Introduction: Use of house mice in biomedical research. ILAR 41(3), 133.
Abstract: The mouse's usefulness as a biomedical model has been greatly enhanced due to the ability to experimentally modify its genome. The English word mouse can be traced back to the Latin mus, then to the Greek mys, and finally to the ancient Sanskirt mush, which means "to steal." Mice are found as commensals in diverse human structured habitats throughout the world and as feral mice in a wide variety of natural habitats. In Europe as early as 1614, the mouse was used by Robert Hooke for studies of oxygen in living systems. At about the turn of the century, mouse fanciers in England and the US were breeding mice for unusual coat color and other characteristics. In the US, "fancy" mice bred by Abbie Lathrop became part of the biologic research programs at Harvard and at the University of Pennsylvania when it was noted that these strains developed tumors. Clarence Little and others developed strains from Miss Lathrop's original colony which include the now commonly used CBA, C3H, C57BL/6 and BALB/c strains. These mice were used to clarify the genetic basis of cancer. As specific strains developed, the use of mice became more common. Clarence Little eventually was involved in founding The Jackson Laboratory and served as its first director. The house mouse, Mus musculus domesticus, has become the mammalian model of choice in research because of its high degree of adaptability and because highly inbred strains became available with traits relevant to human diseases. For example, the mouse 129 strain and its substrains are particularly suitable for derivation of embryonic stem cells that can be genetically manipulated in culture. This is the foundation for "targeted mutation" technology to produce mice with specific phenotypes. Other rodents such as the Norway rat also became common laboratory animal models due to its fertility, size, and tractability. The rat became one of the most frequently used species in behavioral studies, especially those involving learning. However, inbred mouse strains have been used more extensively than any other laboratory mammal to uncover mammalian genes related to specific phenotypes. More information is needed on the behavior of mice so that behavioral phenotypes altered by genetic manipulation can be detected and measured. Complex phenotypes are often altered in unexpected ways by a targeted gene alteration and one must always be aware of potential confounding variables, especially with respect to behavior. Because behavior is at the apex of underlying genetic, developmental, and experimental factors, reduction of variability of these factors will enhance the sensitivity of behavioral tests.
Questions: The questions listed below are some questions I wrote for my study group last year to help us review mouse nomenclature. As I couldn't come up with good questions on this short article, I thought I would submit these questions to the group instead. Many of the mouse strains and stocks used are from the current literature or the ILAR database.
1. Which part of the general transgenic nomenclature format TgX(YYYYYY)#####Zzz refers to the laboratory-assigned number?
e. c and d
2. Which of the following listed below refers to an outbred stock?
3. Which of the following listed below refers to a consomic inbred strain?
4. Which of the following listed below refers to a segregating inbred strain?
5. Which of the following listed below refers to a congenic strain?
6. How was TgN(TcrLCMV)327Sdz most likely created?
a. ES cell
b. DNA pronuclear injection
c. Retroviral vector
e. Spontaneous mutation
7. Which of the following listed below refers to a hybrid mouse strain?
8. Which of the following inbred strains does not breed true?
9. True or False. A conplastic strain is a variation of congenic but a congenic strain is not always conplastic.
10. True or False. The mitochondrial parent in a conplastic strain designation NUCLEAR GENOME-mtCYTOPLASMIC GENOME is always male.
Answers: 1. c. TgX refers to mode, YYYYYY refers to the insert designation, Zzz refers to the laboratory code.
2. a. The big clue is the colon in the designation. Stocks are designated by the name of the holder followed by a colon and then the name of the strain.
3. c. Consomic means differing by a whole chromosome, not a genetic region (congenic) or genetic locus (coisogenic). Designation format is HOST STRAIN-CHROMOSOMEDONOR STRAIN.
4. b. Segregating inbred strains are designated like coisogenic strains (STRAIN NAME-gene name or allele) except the gene locus is maintained in the heterozygous state. Answer a is coisogenic but homozygous. Answer d is congenic.
5. a. Look for the . and combination. HOST STRAIN.DONOR STRAIN-genetic region
6. b. This transgenic mouse was generated by nonhomologous (N) recombination. ES cell is homologous recombination. Retroviral vector would be designated TgR.
