Diacylglycerol kinase ζ interacts with sphingomyelin synthase 1 and sphingomyelin synthase‐related protein via different regions

We previously reported that diacylglycerol (DG) kinase (DGK) δ interacts with DG‐generating sphingomyelin synthase (SMS)‐related protein (SMSr), but not SMS1 or SMS2, via their sterile α motif domains (SAMDs). However, it remains unclear whether other DGK isozymes interact with SMSs. Here, we found that DGKζ, which does not contain SAMD, interacts with SMSr and SMS1, but not SMS2. Deletion mutant analyses demonstrated that SAMD in the N‐terminal cytosolic region of SMSr binds to the N‐terminal half catalytic domain of DGKζ. However, the C‐terminal cytosolic region of SMS1 interacts with the catalytic domain of DGKζ. Taken together, these results indicate that DGKζ associates with SMSr and SMS1 in different manners and suggest that they compose new DG signaling pathways.

In the present study, we comprehensively searched for interactions between DGK isozymes and SMS isozymes. We found that DGKf binds to SMS1 and SMSr but not SMS2. Moreover, DGKf interacts with SMSr and SMS1 in different manners. These results suggest that, beyond our expectations, DGK isozymes and SMS isozymes form a complex network.
Plasmids for expressing N-terminal 39FLAG-tagged human or rat DGK isozymes [44] and for expressing Cterminal V5-tagged human SMS isoforms [42] in mammalian cells were used.

Plasmid constructs
We used the following nomenclature for epitope-tagged proteins: TagX-(protein) and (protein)-TagY means that TagX and TagY are located at the N and C termini of the protein, respectively.
Sf9 cells were maintained in Sf-900 II serum-free medium (Invitrogen, Waltham, MA, USA) in sterile Erlenmeyer flask at 120 r.p.m. and 28°C without CO 2 in the dark. Volume of the medium was kept at 20-30% of flask volume. To generate recombinant baculovirus was generated using pOET3 vector and the flashBAC system (Oxford Expression Technologies) as described previously [43].
Glutathione S-transferase pull-down assays were performed as previously [42]. Purified GST-SMS1-CT or SMSr-NT were incubated with glutathione-Sepharose beads for 30 min at 4°C with constant rocking. The beads were washed five times with buffer containing 20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 0.1% (v/v) Triton X-100, and 1 mM phenylmethylsulfonyl fluoride. Purified Twin-strep tagged DGKf was incubated with the beads for 2 h at 4°C with constant rocking. Then, the beads were washed five times with buffer. The washed beads washed were then boiled in SDS sample buffer, and the extracts were analyzed by western blotting.

DGK activity assay
Diacylglycerol kinase activity was measured using liquid chromatography-tandem mass spectrometry (LC-MS/MS) as previously described [42].

Western blotting
Western blotting was carried out as previously described [42]. Equal quantities of protein were loaded onto a polyacrylamide gel. Separated proteins were transferred onto a polyvinylidene fluoride membrane (Millipore, Burlington, MA, USA) and incubated overnight at 4°C with the following primary antibodies: anti-FLAG (F-1804), V5 (MA5-15253), GFP (sc-9996), Twin-Strep (M211-3), and GST (PM013) antibodies. After washing, the membranes were incubated with a secondary antibody solution (goat antirabbit IgG-HRP or goat anti-mouse IgG-HRP) at room temperature for 1 h, followed by detection using the enhanced chemiluminescence method.

Statistical analysis
Data are represented as the means AE SDs and were analyzed by the Student's t test for the comparison of two groups or one-way ANOVA followed by Tukey's or Dunnett's post hoc test for multiple comparisons using GRAPHPAD PRISM 8 (GraphPad Software, Boston, MA, USA) to determine any significant differences. P < 0.05 was considered significant.

The N-terminal SAMD of SMSr and the Cterminal region of SMS1 interact with DGKf
We next attempted to determine a DGKf-interaction region in SMSr. AcGFP-tagged N-terminal (AcGFP-SMSr-NT) and C-terminal (AcGFP-SMSr-CT) cytosolic regions of SMSr were generated ( Fig. 2A), and their association with 39FLAG-tagged DGKf was determined. We found that the N-terminal cytosolic region of SMSr, which contains SAMD, strongly interacted with DGKf (Fig. 2B,C). Although the C-terminal cytosolic region of SMSr moderately cosedimented DGKf, statistical significance was not detected (Fig. 2B,C).
To narrow the DGKf-interaction area in the Nterminal cytosolic region of SMSr, the DGKf- interaction activity of SAMD alone of SMSr was tested. Figure 2D,E, 3D,E show that the SAMD of SMSr bound to DGKf.

