The impact of murine strain and sex on postnatal development after maternal dietary restriction during pregnancy
- 1Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario M5G 1X5, CanadaDepartments of 2Physiology and 3Obstetrics & Gynaecology, University of Toronto, Toronto, Ontario M5S 1A8, Canada4School of Women's and Infants' Health, The University of Western Australia, Perth, WA 6008, Australia
- Corresponding author S. Lye: 850–600 University Avenue, Toronto, Ontario M5G 1X5, Canada. Email lye{at}mshri.on.ca
Abstract
The objective of this study was to characterize offspring responses to maternal dietary restriction (DR) in two phylogenetically distant strains of mice: A/J and C57BL/6J (B6). Pregnant mice were fed 100% or 70% of ad libitum between 6.5 and 17.5 days (d) gestation. Offspring were fed 100% ad libitum postweaning. All comparisons were made to strain and sex matched controls. Male DR-B6 offspring initially grew slower than controls; however, by 77d and 182d they were significantly heavier (P < 0.05). Further, they had an increase percentage fat mass (+70%, P < 0.01) by 182d and were glucose intolerant at both 80d (P < 0.001) and 186d (P < 0.05). In contrast, weight, %Fat mass and glucose tolerance in DR-A/J males during postnatal life were not different from controls. Female DR-B6 mice showed catch-up growth during the first 77d of life; however, their weight, %Fat mass and glucose tolerance were not different from controls at 80d and 186d. Although female DR-A/J were heavier than controls at 182d (P < 0.05), their %Fat mass and glucose tolerance were not different from controls at 182d and 186d. The observed strain and sex differences offer a unique opportunity to begin to define gene–environment interactions that contribute to developmental origins of health and disease.
An inverse relationship between suboptimal intrauterine and postnatal environments and the development of adult diseases, including the metabolic syndrome (type II diabetes, obesity, insulin resistance, hypertension, stroke and dyslipidaemia) has been reported in a number of different human populations (Barker & Osmond, 1988; Barker, 1998; Osmond & Barker, 2000) and in numerous animal studies in multiple species (Bertram & Hanson, 2001; Matthews et al. 2002; Gluckman & Hanson, 2004). This is particularly important as data suggest that the environment of mother, baby and child is a key contributor to diseases and conditions that account for approximately one-third of the global burden of disease in both the developed and developing world.
Compelling epidemiological and laboratory evidence indicates that the interaction between the fetal environment and genome can modify the risk of postnatal health and disease as well as the individual's ability to cope with postnatal environmental challenges (Gluckman & Hanson, 2004). This suggestion is based on the observation that not all individuals exposed to an adverse antenatal and/or postnatal environment develop the metabolic syndrome (Eriksson et al. 2002, 2003a; Pihlajamaki et al. 2004; Yliharsila et al. 2004). These data have led to the hypothesis that gene–environment interactions underlie the developmental origins of health and disease (DOHaD).
Further evidence suggesting that gene–environment interactions underlie DOHaD comes from studies evaluating polymorphisms in key genes in the metabolic pathways regulating energy metabolism. Polymorphisms in the peroxisome proliferator-activated receptor (PPAR)-γ2 gene modify the relationship between size at birth and adult diseases including insulin sensitivity and metabolism (Eriksson et al. 2002; Laakso, 2004), hypertension (Yliharsila et al. 2004), obesity (Pihlajamaki et al. 2004) and dyslipidaemia (Eriksson et al. 2003b). The well-known association between small body size at birth and insulin resistance and/or hypertension was seen only in individuals with the high-risk Pro12Pro genotype and not the Pro12Ala variant in the PPAR-γ2 isoform-specific exon B which is known to reduce the transcriptional activity of PPAR-γ2 (Eriksson et al. 2003a). Sex has also been shown to modify the relationship between adverse antenatal and/or postnatal environments and the development of adult disease including cardiovascular disease (Eriksson et al. 1999; Forsen et al. 2004) and insulin sensitivity (Flanagan et al. 2000), providing further complexity to these interactions.
