| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Animal and Poultry Sciences Department, Virginia Polytechnic Institute and State University, Blacksburg 24061-0306
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
Key Words: Anesthesia • Festuca • Mice • Resistance • Susceptibility • Toxicity
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
; Cross, 1997). Management practices designed to alleviate its impact have not been universally effective. Selection within (Lipsey et al., 1992) or between cattle breeds (Nutting et al., 1992) for response to endophyte-infected (E+) fescue is a potential genetic solution. Before effective strategies can be designed, however, biology of host resistance to the toxins must be examined.
Hohenboken and Blodgett (1997) selected mouse lines for resistance (R) or susceptibility (S) to the effect on postweaning growth of E+ fescue seed in the diet. After eight generations of selection, the lines differed in the impact of an E+ diet on growth, and R mice had higher hepatic Phase II enzyme activities than S mice (Hohenboken and Blodgett, 1997). Reproduction in the S line was more severely depressed by an E+ diet than was reproduction in the R line (Wagner et al., 2000), and R mice had higher resistance than S mice to the mycotoxin, sporidesmin (Hohenboken et al., 2000).
Our primary objective was to determine whether the R and S lines differ in hepatic Phase I detoxification function as assessed by sleep time following sodium pentobarbital anesthesia (ST). In Exp. I, we sought to answer the following questions: 1) Do R and S line mice differ in ST? 2) Are between-line differences in ST modulated by an E+ or a 50% endophyte-free fescue seed, 50% control (E-) diet? 3) Does prior induction of hepatic Phase I detoxification enzyme systems by ingestion of phenobarbital affect line differences in ST? 4) Are line and diet effects on sleep time retained after mice are changed to a standard diet? Questions in Exp. II were as follows: 1) Do R and S mice differ in ST when managed under a different dietary regimen than in Exp. I? 2) Is ST of individual mice correlated with the impact of the E+ diet on their postweaning growth (the trait used to assess toxicosis response during selection)?
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
Mice used in these experiments were from the R and S lines divergently selected for response to fescue toxicosis as described by Hohenboken and Blodgett (1997). Foundation animals were from the outbred International Cancer Research Institute (ICR) strain (Harlan Sprague Dawley, Inc., Indianapolis, IN). During eight generations of bidirectional selection, mice were evaluated for toxicosis response by the reduction in 2-wk gain when fed an E+ diet compared to gain during the previous 2 wk when their diet was free of the toxin. The selection criterion was an index combining information from the animal’s own phenotype for gain reduction and that of its littermates. In the R line, selection was for nil or minimal reduction in gain, whereas in the S line, selection was for larger gain reduction.
Throughout the current experiments, mice were housed in a single room of the university Laboratory Animal Vivarium and were managed in compliance with institutional laboratory animal policies and regulations. Temperature was maintained at an average of 24°C and fluorescent lights were on from 0700 to 1900 each day. Mice were housed in 15- x 21- x 29-cm transparent plastic cages bedded with a paper fiber product and equipped with continuous-flow nipple watering devices.
Experiment I
A total of 240 male mice, 120 per line, with approximately equal representation from 31 S litters and 33 R litters, were used in the experiment. Following weaning on a single day when litters were between 21 and 26 d of age, mice were randomly assigned four per cage with the provision that cage-mates were from the same line but were not siblings.
Immediately following weaning, the following dietary treatments were randomly assigned to cages of animals in each line: 1) control, finely ground laboratory rodent food (Teklad 70-01, Harlan Sprague Dawley, Madison, WI); 2) E+, a ground and thoroughly mixed diet composed by weight of half rodent food (as above) and half endophyte-infected ‘Kentucky-31’ fescue seed. 3) E+P, the same as Diet 2 but supplemented with 1,000 ppm of phenobarbital. 4) E-, a ground and thoroughly mixed diet composed by weight of half rodent food (as above) and half endophyte-free ‘Forager’ fescue seed. 5) E-P, the same as Diet 4 but supplemented with 1,000 ppm of phenobarbital.
