Online Issues

Journal Vol 47(2)

Cause and Effect Considerations in Diagnostic Pathology and Pathology Phenotyping of Genetically Engineered Mice (GEM)

Colin McKerlie

Colin McKerlie, D.V.M., D.V.Sc., M.R.C.V.S., is a Primary Investigator, the Director of the Pathology Core of the Centre for Modeling Human Disease at the Samuel Lunenfeld Research Institute, Mount Sinai Hospital, the Director of the Canadian Mouse Mutant Repository, and a Scientist in the Integrative Biology Research Program, The Hospital for Sick Children, Toronto, Ontario, Canada.

Abstract

Over the next several decades, biology is embarking on its most ambitious project yet: to annotate the human genome functionally, prioritizing and focusing on those genes relevant to development and disease. Model systems are fundamental prerequisites for this task, and genetically engineered mice (GEM) are by far the most accessible mammalian system because of their anatomical, physiological, and genetic similarity to humans. The scientific utility of GEM has become commonplace since the technology to produce them was established in the early 1980s. Conceptually, however, an efficiently coordinated high-throughput approach that permits correlation between newly discovered genes, functional properties of their protein products, and biological relevance of these products as drug targets has yet to be established. The discipline of veterinary anatomical pathology (hereafter referred to as pathology) is not immune to this requirement for evolution and adaptation, and to address relationships and tissue consequences between tens of thousands of genes and their cognate proteins, novel interdisciplinary technologies and approaches must emerge. Although many of the techniques of pathology are well established, in the context of pathology's contribution to functional annotation of the genome, several conceptually important and unresolved issues remain to be addressed. While an ever-increasing arsenal of genetic and molecular toolsets are available to evaluate and understand the function of genes and their pathophysiological mechanisms, pathology will continue to play an essential role in confirming cause and effect relationships of gene function in development and disease. This role will continue to be dependent on keen observation, a systematic but disciplined approach, expert knowledge of strain-dependent anatomical differences and incidental lesions, and relevant tissue-based evidence. Miniaturization and high-throughput adaptation of these methods must also continue so that they can complement parallel phenotyping efforts, provide pathology-based data in pace with concurrent phenotyping efforts, and continue to find new utility in the collective effort of functional annotation.

Key Words: diagnostic pathology; mouse pathology; phenotype pathology; random mutagenesis; targeted mutagenesis

Anatomical Pathology of Genetically Engineered Mice

Background—Approach and Techniques

For both diagnostic pathology and pathology phenotyping of genetically engineered mice (GEM¹), the primary role of the pathologist is to detect, assess, and interpret morphological abnormalities, and to discriminate normal from abnormal. It is often the common tools and techniques of anatomical pathology that bridge the diagnostic discipline with the science of pathogenesis and pathology phenotype. The intent of this article is to present the reader with an overview of approaches to diagnostic pathology and pathology phenotyping of GEM. It is not meant to be exhaustive in nature, but rather to give the reader an appreciation of the procedures involved in these processes, their application in diagnostic versus phenotyping pathology, and the caveats and limitations in their contributions to functional annotation of the genome.

Necropsy and Gross Pathology

The process begins with an examination of the study protocol, clinical signs, gross observations, and records and information provided on the genetic makeup of the GEM to be evaluated. Consideration and inclusion of data regarding age, gender, diet, husbandry, and infectious disease or health status are important steps. Good and thorough necropsy technique should permit and facilitate evaluation and collection of all organs and lesions (one may not need or want a particular organ initially, but it may prove vital as the diagnoses or phenotyping progresses). Imaging, organ weights, description of all lesions, and robust record collection are crucial to the process. Ideally, the approach should be systematic, efficient, and simple, so that it will generate consistent and reproducible results across multiple cohorts of GEM. The approach should also be adaptable for special techniques (e.g., perfusion), recovery and detailed examination of placentas and embryos, or for training. Excellent reviews are available (Brayton et al. 2001; Ward et al. 2000). Although the glass slides, collected tissue(s), and interpretation can be repeatedly evaluated during diagnoses or phenotyping, many of these early data points at necropsy, unless carefully captured, cannot be reconstructed later. Virtual mouse necropsy tutorials are available (http://www.geocities.com/virtualbiology/ and http://tvmouse.compmed.ucdavis.edu/).

