ALPK1 phosphorylates myosin IIA modulating TNF-α trafficking in gout flares

Abstract

Gout is characterized by the monosodium urate monohydrate (MSU)-induced arthritis. Alpha kinase-1 (ALPK1) has shown to be associated with MSU-induced inflammation and gout. Here, we used bioinformatics, proteomics, cell models, and twenty in vitro human assays to clarify some of its role in the inflammatory response to MSU. We found myosin IIA to be a frequent interacting protein partner of ALPK1, binding to its N-terminal and forming a protein complex with calmodulin and F-actin, and that MSU-induced ALPK1 phosphorylated the myosin IIA. A knockdown of endogenous ALPK1 or myosin IIA significantly reduced the MSU-induced secretion of tumour necrosis factor (TNF)-α. Furthermore, all gouty patients expressed higher basal protein levels of ALPK1, myosin IIA, and plasma TNF-α, however those medicated with colchicine has shown reduced myosin IIA and TNF-α but not ALPK1. The findings suggest ALPK1 is a kinase that participates in the regulation of Golgi-derived TNF-α trafficking through myosin IIA phosphorylation in the inflammation of gout. This novel pathway could be blocked at the level of myosin by colchicine in gout treatment.

Figure 1. Conserved α-helical structures within ALPK1 N-terminal domain.

(A) An enlarged view of ALPK1 N-terminal region and cross-species comparison of N-terminal orthologues (asterisks denote highly conserved amino acid residues that have predicted noncovalent properties with myosin IIA); (B) N-terminal-conserved α-helical-rich regions of human ALPK1 protein can potentially fold into an amphipathic helical structure, well-suited for vesicular transport and curvature fusion sensors. Here, four putative amphipathic α-helical structures are shown that demonstrated partitions of hydrophobic, nonpolar, negative, and positive residues represented as diamonds, circle, triangles, and pentagons, respectively. The area separated by dotted line illustrates the amphipathic properties and the arrow indicates the hydrophobic moment and direction.

Figure 2. ALPK1 vector construction and confirmation.

(A) Schema of Halo-ALPK1 full-length and HA-ALPK1 full-length/truncated forms used in pull-down and immunoprecipitation assays. (B) Schema of various deletion form constructs (shaded) from a full-length ALPK1 (black), the positions (1–3732 bp) of ALPK1 ORFs are as indicated and residue numbers are base pairs. (C) Purification of the Halo-ALPK1 fusion protein using HaloLink resin in mammalian cells, and then pull-down assay using THP-1 cell lysates. Arrows indicate results on the SDS-PAGE gel (left) and wavelength output (right). (D) Protein expressions of ALPK1 full-length and deleted form. The HEK293T cells were transfected with full and serial deletion form of the constructed plasmid and confirmed by an anti-HA antibody on Western blot.

Figure 3. Identification of ALPK1 and three highly scoring protein binding partners: myosin IIA, F-actin, and calmodulin.

(A) Full-length and N-terminal region of Halo-ALPK1 and halo fusion proteins were overexpressed by transient transfection of HEK293F cells and then purified using HaloLink resin. Arrows indicate protein bands unique to the Halo-ALPK1 lanes [~220, 170, 100, 43, 22 kDa positions were identified as: myosin IIA, full-length (FL)-ALPK1, N-terminal (Nt)-ALPK1, F-actin, calmodulin] by LC/MS/MS protein identification. (B) Western blotting using ALPK1, myosin IIA, F-actin, and calmodulin antibodies for the detection of FL-ALPK1, Nt-ALPK1, myosin IIA, F-actin, and calmodulin, and confirmed from the same pull-down assay. (C) FL- and Nt1-ALPK1 specifically interacted with both myosin IIA and F-actin. FL- and Nt2-ALPK1 strongly interacted with endogenous calmodulin. Interaction of ALPK1 serial deletion form with myosin IIA, F-actin and calmodulin was confirmed in 293T cells by coimmunoprecipitation with the transfected HA-tagged ALPK1 serial deletion form and Flag-tagged myosin IIA, followed by SDS-PAGE and Western blotting using anti-HA, anti-Flag, anti-F-actin, or anticalmodulin antibodies. (D) HEK293T cells were cotransfected with HA-tagged ALPK1 serial deletions form and Flag-myosin IIA. The lysates were subjected to coimmunoprecipitation with an antibody against Flag. Presence of ALPK1, F-actin, and calmodulin in the precipitate was detected using anti-HA, anti-F-actin, and anticalmodulin antibodies.

