Exogenous GH is used extensively in the USA to stimulate milk production in dairy cattle but its effectiveness is reduced in undernourished animals. It has been proposed that GH increases milk yield by stimulating IGF-I secretion and that this IGF-I-response is nutritionally sensitive and thus acts as a 'sensor' of energy balance. To investigate this possibility, we placed lactating rats on three planes of nutrition, ad libitum, 50% or 25% of ad libitum for 48 h. Subgroups of these animals were treated for 48 h with bromocriptine, to suppress prolactin secretion, and anti-rat GH, to neutralize GH action. From 24 to 48 h some of the treated animals were assessed for their milk yield response to prolactin or GH. Food restriction reduced milk yield in control rats by approximately 50% and was accompanied by a catabolic state, as judged by lipid mobilization from adipose tissue and by low concentrations of serum insulin, IGF-I, triiodothyronine and thyroxine, and increased serum nonesterified fatty acid concentrations. In animals fed ad libitum, anti-rat GH plus bromocriptine treatment produced an 80% decrease in milk yield and a dramatic fall in the activity of acetyl-CoA carboxylase in mammary tissue. GH was able to stimulate milk yield when given from 24 to 48 h; however, its effectiveness decreased progressively as food intake was reduced. The milk yield response to GH was accompanied by an increase in serum IGF-I concentrations and this response also decreased progressively with reduction of food intake, consistent with the hypothesis that IGF-I determines the milk yield response to GH and thus regulates GH action on the mammary gland in a nutritionally dependent fashion. However, the milk yield response to prolactin and the milk yield of control rats decreased in line with food intake without any changes in serum IGF-I concentrations. This clearly indicates that factors other than IGF-I are responsible for restricting milk yield. In order to assess other possible candidates for this role, we monitored serum glucose, non-esterified fatty acids, insulin triiodothyronine and thyroxine concentrations, but found no evidence for any simple relationship between these parameters and the milk yield response to prolactin and GH. Surprisingly we found that the ability of GH or prolactin to prevent epithelial cell loss in in the mammary gland was completely insensitive to nutrient intake, despite the fact that IGF-I is considered to be an important survival factor for mammary epithelial cells. Finally, we also demonstrated that, at least during short-term food restriction, the lactating rat is capable of mobilizing significant amounts of lipid from adipose tissue, such that it could provide the total output of triglyceride in milk, which is much greater than has previously been proposed.

The factors regulating serum leptin concentration and its relationship to the hyperphagia of lactation have been investigated in rats. Lactation results in hypoleptinaemia and loss, or at least marked attenuation, of the nocturnal rise in serum leptin. Litter removal resulted in a fall in food intake and restoration of the nocturnal rise in serum leptin. Returning the litter to the mother after a 48-h absence increased food intake and began to reinitiate milk production, but the nocturnal serum leptin levels were still increased at 48 h after litter restoration. Adjusting litter size to four, eight, ten or fourteen pups at parturition resulted in different rates of litter growth and food intake during the subsequent lactation, but had no effect on the degree of hypoleptinaemia. Reducing litter size from ten to four pups at mid-lactation resulted in a transient increase in both serum leptin and pup growth rate, while food intake fell to a level found in rats suckling four pups throughout lactation. Reducing milk production by injection of bromocriptine increased serum leptin, but did not restore the nocturnal rise in serum leptin; food intake decreased, but remained much higher than in non-lactating rats. Feeding a varied, high-energy diet resulted in a decrease in the weight of food ingested, but no change in calorie intake, and had no effect on the hypoleptinaemia. These studies suggested that the hypoleptinaemia of lactating rats is due to negative energy balance, but the loss of the nocturnal rise in serum leptin is due to the suckling stimulus. The negative energy balance of lactation does not appear to be caused by a physical constraint on food intake. While the hypoleptinaemia should facilitate the hyperphagia of lactation, other orexigenic signals must also be involved.

