Читать книгу Pathology of Genetically Engineered and Other Mutant Mice - Группа авторов - Страница 90

Gene targeted (null mutant or “knockout” [KO]) mouse embryos, and less often transgenic mouse embryos, develop abnormal organ‐ or tissue‐specific lesions due to the functional abnormalities arising from altering the role of the genetically engineered gene at different stages of normal development. In large‐scale knockout mouse programs, about 30% of all inactivated genes lead to developmental embryonic defects at various stages of gestation [78, 79]. In one study of 242 developmental lethal knockout mouse lines, 44% of lethal events occurred before GD9.5, 17% from GD9.5–12.5, 4% from GD12.5–18.5, and 36% from GD15.5 to weaning [78]. The specific developmental defect(s) in the 242 lines were not determined.

The stage at which injury (lack of normal development of the entire embryo, or specific tissues or organs) occurs will determine the outcome of developing mice (Table 5.1). For harm prior to implantation (i.e. before GD4.5), significant interference to normal development usually will lead to death of the embryo, while modest effects may engender developmental delays but minor consequences or no overt abnormalities. Such delays may undergo complete or partial compensatory repair as the remaining pluripotent stem cells restore lost tissue. Gross abnormalities (“birth defects” seen during gestation) most commonly result from damage incurred after implantation (GD5.0) but before the end of organogenesis (GD15.0), with the character of the malformation determined by the timing of the insult relative to various organ‐specific “critical periods.” For injury after organogenesis is finished (i.e. GD15.0 through the neonatal period), the embryo may exhibit slow growth or develop cell type‐ or tissue‐specific histopathological changes and/or functional abnormalities rather than gross defects; the exceptions to this rule are structures like the cerebellum and hippocampus that undergo major peaks of cell proliferation after birth. Embryonic lethality may result from many different causes, but common aberrations that reliably lead to death are decreased circulation (e.g. heart and/or vascular defects any time during gestation) and reduced oxygen exchange (e.g. erythrocyte deficiencies or hemorrhagic diatheses or placental insufficiency, especially occurring late in gestation).

Gross developmental defects that occur as spontaneous background lesions vary in frequency and severity among wild‐type (WT) mouse strains and lines. Minor incidental anomalies like renal pelvic cavitation (physiological hydronephrosis) and extra or wavy ribs are common (from 5% to 35% of animals) since they do not impact embryonic viability [80]. Major malformations like cleft palate and neural tube defects have low background frequencies (typically 0.5% or less) since they are lethal. An individual may harbor one or several minor or major abnormalities.

Table 5.1 Common embryonic or neonatal defects leading to lethality.

Developmental stage	Developmental days when damage occurs	Target site for mutation or toxicant	Developmental outcome	Developmental timing of lethality
Preimplantation	GD0–3.5	Any cells (totipotent or pluripotent)	Embryonic death (no visible implantation sites)	GD1.0–4.0
Postimplantation	GD4.5–5.0	Any cells (pluripotent)	Embryonic death (no visible implantation sites)	GD5.0–5.5
Gastrulation	GD6.5–9.0	Altered germ layers or body plan	Embryonic death (empty implantation sites)	GD6.0–7.5
Early organogenesis	GD8.0–10.5	Cardiovascular or central nervous system defects, abnormal erythropoiesis	Mid‐gestational death due to cardiovascular or hematopoietic insufficiency	GD10.0–13.0
Late organogenesis	GD11.0–15.0	Defects in many organs, abnormal erythropoiesis, delayed growth	Mid‐ to late‐gestational death due to cardiovascular or hematopoietic insufficiency	GD13.0–16.0
Near‐term growth spurt	GD15.5–18.5	Altered cardiovascular function, abnormal erythropoiesis, delayed growth	Late‐gestational or perinatal death due to cardiovascular or hematopoietic insufficiency (or edema or hemorrhage)	GD16.0 to PND0.5
Early neonatal	PND0	Abnormal cardiovascular, central nervous system, cutaneous, musculoskeletal, or pulmonary function	Insufficiencies in circulation or respiration (inability to breath)	PND0–0.5
Late neonatal	PND0.5–2.0	Abnormal cardiovascular, gastrointestinal, metabolic, or neural functions	Insufficiencies in digestion or suckling (inability to feed)	PND1.0–2.5
Juvenile	PND2.5–28	Abnormal hematopoietic, immune, metabolic, or renal functions	Sustained anemia, abnormal immune function, metabolic or renal insufficiencies	PND3.0–28

GD = gestational day (where GD0 is the vaginal plug‐positive day), PND = postnatal day.

