Lack of Chronic Histologic Lesions Supportive of Sublethal Spontaneous Seizures in FVB/N Mice

Abstract

FVB/N mice with ‘space cadet’ syndrome are prone to audiogenic seizures and are considered excitotoxic ‘sensitive’ mice due to the neuronal damage that accompanies seizures. FVB/N mice found dead demonstrate acute neuronal cell death—attributed to a massive seizure episode—within the hippocampus and cerebrocortical laminae. However, the behavioral features of FVB/N mice and numerous studies using excitotoxins to induce seizure activity indicate that this strain experiences multiple sublethal seizures. To assess whether FVB/N mice develop histologically detectable lesions, we evaluated the brains of 86 aged (154-847 d) FVB/N mice without a history of seizures. The hippocampus and cerebrocortical laminae were evaluated histologically for neuronal atrophy and gliosis. Neuronal atrophy was quantified by counting neurons in the hippocampus (CA3 and dentate gyrus) and cerebral cortex. Gliosis was quantified by using immunohistochemistry for glial fibrillary acidic protein and glial counting in the cerebral cortex. In addition, ventricular area was calculated. Our study revealed no changes in brain weight with age, no neuronal loss or gliosis, no correlation between neuronal or glial cell profile densities and brain weight or age, and no differences in ventricular size between FVB/N and control mice. Neuronal densities in the cerebral cortex and granule cells of the dentate gyrus were lower in FVB/N mice than in control Swiss Webster mice. We conclude that although acute lesions of seizure activity are a previous feature of the FVB/N strain, chronic seizure activity in these mice either is negligible or does not cause morphologic or phenotypic changes.

Mouse models are used widely to study the pathogenesis of various epileptic seizure disorders of humans. Seizures with concomitant histologic lesions have been engineered in rodent models, which have been instrumental in studies of the pathophysiology of epileptic disorders in humans.⁴ Naturally occurring seizure activity in humans and veterinary species frequently originates in the hippocampus and occurs through a common excitatory pathway that includes glutaminergic neuronal depolarization, resulting in massive influx of calcium into the cell, and can ultimately result in neuronal cell death.²⁴ Excitotoxic mechanisms also are mediated by glutamate, leading to cell death in the CA1 and CA3 regions of the hippocampus as well in the amygdala, piriform cortex, cerebellar cortex, and cerebral cortex.¹⁴ Both mouse strain and age affect an animal's response to epileptogenesis.²⁴ Compared with their younger counterparts, aged rodents have an increased susceptibility to seizure disorders,³ are more susceptible to hypoxia-induced cell damage,¹⁰ and experience more severe hippocampal degeneration after seizure activity.²⁴

The FVB/N mouse strain, which is used to study seizures, undergoes both excitotoxin-induced and spontaneous (audiogenic) seizures with resultant neuronal cell death in the hippocampus and cerebral cortex.^6,17 Numerous studies have indicated strain-specific differences in agent-induced seizurogenesis in mice.^{7,9,13,19,21,23,24,26} FVB/N mice exhibit seizures as well as severe neuronal degeneration and necrosis after treatment with excitotoxins, such as kainic acid and pilocarpine.^{13,15,18,21,23} These mice are therefore considered to be ‘excitotoxic sensitive’ or ‘seizure-induced cell-death–susceptible.’^{13,15,18,20-23} In kainic-acid status epilepticus models, FVB/N mice develop permanent morbidity, similar to epilepsy, as a result of neural injury.^26,28 FVB/N mice frequently are studied in comparison with the excitotoxic, seizure-induced cell-death–resistant C57BL/6 mouse strain, which shows only negligible cell death after seizure induction.^13,16,23,28 FVB/N mice have a low-frequency psychomotor seizure threshold of approximately half that of C57BL/6 mice.⁵ A comparison of electroconvulsive threshold between strains showed that female mice of multiple strains display a 10% to 20% lower threshold for seizure induction than do male mice;⁵ this difference is interesting in light of the increased incidence of seizures in female FVB/N mice.^6,17 In addition, aged FVB/N mice are more vulnerable to excitotoxin-induced seizures and subsequent neuronal lesions than are younger animals.¹²

