Biochemical Characterization of AP Lyase and m6A Demethylase Activities of Human AlkB Homologue 1 (ALKBH1)

Abstract

Alkbh1 is one of nine mammalian homologues of Escherichia coli AlkB, a 2-oxoglutarate-dependent dioxygenase that catalyzes direct DNA repair by removing alkyl lesions from DNA. Six distinct enzymatic activities have been reported for Alkbh1, including hydroxylation of variously methylated DNA, mRNA, tRNA, or histone substrates along with the cleavage of DNA at apurinic/apyrimidinic (AP) sites followed by covalent attachment to the 5'-product. The studies described here extend the biochemical characterization of two of these enzymatic activities using human ALKBH1: the AP lyase and 6-methyl adenine DNA demethylase activities. The steady-state and single-turnover kinetic parameters for ALKBH1 cleavage of AP sites in DNA were determined and shown to be comparable to those of other AP lyases. The α,β-unsaturated aldehyde of the 5'-product arising from DNA cleavage reacts predominantly with C129 of ALKBH1, but secondary sites also generate covalent adducts. The 6-methyl adenine demethylase activity was examined with a newly developed assay using a methylation-sensitive restriction endonuclease, and the enzymatic rate was found to be very low. Indeed, the demethylase activity was less than half that of the AP lyase activity when ALKBH1 samples were assayed using identical buffer conditions. The two enzymatic activities were examined using a series of site-directed variant proteins, revealing the presence of distinct but partially overlapping active sites for the two reactions. We postulate that the very low 6-methyl adenine oxygenase activity associated with ALKBH1 is unlikely to represent the major function of the enzyme in the cell, while the cellular role of the lyase activity (including its subsequent covalent attachment to DNA) remains uncertain.

Figure 1

Kinetic characterization of AP lyase activity of ALKBH1 on the basis of 3′-product release. (A) Determination of turnover rates under steady-state conditions. ALKBH1 was incubated at the indicated concentrations with 1 μM 3′-FAM-labeled oligo_AP1 in the presence of 1 mM βME at pH 8 and 37 °C, and product formation was measured at the different time points. We analyzed three to five independent experiments by ImageJ with the averages calculated and plotted. (B) Determination of active protein fraction. The ALKBH1 concentrations used in the assays were plotted against the intercepts of the lines shown in panel A. The resulting slope of 22.0±1.4% represents the active protein concentration. (C) Determination of steady-state turnover rate. We plotted the active ALKBH1 concentrations in the assays against the slopes of the lines shown in panel A. The resulting slope represents the turnover rate of 0.028±0.003 min⁻¹. (D) Determination of product formation under single-turnover conditions. ALKBH1 (15 μM) was incubated with one μM of either ssDNA substrate (oligo_AP1, shown in green) or dsDNA containing one AP site (oligo_AP1 annealed to oligoAP_2, in blue) or two AP sites (oligoAP_1 annealed to oligo_AP3, in black) at pH 8 and 37 °C for the times indicated. The data were fit as described in the Experimental Procedures to estimate the k_obs associated with the chemical step for DNA cleavage by ALKBH1.

Figure 2

The 5′-product released by ALKBH1’s AP lyase activity increases in the presence of βME. (A) Representative PAGE gels showing 5′-product formation in the presence or absence of βME. ALKBH1 (0.2 μM) was incubated with 5′-FAM-labeled oligo_AP1 (1 μM) with or without βME (1 mM) for the indicated time points and analyzed by PAGE. The control samples lacked enzyme to monitor unspecific cleavage. (B) Analysis of 3′-product formation over time in the presence and absence of βME. Three to five independent experiments were analyzed by ImageJ, the amount of product formed relative to the amount of enzyme was calculated for each timepoint, and the average percentages were plotted. The rates of product formation are 0.0091±0.0005 min⁻¹ with βME and 0.0061±0.0004 min⁻¹ without βME.

Figure 3

Determination of the size of native ALKBH1 and its protein-DNA adduct using LC-ESI/MS. (A) LC-ESI/MS analysis of native ALKBH1 (10 μM). (B) LC-ESI/MS analysis of the ALKBH1-DNA adduct. ALKBH1 (10 μM) was incubated with 1 μM oligo_AP1 for 3 h, the reaction was stopped by addition of 0.1 M MX and the sample was subjected to LC-ESI/MS.