We investigated the effects of lactation on diurnal changes in serum leptin and hypothalamic expression of the leptin receptor isoforms, Ob-Ra, -Rb, -Rc, -Re and -Rf in rats. In non-lactating rats, serum leptin concentration was increased at night while hypothalamic mRNA levels of Ob-Rb, -Rc and -Re decreased; by contrast, expression of Ob-Ra and Ob-Rf was unchanged at night. There were significant negative correlations between serum leptin and mRNA expression of Ob-Rb (P<0.001) and Ob-Re (P<0.05), which were independent of time of day. In lactating rats, the nocturnal rise in serum leptin was attenuated. Daytime hypothalamic Ob-Rb mRNA levels were significantly lower than in non-lactating controls, and the normal nocturnal decreases in expression of Ob-Rb, -Rc and -Re were lost. The relationship between serum leptin and Ob-Re expression was not changed by lactation. Lactation had no effect on the expression of Ob-Ra mRNA in the hypothalamus. Decreased daytime Ob-Rb expression could lead to reduced hypothalamic sensitivity to leptin, and thus contribute to increased daytime appetite in lactating rats. Moreover, maintaining high levels of Ob-Re expression could, by increasing hypothalamic leptin-binding protein concentration and reducing local leptin bioavailability, further accentuate hyperphagia. Thus, selective changes in expression of specific isoforms of the leptin receptor may contribute to the hyperphagia of lactation in rats.

Chronic exposure of sheep adipose tissue to growth hormone (GH) in vitro decreases the ability of the adenosine analogue, N6-phenylisopropyladenosine (PIA), to inhibit isoprenaline-stimulated lipolysis by a mechanism which is dependent on both gene transcription and protein serine/threonine phosphorylation. The inhibition is not due to a change in ligand binding to the adenosine receptor, the amounts of the three isoforms of the inhibitory GTP-binding protein, Gi, or the maximum (forskolin-stimulated) adenylate cyclase activity. The ability of GH to modulate the PIA-activated adenosine receptor to stimulate dissociation of heterotrimeric Gi was assessed by measurement of pertussis toxin-catalysed ADP-ribosylation of Gi; GH does not appear to alter the interaction between the activated receptor and Gi. The ability of GH to alter the ability of activated Gi to inhibit adenylate cyclase activity was assessed by measuring the ability of a GTP analogue, guanosine 5'-[beta gamma-imido]triphosphate (p[NH]ppG), to inhibit forskolin-stimulated adenylate cyclase activity; chronic exposure to GH prevented this effect of p[NH]ppG. Thus the attenuation of the inhibition of lipolysis by PIA by chronic exposure of adipocytes to GH appears to be due to an impairment in the interaction between adenylate cyclase and the alpha subunit of one or more isoforms of Gi.

Factors regulating the differentiation of sheep subcutaneous and abdominal (omental or perirenal) preadipocytes from foetal lambs, suckling lambs and fattening sheep have been investigated using a serum-free cell culture system. Differentiation was assessed by changes in the activity of the enzyme glycerol 3-phosphate dehydrogenase. Insulin or IGF-I was essential for differentiation. Dexamethasone, a lipid supplement (Excyte) and the thiazolidinedione, rosiglitazone (BRL 49653) (a peroxisome proliferator-activated receptor-gamma (PPAR-gamma) agonist) all enhanced preadipocyte differentiation, whereas fibroblast growth factor and GH were inhibitory. The most effective combination was insulin, triiodothyronine, dexamethasone and rosiglitazone. Under suboptimal conditions, preadipocytes from fattening sheep differentiated less well than cells from suckling and foetal lambs. Also, under suboptimal conditions, abdominal preadipocytes did not differentiate as well as subcutaneous preadipocytes. However, age and depot differences were minimised when cells were cultured with insulin, triiodothyronine, rosiglitazone and either dexamethasone or lipid. The results suggest that variation in the ability to produce the natural ligand for PPAR-gamma contributes to depot- and age-specific differences in the ability of preadipocytes to differentiate.