Histopathology

Evaluation of tissue collected at necropsy for diagnostic or phenotype review relies on methods common to veterinary or human diagnostic pathology. Routine hematoxylin and eosin (H&E¹) staining of tissue sections on glass slides, microimaging or photography, histochemistry (special stains useful for specific cell, tissue, or substance types), and techniques such as serial sectioning can be used. Effort should be made to maximize the number of tissues evaluated while minimizing the number of glass slides necessary. Efficiencies can be gained in both cost and storage of samples. For example, the Brayton Technique for evaluation of the head provides four sections of decalcified head on a single glass slide that permits evaluation of the following: brain (frontal cortex with corpus callosum and caudate putamen, cerebral cortex with thalamus and hippocampus, and cerebellum with brain stem) and pituitary (for hydrocephalus, hypoplasia, inflammation, neoplasia, degenerative changes); eyes (for retinal atrophy, degeneration or dysplasia, cataract, inflammation, neoplasia); ears (for otitis media, interna); teeth (for dental and periodontal disease, dysplasia); bone and marrow; turbinates, olfactory and respiratory epithelium; salivary and Harderian glands; and other associated tissues (for malformation, inflammation, neoplasia, trauma) (Brayton et al. 2001). The intestinal Swiss Roll is another example of a technique that maximizes histopathological analysis while minimizing expense (Moolenbeek and Ruitenberg 1981). Adaptation of standard histology such as these special techniques can provide a comprehensive tissue survey on a single slide.

Extended Techniques

As with the necropsy, it may be necessary to adapt the approach and techniques of histopathology or to extend them beyond routine H&E and special stains, particularly for problematic diagnostic cases or extended pathology phenotyping of lesions identified on routine evaluation. Histomorphometry can provide quantitative data that may be crucial in determining the significance of pathologies or lesions in GEM, and it is increasingly accessible with commercially available and affordable image analysis software. Immunohistochemistry, utilizing monoclonal and less frequently polyclonal antibodies, may be useful for protein, carbohydrate, or lipid function and expression, identification of targets for biologics, confirmation of pathogens, and investigation of mechanistic pathways. Despite the general acceptance of immunhistochemistry for in situ identification and localization of proteins, it is still important to exercise care in optimizing protocols and interpreting results (Schuh and Harrington 1999). Protocols with insufficient and false characterization of antibodies, lack of reproducibility and appropriate controls, and biologically irrelevant staining are still frequently reported in the literature. Electron microscopy may be appropriate for ultrastructural evaluation of lesions or cellular phenotype.

Changes in expression of genes and their molecular products underlie development and disease and drive what we see as pathology phenotype. Various methods are useful for tissue localization or characterization of the morphological changes including in situ hybridization. These techniques can be applied to fresh, frozen, and fixed tissue collected at necropsy to help in the evaluation of the role of a gene's RNA and DNA. Laser capture microdissection is an additional tool that can be used to extract specific cells from their complex tissue milieu to provide pure populations of targeted cells from specific microscopic regions of tissue sections (Bonner et al. 1997). DNA, RNA, or protein can then be extracted from these captured cell populations, amplified and purified, and made available for downstream molecular analysis.

Diagnostic Pathology of Mouse Models of Mouse Disease

Background

Since the early 1980s, numbers of GEM used in research have increased exponentially, as has the trafficking in these models of human disease. Crowded populations and rampant import, export, and cross-colony breeding all have markedly increased the opportunity for infectious disease, and in some cases have resulted in the re-emergence of previously quiescent mouse pathogens. Diagnostic pathology therefore remains an essential discipline and endeavor, made ever challenging by the myriad of established and evolving spontaneous pathologies of the mouse, infectious diseases of the mouse, and strain-related pathologies (Percy and Barthold 2001; see also “Mouse genetics versus bona fide GEM phenotype—background strain effects” below).

Diagnostic Pathology of Mouse Disease

Spontaneous pathology. The inbred strains of laboratory mice that contribute to the majority of GEM have significant spontaneous pathologies despite the pristine and controlled environments in which so many of them live. For excellent and comprehensive reviews, the reader is referred to several recently published texts (Maronpot et al. 1999; Mohr et al. 1996; Percy and Barthold 2001; Ward et al. 2000). A striking example is the C57BL/6 strain, which is very widely used in the production of GEM. This strain suffers from a high incidence of late-onset amyloidosis, congenital ocular defects (Smith et al. 1994), hydrocephalus (Mori 1968), early-onset periauricular arteritis and organ of Corti degeneration causing deafness (Hultcrantz and Li 1993), and pulmonary histiocytosis/crystal pneumonitis (Guo et al. 2000; Percy and Barthold 2001; Ward et al. 2001). In addition, it is prone to spontaneous alopecia and hypersensitivity dermatitis (Andrews et al. 1994). The pathologist must be aware of these strain-related characteristics to interpret the significance of any changes detected during the diagnostic exercise.

