Endocr Pathol (2008) 19:139–147 DOI 10.1007/s12022-008-9042-2
The Genomic Revolution and Endocrine Pathology Suzana S. Couto & Robert D. Cardiff
Published online: 31 August 2008 # Humana Press Inc. 2008
Abstract The genome has been sequenced. However, the functions of each gene remain to be elucidated through phenotypic analysis. This analysis has been called phenogenomics. That part of phenogenomics related to disease can be called pathogenomics or Genomic Pathology. The initial phases of disease analysis will use genetically modified mice. The proliferation of ambitious programs designed to use mice for phenogenomics has been met with alarm by comparative pathologists who note the lack of qualified genomic pathologists and of training programs in genomic pathology. While endocrine pathology offers a number of excellent examples of the contributions made by pathologists to the scientific literature, it also contains examples of the hazards of working with untrained, unwary personnel. Keywords endocrine pathology . genomics . neuroendocrine . genetically engineered mice . SV40
Introduction The Genomic Revolution proves that all animals have remarkably similar genomes. This similarity has broad implications for all of biology, particularly for medicine. We now know that mutations of the same gene in different
S. S. Couto Memorial Sloan–Kettering Cancer Center, 1275 York Avenue, Room Z- 930, P.O. Box 270, New York, NY 10065, USA R. D. Cardiff (*) Center for Comparative Medicine, University of California, One Shields Ave, Davis, CA 95616, USA e-mail:
[email protected]
species result in the same disease, in short, one gene-one disease, the molecular equivalent of one medicine. The concept has been repeatedly demonstrated with genetically engineered mice (GEM) where the effects of modifications in single genes are tested by observing the resulting phenotype. Genes associated with diabetes in humans “cause” diabetes when placed in mice. This transfer of genes from one species to another has become the “Koch’s Postulate” of modern biology. Although the newly created phenotypes may be appreciated grossly (clinically), confirmation still requires microscopic examination. The similarity of the mouse disease to human disease is “validated” by comparing the functional and structural characteristics of the pathology in both species [1]. Validation clearly requires the unique skills of the pathologists with the ability to interpret structural, in particular microscopic, alterations. Thus, genetic engineering has introduced a new area of comparative pathology, Genomic Pathology. Genomic Pathology seeks to understand and interpret the structural changes resulting from genomic alterations and to relate them to diseases in all species. This requires the ability to compare and contrast the histological anatomy of two or more species (comparative pathology). The literature is replete with the embarrassing misinterpretations by untrained individuals practicing “Do-It-Yourself (DIY) pathology” [2]. The plethora of faulty reports also reflects the lack of pathologists experienced in Genomic Pathology. An intimate knowledge of differences and similarities in the target species is necessary to avoid embarrassing mistakes. Like other areas of pathology, endocrine pathology provides examples of entities that could confuse the inexperienced observer. Relatively few pathologists, regardless of their medical, veterinary, or toxicologic background, have the necessary familiarity or knowledge with GEM. Addi-
140
tional pathologists need to be recruited and trained if society is to benefit from the Genomic Revolution.
Background In antiquity, priest-shamans were responsible for the diagnosis and treatment of both animals and human diseases [3]. This traditional one medicine, still existing in many tribal societies, was broken as medicine became more sophisticated and specialized. Veterinary specialists were required for care and treatment of the transportation animals needed by aristocrats’ wars and amusements. The division between human and animal medicine became codified when separate schools of veterinary and human medicine were developed during the Age of Enlightenment in France and the rest of Europe. The concept of one medicine was re-introduced in the late 1800s by leading physicians such as Virchow and Osler, both of whom were intimately involved with the veterinary profession and infectious disease [2]. However, once again, the diseases of man and animals became separated with the advent of the combustion engine. Laboratory animals, namely, the mouse, became a “model” of human disease and fell into a gray zone, unknown to both physicians and veterinarians. The first were not familiar with rodent medicine, anatomy, and physiology, and the latter were relegated to lab animal husbandry and regulation, rather than science. Research using these models rapidly became the domain of physician-scientists and their PhD colleagues. Pathologists, of medical or veterinary background, remained excluded from the research settings and confined to diagnostics. The modern rebirth of one medicine is credited to the veterinarian and epidemiologist Calvin Schwabe [4]. He, as with Virchow and Osler before him, recognized the relationship between disease and man’s animals. In current times, the term is applicable to newly emerging infectious diseases such as HIV, avian flu, Ebola, Hanta, and a long list of other diseases relating to interactions between man and animals. This version of one medicine has been avidly embraced by both the American Medical Association and American Veterinary Medical Association in July 2007 joint resolutions [2]. However, the sequencing of the genomes of microorganisms, mammals, including human and nonhuman primates, and other species has proven the remarkable genomic similarities between all animals. This demonstration has led to renewed interest in laboratory animals. Instead of a prolonged, frustrating, and often futile search for the molecular origins of spontaneous diseases, the candidate genes can be used to induce the disease. Genetic engineering allows for exogenous genes to be isolated from one species, cloned, grown, and inserted into the genome of another
