The C57BL/6 (B6) mouse is one of the most widely used inbred mouse strains in biomedical research, especially in behavioral studies (Crawley, 2003). Furthermore, B6 mice are commonly used as a background strain for genetically modified mice. However, often little attention is paid to the specific substrain of B6 mice used in such studies. The phenotype of a mutant mouse is not only the result of the targeted gene, but it also reflects interactions with background genes, and other unknown mutations in the genetic background (Crawley et al. 1997). There are several B6 substrains (and even sublines) which differ in several behavioral, pharmacological and genetic aspects (Bothe et al. 2004; Crawley et al. 1997; Grottick et al. 2005; Khisti et al. 2006; Ramachandra et al. 2007; Siegmund et al. 2005). In studies on induction of seizures by various convulsive agents, B6 mice were generally reported to exhibit a lower sensitivity than various other inbred mouse strains (Engstrom & Woodbury 1988; Ferraro et al. 1998, 2002; Kosobud and Crabbe 1990; Schauwecker and Steward 1997). However, in all these studies, C57BL/6J (B6J) mice from the Jackson Laboratory (Bar Harbor, ME, USA) were used, so that it is not known whether other B6 substrains differ from B6J in their sensitivity to convulsants. In a study, in which pilocarpine was used to induce status epilepticus (SE), Borges et al. (2003) described that B6J mice responded with a markedly higher mortality than a substrain of B6 mice obtained from Charles River. In view of the popularity of the pilocarpine model in epilepsy research and the fact that this convulsant is increasingly used to induce SE and epilepsy in mice (Dudek et al. 2006), we directly compared the response of three different B6 substrains from two suppliers (Harlan and Charles River) to pilocarpine. In addition to differences between B6 substrains, we also discovered striking differences in sensitivity to pilocarpine in mice of the same B6 substrain but reared in different barrier rooms of the same vendor.
In rodents, the cholinomimetic convulsant pilocarpine is widely used to induce status epilepticus (SE), followed by hippocampal damage and spontaneous recurrent seizures, resembling temporal lobe epilepsy. This model has initially been described in rats, but is increasingly used in mice, including the C57BL/6 (B6) inbred strain. In the present study, we compared the effects of pilocarpine in three B6 substrains (B6JOla, B6NHsd and B6NCrl) that were previously reported to differ in several behavioral and genetic aspects. In B6JOla and B6NHsd, only a small percentage of mice developed SE independently of whether pilocarpine was administered at high bolus doses or with a ramping up dosing protocol, but mortality was high. The reverse was true in B6NCrl, in which a high percentage of mice developed SE, but mortality was much lower compared to the other substrains. However, in subsequent experiments with B6NCrl mice, striking differences in SE induction and mortality were found in sublines of this substrain coming from different barrier rooms of the same vendor. In B6NCrl from Barrier #8, administration of pilocarpine resulted in a high percentage of mice developing SE, but mortality was low, whereas the opposite was found in B6NCrl mice from four other barriers of the same vendor. The analysis of F1 mice from a cross of female Barrier 8 pilocarpine-susceptible mice with resistant male mice from another barrier (#9) revealed that F1 male mice were significantly more sensitive to pilocarpine than the resistant parental male mice whereas female F1 mice were not significantly different from resistant Barrier 9 females. These observations strongly indicate X-chromosome linked genetic variation as the cause of the observed phenotypic alterations. To our knowledge, this is the first report which demonstrates that not only the specific B6 substrain but also sublines derived from the same substrain may markedly differ in their response to convulsants such as pilocarpine. As the described differences have a genetic basis, they offer a unique opportunity to identify the genes and pathways involved and contribute to a better understanding of the underlying molecular mechanisms of seizure susceptibility.
Materials and methods
The following three B6 substrains were used: C57BL/6JOlaHsd (B6JOla) and C57BL/6NHsd (B6NHsd) from Harlan (Harlan-Winkelmann; Borchen, Germany), and C57BL/6NCrl (B6NCrl) from Charles River (Sulzfeld, Germany). All three substrains originally descended from the same B6 breeding stock and were maintained at the Jackson Laboratory (Bar Harbor), but subsequent history differed, leading to three distinct substrains (Festing 1996). (1) B6JOla: sent in 1974 from the Jackson Laboratory to Laboratory Animals Centre, Carshalton; in 1983 to OLAC (now Harlan UK); in 1997 to Harlan Nederland. (2) B6NHsd: sent in 1974 from the Jackson Laboratory to the National Institutes of Health (NIH), Bethesda, Maryland; Harlan Sprague Dawley, Inc. derived the strain from this breeding nucleus. (3) B6NCrl: sent in 1951 from the Jackson Laboratory to the NIH; in 1974 to Charles River Laboratories. In this study, B6NCrl mice were obtained from two different barriers (#8 and #9) from Charles River in Sulzfeld (Germany). In the following, we refer to the mice from the two different barriers as different ‘sublines’. According to the information from Charles River, both barriers are located in the same building without any obvious indication of environmental differences. After we found striking differences in sensitivity of Barrier 8 and 9 mice to pilocarpine SE (described in the Results section), we obtained B6NCrl mice from three additional barriers (#4, #7 and #11) of the same vendor for comparison.