Seizures in FVB/N mice also can be triggered by audiogenic stimuli and are clinically characterized by chewing automatism, ptyalism, and clonic activity.⁶ The ‘space cadet’ syndrome (SCS) of FVB/N mice is poorly characterized. This seizure disorder is more common in middle-aged to older female mice and is associated with behavioral changes, including social withdrawal, reduced fertility, aggression, and excessive postpartum infanticide.^6,17 In this syndrome, sudden mortality—often without premonitory signs—is attributed to a massive, spontaneous seizure episode. Histologically, the cerebral cortex, hippocampus, and thalamus show neuronal cell death and gliosis.^6,17 The histologic lesions of SCS are very similar to those of humans with status epilepticus or mesial temporal lobe epilepsy.^4,24

The earliest hint of SCS is a report in which 20% of 4- to 12-mo-old female FVB/N mice purchased from a commercial vendor were found dead unexpectedly; a few mice displayed seizure-like activity prior to death.¹¹ The subsequent initial description of a spontaneous seizure disorder with neuronal necrosis in FVB/N mice was in 1998.⁶ That report described 68 FVB/N mice with neuronal necrosis identified over a 4-y period; 17.6% had displayed seizure activity.⁶ Although the incidence of neuronal necrosis in this FVB/N line was not reported,⁶ more than 90% of FVB/N mice with neuronal necrosis were female.^6,17 The mice in the earlier cited study⁶ were obtained from 2 commercial vendors; a subsequent study did not identify the vendor.¹⁷ In the later study, 17 unexpected deaths in a group of 47 FVB/N mice were reported; 4 of these mice exhibited neuronal necrosis in the thalamus that was consistent with SCS.¹⁷ The incidence of neuronal cell death was not reported; it was present in all 4 of the brains examined, but not all the mice underwent brain examination.¹⁷ The current incidence of the SCS is unknown. No subsequent articles on spontaneous seizures and neuronal cell death in FVB/N mice have been published (PubMed accessed 7 August 2015); this lack of additional information suggests that some commercial vendors may have eliminated this deleterious, confounding phenotype in recent years.

Both of the previous studies^6,17 identified concurrent neuronal necrosis, astrocyte hypertrophy, and gliosis, suggesting that the seizure activity in some mice was recurrent (chronic) and sublethal.^6,17 We therefore hypothesized that similar morphologic changes secondary to chronic sublethal seizures would be present in other FVB/N mice and that these lesions could be identified and quantified by using histology and immunohistochemistry. To better understand the sublethal manifestations of SCS in FVB/N mice, we collected 61 female and 25 male aged FVB/N mice, which were euthanized at the end of nonneurologic studies. We failed to identify histologic or immunohistochemical lesions of neuronal cell death or gliosis within the regions of the brain typically affected by seizure activity in FVB/N mice.^6,17 We detected lower neuronal densities in the cerebral cortex and dentate gyrus in our FVB/N mice compared to Swiss Webster control mice. Here we present our data and suggest that either sublethal seizure activity (leading to SCS) is rare in the FVB/N cohort we studied or that it fails to lead to histologic lesions of neuronal necrosis and subsequent gliosis.

