Expression and purification of human diacylglycerol kinase α from baculovirus-infected insect cells for structural studies

Bacterial expression and purification of DGKα-CD

Multiple-constructs approach with different N- and C-terminal boundaries (Gräslund et al., 2008), and several N-terminal fusion-tags (GST, MBP, and small ubiquitin-like modifier (SUMO)) was applied for bacterial expression of DGKα-CD.

To prepare GST-fused constructs, the DNA sequences of DGKα-CD (S332–G722, D344–G722, D369–G722, S332–S735, D344–S735, D369–S735), flanked by BamHI and SalI restriction sites were amplified by PCR from the full-length cDNA for human DGKα, inserted into a pGEX-4T-2 vector (GE Healthcare Life Science, Little Chalfont, UK) and the resulting plasmids were used to transform Escherichia coli strain Rosetta2 (DE3) (Novagen, Madison, WI, USA). The protein construct contained a thrombin-cleavable GST-tag before the DGKα-CD sequence. Cells were cultured in LB media at 37 °C until OD₆₀₀ reached 0.6–0.8. Expression of the recombinant protein was then induced by adding 0.5 mM isopropyl β-D-thiogalactopyranoside (IPTG), and the bacterial culture was continued at 16 °C for overnight. Bacteria harvested by centrifugation were suspended in a lysis buffer (50 mM sodium phosphate, pH 8.0, containing 500 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 1 mM dithiothreitol) and lysed by sonication on ice. Protease inhibitors (20 μg/mL aprotinin, 20 μg/mL leupeptin, 20 μg/mL pepstatin, 1 mM soybean trypsin inhibitor) were added immediately before sonication. To evaluate expression and solubility of the expressed proteins, soluble and insoluble fractions were separated by centrifugation at 15,000×g for 10 min and subjected to SDS-PAGE (10%) followed by Coomassie Brilliant Blue (CBB) staining and immunoblot analysis using anti-GST monoclonal antibody (B-14; Santa Cruz Biotechnology, Dallas, TX, USA). The immunoreactive bands were visualized using peroxidase-conjugated anti-mouse IgG antibodies (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) and the ECL Western blotting detection system (GE Healthcare Life Science, Little Chalfont, UK).

For Sumo-tag, the DNA sequences of DGKα-CD (S332–G722, D344–G722, D369–G722, S332–S735, D344–S735, D369–S735) were cloned via NdeI/SalI sites into pSUMO vector. In addition, three constructs with additional two glutamic acids at the C-terminus (S332–S735EE, D344–S735EE, D369–S735EE) were also cloned into the pSUMO vector. The resulting recombinant protein contains an N-terminal His-tagged Sumo domain, followed by a Sumo-specific protease (Ulp1) cleavage site before the DGKα-CD sequence. His_×6-Sumo-DGKα-CD was expressed in E. coli Rosetta2 (DE3) cells by induction with 0.1 mM IPTG at 16 °C for overnight. After a small-scale expression and solubility test using anti-His_×6 monoclonal antibody (9C11; Wako, Osaka, Japan), large scale expression of the construct D344–S735EE was conducted. After cell-lysis and centrifugation, Ni-affinity chromatography was applied to purify His_×6-Sumo-DGKα-CD (D344–S735EE). The column was washed with 50 mM Tris–HCl, pH 8.0, 500 mM NaCl, 10 and 50 mM imidazole, and the bound proteins were eluted with 300 mM imidazole. An elution fraction containing His_×6-Sumo-DGKα-CD was concentrated using a centrifugal filter (Amicon Ultra-15; Millipore, Burlington, MA, USA), and applied to a Superdex 75 16/60 column for size exclusion chromatography purification.

Maltose binding protein-fused DGKα-CD constructs were prepared by cloning the sequences of DGKα-CD (S332–G722, D344–G722, D369–G722, S332–S735, D344–S735, D369–S735, S332–S735EE, D344–S735EE, D369–S735EE) into a pMAL-c2X vector (New England Biolabs, Ipswich, MA, USA) via BamHI/SalI sites. E. coli strain Rosetta2 (DE3) transformed with the plasmids. Protein expression was induced with 0.1 mM IPTG and bacterial cells were then incubated at 16 °C for overnight. After cell-lysis and centrifugation, the expression and solubility test was conducted by subjecting soluble and insoluble fractions on SDS-PAGE (10%) followed by CBB staining and immunoblot analysis using anti-MBP antibody (Sc-13564; Santa Cruz Biotechnology, Dallas, TX, USA). The construct MBP-DGKα-CD (D369–S735) was expressed in 2 L of LB medium and the MBP-fused protein was purified by affinity chromatography on amylose resin (New England Biolabs, Ipswich, MA, USA). The affinity column was washed with a buffer, 20 mM Tris–HCl, pH 7.4, 200 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 1 mM EDTA, and the bound proteins were eluted with a buffer containing 10 mM maltose. Fractions were subjected to SDS-PAGE (10%) and analyzed by CBB staining and immunoblotting using anti-MBP antibody and anti-DGKα antibody (Yamada, Sakane & Kanoh, 1989).

