method. This arduous method requires
the occurrence of several low-probability events. First, the gene in question must be identified, targeted, and marked precisely. This has been accomplished for an astounding number of murine genes during the last several years (Harris and Ford, 2000; Takahashi et al., 1994). Second, mouse embryonic stem cells must be harvested and cultured. Third, a mutated form of the gene of interest is created (the mutation, or altered order of nucleotides, renders the gene inactive). Fourth, the mutated gene is introduced into the cultured stem cells by using microinjection or electroporation transfection (Tonegawa, 1994); a very small number of the altered genes will be incorporated into the DNA of the stem cells through recombination (Sedivy and Sharp, 1989). Fifth, the mutated embryonic stem cells are inserted into otherwise normal mouse embryos (blastocysts), which are then implanted into a surrogate mother (Boggs, 1990; Le Mouellic et al., 1990; Steeghs et al., 1995). All the descendant cells from the mutated stem cells will have the altered gene; the descendants of the original blastocyst cells will have normal genes. Thus, the newborn animals will have some cells that possess only a copy of the mutant gene and some cells that only possess the normal (wild-type) gene. This type of animal is called a chimera. If the mutated stem cells are incorporated into the germ line (the cells destined to become sperm or ova), some of the gametes will contain the mutant gene. If the chimera is bred with wild-type mice, some of the offspring will be heterozygous for the mutation (possess one copy of the mutant gene). If the heterozygous mice are interbred, about one-fourth of their offspring will be homozygous for the mutation. The homozygous mice become the founders and can be interbred to produce pure lines of mice with the gene of interest “knocked out” (Galli-Taliadoros et al., 1995). As a result, the product that the gene typically encodes will be missing from the progeny (Sedivy and Sharp, 1989).
For technical reasons, most of the stem cells used in targeted-gene deletion studies were derived from mice of the 129/SV strain. The 129/SV stem cells were typically implanted into C57BL/6 blastocysts (Soriano, 1995). The resulting “mixed” offspring are often backcrossed to the C57BL/6 (background) strain. After 10 backcrosses, the mutated strain is considered a congenic strain, identical with the C57BL/6 background strain except at the site of the altered gene.
There are several important advantages of using knockout mice: (1) disabling a gene often results in a precise and “clean” ablation, (2) the effects of the gene product can be abolished without the side effects of drugs, and (3) genetic manipulation may be the only way to determine the precise role of the gene product particularly in behavior. The use of new inducible knockouts, in which the timing and placement of the targeted gene disruption can be controlled, will refine and extend the usefulness of genetically modified animals in neuroscience and behavioral research.
One drawback in the use of knockout animals is lethal mutations. The products of many genes are essential for normal function, and inactivating a gene may prove lethal because of gross morphologic or physiologic abnormalities. For
example, knockout mice with targeted disruption of either the parathyroid hormone-related peptide, ß-1 integrin subunit, or ß-glucocerebrosidase genes die in utero or immediately after birth (Karaplis et al., 1994; Stephens et al., 1995).
Development of Animal Protocols Involving Genetically Modified Animals
The first step in developing a protocol to produce or use genetically modified animals is to determine the disease profile that any particular animal or group of animals is likely to experience during the course of normal life or as a result of experimental use. Some genetically modified animals are created to develop a disease spontaneously, but others may develop a severe or debilitating disease even if the disease is not the intended outcome.
Genetically modified animals are used in a wide array of experimental studies. They can be used in studies of the pathogenesis and therapy of a primary disease, of a concurrent disease or associated clinical problems, or of aging and longevity. However, many of the animals will never be used in experimental studies but rather are maintained as breeders. The PI, IACUC, and veterinarian all need to develop a general health profile of a given strain that is relevant even to nonexperimental animals (breeders and animals intended for but not yet included in a study). The potential adverse effects of the genetic modification itself have to be considered.
For an established strain of genetically modified animal, the literature may provide a good description of the expected phenotype and the course of its development. However, the full repertoire of a gene’s effects may not be envisioned, or a gene’s functions may be unknown at the time a knockout or transgenic animal is created. That uncertainty can make the health-related consequences of developing a knockout or transgenic strain difficult to predict. Many modified mice are generated and maintained for purposes of discovery rather than hypothesis-testing. An example of the “discovery” approach is the use of random mutagenesis, which may create animals whose individual phenotypes are theoretically unpredictable. Because some of the new strains of mice may spontaneously develop problems that are painful or debilitating, assessment strategies and endpoints for these animals must be considered before their generation or their experimental use. Such information is typically solicited by the IACUC as part of the animal-use protocol evaluation.