Infectious pathology. Infectious agents can have a significant impact on the biological response of GEM to disease and to the nature, distribution, and severity of pathologies encountered. Infectious agents that would otherwise be overlooked or unknown in other species are significant in GEM because of their potential and real impact on research results (Percy and Barthold 2001). Opportunistic pathogens such as Helicobacter, Pasteurella, and Pneumocystis are important and increasingly common in colonies of GEM (Barthold 2002). The presence of such infectious agents and the possible complications associated with these infections must be recognized and accounted for as part of the diagnostic exercise.

Pathology Phenotyping of Mouse Models of Human Disease

Background

The classic basic principles approach to diagnostic pathology is dependent on well-established and well-described pattern recognition and rote morphological diagnoses. This approach requires significant adaptation when it is applied to the discovery, description, and interpretation of an induced phenotype. Novel lesions distributed in novel patterns that may or may not represent comparative human pathologies necessitate a fundamental mechanisms of development and disease approach in combination with an element of biology-based and evidence-based creativity. In the context of the multidisciplinary effort to generate, describe, and utilize GEM models efficiently, it is particularly important for the complete dataset contribution from pathology phenotyping to be presented by the pathologist in a meaningful and useful form to the collaborating research community.

Pathology Phenotyping of Genetic Disease in Gene-driven Mutagenesis

Special considerations for GEM phenotyping. Pathology phenotyping of GEM requires rigorous, expert, and formal analyses. Well-established paradigms and techniques can be chosen from the established pathology literature, and adoption of acceptable standards of approach will likely lead to more reproducible or interpretable results of phenotypes discovered and characterized in very different laboratories. Although there is currently no collective resource of standard approaches and guides to interpretation, excellent sources of information in particular areas of pathology are emerging. Examples include recognized standards for tumor classification (Borowsky et al. 2003, 2004; Bult et al. 2000; Galvez et al. 2004; Haley et al. 2005; Nikitin et al. 2004; Shappell et al. 2004), mouse anatomy (http://www.informatics.jax.org/searches/anatdict_form.shtmly), pathology phenotyping (Maronpot et al. 1999; Sundberg and Boggess 1999; Ward et al. 2000), and mouse pathology ontologies (Schofield et al. 2004a,b).

Systematic evaluation and cohort sampling. Comprehensive pathology phenotype assessment is typically used when no to very little information is available on the possible function of the gene. For reasons of economy and efficiency, these comprehensive screens are often divided into tiers or phases. The first tier consists of a general necropsy and histopathology screen across all genotypes and both sexes. The second tier comprises multiple specialized techniques that often require additional but specific cohorts with the objective of providing deeper and more specific pathology phenotype data.

Targeted pathology phenotype assessment is performed when the function of the gene is known, there are predictive data that suggest function, or, in GEM, with tissue-specific expression or deletion. A targeted approach may also be appropriate when there is interest in a specific organ system, tissue type, or pathology phenotype. It typically consists of a focused approach that uses particular techniques to help define and characterize the pathology phenotype within the specific scope or area of research interest. Caution must be taken that preconceptions of gene function used to drive a targeted phenotyping approach do not preclude the omission of important pathology data associated with an off-target tissue or organ. In these so-called “aberrant” phenotypes, the gene-related pathology phenotype can be in a completely unsuspected part of the mouse.

In either approach, a significant challenge in the pathology phenotype analysis of a novel GEM is obtaining sufficient animals for each of the relevant genotypes. Breeding problems and demand for mice in parallel in vivo or in vitro assays often limit the availability of large numbers (the n) of each genotype, or each sex, at particular time points. Compounding this challenge may be the necessity to follow and evaluate the pathology phenotype at multiple age points to assess disease progression fully over the life of the animal.

Mouse “disease” versus bona fide GEM phenotype. The majority of transgenic and targeted GEM are currently being developed using 129 substrains and C57BL/6. The choice of these strains is historical and based on technical successes. However, each of these strains maintains a relatively large number of spontaneous pathologies as described above, and it is important to identify and interpret the presence of any of these changes accurately and appropriately in the context of the phenotyping exercise.