The laboratory mouse has been used extensively to analyse complex genetic traits: indeed, these traits can be dissected in well-defined inbred strains of mice where environmental effects can be controlled (Singer et al. 2004). Comparison of mouse strains using chromosomal single nucleotide polymorphisms reveals that C57BL/6J (B6) and A/J strains are separated by a major bifurcation in a phylogenetic hierarchical tree. This analysis reveals that B6 mice are separated, genetically, into a group of strains that is distinct from the A/J strain and the majority of other inbred strains studied (Cervino et al. 2005). A/J and B6 strains also differ in many important physiological characteristics, including homocysteine (Ernest et al. 2005) and anxiety levels (Singer et al. 2005). Recent studies have documented that exposure of B6 (the most widely used inbred mouse strain) to a high-fat diet during adolescence and adulthood induces features of the metabolic syndrome; however, the same high-fat diet had little or no effect on the A/J strain (Surwit et al. 1995; Singer et al. 2004). The observed obesity and diabetes induced by the high-fat diet in B6 mice do not occur spontaneously (as seen in some single gene mutated mice (Black et al. 1998) making it a good model of human metabolic disease.
Recently we have shown that maternal dietary restriction (DR) during pregnancy in A/J and B6 mice results in strain differences in the fetal response to the adverse environment, even though both strains experienced equal magnitude of intrauterine growth restriction (Knight et al. 2007). DR induced significant reductions in fetal kidney weight and embryo: placenta ratio, as well as increases in placental weight, fetal brain: liver ratio and maternal glucocorticoid levels in B6 compared to A/J (Knight et al. 2007). However, to date the impact of DR on the adult phenotype of these offspring has not been characterized. Therefore the aim of this study was to characterize strain and sex differences in offspring of A/J and B6 mice exposed to maternal DR during pregnancy. We hypothesize that maternal DR during pregnancy in these two strains of mice will result in strain and sex differences in postnatal growth patterns, obesity and glucose tolerance as a result of genetic differences between the strains.
Methods
Animals
All experiments were approved by the Samuel Lunenfeld Research Institute Animal Care Committee in accordance with the guidelines of the Canadian Council for Animal Care. Inbred A/J and B6 mice strains (Jackson, Bar Harbor, ME, USA) were fed sterile 1 g Dustless Precision Pellets® (Bio-Serv, Frenchtown, NJ, USA) and sterile water ad libitum under standard environmental conditions as previously described (Knight et al. 2007). Primiparous female A/J and B6 mice were mated with an experienced male of the same strain. Primiparous females were used to eliminate potentially confounding effects of physiological changes occurring during previous pregnancies. Day 0.5 of gestation was designated as the day a vaginal sperm plug was observed and following visualization of a plug, female mice were housed individually and fed ad libitum.
Maternal DR was imposed from day 6.5 to day 17.5 after which the mice are fed ad libitum. The design of the 30% maternal DR has previously been described in detail (Knight et al. 2007). Briefly, in order to maintain an absolute 30% global reduction, the restriction diets used in this experimental model represented a strain specific 30% reduction from the observed normal pregnancy pattern of food requirements. The normal dietary intake was based on previously obtained dietary intake calculated for each day of pregnancy for each strain (Knight et al. 2007). During pregnancy mice were housed individually from day 6.5 until delivery to ensure that an equal dietary restriction was applied to each individual mouse. Post-partum mothers were fed ad libitum and offspring remained with their mothers until weaning at 21.5 days of age. At weaning, pups were separated into groups of 2–4 same-sex siblings and were fed ad libitum.
Growth and body composition
All offspring were weighed at 2, 21, 28, 35, 42, 49, 56, 63, 70, 77 and 182 days postpartum. Body composition, as determined by dual-energy X-ray absorptiometry, was measured in isoflurane anaesthetized mice at 70 and 189 days of age using the PIXImus small animal densitometer (software version 1.42.006.010; Wipro GE, Madison, WI, USA). Using this method, we measured bone mass density (BMD, g cm−2), bone mineral content (BMC, g), bone area (cm2), total body mass (g), lean mass (g), fat mass (g), and percentage fat mass at a resolution of 0.18 × 0.18 mm pixels. The PIXImus was calibrated routinely with a phantom utilizing known values and a quality assurance test was performed daily. The variability in precision for measuring body composition, by this method, is less than 1% (Nagy & Clair, 2000).
