Late-lactation Holstein cows (n=9/treatment) were used to evaluate effects of continuous adipose tissue TNFα administration on glucose and fatty acid (FA) metabolism. Cows were blocked by feed intake and milk yield and randomly assigned within block to control or TNFα treatments. Treatments (4 mL saline or 14 μg/kg TNFα in 4 mL saline) were infused continuously over 7 d via 2 osmotic pumps in the adipose layer in the tailhead region. Plasma, milk samples, milk yield, and dry matter intake (DMI) data were collected daily. On d 7, pumps were removed and liver and contralateral tailhead adipose biopsies were collected. Results were modeled with fixed effect of treatment and random effect of block; P values > 0.10 were considered non-significant. TNFα did not alter liver TNFα mRNA abundance, plasma TNFα, IL-4, IL-6, or interferon-γ concentrations, DMI, or rectal temperature. Milk fat and lactose concentrations decreased with TNFα (P < 0.05), but milk yield was unchanged and treatments did not alter the proportion of short vs. long-chain FA in milk on d 7. Treatments did not alter plasma NEFA concentration, liver triglyceride content, or adipose mRNA abundance for hormone-sensitive lipase or perilipin. Plasma glucose turnover rate, as measured by disappearance of U-13C-glucose bolus, was not altered by treatment, nor was liver mRNA abundance for phosphoenolpyruvate carboxykinase or pyruvate carboxylase. However, TNFα tended to decrease adipose TNFα mRNA abundance (P=0.09) and increase liver IL-10 mRNA abundance (P=0.05) compared to controls. Messenger RNA expression of IL-10 in adipose and IL-37 in liver tissue increased significantly in cows treated with TNFα (Figure 1; P = .02 adipose; P < 0.05 liver). This TNFα delivery protocol may have allowed for an adaptive anti-inflammatory response to suppress systemic inflammation, which may account for the lack of metabolic responses compared with previous responses to daily subcutaneous TNFα injections.
