Oral Administration of a Novel, Synthetic Ketogenic Compound Elevates Blood β-Hydroxybutyrate Levels in Mice in Both Fasted and Fed Conditions

Abstract

Background/objectives: Elevating ketone levels with therapeutic nutritional ketosis can help to metabolically manage disease processes associated with epilepsy, diabetes, obesity, cancer, and neurodegenerative disease. Nutritional ketosis can be achieved with various dieting strategies such as the classical ketogenic diet, the modified Atkins diet, caloric restriction, periodic fasting, or the consumption of exogenous ketogenic supplements such as medium-chain triglycerides (MCTs). However, these various strategies can be unpleasant and difficult to follow, so that achieving and sustaining nutritional ketosis can be a major challenge. Thus, investigators continue to explore the science and applications of exogenous ketone supplementation as a means to further augment the therapeutic efficacy of this metabolic therapy.

Methods: Here, we describe a structurally new synthetic triglyceride, glycerol tri-acetoacetate (Gly-3AcAc), that we prepared from glycerol and an acetoacetate precursor that produces hyperketonemia in the therapeutic range (2-3 mM) when administered to mice under both fasting and non-fasting conditions. Animal studies were undertaken to evaluate the potential effects of eliciting a ketogenic response systemically. Acute effects (24 h or less) were determined in male VM/Dk mice in both fasted and unfasted dietary states.

Results: Concentration levels of β-hydroxybutyrate in blood were elevated (βHB; 2-3 mM) under both conditions. Levels of glucose were reduced only in the fasted state. No detrimental side effects were observed.

Conclusions: Pending further study, this novel compound could potentially add to the repertoire of methods for inducing therapeutic nutritional ketosis.

Keywords: diabetes; exogenous ketones; hyperketonemia; hypoglycemia; ketogenic diet; ketone metabolic therapy; ketosis; medium-chain triglycerides; nutritional supplements.

Figure 1

Examples of ketone bodies and relevant exogenous ketones: (A) the ketone bodies, β-hydroxybutyrate, acetoacetate, and acetone; (B) β-hydroxybutyrate salt of sodium; (C) 1,3-butanediol; (D) 3-hydroxy butyl 3-hydroxybutanoate ester; and (E) 1,3-butanediol diacetoacetate diester.

Figure 2

Synthesis of propane-1, 2, 3-triyl tris (3-oxobutanoate) (common name, glycerol tri-acetoacetate (1)), abbreviated as Gly-3AcAc.

Figure 3

Experiment 1. Mice were fasted from 2 h before gavage to 5 h after gavage. (A) The average of each group’s blood glucose in mg/dL at each timepoint is shown. For the overall responses, compared to the control group (0 mg/kg), the treatment groups showed significant differences as follows: at 2.5 mg/kg p < 0.01; at 5.0 mg/kg p < 0.0001; at 7.5 mg/kg p < 0.0001. (B) The average of each group’s blood BHB in mM at each timepoint is shown. Significant difference from the control group as follows: at 2.5 mg/kg p < 0.001; at 5.0 mg/kg p < 0.0001; at 7.5 mg/kg p < 0.0001. (C) The average of each group’s glucose ketone index (GKI) at each timepoint is shown. Significant difference from the control group as follows: at 2.5 mg/kg p < 0.0001; at 5.0 mg/kg p < 0.0001; at 7.5 mg/kg p < 0.0001. (D) BHB C_max, the maximum BHB levels measured for each individual test subject are shown. For all line graphs representing data, horizontal bars indicate timepoints when treatment and control measurements are significantly different from each other (p < 0.05). For all column graphs, groups of data points with shared alphabetic symbols are not significantly different from each other.

Figure 4

Experiment 2: Mice were fasted from 2 h before gavage to 24 h after gavage. Horizontal bars indicate timepoints when treatment and control measurements were significantly different from each other (p < 0.05). (A) The average of each group’s blood glucose in mg/dL at each timepoint is shown. For the overall responses, compared to the control group, the treatment group was significantly different: p < 0.05. (B) The average of each group’s blood BHB in mM at each timepoint is shown. For the overall responses, compared to the control group, the treatment group was significantly different: p < 0.05. (C) The average of each group’s glucose ketone index (GKI) at each timepoint. For the overall responses, compared to the control group, the treatment group was significantly different: p < 0.05. (D) BHB C_max, the maximum BHB levels measured for each individual test subject are shown. Groups of data points with shared alphabetic symbols are not significantly different.

Figure 5

Experiment 3: Mice were fasted from 2 h before gavage to 4 h after gavage. (A) The average of each group’s blood glucose in mg/dL at each timepoint. (B) The average of each group’s blood BHB in mM at each timepoint. (C) Serum BHB at 4 h as determined by LCMS. p < 0.005. LCMS quantifies both enantiomers of BHB. (D) Serum AcAc at 4 h as determined by LCMS. p < 0.05. Groups of data points with shared alphabetic symbols are not significantly different.

Figure 5

Experiment 3: Mice were fasted from 2 h before gavage to 4 h after gavage. (A) The average of each group’s blood glucose in mg/dL at each timepoint. (B) The average of each group’s blood BHB in mM at each timepoint. (C) Serum BHB at 4 h as determined by LCMS. p < 0.005. LCMS quantifies both enantiomers of BHB. (D) Serum AcAc at 4 h as determined by LCMS. p < 0.05. Groups of data points with shared alphabetic symbols are not significantly different.

Figure 6

Experiment 4: Food restriction was not implemented. (A) The average of each group’s blood glucose in mg/dL at each timepoint. (B) The average of each group’s blood BHB in mM at each timepoint. (C) The average of each group’s glucose ketone index (GKI) at each timepoint. (D) BHB C_max, the maximum BHB levels measured for each individual test subject are shown, p = 0.016. Groups of data points with shared alphabetic symbols are not significantly different. For all line graphs representing data, horizontal bars or asterisks indicate timepoints when treatment and control measurements are significantly different from each other (p < 0.05).