thermogenesis) whilst shivering was measured in four out of seven lambs in the

Transcription

1 Journal of Physiology (1992), 455, pp With 5 figures Printed in Great Britain EFFECT OF MATERNAL COLD EXPOSURE ON BROWN ADIPOSE TISSUE AND THERMOGENESIS IN THE NEONATAL LAMB BY M. E. SYMONDS, M. J. BRYANT, L. CLARKE, C. J. DARBY AND M. A. LOMAX From the Departments of Biochemistry & Physiology and Agriculture, School of Animal and Microbial Sciences, University of Reading, Whiteknights, PO Box 228, Reading RG6 2AJ (Received 7 November 1991) SUMMARY 1. This study examines the effect of chronic cold exposure during pregnancy, induced by winter shearing twin-bearing ewes 4 weeks before predicted lambing date, on 2 consumption and CO2 production during non-rapid-eye-movement (REM) sleep in lambs maintained for at least 1 h at warm (28-18 'C) and cold (14-5 'C) ambient temperatures at 1, 4, 14 and 3 days of age. This was combined with measurement of the thermogenic activity (GDP binding to uncoupling protein in mitochondrial preparations) of perirenal adipose tissue from lambs immediately after birth and at 33 days of age. 2. Lambs born from shorn (cold-exposed) ewes were 15% heavier (P < -1) and possessed 21 % (P < -1) more perirenal adipose tissue that contained 4% more protein and mitochondrial protein than unshorn (P < -5) controls. Total GDP binding in perirenal adipose tissue was 4 % greater (P < '5) in lambs born from shorn ewes but there was no difference in lipid content of this tissue between the two groups. 3. At 1 day of age, lambs born from shorn ewes exhibited a 16 % higher (P < 5) rate of 2 consumption (per kilogram bodyweight) at the warm temperature and a 4 % greater metabolic response to the cold ambient temperature. All lambs born from shorn ewes responded to cold exposure without shivering (i.e. via non-shivering thermogenesis) whilst shivering was measured in four out of seven lambs in the unshorn group. These differences had disappeared by 4 days of age as a result of a 25% increased (P < -1) rate of 2 consumption in the warm in lambs born from unshorn ewes and a 2 % decrease (P < 5) in the response to the cold in lambs from shorn ewes. Shivering during cold exposure was measured in six out of nine lambs born from shorn ewes indicating a rapid alteration in thermoregulatory responses to cold during the first few days of life. 4. The levels of GDP binding and mitochondrial protein in perirenal adipose tissue fell by one-third in both groups of lambs during the first 33 days of life whereas lipid content either increased or was unchanged. This indicated that brown adipose tissue (BAT) was developing the characteristics of white adipose tissue. In lambs aged 33 days that were born from shorn ewes, perirenal adipose tissue weighed 88 % more MS 9883

2 488 M. E SYMONDS AND OTHERS (P < 1), contained more (P < 5) protein and lipid, and the level of total GDP binding to mitochondrial protein was 1 % greater than in lambs born from unshorn ewes. There were no differences at this age between the two groups of lambs in metabolic rate but four out of eight lambs born from shorn ewes exhibited shivering responses to a cold ambient temperature compared with seven out of eight in the unshorn group. 5. It is concluded that maternal cold exposure during late pregnancy stimulates fetal growth and the thermogenic activity of BAT in the neonatal lamb. At 1 day of age this alternation in BAT was associated with an increased metabolic response to cold achieved entirely via non-shivering thermogenesis while longer term consequences were apparent in the measurement of colonic temperature, number of lambs exhibiting shivering at cold temperatures and perirenal fat deposition. INTRODUCTION Brown adipose tissue (BAT) is uniquely able to generate metabolic heat due to the presence of a mitochondrial uncoupling protein, thermogenin, which plays an important role in mammalian thermoregulatory responses to a cold environment (Canon & Nedergaard, 1985). Newborn lambs rely on non-shivering thermogenesis in BAT to produce the heat necessary to prevent hypothermia in a cold environment (Alexander & Williams, 1968) but the ability to produce heat by non-shivering thermogenesis declines rapidly during the first few weeks following birth as BAT is replaced by white adipose tissue (Thompson & Jenkinson, 1969). Since hypothermia is a major cause of the one to four million lamb deaths occurring each year in the British sheep industry (Slee, 1979) it is surprising that there is a lack of information on the mechanisms regulating BAT development and function in the fetus and neonate (see Symonds & Lomax, 1992). An apparent failure to utilize non-shivering thermogenesis during neonatal development is associated with unexpected death (Symonds, Andrews & Johnson, 1989a). However, there are no publications in which in vivo metabolic responses to cold have been linked to the appropriate measurements of the thermogenic activity of BAT (GDP binding and mitochondrial protein levels, see Trayhurn & Milner, 1989) in lambs during the first month of life. In the ovine fetus BAT is mainly found in the perirenal region and grows rapidly relative to body weight between 7-12 days of pregnancy (Alexander, 1978). After this stage only a small amount of adipose tissue growth occurs in comparison to total fetal growth, and in the case of maternal undernutrition can actually decline (Alexander, 1978). Maternal cold stress during late pregnancy stimulates the in vivo capacity for non-shivering thermogenesis in newborn lambs (Stott & Slee, 1985) and BAT deposition in neonatal rats (Hyvarinin, Pasanen, Hejkura, Heinineva & Laru, 1976). We have shown that maternal cold stress, induced by winter shearing twinpregnant ewes 8 weeks before parturition results in changes in the maternal metabolic environment which improve lamb birth weight and lamb growth rate (Symonds, Bryant & Lomax, 1986, 199). The following study extends these findings by investigating the effects of maternal cold exposure on BAT thermogenic activity and in vivo metabolic response to cold ambient temperatures in newborn lambs. A preliminary report of this work has already been published (Symonds, Bryant & Lomax, 1991).