7. c. The big clue is the "F1" at the end. The female parent is listed first (B6) and then the male (C3) followed by F1.
8. b. Hybrid strains do not breed true and can only be maintained as long as the 2 parental strains exist.
9. True. Conplastic is a variation of congenic.
10. False. The mitochondrial parent is always female.
Behavioral phenotyping of transgenic and knockout mice: experimental design and evaluation of general health, sensory functions, motor abilities, and specific behavioral tests. ILAR 41(3), 136.
Abstract: This article summarizes some of the considerations and tests used in the authors' lab to evaluate sensory and motor function and behavior in gene-targeted mutant mice.
Motor and Sensory Function - open field locomotion, rotarod, visual cliff test, acoustic and tactile startle test, olfactory test (Maybe most important is a good initial physical exam and careful observation of the animal in its home cage.)
Learning and Memory Tasks - Morris swim tasks; fear-conditioning tasks; passive and active avoidance tasks; spatial maze tasks (T-maze, Y-maze, radial mazes, Barnes maze); operant tasks (ex. lever press acquisition); motor learning (ex. repeated sessions on the Rotarod)
Pain Sensitivity Tests - tail-flick, hot-plate, footshock
Anxiety-related behaviors - light/dark exploration, elevated plus maze
Depression-related behaviors - Porsolt swim test
Schizophrenia-related behavior - pre-pulse inhibition
Social behaviors - group huddling, nest building, social dominance, aggression
Sexual and Reproductive Behavior - male, female, parental
Responsivity to Addictive Drugs - IV or drinking water self administration
Problems with the current technology (specifically relating to the study of genes and behavior) include: 1) the mutation is present in all cells, therefore it is not possible to assign a behavioral difference to a specific area of the brain of even to the nervous system and 2) the mutation is present in the animal from the earliest stage of development which may result in compensation by other genes. These problems may be alleviated by developing tissue-specific "conditional" knockouts or temporally-selective "inducible" transgenics.
Questions: 1. What are Von Frey hairs (or filaments)?
2. If you put two animals (unfamiliar to each other) into an unfamiliar environment (i.e. a clean cage), they have a choice of first investigating each other or investigating the new environment. How do rats and mice respond?
Answers: 1. Von Frey hairs are sets of thin wires or plastic filaments used to test the sense of touch. They are calibrated to flex at a given pressure so that a standardized stimulus can be administered.
2. Rats will usually check each other out first. Mice usually check out the new cage.
Structure and limits of animal models: examples from alcohol research. ILAR 41(3), 144.
Abstract: This article is focused on the basics of appropriate animal model selection in investigating alcohol-related processes that are thought to be part of the complex human problem of alcohol use, abuse and alcoholism. The article focuses on the mouse as the current popular animal model of choice.
Following decades of biochemical, pharmacologic, physiologic, behavioral, and social research, the problems of alcohol use and abuse, and of alcoholism can be properly described as a complex system. The establishment of an experimental animal model is an attempt to isolate a part of this complex problem to stdy the effects of a particular elements or elements of the system. One or more elements are chosen to be manipulated, independent variables while other variables are subjected to be controlled by either fixation or randomization. One or more other variables are identified as the outcome or dependent variables to be measured, and still others are assessed as co-variants, important to the final interpretation of results.
Despite efforts made to standardize the conditions of testing and selection of measures between some groups, significant variation in research data can be attributed to the different labs. This is likely as during the study design process, a number of simple decisions must be made about experimental design, husbandry, organismal variables, and the gene pool to be sampled, any of which can impact results. To illustrate how this might be controlled to some degree, two different measures of voluntary alcohol consumption are described: "Alcohol preference" where animals choose between two different drink choices, one containing tap water and the other containing a 10% solution of ethanol, measures being collected over a 14-15 consecutive day, and the containers being rotated to obviate position preferences; and "Alcohol acceptance" consumption from a single container of 10% ethanol is recorded over a 24 hour period after a 24 hour water deprivation. Alcohol preference is thought to be used more widely by the authors, and to have more highly reproducible differences between inbred strains of mice in different labs.
