Binding Affinity Determines Substrate Specificity and Enables Discovery of Substrates for N-Myristoyltransferases

Kinetic parameters (kcat and Km) derived from the Michaelis–Menten equation are widely used to characterize enzymes. kcat/Km is considered the catalytic efficiency or substrate specificity of an enzyme toward its substrate. N-Myristoyltransferases (NMTs) catalyze the N-terminal glycine myristoylation of numerous eukaryotic proteins. Surprisingly, we find that in vitro human NMT1 can accept acetyl-CoA and catalyze acetylation with kcat and Km values similar to that of myristoylation. However, when both acetyl-CoA and myristoyl-CoA are present in the reaction, NMT1 catalyzes almost exclusively myristoylation. This phenomenon is caused by the dramatically different binding affinities of NMT1 for myristoyl-CoA and acetyl-CoA (estimated Kd of 14.7 nM and 10.1 μM, respectively). When both are present, NMT1 is essentially entirely bound by myristoyl-CoA and thus catalyzes myristoylation exclusively. The NMT1 example highlights the crucial role of binding affinity in determining the substrate specificity of enzymes, which in contrast to the traditionally held view in enzymology that the substrate specificity is defined by kcat/Km values. This understanding readily explains the vast biological literature showing the coimmunoprecipitation of enzyme–substrate pairs for enzymes that catalyzes protein post-translational modifications (PTM), including phosphorylation, acetylation, and ubiquitination. Furthermore, this understanding allows the discovery of substrate proteins by identifying the interacting proteins of PTM enzymes, which we demonstrate by identifying three previously unknown substrate proteins (LRATD1, LRATD2, and ERICH5) of human NMT1/2 by mining available interactome data.

: NMT1/2 interacting proteins from publicly available database. This is uploaded as a separate Excel file. Figure S1. HPLC chromatogram of 1 h incubation of 10 µM ARF peptide, without (blue) and with (black) 15 µM NMT1. Arrow indicates the elution of myristoylated ARF peptide.

Why myristoyl-CoA and acetyl-CoA have similar k cat /K m values despite their different preference by NMT1. Previous studies established an ordered Bi-Bi reaction mechanism where
acyl-CoA binds NMT prior to peptide, then acyl peptide release is followed by the dissociation of free CoA (see scheme below) (J. Biol. Chem 266:9732-9739, 1991) Michaelis constant for the acyl-CoA substrate (KCoA) and kcat were defined in Eq a and b (Segel H. I., in Enzyme kinetics: behavior and analysis of rapid equilibrium and steady state enzyme systems, Wiley, New York, 1975). kcat/KCoA for acyl-CoA thus equals k1. The similar kcat/KCoA values for myristoyl-CoA and acetyl-CoA indicates two substrates bind to NMT1 with similar second order binding rate constant k1. The large difference in binding affinity Kd, which is given by k2/k1, indicates that acetyl-CoA has a much larger dissociation rate k2 compared to myristoyl-CoA. Thus, for myristoyl-CoA, k3 (the forward chemical reaction rate) dominates while for acetyl-CoA, k2 (the unbinding from NMT) dominates.
The organic lay was washed by water and brine and dried with anhydrous sodium sulfate. The solution was concentrated and the residue was purified by silica gel column chromatography (20:1 hexane:ethyl acetate) to afford 1-bromopentadecan-2-one (700 mg, 52.1% yield) as a white solid.
Then the reactions were quenched with 50 µL acetonitrile and incubated at room temperature for supernatants of 95 µL were transferred to borosilicate glass vials (Thermo Scientific/MSCERT5000-39TR) and 5 µL 50% trifluoroacetic acid (TFA) was added to each vial.
The samples were analyzed using HPLC with a Kinetex 5u EVO C18 100A column (150 mm × 4.6 mm, Phenomenex). The gradient for HPLC analysis was 100% solvent A ( water with 0.1% TFA) for 2 min, 0 to 35% solvent B (acetonitrile with 0.1% TFA) for 5 min, 35-64% solvent B for 15 min, 64-80% solvent B for 5 min, 89-100% solvent B for 2 min, 100% solvent B for 5 min and 100% solvent A for 5 min. To ensure rates were measured in steady state, the reaction is limited to  20% conversion of peptide substrate and the remaining concentrations of substrates after reaction were at least 10 times of the concentration of NMT1.
For the product detection by MS shown in Figure 1, the reactions were set up with the same method as mentioned above with 100 µM acyl-CoA, 100 µM ARF6 peptide and 32.5 nM NMT1.
The reactions were incubated for 2 hours. The procedure for HPLC analysis was the same as described above. The LC-MS analysis of reaction products was conducted on Shimazu LC-20AD instrument coupled with a Thermo Scientific LCQ Fleet mass spectrometer. The supernatants collected after centrifugation was loaded on a Phenomenex Kinetex 5 μm EVO C18 column (50 mm × 2.1 mm) equilibrated with 100% buffer A (water with 0.1% acetic acid) and 0% buffer B (acetonitrile with 0.1% acetic acid). The elution gradient was 100% solvent A from for 3 min, 0-10% solvent B for 2 min, 10-100% solvent B for 5 min, 100% solvent B for 1 min and 0-100% A for 1 min. Substrates and products were monitored with electrospray in positive mode.