3 BRO TT. ADIPOSE TISSL E IN THE ianeo.natal LAMB 489 METHODS Animals and diet Eighteen Bluefaced Leicester x Swaledale ewes of known tupping date and diagnosed as being pregnant with twins using a real-time ultrasound echograph were entered into this study. All ewes were housed individually at ambient temperature 5 weeks prior to lambing and 1 week later ten ewes were shorn. The mean minimum and maximum ambient temperatures recorded at 9. h for the remaining 4 weeks of pregnancy were 6.8 (S.D. 2 3) and 12 3 (S.D. 3 5) C respectively. Ewes were fed daily at 8.45 h, a diet comprising barley concentrate ( kg) and chopped hay (1-1 2 kg), which contained sufficient energy to meet 6% of the requirements for maintenance and pregnancy for the final 4 weeks of pregnancy in unshorn ewes (Agricultural Research Council, 198). We have previously shown (Symonds et al. 1986) that at this restricted level of feed intake, winter shearing significantly increases lamb birth weight in twin-bearing ewes. All ewes produced live twins normally at term (147th day of pregnancy) with two exceptions: one unshorn ewe exhibited clinical symptoms of pregnancy toxaemia from day 134 of pregnancy and was therefore given an intramuscular injection of oxycytocin on day 138 of pregnancy in order to induce labour. This animal subsequently lambed 3 days later, producing healthy twins weighing 5-94 and 5-1 kg which were included in the study. The second exception was a shorn ewe which gave birth to twins on day 153 of pregnancy following a protracted labour, but one lamb failed to survive and its twin did not establish normal drinking patterns so the results from these lambs are not included. One twin from six shorn and six unshorn ewes and both twins from one shorn and one unshorn ewe were humanely slaughtered by intravenous administration of barbiturate (1 mg kg-1 sodium pentobarbitone: Euthalol) within 2 h of birth and perirenal-abdominal adipose tissue depots were removed immediately, placed in an ice-cold solution of 1 mm-tris(hydroxymethyl)aminomethane (Tris) buffer (ph 74), containing 25 mm-sucrose and 1 mm-ethylenediaminotetracetic acid (EDTA), and then frozen at -2 'C. The remaining lambs (n = 1 born from shorn ewes; n = 8 born from unshorn ewes) were then separated from their mother within 24 h of life, after they had gained a satisfactory colostrum feed, and housed individually with the mean minimum and maximum daily ambient temperatures being 16 6 (S.D. 2 2) and 22 4 (S.D. 4-3) 'C respectively. All lambs were fed dailv a 2 1 volume of milk containing 2 g milk replacer (Volac Lamlac, Royston. Herts SG8 5QX) and from day 8 creep feed pellets (1 g day-1) were added to the diet plus chopped hay (1 g dayv1) on day 15. On day 33 of life seven lambs born from shorn ewes and six lambs born from unshorn ewes were humanely slaughtered and perirenal adipose tissue sampled. Experimental design.mlaternal. At an average of 2 weeks before predicted lambing date (day 133 of pregnancy) catheters were inserted into a jugular vein of nine shorn and six unshorn ewes. The next day, 15 ml of blood samples were taken at 1 h intervals between 9.3 and 16.3 h, the plasma collected and stored at -2 'C. The mean ambient temperature over the sampling periods was 9 'C (S.D. 3-6). Neonatal. Metabolic rate was measured in lambs sequentially when aged 1, and 3 days. During each study the lamb was maintained at a warm temperature approximating thermoneutrality (Symonds et al. 1989a; day 1, 28 TC; day 4, 25 TC; day 14, 22 'C and day 3, 19 'C) for at least 1 h and measurements were made over a 15-3 min period once a non-rapid-evemovement (REM) sleep epoch was established. Sleep state was determined from respiratory pattern measurements made using inductance plethysmography. The temperature was then decreased to a cold ambient temperature for each age group (Symonds et al a; day 1, 14 'C; day 4, 11 C; day C and day 3, 5 'C) and this procedure repeated. These measurements were made whilst the lamb was placed in an indirect calorimeter, housed in a purpose-built temperaturecontrolled room (Ileo Refrigeration Ltd. Osney Mead. Oxford OX2 OER). Throughout the experiment the animal was free to stand up or lie down and had access to its feed. Oxygen (2) consumption and carbon dioxide (CO2) production were measured as described by Symonds et al. 1989a, b) using the analytical system of Symonds, Bryant & Lomax (1986). Colonic temperature was measured using an electronic thermometer (Type 3GlD, Light Laboratories, Brighton [N2 IRD) and the occurrence of shivering was assessed from observations of interference on the respiratory pattern measured using inductance plethvsmography (Symonds et al. 1989a). Seven lambs born from shorn ewes and five lambs born from unshorn ewes had catheters inserted into a jugular vein on the first day of life at least 1 h before any measurements were made. These catheters remained in place for the remainder of the experiment unless they became blocked, in which case