All substrains were purchased from either Harlan or Charles River at an age of about 5–6 weeks. Female mice were used for most experiments of the present study to allow comparison with previous studies for other mouse strains (Gröticke et al. 2007; Löscher et al. 1986; Löscher et al. 1991; Löscher and Lehmann 1996; Potschka and Löscher 1999), and to ease housing in groups for the long period needed to perform experiments in models of epilepsy. However, it is known that differences in estrous cycle stage can influence seizure susceptibility, so that additional experiments were performed in male B6 mice.
After arrival, animals were housed in groups under controlled conditions (temperature: 21 ± 0.5°C; humidity: 55 ± 5%), and a 12-h light–dark cycle with lights on at 6.00 a.m. and food and water ad libitum. The animals were allowed to adapt to the new environment for 1–3 weeks before starting the experiments. All substrains were tested in parallel at the same time of the year to avoid seasonal effects. Furthermore, on each experimental day, mice of different substrains were tested together at the same time of the day to avoid that substrain differences were secondary due to circadian effects. All the experiments were performed between 8:00 a.m. and 1:30 p.m. to minimize variation because of circadian rhythms (see more details below). All possible steps were taken to avoid animals suffering at each stage of the experiment. The procedures used in this study had the approval of the Institutional Animal Care and Use Committee and were carried out in accordance with the European Council Directive of November 24th, 1986 (86/609/EEC).
Comparison of different dosing protocols for pilocarpine
Two different dosing protocols were used in all B6 substrains. (1) i.p. injection of single, high doses (300–400 mg/kg) of pilocarpine; and (2) repeated low-dose treatment by i.p. application of 50 or 100 mg/kg pilocarpine every 20 min until onset of SE. The latter protocol was based on previous experiments of our group with pilocarpine in NMRI (Naval Medical Research Institute; obtained from Harlan) mice (Gröticke et al. 2007). In order to avoid peripheral cholinergic effects, methylscopolamine (1 mg/kg) was administered 30 min before the application of pilocarpine. SE was defined as continuous limbic seizure activity (see Results for more detailed description), which typically lasted for several hours if not terminated earlier by diazepam (10 mg/kg i.p.). For the present study, all mice that developed SE received diazepam (10 mg/kg i.p.) after 90 min of SE.
In order to minimize circadian effects on seizure threshold, mice of the three substrains were usually brought to the laboratory at 8:00 a.m., and were allowed to adapt to the laboratory conditions for 30 min. Then, methylscopolamine was injected at 8:30 a.m. in mice of the three substrains tested on this day, followed 30 min later by injection of pilocarpine. The experiment was usually completed by 12:00 a.m. except in mice that did not develop SE, in which observation was continued until 1:30 p.m. (see Results). Seizure severity was rated by the Racine scale (Racine 1972).
Maximal electroshock seizure threshold
After we found differences in sensitivity to pilocarpine in B6NCrl mice from two different barriers of Charles River (see Results), the seizure susceptibility of these mice was also compared in the maximal (tonic hindlimb extension) electroshock seizure (MES) threshold test as described previously (Löscher & Lehmann 1996). The MES threshold was determined via corneal copper electrodes (covered with leather and wetted with saline) by means of a stimulator (Witt GmbH, Berlin, Germany) that delivered a constant current (adjustable from 1 to 200 mA regardless of impedance of the test object; self-adjusting stimulus voltage max. 7000 V) with sinusoidal pulses (50/s) for 0.2 s. During stimulus application by the corneal electrodes, the animal was restrained only by hand and released at the moment of stimulation in order to permit observation of the seizure throughout its entire course. The stimulus intensity was varied by an up-and-down method in which the current was lowered or raised by 0.06-log intervals if the preceding animal did or did not exert hind limb extension, respectively. The starting current was 20 mA. The data thus generated in groups of mice were used to calculate the threshold current inducing hind limb extension in 50% of the mice (CC50 with confidence limits for 95% probability) by the method of Kimball et al. (1957). Each mouse was used for only one threshold determination. Eight mice were used per group.
The open field is a popular test for evaluating exploratory and anxiety-like behavior (Prut & Belzung 2003). The procedure consists of subjecting an animal to an unknown environment from which escape is prevented by surrounding walls. In such a situation, rodents spontaneously prefer the secure and darker periphery of the apparatus over the bright and open central parts of the open field. In the present study, the test was performed in a white, square open field (60 × 60 × 40 cm). The animals were placed individually in the center of the open field. Behavior was observed for 10 min. Before each trial, the field was cleaned thoroughly with 0.1% acetic acid solution and was dried with paper towels. The open field was divided into three zones: center (20 × 20 cm; light intensity 197 lux), internal ring (10 cm), and outer ring (10 cm; 160 lux). For each mouse, the total distance moved, the velocity, and the time spent in each zone were measured by a computerized tracking system (EthoVision; Noldus; Wageningen, Netherlands).