Materials and Methods

FVB/N mice scheduled for euthanasia were obtained from 2 campus breeding colonies; no attempt was made to obtain mice from commercial vendors or other colonies; we examined 86 (61 female and 25 male) FVB/N mice and 10 (female) Swiss Webster mice in this study. Female FVB/N mice ranged from 161 to 847 d in age, averaging 393 d. Male FVB/N mice ranged in age from 154 to 557 d, averaging 242 d of age. Twenty-nine (27 female and 2 male) mice older than 365 d of age were examined. Mice were purchased from Taconic (Hudson, NY), and every 6 to 9 mo, the breeding stock was replaced by newly purchased mice. No more than 4 generations elapsed without the introduction of new mice into the breeding scheme. Some of the mice we evaluated were transgenic, carrying a mammary-specific promoter and transgene. Mice were collected over a 15-mo period and therefore included animals from multiple generations. Mice were housed and handled in accordance with the Guide for Care and Use of Laboratory Animals in AAALAC-accredited facilities.⁸ Mice had unrestricted access to Teklad Global 2014 Diet (Harlan, Madison, WI). Serologic testing failed to identify antibodies to epizootic diarrhea of infant mice virus, Mycoplasma pulmonis, mouse hepatitis virus, mouse parvovirus, mouse minute virus, Theiler murine encephalomyelitis virus, Sendai virus, ectromelia, lymphocytic choriomeningitis virus, mouse adenovirus types 1 and 2, polyoma virus, reovirus 3, and pneumonia virus of mice. Occasional positive tests for murine norovirus occurred during the course of this study. Mice were euthanized by carbon dioxide inhalation (according to an IACUC-approved protocol) and were necropsied within 1 h of death. Age, sex, body weight, brain weight, and any gross abnormalities were recorded at necropsy. Complete necropsy and histopathologic analyses of the brain and any grossly-identified comorbid condition were performed. Female Swiss Webster mice, the strain from which FVB/N mice were derived, were obtained from Taconic (Hudson, NY) and euthanized at 134 d of age as a regularly scheduled part of the campus sentinel program unrelated to these breeding colonies; they had no observed gross or serologic abnormalities.

Each brain was removed from the calvarium, weighed, and fixed in 10% normal-buffered formalin for at least 24 h before being cut into 13 or 14 sections by using a rodent brain matrix for mice (ASI Instruments, Warren, MI). Sections were routinely processed, cut 5 μm thick, and stained with hematoxylin and eosin. Neuronal densities were calculated by counting the neurons present within equal-size photomicrographs of stained sections; cortical density data represent all neurons within the cortical section, but those of the CA3 region of the hippocampus include only those neurons within the photomicrograph of the histologically defined structure. For the dentate gyrus, only the granule cells in an approximately 350-μm–long segment of the dorsal portion were quantified.

Sections for immunohistochemistry were placed onto Probe-On Plus slides (Thermo Fisher Scientific, Pittsburg, PA). Immunohistochemistry for glial fibrillary acidic protein (GFAP) was performed by the University of Wisconsin Carbone Cancer Center Experimental Pathology Laboratory (Madison, WI) using a polyclonal rabbit antibovine GFAP antibody (Dako, Carpinteria, CA). Briefly, the primary antibody was added at a 1:750 dilution to the slides and incubated for 2 h at room temperature, followed by incubation with the secondary antibody (1:200, goat antirabbit conjugated with biotin, Dako) for 30 min and detection by using the Vectastain ABC reagent (avidin–biotinylated enzyme complex, Vector Laboratories, Burlingame, CA) with 3,3′-diaminobenzidine as the enzyme substrate for color development, as according to the manufacturer's recommendations. Automated quantification of GFAP immunoreactivity in a defined region (10,000 × 25,000 µm) overlying the hippocampus of a coronal temporal lobe section at the level of the third ventricle and habenular nuclei was performed on whole-slide images scanned at 40× magnification (ScanScope XT, Leica Biosystems, Buffalo Grove, IL) by using the Color Deconvolution Algorithm in the Image Analysis Toolbox (Aperio Technologies, Vista, CA). Two pathologists collaborated to delineate staining intensity into percentages of strong, intermediate, and weak positivity through custom-programming of the software; the total percentage positivity was calculated by standard program parameters.

Bilateral lateral ventricle measurements were calculated from photomicrographs of stained sections by drawing the outline of each lateral hippocampus at the level of the third ventricle and habenular nuclei and then using ImageJ (image.nih.gov/ij/) to calculate the area (in μm²).