In selecting animal groups for research programs, the central issues are how well the animal represents the problem being studied and how reproducible the results can be. Selecting inbred strains of mice can provide uniform and consistent genotypes across labs across the country, however this is at the cost of reduced generalizability. Systematically generated and well maintained genetically heterogenous stocks offer a large scope for generalization, but no single genotyp within such a group is replicable. Other important criteria that are critical in evaluating the adequacy of a model system are 1) validity - the degree to which the model system actually measures what it is designed to measure. Validity should target one or more features of a selected phenomenon and is the very core of the relavenace of the model. However, the issue of validity of any chosen measure is intimately entwined with how we perceive what it is we want to measure, and perceptions tend to be dynamic over time with the continued growth of the total body of empirical information. Therefore at any particular time, some degree of validity can be obtained, but the concept itself is constantly evolving.; and 2) reliability - the concept of reliability relates to the accuracy of the measurement. The concept assumes there are inherent errors in operations of measurement, an lends to an aura of uncertainty about any measured value. The more reliable measures are those with the narrowest error distribution; 3) temporal stability - since complex systems are maintained dynamically (as opposed to statically), temporal fluctuations in values of measured variables and their relationships to other variables can be expected. The concept of individual variability or "fluctuance" is garnering increasing interest as a dearth of alcohol related data illustrate the wide variability among each animal's response over a period of time. There is conjecture that individual fluctuance is due to previous alcohol exposure either in a positive or negative manner. Evidence shows that a short measurement period may only reflect the true state of the animal's response at that particular time, as the animal's response will change given time, exposure to alcohol, and homeostatic feedback. The authors suggest that fluctuance be interpreted as a measure's sensitivity to environmental nuances or to subtle changes in the animal rather than its proneness to error and conclude that the experimenter must demonstrate whether fluctuance is sufficiently small that a single measure point in time will suffice for the processes being evaluated. 4) development- introduces variability within a measurement system as the study data indicates that a model system will act and interact in quite different ways at different stages of life. An illustration of this point is that a quantitative trait locus (QTL) on mouse chromosome 15 influences mouse alcohol acceptance at about 100 days of age, but has no detectable influence at about 300 days of age. The recognition that these changes are occurring in adulthood is departure from the common presumption that all development takes place early in life.
The genetic components of several alcohol related behaviors have been measured extensively in mice, and a unique QTL pattern has emerged for each alcohol related phenotype that has been studied. However there is accumulating evidence that QTL's are associated with more than one phenotype. The complex relationship between measures of behavior and the genes underlying the behavior with alcohol consumption have not yet been explored, although comparisons have been drawn between already explored cocaine-related behavioral QLT's.
The problems with alcohol use, alcohol abuse and alcoholism clearly constitute complex systems with many influential variables that have been identified with the basic genetic, physiologic and biochemical attributes of the individual, and a broad selection of environmental circumstance in which an individual lives. The adequacy of an animal model of such a complex system can only be evaluated in the total context of the model system being used. Several aspects of generalization are pertinent to this evaluation, and no single marker variable is adequate to characterize such a complex system. Information from as many overlapping measures as possible must be made, and concepts of alcohol problems will emerge from the convergence of information derived from many animal models. The target phenomenon of a model will have complex geneic architecture, with a large number of genes affecting the phenotype.
Questions: 1 The most central two issues when selecting an animal model are:
2. Validity refers to:
3. Reliability refers to:
4. What perent of mouse genes match human deisease related genes?
Answers: 1. How well the animal model REPRESENTS the problem being studied and how REPRODUCIBLE the measures can be.
2. The term validity refers to how much the model system actually measures what it is designed to measure.
3. The concept of reliability relates to the accuracy of the measurement
4. 81% of human disease related genes are matched by the mouse.
Aggression in knockout mice. ILAR 41(3), 153.
