Duodenal PKC-δ and Cholecystokinin Signaling Axis Regulates Glucose Production

The protocols were approved by the Institutional Animal Care Committee of the University Health Network. Male SD rats (250–300 g) were housed in individual cages and maintained on a standard light-dark cycle. The rats had access to chow and water ad libitum and were given 7 days to acclimatize before the designated surgeries were performed.

Duodenal cannulation was performed where a catheter was inserted into the proximal duodenum (∼1.5–2 cm downstream of the pyloric sphincter) (1,2). Intravenous catheters were placed into the internal jugular vein and carotid artery for infusion and blood sampling, respectively (1,2). Rats were monitored for daily food intake and body weight for 3 to 4 days after surgery and only underwent clamp studies when food intake and body weight were recovered back to at least 90% of basal values. A subgroup of rats with duodenal and intravenous catheters implanted were placed on a lard-oil enriched high-fat diet for 3 days (Purina Mills [TestDiet]; No. 0009389) that induces duodenal lipid/CCK resistance (1,2).

The clamp was performed as described (1,2) (Fig. 1B). The night before, rats were restricted to ∼58 kcal of food to ensure the same nutritional postabsorptive status. At t = 0 min, a primed continuous infusion of [3-³H]glucose (bolus 40 μCi; 0.4 μCi/min) was given, and this rate was maintained throughout to assess changes in steady-state glucose kinetics based on tracer-dilution methodology. The clamp was started at t = 90 min whereby a continuous infusion of insulin at 1.2 mU/kg/min and somatostatin at 3 μg/kg/min were initiated. Somatostatin suppressed the plasma C-peptide level by >80% (basal: 0.63 ± 0.05 nmol/L vs. clamp: 0.12 ± 0.03; n = 14/group). A 25% glucose solution was infused at a variable rate to maintain euglycemia. The glucose infusion was adjusted every 10 min from t = 120–200 min. Intraduodenal infusions were initiated at t = 150 min and maintained until t = 200 min at 0.01 mL/min. Blood samples were taken at t = 90, 180, and 200 min to determine plasma insulin and C-peptide levels. Of note, during the clamps, plasma insulin levels (saline: 0.7 ± 0.2; 1-oleoyl-2-acetyl-sn-glycerol [OAG]: 0.7 ± 0.1; saline ± MK-329: 0.7 ± 0.1; OAG ± MK-329: 0.6 ± 0.1) were comparable among individual groups.

The following reagents were administered intraduodenally: 1) saline; 2) OAG–PKC activator (Calbiochem; 250 μmol/L); 3) CCK-8 (sulfated)–CCK receptor agonist (Sigma-Aldrich; 35 pmol/kg/min); 4) MK-329–CCK-A receptor antagonist (Tocris Bioscience; 3.2 μg/kg/min); 5) CR-1409–CCK-A receptor antagonist (Sigma-Aldrich; 29 μg/kg/min), and 6) rottlerin–PKC-δ specific inhibitor (Calbiochem; 60 μmol/L).

Adenovirus (Ad) application into the duodenum was performed as described (6). Ad expressing LacZ or the dominant negative form (DN) of PKC-δ (pfu/mL = 4 × 10⁸; 0.2 mL; gift from Dr. J. Soh, Biomedical Research Centre for Signal Transduction Network, Incheon, Korea) was injected into the duodenal lumen. A duodenal catheter was subsequently inserted through the injection site, followed by vascular cannulation using the same procedures described earlier. Animals were allowed to recover for 3 days before the clamp procedures. We have validated previously that duodenal injection of Ad expressing LacZ selectively infected intestinal cells, but not hepatocytes, to a detectable level (3).

Duodenal tissues were collected for immunofluorescence staining after animals were killed. After sectioning, the sections were postfixed with 4% PFA and Tween-20 (0.05%) and were blocked with 10% goat serum in phosphate-buffered saline for 1 h. To check for PKC-δ and CCK immunoreactivity, the sections were incubated with mouse anti–PKC-δ antibody (1:100) and rabbit anti–CCK-8 antibody (1:200) overnight at 4°C, respectively. The sections were incubated with goat anti-mouse antibody (1:500) and anti-rabbit antibody (1:200) for 1 h at room temperature. Immunolabeling was detected with the aid of a fluorescence microscope.

Plasma glucose concentrations were measured using the glucose oxidase method (Glucose analyzer GM9; Analox Instruments). Plasma insulin and C-peptide concentrations were measured using RIA (Linco Research).