4 49 M. E. SYMONDS AND OTHERS they were removed and a new catheter inserted on the day before a study. All remaining lambs were catheterized when aged 3 days. Catheters were filled with fresh saline containing heparin (25 units ml-1) two or three times a week, and on each study day the content of heparin was reduced to 5 units ml-'. After a minimum of 15 min in non-rem sleep a 5-1 ml venous blood sample was taken without arousing the lamb. The plasma was collected and stored at -2 'C. Values obtained at 1 day of age for a set of twins born from a shorn ewe are not included because the plasma concentration of glucose in these animals was found to be approximately 1-5 mm. This compares to a mean value of 5-5 mm in other lambs born to shorn ewes studied at this age. It is therefore likely that the supply of milk to these lambs before removal from the mother was limited, hence they became hypoglycaemic. Once feeding was established in the laboratory, glucose levels returned to normal in these lambs and their growth and development was similar to other lambs born from shorn ewes, therefore subsequent data obtained are included in the mean results for these animals. Laboratory procedures Mitochondrial preparations were made from frozen adipose tissue based on the method of Casteilla, Forest, Robelin, Ricquier, Lombet & Ailhaud (1987). Two grams of tissue were thawed and homogenized in the same Tris buffer used for sampling to give a final volume of 4 ml. The homogenate was then centrifuged at 8 g for 1 min and the supernatant passed through two layers of surgical gauze in order to remove lipid tissue, and then centrifuged at 1 9 for 3 min. The supernatant was discarded and the mitochondrial pellet resuspended in a total volume of between 5 and 1- ml Tris buffer and frozen at -2 'C until assayed for guanosine diphosphate (GDP) binding. Aliquots of the homogenate and mitochondria were also frozen for measurement of protein content (Lowry, Rosebrough, Farr & Randall, 1951) and cytochrome c oxidase activity (Wharton & Tzagoloff, 1967) in order to assess the recovery of mitochondrial protein. BAT thermogenic capacity was then assessed from the in vitro activity of the mitochondrial proton conductance pathway from the binding of radiolabelled GDP to mitochondrial preparations (Cooper, Dascombe, Rothwell & Vale, 1989). Before each assay the mitochondrial suspension was diluted to 1 mg protein ml-' with Tris buffer. GDP binding was assessed in triplicate over a 7 min period at room temperature using -1 ml of mitochondrial suspension added to -85 ml incubation buffer consisting of 1 mm-sucrose, 2 mm-tes (N-tris(hydroxymethyl) methyl-2-aminoethanesulphonic acid: 2-([2-hydroxy- 1, 1-bis(hydroxymethyl)ethyl]amino) ethanesulphonic acid), 1 mm-edta, 1/tM-choline chloride, -8% bovine serum albumin (free fatty acid free), 5, M- rotenone, 2 ItM-GDP plus 64 pmol [3H]GDP (specific activity 1 Ci mmol-1; Amersham International plc) and 72 pmol [14C]sucrose (specific activity 5 mci mmol-1; Amersham International plc) added as a marker for the extra mitochondrial volume. None-specific binding was measured using the same incubation buffer but containing 2 ItM-GDP. The reaction was stopped by placing the tubes in an MSE microcentaur centrifuge for 3 min at 13 r.p.m. and the supernatant was removed. Pellets were then dissolved by the addition of -1 ml -75 M-sodium hydroxide and incubating at 55 'C for 15 min, 5 ml was then taken and mixed with 5 ml Ecosant A (National Diagnostics) and both [3H] and [14C] counted using a Packard Tri-carb scintillation counter. The amount of [3H]-GDP bound to mitochondrial pellets was corrected for extra mitochondrial binding measured using [14C]sucrose and specific binding was calculated by subtraction of non-specific binding. The dry matter content of fat was measured by freeze-drying a 5 g sample which was then used for lipid analysis via ether extraction. Plasma concentrations of D(-)-3-hydroxybutyrate, glucose, lactate and non-esterified fatty acids (NEFA), packed cell volume and blood concentration of haemoglobin were all measured as described by Symonds et al. (1986, 1989c) and Symonds, Bryant, Shepherd & Lomax (1988a). Statistical analysis Statistical analysis with respect to differences between shorn and unshorn ewes, or lambs born from shorn and unshorn ewes were assessed by Student's t test. Changes with respect to ambient temperature or age within each group of lambs were assessed by Student's paired t test.

5 BROWN ADIPOSE TISSUE IN THE NEONATAL LAMB 491 RESULTS Lamb growth, ewe live weight and feed intake The mean birth weight of lambs born from shorn ewes was 15 % higher than those from unshorn controls (shorn kg, n = 18; unshorn kg, n = 16; P < -1). There was no difference in the intake of concentrate (shorn g, n = 1; TABLE 1. Mean packed cell volume, jugular venous blood concentration of haemoglobin, and plasma concentrations of glucose, lactate, non-esterified fatty acids and 3-hydroxybutyrate in shorn (n = 9) or unshorn (n = 6) twin-bearing ewes on day 133 of pregnancy Shorn Unshorn Packed cell volume (%) Haemoglobin (g 1-l) Glucose (mm) Lactate (mm) NEFA (mm) Hydroxybutyrate (mm) Values are means+s.e.m. unshorn g; n = 8) between the two groups but shorn ewes consumed 19% more hay (shorn g, n = 1; unshorn g, n = 8; P < -5) over the final 4 weeks of pregnancy. The results from our previous study (Symonds et al. 1986) would predict that metabolizable energy intake was similar in the two groups since although hay intake was higher, there is a decrease in apparent digestibility of gross energy in shorn ewes. There was no difference in the growth rate of lambs born from either shorn or unshorn ewes, over the first month of life. All animals gained weight over this period with the mean body weights for lambs born from shorn ewes (n = 1) at 4, 14 and 3 days of age being , and kg respectively and the mean weights for lambs born from unshorn ewes (n = 8) at these ages were , and kg. Maternal plasma metabolic concentrations There was no difference in the mean packed cell volume or blood haemoglobin concentrations between shorn and unshorn groups as measured 2 weeks before predicted lambing date (Table 1). Shorn ewes did, however, exhibit a 17 % higher plasma concentration of glucose whilst 3-hydroxybutyrate (P = -8) levels were half those recorded in the unshorn group despite no difference in the plasma concentration of NEFA or lactate between shorn and unshorn animals. Neonatal perirenal adipose composition When perirenal adipose tissue was sampled within 2 h of birth it was found that lambs born from shorn ewes possessed 21 % more perirenal adipose tissue which contained 4 % more protein and mitochondrial protein compared to unshorn controls (Table 2). GDP binding to mitochondrial protein was found in all samples and was similar in both groups of lambs. The total amount of GDP binding was 4 % higher in lambs born from shorn ewes as a result of the increased level of mitochondrial protein. There was no difference in the lipid or dry matter content of perirenal adipose tissue when sampled from the two groups within 2 h of birth. By