Discussion
In the present study, we demonstrated for the first time that DGKf interacts with SMS1 and SMSr but not SMS2 (Figs 1 and 5). DGKd1 and d2 also bound to only SMSr but not SMS1 or SMS2 (Figs 1 and 5), as previously reported [42]. Moreover, DGKa, b, c, g1, g2, j, e, ι, and h failed to show interactions with SMSr and SMS1 (Fig. 1). Therefore, the interaction between DGKf and SMSr and the association between DGKf and SMS1 are highly selective. We previously reported that DGKd associates with SMSr via the interaction between DGKd-SAMD and SMSr-SAMD [42]. Although DGKf does not have SAMD [1-5], unlike DGKd, the protein interacted with SMSr. Notably, the interaction occurred between DGKf-CD-b and SMSr-SAMD (Figs 2 and 3). We searched for a SAMD-like region in CD-b of DGKf. However, such a region was not found. Because SAMD has two interfaces to form oligomer structures [47], DGKf-CD-b may interact with another interface of SMSr-SAMD, which is different from the SMSr-SAMD-DGKd-SAMD interface. However, the binding mechanisms between DGKf-CD-b and SMSr-SAMD are still unclear.
Although SMSr is a DG-generating enzyme, its CPES activity is very low [43]. We recently found that SMSr has high PAP, PI-PLC, PE-PLC, and PC-PLC activities, which produce DG, instead of CPES activity [43]. SMS1 generates DG and sphingomyelin through the transfer of phosphocholine from PC to ceramide [36,37]. DG is known to quickly diffuse across the lipid bilayer by flip-flop [49]. Therefore, it is considered that the DG generated by SMSr and SMS1 immediately transverses the Golgi and endoplasmic reticulum membranes from the lumen side to the cytosol leaflet and, consequently, is supplied to DGKf, which exists in the cytoplasm, as illustrated in Fig. 6. Moreover, DGKf (https://www.proteinatlas. We previously demonstrated that DGKd-SAMD associates with SMSr-SAMD and that SMSr activates DGKd [42]. See Results and Discussion for the full description. MG-PLC, multiglycerophospholipid PLC hydrolase [43]. org/ENSG00000149091-DGKZ/tissue), SMSr (https:// www.proteinatlas.org/ENSG00000156671-SAMD8/tissue), and SMS1 (https://www.proteinatlas.org/ENSG0000019 8964-SGMS1/tissue) are ubiquitously expressed in a variety of tissues [50]. These results indicate that DGKf and SMSr/SMS1 can functionally link to each other. In summary, in the present study, we demonstrated for the first time that DGKf interacts with SMS1 and SMSr but not SMS2 (Fig. 6). DGKd also associates with SMSr via their SAMDs [42] (Fig. 6). Therefore, it is likely that DGK isozymes and SMS isozymes form a complex network. Intriguingly, DGKf interacts with SMS1 and SMSr in different manners. These data suggest that SMSr and SMS1 are promising candidates for DG supply enzymes upstream of DGKf and that they compose novel and distinct DG-signaling pathways. However, further studies will be required to analyze whether SMS1 and SMSr are functionally linked to DGKf. Moreover, we need to find candidates for DG supply enzymes upstream of other isozymes (eight isozymes: a, b, c, g, j, e, ι, and h).

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article. Fig. S1. Sequence alignment of SMSr-SAMD and SMS1-SAMD. (A) Sequence alignment of SMSr-SAMD (aa 12-78) and SMS1-SAMD (aa 7-70). Sequence alignment was created using Clustal Omega provided by EMBL's European Bioinformatics Institute (EMBL-EBI). Compared with SMSr-SAMD, white letters on a black background indicate fully conserved residues, and black letters on a gray background indicate strongly similar residues. (B) Amino acid identities between the SAMDs of SMSr and SMS1. Amino acid identity and similarity were determined using Pairwise Sequence Alignment provided by the European Molecular Biology Open Software Suite (EMBOSS). . Multiple sequence alignment was created using Clustal Omega provided by EMBL's European Bioinformatics Institute (EMBLEBI). Compared with SMS1-CT, white letters on a black background indicate fully conserved residues, and black letters on a gray background indicate strongly similar residues. (B) Amino acid identities between the C-terminal regions of SMS1, SMS2, and SMSr. Amino acid identity and similarity were determined using Pairwise Sequence Alignment provided by the European Molecular Biology Open Software Suite (EMBOSS).