On the basis of the results of our experiments in different B6 strains (see Results section), showing that B6NCrl mice from Barrier 8 were exceptionally sensitive to SE induction by pilocarpine, we assumed that a mutation must have arisen in these mice. We therefore cross-bred Barrier 8 and Barrier 9 mice to produce F1 hybrids, which were then tested with pilocarpine. For this purpose, 9 B6NCrl females from Barrier 8 were mated with 9 B6NCrl males from Barrier 9, and the offspring was tested at an age of 6–8 weeks with pilocarpine, using the same protocol as in the parenteral generation.
Statistical analyses to determine if substrains and sublines differed significantly in the incidence of SE or mortality after pilocarpine were performed by Fisher's exact test (2 × 2). In case of comparisons between more than two groups, the chi-square test was used to analyze whether the observed frequencies differed significantly among groups, followed by pair-wise comparisons with Fisher's exact test (using α correction) to determine which groups were different. Group comparisons of dose to first seizure, latency to first seizure, dose to SE, latency to SE and the average number of seizures before SE (or without SE) were performed by Student's t-test. In case of more than two groups, group comparisons were performed by one-way analysis of variance (anova) followed by post hoc testing with the Bonferroni test. All tests were used two-tailed and a P < 0.05 was considered significant.
Definition of SE
In all B6 substrains tested in this study, the same criteria for defining SE were used.
In general, SE is defined by continuous or intermittent seizures without full recovery of consciousness between seizures (Lowenstein 1999). In the B6 substrains, SE onset was typically characterized by a generalized convulsive (stage 4 or 5) seizure that switched over in continuous limbic seizure activity characterized by head nodding and gnawing. The continuous limbic seizure activity, during which mice were typically immobile, was interrupted by clonic forelimb seizures, generalized tonic-clonic seizures with the loss of righting reflexes and, occasionally, running and jumping seizures. After such convulsive seizures, mice resumed the continuous limbic seizure activity. Previous electroencephalogram (EEG) recordings with cortical electrodes in NMRI mice (Gröticke et al. 2007) and in few B6 mice showed that this continuous limbic seizure activity was associated with paroxysmal EEG alterations. Such continuous seizure activity lasted for hours if mice did not die before. In the present experiments, SE was generally interrupted after 90 min by diazepam. Only few mice (<10%) with SE evaluated in this study died during SE before interruption by diazepam. Mice that did not develop SE did either not exhibit any seizures in response to pilocarpine or only individual convulsive (score 4 or 5) seizures that did not progress into continuous limbic seizure activity (for more details see below). If such individual convulsive seizures were associated with tonic hindlimb extension, this often led to death of the animals by respiratory failure.
Comparison of three B6 substrains with bolus injection of pilocarpine
In a first experiment, single i.p. doses of pilocarpine were tested in the three B6 substrains (Table 1). The range of doses (300–400 mg/kg) covered the dose range previously reported for B6 mice (Borges et al. 2003; Chen et al. 2005; Peng et al. 2004; Shibley & Smith 2002). Within a few minutes after pilocarpine injection, immobility, tremor, straub tail and occasional clonic seizures occurred in all B6 substrains. However, the incidence of SE with continuous seizure activity markedly differed among B6 substrains. In B6JOla and B6NHsd, only few mice developed SE at doses up to 400 mg/kg, but mortality was high, so that the doses were not increased further. In contrast, 70–80% of B6NCrl mice developed SE after 300–325 mg/kg of pilocarpine and most mice survived. On the basis of this clearly distinguishable substrain difference obtained with high single doses of pilocarpine, we used a ramping up dosing design to characterize the substrain difference in more detail.
