Off-target depletion of plasma tryptophan by allosteric inhibitors of BCKDK

Abstract

The activation of branched chain amino acid (BCAA) catabolism has garnered interest as a potential therapeutic approach to improve insulin sensitivity, enhance recovery from heart failure, and blunt tumor growth. Evidence for this interest relies in part on BT2, a small molecule that promotes BCAA oxidation and is protective in mouse models of these pathologies. BT2 and other analogs allosterically inhibit branched chain ketoacid dehydrogenase kinase (BCKDK) to promote BCAA oxidation, which is presumed to underlie the salutary effects of BT2. Potential "off-target" effects of BT2 have not been considered, however. We therefore tested for metabolic off-target effects of BT2 in Bckdk^-/- animals. As expected, BT2 failed to activate BCAA oxidation in these animals. Surprisingly, however, BT2 strongly reduced plasma tryptophan levels and promoted catabolism of tryptophan to kynurenine in both control and Bckdk^-/- mice. Mechanistic studies revealed that none of the principal tryptophan catabolic or kynurenine-producing/consuming enzymes (TDO, IDO1, IDO2, or KATs) were required for BT2-mediated lowering of plasma tryptophan. Instead, using equilibrium dialysis assays and mice lacking albumin, we show that BT2 avidly binds plasma albumin and displaces tryptophan, releasing it for catabolism. These data confirm that BT2 activates BCAA oxidation via inhibition of BCKDK but also reveal a robust off-target effect on tryptophan metabolism via displacement from serum albumin. The data highlight a potential confounding effect for pharmaceutical compounds that compete for binding with albumin-bound tryptophan.

Keywords: Albumin; BT2; Branched chain amino acid; Branched chain ketoacid dehydrogenase kinase; Child-centered care; Tryptophan; Tryptophan 2,3-dioxygenase.

Figure 1

Inhibitors of BCKDK activate tryptophan catabolism in a BCKDK-independent manner. (A) BT2 or sodium phenylbutyrate (NaPB) inhibit BCKDK to activate BCAA oxidation. (B) BT2 (40 mg/kg) was injected through a jugular vein catheter of wild-type C57Bl/6 mice, and arterial plasma was sampled via a catheter in the carotid artery. Metabolite levels measured by LC-MS were normalized to the plasma sample collected before BT2 injection (n = 3). Kynurenic acid was not detected in the 0-minute samples, so its value was imputed from the limit of detection in other experiments. (C) Whole-body Bckdk^−/− knockout (KO) mice were fasted 5 h then given BT2 (40 mg/kg, i.p.) or saline, and plasma metabolites were measured after 30 min by LC-MS (n = 4). Red and green highlighted metabolites are BCAA- or tryptophan-related, respectively. The fill color indicates the fold difference (log2) between the abundance of the compound in each sample and the mean of vehicle-treated samples for each genotype. (D) Acute oral gavage of BT2 (40 mg/kg) reduced plasma tryptophan to a similar extent in both control (n = 5) and Bckdk KO mice (n = 7) (time effect p = 0.0278 and genotype effect p = 0.0307 in mixed-effects model). The values are normalized to the average in WT littermate control mice at baseline. (E) NaPB injection (200 mg/kg, i.p.) also depleted plasma tryptophan in both WT (n = 8) and KO mice (n = 5). The values are normalized to the corresponding averages at baseline. BCKDK = branched chain ketoacid dehydrogenase kinase; BCKDH = branched chain ketoacid dehydrogenase; BCAA = branched chain amino acid; BCKA = branched chain ketoacid; TCA = tricarboxylic acid cycle; α-KMV = alpha-ketomethylvalerate.

Figure 2

BT2 lowers plasma tryptophan independent of rate-limiting tryptophan catabolic enzymes. (A) Tryptophan catabolic pathway to kynurenic acid. (B) Acute gavage with BT2 of Tdo2^−/− mice (n = 5–11). (C) Immunoblot of liver protein 90 min after 40 mg/kg gavage of BT2 in Tdo2^−/− and littermate controls demonstrates TDO is not required for inhibition of BCKDH phosphorylation by BT2. (D) Acute gavage with BT2 treatment of Ido1^−/−, Ido2^−/− and Ido1/2 double knockout mice lowered plasma tryptophan (n = 4–8). (E) The kynurenine aminotransferases Kyat1, Aadat, and Kyat3 were simultaneously deleted in the liver (triple KO) by CRISPR/Cas9-mediated knockout. (F) Reduced liver mRNA abundance in triple KO animals. (G) Efficient BT2-induced decrease in plasma tryptophan in triple KO animals. Low dose = 1.5E12 genome copies (gc) and high dose = 3E12 gc; n = 3–4. Letters indicate statistically significant differences by one-way ANOVA and Tukey's multiple comparisons test (ie. a and b are statistically significantly different with p < 0.05).

Figure 3

BT2 displaces tryptophan from serum albumin in vivo. (A) Equilibrium dialysis of serum can identify metabolites that BT2 displaces from serum proteins. After dialysis, the ratio of metabolite concentration in the serum compared to that in PBS indicates the retention of the metabolite in the serum chamber due to protein binding. (B) BT2 binds serum proteins avidly. (C) Tryptophan binds serum proteins and is displaced by BT2 at 0.3 mM or higher concentrations. (D) Oleic acid binds to serum protein, and is not displaced by BT2. (E) Lactate does not bind to serum protein (n = 3). (F) NaPB and naproxen also displace tryptophan from serum protein. (G) Acute oral gavage of naproxen (100 mg/kg) reduced plasma tryptophan in WT mice (n = 4). The values are normalized to the averages at baseline. (H) Equilibrium dialysis of serum collected 90 min after gavage of 40 mg/kg BT2 or vehicle from Tdo2 KO and littermate control mice reveals BT2 displaces tryptophan but not oleic acid from serum protein. (I) BT2 is not retained in the plasma, and does not lower tryptophan levels, in albumin KO mice: mice were administered 40 mg/kg BT2 by oral gavage, followed by plasma tryptophan and BT2 measurements at the indicated times (n = 5–6). (J) Tryptophan increases Tdo2 protein expression in mouse primary hepatocytes, but BT2 does not. (K) in silico modeling of binding of tryptophan (TRP, left) and BT2 (right) to human albumin (using PDB 1AO6) [62]. (L) Illustration of BT2 displacing tryptophan from albumin, followed by free tryptophan stabilization of TDO2 and subsequent clearance. In the absence of TDO2, tryptophan is likely cleared by IDO (not pictured).