Frequencies of both minor and major defects may be increased by genetic manipulation, infections, or exposure to teratogens. Nearly 5000 abnormal phenotypes have been identified in mutant mouse embryos [79, 81], most of which are associated with monogenic null mutations (“knockouts”). Induced developmental defects have been reported in approximately 30% of genetically engineered knockout mice [78]. Bacterial and viral infections that can cross the placenta may lead to substantial tissue destruction of the embryo, placenta, or both. Numerous toxicants are capable of producing developmental phenotypes in rodents including agricultural and industrial chemicals, drugs, metals, and physical agents (heat and radiation) [82]. The kind of phenotype elicited by a genetic mutation or toxicant exposure will depend on where and when during development the gene normally performs its function. For example, in the brain anomalies in the cerebral cortex tend to arise from genetic defects that arise during mid‐gestation while abnormalities in the cerebellum – which develops late in gestation and after birth in rodents [42] – typically involve genetic dysfunction late in gestation or after birth. Knowledge of normal spatial and temporal patterns of gene expression during development is instrumental in allowing developmental pathologists to define potential target organs and more efficiently characterize developmental phenotypes. In general, the frequency and severity of developmental phenotypes will vary among litters, and among individuals of comparably treated litters. The variation in stages of embryo degeneration, necrosis and death, no matter the etiology, may be seen grossly [83] (Figure 5.15).

Microscopic lesions of many cell types, and tissues and organs may be identified when phenotyping the tissues from developing mice. Identification of normal cell and tissue features for developing rodents is simple thanks to several detailed developmental atlases [7, 11, 12, 14, 15,84–86]. A full appreciation of histopathologic changes for developmental phenotypes often requires that the histogenesis of the lesions be explored by serial evaluation of embryos at several times (e.g. serial days). For example, cardiac defects that initially produce gross structural malformations and/or histopathological abnormalities and altered circulation in the heart may lead over time to secondary lesions in other highly vascularized organs such as liver (congestion followed by necrosis) and lungs (right‐sided congestion) (Figure 5.16).

Figure 5.15 Comparison of variable lesion progression in three litters of wild‐type GD13.5 mouse embryos at four days following maternal Zika virus infection. Several phenotypes are evident, and often cluster in particular litters. Acute death soon after exposure (based on conceptus size) resulted in embryonic resorption (as shown in the bottom row by small, oval, discolored implantation site remnants [arrowheads with dotted lines]). Slightly later embryonic death led to diffuse embryonic necrosis or autolysis rather than resorption (indicated in the middle row by uniform pallor, the absence of a pigmented eye primordium, and indistinct body contours). Other embryos appear to be viable (based on their similar, age‐appropriate size and anatomic features), but some exhibit focal acute hemorrhage (arrowheads with solid lines) as either a delayed effect of the virus or more likely an artifact of slight trauma encountered during necropsy.

Source: Szaba et al. [83] by permission of the Authors under a Creative Commons 4.0 International License.

Figure 5.16 Developmental phenotypes resulting from circulatory disturbances may present as macroscopic or microscopic lesions. Panel (a): Relative to a wild‐type (WT) littermate, knockout (KO) neonates at PND0 that lack the gene for non‐muscle myosin heavy chain II‐B (Myh10^tm2Rsad) have congestive heart failure that leads to a reduced body size, pallor (indicative of poor perfusion), and a dark band (asterisk) in the abdomen (due to chronic hepatic congestion). Relative to the WT animal (Panel (b)), the microscopic lesions in the KO animal (Panel (c)) stem from myocardial dysgenesis (shown by the larger size and random arrangement of ventricular myofibers). Stain: H&E.