Abstract: Aggression is defined as overt behavior with the intention of inflicting physical damage on another individual. Agonistic is the term used to describe the entire repertoire of both aggressive and submissive actions within the context of a social interaction. Aggression and submission may be opposite ends of a continuum, or two independent but interaction aspects of behavior. This distinction is important in designing studies of the genetic base for these behaviors. Domestic mice (Mus musculus) are used to study aggression, usually by putting them into artificial situations that promote aggression such as isolation, electrical shocks, or introducing a novel mouse to a cage. Types of aggression in mice are classified into Maternal aggression, Intermale aggression, territorial aggression, Predatory agg., Learned agg. and irritable agg. Fear induced aggression is another type which the authors feel is better classified as defense. Tests of aggressive behavior in mice with targeted deletion of specific genes have been limited to isolation-induced and resident-intruder tests of males and maternal aggression of females. (references are cited in Table 2, p. 154). Methodology of aggression testing: many tests use the model of offensive behavior initiated on the part of the aggressive animal. Male mice can be made more aggressive by isolation for several weeks. They are then placed in an unfamiliar of neutral area with non-aggressive male group housed mice, or in the group's home cage. Behaviors such as pursuit, sideways threat, attack bites, tail rattles, grooming, rearing, and walking are recorded by frequency, onset, and termination. Observations may be visual or remote. In knockout mice, the resident-intruder model is the most used. Previous experience of the test mice can affect their responses, thus careful control if stimuli is important. In female mice parturient females show aggression toward strange intruders during the early part of the lactation period. Usually these attacks have short latency and high intensity, without threat behaviors such as are seen in males. Brain regions Associated with Aggression: Several brain regions have been identified to be involved in the inhibition of aggression, while others may increase aggression, and these regions may not be identical between species. In mice removal of the temporal lobes leads to passivity. Lesions of limbic areas (olfactory bulbs, lateral septum, medial accumbens, dorsal and median raphe, and amygdala) enhance defensiveness and predations, but not social aggression. in cats the hypothalamus and periaqueductal gray matter (PAG) subserve defensive rage and predatory attack, respectively. Ablation of the PAG abolishes defensive rage in cats but not rats. In both rat and cat simulation of the amygdala induces immediate attack behavior experimentally. Studies of knockout mice, in conjunction with additional pharmacologic studies will be useful in further studies of aggression. Pharmacology of Aggression: The authors note that understanding of neurotransmitters regulating aggression is incomplete and that the same neurotransmitter may have different effects in different species. This must be remembered when using animal models to study pharmacologic effects to be related to humans. The major neurotransmitter in aggression , intermale and maternal, is serotonin (5-hydroxytryptamine or 5-HT); while others include noradrenaline, acetylcholine, gamma-aminobutyric acid, and dopamine. Depletion of serotonin leads to increased aggression and addition of serotonin can suppress aggression in a variety of species and social situations. Non-selective 5-HT agonists and antagonists were used in early studies and led to inconclusive results. More specific testing strongly suggests an important role for 5-Ht through two receptor subtypes. Gene Targeting Technology - How to "knock-out" a specific gene: In order to remove or knock-out a gene in the mouse genome it must be identified, targeted, marked precisely, and the nucleotide sequence in two chromosomes rearranged or cut to remove the gene. Mouse embryonic stem cells (ES cells) are then harvested and cultured into which a mutated form of the gene is introduced by micro injection or electroporation transfection. Homologous recombination will incorporate a few of the altered genes into the DNA of the ES cells. Mutated ES cells are inserted in mouse blastocysts and implanted in surrogate mothers. Chimeric mice are produced, in which a few may carry the mutation in germ cells. The chimeric mice are bred with wild-type mice, homozygotes for the mutation (approx. 