|Substrain of C57BL/6||Type of application||Dose (mg/kg)||Number of mice tested||Mice with SE||Mice surviving SE||Total mortality|
|C57BL/6JOlaHsd||Single dose i.p.||300||10||0||—||1 (10%)|
|C57BL/6JOlaHsd||Single dose i.p.||325||5||0||—||0|
|C57BL/6JOlaHsd||Single dose i.p.||350||10||0||—||4 (40%)|
|C57BL/6JOlaHsd||Single dose i.p.||400||5||2 (40%)||1 (20%)||4 (80%)|
|C57BL/6JOlaHsd||Repeated injections (50 mg/kg every 20 min) till SE||400–550||5||0||—||5 (100%)|
|C57BL/6JOlaHsd||Repeated injections (100 mg/kg every 20 min) till SE||300–1000||19||3 (16%)||0||16 (84%)|
|C57BL/6NHsd||Single dose i.p.||300||5||1 (20%)||1 (20%)||1 (20%)|
|C57BL/6NHsd||Single dose i.p.||325||5||1 (20%)||1 (20%)||0|
|C57BL/6NHsd||Single dose i.p.||350||6||2 (33%)||2 (33%)||3 (50%)|
|C57BL/6NHsd||Repeated injections (50 mg/kg every 20 min) till SE||500–550||5||0||—||1 (20%)|
|C57BL/6NHsd||Repeated injections (100 mg/kg every 20 min) till SE||300–1000||11||1 (9%)||0||10 (91%)|
|C57BL/6NCrl (#8)||Single dose i.p.||300||5||4 (80%)*||4 (80%)*||0|
|C57BL/6NCrl (#8)||Single dose i.p.||325||10||7 (70%)*||6 (60%)*||3 (30%)|
|C57BL/6NCrl (#8)||Repeated injections (50 mg/kg every 20 min) till SE||300–550||5||3 (60%)||2 (40%)||1* (20%)|
|C57BL/6NCrl (#8)||Repeated injections (100 mg/kg every 20 min) till SE||300–1000||10||9 (90%)*,+||8 (80%)*,+||1 (10%)*,+|
|C57BL/6NCrl (#9)||Repeated injections (100 mg/kg every 20 min) till SE||300–1400||23||7 (30%)#||6 (26%)#||17 (74%)#|
Comparison of three B6 substrains with a ramping up dosing protocol with pilocarpine
This protocol is based on the previous experience with repeated i.p. injections of low doses of pilocarpine in NMRI mice (Gröticke et al. 2007). In this protocol, 100 mg/kg are injected every 20 min until the development of SE, which allows a more individual dosing of pilocarpine compared to the injection of fixed doses in each animal per group. Similar to our recent experiments in NMRI mice, in the present study in B6 mice it usually took three injections to a first seizure (typically a score 4 or 5 seizure), but for the development of SE it was important to continue the injections of pilocarpine after the occurrence of this first seizure, until SE starts. Typically, 1–2 additional injections were needed after the first seizure to induce SE. If SE was not induced after 7–8 injections in an individual animal, usually additional injections failed to induce SE, but mice died in individual convulsive (tonic-clonic) seizures, due to respiratory arrest. Thus, the maximum number of repeated pilocarpine injections was restricted to about 8–14. In our previous experiments in NMRI mice, this protocol was associated with a high percentage of mice developing SE but low mortality (Gröticke et al. 2007). However, in B6JOla and B6NHsd mice, only a few animals developed SE by this method, but mortality was high (Table 1). We therefore modified the protocol and injected only 50 mg/kg pilocarpine every 20 min, but SE could not be induced in the B6JOla and B6NHsd substrains (Table 1). In contrast, 90% of B6NCrl mice developed SE with the ramping up protocol when 100 mg/kg pilocarpine was injected every 20 min until onset of SE. In B6NCrl mice, the dosing protocol was less effective with 50 mg/kg, so that all subsequent experiments were performed with the 100 mg/kg version of the ramping up method. By using this protocol, the following sequence of behavioral symptoms was observed: immobility (after about 1–6 min following the first injection), tremor (1–11 min), straub tail (2–29 min), first seizure (7–225 min; see Fig. 1d) and SE (17–164 min; see Fig. 1f). In case of mortality, the mice usually died after 8–190 min following the first injection of pilocarpine. As described above, SE was characterized by continuous limbic seizure activity, which was occasionally interrupted by clonic forelimb seizures, generalized clonic-tonic seizures or running and jumping seizures.
In Fig. 1a, data from all experiments with pilocarpine in the three B6 substrains are summarized, including both the results of bolus and ramping up experiments. As illustrated in this figure, striking substrain differences in the percentage of mice developing SE, surviving SE and total mortality were observed. In both B6JOla and B6NHsd, only very few animals developed SE, independent of the protocol used, but mortality was high, whereas the reverse was true in the B6NCrl substrain, resulting in highly significant differences between B6NCrl mice and the two other B6 substrains. In order to evaluate whether B6JOla and B6NHsd were generally insensitive to pilocarpine or whether the low incidence of SE in these substrains was related to a lack of progression from first isolated seizures to SE, we examined the data in more detail. Only 28 of 54 B6JOla mice treated with pilocarpine exhibited one or more single seizures after pilocarpine, compared to 26/32 B6NHsd mice (P = 0.0210) and 29/30 B6NCrl mice (P < 0.001), respectively (Fig. 1b). The dose of pilocarpine to first seizure was comparable in B6JOla and B6NHsd mice, but significantly lower in B6NCrl mice (Fig. 1c). The latency to first seizure (usually a score 4 or 5 seizure) was not significantly different in the three substrains (Fig. 1d). Similarly, no significant difference was determined for dose to SE (Fig. 1e) or latency to SE (Fig. 1f) in the three B6 substrains. In mice that developed SE, significant inter-substrain differences in the number of isolated seizures before SE were determined. Average number of seizures before onset of SE was 1.0 (range 0–2) in B6JOla, 2.2 (1–5) in B6NHsd, but 3.7 (2–5) in B6NCrl [analysis of data by anova: F2,30 = 18.55;P < 0.0001]. Post hoc analysis by the Bonferroni test indicated that B6NCrl mice differed significantly from the two other substrains (B6NCrl vs. B6JOla, t = 5.688,P < 0.001; B6NCrl vs. B6NHsd, t = 3.156,P < 0.01 [df = 30]). In mice, which exhibited isolated seizures but no SE, the average number of seizures was 1.2 (range 1–3) in B6JOla, 2.4 (1–4) in B6NHsd, and 2.3 (1–4) in B6NCrl [anova: F2,48 = 10.65; P = 0.0001]. Post hoc analysis by the Bonferroni test indicated significant differences between B6JOla and the two other strains (B6JOla vs. B6NHsd, t = 4.436,P < 0.001; B6JOla vs. B6NCrl, t = 2.556,P < 0.05 [df = 48]).