Mouse genetics versus bona fide GEM phenotype—background strain effects. Background strain effects can also be problematic. There are vast differences in normal anatomy among different strains of GEM. The pathologist or investigator must be unambiguously informed about what is normal, and this information depends variably on the availability of normal “wild-type” controls and on a comprehensive familiarity with normal strain-associated differences in tissue morphology. For example, C3H mice have high cortical bone density compared with C67BL/6 mice (Rosen et al. 1997)—a potentially important differentiator when comparing the utility of a new model of osteoporosis. Likewise, the significantly larger alveolar size in C3H/HeJ mice versus C57BL/6J mice must be considered when interpreting changes in lung development or injury in GEM models of lung disease (Soutiere et al. 2004). A particularly dramatic anatomical example is the high incidence of hypoplastic or absent corpus callosum in several substrains of 129 mice and the BALB/cJ strain as the result of retarded embryonic formation of the hippocampal commissure (Livy and Wahlsten 1991, 1997; Wahlsten 2001; Wahlsten et al. 2001).

The Mouse Phenome Database (MPD¹) hosted by The Jackson Laboratory (Paigen and Eppig 2000; ) is an international collaborative effort to promote the quantitative phenotypic characterization of a defined set of JAX® mouse strains under standardized conditions for a wide range of phenotypes of biomedical relevance. Numerous measurement categories at MPD are relevant and useful to the interpretation of pathology phenotype in GEM.

Pathology Phenotyping of Genetic Disease in Phenotype-driven Mutagenesis

Special considerations for high-throughput phenotyping. As a complement to targeted gene knockout and knockin strategies that are limited to known genes, phenotype-driven chemical mutagenesis with effective phenotype-based screens makes no a priori assumption about gene sequence. This phenotype-driven approach starts with a phenotype, and only at the end is the responsible gene identified (Balling 2001; Beckers and Hrabe de Angelis 2001; Brown and Balling 2001; Justice 2000; Justice et al. 1999; Rossant and McKerlie 2001).

Mice are exposed to a mutagen that acts in a random shotgun-like approach across the entire genome. Consecutively, animals are screened in a systematic way to identify individuals that display disease phenotypes. Successful screening for mutations causing phenotypes requires the mutant phenotype to vary significantly from the background. However, screening involves the analysis of large numbers of mice, so the ideal phenotype screen must be broad and inexpensive (Justice et al. 1999). Visible phenotypes that affect external features (e.g., eye, coat, size, neuromuscular function) are simple to identify, whereas screens for clinical chemistry, learning and memory, skeletal development, and other factors may require more sophisticated protocol and equipment. Regardless of the screen used, following the isolation of an outlier, test breeding and repeat screening of progeny are used to confirm the heritability of the trait. Standard genetic and molecular tools of gene mapping are then used to isolate and identify the mutated gene, making the connection between gene and function.

The most potent mutagen in mice is N-ethyl-N-nitrosourea (ENU¹), and the unique strength of ENU mutagenesis is its great efficiency. ENU has been shown to create a new loss-of-function allele in any given locus, typically once in every 700 gametes (Hitotsumachi et al. 1985). This efficiency, combined with the development of increasingly sophisticated phenotype screens, has led to the establishment of several large ENU mutagenesis initiatives worldwide that are generating significant numbers of exciting new mutants (Hrabe de Angelis et al. 2000; Nolan et al. 2000; Svenson et al. 2003). Yet regardless of its confirmed efficiency, ENU mutagenesis for the creation of models of development and disease ultimately relies on the generation and screening of tens to hundreds of thousands of mice using ever more inventive screening protocols for phenotype. Additionally, the standard for a successful ENU screen is for it to be fast, cheap, robust, and preferably noninvasive—not criteria typically assigned to pathology. Nevertheless, continued development and refinement of phenotyping methods are critical if we are to exploit the power of phenotype-based approaches fully for annotation of the genome.

Is pathology screening for phenotype-driven mutagenesis useful? ENU programs have clearly shown their ability to generate new mouse models of human development and disease, and the point mutations in genes induced by ENU continue to display a range of mutant effects successfully, from complete or partial loss-of-function to gain-of-function. This approach has several advantages: It generates multiple allelic variants and most often inactivates or alters the function of individual protein domains rather than eliminating the protein altogether, making it more difficult for related proteins to compensate; it closely mimics the effects of drugs or natural genetic variants; it illuminates active sites in proteins (Nelms and Goodnow 2001).