1/4) are then used to establish the knock-out line. Behavioral testing uses wild type, heterozygotes, and homozygotes in order to compare behaviors and estimate the role of the gene. (Very abbreviated version, worth reviewing this process in molecular biology texts.) Advantages and Disadvantages of Knock-out Technology in Behavioral studies: Disadvantages: lack of the gene may disrupt normal function, this complicates behavioral observations; tests study effect of the missing gene, not direct effects of the gene; compensatory mechanisms may be activated when a gene is missing; all of these can be overcome by testing many pharmacologic and lesional parameters and using standardized behavioral testing methods. Inducible or conditional knockouts may also be useful since the gene can be turned on and off. One example discussed is a Cre-lox bacteriophage site-directed recombination method. Advantages: Disabling a gene is often a clean ablation, it is done without the effects of drugs, and genetic manipulations may be the only way to study the role of endogenous substances. Inducible knockouts are expected to be a very useful tool for future studies. Aggressive Phenotypes: Several models which have been developed using Knock-out Mice are briefly discussed. These include 1. Monoamine Oxidase A - when knocked out 5-HT is expected to increase; the mice however had increased 5-HT and also increased aggression with trembling, difficulty righting, and increased fear response before adulthood, while adults show a decreased fear response. 2. 5-Hydroxytryptamine 1B Receptor - the 5-HT 1B receptor in mice is the homologue of the 5-Ht 1D receptor in humans and modulates the release of other neurotransmitters. Knock-out mice missing the 5-HT 1B receptor show increased aggression toward an intruder, and increased maternal aggression, as well as inter-littermate aggression. Use of eltoprazine, a nonselective 5-HT agonist reduces the aggressive behavior, suggesting that other receptors are also involved including the 5-HT 1A receptor. 3. Alpha Calcium-Calmodulin Kinase II - homozygous mutants showed reduced offensive and defensive aggression. This substance activates tryptophan hydroxylase which is the rate limiting step in 5-HT synthesis, so knock-outs show reduced 5-HT release. These mice show reduced fear. 4. Nitric oxide synthetase (NOS) knockouts - three varieties: endothelial NOS, inducible form in neutrophils, and neural NOS. Neural form knock-outs: high intermale aggression. In most studies the gene is missing throughout life. Other mice can be treated pharmacologically to stop function of nNOS for comparison (using 7-nitroindazole to inhibit nNOS formation), these mice showed increased aggression vs wild type controls. Testosterone levels were studied as a possible contributing factor and were not different; data suggest that testosterone is necessary but not sufficient to promote increased aggression in the mutant mice, castration reduced aggression, while androgen replacement restored it. Female mutants did not show inappropriate aggression, but showed decreased maternal aggression. Results suggest that NO from neurons has important but opposite effects in mediation of aggression in male and female mice. Endothelial Isoform NOS Knock-outs - eNOS knock-out mutants show a mildly elevated blood pressure, animals were very docile and normalization of blood pressure did not change their behavior. The nNOS and eNOS may interact in wild type mice to moderate aggressive behavior. 5. Oxytocin - this hormone has been reported to mediate aggressive and affiliative behaviors in many species. Homozygous knock-outs showed significant reduction in duration of aggressive behavior vs. heterozygous and wild-type mice, with no change in general anxiety levels. 6. Neural Cell Adhesion Molecule - (NCAM) result with knock-out was increased aggression in resident-intruder tests. The mice were shown to have an increased emotional response to threatening stimuli. Testosterone levels were similar to wild-type mice, but cortisone rises were higher. 7. Estrogen receptor - reduced aggression in males, increased aggression in females toward other females. Since estrogen influences the CNS during development, the effects in the mutant mice may be genetic or ontogenetic. 8. Adenosine A2a Receptor - a major target of caffeine, more bite wounds than normal seen in group housed males, increased aggression which may be due to neuromodulatory effects on the release of other neurotransmitters. 9. Interleukin-6 - lack of the gene appears to increase aggression, while increased expression decreases it. Dopamine levels in the brain were modified by lack of IL-6. 10. Adrenergic alpha 2c Receptors - role not yet clear 11. Neurokinin-1 Receptor - substance P modulates sensitivity to pain by activation of these receptors. KO mice were less aggressive. Defensive rage in cats is reduced in this receptor is blocked. 12. Enkephalin - endogenous opioid, lack lead to increased aggression. 13. Mice lacking functional tailless protein show increased aggression. Note: Enzymes: Monoamine oxidase, alpha-calcium calmodulin Kinase II, NOS (n and e), Hormones: oxytocin, estrogen, testosterone Cytokines: Interleukin - 6 Receptors: 5-HT Receptors, estrogen receptor, adenosine A2a receptor, adrenergic alpha 2c receptors, and neurokinin-1 receptor Other: endogenous opioid - Enkephalin, functional tailless protein, neural cell adhesion molecule. Brain regions influenced by each of the above overlap but also differ, so by studying a variety of knock-out animals with different effects on the same brain regions, the larger picture of how aggression is regulated may be elucidated. Summary: Many knock-out mice may show obviously altered behavior and many can be expected to show increased aggression. This should be considered in housing and also standardized behavior testing can be implemented to characterize these mice. The mice listed are a sampling of known knock-outs with altered behavior. Thousands of such mice are being created but few are examined for specific behavioral changes this may be a great oversight in characterizing useful models.