6 492 M. E. SYMONDS AND OTHERS 1 +l C> S. 6 N 6 +l 9 o 9 11 r 1 ce N ca c O- t- r q ~ S +l o CO t- - 1 c3 4iD.5 t DZ L- 1- S Sce 4-~ B~._ od _1.e o _ SI C. o._e o1 1 r Il 1 't N N N - CO r- 1 r -4 cs oc CO S Ca -.. GS C,-i._- _:- -4D o = E N oo X CO W -4._ - En 1I 1 1 1o CO 1 es C 1 9l l L. ic) e r.o cet X 'V,_e._ o oo E = O E, ~ 4 MC 11 ce n2n e

7 BROWN ADIPOSE TISSUE IN THE NEONATAL LAMB days of age the magnitude of these differences had increased to the extent that perirenal adipose tissue sampled in lambs born from shorn ewes weighed 88 % more and contained 89 % more protein, 73 % more mitochondria protein and had twice the amount of lipid than the same fat depot sampled in lambs born from unshorn ewes _ :,S'; ox f. k xe 2 Age (days Cod 2 AgeL 1 Idys E~Ae dys g (as Fig. 1. Mean oxygen consumption and carbon dioxide production in lambs born from shorn (. L]) or unshorn (@, U) pregnant ewes when measured during non-rem sleep at warm or cold ambient temperatures over the first month of life. Values are means+5.e.m. In the shorn group at day 1, n = 7; day 4, n = 9; day 14, n = 9 and day 3, n = 8; and in unshorn group at day 1, n = 7 and days 4-3, n = 8. During the first month of life the weight of perirenal adipose tissue fell by 37 % in lambs born from unshorn ewes but increased by 12 % in those born from shorn ewes as a result of a higher lipid content of the tissue. In all lambs the level of protein and GDP binding had fallen to approximately one-third of the value recorded at birth with the total amount of GDP binding being 1 % of values recorded at birth. Neonatal metabolic responses to warm andl cold ambient temperatures during non- REM sleep At 1 day of age lambs born from shorn ewes had a 16% higher (P < 5) rate of 2 consumption at the warm temperature than those born from unshorn animals (Fig. 1). Metabolic rate increased after a short period of cold exposure in both groups of lambs with thisresponse being 4 % greater in lambs born from shorn ewes (shorn , n = 7; unshorn ml mink (kg body weight)', n = 7) although this difference was not statistically significant o (P = l). However, whereas all of the lambs born from shorn ewes responded to cold exposure without shivering (i.e. via non-shivering thermogenesis), shiver beiwas measured in four out of seven lambs in

8 494 M. E. SYMONDS AND OTHERS the unshorn group. These differences between the two groups of lambs had disappeared by 4 days of age as a result of a 25% increased (P < 1) rate of 2 consumption in the warm in lambs born from unshorn ewes compared to day 1 and a 2 % decrease (P < 5) in the metabolic response to cold exposure in the shorn 2 Warm 2- Cold x O.,.., O~~~~~~~~~~~~~~. 1 1~~~~~ 1 Age (days) ~~~~~Age(days) Fig. 2. Mean oxygen consumption from fat oxidation in lambs born from shorn (, LI) or unshorn (-, ) pregnant ewes when measured during non-rem sleep at warm or cold ambient temperatures over the first month of life. Values are means + 5.E.M. In the shorn group at day 1, n = 7, day 4, n = 9; day 14, n = 9 and day 3, n = 8; and in the unshorn group at day 1, n= 7 and days 4-3, n= 8. group. In this latter group six of the nine animals studied at this age adopted shivering thermogenesis, compared with four of the eight lambs born from unshorn ewes. There were no further differences in metabolic rate as measured at warm or cold ambient temperatures between the two groups of lambs at 14 or 3 days of age, although the increase in oxygen consumption during cold exposure fell by one-fifth between day 1 and day 3 of life. All animals responded to cold exposure via shivering thermogenesis at 14 days of age, but only four of the eight lambs born from shown ewes exhibited shivering thermogenesis at 3 days of age compared to seven out of eight in the unshorn group. The above changes in 2 consumption tended to be paralleled by similar responses in CO2 production (Fig. 1). There were, however, some differences in the magnitude of the increase in 2 production thereby indicating changes in the proportion of fat and carbohydrate being oxidized. This is illustrated in Fig. 2 in which the rate of fat oxidation, as calculated from the respiratory quotient (2 consumption/co2 production; Symonds et al. 1989b) is presented. When studied in the warm at 1 day of age, lambs born from shorn ewes exhibited double (P < ~1) the rate of fat oxidation observed in those born from unshorn ewes. There were no further differences in fat oxidation measured in the warm between the two groups at subsequent ages. The rate of fat oxidation increased significantly during cold