The pathology phenotyping screen of one large ENU center accumulated data over a 3-yr screening period. The objectives of the pathology screen were two-fold: (1) to perform a gross and histopathology screen of all first generation (G1¹) males in an effort to identify dominantly expressed tissue phenotypes that had not been detected by any of the in-life phenotyping assays; and (2) to apply routine and extended pathology techniques to the comprehensive phenotype characterization of confirmed heritable mutant lines that were discovered by in-life phenotyping assays and were in the process of being positionally cloned. After more than 5,500 mice were screened for pathology phenotype, 76 of the 8- to 12-wk-old G1 mice tested had a detectable pathology phenotype (C. McKerlie, unpublished data). Although this “hit-rate” demonstrated bona fide pathology phenotypes, including neoplasia, in approximately 1.4% of the mice screened, the proportion of the phenotypes that would be confirmed heritable by in vitro fertilization has not been established. Published values for recovery of mice with abnormal phenotypes that are confirmed as heritable mutations are typically 1 to 2% in dominant screens (Hrabe de Angelis et al. 2000; Nolan et al. 2000).

Clearly, as with any phenotype screen, pathology has its challenges. Rare mutant animals of interest must be readily detected against a background of hundreds of phenotpyically normal individuals. Mutations in genes critical to development and disease are frequently not immediately apparent by a spontaneous clinical disorder or a routine and systematic anatomical evaluation, and may in fact affect rare cell types or processes that are not obvious without specific histomorphometric, ultrastructural, or indeed biochemical evaluation. High-throughput pathology-based screens miss these phenotypes. Genes and mutations that require environmental, genetic, or pharmacological challenge to reveal their resultant phenotype are difficult to detect without in-depth pathology screening in combination with the appropriate epigenetic or environmental stressor. Mutations contributing to spontaneous disorders that require multiple genetic abnormalities to be present before the defect becomes penetrant as an anatomical phenotype, or that require a prolonged period of time (late-onset phenotypes), challenge pathology phenotyping in adult mice. Likewise, the estimated 10 to 20% of mammalian genes that are essential for embryonic development require the adaptation and miniaturization of pathology techniques applied to the mouse embryo and neonate. Finally, recovery of mutations responsible for pathology phenotypes, particularly in dominant screens where each G1 animal being evaluated is a unique genome, relies on robust and efficient germ cell collection and in vitro fertilization.

Is it practical? Several issues should be considered when selecting any in vivo assay for phenotype screening.

Critical assay attributes such as accuracy, reproducibility, and diagnostic discrimination may not be similar among laboratories. Initiatives such as the European Mouse Phenotyping Resource of Standardized Screens (EMPReSS), which is being developed and promoted by EUMORPHIA (http://www.eumorphia.org/), is an excellent example of emerging efforts to create standards for phenotyping.
Surgical or other manipulation may independently and variably influence the specific parameter being screened for.
The suitability of a phenotyping assay is influenced by the developmental stage and/or age of the mouse or cohort of mice being evaluated.
The criteria for selection and availability of appropriate control animals (e.g., littermates and hybrid or inbred wild-type) may be difficult to establish.
Local expertise, technical difficulty of the process, and cost are paramount considerations.

For these reasons, it is clear that pathology screening in phenotype-driven mutagenesis has advantages and limitations. Therefore it is important for pathology techniques that are selected or included in a phenotype-driven screen to be tailored to the specific hypotheses and questions being addressed, or the local disease areas of interest that drive the screening process. Attention to these caveats will become increasingly important as the large mutagenesis centers continue to transform into phenotyping centers with specialty in one or a few areas.

Can pathology keep up? It cannot go unnoticed that there are substantial cost factors and effort associated with the development of a comprehensive approach to diagnostic pathology and pathology phenotyping. In addition, it is necessary for pathology phenotyping assays to achieve some degree of standardization across the international research community. Optimization, acceptance, and implementation of standards in nomenclature, ontologies, protocols, and diagnostic criteria will experience all of the challenges of egotistical inertia and individual investment in local or regional historical standards that characterize the discipline of pathology.

The volume of pathology-derived phenotype data being generated will also be problematic. To advance the use of these data, we must move beyond cataloguing the components of a given GEM with a limited representation of the associated pathology phenotype. Meaningful integration and visualization of complete pathology phenotype datasets in real-time and accessible format are required. The tasks involved include display, visualization, and query of two- and three-dimensional gross-, histo-, and molecular pathology repositories (i.e., high-resolution whole slide scans). An approach is required that combines the disciplines of biology, pathology, genomics, imaging, scientific visualization, human-computer interfaces, information technology, and data integration.