Questions: 1. T or F Knock-out mice allow us to test the direct effect of a gene on behavior.
2. Define aggression and agonistic behavior as used in this article.
3. List two common types of aggression tests used with mice.
4. List three or more models for aggression using knock-out mice.
5. What is the difference between a knock-out mouse and a transgenic mouse?
Answers: 1. False, knock-out animals test effects of the lack of a gene, not its direct effect.
2. Aggression: overt behavior with the intention of inflicting physical damage on another individual. Agonistic behavior: the entire behavioral repertoire of both aggressive and submissive actions within the context of a social interaction.
3. Isolation induced aggression, resident- intruder, maternal.
4. (See list of 13 above)
5. In the knock-out mouse a gene is inactivated or removed at a specific site. In a transgenic mouse a gene is added and the site is usually not specific.
Developing Standardized Behavioral Tests for Knockout and Mutant Mice. ILAR 41(3), 163.
Abstract: There are many factors (genetic and otherwise) that may result in an alteration of behavior and there are many test systems used to evaluate different aspects of behavior. Unfortunately there currently is little or no standardization of methodology (which tests and how they are conducted) and this makes it difficult to compare behavioral assessments of different strains performed by different laboratories. This article describes some nongenetic variables which can affect behavior (or behavioral tests) and gives some examples of test batteries that can be used to examine different aspects of behavior.
Nongenetic variables may be either "background" or "procedural". Background considerations include: source of the animals; health status; physical housing conditions; type of food, water, medications; litter size and sex composition; effects of maternal care; social experiences; and environmental enrichment. Procedural considerations include: selection of control strains; age of testing; sex to be tested; time of day:night cycle tests are performed; how many tests are give to each animal and how many subjects per group; how many test paradigms are used; how are animals handled before and during testing; what apparatus is used for testing; and what is the testing room environment. Examples of how the factors can influence test results are given in the article.
Animals must also be evaluated for sensory and motor deficits. A blind or deaf animal or one with locomotor difficulties may not perform well in certain behavioral testing tasks and this must be considered. Otherwise non-cognitive sensory-motor abnormalities may be confused with a deficiency in learning or memory.
Another non-cognitive factor that can affect behavioral tests (including learning and memory tasks) is "emotion". Some strains of mice show more fear and anxiety than others. An example emotional/defensive behavior test battery includes the following: locomotion/exploration in the open field; elevated plus maze; elevated zero maze; light:dark box; Holeboard test; social interaction test; social conflict test; exploratory/defensive behavior in the visible burrow system. (Some of the tests listed in the paper are briefly described but for many, the reader is given only a reference.)
The understanding of the neurobiologic basis for learning and memory is incomplete but it is known that there are multiple memory systems involving at least 5 neural systems including the hippocampus, amygdala, dorsal striatum, rhinal cortex, and cerebellum. An example test battery based on these multiple memory systems includes: Habituation; Hebb-Williams maze; Morris water maze; Win-shift task on the 8-arm radial maze; conditioned fear response; step down passive avoidance; cued discrimination task of the 8-arm radial maze; object exploration and memory; and delayed nonmatching to sample. (Again, most of these tests are not described.)
Finally, two more test batteries are given to expand the behavioral testing; one to look at animals at various physical and behavioral developmental stages from birth to adulthood and the other to look at aged animals.
In conclusion, the authors reemphasize the vast number of test paradigms available and the need to develop standardized test batteries based on the best current information. Developing standardized behavioral tests for knockout and mutant mice.
Questions: 1. What is an ethogram?
2. Why may DBA/2 and C57BL/6J mice have a reduced response in the auditory (or acoustic) startle test?
Answers: 1. An ethogram is an inventory of the behavioral repertoire of a given species. For example the ethogram of a mouse might consist of a list of behaviors such as sleeping/resting, locomotion, grooming, nestbuilding, food ingestion, etc. The behaviors should also be quantitated in some fashion (ex. percentage of time involved in that activity).
2. These strains may become deaf at an early age