9 BROWN ADIPOSE TISSUE IN THE NEONATAL LAMB exposure in all animals on days 1 and 4 of age, but this response declined with age. Differences between the groups were apparent at 4 days of age when lambs born from unshorn ewes exhibited a 72 % higher (P = 9) rate of fat oxidation than in the shorn group, and at 14 days when fat oxidation was 6% lower (P < -1) in lambs from unshorn ewes. 41- Warm 41 Cold Q) I~~~~~ a C Age (days) Age (days) Fig. 3. Mean colonic temperature in lambs born from shorn (O. El) or unshorn (I, U) pregnant ewes when measured during non-rem sleep at warm or cold ambient temperatures over the first month of life. Values are means + S.E.M. In the shorn group at day 1, n = 7; day 4, n = 9; day 14, n = 9 and day 3, n = 8; and in the unshorn group at day 1, n = 7 and days 4-3, n = 8. Colonic temperature There was an increase in colonic temperature of all lambs measured at the warm ambient temperature during the first month of life (Fig. 3). Colonic temperature tended to be higher in lambs born from unshorn ewes when studied in the warm with this difference being significant (P < -5) at 4 days of age. This difference was accentuated during cold exposure with colonic temperature being significantly (P < 5) higher in lambs from unshorn ewes on days 4 and 14 of age. Colonic temperature increases in response to cold stress were significant (P < 5) on days 1 and 4 in lambs from shorn ewes and on day 14 of age in lambs from the unshorn group. Neonatal plasma metabolite concentrations In both groups of animals there was a significant (P < 5) fall in NEFA levels after 1 day of age when measured in the warm, an effect not observed until 14 days of age when studied at cold ambient temperatures (Fig. 4). However plasma NEFA concentrations were not significantly elevated by cold exposure and there were no differences between lambs born from shorn or unshorn ewes. The concentration of glucose in plasma increased between 1 and 4 days of age although this effect was only significant (P < 5) in lambs born from shorn ewes (Fig. 5). Cold exposure had no

10 ~ m 35 an _~ '5 3 ` _ ).c 'Z 2-5 9' o C' M 5 Cu 15-5 M. E. SYMONDS AND OTHERS Warm Fig. 4. Mean jugular venous plasma concentration of non-esterified fatty acid in lambs born from shorn (O. El) or unshorn (I, U) pregnant ewes when measured during non- REM sleep at warm or cold ambient temperatures over the first month of life. Values are means+s.e.m. In the shorn group at day 1, n = 5; day 4, n = 9; day 14, n = 9 and day 3, n = 7; and in the unshorn group at day 1, n = 5 and days 4-3, n = Cold Age (days) Age (days) o - mm E Mc '3 E:co ?4. 2 Warm Cold Age (days) Age (days) Fig. 5. Mean jugular venous plasma concentration of glucose in lambs born from shorn (, El) or unshorn (I, U) pregnant ewes when measured during non-rem sleep at warm or cold ambient temperatures over the first month of life. Values are means + S.E.M. In the shorn group at day 1, n = 5; day 4, n = 9; day 14, n = 9 and day 3, n = 7; and in the unshorn group at day 1, n = 5 and days 4-3, n = 7. effect on the plasma concentration of glucose in lambs born from shorn ewes, but a 75 % increase (P < -5) in glucose levels was recorded at 1 day of age in the unshorn group.

11 BROWN ADIPOSE TISSUE IN THE NEONATAL LAMB 497 DISCUSSION The present study clearly indicates that maternal metabolic adaptations to chronic cold exposure can alter fetal development, with BAT being particularly sensitive to this effect. This change in BAT function appears to remain for at least the first month of life and indicates that maternal influences on the fetus can significantly affect neonatal adaptation to the extra-uterine environment. Maternal adaptation to chronic cold exposure Chronic cold exposure, induced by winter shearing ewes 8 weeks before lambing, has been shown to stimulate oxidation of maternal fat reserves to meet a higher rate of whole-body heat production (Symonds et al. 1986, 1989c). This occurs in conjunction with a stimulation of maternal glucose entry rate and can result in higher circulating glucose levels in shorn animals over the final 2 weeks of pregnancy (Symonds et al. 1988a, b). In the present study chronic cold exposure was induced by winter shearing 4 weeks before predicted lambing date and resulted in shorn ewes exhibiting a higher plasma concentration of glucose and producing heavier lambs compared with unshorn controls. It has been shown that acute cold exposure of sheep increases the plasma concentration of glucose in both mother and fetus as well as stimulating fetal growth (Thompson, Bassett, Sansom & Slee, 1982). Intravenous infusion of glucose into fetuses of single-bearing ewes during the final 4 weeks of pregnancy, results in fetal plasma concentrations of glucose and insulin doubling and also stimulates fetal growth, to the extent that lamb birth weight is 18 % higher than in saline-infused controls (Stevens, Alexander & Bell, 199). This increase in birth weight is similar in magnitude to that observed from shorn ewes in the present study and suggests that an increased rate of maternal supply to the fetus may be responsible for the higher lamb birth weight in shorn animals. Stevens et al. (199) also reported that the weight of internal fat depots was 73 % greater in glucoseinfused fetuses, although no further tissue analysis was performed. This contrasts with the response observed in lambs born from shorn ewes in which the increased weight of perirenal adipose tissue was of similar magnitude to the overall rise in lamb birth weight and may be due to the higher feed intakes in the study by Stevens et al. (199). Brown adipose tissue and neonatal metabolic rate The thermogenic activity of BAT uncoupling protein, thermogenin, can be assessed by the level of GDP binding to mitochondrial protein (see Trayhurn & Milner, 1989), and when made in conjunction with an assessment of the total amount of this protein in BAT gives an index of the total thermogenic activity of this tissue. The present study is the first to be undertaken in the ovine species in which the thermogenic activity of BAT has been compared with measurements of whole-body metabolic rate at warm and cold ambient temperatures. The results demonstrate that maternal metabolic responses to cold exposure influenced fetal BAT development such that total BAT thermogenic activity, measured within 2 h of birth, was increased by 4 %. This increase was entirely due to higher levels of mitochondrial protein rather than any change in GDP binding. Stimulation of the