Expression of DGKα in insect cells using baculovirus expression vector system

The construct of DGKα-CD (D364–S735) or full-length DGKα with N-terminal His_×6 tag was PCR-amplified and cloned into the pOET3 vector (Oxford Expression Technologies, Oxford, UK) via SalI/NotI sites. The resulting DNA sequences were verified to be correct by DNA sequencing. The flashBAC system (Oxford Expression Technologies, Oxford, UK) was used to generate a recombinant baculovirus and the virus stock was amplified by several rounds of infection of Sf9 cells cultured in Sf-900 II serum free medium (Invitrogen, Carlsbad, CA, USA) at a low multiplicity of infection (MOI). Plaque assays were performed to determine titers of the amplified virus stocks. Both DGKα-CD and full-length DGKα were expressed in Sf9 cells by infecting the cells (at 2 × 10⁶ cells/mL) with the baculovirus stock at MOI of 2. Cells were cultured at 28 °C with shaking for 24 h and pelleted by centrifugation at 1,500×g, 4 °C for 20 min and washed with sterile phosphate buffered saline before storage at −80 °C.

Purification of DGKα expressed in insect cells

The cell pellets were thawed and suspended in a lysis buffer containing 50 mM Tris–HCl, pH 8.0, 0.5M NaCl, 20 mM imidazole, 20% glycerol, 5 mM CaCl₂, 5 mM MgCl₂, 5 mM β-mercaptoethanol, 1% Nonidet P-40 (NP-40), 5 mM adenosine 5’-diphosphate (ADP), 5 U/mL benzonase (EMD Millipore, Burlington, MA, USA) and a EDTA-free protease inhibitor cocktail tablet (Roche, Basel, Switzerland). The cell suspension was lysed by sonication on ice followed by centrifugation at 25,000×g, 4 °C for 1 h. The supernatant was incubated with 2 mL of Ni-NTA agarose (Qiagen, Venlo, The Netherlands) for 2 h at 4 °C, and then the mixture was packed into a column by gravity. The column was washed with wash buffer 1 (50 mM Tris–HCl, pH 8.0, 0.5M NaCl, 20 mM imidazole, 20% glycerol, 5 mM CaCl₂, 5 mM MgCl₂, 5 mM β-mercaptoethanol, 1% NP-40) and wash buffer 2 (50 mM Tris–HCl, pH 8.0, 0.5M NaCl, 20 mM imidazole, 20% glycerol, 5 mM CaCl₂, 5 mM MgCl₂, 5 mM β-mercaptoethanol and 10% ethanol). Subsequently, the bound proteins were eluted with step-wise increase of imidazole concentration (50, 100, and 200 mM) in a buffer consisting of 50 mM Tris–HCl, pH 8.0, 0.5M NaCl, 20 mM imidazole, 20% glycerol, 5 mM CaCl₂, 5 mM MgCl₂, 5 mM β-mercaptoethanol. Collected fractions were analyzed by SDS-PAGE with Coomassie blue staining and immunoblot analysis using anti-DGKα antibody (Yamada, Sakane & Kanoh, 1989).

Fractions containing full-length DGKα were further purified using size-exclusion chromatography on a Superdex 200 column 16/60 equilibrated with 20 mM Tris–HCl, pH 7.4 with 200 mM NaCl, 3 mM CaCl₂, 3 mM MgCl₂, 0.5 mM dithiothreitol, and 5% glycerol. Resulting fractions were analyzed by SDS-PAGE followed by Coomassie blue staining. Protein quantification was done by Bradford assay or using the extinction coefficient, Ε_0.1% = 1.14. Gel filtration standards (Bio-Rad, Hercules, CA, USA) containing thyroglobulin (670 kDa), γ-globulin (158 kDa), ovalbumin (44 kDa), myoglobin (17 kDa), and vitamin B12 (1.3 kDa) were used to determine the molecular mass of proteins.

