Enzymatic characterization of a human acyltransferase activity

Abstract

Background: Non-histone protein acylation is increasingly recognized as an important posttranslational modification, but little is known as to the biochemical properties of protein serine acylating enzymes.

Methodology/principal findings: We here report that we have identified a metal-stimulated serine octanoyltransferase activity in microsomes from human erythroleukemic (HEL) cells. The HEL acylating enzyme was linear with respect to time and protein, exhibited a neutral pH optimum (stimulated by cobalt and zinc), and inhibited by chelating reagents. Hydroxylamine treatment removed most, but not all, of the attached radioactivity. A salt extract of microsomal membranes contained the major portion of enzyme activity, indicating that this acyltransferase is not an integral membrane protein. Sucrose density fractionation showed that the acyltransferase activity is concentrated in the endoplasmic reticulum. In competition experiments, the acyltransferase was well inhibited by activated forms of fatty acids containing at least eight to fourteen carbons, but not by acetyl CoA. The zinc-stimulated HEL acyltransferase did not octanoylate proenkephalin, proopiomelanocortin, His-tagged proghrelin, or proghrelin lacking the amino-terminal His-tag stub of Gly-Ala-Met. The peptides des-acyl ghrelin and ACTH were also not acylated; however, des-acyl ghrelin containing the N-terminal tripeptide Gly-Ala-Met was acylated. Mutagenesis studies indicated a requirement for serine five residues from the amino terminus, reminiscent of myristoyl transferase, but not of ghrelin acylation. However, recombinant myristoyl transferase could not recapitulate the hydroxylamine sensitivity, zinc-stimulation, nor EDTA inhibition obtained with HEL acyltransferase, properties preserved in the HEL cell enzyme purified through four sequential chromatographic steps.

Conclusions/significance: In conclusion, our data demonstrate the presence of a zinc-stimulated acyltransferase activity concentrated in the endoplasmic reticulum in HEL cells which is likely to contribute to medium-chain protein lipidation.

Figure 1. The microsomal fraction of HEL cells contains acyltransferase activity.

(A) Acyltransferase activity was tested using [¹⁴C]octanoic acid transfer to either His-tagged proghrelin or GAM-proghrelin and the P2 microsomal fraction from HEL cells as an enzyme source. The reactions were carried out under the standard reaction conditions at 37 C for 2 h and then analyzed on 16.5% polyacrylamide gels. Left panel, Coomassie-stained gel of reaction mixtures to demonstrate the presence of equal quantities of His-tagged proghrelin and GAM-proghrelin; right panel, autoradiogram of the same reaction mixtures to identify the [¹⁴C]octanoylated band. Lane 1, His-tagged proghrelin alone; lane 2, His-tagged proghrelin with P2 microsomal fraction; lane 3, GAM-proghrelin; lane 4, GAM-proghrelin with P2 microsomal fraction; lane 5, P2 microsomal fraction alone. An arrow and asterisk indicate [¹⁴C]octanoylated GAM-proghrelin and endogenous substrate protein, respectively. (B) The removal of [¹⁴C]octanoic acid from [¹⁴C]octanoylated GAM-proghrelin was performed using either 1 M Tris-HCl, pH 8.0 or 1 M NH₂OH, pH 8.0 (B).

Figure 2. Acyltransferase activity can be extracted from HEL cell microsomes with high salt.

Proteins in the P2 pellet were extracted according to the procedures described in Materials and Methods, dialyzed, and assayed for enzyme activity under the standard reaction conditions. Proteins were extracted under the following conditions: Set 1, using 10 mM Tris-HCl; set 2, 1 M NaCl; set 3, 1% Triton X-100; set 4, 1% sequential extraction first with Triton X-100, then the pellet extracted with 1 M NaCl; set 5, simultaneous extraction with 1% Triton X-100 and 1 M NaCl. S, supernatant; P, pellet. Results are given as dpm of total octanoyltransferase activity (reaction dpm multiplied by the total protein in each fraction). Samples were assayed in duplicate and the mean and standard deviation are shown.

Figure 3. Acyltransferase activity is enriched in endoplasmic-reticulum-containing fractions.

