Neonatal calf mortality losses are second only to infertility in contributing to low reproductive efficiency rates in cattle. Mortality losses of calves during the neonatal period typically range from 3 to 14% and can at times exceed 25% (Cundiff et al. 1986). One of the primary etiologic factors associated with neonatal calf mortality is hypothermia induced by cold ambient temperatures (Azzam et al. 1993). A major thermoregulatory mechanism for survival of neonatal ruminants during cold stress is heat production by brown adipose tissue (BAT) ⁵. Neonatal ruminants are capable of generating heat from BAT for the first 2-3 wk of life by uncoupling oxidative phosphorylation from mitochondrial respiration (Carstens 1994). Maximal thermogenic response to cold is more than twofold higher than thermoneutral metabolism in neonatal calves (Robinson and Young 1988), and approximately one-half of maximal thermogenic response in newborn lambs is derived from nonshivering thermogenesis (Stott and Slee 1985).

Maternal prepartum malnutrition during late gestation adversely affects survival of newborn calves (Corah et al. 1975, Hight 1966). Prepartum malnutrition of the dam may impair cold tolerance of newborn calves by limiting deposition and thermogenic capacity of BAT (Alexander 1978, Tyzbir 1984). Additionally, the availability of energy substrates to support thermogenesis in newborn calves may be reduced by prepartum malnutrition (Mellor and Murray 1985). Therefore we examined the effects of prepartum protein restriction on BAT thermogenesis, lipid biosynthesis, cellularity and uncoupling protein (UCP) gene expression in BAT from newborn calves. A second objective was to characterize and contrast the metabolism and morphology of BAT and white adipose tissue (WAT) of neonatal calves.

MATERIALS AND METHODS

Table 1. Ingredient and chemical composition of experimental diets [View Table]

1

1

At 12 h of age the calves were killed with an overdose of sodium pentobarbital and exsanguinated. A sample of the perirenal adipose tissue was excised immediately and transported to the laboratory in Krebs-Henselheit Ca²⁺-free bicarbonate buffer containing 5 mmol/L glucose. Additional samples were frozen immediately in liquid nitrogen and stored at 70°C until further analysis. Remaining perirenal adipose tissue, heart, liver, lungs and trachea, empty gastrointestinal tract, kidneys, right adrenal gland, and spleen were excised and weighed.

All procedures and protocols involving the use of these animals were approved by the Institutional Agricultural Animal Care and Use Committee at Texas A&M University.

Fixed cells were filtered through 240-, 64- and 20-µm nylon mesh screens using 0.01% Triton in 0.154 mol/L NaCl. The cells from the 20-µm screen were counted and sized by a model ZM Coulter Counter, equipped with a model 256 channelizer (Coulter Electronics, Hialeah, FL), using a 280-µm aperature. The cells from the 64-µm screen were counted on the same equipment with a 400-µm aperature. Adipocytes were divided into 10-µm intervals; those counted with the smaller aperature were observed in channels between 20 and 60 µm, whereas those counted with the larger aperature were in channels ranging from 60 to 190 µm. Cells occurring in the 60-µm channel from both aperatures were summed.