Figure 4. MSU-induced phosphorylation of myosin IIA is ALPK1 dependent.

(A) After 16-h treatments with uric acid (0, 25, and 50 mg/dl) or MSU crystals (0, 50, and 100 μg/mL), uric acid did not increase the protein expression of ALPK1, however MSU did (p < 0.05). (B) Uric acid and MSU did not affect myosin IIA expression. (C) Both MSU treatment and ALPK1 increased myosin IIA phosphorylation (lane 3) in THP-1 cells using antiphospho-serine/threonine antibody. (D) Phosphorylated myosin IIA was confirmed at the 220-kDa position, from comparing the depletion and no depletion of myosin IIA by MYH9 siRNA, on the Pro-Q staining SDS-PAGE gel (left) and protein-specific Coomassie Blue staining gel (right). A control without MSU stimulus is shown in Fig. S3. (E) Without MSU stimulus, the phosphorylation of myosin IIA is shown to require ALPK1 and ATP (lane 1) on the Pro-Q staining assay. The weak myosin IIA bands for example in second lane that has no myosin were trace myosin IIA bound to ALPK1 FL and Nt1, refer to Fig. 3C, from purification that was co-eluted.

Figure 5. TNF-α secretion is relatively controlled by ALPK1 and myosin IIA in MSU-induced monocytes.

(A) Protein knockdown of si-ALPK1 and si-MYH9 by immunofluorescence (left) and quantification of intensity of colour (right) showed they were effective but not complete knockdown. (B) MSU treatment (compared to none) has generally increased intracellular TNF-α levels, including when ALPK1 was knockdown, which could be explained by the limitation in knockdown efficiency (see Fig. 5 A) or other protein kinases that phosphorylated myosin IIA were not eliminated. The most intracellular TNF-α increase was seen with myosin IIA knockdown. (C) MSU-induced TNF-α secretion was reduced with the knockdown of the endogenous ALPK1 or myosin IIA, however no effect was observed for IL-1β secretion. (D) TNF-α secretion was unaffected by the depletion of another kinase that may phosphorylate myosin IIA, myosin light-chain kinase (MLCK), using siRNA for 24 h prior treatment with MSU. Data are representative of results performed in triplicates. *p < 0.0001.

Figure 6. Gouty patients express higher protein levels of ALPK1, myosin, and plasma TNF-α.

(A) In vitro human assays of freshly isolated monocytes from twenty gout patients and ten healthy controls were analysed for the levels of ALPK1, myosin IIA, and plasma TNF-α. Eight gout patients (no. 3, 5 12, 14, 17, 18, 19, 20; details refer to in Supplementary Table 3) who were medicated with a combined benzbromarone and colchicine showed lower plasma TNF-α than the remaining twelve gout patients who were nonmdicated. (B) Gout patients were pooled to show they expressed more ALPK1, however no difference was observed between medicated and nonmedicated groups (left), gout patients expressed more myosin IIA, which was reduced in the medicated group (central), gout patients have higher levels of plasma TNF-α (right), especially in the nonmedicated (60.5 ± 8.7 pg/mL) than medicated patients (39.2 ± 9.7 pg/mL) or healthy controls (18.8 ± 12.6 pg/mL). (C) In vitro effect of medication (benzbromarone, colchicine, or combined of two medications) showed a high dose colchicine (1 mM) had reduced myosin IIA. There was no effect on ALPK1 expression. The THP-1 cells had been incubated with 100 nM PMA for 3 h and then treated with 100 μg/ml MSU for 16 h.

Figure 7. Possible role of ALPK1 in gouty inflammation.

(1) MSU crystals stimulate ALPK1 gene overexpression in the monocyte. (2) One of α-kinase’s functions is bind with myosin IIA via N-terminal (Nt) while C-terminal (Ct) phosphorylates the motor protein in presence of ATP. (3) Activated myosin IIA at the Golgi membrane. (4) Transport of Golgi-derived TNF-α vesicles. (5) Transport towards plasma membrane and contribute to secretion of TNF-α. (6) ALPK1 is unaffected by colchicine, so MSU persists in stimulating ALPK1 expression. However, colchicine (blue line) disrupts microtubule formation, which myosin motor proteins act upon, thus less myosin is recruited, resulting in less vesicular delivery of TNF-α. An aberrant ALPK1 or knockdown may decrease phosphorylation of myosin IIA and lower TNF-α secretion.