Fescue seed from the same endophyte-infected Kentucky-31 lot was used to formulate both the E+ and the E+P diets; likewise, the E- and the E-P diets both included seed from the same endophyte-free ‘Forager’ source. Neither E+ nor E- seed was analyzed for ergovaline content. Phenobarbital was added to Diets 3 and 5 to induce hepatic Phase I enzymes of the cytochrome P450 class (Conney, 1967; Omiecinski et al., 1985). There were six cages (24 mice) per line by diet subclass. Mice were provided ad libitum access to diets, and fresh food was provided three times per week. It was not possible to quantify feed consumption or refusals in this or the second experiment, to be described.
After 4 wk on the experimental diets, mice were administered sodium pentobarbital anesthesia, half the cages on each line by diet combination (120 mice) were randomly chosen on one morning and the other half on the next. Following the protocol of Lovell (1986 a,b), each mouse was anesthetized by i.p. injection of sodium pentobarbital at 40 mg/kg of weight. The anesthetic was dissolved in physiological saline and administered as 0.1 mL/10 g of BW. Clock time was recorded at injection and again when the mouse could no longer accomplish the righting reflex, defined as the ability of the mouse to right itself two times within 30 s. The mouse was then placed on its back in a bedded standard cage and clock time was recorded when the righting response was regained. Sleep latency was the time in minutes between injection and loss of the righting reflex, and ST was the elapsed time between losing and regaining the righting reflex.
Following the second day of testing, all mice were switched to a diet of pelleted rodent food (Teklad 70-01, Harlan Sprague Dawley). A second ST test from sodium pentobarbital anesthesia took place 1 wk later on one-fourth of the population (60 mice), consisting of one randomly chosen mouse from each cage. Following that trial, those males were killed by CO2 asphyxiation. One week later, an additional 60 mice, one randomly chosen from each cage, were administered the ST test and then killed. This continued for two more weeks, decreasing the number of mice per cage by one each week until each mouse had been tested twice, with intervals between first and second testing of 1, 2, 3, or 4 wk.
Experiment II
A total of 280 individuals, 70 per line by sex subclass, with approximately equal representation from 35 S litters and 34 R litters, were used in the experiment. Mice were weaned on a single day when litters were between 21 and 26 d of age and randomly assigned four individuals per cage, with the provisions that cage-mates were of the same line and sex but were not siblings.
All mice were fed the E- diet from Exp. I for the first 2 wk following weaning, the E+ diet from Exp. I for the following 2 wk, and pelleted laboratory rodent food (Teklad 70-01, Harlan Sprague Dawley) for the next 2 wk. Seed lots used to formulate the E+ and E- diets were different from those used in Exp. I. Again, seed was not analyzed for ergovaline content. Using the same protocol for ST testing as was used in Exp. I, mice were then administered sodium pentobarbital anesthesia (a randomly chosen one-third of the cages of each line by sex combination on three consecutive mornings).
Mice were weighed individually at weaning and weekly for the duration of the experiment. Toxicosis response of each individual was quantified as the proportional reduction in weight gain during the 2 wk that E+ fescue seed was present in the diet, compared to weight gain during the previous 2 wk when E- fescue seed composed the same diet proportion. (Throughout the divergent selection that produced the R and S lines, this variable was used to assess individual mouse response to endophyte infected fescue challenge.)
Statistical Analysis
Chi-squared analysis was used to test whether the proportions of mice in Exp. I that did not succumb to anesthesia differed significantly among diets or lines.
Sleep time has a positive skewed distribution, with fewer observations above than below the mean. Accordingly, ST observations were subjected to logarithmic transformation prior to analyses of variance to more nearly normalize the distribution and to reduce heterogeneity of variance. Our conclusions on statistical significance of effects and interactions are based on analysis of the log-transformed variable. For ease of interpretation, however, numerical results are presented and discussed in minutes, the original unit of measurement.
The mathematical model for ANOVA of the first sleep latency and first log ST for all mice included fixed effects for line, diet and the line by diet interaction. The model was as follows:
 |
where SL1 and log ST1 = sleep latency and log sleep time 1 (when the entire population underwent anesthesia), µ = the mean, L = line, and D = dietary treatment.