For Northern blot analysis, 40 µg of total RNA was denatured at 68°C, separated by electrophoresis on a 1.0% agarose gel containing formaldehyde and capillary transferred to nylon Hybond N⁺ membrane (Amersham Life Science, Arlington Heights, IL). Transfer efficiency was checked by ethidium bromide UV visualization. The membrane was baked at 80°C for 2 h and UV crosslinked (UV Stratalinker 1800, Stragene Cloning System, La Jolla, CA). The blot was prehybridized with 0.1 g/L salmon sperm DNA for 2 h at 55°C. A UCP mRNA probe was generated by polymerase chain reaction (PCR). The template DNA was the bovine calf UCP 1.4-kb cDNA (generously provided by L. Casteilla, Centre de Rechere, CNRS, Meudon-Bellevue, France) linearized with EcoR1. The primers were 5-CTC AGC GGG CCT AAC GAC-3 and 5-GTT TGT TTT TCA CCA GGG-3, which produced a PCR product approximately 350 bp in size. The PCR-generated UCP probe was radiolabeled with [-³²P]dCTP by random primer method (Gibco BRL Life Technologies, Grand Island, NY) and hybridized to the RNA. The blot was rinsed once with 0.1% SDS in 2× SSC (300 mmol/L NaCl, 30 mmol/L trisodium acetate) at 42°C for 15 min and applied to Kodak X-AR5 X-ray film (Eastman Kodak, Rochester, NY) for 10 d. To test the efficacy of the PCR-generated UCP probe, perirenal adipose tissue was obtained from two newborn and two 7-d cold-adapted calves from a separate study. As a negative control, RNA was extracted from bovine longissimus dorsi muscle (from an adult animal) that had been snap-frozen in liquid nitrogen. The RNA was extracted, and Northern blot analysis was performed as described above.

A commercial slot blot apparatus (Schleicher & Schuell, Keene, NH) was used to quantify the relative amounts of UCP mRNA in the perirenal adipose tissue samples. Ten micrograms of total RNA were incubated at 65°C for 5 min in three volumes (v/v) of the following solutions: 500 µL formamide, 162 µL formaldehyde (37% solution) and 100 µL 10× MOPS [0.02 mol/L 3-(N-morpholino)propanesulfonic acid, 0.005 mol/L sodium acetate, 0.005 mol/L ETDA, pH 7.0]. Samples were chilled on ice, and one volume of cold 20× SSC (3 mol/L NaCl, 0.3 mol/L trisodium acetate, pH 7.0) was added. The RNA was applied to a nylon Hybond N⁺ membrane in the slots of the manifold. The blot was processed as described above except that a duplicate blot also was hybridized with a radiolabeled cDNA for the 28S rat ribosomal RNA. After hybridization, rinsing and autoradiography, the slot blot and Northern blot were scanned using the LKB 2202 Ultroscan Laser Densitometer (Bromma, Sweden), and the intensities of the bands were determined. For slot blots, density of UCP mRNA bands was corrected by ribosomal RNA.

RESULTS

Table 2. Characteristics of heifers and calves fed adequate- or restricted-protein diets1 [View Table]

Birth weights of newborn calves were not affected (P > 0.10) by prepartum protein restriction (Table 2). Additionally, prepartum protein restriction did not influence calving ease scores or calf vigor scores at birth (data not shown). Thermoneutral metabolic rates were the same for calves born to heifers fed adequate and restricted protein. Consistent with the lack of treatment effect on thermoneutral metabolic rates, visceral organ weights, with the exception of lung plus trachea weights, were not affected (P > 0.10) by prepartum protein treatment (Table 3). Lung plus trachea weights were 12% less (P < 0.01) in calves born to restricted-protein heifers than in calves born to adequate-protein heifers. Peak metabolic rates also were not affected (P > 0.10) by prepartum protein treatment and on average were twice the thermoneutral metabolic rates.

Table 3. Organ weights and composition of brown adipose tissue of calves fed adequate- or restricted-protein diets1 [View Table]

Northern analysis of the RNA from the newborn and cold-adapted calves indicated a single band of identity of approximately 2 kb (Fig. 1). This is consistent with previous results of others (Brander et al. 1993). There was a dramatic loss of UCP mRNA in the 7-d postnatal calves, even though the calves were subjected to 4°C for the 7-d period. The UCP-PCR probe did not recognize any mRNA species in the longissimus dorsi muscle sample (Fig. 1). Slot blot analysis was used to quantify UCP mRNA in perirenal adipose tissue. The UCP:28S rRNA ratios for perirenal adipose tissue were not different for calves born to adequate- and restricted-protein heifers (Table 3).