Reversible inhibition assays. Reversible inhibition of NMT1 by myristoyl-CoA analog S-
(2-oxo)pentadecyl-CoA was carried out by varying both concentrations of myristoyl-CoA and S-(2-oxo)pentadecyl-CoA with 12.5 µM ARF6 peptide in 50 mM Tris-HCl, pH 8.0, by monitoring the rate of myristoylated ARF6 peptide formation using HPLC. Reversible inhibition of NMT1 by acetyl-CoA analog S-acetonyl-CoA was investigated by varying both concentrations of acetyl-CoA and S-acetonyl-CoA with 100 µM ARF6 peptide in 50 mM Tris-HCl, pH 8.0, by monitoring the rate of acetylated ARF6 peptide formation using HPLC. The substrate analogs were dissolved in the reaction buffer and their concentrations were determined using a CoA extinction coefficient of 1.54 × 10 4 M -1 cm -1 at 260 nm. The reaction and HPLC procedures were the same as described above.
Data analysis. Kinetics data were fit with KaleidaGraph software (Synergy Software, Reading, PA). Apparent steady-state kinetic data were fit to Eq 1, the Michaelis-Menten equation, where S is the concentration of substrate acyl-CoA. Steady-state kinetic data were fit to Eq 2, which describes an ordered steady-state kinetic mechanism where vo represents the initial velocity, e is the concentration of the enzyme, kcat is the first-order macroscopic rate constant for enzyme turnover at saturating concentration of both acyl-CoA and ARF6 peptide, KCoA and Kpep are the Michaelis constants for acyl-CoA and ARF6 peptide, respectively, and Kia is the dissociation constant of the enzyme and acyl-CoA.
The kinetic data with substrate analog inhibitors were fit to Eq 3, 4 and 5, where S is the concentration of substrate acyl-CoA, I is the concentration of inhibitor and Ki is the dissociation constant for the inhibitor. Eq 3 describes a competitive inhibition pattern of the inhibitor versus acyl-CoA substrates. Eq 4 and 5 describe a non-competitive and uncompetitive pattern respectively. hour. 10 µL of 2 × SDS protein loading dye was added and the mixtures were heated at 95 o C for 10 min before centrifugation at 17,000 × g for 2 minutes. The supernatants were collected and treated with 250 mM hydroxylamine at 95 o C for 7 minutes. 12 uL of each sample was loaded on a 12% polyacrylamide gel and allowed to be resolved at 220 V for 1 hour. In-gel fluorescence was detected with Typhoon FLA7000 (GE Healthcare Life Science). 3 µL of each sample was collected for western blotting analysis for protein loading.