12 498 M. E. SYMONDS AND OTHERS synthesis of thermogenin is likely to be the major factor contributing to the increased thermogenic activity of BAT sampled in lambs born from shorn ewes. In adult rodents the principal factors regulating thermogenin synthesis and BAT thermogenic activity are noradrenaline (Mory, Bouillard, Combes-George & Riquier, 1984), insulin (Geloen & Trayhurn, 199) and production of triiodothyronine within BAT (Bianco & Silva, 1987). It is not known if these factors influence fetal BAT development but they represent possible mechanisms which could be influenced by changes in nutrient supply to the fetus as a result of maternal cold exposure. The alterations in mitochondrial protein were accompanied by higher BAT weight and protein but not lipid content in newborn lambs from shorn compared with unshorn ewes. The increased thermogenic activity of BAT in newborn lambs from shorn ewes was associated in vivo with an enhanced use of non-shivering thermogenesis in the cold since shivering was only observed in lambs born from unshorn ewes (four out of seven) on day 1 of life. The lack of shivering during cold exposure in lambs born from shorn ewes suggests an increased ability to generate heat via non-shivering thermogenesis in BAT (Alexander & Williams, 1968), a proposal that is supported by the 4 % greater increase in oxygen consumption during cold exposure compared to lambs born from unshorn ewes. Metabolic rate on day 1 of life (Fig. 1) was also 16% higher in lambs from shorn ewes measured in warm ambient temperatures and this was associated with a significant increase in fat oxidation in this group (Fig. 2). We have recently observed that in lambs which are reared in a cold environment over the first week of life there is an increase in BAT thermogenic activity activity which is associated with higher metabolic rates at both warm and cold ambient temperatures (Darby, Clarke, Lomax & Symonds, 1992). Since it is unlikely that BAT thermogenesis makes an appreciable contribution to metabolic rate in the warm environment (28 C) it is possible that factors influencing BAT also regulate basal metabolism. Differences between the two groups of lambs were not confined to BAT metabolism because lambs born from unshorn ewes appeared to be more reliant on hepatic glycogen as an energy source on day 1 of life since there was a 75 % increase in the plasma concentration of glucose during cold exposure, a response not observed in the shorn group. Lambs born from shorn ewes may therefore benefit from increased utilization of BAT lipid stores as a source of heat production rather than using glycogen stores, for shivering thermogenesis, with the latter situation being associated with a greater incidence of hypothermia (Mellor & Cockburn, 1986). In lambs born from unshorn ewes there was a rise in metabolic rate between days 1 and 4 of life (Fig. 1) in both cold and warm ambient temperatures. Mount (1969) reported that a similar rise in metabolic rate in the neonatal pig is related to the intake of milk. In contrast, the lambs born from shorn ewes showed no alteration in metabolic rate in the warm over this period, due to a higher value measured on day 1 of life, and the metabolic response to cold exposure on day 4 was markedly reduced to values similar to those in lambs from unshorn ewes (Fig. 1). The higher capacity for non-shivering thermogenesis in 1-day-old lambs born from shorn ewes was therefore rapidly lost since by day 4 of life, six out of nine lambs adopted shivering thermogenesis in the cold. At 14 days of age all lambs responded to cold temperature by shivering indicating a transition from non-shivering to shivering thermogenesis as