It is clear from the international initiative now underway to mutate every gene in the mouse genome that over the next several years, the rate-limiting step in functional annotation will not be the availability of GEM but may instead be the availability of well-trained and expert mouse biologists. Pathology in particular appears to be affected by a combination of insufficient training programs and a lack of experienced pathologists who have expertise in the subspecialty of general mouse pathology, spontaneous, infectious, and developmental pathology of the mouse, and comparative pathology of GEM and humans.

Nevertheless, there are exciting changes and developments in other disciplines that can also help pathology keep up. The parallel research and application of imaging modalities such as magnetic resonance imaging (MRI¹) will provide complementary tools to focus and expedite a whole-mouse tissue phenotyping exercise to a narrower, more practical subset of tissues (Maronpot et al. 2004). Tooling and optimization of ultrasound for GEM phenotyping has provided tremendous insight into cardiovascular function and early embryonic development (Foster et al. 2000). As with MRI, the evolution of ultrasound for the mouse in partnership with pathology phenotyping will reap mutual rewards for efficiency and validation of phenotype. The maturity of germ cell (i.e., sperm freezing) and in vitro fertilization will markedly improve the accessibility and mobility of GEM, thereby increasing the availability and use of distributed and limited pathology expertise and resources.

Conclusions

It is clear that diagnostic pathology of the mouse and pathology phenotyping of human development and disease in GEM extends far beyond simple recognition and mimicry of the human pathological condition. What is now needed for pathology phenotyping of targeted GEM, or for high-throughput pathology screening and phenotyping of GEM from ENU programs, is in-depth, detailed, and systematic phenotyping that delivers biologically relevant information. To contribute fully, the pathologist is asking, “How could this gene or gene product be involved in this pathology?“ rather than, “Which gene or disease is responsible for this pathology?” Using this approach, it is indeed possible for pathologists to claim contribution toward the mouse geneticists' motto, “Give me a mutant, and I’ll give you a gene” (Balling 2001).

Challenging questions for the immediate future include the following:

How will the pathology community deal with such an influx of genes with unknown function, or such an influx of pathology phenotypes with an unknown gene?
How does pathology best contribute to the effort of assigning function to these genes and assembling pathways that are relevant to human development and disease?

Practitioners, trainees, students, and collaborators of diagnostic pathology and pathology phenotyping of GEM are poised to make tremendous contributions toward resolving each of these questions. Indeed, the brightest future for mouse pathology in an era of functional annotation of the genome requires a strong link between diagnostic work and the science of pathogenesis and pathology phenotype in GEM. To that end, it is critical to prevent any increase in the gap between these branches of pathology. It is incumbent upon pathologists, investigators, and other science policy decision makers to avoid missing this great opportunity.

Abbreviations used in this article: ENU, N-ethyl-N-nitrosourea; G1, first generation; GEM, genetically engineered mice; H&E, hematoxylin and eosin; MPD, Mouse Phenome Database; MRI, magnetic resonance imaging.

References

Andrews AG, Dysko RC, Spilman SC, Kunkel RG, Brammer DW, Johnson KJ. 1994. Immune complex vasculitis with secondary ulcerative dermatitis in aged C57BL/6NNia mice. Vet Pathol 31:293-300.

Balling R. 2001. ENU mutagenesis: Analyzing gene function in mice. Annu Rev Genomics Hum Genet 2:463-492.

Barthold SW. 2002. Muromics: Genomics from the perspective of the laboratory mouse. Comp Med 52:206-223.

Beckers J, Hrabe de Angelis M. 2001. Large-scale mutational analysis for the annotation of the mouse genome. Curr Opin Chem Biol 6:17-23.

Bonner RF, Emmert-Buck M, Cole K, Pohida T, Chuaqui R, Goldstein S, Liotta LA. 1997. Laser capture microdissection: Molecular analysis of tissue. Science 278:1481-1482.

Borowsky AD, Galvez JJ, Munn RJ, Cardiff RD. 2003. Comparative pathology of mouse models of human cancers. Comp Med 53:248-258.

Borowsky AD, Munn RJ, Galvez JJ, Cardiff RD. 2004. Mouse models of human cancers (Part 3). Comp Med 54:258-270.

Brayton C, Justice M, Montgomery CA. 2001. Evaluating mutant mice: Anatomic pathology. Vet Pathol 38:1-19.

Brown SDM, Balling R. 2001. Systematic approaches to mouse mutagenesis. Curr Opin Genet Dev 11:268-273.