Statistical analyses were performed by using standard statistical formulas; actual standard deviations are reported. The 2-tailed t test type 3 (for unequal variances) was applied to the brain weight data for male and female mice, but the 2-tailed t test type 2 (for equal variances) was used for brain weight data within equal numbers of old and young female brain weights. Pearson correlation analysis was performed to analyze neuronal and glial counts as compared with increasing brain weight, and R² values are reported. Mann–Whitney nonparametric analysis was applied to compare the control group with FVB/N mice in regard to neuronal density and brain weight, neuronal density and age, and lateral ventricle area.

Results

Clinical and nonneurologic findings.

Among the 86 FVB/N mice evaluated, 2 animals (2.3% of the study cohort) exhibited possible neurologic abnormalities before death: a 308-d-old female mouse was reported to run in circles, and a lactating and pregnant 201-d-old dam was reported to have tremors and an odd posture (holding her tail up). None of the 86 mice had a history of observed seizure activity. All comorbid conditions were subject to histologic diagnosis. Neither of the neurologic cases had histologic lesions that explained the clinical signs. Another 9 mice had tumors (mammary tumors, 3; lymph node lymphomas, 3; uterine histiocytic sarcoma, cervical and uterine hemangioma, and clitoral gland sarcoma: 1 case each). In addition, 3 female FVB/N mice were pregnant, another had retained fetuses, one had hydrometra, and one had mouse urologic syndrome. No gross abnormalities were noted in any male mouse.

Brain weight.

The brain weight (mean ± 1 SD) in female FVB/N mice was 0.501 ± 0.058 g (range, 0.396 to 0.694 g); in male mice, it was 0.492 ± 0.040 g (range, 0.425 to 0.594 g). Our data do not recapitulate reported averages for brain weights in 12.5- to 16-wk-old female FVB/N mice (0.431 to 0.486 g), but they do correlate well with data reported for male mice of this strain (0.424 to 0.494 g).¹⁶ The average brain weight in control female Swiss Webster mice was 0.473 ± 0.025 g (0.463 to 0.526 g).

Brain weight and age.

Compared with control Swiss Webster mice, aged FVB/N mice showed no significant decrease (or increase) in brain weight (Figure 1). Brain weights of our FVB/N cohort were not statistically different from those of the Swiss Webster control cohort. The brain:body weight ratio was significantly (P < 2 × 10⁻⁴) higher in female mice (1.62; n = 61) compared with male mice of comparable age (1.30; n = 25). However, brain:body weight ratios are considered less reliable due to the widely variable body condition of study animals, and there was no relationship between the brain:body weight ratio and age (data not shown). The brain weight for the 10 oldest (average, 687 d) female FVB/N mice (0.495 ± 0.064 g; range, 0.453 to 0.510 g) did not differ from that for the 10 youngest (average, 184 d) female FVB/N mice (0.486 ± 0.017 g; range, 0.396 to 0.651 g).

Figure 1.

Brain weight compared with age. In FVB/N mice (age, 154 to 857 d), the average brain weight was 0.501 g for female mice and 0.492 g for male mice. Swiss Webster mice (n = 10; age, 134 d) were used as controls.

Histologic examination.

No significant microscopic lesions were noted in any brain from the current cohort (Figure 2), although an archival case of acute seizure in a female FVB/N mouse (Figure 2 C through F) demonstrated acute neuronal necrosis in the hippocampus and cerebral laminar cortex consistent with previous descriptions of SCS.^6,17 Lesions expected of chronic sublethal seizures include neuronal atrophy in these areas, local gliosis, and possibly ventricular dilation. None of the mice in the study showed neuronal cell death, atrophy, or gliosis (Figure 2 G and H).

Figure 2.

Brain histology. Selected representative images of hippocampus (panels A, C, E, G, and I) and cerebral cortex (panels B, D, F, H, and J) from an unaffected Swiss Webster control mouse (panels A and B), a previous FVB/N mouse with acute space cadet syndrome and neuronal necrosis (arrows; panels C through F), and an FVB/N mouse from the current study cohort (panels G through J). Hematoxylin and eosin stain (A through H); immunohistochemistry for GFAP (I and J). Bar, 50 µm (A through D, G through J), 20 μm (E and F).

Neuronal density.