13 BROWN ADIPOSE TISSUE IN THE NEONATAL LAMB a result of a decrease in BAT that occurs in lambs over the first weeks of life (Thompson & Jenkinson, 1969) as BAT develops white adipose tissue characteristics (Gemmel, Bell & Alexander, 1972). Analysis of perirenal adipose tissue at 33 days of age confirmed this loss of BAT since in both groups of lambs there was a 7-8 % reduction in the level of mitochondrial protein and a 9 % decrease in the amount of GDP binding, thereby causing the total thermogenic capacity of the tissue to be 7 % of the value recorded within 2 h of birth. The thermogenic activity of the perirenal adipose tissue in lambs from shorn ewes was still 5% higher than those from unshorn controls which could have contributed to the different behavioural response to acute cold exposure at 3 days of age when all lambs born from unshorn ewes exhibited shivering thermogenesis, compared with only half of the shorn group. In the unshorn group there was a 37 % decrease in perirenal adipose tissue weight and a small fall in lipid content between 1 and 33 days of age. This contrasted with the lambs born from shorn ewes in which the weight and lipid content of perirenal adipose tissue increased such that at 33 days of age this tissue contained twice the lipid present in lambs born from unshorn ewes. This effect was observed despite the identical feed intake and ambient conditions in which the two groups of lambs were reared and suggests that the thermogenic activity of BAT in the perirenal region at birth is related to the amount of lipid stored in white adipose tissue in this region after 1 month of life. An increase in fetal brown adipocyte proliferation during pregnancy due to maternal cold stress may result in an increased number of adipocytes which become filled with lipid over the first month of life. The high plasma concentrations of NEFA (1-5 mm) measured in all lambs when studied on day 1 of life are likely to be caused by the stimulation of fat mobilization which occurs at birth (Sack, Beaudry, Delamater, Oh & Fisher, 1976). NEFA levels fall dramatically over the first 2 weeks of life irrespective of ambient conditions and there is a fall in the rate at which fat oxidation is stimulated in response to cold exposure, although this effect is delayed in lambs born from shorn ewes. The absence of any significant change in the plasma concentration of NEFA during cold exposure is not unexpected because even the onset of summit metabolism fails to elicit such a response in lambs studied over the first month of life (Alexander, Mills & Scott, 1968). This is likely to be due to the rapid utilization of NEFA released from triglycerides as a result of oxidation in BAT (Cannon & Nedergaard, 1985) or by shivering muscle tissue (Bell & Thompson, 1979). Neonatal colonic temperature Between 4 and 3 days of age there was a decrease in metabolic rate and a fivefold decrease in the metabolic response to cold exposure in all lambs as a result of an increased thermal efficiency (Symonds et al a). There were no differences in metabolic rate between shorn and unshorn groups but colonic temperatures measured at both warm and cold temperatures were significantly higher in lambs born from unshorn ewes and remained lower in the cold until day 14 of life (Fig. 3). This suggests that maternal influences on fetal development have long-term implications for thermoregulation either by slowing the increase in the skin thickness, fleece depth and total body insulation that occurs during the first 4 days of life (Stott, 1985) or by altering the hypothalamic temperature to which thermoregulatory mechanisms 499

14 5 M. E. SYMONDS AND OTHERS respond (Boulant, 1981). Neonatal hypothalamic temperatures may be influenced by maternal temperature since it is known that fetal and maternal temperatures are closely linked both throughout the day (Gluckman, Gunn, Johnston & Quinn, 1984) and following acute changes in maternal metabolism such as exercise (Lotgering, Gilbert & Longo, 1983). Colonic temperature in unshorn ewes are significantly higher than shorn ewes during the last month of pregnancy (Russel, Armstrong & White, 1985) and this may be responsible for a lower hypothalamic 'set point' thereby resulting in lambs from unshorn ewes exhibiting higher colonic temperatures between days 4 and 14 of life. The lack of difference between the groups in colonic temperatures on day 1 of life is likely to reflect the dramatic changes in metabolic rate that occur in lambs over the first day of life (Mercer, Andrews & Szekely, 1979). The present study suggests that exposure of the pregnant ewe to a cold environment results in maternal signals to the fetus which cause the lamb to be born with certain survival advantages. These include an increase in lamb birth weight, an increase in the thermogenic activity of BAT, an increased non-shivering thermogenic response to cold and a decrease in glycogen mobilization during cold exposure. Although these responses to cold appeared to have largely disappeared by day 4 of life, it is evident that maternal influences on the fetus can affect neonatal development since effects on colonic temperature and fat deposition in perineal adipose tissue were still observed several weeks after birth. This work was funded by the XWellcome Trust and an AFRC research grant. L. C. is supported by an MRC Studentship. REFERENCES AGRICULTURAL RESEARCH COUNCIL (198). Requirements for energy. In The Nutrient Requirements of Ruminant Livestock, pp Commonwealth Agricultural Bureaux, Slough. ALEXANDER, G. (1978). Quantitative development of adipose tissue in foetal sheep. Australian Journal of Biological Sciences 31, ALEXANDER, G., MILLS, S. C. & SCOTT, T. W. (1968). Changes in plasma glucose, lactate and free fatty acids in lambs during summit metabolism and during treatment with catecholamines. Journal of Physiology 198, ALEXANDER, G. & WILLIAMS, D. (1968). Shivering and non-shivering thermogenesis during summit metabolism in young lambs. Journal of Physiology 198, BELL, A. W. & THOMPSON, G. E. (1979). Free fatty acid oxidation in bovine muscle in vivo: effects of cold exposure and feeding. American Journal of Physiology 237, E BIANCO, A. C. & SILVA, J. E. (1987). Optimal response of key enzymes and uncoupling protein to cold in BAT depends on local T3 generation. American Journal of Physiology 253, E BOULANT, J. A. (1981). Hypothalamic mechanisms in thermoregulation. Federation Proceedings 4, CANNON, B. & NEDERGAARD, J. (1985). The biochemistry of an inefficient tissue: brown adipose tissue. Essays in Biochemistry 2, CASTEILLA, L., FOREST, C., ROBELIN, J., RICQUIER, P., LOMBET, A. & AILHAUD, G. (1987). Characterization of mitochondrial-uncoupling of protein in bovine fetus and newborn calf. American Journal of Physiology 252. E COOPER, A. L., DASCOMBE, M. J., ROTHWELL, N. J. & VALE. M. J. (1989). Effects of malaria on 2 consumption and brown adipose tissue activity in mice. Journal of Applied Physiology 67, DARBY, C. J., CLARKE, L., LOMAX, M. A. & SYMONDS, M. E. (1992). Effect of rearing neonatal lambs in a cold and warm environment on thermogenesis during slow wave sleep. Proceedings of the Nutrition Society (in the Press).

