Methods
Anesthesia and surgical preparation
After a fasting period of 4 hours, mice were anesthetized with 2% isoflurane in air followed by an intraperitoneal (i.p.) injection of ketamine, midazolam and atropine. To ensure adequate fluid resuscitation and perioperative antibiotic treatment, animals received a subcutaneous (s.c.) injection of Lactated Ringer´s solution containing glucose, ceftriaxon, and clindamycine. A laparotomy was performed, and the cecum was ligated. A single puncture using a 18-gauge needle was performed, and the cecum was lightly squeezed to expel a small amount of feces from the puncture site to ensure a full-thickness perforation. The abdomen was closed thereafter. Postoperatively, water and food were provided ad libitum. After a period of 8 hours, all mice received a second s.c. injection of fluids and antibiotics together with buprenorphine to provide adequate postoperative analgesia. 15 hours post-CLP animals were anesthetized by a second i.p. injection and placed on the procedure bench equipped with heating pads and lamps to maintain body temperature between 36.5-37.5 °C. Thereafter, they were tracheostomized to allow for endotracheal intubation and mechanical ventilation. Normoventilation was controlled by monitoring the endexpiratory CO2 concentration. A central venous line was placed in the right jugular vein, and anesthesia was maintained by continuous i.v. ketamine and fentanyl, the infusion rates being identical in the three experimental groups [1,18]. Since we had previously shown that CLP-challenged wild-type mice died from refractory hypotension with fluid resuscitation alone, continuous intravenous colloids and norepinephrine were administered simultaneously using a fixed catecholamine concentration (0.033 µg·mL-1 hetastarch) [1]. This approach allowed a sustained 50 – 100% increase in cardiac output (CO) at a mean arterial pressure (MAP) of that seen in sham-operated animals [1]. Practically, in septic animals the infusion rate of this solute was titrated to maintain the MAP > 65 mmHg throughout the entire observation period and thereby to achieve normotensive and hyperdynamic hemodynamics as indicated by a sustained increase in CO. Normoglycemia was maintained by continuous i.v. glucose (2 mg·g-1·h-1) which comprised 50% non-radioactive labelled 1,2,3,4,5,6,-13C6-glucose (Campro-Scientific GmbH (ISOTEC), Berlin, Germany).
A 1.4-F catheter with pressure and conductance sensors (SPR 864, Millar Instruments, Houston, TX, USA) was introduced into the heart via the right carotid artery. The conductance catheter was interfaced with a pressure-volume analogue signal for the assessment of systemic hemodynamics and cardiac output (CO) in mice as described previously [18]. Thereafter, perivascular flow probes were placed around the superior mesenteric artery (SMA) and the portal vein (PV) to obtain regional blood flow data using a multi-channel ultrasonic transit-time flowmetry. Microvascular perfusion, capillary hemoglobin concentration ([µHb]) and oxygen saturation (Hb-O2) were determined simultaneously using combined laser Doppler flowmetry and remission spectroscopy technique as described in detail previously [1]. Hemodynamic parameters recorded over 6 hours were mean arterial pressure (MAP), heart rate (HR), cardiac output (CO), SMA and PV as well as the microvascular parameters.
Estimation of hepatic glucose production rate
Hepatic glucose production was estimated from the levels of intrahepatic glucose isotope enrichment. Katz et al. previously showed that plasma and liver tissue isotope labelling values are close to equilibrium [56, 57]. In fact, in separate sham-operated and CLP mice simultaneous measurement of the isotope enrichment in liver tissue specimen and in consecutive blood samples taken from the tail-tip yielded virtually identical results, and the blood isotope enrichment did not show any time dependent variation (77.2, 77.4, 77.2 % tracer/tracee ratio at 2, 4 and 6 hours after the start of the glucose isotope infusion, rescpectively). These findings imply stationary conditions for the isotope labelling levels both in the intravascular and the intrahepatic compartment. Consequently, based on these results established tracer dilution concepts [23, 24] were applied on levels of the intrahepatic glucose isotope enrichment to estimate the rate of hepatic glucose production. For this purpose, glucose was isolated from the liver tissue samples and converted to a penta-(trifluoro-acetyl) derivate (wildtype controls, n=12; iNOS-/-n=8, wildtype+GW274150, n=8). Under positive chemical ionisation molecular fragments were analysed for the mass range m/z 319 – 326 using a gas chromatography-mass spectrometry system (Agilent 6890GC/5973MS, Palo Alto, CA). The mass distribution derived from these ions was separated using standard deconvolution approaches (Lee W-NP, Byerley LO, Bergner EA, Edmond J (1991) Mass isotopomer analysis: theoretical and practical considerations. BiolMass Spect 1991, 20:451-458) into a component arising from unlabelled and labelled recycling glucose, which mainly contributes to the ions at m/z 320 and m/z 321 arising from the infused tracer, and a component arising from the infused tracer. The level of the latter was used to estimate glucose production.