As a next step, we wanted to confirm the observations made for the B6NCrl substrain and purchased another batch of mice from the same vendor (Charles River). But unexpectedly, only 30% of the animals developed SE and mortality was higher (>70%), as shown in Table 1. Only then we realized that the B6NCrl mice used in this experiment were derived from another barrier (#9) than the B6NCrl mice previously used (#8). We therefore decided to directly compare additional groups of age-matched B6NCrl from the two barriers of the same vendor for their sensitivity to pilocarpine.
Comparison of B6NCrl mice from two different barriers of the same vendor
For the pilocarpine experiment and all further experiments, age-matched B6NCrl mice from Barrier 8 (B6NCrl#8) and Barrier 9 (B6NCrl#9) were compared. The mice from the two barriers were obtained with one shipment and were tested together on the same days to avoid any seasonal or circadian effects. As shown in Fig. 2a, the direct comparison of mice from the two barriers confirmed the previous observations shown in Table 1. Compared to Barrier 8 mice, significantly less animals from Barrier 9 developed SE, whereas mortality was higher. For further characterization of these differences between B6NCrl mice from the two barriers, data from all pilocarpine experiments with the ramp design were evaluated together, resulting in group sizes of 27 (Barrier 8) or 40 (Barrier 9) animals, respectively. As shown in Fig. 2b, the difference in SE incidence and mortality between the two sublines were highly significant (P = 0.0001 and 0.0002, respectively). Twenty-two (82%) of the 27 B6NCrl#8 mice developed SE compared to 13 (33%) of the 40 B6NCrl#9 mice. In contrast to these differences in SE induction by pilocarpine, isolated seizures were induced by pilocarpine in all (except one) mice of both barriers. However, the individual dose of pilocarpine to the first seizure was significantly higher in B6NCrl#9 than B6NCrl#8 mice (Fig. 2c), although this difference was not as striking as the difference in SE induction (Fig. 2a). There were also significant trends for higher latency to first seizure (Fig. 2d), individual dose to onset of SE (Fig. 2e) and latency to onset of SE (Fig. 2f) in B6NCrl#9 mice. When the number of isolated seizures before onset of SE was compared in B6NCrl mice from the two barriers, the average seizure number was 3.3 (range 1–5) in B6NCrl#8 vs. 5.5 (3–9) in B6NCrl#9 mice, which was significantly different (P < 0.001). In contrast, the number of seizures in mice without SE did not differ (3 vs. 3.5 on average in the two groups, respectively).
When comparing these data of Barrier 8 and Barrier 9 mice in Fig. 2, the B6NCrl mice obtained from Barrier 9 exhibited a greater variability with respect to several parameters: dose to first seizure, latency to first seizure, dose to SE or latency to SE. The variability was substantially greater than in B6NCrl mice obtained from Barrier 8 or compared to the other B6 strains.
In view of the lower sensitivity of B6NCrl#9 to SE induction by pilocarpine, we next tested whether this difference to B6NCrl#8 mice extended to another seizure stimulus. Therefore, we determined the MES threshold in the two sublines of animals. CC50 was 14.8 (11.8–18.5) mA in B6NCrl#9 mice compared to 14.7 (12.7–17.1) mA in B6NCrl#8 mice, which was not significantly different. The majority of the mice from both barriers died during or after the tonic-clonic seizures induced by electroshock.
Finally, we evaluated whether the two sublines of B6NCrl mice differed behaviorally, using the open field test. No significant differences between mice from the two barriers were seen in any parameter evaluated in this model, including locomotor activity (distance moved in the field) or time spent in the aversive center of the open field (not illustrated).