Glucose tolerance
Glucose tolerance testing was conducted on conscious mice restrained in 50 ml Falcon tubes with added air holes at the ages of 80 and 186 days. Animals were starved for 15 h prior to the testing period but were allowed water ad libitum. Mice underwent intraperitoneal injection with 1.5 mg glucose per gram body weight using a 75 mg ml−1 glucose stock solution. Blood was collected to measure glucose using the saphenous vein puncturing method (Hem et al. 1998) prior to the injection of glucose and 30 and 60 min post-injection. Blood glucose was measured using a One Touch Basic Glucometer (LifeScan Canada Ltd, Burnaby, BC, Canada).
Statistical analysis
Multivariable regression analyses were used to evaluate experimental outcomes using repeated measures techniques where appropriate. Variable selection in multivariable modelling was based on stepwise procedures (forwards and backwards selection) including the assessment of interaction terms to arrive at the most parsimonious model. Variables considered in the multivariable models included strain, maternal diet, sex, mother, maternal weight at term, number of offspring in the litter, proportion of males in each litter, number of dead offspring in the litter, and biologically relevant interactions. The final model utilized to assess growth trajectories and body composition included strain, maternal diet and sex modelled as fixed effects and individual offspring as random effects repeated over time. For glucose tolerance, area under the curve (AUC) was first calculated for each individual offspring test and then multivariable regression modelling was utilized with strain, maternal diet and sex modelled as fixed effects and individual offspring as random effects repeated over time. The analyses comparing weight, body composition and glucose tolerance in offspring were based on 85 pups from 25 litters who achieved 185 days of age. SAS statistical software (v. 9.1; SAS Institute Inc., Cary, NC, USA) was used for data analysis and P values < 0.05 were considered statistically significant.
Results
There was no significant effect of strain (P = 0.873) or diet (P = 0.766) on the number of pups per litter nursed from 24 h until weaning (pups per litter: A/J 3.8 ± 0.4; DR-A/J 5.1 ± 0.8; B6 4.8 ± 0.8; DR-B6 3.3 ± 0.9), nor was there any interaction between these effects (P = 0.279). The proportion of male: female pups surviving within each litter was not influenced by maternal diet (P = 0.613); however, the difference between strains approached significance (P = 0.058) with relatively more male fetuses surviving in A/J compared to B6 mice (male: females ratio: A/J 1.7 ± 0.4; DR-A/J 1.2 ± 0.5; B6 0.7 ± 0.4; DR-B6 0.8 ± 0.4). DR did result in an increase in loss of litters due to cannibalism compared to controls fed ad libitum (P = 0.048); however, there was no impact of strain on these effects.
Postnatal growth patterns in offspring of mothers exposed to maternal DR during pregnancy were influenced by strain, maternal diet and sex (Fig. 1). In A/J males (Fig. 1A) and females (Fig. 1B), maternal DR did not significantly alter the postnatal growth trajectory in the first 35 days of life when compared to strain and sex matched controls. After 35 days postnatal age, DR-A/J males continued to have the same growth trajectory as strain and sex matched controls whereas DR-A/J females were significantly heavier than controls at 63, 70 and 182 days (all P = 0.050). In contrast to A/J mice, both B6 male (Fig. 1C) and B6 female (Fig. 1D) DR offspring had a delay in postnatal growth with both sexes weighing significantly less than strain and sex matched controls during the first third of their postnatal life (day 21 and day 28 P = 0.050) but by 63 days postnatal age they had similar weights to controls. Beyond day 63, female DR-B6 offspring had the same growth trajectory as controls, whereas male DR-B6 offspring were significantly heavier than controls at 70 and 182 days (P = 0.050).
To further evaluate the observed strain and sex differences observed in postnatal bodyweight of offspring, body composition was examined using PIXImus small animal densitometery (Table 1). Significant strain and sex differences in percentage fat mass (after controlling for maternal DR during pregnancy) were seen at both 77 and 182 days postnatal age (all P < 0.001): higher percentage fat mass was observed in A/J compared to B6 and in female compared to male offspring at both postnatal ages. A significant interaction between strain, sex and maternal DR was seen at 182 days postnatal age (P = 0.042). Male DR-B6 offspring at 182 days postnatal age had a 70% increase in percentage fat mass when compared to strain and sex matched controls (DR-B6-DR male 28.7 ± 2.2%, B6 male 16.8 ± 1.2%, P = 0.002, Fig. 2). This response to maternal DR during pregnancy at 182 day of age was not seen in B6 female mice nor was it seen in AJ mice (Fig. 2).