Bult CJ, Krupke DM, Sundberg JP, Eppig JT. 2000. Mouse tumor biology database (MTB): Enhancements and current status. Nucl Acid Res 28:112-114.

Foster FS, Pavlin CJ, Harasiewicz KA, Christopher DA, Turnbull DH. 2000. Advances in ultrasound biomicroscopy. Ultrasound Med Biol 26:1-27.

Galvez JJ, Cardiff RD, Munn RJ, Borowsky AD. 2004. Mouse models of human cancers (Part 2). Comp Med 54:13-28.

Guo L, Johnson RS, Schuh JCL. 2000. Biochemical characterization of endogenously formed eosinophilic crystals in the lungs of mice. J Biol Chem 275:8032-8037.

Haley P, Perry R, Ennulat D, Frame S, Johnson C, Lapointe J-M, Nyska A, Snyder PW, Walker D, Walter G. 2005. STP position paper: Best practice guidelines for the routine pathology evaluation of the immune system. Tox Pathol 33:404-407.

Hitotsumachi S, Carpenter DA, Russell WL. 1985. Dose-repetition increases the mutagenic effectiveness of N-ethyl-N-nitrosourea in mouse spermatogonia. Proc Natl Acad Sci U S A 82:6619-6621.

Hrabe de Angelis MH, Flaswinkel H, Fuchs H, Rathkolb B, Soewarto D, Marschall S, Heffner S, Pargent W, Wuensch K, Jung M, Reis A, Richter T, Alessandrini F, Jakob T, Fuchs E, Kolb H, Kremmer E, Schaeble K, Rollinski B, Roscher A, Peters C, Meitinger T, Strom T, Steckler T, Holsboer F, Klopstock T, Gekeler F, Schindewolf C, Jung T, Avraham K, Behrendt H, Ring J, Zimmer A, Schughart K, Pfeffer K, Wolf E, Balling R. 2000. Genome-wide, large-scale production of mutant mice by ENU mutagenesis. Nat Genet 4:444-447.

Hultcrantz M, Li HS. 1993. Inner ear morphology in CBA/Ca and C57BL/6J mice in relationship to noise, age and phenotype. Eur Arch Otorhinolaryngol 250:257-264.

Justice MJ, Noveroske JK, Weber JS, Zheng B, Bradley A. 1999. Mouse ENU mutagenesis. Hum Mol Genet 8:1955-1963.

Justice MJ. 2000. Capitalizing on large-scale mouse mutagenesis screens. Nat Rev Genet 1:109-115.

Livy DJ, Wahlsten D. 1991. Tests of genetic allelism between four inbred mouse strains with absent corpus callosum. J Hered 82:459-464.

Livy DJ, Wahlsten D. 1997. Retarded formation of the hippocampal commissure in embryos from mouse strains lacking a corpus callosum. Hippocampus 7:2-14.

Maronpot RR, Boorman GA, Gaul BW, eds. 1999. Pathology of the Mouse: Reference and Atlas. 1st ed. Vienna IL: Cache River Press.

Maronpot RR, Sills RC, Johnson GA. 2004. Applications of magnetic resonance microscopy. Tox Pathol 32(Suppl 2):42-48.

Mohr U, Dungworth DL, Capen CC, Carlton WW, Sundberg JP, Ward JM, eds. 1996. Pathobiology of the Aging Mouse. Vols 1 and 2. 1st ed. Washington DC: ILSI Press.

Moolenbeek C, Ruitenberg EJ. 1981. The “Swiss Roll”: A simple technique for histologic studies of the rodent intestine. Lab Anim 15:57-59.

Mori A. 1968. Hereditary hydrocephalus in C57BL mouse. Brain Nerve 20:695-700.

Nelms KA, Goodnow CC. 2001. Genome-wide ENU mutagenesis to reveal immune regulators. Immunity 15:409-418.

Nikitin AY, Alcaraz A, Anver MR, Bronson RT, Cardiff RD, Dixon D, Fraire AE, Bagrielson EW, Gunning WT, Haines DC, Kaufman MH, Linnoila RI, Maronpot RR, Rabson AS, Reddick RL, Rehm S, Rozengurt N, Schuller HM, Shmidt EN, Travis WD, Ward JM, Jacks T. 2004. Classification of proliferative pulmonary lesions of the mouse: Recommendations of the Mouse Models of Human Cancers Consortium. Cancer Res 64:2307-2316.