The neuron profile density was manually quantified in 2 (bilateral) fields each of the cerebral laminar cortex, granule cells of the dentate gyrus, and the CA3 region of the hippocampus of 35 FVB/N (22 female, 13 male) mice and 5 control mice (Figures 3 and 4). The density of granule cells in the dentate gyrus was used as a control, because no neuronal cell loss at this site has been previously reported in FVB/N mice. In FVB/N mice, the neuronal density (mean ± 1 SD) was 155 ± 25 neurons/field (range, 100 to 195 neurons/field) in the cerebrocortex, 109 ± 15 neurons/segment (range, 81 to 134) in the dentate gyrus, and 44 ± 7 neurons/region (range, 31 to 78) in the CA3 region of the hippocampus. For control mice, the neuronal density was 204 ± 37 neurons/field (range, 145 to 237 neurons/field) in the cerebrocortex, 184 ± 33 neurons/segment (range, 135 to 213 neurons/segment) in the dentate gyrus, and 47 ± 12 neurons/region (range, 35 to 64) in the CA3 region. Age was not significantly correlated with neuronal density in any part of the brain, but a comparison of brain weight and neuron density showed a weak correlation (R² = 0.4295) only between decreasing density of cortical neurons and increasing brain weight. However, neuronal densities in the dentate gyrus and cerebral cortex were lower in FVB/N mice, both female and male, than in Swiss Webster control mice (female: dentate gyrus, P = 0.0009; cerebral cortex, P = 0.0149; male: dentate gyrus, P = 0.0016; cerebral cortex, P = 0.0301). These significant differences remained when neuronal densities were compared with age as well as brain weight (Figures 3 and 4).

Figure 3.

Neuronal density compared with brain weight. Each data point represents the average of neuron counts from each half of a bilateral coronal section for an individual mouse; data from 35 representative FVB/N cohort mice and 5 Swiss Webster control mice are shown. The neuronal profile densities for male and female FVB/N mice within the cerebral cortical regions, granule cells of the dentate gyrus (DG), and CA3 of the hippocampus are represented separately. Similar data for Swiss Webster controls are presented also.

Figure 4.

Neuronal density compared with age. Each data point represents the average of neuron counts from each half of a bilateral coronal section for a particular animal; data from 35 representative FVB/N cohort and 5 Swiss-Webster control mice is shown. The neuronal profile densities for male and female FVB/N mice within the cerebral cortical regions, granule cells of the dentate gyrus (DG), and CA3 of the hippocampus are represented separately. Similar data for Swiss Webster control mice are presented also.

Glial cell counts.

Immunohistochemistry for GFAP was performed on 6 groups of mice: FVB/N mice with high and low brain weights; young, intermediate-aged, and geriatric FVB/N mice; and a young, female Swiss Webster control mouse (Figure 2 I and J). Sections were examined visually for focal areas of increased GFAP immunoreactivity, suggesting gliosis; none was found in any animal examined. Positively stained glia within representative areas of the cerebral cortex were counted; there was no correlation between glial profile density and brain weight (R² = 0.3861; data not shown). Quantification of the total percentage of positive GFAP immunoreactivity also failed to identify significant variations between groups (Figure 5). The average strong and total positive immunoreactivities were 23% and 89%, respectively, for the high-brain–weight mice; 16% and 81% for the low-brain–weight mice; 16% and 82% for the youngest mice; 15% and 77% for the intermediate-age mice; and 18% and 90% for the oldest mice. There was no correlation between strong or total GFAP reactivity and brain weight, thus confirming the lack of gliosis according to a technique with greater sensitivity than the human eye (Figure 5). Finally, there was no relationship between glial count in the cerebral cortex and the brain weight or age of the mouse.

Figure 5.