The number of adipocytes per gram of tissue and mean volumes were not affected significantly by prepartum protein restriction, nor were there any differences between fat depots (Table 4). Sternum adipocytes were smaller (P < 0.01) than perirenal adipocytes. This was evident in the relative diameter and volume proportions (Fig. 3). Over 90% of the perirenal and sternum adipocytes were 20-30 µm in diameter, and there were no sternum adipocytes with diameters greater than 90 µm (Fig. 3a). There was a much broader distribution of perirenal adipocyte diameters, with a small proportion of adipocytes as large as 190 µm. Thus adipocytes with diameters of 20-50 µm made up a greater proportion of the total adipocyte volume in sternum adipose tissue than in perirenal adipose tissue (Fig. 3b).

Table 4. Cellularity and lipogenesis of brown and white adipose tissue from newborn calves born to adequate- and restricted-protein heifers1 [View Table]

DISCUSSION

Thermoneutral metabolic rates of newborn calves in the present study were not affected by prepartum protein restriction. This result was unexpected, given that the reductions in maternal body weights and body condition scores imposed by protein malnutrition were more severe than those reported by Carstens et al. (1987). The lack of a treatment affect on thermoneutral metabolic rate is supported by the findings that proportional weights of liver, heart, kidney, spleen and gastrointestinal tract of newborn calves also were unaffected by prepartum protein restriction. Koritnik et al. (1981) found that prepartum nutritional restriction of ewes sufficient to reduce birth weights by 24% caused a 15% reduction in the proportional liver weight. A decrease in the proportional mass of visceral organs due to prepartum malnutrition would be expected to reduce the thermoneutral metabolic rate, because visceral organ tissues have higher metabolic rates relative to the whole body (Drouillard et al. 1988, Koong et al. 1985).

Alexander (1978) fed high and low energy diets to pregnant ewes, beginning on d 90 of gestation, and found that prepartum energy restriction reduced the proportional weight of perirenal adipose tissue (the primary BAT depot in newborn ruminants) by 17% in single and 24% in twin fetuses at 125 d of gestation. A reduction in the quantity of fetal perirenal adipose tissue strongly suggests that nonshivering thermogenic capabilities would have been impaired by maternal malnutrition in this study, given that very little WAT is present in perirenal adipose tissue depots (Alexander et al. 1975). Furthermore, Tyzbir et al. (1984) demonstrated that prepartum protein restriction of rats reduced BAT mass 40-50% as well as BAT mitochondrial thermogenic capacity in newborn rat pups, even though birth weights were not affected by prepartum protein restriction. In contrast, NE-induced peak metabolic rate was the same in calves born to adequate- and restricted-protein heifers in the present investigation. Consistent with the lack of a treatment effect on peak metabolic rates, prepartum protein restriction did not affect perirenal adipose tissue mass or composition. Nor did prepartum protein restriction alter UCP gene expression in BAT.

The birth weights of the Wagyu × Angus calves used in the present investigation (26 kg) were considerably smaller than the birth weights of the Hereford × Angus calves (32 to 38 kg) used in previous experiments that reported reduced thermoneutral metabolic rates in response to prepartum malnutrition (Carstens et al. 1987, Ridder et al. 1991). The small birth weight may be a breed type effect; Smith et al. (1992) reported that birth weight decreases as the percentage of Wagyu increases in crossbred calves. Heat exposure during late gestation also may have contributed to reduced birth weights in the present investigation, because the calves were born during the second and third weeks of September in central Texas. Prolonged heat exposure has been shown to reduce uterine and umbilical blood flows (Reynolds et al. 1985), resulting in reduced birth weights of sheep (Bell et al. 1989) and cattle (Collier et al. 1982). In spite of differences in birth weights, breed types and methods of measurement (water immersion versus metabolic chambers), thermoneutral metabolic rates of the Wagyu × Angus calves of the present investigation were similar to rates previously observed in Hereford × Angus calves (~143 J·kg¹·min¹; Carstens et al. 1987).