Comparison of female and male B6NCrl mice
All experiments described above were performed in female B6 mice. Because it is known that differences in estrous cycle stage can influence seizure susceptibility, additional experiments were performed in age-matched female and male B6NCrl mice from Barrier 9, which were obtained with one shipment and were tested together to avoid any seasonal or circadian effects. As shown in Table 2, male B6NCrl of Barrier 9 mice tended to be even more resistant to SE induction by pilocarpine than female B6NCrl mice from this barrier, although the differences between genders were not statistically significant. Average seizure number before onset of SE (4.0 ± 0.53 vs. 5.0 ± 0.58) or number of seizures in mice without SE (5.2 ± 0.83 vs. 4.06 ± 0.36) did not differ significantly between female and male mice. Furthermore, dose to first seizure, dose to SE and latency to SE did not differ significantly between genders (not illustrated), but latency to first seizure was significantly lower in male (55.6 ± 1.6 min) vs. female mice (73.5 ± 5.8 min; P = 0.0077).
|Substrain of C57BL/6||Type of application||Dose (mg/kg)||Number of mice tested||Mice with SE||Mice surviving SE||Total mortality|
|C57BL/6NCrl (#9)||Repeated injections (100 mg/kg every 20 min) till SE||300–900||20||9 (45%)||8 (40%)||12 (60%)|
|C57BL/6NCrl (#9)||Repeated injections (100 mg/kg every 20 min) till SE||300–1000||20||3 (15%)||3 (15%)||18 (90%)|
|C57BL/6NCrl (#4)||Repeated injections (100 mg/kg every 20 min) till SE||300–1200||20||6 (30%)||6 (30%)||13 (65%)|
|C57BL/6NCrl (#7)||Repeated injections (100 mg/kg every 20 min) till SE||300–1000||20||6 (30%)||6 (30%)||14 (70%)|
|C57BL/6NCrl (#11)||Repeated injections (100 mg/kg every 20 min) till SE||400–1100||20||7 (35%)||6 (30%)||15 (75%)|
Comparison of B6NCrl mice from additional barriers of the same vendor
For evaluating whether the high sensitivity of B6NCrl mice of Barrier 8 was an exception within the B6NCrl substrain, we tested B6NCrl mice from three additional barriers (#4, 7 and 11) of the same vendor. As shown in Table 2, B6NCrl mice from these barriers were as resistant to SE induction by pilocarpine as mice from Barrier 9. This observation strongly indicated that a mutation might have arisen in Barrier 8 animals causing the high sensitivity to SE induction. This hypothesis was thus further assessed in studies with F1 hybrids.
Analysis of F1 hybrid mice obtained from crosses of Barrier 8 and Barrier 9 B6NCrl mice
Female Barrier 8 mice were crossed to male Barrier 9 B6NCrl mice to produce F1 hybrids that were subsequently injected with pilocarpine. In total, 28 female and 32 male F1 hybrids were studied. As shown in Fig. 3a, female F1 hybrids did not significantly differ in sensitivity to SE induction from Barrier 9 resistant female mice. However, male F1 hybrids were significantly more sensitive to SE induction (and less sensitive to mortality) by pilocarpine than the male Barrier 9 parental mice (Fig. 3b). In addition, significant differences were determined between male and female F1 mice. Dose to first seizure and latency to first seizure were significantly lower in male F1 compared to female F1 hybrids (Fig. 3c,d). Furthermore, dose to SE and latency to SE were significantly lower in F1 males compared to F1 females (Fig. 3e,f). Average seizure number before onset of SE (4.14 ± 0.25 vs. 3.71 ± 0.14) or number of seizures in mice without SE (3.71 ± 0.3 vs. 3.2 ± 0.22) did not differ significantly between female and male F1 hybrids.
The three major findings of this study are that (1) a specific substrain of B6 mice, B6NCrl, is much more sensitive to SE induction by pilocarpine than two other B6 substrains, B6JOla and B6NHsd; (2) B6NCrl mice from different barrier rooms of the same supplier (Charles River) strikingly differ in their sensitivity to SE induction by pilocarpine; and (3) the high sensitivity of Barrier 8 B6NCrl mice is heritable, indicating that genetic mutation(s) represent the basis for these subline differences. To our knowledge, the only previous indication for substrain differences in sensitivity to SE induction by pilocarpine came from experiments of Borges et al. (2003) who reported that B6 mice from Charles River (B6NCrl) were more sensitive than B6 mice from Jackson Laboratories (B6J). Using single doses of pilocarpine in the range of 247–335 mg/kg i.p. in 99 B6J mice, only 12 mice exhibited SE and survived; total mortality was 64%, which is similar to the 56% mortality rate observed in the present experiments in B6JOla mice from Harlan. In contrast to B6J mice from the Jackson Laboratory, a larger percentage (58%) of B6NCrl mice from Charles River developed SE and survived (15 of 26 mice total) in the experiments reported by Borges et al. (2003). Neuro-pathological changes in the hippocampus examined in mice surviving SE were similar in the B6J and B6NCrl substrains and were characterized by the loss of pyramidal cells in CA1 and CA3 and the loss of neurons in the dentate hilus (Borges et al. 2003). Furthermore, spontaneous seizures were determined after SE in both B6 substrains (Borges et al. 2003).