(A) Schematic representation of fractions obtained using differential centrifugation. Details of this experiment are described in Materials and Methods. “Sup” and “ppt” indicate the supernatant and pellet, respectively, obtained from each centrifugation. (B) The total acyltransferase activity in each fraction was calculated using 2.5 µg protein under standard assay conditions and correcting for the total protein in each fraction. Two and a half µg of protein from each fraction were subjected to SDS-PAGE, and analyzed by Western blotting using each subcellular marker; calreticulin, ER; TGN-46, Golgi; catalase, peroxisome; prohibitin-1, mitochondria. (C) The P2 microsomal fraction was subjected to sucrose gradient centrifugation. Two and a half µg protein (10 µg for the TGN fraction) from each fraction were subjected to SDS-PAGE and analyzed by Western blotting using antisera against proteins with known subcellular localizations. The band intensities of marker proteins were measured using an Alphaimager 3300, and the total band intensity of each marker protein in each fraction was calculated as described in the Materials and Methods. Acyltransferase samples were assayed in duplicate, and the mean and standard deviation are shown. Results are given as total dpm of octanoyltransferase activity in each fraction, and as arbitrary units of total band intensity per fraction. One of four independent fractionation experiments is shown; all gave essentially the same results.

Figure 4. Zinc directly enhances ERAT activity.

All experiments in this figure were performed with aliquots of the most active fraction from Mini-S chromatography as of acyltransferase enzyme source. To characterize the biochemical properties of ERAT, 2 µg of GAM-proghrelin were incubated for 2 h with either 0.1 mM ZnCl₂, 5 mM EDTA, 1 mM TLCK or 5 mM DTT (A), or with the different concentrations of ZnCl₂ indicated in (B). In order to determine whether zinc acts directly or indirectly on ERAT, GAM-proghrelin was pre-incubated with either 0.1 mM ZnCl₂ or 5 mM EDTA (C). After this preincubation, [¹⁴C]octanoyl CoA was added either with or without additional zinc, as shown. Acylation reactions were then carried out at 37C for 1 h. All samples were tested in duplicate, and the mean and standard deviation are shown. Results are given as dpm of [¹⁴C]octanoic acid transferred to GAM-proghrelin per reaction.

Figure 5. ERAT is time- and protein-dependent.

Enzyme activity was determined at the indicated times (A), while the dependence of activity on the amount of HEL cell P2 protein and GAM-proghrelin is shown in (B) and (C), respectively. Samples were tested in duplicate under the standard reaction conditions, and the mean and standard deviation are shown. Results are given as dpm of [¹⁴C]octanoic acid transferred to GAM-proghrelin.

Figure 6. ERAT has a neutral pH optimum.

Enzyme activity was determined at pHs between 5.0–9.0 under standard conditions using aliquots of active fractions from Mini-S chromatography. The indicated pH conditions were reached using either 50 mM Tris-HCl buffer (solid line) or 50 mM sodium phosphate buffer (broken line) in independent experiments. Samples were tested in duplicate at each pH point and the mean and standard deviation are shown. Results are given as the dpm of [¹⁴C]octanoic acid transferred to GAM-proghrelin.

Figure 7. ERAT acylates GAM-proghrelin, requires certain amino acids at the N-terminus of the acylated protein, and is specific for Ser5.

(A) ERAT acylates GAM-proghrelin, but not His-tagged proghrelin, nor other precursors. Two µg of each peptide precursor were incubated with HEL cell P2 protein in the presence or absence of 0.1 mM Zn⁺⁺. Lanes 1 and 2, His-tagged proghrelin; lanes 3 and 4, GAM-proghrelin; lanes 5 and 6, mouse POMC; lanes 7 and 8, rat proenkephalin; lanes 9 and 10, ACTH. Left panel, Coomassie-stained gel of reaction mixtures showing the presence of equal quantities of each substrate; right panel, autoradiogram of the same gel, to identify [¹⁴C]octanoylated bands. (B) ERAT requires certain amino acids at the N-terminus of the acylated protein. Two µg of the peptides des-acyl ghrelin (lane 1) or GAM-des-acyl ghrelin (lane 2) were tested as potential substrates for ERAT. Left panel, autoradiogram of reaction mixtures to identify [¹⁴C]octanoylated bands; right panel, Coomassie-staining of the same gel to demonstrate the quantity of peptide. (C) ERAT is specific for Ser5: lack of acylation of the GAM-proghrelin mutants and preproghrelin. Acyltransferase activity was tested using GAM-proghrelin (lane 1), each mutant (S5A, lane2; S6A, lane 3; and S5,6A, lane4) and preproghrelin (lane 5) under the standard reaction conditions. Lower panel, autoradiogram to identify [¹⁴C]octanoylated bands; upper panel, Coomassie staining of the same gel in order to demonstrate the presence of equal quantities of each substrate.

Figure 8. ERAT is inhibited by addition of competing acyl CoAs containing at least 8–14 carbons.

Duplicate reactions containing various acylated coenzymes were incubated with 2 µg of GAM-proghrelin and the P2 fraction under standard reaction conditions. Samples were analyzed in duplicate at each point and the mean and standard deviation are shown. Results are given as the percentage of octanoic acid transfer as compared to control reactions lacking added competing acyl CoAs.