Immunoreactivity for glial fibrillary acidic protein (GFAP) in 14 mice were selected from 6 groups: 2 (one female, one male) FVB/N mice with high brain weights; 4 (3 female, one male) FVB/N mice with low brain weights; 2 (one female, one male) of the youngest (average age, 158 d) FVB/N mice; 2 (one female, one male) midage (average age, 322 d) FVB/N mice; 3 (2 female, one male) of the oldest (average age, 656 d) FVB/N mice; and one young (134 d) female Swiss Webster control mouse. Total GFAP positivity and percentage strong positivity are displayed for each animal.

Ventricular area.

Ventricular sizes in FVB/N mice ranged from 922 to 597,658 μm² (mean ± 1 SD, 153,264 ± 153,260 μm², n = 49) and from 34,625 to 213,855 μm² (100,137 ± 49,831 μm², n = 5) in control mice (P = 0.4476). In addition, as compared with control animals, ventricular size and choroid thickness did not differ in any of the 13 or 14 brain sections of brains, and there was no relationship between ventricular size and brain weight or age (data not shown).

Discussion

This study was performed by collecting FVB/N mice at the end of studies unrelated to neurologic disease. We evaluated a large cohort of aged FVB/N mice to determine whether they had measurable morphologic changes resulting from hypothesized recurrent sublethal seizures. Phenotypic analysis included evaluation of brain weight; histologic assessment of 13 or 14 coronal brain sections; determination of neuronal density in the cerebral cortex, hippocampus, and granule cells of the dentate gyrus; measurement of ventricular size, determination of glial cell density in the cerebral cortex; and quantification of GFAP immunoreactivity. We were unable to detect histologic lesions of neuronal cell death or gliosis attributable to sublethal seizure activity in 86 aged (range, 5.1 to 28.2 mo; average, 349 d) FVB/N mice.

We hypothesized that brain weight would change, either decreasing with age as the cerebral cortex underwent physiologic atrophy, decreasing as neuronal loss occurred due to recurrent sublethal seizure activity, or increasing due to postnecrotic gliosis in FVB/N mice undergoing multiple sublethal seizures.¹⁷ We could not detect any significant change in brain weight with increasing age in our large cohort of aged FVB/N mice. This result may reflect the sensitivity (0.0001 g) of the scale used to weigh the brains. Our results indicate that the brain was heavier in female FVB/N mice than in their male counterparts, consistent with reported averages.¹⁶ However, the average brain weight for our FVB/N female mice is markedly higher than that reported in the Mouse Phenome Database.¹⁶ This apparent discrepancy may be due to the larger age range in our study (154 to 847 d) or the smaller number of animals sampled and reported in the Mouse Phenome Database (n = 5 to 11 for FVB/N mice of both sexes).

Our histologic evaluation included morphologic examination of numerous (more than 700) coronal brain sections, as well as detailed analyses of neuronal and glial cell densities within areas of the brain where we expected to find neuronal cell death or atrophy attributable to seizure activity (hippocampus, cerebrocortical laminae); the dentate gyrus (granule cells) was used as a control, because we expected no neuronal loss in that location. We hypothesized that neuronal density would decrease in older FVB/N mice with potentially greater numbers of seizures and that gliosis would increase as a postnecrotic reaction to neuronal cell death. However, neuronal densities did not significantly change with age or brain weight in the cerebral cortex, dentate gyrus, or hippocampus of our cohort. Indeed, there was a nonsignificant trend toward decreased neuron profile density in the cerebral cortex of mice with heavier brains (R² = 0.4295).

Both male and female FVB/N mice exhibited a reduced density of granule cells in the dentate gyrus and cerebral cortex, as compared with Swiss Webster control mice. Given the similarity in brain weight between FVB/N and Swiss Webster mice, the physiologic or pathologic significance of this difference in the neuronal density in these regions is unclear. The difference in neuronal density of the granule cells of the dentate gyrus is not likely a reflection of seizure activity. Indeed, dentate gyrus granule cells are resistant to seizure-induced cell death, and changes in the dentate gyrus have not been reported in previous studies of SCS.^6,17 In addition, neuronal density in the CA3 region of the hippocampus, an area highly sensitive to seizure-induced cell death, did not differ between FVB/N and control mice. We hypothesize that this difference in neuronal cell density in the dentate gyrus and cerebral cortex reflects an inherent strain variation. The outbred Swiss Webster strain is the parental strain of the inbred FVB/N strain, but Swiss Webster mice do not experience audiogenic or spontaneous seizure activity (that is, SCS). The 2 strains diverged more than 80 y ago.¹ Perhaps the reduction in cerebrocortical and dentate gyrus granular neurons in FVB/N mice provides a morphologic explanation for their inherent susceptibility to cell death during seizures, in addition to their genetic predisposition to seizures. Further studies examining a potential difference in neuronal density between excitotoxic sensitive FVB/N mice and excitotoxic resistant C57BL/6 mice may reveal additional clues to the pathogenesis of seizure-induced cell death in FVB/N mice.