The high susceptibility of B6NCrl mice to SE induction by pilocarpine found in the study of Borges et al. (2003), and the present study is substantiated by the study by Chen et al. (2005) in which >80% of B6NCrl mice developed SE after i.p. doses ranging from 272 to 340 mg/kg and 81% of the mice survived. Following SE, spontaneous recurrent seizures and hippocampal damage were observed in B6NCrl mice by Chen et al. (2005). Thus, overall, the B6NCrl substrain seems to be better suited for the pilocarpine model of temporal lobe epilepsy than other B6 substrains tested in this respect.
Several previous studies have found behavioral and genetic differences between B6J and B6N substrains of B6 mice. Stiedl et al. (1999) reported that B6JOla and B6NCrl differed in their course of extinction of conditioned fear, with B6NCrl showing a slower decline of fear in response to a conditioned stimulus than B6JOla. In B6JOla from Harlan, Specht & Schoepfer (2001) found a spontaneous chromosomal deletion of the gene that encodes the presynaptic protein alpha-synuclein, which has been implicated in synaptic transmission and the etiology of a range of neurodegenerative disorders (Windisch et al. 2007), whereas the alpha-synuclein gene is present in B6NCrl mice and B6JCrl mice (Siegmund et al. 2005). This prompted Siegmund et al. (2005) to study whether this inter-substrain difference in alpha-synuclein expression is involved in the different extinction of conditioned fear in B6 substrains, but no relationship was found. Grottick et al. (2005) reported dramatic differences in prepulse inhibition of the startle response (PPI) in B6J and B6NHsd mice and sought to identify the molecular mechanisms underlying these differences in sensorimotor gating by a microarray-based approach. In the two substrains, differences in glutamatergic and GABAergic signaling were found that may explain the different PPI but may also be involved in the different response to pilocarpine determined in the experiments of Borges et al. (2003) and in the present study. In a study in which the relationship of 102 mouse strains, including B6 substrains, was assessed by using a panel of 1638 single nucleotide polymorphisms (SNPs), substrains derived from the B6 colony at the NIH differed in only five SNPs from the B6J strain at Jackson Laboratory (Petkov et al. 2004), suggesting that the genetic drift between B6J and B6J-derived substrains is minor.
Determination of genotypes for 342 microsatellite markers in B6J (obtained from Jackson Laboratory) and B6N (obtained from Taconic Farms, Germantown, NY) identified only 12 microsatellite differences between the two B6 substrains and no substrain difference was found in three behavioral tests, i.e. rotarod, open field and fear conditioning (Bothe et al. 2004). However, significant differences between the B6J and B6NCrl substrains were found in ethanol drinking and dependence (Khisti et al. 2006). The difference in ethanol preference between the two substrains was inversely correlated with ethanol-induced dopamine release in the ventral striatum (Ramachandra et al. 2007). B6J mice exhibited significantly greater ethanol preference and less ethanol-stimulated dopamine release compared to B6NCrl mice (Ramachandra et al. 2007).
To our knowledge, the experiments presented here describe for the first time that behaviors studied in B6 mice are not only affected by the B6 substrain used but that different responses may also occur in the same substrain obtained from different barriers of the same vendor. As demonstrated by the experiment on SE induction by pilocarpine, the differences in SE induction and mortality determined between B6NCrl mice from Barriers 8 and 9 of the same vendor (Charles River) were as pronounced as the differences determined between B6J and B6N substrains from different vendors. In other words, if we would have compared only B6NCrl mice from Barrier 9 with the B6JOla and B6NHsd substrains, our conclusion would have been that all three substrains are resistant to SE induction by pilocarpine, whereas the B6NCrl mice from Barrier 8 were much more sensitive to SE induction by pilocarpine than any of the other substrains.
There are three possibilities to explain the differences found between B6NCrl mice from the two barriers of the same vendor. First, a spontaneous genetic alteration that might have arisen in one colony or the other may explain the difference. Second, the difference could be a consequence of an environmental effect initiated at the vendor, or, third, the difference could be a consequence of a complex gene-environment interaction that may involve multiple factors both genetic and from the environment. Therefore, we performed several experiments to directly address the possibility of a genetic variation. First, B6NCrl mice from three additional barriers of the same vendor (Charles River) were tested. They all exhibited similar resistance to SE induction by pilocarpine as B6NCrl mice from Barrier 9, demonstrating that Barrier 8 mice differed from the other four barriers of this vendor. This strongly indicates that the high sensitivity to SE induction of Barrier 8 mice was due to a genetic variation that has arisen only in this subline of B6NCrl mice. To further test this hypothesis, we generated F1 mice from matings of female pilocarpine-sensitive Barrier 8 mice with male resistant Barrier 9 mice. In the F1 generation, a significant difference was observed between male F1 hybrid and the resistant Barrier 9 parental subline. On the other hand, female F1 mice were not significantly different compared to Barrier 9 resistant female mice. These results strongly indicate that the phenotypic differences between Barrier 8 and Barrier 9 mice are caused by a recessive genetic variation which occurred in the Barrier 8 subpopulation, and that the allele which causes pilocarpine-sensitivity resides on the X-chromosome. It should be noted that for the described phenotype assay, mice had to be tested in groups. Thus, to further prove genetic inheritance and to map the underlying genetic alterations, it will be necessary to develop phenotype assays which can distinguish between individual susceptible and resistant carriers. As the breeder has discontinued the Barrier 8 breeding, we have started to establish an independent breeding colony of Barrier 8 mice at our department. This should allow us to test the above hypotheses further in the future.
