Small Molecule Inhibitors of TET Dioxygenases: Bobcat339 Activity Is Mediated by Contaminating Copper(II)

Abstract

Ten eleven translocation (TET) dioxygenases 1-3 are non-heme Fe(II) and α-ketoglutarate dependent enzymes that catalyze oxidation of 5-methylcytosine (5mC) in DNA to hydroxymethyl-C, formyl-C, and carboxy-C. This typically leads to gene activation and epigenetic remodeling. Most known inhibitors of TET are α-ketoglutarate mimics that may interfere with other α-ketoglutarate dependent enzymes. Recently, a novel cytosine-based inhibitor of TET, Bobcat339, was reported to have mid-μM inhibitory activity against TET1 and TET2. The molecule is now sold as a TET inhibitor by several vendors. We independently prepared Bobcat339 in our laboratory and observed that it had minimal inhibitory activity against human TET1 and TET2 via a quantitative LC-ESI-MS/MS assay. Furthermore, the inhibitory activity of commercial Bobcat339 preparations was directly correlated with Cu(II) content. We therefore conclude that Bobcat339 alone is not capable of inhibiting TET enzymes at the reported concentrations, and that its activity is enhanced by contaminating Cu(II).

Scheme 1. TET-Mediated Conversion of 5mC to 5hmC, 5fC, and 5caC

Figure created with BioRender.com.

Scheme 2. Catalytic Cycle of TET Dioxygenases

(a) Decarboxylation of the α-ketoglutarate cofactor, (b) heterolysis of the dioxygen bridge, (c) hydrogen abstraction, (d) OH- rebound, and (e) expulsion of succinate and regeneration of the active site.

Figure 1

Activity of in-house synthesized Bobcat339 against TET1. (A) Schematic overview of the quantitative HPLC-ESI-MS/MS assay of TET activity. (B) Structure of Bobcat339. (C) Percent conversion of 5mC-containing DNA duplexes to 5hmC and 5fC by TET1 in the presence or absence of Bobcat339. 5mC containing DNA oligomers were incubated with recombinant human TET1 catalytic domain and α-ketoglutarate cofactor. The formation of 5hmC and 5fC was quantified by isotope dilution HPLC-ESI-MS/MS at multiple time points (5, 20, 60 min.). Data shown represent experimental triplicates ± SD.

Figure 2

Normalized activity of TET1 and TET2 in the presence of Bobcat339 preparations synthesized in our laboratory or purchased from different vendors. 5mC containing DNA was incubated with the TET1 or TET2 catalytic domain for 30 min in the presence or in the absence of Bobcat339 (125 μM) in DMSO. The formation of hmC was detected by a quantitative LC-MS assay and is displayed as % activity relative to DMSO controls All data are averaged experimental triplicates normalized to DMSO ± SD (exception to House + Cu(II) for TET2 where n = 2). Statistical analysis was performed by 2-way ANOVA with multiple comparisons and significance reported in relation to the DMSO control (***P < 0.001).

Figure 3

Titration of Cu(II) reveals limited effect of Bobcat339 on Cu(II) inhibition of TET1. (A) Inhibition of hmC production by TET1 catalytic domain in the presence of of Bobcat339 from MedKoo, Bobcat339 made in house, and Cu(II) alone. Inhibitory activity was assessed by ESI-LCMS/MS quantitation of 5hmC formed upon incubation of 5meC containing DNA duplex with TET1. Experimental duplicates were averaged, normalized to DMSO control, and fit with a non-linear curve. (B) Dose-dependence of Cu(II) inhibition of TET1 in the presence or absence of 125 μM Bobcat339. Activity was assessed by HPLC-ESI-MS/MS quantitation of 5hmC formed. Data are normalized to DMSO control and shown as average ± SD. Statistical analysis was performed by 2-way ANOVA with multiple comparisons and significance reported between the Cu(II) doses with and without Bobcat339 (*P < 0.05, **P < 0.01).

Figure 4

Effects of Bobcat339 on global hmC levels in DNA of cultured human cells. (A) 5hmC quantification in genomic DNA of Hep3B cells via dot blotting after treatment with Bobcat339 sourced from MedKoo or Sigma Aldrich (48 hours at 50 μM). (B) Quantified hmC blotting intensity via densitometry is normalized to maximal intensity. Statistical comparison was performed with one-way ANOVA (*P < 0.05). (C) HPLC-ESI-MS/MS quantitation of 5hmC from the same DNA samples confirms the trends observed via dot blotting. Data are averaged experimental triplicates normalized to DMSO ± SD. Statistical analysis was performed by one-way ANOVA (**P < 0.01).