Endocr Pathol (2008) 19:139–147
species to cause a specific disease. The diseases induced in the mouse using human genes frequently produce morphological mimics of the human disease (Fig. 1) [5]. These genetically engineered animals prove that the genes suspected of causing disease in humans cause the same disease in animals fulfilling modern times “Koch’s postulate”. Thus, genomic biology has given the concept of One Medicine a dramatic new meaning, that of the ultimate molecular reductionism: one gene, one disease, one medicine. Disease has been reduced to a series of molecular interactions and Genomic Medicine offers the promise to cure. The current challenge is to understand the phenotypic effects of these genomic abnormalities on the disease state. Disease is the abnormal function and structure of the organism that results from molecular malfunction but is not the molecular malfunction per se. Thus, pathologists have a critical role in the new medicine. We are the only discipline skilled in interpretation of the microscopic structure of disease. We are also schooled in integrating microscopic structure with the molecular, biochemical, and clinical aspects of disease. Pathologists will need, however, to return to their roots in comparative pathology to effectively function in this new world of genomic pathology. The new genomic pathologist will need to recognize and understand the normal and abnormal structure of more than one species. Furthermore, they will need to understand genomic biology and its implications in sufficient depth to integrate the microscopic pathology with the molecular aspects of the disease. An even more exciting frontier will subsequently appear as we learn to apply the genomic pathology of the current induced mutations in laboratory animals to the spontaneous mutations that occur in humans.
Examples of Subtleties in Endocrine Pathology The endocrine system represents a particular problem because of widespread influences in a number of end organs. Investigators unaware of these effects are susceptible to misinterpretation of end-organ pathology. Examples of pituitary and interstitial cell pathology are informative. The Mouse Pituitary Some strains of mice develop pituitary abnormalities, including adenoma, as a result of normal aging. One strain, Friend virus B type (FVB), is important because it is used for the majority of studies of the mouse mammary gland. Mahler et al. [6] described pituitary adenomas in aging FVB. Nieto et al. [7] described persistent mammary hyperplasia in FVB females after three pregnancies. Finally, Wakefield et al. [8] found that the nulliparous females in
Endocr Pathol (2008) 19:139–147
141
Fig. 1 Images showing adenocarcinomas of mouse (A) and human (B) prostates. The mouse tumor is a reasonable phenocopy of the human with back-to-back glands lined by epithelial cells with pleomorphic nuclei and prominent nucleoli. The mouse adenocarcinoma was created
using a probasin-myc construct to overexpress myc [5]. Note the poorly formed glands in both species. (Slide A courtesy of Dr G. Thomas and Slide B courtesy of Dr A.D. Borowsky)
the NIH FVB colony all developed pituitary abnormalities by 12 months of age with associated abnormalities of the mammary gland. Wakefield’s observations explained an unexpected mammary phenotype thought to be induced by a specific transgene. One can only speculate about other mammary phenotypes in elderly FVB females. Subsequently, we have requested that any collaborator with mammary phenotypes in mice older than 12 months also submit pituitaries. As a result, five other FVB colonies have been identified with pituitary abnormalities [9]. Fortunately, the mammary tumor phenotypes in GEM are unique with signature phenotypes for each genotype. Thus far, the bulk of the pituitary-induced mammary tumors are adenosquamous carcinomas (aka keratoacanthomas) (Fig. 2) [9]. However, we have seen a range of adenocarcinomas and papillary carcinomas (unpublished). In some cases, the pituitary abnormalities have not precluded publication [9]. In others, pituitary adenomas have tragically ended the project for an unfortunate fellow.
of unforeseen results that scientists may encounter along the way. This use-case scenario makes the need for well-trained genomic pathologists even more obvious. In this instance, many of the models based on the insertion of the SV40 Tag gene produce undifferentiated tumors that are actually of neuroendocrine origin. Some of these were originally misdiagnosed as adenocarcinomas by unwary pathologists or by eager but untrained observers. The following brief review relates how the SV40 Tag-neuroendocrine phenotype occurs in a wide variety of tissues under the influence of different promoters.