A secondary goal of this investigation was to further characterize the metabolism and ultrastructure of BAT and to contrast BAT and WAT in newborn calves. The brown adipocytes from newborn calves did not display the typical multilocular feature that is characteristic of BAT in other species. Instead, bovine brown adipocytes contained a large central lipid vacuole with few peripheral lipid inclusions. This is consistent with the findings of Alexander et al. (1975), who examined perirenal brown adipocytes in newborn calves at lower magnification and described adipocytes as dominated by a large lipid vacuole with smaller lipid inclusions in the marginal cytoplasm of some of the cells. Our results support the observation of Napolitano (1963) that the major criterion used to characterize brown adipocytes morphologically should be the appearance and differentiation of mitochondria rather than occurrence of multilocular lipid droplets.

Subcutaneous adipose tissue overlying the sternum also contained unilocular adipocytes with few cytoplasmic inclusions. Unlike perirenal adipose tissue, these adipocytes contained only a small number of mitochondria with poorly developed cristae; thus, sternum adipose tissue represents a white adipose tissue depot in these calves. These results are similar to those of Alexander et al. (1975) for newborn calves, although the earlier work described adipocytes containing a few small lipid droplets in addition to the large central vacuole.

The relative cell diameter and volume proportions derived from the particle sizing method indicated that the white adipocytes were more uniform in diameter and that sternum WAT did not contain adipocytes with diameters larger than 90 µm. This is in contrast with BAT from both treatment groups, which contained adipocytes as large as 190 µm. In lambs, perirenal adipose tissue mass increases dramatically during the last 3 wk of gestation, whereas subcutaneous adipose tissue mass declines (Alexander 1978). The larger brown adipocytes, relative to white adipocytes, may reflect this same phenomenon in calves.

The fact that lipogenesis was not different between treatment groups supports the finding that the BAT lipid mass was the same in calves born to adequate- and restricted-protein heifers. The rate of lipogenesis from acetate reported here for BAT (55 nmol·10⁵ cells¹·h¹) was not different from what we have observed in perirenal adipose tissue from adult cattle (50-70 nmol·10⁵ cells¹·60 min¹; unpublished observations). Bovine perirenal BAT may be considerably more active prenatally; Vernon et al. (1981) reported that lipogenesis in ovine perirenal adipose tissue declined by nearly 98% during the last 4 wk of gestation. The rate of fatty acid synthesis from acetate in subcutaneous WAT was less than 10% of that in BAT. This may be indicative of a decline in subcutaneous adipose tissue mass during the latter phases of gestation. In contrast to perirenal BAT, subcutaneous adipose tissue becomes considerably more active postnatally as cattle achieve physiological maturity (Smith et al. 1984).

It is apparent that acetate is the primary lipogenic precursor in BAT which is consistant with results reported for WAT from adult ruminants (Hanson and Ballard 1967, Robertson et al. 1981). Robertson et al. (1981) reported that glucose contributed 17% of the acetyl units for fatty acid biosynthesis in perirenal adipose tissue from near-term fetal lambs and only 2% from adult sheep, indicating that age is important in determining carbon source for fatty acid biosynthesis. The rate of acetyl unit synthesis from glucose was 11% of that from acetate in BAT in the current study. This was considerably less than in WAT, where the rate of acetyl unit synthesis from glucose was over twice that from acetate. In BAT from newborn calves of the present investigation, and in WAT from adult steers (Smith and Crouse 1984, Smith and Prior 1986), the contribution of glucose to fatty acid biosynthesis is substantially less than that from acetate.

In summary, severe protein restriction of pregnant heifers during the last trimester dramatically affected the heifers but had no measurable effect on BAT metabolism or thermogenesis in the calves. However, we cannot rule out the possibility that larger calves from other, more typical breed types would respond differently to prepartum protein restriction. The data suggest that bovine fetuses that are small in proportion to the dams are provided a natural protection against prepartum protein malnutrition of the dams.

ACKNOWLEDGMENTS