Offspring lean mass at 77 and 182 days postnatal age (Table 1) was significantly influenced by sex (77 days P = 0.030; 182 days P < 0.001) and maternal DR during pregnancy (77 days, P = 0.002; 182 days, P < 0.001), whereas there was no significant effect of strain at either postnatal age (77 days, P = 0.838; 182 day, P = 0.208). The interaction between strain, sex and maternal DR was significant at 77 (P = 0.006) and 182 (P < 0.001) days postnatal age with complex and divergent responses seen between the sexes and strains seen at both gestational ages (Fig. 2). At 77 days, male DR-A/J offspring had a small increase in lean mass (+2%) when compared to strain and sex matched controls whereas female DR-A/J offspring had a larger increase in lean mass (+17%). This contrasts with B6 mice at the same postnatal age where the effect of sex was reversed: a much larger increase in lean mass was seen in the DR male offspring (+11%) than female offspring (+6%). At 182 days the magnitude of the effect of DR on lean mass was also influenced by sex and strain; again larger effects were seen in female A/J mice and male B6 mice. DR-A/J male offspring had a reduction in lean mass (−6%) compared to strain and sex matched controls whereas female DR-A/J had a large reduction in lean mass (−16%). Conversely, DR-B6 males had a large reduction in lean mass (−17%) with female DR-B6 only having a small reduction in lean mass (−8%).
As a further measure of body composition, BMD, BMC, and bone area were examined (Table 1). A strain (P = 0.004), maternal dietary treatment (P = 0.072) and sex (P = 0.031) effect was seen in BMD at 182 but not at 77 days of age. Further, there was a significant interaction between strain, diet and sex at 182 days of age (P = 0.005). The only group of offspring to have a significant change in BMD was the female DR-A/J offspring, which had a 9% increase in BMD at 182 days when compared to strain and sex matched controls (P = 0.015).
Examination of BMC at 77 days postnatal age showed a significant strain (P < 0.001) and sex (P = 0.05) effect with higher BMC in B6 than A/J mice and higher BMC in male than female offspring. There was no effect of maternal DR on BMC at 77 days postnatal age. Conversely, by 182 days, strain (P < 0.001), sex (P = 0.073) and maternal DR (P = 0.001) were significantly associated with BMC. All offspring from mothers exposed to DR had a reduction in BMC at 182 days; however, the magnitude of this effect was significantly greater in male B6 (− 17%) than female B6 or any of the A/J offspring (P < 0.001).
Similar to BMC, bone area in offspring, at 77 days of age was associated with mouse strain (P < 0.001) and sex (P = 0.056) whilst there was no association with maternal diet (P = 0.658). At 182 days postnatal age, strain (P < 0.001), sex (P = 0.087) and maternal DR were associated with BMC. Increases in bone area were seen in male DR-B6 (P < 0.001), female DR-B6 (P = 0.068) and female DR-A/J (P = 0.045) but not male DR-A/J when compared to strain and sex matched controls.
In response to the glucose challenge, there was a significant strain and sex effect on AUC at both 80 (P < 0.001; P = 0.083, respectively) and 186 days (P = 0.044; P = 0.005): A/J offspring had a higher AUC than B6 offspring at both postnatal ages. There was no effect of maternal DR on AUC at 80 or 186 days (all, P = 0.150). However, a significant interaction between strain, sex and maternal DR was identified at both 80 days (P = 0.007) and 186 days (P = 0.098). When compared to strain and sex matched controls (Fig. 3A and B), at 80 days postnatal age increases in AUC were seen in male DR-B6 (+51%, P = 0.001) and female DR-B6 (+ 23%, P = 0.089) whilst no change was seen in male DR-A/J and a decrease was seen in female DR-A/J (−25%, P = 0.059). At 186 days, male DR-B6 mice had significantly increased AUC (+64%, P = 0.02) compared to strain and sex matched controls (Fig. 3C) whereas no change in AUC was seen in B6 female offspring or A/J offspring of either sex exposed to the same environment during pregnancy (Fig. 3C and D).