Testicular Abnormalities The prostate is a major end organ for the male mouse. Many strains of mice and rats have interstitial (Leydig) cell hyperplasia or adenomas. These are frequently associated with abnormalities of the prostate and male accessory organs [10]. We insist on examination of the testes as part of the evaluation of the prostate. The interpretation of prostate pathology is risky without examination of the testes. Unexpected Neuroendocrine Tumors Genetic engineering often has surprising outcomes. The emergence of unexpected neuroendocrine tumors in some “adenocarcinoma of mouse” models is an excellent example
Fig. 2 Slides captured on a flatbed scanner to illustrate a sagittal section of a FVB female mouse head with a large pituitary prolactinoma (upper panel, arrow). The same mouse has multiple adenosquamous carcinomas in the mammary gland (lower panel, arrow)
142
The transgenic adenocarcinoma of mouse prostate (TRAMP) model of prostate cancer in which SV40 Tag expression is driven by the rat probasin promoter has been widely used to model prostate cancer in humans. An extensive literature has appeared using this premise. It now turns out that the fast-growing, aggressive, and metastatic tumors in this model are primarily undifferentiated and express neuroendocrine markers such as synaptophysin (Fig. 3) [11–13]. The remaining, well-differentiated, glandular epithelium develops slow-growing hyperplasias that rarely, if ever, progress to adenocarcinomas. Interestingly, the neuroendocrine phenomena was observed and correctly documented using SV40 Tag behind a modified probasin promoter 3 years after the initial publication of the TRAMP model [14, 15]. Almost simultaneously, neuroendocrine tumors were found in mice with the cryptidinpromoted SV40 Tag [16]. Genomic pathologists experi-
Fig. 3 Images of neuroendocrine tumors of the mouse (A, C, and E) and human (B, D and F). Note the absence of well-formed glands and small, compact nuclei with scanty cytoplasm in both tumors. Compare with the glandular organization of the adenocarcinomas in Fig. 1. The tumors were stained using immunohistochemistry for keratin 8 (C and D) and synaptophysin (E and F). The immunohistochemistry for keratin 8 demonstrates perinuclear dot-like staining (arrows) characteristic of some neuroendocrine tumors
Endocr Pathol (2008) 19:139–147
enced with more than one model quickly recognized that neuroendocrine tumors occurred in the vast majority of SV40 Tag models independent of the promoter [17]. The comparisons provided by genomic pathology workshops incriminated the large T antigen in neuroendocrine tumor development [17]. This is confirmed by studies of secretin-SV40 transgenic mice that develop endocrine pancreatic cancer. These mice develop tumors only when the secretin promoter drives the expression of the intact early region of SV40 (both T and t antigens). Mice expressing a mutant SV40 gene producing only small t antigen do not develop tumors [18]. Recently, the different histology in the SV40 Tag-based models has been shown to also depend on the cell of origin where neuroendocrine tumors arise from transformed progenitor cells, and hyperplasias as well as benign tumors arise from terminally differentiated epithelium [13].
Endocr Pathol (2008) 19:139–147
The TRAMP model also illustrates the well-known fact that promoters are “leaky.” The effect of the SV40 transgene in different cell types may seem puzzling, since the native probasin gene is expressed exclusively by prostatic luminal cells in wild-type mice. However, the truncated version of the probasin promoter used in GEM lacks such specificity. Immunohistochemical studies of transgene expression in the reproductive tract of TRAMP mice shows that SV40 is not limited to luminal cells but is also expressed in prostatic basal epithelial cells, in seminal vesicle mesenchyme, and the stromal smooth muscle cells. A transgenic mouse model of gastric adenocarcinoma provides another example of “leaky promoter” [19]. Mice were engineered to express SV40 Tag behind the Atp4b gene, thought to be restricted to progenitors of acidproducing gastric parietal epithelial cells. The resulting tumors were undifferentiated and expressed neuroendocrine markers. Since pre-parietal cells do not express neuroendocrine markers, these tumors were thought to arise from transdifferentiation. In the Atp4b-SV40 model, neuroendocrine tumorigenesis in the stomach depends on maintenance of the transformed neural precursor cells in an undifferentiated state. In this study, the connection between neuroendocrine tumors and transformation of progenitor cells by SV40 recapitulates what was observed in the TRAMP model. SV40 also leads to neuroendocrine tumors in the colon when driven by the intestinal trefoil factor (ITF). Endogenous ITF is expressed in goblet cells of all segments of the small intestines and colon. The truncated ITF promoter, however, is only expressed in the proximal colon of transgenic mice and in the duodenum of one line. Although transformed cells in the duodenum of these mice become hyperplastic and loose goblet cell morphology, they do not give origin