Source: Tullio et al [98] with permission of the U.S. National Academy of Sciences.

Several gross and microscopic patterns occur commonly in development phenotypes that arise during gestation (Figure 5.17). Cell degeneration (potentially reversible) and/or necrosis (irreversible) may be evident and often presents as organs with aberrant cellular features (for degeneration) or that are small (for either, but especially for necrosis). Heterotopiae are clusters of ectopic (displaced) cells that usually indicate prior abnormalities in migration. Kinetic defects appear as a surplus or dearth in either proliferation or programmed cell death (PCD) for a particular population. Any or all of these intrinsic mechanisms as well as other extrinsic tissue changes like hemorrhage may produce gross defects. In general, inflammation is not a prominent tissue response in mouse embryos since the immune system is functionally immature until fairly late in gestation or after birth. The gross appearance of the uterus and implantation sites (Figure 5.17a) as well as the microscopic lesion spectrum in a dying embryo varies with the tissue or organ involved (Table 5.1), but once viability is lost, the microscopic pattern invariably will include first focal or multifocal necrosis in various organs (Figure 5.17e) and then diffuse end‐stage necrosis of all organs (Figure 5.17d,f). Nonviable conceptuses grossly will appear as small implantation sites either lacking an embryo (a finding termed a “resorption,” and typically appearing as a pale, dark green, or purple tissue mass) or with cloudy intra‐amniotic fluid surrounding a white, friable embryo (Figure 5.17b,d).

Many different developmental disturbances may lead to neonatal lethality. Death within one to two hours of birth is a reliable indicator of functional or structural defects that prevent effective oxygen uptake and distribution. Commonly affected systems in this regard include the heart and vessels, lungs, hematopoietic tissue (red blood cells), and brain (respiratory centers). Mice can be born without an intact brain, endocrine glands, limbs, lungs, and skin because these organs are not needed for viability in utero, but only after birth. Acute but slower death (i.e. 6–36 hours after birth) typically results from impaired energy metabolism (especially glucose utilization) or suckling difficulties (Figure 5.18). The pathogenesis of these effects may be simple and obvious (e.g. small or no lung lobes), but often the cause is associated with subtle (e.g. reduced neurons in brainstem centers that control respiration) or no (e.g. altered neurotransmission in synapses of respiratory muscles) structural lesions. For this reason, investigations of developmental lethal phenotypes require a team approach.

For mutant mouse lines, the time of death may be probed further by evaluating the genotype of viable embryos and neonates. The first clue that a given genotype may be responsible for an embryonic or neonatal lethal phenotype often is an altered Mendelian ratio and/or a decrease in the expected litter size (Figure 5.17a) [87]. The developmental stage at which lethality occurs may be estimated by the gross appearance of abnormal individuals relative to the expected appearance of developmental stage‐matched embryos or neonates. If all implantation sites contain structurally normal embryos with a wild‐type (WT) or heterozygous (HET) genotype, then the null mutant (“knockout” [KO]) embryos likely died prior to implantation (i.e. before GD4.5). If some implantation sites contain no embryo or a small dead embryo, then the KO embryos likely expired during initial early organogenesis before generation of the DP (i.e. between GD8.5–9.5). If implantation sites at mid‐gestation (i.e. GD12.5) include apparently normal KO embryos, then the embryo lethal phenotype likely is expressed near or just after birth. Dead pups often are eaten immediately by the dam, so harvesting embryos right before birth (i.e. GD17.5–18.5) might be required to confirm that animals were viable throughout gestation. Given this spectrum of time‐dependent defects, most developmental pathologists tend to perform the initial evaluation to characterize a genetically engineered lethal phenotype by examining litters toward the end of gestation (between GD16.0–18.5; Figure 5.19) or at weaning (PND21; Figure 5.20) [22, 88, 89]. If necessary to pinpoint the time of death and/or characterize the phenotype, follow‐up examinations usually are conducted by working backward at two‐day intervals during gestation (GD14, GD12, GD10, etc.) or two‐ to three‐day intervals between birth and weaning (PND18, PND15, PND12, PND9, PND6, PND4, PND2) until viable but affected individuals are observed.