Nolan P, Peters J, Strivens M, Rogers D, Hagan J, Spurr N, Gray IC, Vizor L, Brooker D, Whitehill E, Washbourne R, Hough T, Greenaway S, Hewitt M, Liu X, McCormack S, Pickford K, Selley R, Wells C, Tymowska-Lalanne Z, Roby P, Glenister P, Thornton C, Thaung C, Stevenson JA, Arkell R, Mburu P, Hardisty R, Kiernan A, Erven A, Steel KP, Voegeling S, Guenet JL, Nickols C, Sadri R, Nasse M, Isaacs A, Davies K, Browne M, Fisher EM, Martin J, Rastan S, Brown SD, Hunter J. 2000. A systematic, genome-wide, phenotype-driven mutagenesis programme for genetic function studies in the mouse. Nat Genet 4:440-443.

Paigen K, Eppig JT. 2000. A mouse phenome project. Mamm Genome 11:715-717.

Percy DH, Barthold SW. 2001. Mouse. In: Percy DP, Barthold SW, eds. Pathology of Laboratory Rodents & Rabbits. 2nd ed. Ames: Iowa State Press. p 3-106.

Rosen CJ, Dimai HP, Vereault D, Donahue LR, Shultz KL, Farely J. 1997. Circulating and skeletal IGF-1 concentrations in two inbred strains of mice with different bone densities. Bone 21:217-233.

Rossant J, McKerlie C. 2001. Mouse-based phenogenomics for modeling human disease. Trends Mol Med 7:502-507.

Schofield PN, Bard JB, Boniver J, Covelli V, Delvenne P, Ellender M, Engstrom W, Goessner W, Gruenberger M, Hoefler H, Hopewell J, Mancuso M, Mothersill C, Quintanilla-Martinez L, Rozell B, Sariola H, Sundberg JP, Ward A. 2004a. Pathbase: A new reference resource and database for laboratory mouse pathology. Rad Prot Dosim 112:525-528.

Schofield PN, Bard JB, Booth C, Boniver J, Covelli V, Delvenne P, Ellender M, Engstrom W, Goessner W, Gruenberger M, Hoefler H, Hopewell J, Mancuso M, Mothersill C, Potten CS, Quintanilla-Fend L, Rozell B, Sariola H, Sundberg JP, Ward A. 2004b. Pathbase: A database of mutant mouse pathology. Nucl Acids Res 32:D512-D515.

Schuh JC, Harrington KA. 1999. Mechanisms of disease and injury: Utilization of mutants, monoclonals, and molecular methods. Toxicol Pathol 27:115-120.

Shappell SB, Thomas GV, Roberts RL, Herbert R, Ittmann MM, Rubin MA, Humphrey PA, Sundberg JP, Roxengurt N, Barrios B, Ward JM, Cardiff RD. 2004. Prostate pathology of genetically engineered mice: Definitions and classification. The Consensus Report from the Bar Harbor Meeting of the Mouse Models of Human Cancer Consortium Prostate Pathology Committee. Cancer Res 64:2270-2305.

Smith RS, Roderick TH, Sundberg JP. 1994. Microphthalmia and associated abnormalities in inbred black mice. Lab Anim Sci 44:551-560.

Soutiere SE, Tankersley CG, Mitzner W. 2004. Differences in alveolar size in inbred mouse strains. Resp Physiol Neurobiol 140:283-291.

Sundberg JP, Boggess D, eds. 1999. Systematic Characterization of Mouse Mutations. 1st ed. Boca Raton: CRC Press.

Svenson KL, Bogue MA, Peters LL. 2003. Identifying new mouse models of cardiovascular disease: A review of high-throughput screens of mutagenized and inbred strains. J Appl Physiol 94:1650-1659.

Wahlsten D. 2001. Standardizing tests of mouse behaviour: Reasons, recommendations, and reality. Physiol Behav 73:695-704.

Wahlsten D, Crabbe JC, Dudek BC. 2001. Behavioral testing of standard inbred and 5HT_1B knockout mice: Implications of absent corpus callosum. Behav Brain Res 125:23-32.

Ward JM, Mahler JF, Maronpot RR, Sundberg JP, eds. 2000. Pathology of Genetically Engineered Mice. 1st ed. Ames: Iowa State University Press.

Ward JM, Yoon M, Anver MR, Haines DC, Kudo G, Gonzalez FJ, Kimura S. 2001. Hyalinosis and Ym1/Ym2 gene expression in the stomach and respiratory tract of 129S4/SvJae and wild-type and CYP1A2-Null B6,129 mice. Am J Pathol 158:323-332.