Discussion
In this study we have shown that murine strain and sex influence postnatal growth patterns, body composition and glucose tolerance of offspring from mothers exposed to DR during pregnancy. Male B6 offspring appear to have the greatest adverse effects from maternal DR compared to female B6 and male and female A/J mice. Although B6 female and A/J female offspring of mothers exposed to DR during pregnancy were heavier than strain and sex matched controls, these mice did not have any significant changes in percentage fat or glucose tolerance. Overall the A/J mice were relatively resistant to the postnatal effects of maternal DR during pregnancy with no significant changes in percentage fat or glucose tolerance seen during the 186 days of postnatal life. The strain differences in postnatal response to maternal DR are consistent with our previous data suggesting that B6 mice were more susceptible than A/J mice to the adverse effects of maternal DR during pregnancy (Knight et al. 2007). The mechanisms responsible for the strain differences in response to DR during pregnancy are complex and not fully determined. Exposure to elevated glucocorticoid levels is one mechanism by which a series of stressors including, under-nutrition, unbalanced nutrition, utero-placental insufficiency and excess glucocorticoid exposure could impact on fetal programming (Gluckman & Hanson, 2004). Glucocorticoid exposure has been documented as a mechanism influencing fetal programming. Our previous observation that during DR, B6 fetuses are exposed to higher glucocorticoid levels that A/J raises the possibility that this might influence the spectrum of metabolic phenotype we observed in this study.
In the present study, male B6 offspring from mothers exposed to 30% DR during pregnancy demonstrated catch up growth by 63 days of postnatal age after which they maintained a significantly heavier weight than strain matched controls. DR B6 offspring also exhibited impaired glucose tolerance at 80 and 186 days postnatal age. These results contrast with those of a similar study in B6 mice (Yura et al. 2005) in which a later and shorter nutritional restriction (30% maternal DR between day 10.5 and day 18.5 of gestation) resulted in earlier catch up growth (by 21 days) but no obesity (to 133 days postnatal age). These offspring did, however, exhibit impaired glucose tolerance at 119 days postnatal age. Obesity in these mice was only seen when offspring were exposed to late DR followed by exposure to a high fat diet after weaning. These data suggest that the timing of maternal DR during pregnancy may modulate the phenotype induced later in life. Other studies have shown that the timing of the dietary challenge impacts phenotype in sheep (Symonds et al. 2004; Maclaughlin et al. 2005), rats (Lesage et al. 2001; Gluckman & Hanson, 2004) and humans (Ravelli et al. 1999; Roseboom et al. 2001). A similar discordance between obesity and glucose tolerance was associated with the timing of exposure of pregnant women to DR during the Dutch winter famine (Ravelli et al. 1998; Ravelli et al. 1999). Thus, DR during mid to late pregnancy (but not early pregnancy) led to impaired glucose tolerance in the offspring in later life (Ravelli et al. 1998). In contrast, DR during mid to late pregnancy had no effect on obesity while DR in early pregnancy induced obesity in later life but only in female offspring (Ravelli et al. 1999). Taken together these data suggest that important temporal and sex influences on DR induced fetal programming.
The genetic (strain) differences we report here on B6/A/J offspring responses to maternal DR during pregnancy are somewhat analogous to the outcomes when adults from these strains were exposed to a high-fat diet. In those studies B6 mice developed symptoms of the metabolic syndrome to a much greater extent than A/J in response to an adult high fat diet (Surwit et al. 1995; Singer et al. 2004). Subsequently, chromosome substitution in A/J and B6 mice generated a substrain panel which was used to identify 150 quantitative trait loci (on 17 of the 22 chromosomes) affecting serum levels of sterols and amino acids, diet-induced obesity and anxiety (metrics of the metabolic syndrome) in postnatal over-nutrition studies (Takahashi et al. 1999; Singer et al. 2004). It will be of interest to determine whether similar loci might underlie the strain differences in the ontogeny of the metabolic syndrome in offspring of mothers exposed to DR during pregnancy.
We found marked sex differences in response to maternal DR during pregnancy. Similar sex differences in response to an adverse in utero environment have previously been reported in rats (Desai et al. 1997; Sugden & Holness, 2002; Desai et al. 2005), guinea pigs (Liu et al. 2001; Kapoor et al. 2006) and also in human cohort studies (Eriksson et al. 1999; Flanagan et al. 2000; Forsen et al. 2004). Males who were lighter at birth have lower insulin sensitivity but higher insulin secretion and glucose effectiveness in adulthood whereas there was no correlation between size at birth and insulin sensitivity in women (Forsen et al. 2004). In rats, maternal protein restriction during pregnancy (Desai et al. 1997) or offspring protein restriction during early postnatal life (Sugden & Holness, 2002) resulted in insulin resistance and hyperinsulinaemia in male but not female adult rats. Importantly, we also found interactions between strain and sex that might indicate genetic influences in response to maternal DR.
