Selenoprotein K Increases Efficiency of DHHC6 Catalyzed Protein Palmitoylation by Stabilizing the Acyl-DHHC6 Intermediate

Abstract

Selenoprotein K (SELENOK) is a selenocysteine (Sec)-containing protein localized in the endoplasmic reticulum (ER) membrane where it interacts with the DHHC6 (where single letter symbols represent Asp-His-His-Cys amino acids) enzyme to promote protein acyl transferase (PAT) reactions. PAT reactions involve the DHHC enzymatic capture of palmitate via a thioester bond to cysteine (Cys) residues that form an unstable palmitoyl-DHHC intermediate, followed by transfer of palmitate to Cys residues of target proteins. How SELENOK facilitates this reaction has not been determined. Splenocyte microsomal preparations from wild-type mice versus SELENOK knockout mice were used to establish PAT assays and showed decreased PAT activity (~50%) under conditions of SELENOK deficiency. Using recombinant, soluble versions of DHHC6 along with SELENOK containing Sec₉₂, Cys₉₂, or alanine (Ala₉₂), we evaluated the stability of the acyl-DHHC6 intermediate and its capacity to transfer the palmitate residue to Cys residues on target peptides. Versions of SELENOK containing either Ala or Cys residues in place of Sec were equivalently less effective than Sec at stabilizing the acyl-DHHC6 intermediate or promoting PAT activity. These data suggest that Sec₉₂ in SELENOK serves to stabilize the palmitoyl-DHHC6 intermediate by reducing hydrolyzation of the thioester bond until transfer of the palmitoyl group to the Cys residue on the target protein can occur.

Keywords: ANK repeat and PH domain-containing protein 2 (ASAP2); cluster of differentiation (CD36); inositol 1,4,5-triphosphate receptor; palmitic acid; palmitoyl-CoA; palmitoylation; selenium; thioester.

Figure 1

A diagram illustrating protocol for detecting palmitoylation of peptides by murine splenocyte microsomes. (A) Splenocytes were isolated from either WT or SELENOK mice, and microsomes were prepared containing ER membranes with or without SELENOK, respectively. Substrates including palmitoyl-CoA and FITC-peptide were added to microsomes for different periods of time after which the reaction mixtures were subjected to reverse-phase TLC. Fluorescent bands corresponding to palmitoylated peptides were identified using UV transillumination. (B) For each experiment, equivalent levels of DHHC6 were confirmed using Western blotting.

Figure 2

Palmitoylation of target peptides is decreased in SELENOK deficient microsomes. (A) A time course for peptide palmitoylation was performed using a TLC-based assay. A representative image for fluorescence imaged with a uv transilluminator. Note the imasge was was captured at high exposure to clearly show the palmitoylated peptide bands, which resulted in some red overexposed regions in the native peptide band. (B) Three independent experiments were conducted to compare microsomes from WT and SELENOK KO splenocytes. Values at each time point indicate percent of peptide that was palmitoylated = (FITC/(Native + FITC)) × 100. Means ± S.E. are shown.

Figure 3

Synthesis of a soluble cytosolic DHHC6 that binds to SELENOK. (A) Illustration of full-length DHHC6 with the SH3 and catalytic domains that were used to construct a modified version of DHHC6, which contains both domains connected by a linker domain shown in blue. (B) The DNA sequence used to express the cytosolic DHHC6 along with amino acid sequence showing linker region in blue. (C) StrepTactin beads were used to pull down 1 μg streptavidin-tagged SELENOK containing either Ala, or Cys, or Sec at amino acid position 92. These beads were divided into two aliquots that were not or were incubated overnight at 4 °C with the cytosolic DHHC6 in a 1:1 molar ratio. Eluted proteins were analyzed by Coomassie blue staining of PAGE, and results show all three versions of SELENOK bound to cytosolic DHHC6. Note that some degradation of the recombinant SELENOK occurred during the overnight incubation giving rise to double bands. Also, these are recombinant proteins and not cell lysates, which leads to no detectable background in lanes corresponding to the “no Cytosolic DHHC6 added” conditions.

Figure 4

Sec-containing SELENOK stabilizes acyl-DHHC6 intermediate formed during the autopalmitoylation of the enzyme. Cytosolic DHHC6 bound to different versions of SELENOK (U92 → A, U92 → C, and U92) bound to StrepTactin beads were incubated for 60 min with fluorescent NBD-palmitoyl-CoA in buffer of pH 6.8, 7.4, and 8.2. Beads were then washed, proteins eluted and analyzed by SDS-PAGE. (A) The gel was exposed to uv using a transilluminator and the acylated DHHC6 visualized at ~31 kDa with highest levels found in the lanes corresponding to Sec-containing SELENOK at pH 7.4 and 8.2. (B) Coomassie Blue staining of the gel revealed two bands for DHHC6 ~31 kDa. Based on the pattern within the pH ranges, the upper band corresponds to acylated DHHC6 and the lower band to the nonacylated DHHC6. (C) Coomassie Blue staining of the gel in the region of SELENOK (~14 kDa) shows that the protein was eluted from beads at equivalent levels. Note that some degradation of the recombinant SELENOK occurred during the overnight incubation giving rise to double bands.

Figure 5

Sec-containing SELENOK is most effective at stabilizing the acyl-DHHC6 intermediate. The autopalmitoylation of cytosolic DHHC6 was performed with different bead-bound versions of SELENOK (U92 → A, U92 → C, and U92) under different pH conditions. Western blotting was performed to detect NBD-palmitate-DHHC6 as a green fluorescent band and the DHHC protein that was detected with anti-DHHC6 and IR-secondary antibodies as a red fluorescent band.

Figure 6

Acyltransferase activity of DHHC6 is most efficient when bound to Sec-containing SELENOK. The NBD-palmitoyl-DHHC6 intermediates formed at pH 7.4 were incubated with CD36 peptide (MGCDRNCK) for different periods of time. Reactions were terminated with SDS on ice and reaction products run on reverse-phase TLC, followed by analysis on a uv transilluminator to detect NBD-palmitoyl-peptide. (A) The first experiment was run at three time periods of 2, 5, and 10 min. (B) The second experiment was run for 10 and 15 min.