Two questions were relevant from data collected on the second ST measurement of the mice. First, did line and diet effects on sleep time persist after the mice had been switched to a pelleted rodent food diet for various lengths of time? To answer this question, the data were subjected to ANOVA with the following mathematical model:
 |
where ST2 = sleep time the second time each mouse underwent anesthesia, T = the time in weeks between the first and second ST test, and other factors are as defined above.
Secondly, what was the correlation between a mouse’s first and second ST observation, and did this relationship change as time between the first and second measurement increased? To answer this question, data for ST1 and ST2 were analyzed simultaneously according to the following model:
 |
and the residual correlation was computed between ST1 and ST2. This analysis was done separately for groups whose second ST test followed the first by 1, 2, 3 and 4 wk. Because these correlations were of similar magnitude, the relationship between ST1 and ST2 also was quantified as the linear regression coefficient of ST2 on ST1 from the following model:
 |
where ß is the regression coefficient of ST2 on ST1.
In Exp. II, we sought to determine whether line differences in ST following sodium pentobarbital anesthesia, as had been found in Exp. I, would be expressed under different experimental conditions. The model for ANOVA of log ST included fixed effects for line, sex, and the line x sex interaction: log ST = µ + L + S + (L x S) + error. Experiment II also was designed to see if an individual animal’s ST was correlated with the impact of the E+ diet on its postweaning growth. To answer this question, we computed the residual correlation between ST and the percentage reduction in gain associated with E+ feeding when gain reduction and ST were analyzed simultaneously with the above model.
|
Results and Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
In Exp. I, 10 of 120 R mice and 9 of 120 S mice were not completely anesthetized by the standard dose of sodium pentobarbital. Although not statistically significant by chi-squared analysis, a substantial proportion of nonresponders (12 of 19) had been on diets supplemented with phenobarbital. In Exp. II, 16 of 280 mice were not completely anesthetized, eight because of injection or observational errors. Remaining nonresponders were distributed similarly between lines and sexes.
Sleep Latency
Average sleep latency was 5.6 min (SD = 4.9) in Exp. I, and 4.7 min (SD = 1.8) in Exp. II. Lines did not differ significantly for this trait in either experiment. Sleep latency was not affected by diet or the line x diet interaction in Exp. I, or by sex or the line x sex interaction in Exp. II.
Experiment I
First and second ST averaged 23.6 and 25.3 min with CV of 66% and 63%, respectively. As shown in Table 1, S mice slept nearly 5 min longer on average than R mice (P < 0.01) during their first ST test. The line x diet interaction was not statistically significant. That is, the difference in average ST between lines was relatively consistent across diets (Figure 1). Dietary treatments exerted an important effect on ST1 (Table 1, P < 0.01, and Figure 1). Sleep time of mice fed E+ did not differ from that of mice fed E- or Control, but the E- group had longer ST1 than the Control group. Mice on diets supplemented with phenobarbital had shorter ST than all other groups.
|
|
, P < 0.01). Although effects of the previous diets on ST2 were not statistically significant, groups supplemented with phenobarbital (E-P and E+P) had slightly longer ST2 than corresponding nonsupplemented (E- and E+) groups (Table 1). In contrast, the phenobarbital-supplemented groups (E-P and E+P) had much shorter ST1 than mice on the E- and E+ diets. Time interval between ST1 and ST2 did not affect ST2 significantly. Average ST2 for time intervals between tests of 1, 2, 3 and 4 wk were 22.7, 26.9, 26.6, and 24.4 min, respectively. The average standard error of these least squares means was 2.1 min. Residual correlations between ST1 and ST2 equaled 0.43, 0.41, 0.31, and 0.29 for groups whose first and second ST observations were separated by 1, 2, 3 and 4 wk, respectively, for an average value of 0.36. Thus, ST is moderately repeatable across these time intervals. When the earlier ST was fit as a covariate in the analysis of the later ST, the regression of ST2 on ST1 equaled 0.44 ± 0.08 min/min.