Statistical analysis was performed using ANOVA followed by Tukey t test post hoc analysis. Significance was accepted at P < 0.05. Data are presented as means + SEM. The time period 60–90 min was averaged for the basal condition, and the time period 180–200 min was averaged for the clamp condition.

Oleate activates PKC and stimulates CCK release within the same CCK-secreting cells (8,9). If duodenal CCK signaling is necessary for PKC-δ activation to regulate glucose, PKC-δ should be colocalized with CCK in the duodenal mucosa. Immunostaining of CCK indicated that CCK was expressed along the duodenal mucosa (Fig. 1C) where the I cells that release CCK (13) are found. PKC-δ was expressed around the same area that contained CCK (Fig. 1C), and PKC-δ appeared to colocalize with CCK (Fig. 1C), suggesting a mechanistic link between PKC-δ and CCK.

To evaluate whether duodenal CCK signaling is required for duodenal mucosal PKC-δ activation to regulate glucose kinetics, we activated duodenal mucosal PKC-δ and inhibited the duodenal CCK-A receptor via intraduodenal coinfusion of OAG with either MK-329 or CR-1409 in normal rats during the clamps (Fig. 1A and B). When intraduodenal OAG was infused at a dose that activated duodenal mucosal PKC-δ activity (6), the glucose infusion rate required to maintain euglycemia was increased (Fig. 1D) as a result of an inhibition of glucose production (Fig. 1E and F). Coinfusion of either intraduodenal MK-329 or CR-1409 with OAG fully negated the ability of OAG to increase glucose infusion rate (Fig. 1D) and lower glucose production (Fig. 1E and F), whereas duodenal MK-329 or CR-1409 alone had no effects (Fig. 1D–F). Glucose uptake was comparable in all groups (Fig. 1G). These data indicate that activation of the duodenal CCK-A receptor is required for PKC-δ activation to lower glucose production, suggesting that PKC-δ activation lies upstream of CCK signaling.

It is equally possible that duodenal PKC-δ signaling lies downstream of CCK (Supplementary Fig. 1). We first tested the metabolic effects of duodenal CCK signaling (Supplementary Fig. 1), and consistent with a previous report (2), intraduodenal CCK-8 infusion increased the glucose infusion rate (Fig. 2A) and lowered glucose production (Fig. 2B and C) but did not alter glucose uptake (Fig. 2D). The intraduodenal PKC-δ inhibitor rottlerin infused at a dose that blocked the metabolic effects of OAG (6) failed to negate the ability of intraduodenal CCK-8 to regulate glucose kinetics (Fig. 2A–D).

To alternatively exclude the downstream signaling possibility of duodenal PKC-δ to CCK, duodenal mucosal PKC-δ activity was knocked down with a molecular approach (6) (Supplementary Fig. 1). We injected the Ad expressing LacZ (control) or the DN of PKC-δ directly into the lumen of the duodenum. Duodenal mucosal PKC-δ activity was knocked down by ∼60% with intestinal Ad DN PKC-δ, and such inhibition of PKC-δ negated the ability of duodenal OAG to regulate glucose production (6). Intraduodenal CCK-8 was equally potent to increase the glucose infusion rate (Fig. 2E) and lower glucose production (Fig. 2F and G) in duodenal Ad LacZ–injected rats as compared with normal rats (Fig. 2A–C). The metabolic effects induced by intraduodenal CCK-8 preserved in rats injected with duodenal Ad DN PKC-δ (Fig. 2E–G). Glucose uptake was comparable in all groups (Fig. 2H). These data together indicate that duodenal PKC-δ signaling is not required for (or not downstream of) CCK to regulate glucose production.

Given that selective activation of duodenal mucosal PKC-δ enhances CCK signaling to regulate glucose production in normal rats, we tested whether activation of duodenal PKC-δ is sufficient to enhance CCK signaling to overcome CCK resistance in response to high-fat feeding (2). Intraduodenal OAG increased the glucose infusion rate (Fig. 3A) and lowered glucose production (Fig. 3B and C) independent of changes in glucose uptake (Fig. 3D) in normal rats. In contrast, intraduodenal OAG failed to alter glucose kinetics (Fig. 3A–C) in rats fed with a high-fat diet. Thus, direct activation of duodenal PKC-δ fails to overcome CCK resistance in high fat-fed rats.