to tumors. Colonic neuroendocrine carcinomas are the only malignancy in the trefoil-SV40 mouse, originally designed to model mucinous colorectal adenocarcinoma [20]. Not much is known about the mechanism by which SV40 Tag induces neuroendocrine carcinomas. It appears that large T antigen and undifferentiated cells are prerequisites, pointing toward a role for simultaneous blockage of p53 and Rb in progenitor epithelial cells of certain organs. In spite of the lack of mechanistic data, the recognition of the morphologic link between SV40 Tag expression and neuroendocrine differentiation served to “educate” wellinformed genomic pathologists. The trained eye now knows to recognize neuroendocrine carcinoma arising from simultaneous p53 and Rb disruption even in non-Tag models. Small cell carcinoma can be induced in the lung by conditional inactivation of both Rb1 and Trp53 in pulmonary epithelium [21]. Neuroendocrine small cell carcinomas also arise in the lungs of double transgenic CC10-SV40 Tag/hASH1 (human achaete-scute homolog 1) but not in
143
single transgenic CC10-hASH1 mice [22, 23]. Furthermore, deletion of Rb and Trp53 predispose mice to developing thyroid medullary carcinomas, yet another neuroendocrine tumor [24]. These observations reinforce the causal relationship between loss of p53 and Rb and neuroendocrine carcinoma development and further inform the pathologist. The scientific community using SV40 models of adenocarcinoma should be, therefore, aware of the possibility of neuroendocrine differentiation previously recognized by the well-trained genomic pathologist. In the last decade, poorly differentiated carcinomas arising in the mammary gland, urinary bladder, thyroid, and ovaries of SV40 transgenic mice have been reported in the literature [25–28]. It is disconcerting to note that neuroendocrine differentiation may not have been investigated in these models. The lack of differentiation, aggressiveness, and cell morphology in the resulting tumors suggest a neuroendocrine origin. SV40 and Anticipated Neuroendocrine Tumors The reverse of the Tag models can be informative but misleading. It may not be a coincidence to find that many of the intentional neuroendocrine tumor models deploy the SV40 transgene. The CR2-Tag model, for instance, brings SV40 Tag driven by the cryptdin-2 promoter, which is considered specific to neuroendocrine cells of the prostate [29]. This model was specifically designed to clarify neuroendocrine cell biology and tumorigenesis. Other examples of expected neuroendocrine tumor models are the RIP-Tag2 and the pTet-on/pTRE-SV40 mice, both modeling islet cell tumors, and the c-kit - SV40 mice that develop multiple neuroendocrine tumors [30, 31]. The difficulty here lies in separating what molecular events are truly expected of neuroendocrine differentiation in a certain organ from the fortuitous effects of SV40 alone, regardless of the promoter used. On the other hand, the molecular pathogenesis of these models will likely reflect the biology of SV40 T antigen and may be useful when translated to other, non-Tag models of neuroendocrine carcinoma. A trained genomic pathologist is able to both recognize morphologic traits that may suggest a certain molecular lesion and understand the molecular basis of a broad spectrum of lesions. This two-way knowledge is what makes genomic pathologists a vital link between creator and creation in the GEM world. DIY Pathology Many investigators wish to perform their own pathology. A recent survey from Europe suggests that the vast majority of necropsies are performed by laboratory staff [10]. This is
144
consistent with our more empirical observations. Most materials sent to us are obtained and submitted by untrained laboratory technicians or the newest graduate student. The comedy of errors begins there. Autolyzed samples unfixed in spent paraformaldehyde, poorly sampled and mislabeled specimens abound. Sometimes, it reaches tragic irretrievable proportions as when a graduate student appeared with 7 years worth of poorly fixed, uninterpretable mouse uteri. In another instance, transgenic mice with an expected thyroid phenotype were killed and submitted when tumors could be palpated, according to protocol. Unfortunately, the tumors turned out to be enlarged submandibular lymph nodes, the thyroids were not sampled, and the transgenic mice were wasted. These types of errors are sometimes compounded by DIY pathology practiced in many institutions. DIY has created a literature filled with obvious errors, which are not only humiliating but may also have important repercussions. An account of scientific articles containing DIY pathology has been published [11]. Other unpublished examples from our files include a tumor sent as either an undifferentiated carcinoma or a carcinoid of the intestine. Fortunately, we identified the lesion as a granuloma induced by a round worm. The phenotype disappeared when the parasitic infestation was cleared. Another example was a strangely localized facial “multiple myeloma” (plasma cell tumor) diagnosed by an excited researcher studying hematological disorders. When later examined the “tumor” contained large colonies of Staphylococcus sp. amidst the plasma cells. Treating the bacterial infection, also known as Botryomycosis, eliminated the supposed myeloma. The thyroid researcher tends to see only thyroid, the blood researcher tends to see only hematological disorders, and all seem to think mouse is only model. Whereas the trained genomic pathologist can recognize a variety of lesions, molecular, and infectious in a wide range of species and is able to see the transgenic mouse as a whole providing correct interpretation of phenotype. Many of the double and triple transgenic mice are generated in order to accelerate a certain tumor phenotype. Nonetheless, most of the reports in the literature mention “tumor” without any distinction between benign or malignant cancer. Tumor incidence is, in such cases, insufficient information. In some cases, “tumors” are diagnosed based solely on macroscopic appearance with many an ovarian cyst being misdiagnosed as a tumor and lymphoma, the “wild card” of the mouse world, disguised as a variety of metastatic lesions. For example, the Pten heterozygous mouse, extensively studied due to the role of this important tumor suppressor gene in the pathogenesis of many tumors, develops thyroid and adrenal masses. Although the incidence of thyroid masses is provided in the literature, the type of tumor
Endocr Pathol (2008) 19:139–147
(follicular, papillary, or medullary) is often neglected and the distinction between nodular hyperplasia, adenoma, and carcinoma often not made [32–34]. The same is true for adrenal tumors in this mouse, all pheochromocytomas, which also happen to be somewhat common in aging mice. As irony would have it, Pten mice also develop lymphoma at higher rates than wild-type mice, and lymph nodes are located dangerously close to both the thyroid and the adrenal glands. Researchers have crossed Pten mutant mice with many other mutants in search of new or more severe phenotypes. How can the outcome of a compound genetic lesion in double or triple transgenic mice be judged when data on the original animal are either missing or faulty? The skill and experience to grade tumors and to recognize subtle and often useful morphologic changes caused by the added genetic lesion in these compound mutants distinguishes the well trained genomic pathologist from the amateur morphologist practicing DIY pathology. Another important consideration regarding the increasingly popular double and triple mouse mutant crosses is the background strain in which each GEM is created. Besides considering spontaneous background lesions in the interpretation of experimental outcome and gene function, one also needs to evaluate the complex interplay between genetic background and the gene of interest. The high incidence of pituitary tumors in the FVB strain and their relationship with mammary tumors, as discussed above, is an excellent example of background lesion interfering with experimental outcome. Another example in the realm of endocrine pathology is the development of pheochromocytomas in Nf1n31 transgenic mice, which only happens in a mixed background [35]. NPcis mice, which are heterozygous for the Nf1 gene and used as a model for Neurofibromatosis type 1, provide further evidence of the importance of background strain. When messenger RNA is analyzed quantitatively, the magnitude of change in expression levels of Nf1 between NPcis and wild-type littermates is similar to that seen between wild-type mice on the B6 and 129 strains. This means that brain tissue from a B6 NPcis mouse expresses the same amount of Nf1 as brain tissue from a wild-type 129 mouse. Nf1 expression level correlates to astrocytoma susceptibility in this model; therefore, background strain deeply impacts interpretation of experimental outcome in haploinsufficiency studies [36]. The TRAMP model, also discussed above, offers yet another example. In the C57BL/6 background, the typical neuroendocrine carcinomas caused by SV40 Tag occur much later and in much lower frequency than in FVB mice carrying the same oncogene. This difference in neuroendocrine tumor formation between these two strains accounts for the fact that B6 TRAMP mice live twice as long as their FVB counterparts [11]. Although the molecular basis for the strain differences mentioned above remains unknown, the cases serve to
Endocr Pathol (2008) 19:139–147
illustrate the fact that background strain can confound the Koch’s postulate of phenogenomics. Here again, the need for trained genomic pathologists to be aware of the possible effects of mouse background strain in human disease models becomes acute. Education As documented in the preceding review, genomic sciences, and specifically endocrine scientists, frequently have been poorly served by pathology. Endocrine pathology is but an illustrative microcosm of the broader problem. Comparative pathology has been ignored during the genomic revolution as molecular biologists have placed their faith in molecular phenotyping. The failure to integrate morphological structure of disease into genomic analyses has been disastrous. The failure also resides with those of us who are