Experiment II
Results are not tabulated but are presented herein. Sleep time averaged 24.3 min with a CV of 47%. Susceptible mice slept an average of 26.3 ± 1.0 min and R mice an average of 22.3 ± 1.0 min. This 4 min difference (P < 0.01) was similar to the 4.9 min difference between R and S mice for ST1 in Exp. I. Sleep time was not significantly affected by sex or the line x sex interaction. The residual correlation between an individual animal’s ST and the depression in its growth rate caused by endophyte-infected fescue seed in the diet was 0.04.
Discussion
Our primary objective was to determine whether selection for response to fescue toxicosis had caused differentiation between lines in ST following sodium pentobarbital anesthesia. It did; R line mice slept 19, 27, and 15% fewer minutes than S line mice in three comparisons. Compared to the S line, R line mice may have had higher constitutive concentrations of hepatic drug metabolizing enzymes, higher synthesis of hepatic drug metabolizing enzymes following induction and/or a higher biotransformation rate (Parke et al., 1991; Nebert and Dieter, 2000; Parkinson, 2001), hastening clearance of sodium pentobarbital from the liver and subsequent recovery from anesthesia.
Contrary to this inference, Hohenboken and Blodgett (1997) reported that R and S line mice did not differ in activities of hepatic cytochrome P450 and b5 enzymes. In their experiment, activities of these enzymes also did not differ between mice on the E+ or an E- diet, nor were they significantly affected by the interaction between diet and genetic line. Assessment methods in their experiment did not allow differentiation among activities of individual enzymes within the P450 and b5 families. Differential induction or expression between lines of some enzymes within P450 or b5 families could therefore have escaped detection.
Smith et al. (1980) reported a negative correlation between breed group means for pentobarbitone ST and liver injury following exposure to sporidesmin toxin in sheep. Merinos had higher resistance to sporidesmin intoxication and shorter ST than Romneys, Border Leicesters, or Romney x Border Leicester crossbreds. Earlier, Fairclough et al. (1978) reported that microsomal fractions from livers of Merino ewes metabolized sporidesmin more efficiently than microsomal fractions from livers of Romney rams and lambs, which may contribute to the breed difference in resistance to the toxin.
Our second objective was to determine whether line differences in ST were modulated by endophyte-infected fescue seed and/or phenobarbital supplementation of the diet. They were not, as evidenced by the lack of significance or importance of the line x diet interaction effect on ST1. Sleep times of mice from both lines on both the E+ and E- diets were decreased by phenobarbital, a potent inducer of cytochrome P450 enzymes in the liver (Conney, 1967; Omiecinski et al., 1985).
Our third objective was to determine whether ST of individual mice was correlated with the reduction in growth rate caused by endophyte-infected fescue seed in the diet (the criterion used to assess susceptibility to fescue toxicosis during selection). We hypothesized that mice with longer ST (and by inference, less effective drug metabolism) within line x diet subclasses would have larger gain reductions than mice with longer ST. This expectation was not met. The correlation between ST and gain reduction equaled 0.04, there was virtually no relationship between ST of individual animals and the impact of the E+ diet on their postweaning growth, and neither variable could accurately be predicted from the other. Similarly, Smith et al. (1980) were not able to show a within-breed-group correlation between pentobarbitone ST and sporidesmin resistance in sheep, even though group averages were negatively correlated for the two traits.
This lack of association between gain reduction and ST may be due to error variance in both indicators of susceptibility. That is, individual weight gain reduction and ST may both reflect toxicosis resistance with poor precision. Selection for divergence in toxicosis response (Hohenboken and Blodgett, 1997; Wagner et al., 2000) may have been effective despite this limitation, because breeding values were estimated using an index that combined information from each individual and all of its full sibs (Falconer, 1981). Such methods enhance effectiveness of selection, particularly for lowly heritable traits.
|
Implications |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
1 Correspondence—phone: 540-552-2817; fax: 540-231-3010; E-mail; whohenbo{at}vt.edu.
Received for publication January 27, 2003. Accepted for publication June 2, 2003.
|
Literature Cited |
|---|
|
TopAbstractIntroductionMaterials and MethodsResults and DiscussionImplicationsLiterature Cited |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