Maternal nutrition and intrauterine growth have also been reported to influence bone mass and skeletal disorders, in particular the relationship between osteoporotic fractures later in life, birthweight, weight in infancy and adult bone mass (Javaid & Cooper, 2002). Interestingly, there is also a link between lowered BMC, osteoporosis and type II diabetes in humans (Isaia et al. 1987). Our observation of similar findings (altered BMC, obesity and glucose intolerance) in mice that were strain and sex dependent, again points to an underlying gene–environment interaction.
We are aware that DR during pregnancy can impact the subsequent lactational performance of the mother. In this study we did not experimentally manipulate litter sizes or cross foster pups after birth since in pilot studies manipulation of the litter dramatically increased the number of pups rejected and/or consumed by the dam. Our outcomes are thus likely to be influenced by both pregnancy and early postnatal nutrient deprivation. Nonetheless, our multivariable statistical analyses demonstrated that litter size or male: female litter ratio had no impact on outcomes. Future studies, using cross-fostering, might address the relative impact of pre- or postnatal nutrient restriction.
In conclusion, we have demonstrated strain and sex differences in B6 and A/J mice in response to an adverse intrauterine environment, modelled through a 30% maternal dietary restriction between 6.5 and 18.5 days of gestation. Male B6 mice appear to be the most adversely affected, demonstrating catch-up growth, increased total weight, obesity, impaired glucose tolerance and decreased lean mass, bone mineral content and area. These alterations were absent in male A/J offspring and minimal in both B6 and A/J female offspring subjected to maternal DR. The observed strain and sex differences demonstrate a susceptibility to DOHaD in the male B6 genome and/or a relative resistance in the others. The underlying cause of this genetic susceptibility remains to be determined, but we speculate that the exposure of DR B6 fetuses to higher levels of glucocorticoids during pregnancy contributes to the development of the altered metabolic phenotype we observed. If so, polymorphisms within genes that regulate the hypothalamic–pituitary–adrenal axis may underlie this phenomenon. The findings of our study will provide the basis for future exploration to evaluate these gene–environment interactions underlying the developmental origins of obesity and impaired glucose tolerance.
Acknowledgements
Funded by Canadian Institutes of Health Research #MOP81238.
Footnotes
-
(Received 13 December 2006; accepted after revision 5 March 2007; first published online 8 March 2007)
- 2007 The Author. Journal compilation © 2007 The Physiological Society
References
Figure 1. Growth trajectories of offspring from 2 days postpartum until 182 days of age measured in body weights In A/J males (A) there were no significant differences between maternal control and restricted offspring but maternally restricted A/J females (B) were significantly larger by 63, 70 and 182 but not 77 days of age compared to female controls (*P = 0.05, **P = 0.008). In B6 both males (C) and females (D) maternal restriction resulted in offspring that initially grew slower than control strain and sex matched controls where maternally restricted offspring were significantly smaller (*P = 0.05 and *P = 0.03, respectively). Later both experienced catch-up growth at 75 days of age, but only male DR-B6 were significantly heavier by 182 days of age than strain and sex matched controls (**P = 0.01). Data are presented as means and standard errors of the mean.
Figure 2. A comparison of the percentage change in fat and lean body mass at 77 and 182 days of age of offspring after maternal DR relative to control offspring Data presented as percentage change of DR: control treated offspring.
Figure 3. Glucose tolerance testing represented as area under the curve (AUC) measured at 80 and 186 days of age in males (A and C) and females (B and D) When comparing the effects of dietary restriction between strain and sex only maternally restricted B6 males at both 80 and 186 days of age elicited a significant increase in AUC, indicating a greater intolerance, when compared to strain and sex matched controls (***P = 0.001; **P = 0.02, respectively). Data are presented as means and standard errors of the mean.
Table 1. Body Composition measured by PIXImus densitometry under isoflurane anaesthetic