We also hypothesized that neurons lost due to seizure activity would result in fibrous astrocyte gliosis or scarring, leading to an increase in astrocytic projections in these areas. This feature was previously reported in FVB/N mice with SCS.^6,17 We evaluated GFAP positivity in different groups of FVB/N and control mice, but GFAP immunoreactivity did not differ within or between these groups or compared with that in the control mouse.

Our hypothesis that FVB/N mice undergo repeated sublethal seizures was based on 3 lines of evidence: reports of their behavioral features which were attributed to spontaneous sublethal seizures;^2,6,17 their inherent susceptibility to seizures;^3,7,9 and previous reports of neuronal cell death in mice found dead.^6,17,27 We therefore sought to determine whether FVB/N mice naturally experience sublethal seizures with resultant neuronal atrophy or glial scarring in the same brain regions. We were unable to identify histologic lesions of neuronal atrophy or gliosis in 86 FVB/N mice (average age, 349 d). Two explanations of this failure are possible. First, our cohort of FVB/N mice may not experience seizures; indeed, no clinical signs suggestive of seizures had been reported. Second, perhaps sublethal seizure activity does not result in morphologic abnormalities, or perhaps our methods were not sufficiently sensitive to detect any changes. Indeed, some neuroscientists feel that it is extremely difficult to rule out neuronal death after seizure activity.² The histologic lesions of recurrent sublethal seizure activity are variable and poorly described in any species; histologic diagnosis can be challenging in cases with mild disease or lesions. In domestic animals with epilepsy, clinical disease and postmortem histologic findings are poorly correlated, and frequently no morphologic lesions are noted in these animals (although only a small proportion of the brain is examined in most cases).²⁵

We hypothesized that we would detect histologic and immunohistochemical changes associated with sublethal seizure activity in FVB/N mice. We carefully chose an aged experimental cohort to maximize the likelihood that these mice would have experienced at least one seizure in their lifetimes, yet we were unable to detect histologic lesions consistent with sublethal seizure activity. Although we could not examine absolutely every neuron, we examined a greater proportion of each mouse's brain than is generally achievable for humans or most veterinary species, and our coronal sections showed increased standardization due to the use of the brain matrix-cutting tool. Importantly, the acute lesions of SCS have been well described,^6,17 allowing us to focus on specific areas of the brain. We therefore feel that our failure to identify histologic lesions suggestive of chronic seizure activity likely reflects selective breeding to eliminate the SCS phenotype rather than indicates an inability to detect neuronal cell death or gliosis.

Our negative data are consistent with the hypothesis that SCS arose spontaneously in FVB/N mice derived from the NCr stock^17,28 and suggest that this syndrome is now extinct in at least one commercial breeding colony. Hippocampal and cerebrocortical neuronal necrosis were consistently identified in a large study of younger FVB/N mice (2 to 6 mo of age; average age, 174 d) that was published more than 15 y ago.⁶ Our diagnostic services have not diagnosed acute neuronal necrosis consistent with SCS in FVB/N mice in the past 3 y; in contrast, one of our services made that diagnosis 9 times during 2005 through 2012. The lack of recent acute cases, as well as the lack of chronic sublethal lesions in this large cohort of FVB/N mice, suggests that the SCS phenotype has been eliminated from at least one commercial FVB/N vendor.