pathologists. Veterinary pathology has largely concentrated on domestic, exotic, and herd animals. Medical pathology increasingly concentrated on the diagnostic pathology of human disease. Comparative pathology and laboratory mice have fallen between the two disciplines with very few pathologists interested, or trained, in comparative pathology of the mouse, in particular genetically modified mice. Very few resident training programs in either discipline have sufficient knowledgeable faculty to provide the requisite training. Meanwhile, the human and mouse (and 720 other) genomes have been sequenced. The next obvious step is to understand the function of each of the 20,000 +/− genes. The functional characterization of each gene requires phenotyping. The phenotyping will be done using the genetically engineered laboratory mouse. In anticipation of this step, giant collaborative projects have been developed to knockout each gene creating estimations of between 20,000 and 200,000 new lines of mice. Who will phenotype 20,000 mutant mice? The same scientists that brought us DIY pathology? With the current shortage of qualified pathologists, science is threatened with disastrous chaos. The shortage of pathologists has been recognized and discussed in several other publications [2, 37–39]. The short-comings of current educational systems have been discussed elsewhere [2]. Some admirable workshop programs to rectify the shortages have been mounted. However, our concepts of pathology education are mired in traditional “across-the-microscope apprenticeship” formats. Moreover, well-trained, dedicated comparative pathologists are rare and geographically dispersed. As a result, most training programs do not have faculty that is truly up to the task. We do not have time to revise our entire educational system or develop programs to recruit the large number of qualified pathologists needed. As professionals, we are responsible for our own field and need to own up to
145
that responsibility by ending our whining and devising solutions. One solution is to create educational programs that make use of the informatics age and electronic media. If the apprentice cannot come to the master, why not make the master available to the apprentice? With the advent of widespread whole-slide imaging (WSI) and telecommunication system, we should be able to prepare the next generation with our accumulated knowledge and resources over the Internet. The modern technology allows interactive sharing of images and knowledge, real-time or asynchronous, over the entire globe. High-quality WSI microscopic slides are being captured, annotated, archived, and displayed over the internet. The apprentices can now view them on their web browser at any time, at any magnification, and at any field on the slide. These slides are being broadcast over the Internet using live, online applications allowing the master to teach the apprentice or the moderator to elicit the simultaneous opinions of dispersed experts. These informal education sessions can be formalized and used to teach and prepare the next generation. Formal educational programs designed for different levels of training and expertise are being designed. Since most of the initial contacts with diseased mice are by trainees and technical staff without the benefit of pathologists, it is incumbent upon us to train the neophytes who frequently necropsy, sample, and process the organs in proper technique and the rudiments of pathobiology. Principal investigators who lead laboratories also need to learn the basics of pathology and, frequently, need an understanding of the diseases appearing in their mice. Certified pathologists recruited into the field, medical and veterinary alike, will need to learn about the peculiarities of genetically modified mice and the human diseases they are suppose to emulate. Genomic pathologists, practicing in isolation of their institution, need to be able to consult with their peers and have an easy mechanism for keeping abreast of current developments. All participants should have the opportunity to check their impressions with archived examples and benefit from second, informed opinions from the experts. We believe that all of the above is possible and have initiated the Center for Genomic Pathology as a repository and educational resource (http://ctrgenpath.org). We envision the Center as a first-order, modern solution for our current dearth of experts in genomic pathology and a rallying point for the next generation. In order to achieve our goals, the Center for Genomic Pathology will need your support. We are open and encourage your suggestions and commentary. Our growth and development will depend on funding and your efforts. Please rally to our call for action and join the movement.
146 Acknowledgments This work was, in part, supported by grants U01 CA084294 and NCI 2 P30 CA093373 from the US National Cancer Institute and U42 RR14905 from US National Center for Research Resources of the National Institutes of Health. The authors appreciate the contribution of slides by Drs A.D. Borowsky and G.V. Thomas and the image preparation by R.J. Munn.