Lipid metabolism within the duodenum triggers a neuronal network to lower glucose production in normal but not in high fat-fed rats (1). Given that diabetes and obesity are characterized by an excessive rate of glucose production (14,15), studies aimed to dissect the underlying mechanisms of duodenal lipid sensing could reveal novel therapeutic molecules that lower glucose production in diabetes and obesity (16).

We here demonstrate that activation of the duodenal CCK-A receptor is necessary to mediate the metabolic effects of duodenal mucosal PKC-δ activation, thereby revealing a duodenal PKC-δ and CCK signaling axis that lowers glucose production in the presence of circulating basal insulin levels. Because activation of duodenal mucosal PKC-δ (6) or CCK-A receptors (2) is sufficient and necessary for lipid sensing to trigger a gut–brain–liver axis to lower glucose production in normal rats through the hepatic vagus, our data collectively indicate that metabolism of lipids within the duodenum activates mucosal PKC-δ and stimulates the release of CCK from the mucosal I cells, leading to the activation of CCK-A receptors (i.e., lipids→PKC-δ→CCK→CCK-A receptors) and lowering of glucose production. The mechanistic link(s) of the ability of PKC-δ to stimulate CCK release remains unknown. However, the potential involvement of the soluble NSF attachment protein receptor proteins (i.e., Munc18–1 and vesicle-associated membrane protein-2) warrants future investigation since PKC-δ enhances insulin secretion in association with increased phosphorylation of Munc18–1 in pancreatic β-cells (17) and vesicle-associated membrane protein-2 mediates CCK secretion in secreting tumor cell-1 cell lines (18).

Although the physiological relevance of the ability of the duodenal PKC-δ→CCK signaling axis to regulate glucose is supported by the fact that direct inhibition of either duodenal PKC-δ (6) or CCK-A receptors (2) disrupts glucose homeostasis during refeeding, the nutritional value of this newly described signaling axis remains to be explored. Experiments aimed to assess whether different types of long-chain fatty acid such as oleate or palmitate (via intraduodenal infusion or oral ingestion) trigger the PKC-δ→CCK signaling axis and regulate glucose production would begin to address the nutritional, physiological, and beneficial impact of this signaling pathway.

Rats acquired duodenal CCK resistance in response to high-fat feeding (2). Although the site(s) of the defect is unclear, evidence suggests that it lies within the signaling cascade of the duodenal CCK-A receptor since direct administration of the receptor agonist CCK-8 into the duodenum fails to lower glucose production in high fat-fed rats (2). Consistent with the fact that duodenal PKC-δ activation is upstream of CCK signaling and that diet-induced CCK resistance is postulated to lie within the downstream signaling cascade of CCK-A receptors, direct activation of duodenal mucosal PKC-δ still fails to overcome intestinal CCK resistance to lower glucose production. These findings strengthen the argument that duodenal lipid resistance is not a result of the inability of lipid to trigger signaling events (i.e., PKC-δ activation) to stimulate the local release of CCK from the mucosa, but rather, the downstream signaling cascade of CCK/CCK-A receptor.

Of note, although activation of duodenal PKC-δ is not required for CCK (i.e., CCK→CCK-A receptors→PKC-δ) to regulate glucose production, our present data do not rule out the possibility that the CCK-A receptor may activate other isoforms of PKC via phospholipase C to regulate glucose production, which warrants future investigation. In summary, the current set of data unveils a novel duodenal PKC-δ and CCK signaling axis that regulates glucose production in vivo.

D.M.B. and J.T.Y.Y. were supported by postdoctoral fellowships and B.A.R. by a graduate scholarship from the Banting and Best Diabetes Centre. A.K. and G.W.C.C. were supported by the Canadian Institutes of Health Research and the Banting and Best Diabetes Centre graduate scholarships. This work was supported by a research grant to T.K.T.L. from the Canadian Institutes of Health Research (MOP-82701). T.K.T.L. holds the J.K. McIvor Endowed Chair in Diabetes Research and the Canada Research Chair in Obesity at Toronto General Research Institute (TGRI) and the University of Toronto.

No potential conflicts of interest relevant to this article were reported.

D.M.B. conducted and designed the experiments, performed data analyses, and wrote the manuscript. J.T.Y.Y., B.A.R., A.K., and G.W.C.C. assisted with experiments. T.K.T.L. supervised the project, designed the experiments, and edited the manuscript.

The authors are grateful to P.Y.T. Wang from the TGRI, University Health Network, for excellent technical assistance.