15 BROWN ADIPOSE TISSUE IN THE NEONATAL LAMB GELOEN, A. & TRAYHURN, P. (199). Regulation of the level of uncoupling protein in brown adipose tissue by insulin. American Journal of Physiology 258, R GEMMEL, R. T., BELL, A. W. & ALEXANDER, G. (1972). Morphology of adipose cells in lambs at birth and during subsequent transition of brown to white adipose tissue in cold and warm conditions. American Journal of Anatomy 133, GLUCKMAN, P. D., GUNN, T. R., JOHNSTON, B. M. & QUINN, J. P. (1984). Manipulation of the temperature of the fetal lamb in utero. In Animal Models in Fetal Medicine, ed. NATHANIELSZ, P. W., pp Perinatology Press, New York. HYVARINEN, H., PASANEN, S., HEJKURA, H., HEININEVA, R. & LARU, H. (1976). Effects of a cold environment on energy-related enzyme activities in the postnatal rat. Growth 4, LOTGERING, F. K., GILBERT, R. D. & LONGO, L. D. (1983). Exercise responses in pregnant sheep: blood gases, temperatures and fetal cardiovascular system. Journal of Applied Physiology 55, LOWRY,. H., ROSEBROUGH, N. J., FARR, A. L. & RANDALL, R. J. (1951). Protein measurement with the Folin phenol reagent. Journal of Biological Chemistry 193, MELLOR, D. J. & COCKBURN, F. (1986). A comparison of energy metabolism in the new-born infant, piglet and lamb. Quarterly Journal of Experimental Physiology 71, MERCER, J. B., ANDREWS, J. F. & SZEKELY, M. (1979). Thermoregulatory responses in new-born lambs during the first thirty-six hours of life. Journal of Thermal Biology 4, MORY, G., BOUILLAND, F., COMBES-GEORGE, M. & RIQUIER, D. (1984). Noradrenaline controls the concentration of the uncoupling protein brown adipose tissue. FEBS Letters 166, MOUNT, L. E. (1969). The respiratory quotient in the newborn pig. British Journal ofnutrition 23, RUSSEL, A. J. F., ARMSTRONG, R. H. & WHITE, I. R. (1985). Studies on the shearing of housed pregnant ewes. Animal Production 4, SACK, J., BEAUDRY, M., DELAMATER, P. V., OH, W. & FISHER, D. A. (1976). Umbilical cord cutting triggers hypertriiodothyroninemia and non-shivering thermogenesis in the newborn lamb. Pediatric Research 1, SLEE, J. (1979). Mortality and resistance to hypothermia in young lambs. In Biometeorological Survey, vol. IB, ed. TROMP, S. W. & BOUMA, J. J., pp Heyden, Philadelphia. STEVENS, D., ALEXANDER, G. & BELL, A. W. (199). Effect of prolonged glucose infusion into fetal sheep on body growth, fat deposition and gestation length. Journal of Developmental Physiology 13, STOTT, A. W. (1985). Effects of previous cold exposure on the cold resistance of young lambs. In Progress in Biometeorology, vol. 2, ed. TROMP, S. W., pp Swets & Zeitlinger B.V., Lisse, Netherlands. STOTT, A. W. & SLEE, J. (1985). The effect of environmental temperature during pregnancy on thermoregulation in the newborn lamb. Animal Production 41, SYMONDS, M. W., ANDREWS, D. C. & JOHNSON, P. (1989 a). The control of thermoregulation in the developing lamb during slow wave sleep. Journal of Developmental Physiology 11, SYMONDS, M. E., ANDREWS, D. C. & JOHNSON, P. (1989b). The endocrine and metabolic response to feeding in the developing lamb. Journal of Endocrinology 123, SYMONDS, M. E,, BRYANT, M. J. & LOMAX, M. A. (1986). The effects of shearing on the energy metabolism of the pregnant ewe. British Journal of Nutrition 56, SYMONDS, M. E., BRYANT, M. J. & LOMAX, M. A. (1988b). Metabolic adaptation during pregnancy in winter-shorn sheep. Journal of Agricultural Science 111, SYMONDS, M. E., BRYANT, M. J. & LOMAX, M. A. (1989c). Lipid metabolism in shorn and unshorn pregnant sheep. British Journal of Nutrition 62, SYMONDS, M. E., BRYANT, M. J. & LOMAX, M. A. (199). Metabolic adaptation during lactation in winter-shorn sheep. Journal of Agricultural Science 114, SYMONDS, M. E., BRYANT, M. J. & LOMAX, M. A. (1991). The effect of chronic cold exposure during late pregnancy on brown adipose tissue and thermogenesis in the neonatal lamb. Journal of Physiology 438, 236P. SYMONDS, M. E., BRYANT, M. J., SHEPHERD, D. A. L. & LOMAX, M. A. (1988a). Glucose metabolism in shorn and unshorn pregnant sheep. British Journal of Nutrition 6, SYMONDS, M. E., & LOMAX, M. A. (1992). Maternal and environmental influences on thermoregulation in the neonate. Proceedings of the Nutrition Society (in the Press). 51

16 52 M. E. SYMONDS AND OTHERS THOMPSON, G. E., BASSETT, J. M., SAMSON, D. E. & SLEE, J. (1982). The effect of cold exposure of pregnant sheep on foetal plasma nutrients, hormones and birth weight. British Journal of Nutrition 48, THOMPSON, G. E., & JENKINSON, D. McE. (1969). Nonshivering thermogenesis in the newborn lamb. Canadian Journal of Physiology and Pharmacy 47, TRAYHURN, P. & MILNER, R. E. (1989). A commentary on the interpretation of in vitro biochemical measures of brown adipose tissue thermogenesis. Canadian Journal of Physiology and Pharmacology 67, WHARTON, D. C. & TZAGOLOFF, A. (1967). Cytochrome oxidase from beef heart mitochondria. Methods in Enzymology 1,