For developmental toxicity studies, the developmental outcome also is predicted chiefly by the timing of exposure relative to critical periods of development. However, other factors also play a role in the vulnerability of developing mice. For example, some strains have an inherently higher sensitivity to some forms of birth defects because their background incidence is high (e.g. A/J mice and cleft palates [90, 91], SWV [Swiss Webster Vancouver] mice and neural tube closure defects [92]). This strain‐specific sensitivity usually is for a specific spectrum of birth defects and/or teratogenic agents, and not for all kinds of malformations and teratogens generally. Similarly, male and female embryos may respond differently to toxicants based on their proximity to littermates of the opposite sex during gestation and the nutritional status of the dam. Therefore, the pathologist will need to be familiar with the background strain sensitivity as well as the specific intrauterine environment experienced by the embryo prior to performing developmental pathology evaluations.

In the simplest scenario, a lethal developmental phenotype will reliably produce a common spectrum of lesions affecting the same organs in all individuals with a given genetic mutation or that were exposed to a particular dose of toxicant. However, in many cases (for mutant lines and infectious diseases more so than toxicant exposures), lethal phenotypes exhibit variable penetrance, with some embryos exhibiting substantial defects and/or death early in development while a few live well after birth. Typically, no explanation can be defined to account for the differing penetrance, although factors like the genetic background, gene dosage, and maternal age commonly are implicated. The pathologist's role in such investigations is to properly describe the lesions associated with lethality, and to record any shifts in the lesion constellation noted in the long‐term survivors.

Figure 5.17 Embryonic death is demonstrated by a spectrum of macroscopic and microscopic changes. Panel (a): Gravid uterus at GD10.0 in which all but three implantation sites exhibit various degrees of regression (termed “resorption”). Panel (b): An isolated implantation site at GD15.0 in which a dead (pale) embryo is obscured by opaque, red‐tinged amniotic fluid (i.e. containing increased protein and cells, including blood). A viable wild‐type GD14.5 embryo (Panel (c)) with tan skin and many branching blood vessels is contrasted to a white, avascular, wild‐type littermate (Panel (d)) that died at approximately GD13.5 (as shown by the obvious albeit short digits on the forepaw and digital rays (which arise about GD12.8) but very minimal digits on the hind paw). The grossly visible pallor corresponds to diffuse necrosis microscopically. Panel (e): Focal necrosis and acute hemorrhage in a GD13.5 embryo are the first harbingers of imminent embryonic death. Panel (f): Diffuse necrosis leading to death in a GD13.5 embryo (asterisk) is a consequence of widespread primary placental necrosis and acute hemorrhage at four days following maternal Zika virus infection; placental failure in this model causes embryonic death but not embryonic infection. Stain (panels (e) and (f)): H&E.

Sources: Ward and Devor‐Henneman [44] with permission of Iowa State University Press, Bolon and Ward [39] with permission of CRC Press, and Szaba et al. [83] by permission of the Authors under a Creative Commons 4.0 International License.

Figure 5.18 A common cause of neonatal lethality is inability to suckle, which is shown by the inability to discern a “milk spot” (milk‐engorged stomach [arrow]) through the left abdominal wall. Relative to wild‐type (WT) littermates, knockout mice lacking the gene for cardiotrophin‐like cytokine (Clcf1^tm1Zou) die by PND1 from dehydration and starvation resulting from deficient numbers of brainstem neurons needed to mediate nursing.

Source: Zou et al. [99] by courtesy of Sage Publications.

Figure 5.19 An outcome‐oriented decision tree for evaluating embryonic lethal phenotypes in developing mice. The initial analysis is performed near term (gestational day [GD] 17.5–18.5). If necessary, one or more follow‐up experiments may be needed earlier in gestation, usually either moving backward in two‐day intervals (GD16, GD14, GD12, GD10…) or selecting a new time point based on the external features of regressing implantation sites or dead embryos.