Dual Role of ACBD6 in the Acylation Remodeling of Lipids and Proteins

Abstract

The transfer of acyl chains to proteins and lipids from acyl-CoA donor molecules is achieved by the actions of diverse enzymes and proteins, including the acyl-CoA binding domain-containing protein ACBD6. N-myristoyl-transferase (NMT) enzymes catalyze the covalent attachment of a 14-carbon acyl chain from the relatively rare myristoyl-CoA to the N-terminal glycine residue of myr-proteins. The interaction of the ankyrin-repeat domain of ACBD6 with NMT produces an active enzymatic complex for the use of myristoyl-CoA protected from competitive inhibition by acyl donor competitors. The absence of the ACBD6/NMT complex in ACBD6.KO cells increased the sensitivity of the cells to competitors and significantly reduced myristoylation of proteins. Protein palmitoylation was not altered in those cells. The specific defect in myristoyl-transferase activity of the ACBD6.KO cells provided further evidence of the essential functional role of the interaction of ACBD6 with the NMT enzymes. Acyl-CoAs bound to the acyl-CoA binding domain of ACBD6 are acyl donors for the lysophospholipid acyl-transferase enzymes (LPLAT), which acylate single acyl-chain lipids, such as the bioactive molecules LPA and LPC. Whereas the formation of acyl-CoAs was not altered in ACBD6.KO cells, lipid acylation processes were significantly reduced. The defect in PC formation from LPC by the LPCAT enzymes resulted in reduced lipid droplets content. The diversity of the processes affected by ACBD6 highlight its dual function as a carrier and a regulator of acyl-CoA dependent reactions. The unique role of ACBD6 represents an essential common feature of (acyl-CoA)-dependent modification pathways controlling the lipid and protein composition of human cell membranes.

Keywords: acylation; lipid droplets; lysophospholipids; myristoylation; palmitoylation.

Figure 1

Thioesterification requirements for acyl-donors and competitors’ formation. (A) Measurements of the formation of the myristoyl-peptide were performed in the absence or presence of the drug IMP-1088 (10–100–1000 nM) or the fatty acid analog 2-OH Myr (1 mM). Reactions were performed with 250 nM NMT2 enzyme in the presence of 50 µM Myr-CoA and the indicated concentration of drugs for 20 min at 37 °C. The peptide and myristoyl-peptide were extracted and quantified by HPLC on a C18 column, as previously described [2]. The chromatogram traces obtained at 274 nm at the time of the elution of the myr-peptide (≈34.6 min) from the C18 column were overlapped, as indicated. (B). Measurements of the formation of the myristoyl-peptide were performed in the absence of the NMT substrate Myr-CoA but in the presence of the myristate precursor Myr and human acyl-CoA synthetase ACSL6 enzyme. Reactions were performed from 0 to 24 min with NMT2 (250 nM), Myr (20 µM) and ACLS6 (150 nM) in the presence of ATP and CoASH. Inset: Control reactions were performed in the absence of either NMT2, ACSL6, ATP or CoASH, Formation of the myr-peptide was quantified after 24 min incubation. (C). Measurements were performed as described in panel B but in the absence (open circle) or the presence of the fatty acid analog 2-OH Myr (100 µM) (filled circle). Control reactions were performed in the absence of the acyl-donor precursor Myr (asterisk). (D). Measurements were performed as described in panel C, but myristate was replaced by Myr-CoA (10 µM). The rates of myr-peptide formation were calculated and are presented as the amounts of myristoyl-peptide formed per min. Reactions were performed with the indicated concentration of 2-OH Myr in the absence or presence of ACSL6. Error bars in (B–D) represent the standard deviations of values obtained from at least three reactions: n.s., non-significant; **, p = 0.01; ****, p < 0.0001. A cartoon representation of the fatty acids and of their CoA ester derivatives produced by the action of ACSL6 in the presence of ATP and CoASH is shown in the inset above Panel (B–D). Panel (E). Schematic representation of the reactions presented in panels (A) (top traces), (B–D).

Figure 2

Compounds affecting NMT activity and their precursors. The fatty acids palmitic acid (C_16:0), myristic acid (C_14:0), 2-hydroxymyristic acid (2-OH C_14:0), 12-azidododecanoic acid (12-ADA) and alkynyl myristic acid (YnMyr) are substrates of the cellular acyl-CoA synthetase enzymes. Their CoA thioester derivatives can bind the Myr-CoA binding site of NMT enzymes. The acyl-chain of C₁₆-CoA and 2-OH C₁₄-CoA are not a substrate for the acyl-transferase step and block myristoylation. The drug IMP-1088 (and other compounds in that class) occupies the peptide binding pocket of NMT and prevents myristoylation.

Figure 3

Disruption of the ACBD6 gene in HeLa cells. An out-of-frame deletion of the region encoding the entire ACB domain of the ACBD6 gene was obtained in HeLa cells. (A). The result of PCR reactions of cDNAs obtained from RNAs isolated from the ACBD6.KO and the parent HeLa cells with ACBD6 primers 1 and 2 is shown. The full-length ACBD6 cDNA (≈860 base pairs) present in normal cells is absent in the ACBD6.KO cells and a single cDNA (≈530 bp) was detected establishing that no unaltered copy of the ACBD6 gene was present in the ACBD6.KO cells. The ladder was the 1 Kb Plus DNA Ladder (ThermoFisher Scientific). (B). Deletion of the region coding the ACB domain was confirmed by sequencing of the cDNA. The deletion results in the out-of-frame fusion of the codons encoding Serine 15 and Tryptophan 125 preventing formation of an ACB truncated product. Primers used for the PCR amplification presented in panel A are indicated as 1 and 2. (C). Western-blot detection were performed with 100 µg of cell lysate proteins obtained from the parent HeLa cells and three ACBD6.KO lines. Proteins were separated on denaturing PAGE-gradient gel. After electrophoresis, proteins were transferred on PVDF membrane and probed with monoclonal antibodies against ACBD6, NMT2 and, β-actin. The protein ladder Blue Prestained Protein Standard (New England BioLabs, Ipswich, MA, USA) is indicated on the left.