When submitting an animal-use protocol to develop a genetically modified animal, neuroscientists must include an estimate of the number of animals to be
used (not including experimental manipulations). Determining the number can be challenging because the process to develop a genetically modified animal is subject to unpredictable outcomes. A detailed presentation of a method for estimating the number of animals needed to develop a genetically modified founder mouse can be found in Appendix B.
After founder mice have been identified, 80–100 mice may be needed to maintain and characterize a line. That assumes that up to five breeder pairs per line are needed, that there is no unusual infertility, and that adequate numbers of weanlings are produced for genotypic and phenotypic characterization (ARENA-OLAW, 2002) Appendix B also contains extensive information on calculating animal estimates for colony breeding and experimental use.
Breeding of a congenic strain by using “speed congenics” requires a significant number of animals. Speed congenics is the process by which the DNA of each mutated animal is screened to select animals with the most genetic similarity to the background strain; this reduces the number of back-crosses necessary to develop the congenic strain. Usually, at least 750 mice are required, assuming a breeding colony of 10–12 breeding pairs and adequate progeny for phenotypic and genotypic characterization. If the homozygous mutant is infertile, the congenic strain must be developed by using intercross matings, and the number of mice needed is about 1,200 (ARENA-OLAW, 2002).
Because development and maintenance of genetically modified animal colonies require large numbers of animals, animals may be produced that are determined not to be useful for a particular project. Those animals may be useful for another project and should either be transferred to that project or culled from the colony.
The debility that genetically modified animals may experience is a cause of concern. It is important to provide as much support and comfort for mutant animals as possible. Some strains may require specific husbandry interventions to enable or promote well-being. For example, mice with targeted deletion of the gene for neuronal nitric oxide synthase (NOS-1 -/-) develop defects that model the clinical idiopathic voiding disorders that can affect to 10–15% of men and women. These mice have hypertrophic dilated bladders, dysfunctional urinary outlets, and increased urinary frequency (Burnett et al., 1997). They require extra bedding and more frequent cage changes than wild-type mice. Other examples of special husbandry interventions are those prone to audiogenic seizures, which must be housed in quiet environments, and those with ataxia or paralysis, which may require special provisions to enable access to food and water.
Close scrutiny of genetically modified animals during routine daily observation by the animal-care personnel may be warranted. Animal-care personnel often discover disabilities and abnormalities in genetically modified animals (such
as motor deficits or anorexia) and should be trained to recognize them. Additional training of the animal-care staff to include practical information on the special needs and common problems associated with specific strains is recommended.
The dramatic growth in the use of genetically modified rodents, primarily mice, creates substantial challenges for timely and effective assessment of animal health and well-being. Many institutions house large populations of genetically modified mice with a wide array of deficits that affect physiologic homeostasis and behavior. The popularity of high-density, individually ventilated caging systems for housing these valuable mice adds barriers and challenges for effective observation and increases the importance of a careful and systematic examination of individual animals during scheduled cage-maintenance activities.
General Health Assessment
The general health of novel genetically modified animals should be assessed soon after their availability and before the onset of complex behavioral analyses (Crawley, 1999). Identifying potential health problems early is critical to providing appropriate care. Undetected health problems can also skew the assessment of more complex behaviors—such as learning and memory, aggression, mating, and parenting, so it is essential to identify problems before behavioral phenotyping (see “Behavioral Screening of Genetically Modified Animals” in Chapter 9).