References 1. Cardiff RD, Rosner A, Hogarth MA, Galvez JJ, Borowsky AD, Gregg JP. Validation: the new challenge for pathology. Toxicol Pathol 32(Suppl 1):31–39, 2004. doi:10.1080/01926230490424662. 2. Cardiff RD, Ward JM, Barthold SW. ‘One medicine—one pathology’: are veterinary and human pathology prepared? Lab Invest 88:18–26, 2008. doi:10.1038/labinvest.3700695. 3. Schwabe CW. Cattle, priests, and progress in medicine. Minneapolis: Univ. of Minnesota Press, 1978. 4. Schwabe CW. Veterinary medicine and human health. Baltimore: Williams and Wilkins, 1984. 5. Ellwood-Yen K, Graeber TG, Wongvipat J, et al. Myc-driven murine prostate cancer shares molecular features with human prostate tumors. Cancer Cell 4:223–238, 2003. doi:10.1016/S1535-6108 (03)00197-1. 6. Mahler JF, Stokes W, Mann PC, Takaoka M, Maronpot RR. Spontaneous lesions in aging FVB/N mice. Toxicol Pathol 24:710– 716, 1996. 7. Nieto AI, Shyamala G, Galvez JJ, Thordarson G, Wakefield LM, Cardiff RD. Persistent mammary hyperplasia in FVB/N mice. Comp Med 53:433–438, 2003. 8. Wakefield LM, Thordarson G, Nieto AI, et al. Spontaneous pituitary abnormalities and mammary hyperplasia in FVB/NCr mice: implications for mouse modeling. Comp Med 53:424–432, 2003. 9. Kwak EL, Kim S, Zhang J, Cardiff RD, Schmidt EV, Haber DA. Mammary tumorigenesis following transgenic expression of a dominant negative CHK2 mutant. Cancer Res 66:1923–1928, 2006. doi:10.1158/0008-5472.CAN-05-1237. 10. Borowsky AD, Dingley KH, Ubick E, Turteltaub KW, Cardiff RD, Devere-White R. Inflammation and atrophy precede prostatic neoplasia in a PhIP-induced rat model. Neoplasia 8:708–715, 2006. doi:10.1593/neo.06373. 11. Chiaverotti T, Couto SS, Donjacour A, et al. Dissociation of epithelial and neuroendocrine carcinoma lineages in the transgenic adenocarcinoma of mouse prostate model of prostate cancer. Am J Pathol 172:236–246, 2008. doi:10.2353/ajpath.2008.070602. 12. Huss WJ, Gray DR, Tavakoli K, et al. Origin of androgeninsensitive poorly differentiated tumors in the transgenic adenocarcinoma of mouse prostate model. Neoplasia 9:938–950, 2007. doi:10.1593/neo.07562. 13. Zhou Z, Flesken-Nikitin A, Nikitin AY. Prostate cancer associated with p53 and Rb deficiency arises from the stem/progenitor cellenriched proximal region of prostatic ducts. Cancer Res 67:5683– 5690, 2007. doi:10.1158/0008-5472.CAN-07-0768. 14. Greenberg NM, DeMayo F, Finegold MJ, et al. Prostate cancer in a transgenic mouse. Proc Natl Acad Sci U S A 92:3439–3443, 1995. doi:10.1073/pnas.92.8.3439. 15. Kasper S, Sheppard PC, Yan Y, et al. Development, progression, and androgen-dependence of prostate tumors in probasin-large T antigen transgenic mice: a model for prostate cancer. Lab Invest 78:319–333, 1998. 16. Garabedian EM, Humphrey PA, Gordon JI. A transgenic mouse model of metastatic prostate cancer originating from neuroendocrine cells. Proc Natl Acad Sci U S A 95:15382–15387, 1998. doi:10.1073/pnas.95.26.15382.