Figure 4

Impaired NMT activity in ACBD6.KO cells challenged with NMT inhibitors. HeLa and ACBD6.KO cells were grown in 96-well plates and exposed to the indicated concentration of IMP-1088 (A) and 2-hydroxymyristate (2-OH Myr) (B). Growth of cells in the absence or the presence of 10 µM 2-OH Myr are presented in (C). Cells were fixed in 10% ice-cold TCA, stained with the SRB dye, and absorbance was read at 560 nm [3]. In (A,B), the values obtained in the presence of the drugs are reported relative to the values obtained in their absence (p < 0.0001). Error bars represent the standard deviations of values obtained from 12 measurements. (D) The values of the growth inhibition difference of the ACBD6.KO cells relative to HeLa obtained in the presence of IMP-1088 at the indicated concentrations (A) were plotted as a function of the values obtained in the presence of 2-OH Myr (B). The dashed line represents the plot of the theoretical values in a scenario of identical sensitivity for the two inhibitors.

Figure 5

In vivo N-myristoylation of proteins is reduced in the absence of ACBD6. (A,B). HeLa and ACBD6.KO cells were grown in flask and exposed to the labeling probe 12-ADA (azido-myristate) at a concentration of 5 µM for 1–4–18 h, as indicated. Cells were harvested, lysed, and azido-myristoylated-proteins were detected by Click chemistry (see Method). Proteins were separated on denaturing SDS-gradient gel, transferred on PVDF membrane, and detected with streptavidin-HRP. β-actin was used as a loading reference and was detected with a monoclonal antibody (A). Intensity of a major band (asterisk) and all the visible bands were quantified in Hela (circle) and ACBD6.KO (square) (B). Values are reported relative to the intensity of the β-actin signal detected in each sample as a function of time. The values obtained for the ACBD6.KO cells are also reported relative to the values obtained with HeLa cells (filled red square; data plotted on the right y axis). Error bars represent the standard deviations of values obtained from 3 measurements: **, p < 0.05; ***, p < 0.005. (C,D). HeLa and ACBD6.KO cells were exposed to the labeling probe 15-azido-pentadecanoic acid (azido-palmitate) at a concentration of 10 µM for 1–4–18 h, as indicated. Proteins were detected as indicated above. The protein ladder Blue Prestained Protein Standard (New England BioLabs, Ipswich, MA, USA) is indicated on the right in (A,C). The values obtained for the ACBD6.KO cells are reported relative to the values obtained with HeLa cells. Error bars represent the standard deviations of values obtained from 3 measurements.

Figure 6

Decrease activity of the lipid acylating enzymes. (A). HeLa and ACBD6.KO cells, grown in 96-well plates, were exposed to 5 µM [¹⁴C]C_16:0 for the indicated times (10 to 120 min). Incorporation of the radiolabel was quantified with a scintillation counter and is reported relative to the protein content, quantified with SRB staining. Error bars represent the standard deviations of values obtained from 8 measurements. (B–D). HeLa and ACBD6.KO cells were grown in T75 flasks. Cells were harvested, lysed, and debris was removed by centrifugation. (B), the rate of esterification of [¹⁴C]C_16:0 (10 µM) into [¹⁴C]C₁₆-CoA in the presence of ATP, CoASH and of 8 µg protein lysate was determined from 0 to 10 min. Values obtained with ACBD6.KO are reported relative to the values obtained with HeLa. In (C), acylation of lysoPA (10 µM) by 20 µg protein lysate was monitored from 0 to 10 min. In (D), acylation of lysoPC (10 µM) by 8 µg protein lysate was monitored from 0 to 16 min. In each assay, the amount of product formed (yield) was calculated at the last time point. Error bars in (B–D) represent the standard deviations of values obtained from 6 measurements: n.s., non-significant; ****, p < 0.000001.

Figure 7

Analysis of the acyl-CoA dependent acylation of lysoPA and lysoPC. Products of the reactions presented in Figure 6 and Figure 8 were separated by thin-layer chromatography as previously described [51]. Formation of PC and of PA by lysates of ACBD6.KO and HeLa are shown in the top and middle panels, respectively. Formation of PC by lipid droplets (LDs) isolated from ACBD6.KO and HeLa is presented in the bottom panel.

Figure 8

Reduced LPCAT activity impaired LD formation in ACBD6.KO cells. (A). HeLa and ACBD6.KO cells were grown in 96-well plates in the absence or presence of 200 µM oleic acid (OA) for 24 h. LD content was quantified with a Fluorometric Lipid Droplet staining (see Section 2) and was normalized to the protein cellular content, quantified with SRB staining. Values obtained with the ACBD6.KO cells are reported relative to the values obtained with HeLa cells. Error bars represent the standard deviations of values obtained from 21 measurements: ****, p < 0.000001. (B). Re-acylation of lysoPC (20 µM) by LDs isolated from ACBD6.KO and HeLa cells were performed with 5 µM [¹⁴C]C₁₆-CoA and 0.8 μg LD proteins. The rate of PC formation was calculated from 0 to 10 min and values obtained with ACBD6.KO are reported relative to the values obtained with HeLa cells. Thin-layer chromatography analysis of the reactions is presented in the bottom panel of Figure 7. Error bars represent the standard deviations of values obtained from 3 measurements. n.s.—not significant.