For mice, a general health assessment starts with a brief evaluation of body mass, core body temperature, and appearance of the pelage (fur). Neurologic reflexes should be assessed, including the righting reflex, the eye blink, and the ear and whisker twitch in response to tactile stimuli (Crawley, 1999). Any of the following symptoms should be recorded, treated if necessary, and considered when behavioral phenotyping is later conducted: self-mutilation, guarding, vocalization (with or without stimuli associated with pain), hunched posture, inactivity, lethargy, rough hair coat, no response to mild stimuli, increased heart or respiratory rate, anorexia for longer than 24 hours, weight loss greater than 20%, decrease in weight gain compared with aged-matched controls, and lesions (such as swelling, redness, and abnormal discharges). Any obvious deviations from the typical naturally occurring behaviors (ethogram) of mice should be noted. The mouse ethogram includes such behaviors as sleeping, resting, locomotion, grooming, ingestion of food and water, nest-building, exploration, foraging, and fear, anxiety, and defensive behavior (Brown et al., 2000).
After an initial health assessment, daily observation of the genetically modified animal should include an assessment of general activity levels, posture, haircoat condition, the presence of scratching or self-mutilation, and the general condition of the cage. When the cage is manipulated, as during cleaning, animals can be more closely examined for additional characteristics, such as response to handling; unexpected vocalization; ulceration; masses; abnormalities of the eyes,
ears, nose, and mouth; palpable hyperthermia or hypothermia; and general body condition.
Subsequent to the general health assessment, sensory and motor testing should be carried out followed by behavioral testing (e.g., anxiety behaviors). Behavioral assessment should proceed as soon as sufficient numbers of transgenic animals are available to identify sensory, motor, or motivational deficits that may compromise the well-being of the animals. Behavioral screening is discussed at length in Chapter 9, “Behavioral Screening of Genetically Modified Animals.”
Pain, Distress, and Endpoints
The elimination of all pain and distress from all affected animals is unlikely, inasmuch as the diseases being modeled in genetically modified animals are often associated with pain or distress that cannot be relieved in human patients. Achieving a balance between animal well-being and research objectives is essential to obtaining valid answers to questions about the causes, treatment, and preventions of such diseases in humans.
When a neuroscientist initiates assessment of a new genetically modified animal, information about clinical abnormalities associated with the phenotype and special husbandry requirements usually are not available. The investigator must, however, include general humane endpoints in case a severe debilitating phenotype develops and should provide the IACUC with this information in writing when the new mutant has been developed or at the next annual review of the animal-use protocol.
When a genetically modified animal strain has been characterized, the standard of “normal” for a mutant animal may or may not be the same as that for a nonmutant animal (ARENA-OLAW, 2002); therefore, clinical signs that would be used as humane endpoints in normal animals may be inappropriate as endpoints in genetically modified animals. If the mutant phenotype does not affect the general welfare or clinical well-being of an animal, the same standard of “normal” may be used for mutant and nonmutant animals. In the case of mutants whose phenotype involves clinical abnormalities, the standard for “normal” may have to be modified to include the expected phenotype. For example, 8-month-old mice lacking the gene for a key enzyme that encodes ganglioside biosynthesis (GM2/GD2 synthase) develop substantial neuropathologies, motor incoordination and an abnormal gait (Chiavegatto et al., 2000). As these mice age, muscular weakness progresses, and the standard of “normal” for GM2/GD2 synthase knockout mice includes difficulties in locomotion, which in a nongenetically modified animal might be one criterion of a humane endpoint.
Humane endpoints for mutant animals should be established on the basis of the ability of the mutant to access and consume food and water, the response of the mutant to stimuli, and the general condition of the mutant (for example, it is excessively underweight, it shows progressive weight loss, it doesn’t groom, it has a hunched posture, or it has sensorimotor deficits).
The specific use of a genetically modified animal will influence the type of endpoint that is described in the animal-use protocol and the circumstances in which an endpoint decision will be implemented. For example, an animal that develops a clinical problem while in a study of the prevention of disease development could potentially be euthanized earlier than one involved in a study of disease therapy. A nonexperimental animal (a breeder or an animal intended for but not yet part of a study) that develops a substantial clinical problem should be euthanized. A maximum holding period should be set to avoid the development of predictable problems in strains of mice that have debilitating phenotypes.
Endpoint issues generally apply to the entire life of genetically modified animals. Therefore, endpoints become relevant both in the context of experimental procedures and with regard to the potential pain or distress that is caused by the genetic modification itself. Care must be taken to provide general endpoints in the animal-use protocol for the period in which the initial colony is being developed and the phenotype of the animals is first characterized, as well as for experimental and nonexperimental animals.