New mutations in sublines of inbred mice have been described recently for epilepsy, for example, within the B6J substrain of B6 mice (Yang et al. 2003) and in the C3H/HeJ strain (Beyer et al. 2008). Large breeding colonies at vendors are historically organized in the form of subcolonies. Here, the possibility exists that a spontaneous mutation occurs in a founder animal and becomes fixed in the resulting subcolony. Although this may be a rare event, once discovered, it offers a unique possibility to identify new susceptibility genes.
Compared to a variety of other inbred and outbred mouse strains, B6 mice are generally considered seizure resistant in that higher doses of various convulsants are needed to induce seizures in this strain (Engstrom & Woodbury 1988; Ferraro et al. 1997, 1998, 1999, 2002, 2004, 2007; Kosobud & Crabbe 1990). By mapping murine loci for seizure response, Ferraro and colleagues reported that the difference in seizure susceptibility between seizure-resistant B6 and seizure-sensitive DBA/2J mice is a polygenetic phenomenon with loci of significant effects on chromosomes 1, 5, 7 and 15 (Ferraro et al. 1997, 1999, 2004, 2007). All these studies were performed in B6J mice from the Jackson Laboratory. The present and previous studies (Borges et al. 2003; Chen et al. 2005) suggest that B6NCrl mice are more sensitive to convulsants than B6J mice. In this respect, it is also interesting to note that the MES threshold of B6NCrl mice was low in the present study and not different from that of seizure-sensitive NMRI mice (unpublished data).
It is important to note that the main difference between B6 substrains and sublines determined in the present study was not related to the dose of pilocarpine inducing convulsions, but to the consequences of convulsive doses of pilocarpine, i.e. induction of SE and mortality. In experiments on mapping genome loci for seizure response to kainate in B6J and DBA/2 J mice, two types of genetic influence upon the seizure phenotype were identified: one that increases susceptibility to kainate-seizures and another that increases severity of the seizure syndrome once it has been initiated (Ferraro et al. 1997). Schauwecker & Steward (1997) reported that kainate, 30 mg/kgs.c., induced comparable seizures in B6 mice (obtained from Hilltop Labs, Philadelphia) and the FVB/n and 129/SvEMS inbred strains, but that, in contrast to these other mouse strains, B6 mice did not exhibit excitotoxic cell death in the hippocampus after kainate-induced seizures. These results indicated that B6 mice carry genes that convey protection from glutamate-induced excitotoxicity (Schauwecker & Steward, 1997). However, this protection does not extend to pilocarpine-induced excitotoxicity, because pilocarpine-induced SE results in hippocampal damage and spontaneous seizures in B6 mice (Borges et al. 2003; Chen et al. 2005; Peng et al. 2004; Shibley & Smith 2002).
In addition to substrain differences in SE induction by pilocarpine, we also found significant differences in seizure induction. Thus, B6JOla mice were significantly more resistant to seizure induction than B6NHsd and B6NCrl#8 mice. In contrast, B6NCrl from the different barriers did not differ in seizure induction but only in SE induction by pilocarpine. Thus, these B6 substrains and sublines can be used both to explore mechanisms of seizure susceptibility and mechanisms of SE induction.
In conclusion, the present study demonstrates that substrains of B6 mice from different vendors markedly differ in their response to the convulsant pilocarpine. In studies using B6 mice, the exact nomenclature of the B6 substrain is often not indicated, which makes it difficult to interpret the results of such studies, considering the various behavioral and genetic differences between B6 substrains that have been reported (see above). Thus, although B6 substrains are closely related mouse lines with limited genetic differences, such differences may critically affect the phenotype in behavioral studies and interact with targeted mutations that are based on a B6 background. This situation is getting even more complex by the present finding that sublines of the same B6 substrain may differ between different barriers of the same vendor. On the other hand, B6 substrain or subline differences as obtained in the present study with pilocarpine offer the possibility of identifying genetic loci of critical relevance to seizure phenotypes in mice and may further our understanding of basic processes that are involved in the evolution of single seizures to a self-sustaining SE.
We thank Dr Thomas N. Ferraro (Center for Neurobiology and Behavior, Department of Psychiatry, University of Pennsylvania, Philadelphia) for helpful discussions during this study and the preparation of the manuscript, and Dr Martin Beyerbach (Department of Biometry, Epidemiology and Information Processing, University of Veterinary Medicine, Hannover, Germany) for statistical advice. Christine J. Müller acknowledges the financial support obtained from the Center for Systems Neuroscience (Hannover, Germany).