Endocr Pathol (2008) 19:139–147 17. Shappell SB, Thomas GV, Roberts RL, et al. 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, 2004. doi:10.1158/0008-5472. CAN-03-0946. 18. Ratineau C, Ronco A, Leiter AB. Role of the amino-terminal domain of simian virus 40 early region in inducing tumors in secretin-expressing cells in transgenic mice. Gastroenterology 119:1305–1311, 2000. doi:10.1053/gast.2000.19278. 19. Syder AJ, Karam SM, Mills JC, et al. A transgenic mouse model of metastatic carcinoma involving transdifferentiation of a gastric epithelial lineage progenitor to a neuroendocrine phenotype. Proc Natl Acad Sci U S A 101:4471–4476, 2004. doi:10.1073/ pnas.0307983101. 20. Gum JR Jr., Hicks JW, Crawley SC, et al. Mice expressing SV40 T antigen directed by the intestinal trefoil factor promoter develop tumors resembling human small cell carcinoma of the colon. Mol Cancer Res 2:504–513, 2004. 21. Meuwissen R, Linn SC, Linnoila RI, Zevenhoven J, Mooi WJ, Berns A. Induction of small cell lung cancer by somatic inactivation of both Trp53 and Rb1 in a conditional mouse model. Cancer Cell 4:181–189, 2003. doi:10.1016/S1535-6108 (03)00220-4. 22. Linnoila RI, Sahu A, Miki M, Ball DW, DeMayo FJ. Morphometric analysis of CC10-hASH1 transgenic mouse lung: a model for bronchiolization of alveoli and neuroendocrine carcinoma. Exp Lung Res 26:595–615, 2000. doi:10.1080/01902140150216693. 23. Wikenheiser KA, Clark JC, Linnoila RI, Stahlman MT, Whitsett JA. Simian virus 40 large T antigen directed by transcriptional elements of the human surfactant protein C gene produces pulmonary adenocarcinomas in transgenic mice. Cancer Res 52:5342–5352, 1992. 24. Knostman KA, Jhiang SM, Capen CC. Genetic alterations in thyroid cancer: the role of mouse models. Vet Pathol 44:1–14, 2007. doi:10.1354/vp.44-1-1. 25. Connolly DC, Bao R, Nikitin AY, et al. Female mice chimeric for expression of the simian virus 40 TAg under control of the MISIIR promoter develop epithelial ovarian cancer. Cancer Res 63:1389–1397, 2003. 26. Green JE, Shibata MA, Yoshidome K, et al. The C3(1)/SV40 T-antigen transgenic mouse model of mammary cancer: ductal epithelial cell targeting with multistage progression to carcinoma. Oncogene 19:1020–1027, 2000. doi:10.1038/sj.onc.1203280. 27. Grippo PJ, Sandgren EP. Highly invasive transitional cell carcinoma of the bladder in a simian virus 40 T-antigen transgenic mouse model. Am J Pathol 157:805–813, 2000. 28. Ledent C, Dumont J, Vassart G, Parmentier M. Thyroid adenocarcinomas secondary to tissue-specific expression of simian virus-40 large T-antigen in transgenic mice. Endocrinology 129:1391–1401, 1991. 29. Hu Y, Ippolito JE, Garabedian EM, Humphrey PA, Gordon JI. Molecular characterization of a metastatic neuroendocrine cell cancer arising in the prostates of transgenic mice. J Biol Chem 277:44462–44474, 2002. doi:10.1074/jbc.M205784200. 30. Bosse P, Bernex F, De Sepulveda P, Salaun P, Panthier JJ. Multiple neuroendocrine tumours in transgenic mice induced by c-kit-SV40 T antigen fusion genes. Oncogene 14:2661–2670, 1997. doi:10.1038/sj.onc.1201122. 31. Shen Q, Sun Q, Wei X, et al. Generation and characterization of islet cell tumor in pTet-on/pTRE-SV40Tag double-transgenic mice model. J Biosci Bioeng 104:14–21, 2007. doi:10.1263/ jbb.104.14. 32. Di Cristofano A, Pesce B, Cordon-Cardo C, Pandolfi PP. Pten is essential for embryonic development and tumour suppression. Nat Genet 19:348–355, 1998. doi:10.1038/1235.
Endocr Pathol (2008) 19:139–147 33. Podsypanina K, Ellenson LH, Nemes A, et al. Mutation of Pten/ Mmac1 in mice causes neoplasia in multiple organ systems. Proc Natl Acad Sci U S A 96:1563–1568, 1999. doi:10.1073/ pnas.96.4.1563. 34. Stambolic V, Tsao MS, Macpherson D, Suzuki A, Chapman WB, Mak TW. High incidence of breast and endometrial neoplasia resembling human Cowden syndrome in pten+/− mice. Cancer Res 60:3605–3611, 2000. 35. Tischler AS, Shih TS, Williams BO, Jacks T. Characterization of pheochromocytomas in a mouse strain with a targeted disruptive mutation of the neurofibromatosis gene Nf1. Endocr Pathol 6:323–335, 1995. doi:10.1007/BF02738732.
147 36. Hawes JJ, Tuskan RG, Reilly KM. Nf1 expression is dependent on strain background: implications for tumor suppressor haploinsufficiency studies. Neurogenetics 8:121–130, 2007. doi:10.1007/ s10048-006-0078-5. 37. Barthold SW, Borowsky AD, Brayton C, et al. From whence will they come? - A perspective on the acute shortage of pathologists in biomedical research. J Vet Diagn Invest 19:455–456, 2007. 38. Cardiff RD. Pathologists needed to cope with mutant mice. Nature 447:528, 2007. doi:10.1038/447528c. 39. Valli T, Barthold SW, Ward JE, et al. Over 60% of NIH extramural funding involves animal-related research. Vet Pathol 44:962–963, 2007, (author reply 963–964).