Figure 4

Kinetics of adduct formation of ALKBH1 with its 5′-product and comparison to release of the 3′-product. (A) Representative PAGE gel showing 3′-product formation over time. ALKBH1 (5 μM) was incubated with 3′-FAM-labeled oligo_AP1 (1 μM) at 37 °C and product formation was analyzed at the indicated time points by PAGE. The control sample lacks protein and monitors unspecific cleavage of the AP site. (B) Representative SDS-PAGE gel illustrating adduct formation of ALKBH1 with 5′-product over time. ALKBH1 was incubated with 5′-FAM-labeled oligo_AP1 as described in panel A and adduct formation was analyzed by SDS-PAGE. (C) Adduct and product formation over time. Gels from three to five independent experiments were analyzed by ImageJ, and averages were calculated and plotted. Assays were carried out in the absence of βME.

Figure 5

Adduct formation of ALKBH1 and selected variants with the 5′-DNA product. (A) ALKBH1 WT and variants (5 μM) were incubated with 5′-FAM labeled oligo_AP1 for 1 h at 37 °C, the reaction was stopped by addition of MX, and adduct formation was analyzed by SDS-PAGE. (B) NTCB cleavage of the protein-DNA adduct for WT and variant forms of ALKBH1. ALKBH1 and its variants were incubated with DNA as described in panel A and then subjected to chemical cleavage at Cys residues by NTCB as described in the Experimental Procedures. E stands for endonuclease III, another AP lyase, which was incubated with the substrate DNA in order to cleave the oligonucleotide at the abasic site and visualize the size of the product on the gel. The samples were analyzed by Tris-tricine SDS-PAGE. For all gel analysis, FAM-labeled DNA was detected using a fluorescence scanner. (C) Scheme depicting ALKBH1 with the positions indicated for its 22 Lys and 7 Cys residues. Underneath are the corresponding protein fragments for the adducts seen on the Tris-tricine gel of panel B.

Figure 6

m⁶A Demethylation activity of ALKBH1. (A) Scheme of the assay. A m⁶A-containing oligonucleotide is incubated with ALKBH1 in the presence of Fe(II) and 2OG. After the assay, ALKBH1 is inactivated by heating to 95 °C and the ssDNA substrate is annealed to the complementary oligonucleotide by slowly cooling to room temperature. The reaction is then digested with DpnII, a methylation-sensitive restriction enzyme that cleaves the ds oligonucleotide only if the methyl group on the adenine has been removed. (B) Representative PAGE gel reporting the m⁶A-demethylase activity of ALKBH1. Different concentrations of ALKBH1 were incubated with oligo_m6A1 (1 μM), Fe(II) (1 mM), and 2OG (2 mM) for 1 h, then treated according to the scheme shown in panel A and analyzed by PAGE. (C) Time course of the reaction. ALKBH1 (1 μM) was incubated with 1 μM oligo_m6A1, cofactor, and cosubstrate, and the reaction was stopped at the indicated time points. Three independent experiments were used to calculate the average demethylation activity and the error bars represent the standard deviation. (D) The m⁶A-demethylase activity of ALKBH1 requires Fe(II) and 2OG, and is abolished in the iron-ligand variants H231A, D233A, and H287A. WT ALKBH1 (1 μM) was incubated with or without Fe(II) (1 mM) and 2OG (2 mM) and oligo_m6A1 (1 M) for 1 h at 37 °C and analyzed as before. The variants were analyzed similarly. (E) ALKBH1 demethylates ssDNA containing m⁶A, but not hemimethylated DNA. ALKBH1 (1 μM) was incubated with either ss oligo_m6A1 or oligo_m6A1 annealed to oligo_m6A2 containing one m⁶A site (1 μM) for 1 h and then treated and analyzed as described in panel B.

Figure 7

Comparison of m⁶A-demethylase and AP lyase activities of WT and variant forms of ALKBH1. ALKBH1 and selected variants (1 μM) were incubated with either oligo_m6A1 or oligo_AP1 (1 μM) for 30 min at 37 °C, then analyzed as described in Experimental Procedures. Three to five independent assays were used to calculate averages and standard deviations. The activities of the WT protein (0.005±0.001 min⁻¹ for m⁶A demethylation and 0.012±0.001 min⁻¹ for AP lyase) were set to 100% for comparison the activities of the variant proteins.

Scheme 1

Enzymatic activities reported for mammalian Alkbh1 in vitro^{a a}Key: (A) Demethylation of m³C in DNA. (B) Demethylation of methylated lysine in histone H2A (demonstrated only in a cell-based assay). (C) Demethylation of m⁶A in DNA. (D) Double hydroxylation of m⁵C in mitochondrial tRNA. (E) Demethylation of m¹A in tRNA. (F) Lyase cleavage of AP sites in DNA (this reaction is independent of Fe(II) and 2OG).

Scheme 2