In vitro DGKα activity assay

Activity of full-length DGKα was determined using the octyl-β-d-glucoside mixed micelle assay combined with the ADP-Glo^™ kinase assay kit (Promega, Madison, WI, USA), as previously described (Sato et al., 2013; Liu et al., 2016). Briefly, the substrate micelle mixture containing 50 mM n-octyl-β-d-glucoside (Dojindo, Kumamoto, Japan), 10 mM (27 mol%) phosphatidylserine (PS; Sigma-Aldrich, St. Louis, MO, USA), 2 mM (5.4 mol%) 1,2-dioleolyl-sn-glycerol (DG; Sigma-Aldrich, St. Louis, MO, USA), 0.2 mM adenosine 5’-triphosphate (ATP) in a final buffer consisting of 50 mM MOPS, pH 7.4, 100 mM NaCl, 20 mM NaF, 10 mM MgCl₂, 1 μM CaCl₂ was mixed with 5 μL of purified DGKα to initiate enzymatic reaction. The reaction mixtures were incubated at 30 °C for 30 min. Subsequently, 25 μL of ADP-Glo reagent was added and incubated at room temperature for 40 min to terminate the enzyme reaction and deplete the remaining ATP. Kinase Detection Reagent (50 μL) was then added to convert the ADP produced from the kinase reaction into ATP for a luciferase/luciferin reaction. The reaction was performed at room temperature for 40 min and the luminescence from the luciferase/luciferin reaction was measured with a GloMax microplate reader (GloMax; Promega). A standard curve for ADP was generated by fitting a various concentration of ADP ranging from 25 to 200 μM and the corresponding luminescence signals relative luminescence unit by linear regression, and was used to convert the luminescence intensities from DGKα reaction into ADP concentrations. To determine kinetic constants, the activity assay was performed under a series of concentrations of ATP (20 μM–1 mM) and DG (0–5.4 mol%), respectively. DGKα purified by size exclusion chromatography was added to 100 ng for each reaction and the assays were done in triplicate for each ATP or DG concentrations. The K_m value was obtained by fitting the kinase activity of DGKα with the Michaelis–Menten equation using Prism 7 (GraphPad Software, La Jolla, CA, USA). To test the calcium dependency of DGKα activity, the enzyme activity was measured under the conditions containing either EGTA (3 mM) or CaCl₂ (0.6 mM).

Inhibitor activity assay

Inhibitory activity of a previously identified inhibitor, CU-3 (Liu et al., 2016), against DGKα was measured with the octyl-β-d-glucoside mixed micelle assay followed by the ADP-Glo assay. A concentration series of CU-3 (0.02–10 μM) was incubated with the purified DGKα for 30 min at room temperature before adding to a reaction mixture for the assay. Half maximal inhibitory concentration (IC50) was determined by fitting the CU-3 dependent decrease of DGKα activity with the variable slope model in Graphpad Prism 7 software.

Circular dichroism spectroscopy

Circular dichroism spectrum were recorded at ambient conditions between 190 and 250 nm on a Jasco J-805 spectrometer (JASCO Corporation, Tokyo, Japan) using a cell with path length of 0.2 mm, 20 nm/min scan speed and a bandwidth of 1 nm. DGKα was prepared at 0.32 mg/mL (3.75 μM) in 20 mM Tris–HCl buffer, pH 7.5, 10 spectra were averaged and a spectrum obtained for the buffer was subtracted. Spectral data were analyzed using the program Contin-LL (Provencher & Glöckner, 1981) suited in the DICHROWEB platform (Whitmore & Wallace, 2004).

A full-length form of DGKα was expressed in baculovirus-infected insect cells and purified as a monomer

We have previously reported that DGKα-CD possess enzymatic activity comparable to that of the full-length enzyme when expressed in COS-7 cells (Sakane et al., 1996), indicating that its substrate (ATP and DG) binding sites locate in the CD and DGKα-CD is an essential target for inhibitor development. Full-length DGKα also contains cysteine-rich C1 domains (Fig. 1A), which might be detrimental for correct folding in heterologous expression hosts. Therefore, we have first attempted to express DGKα-CD in E. coli by revamping the previous approach by Petro & Raben (2013). In addition to N-terminal GST and MBP-tags, which were previously utilized (Petro & Raben, 2013), we have used Sumo domain fusion-tag for the enhancement of expression and solubility (Butt et al., 2005; Marblestone et al., 2006). To further increase the chance for expression of soluble proteins, we have also applied a multiple-construct approach (Gräslund et al., 2008) to prepare DGKα-CD constructs which have different N- and C-terminal boundaries (S332–G722, D344–G722, D369–G722, S332–S735, D344–S735, D369–S735). Each of those constructs was fused with the GST, MBP, and Sumo-tags. Despite our efforts, those constructs resulted in either insoluble inclusion body formation (with GST-tag), or insufficient translation and proteolytic degradation (with MBP-tag), or soluble microscopic aggregation (with Sumo-tag) (Fig. S1).

To circumvent the difficulty associated with bacterial expression system, we have used baculovirus-infected Sf9 cells to produce DGKα-CD. The construct of DGKα-CD (D364–S735) with N-terminal His_×6 tag was cloned into pOET3 transfer vector harboring the late AcMNPV p6.9 promoter, which provides earlier expression compared to the polyhedrin promotor. The recombinant DGKα-CD was expressed in cultured insect cells using the flashBAC baculovirus vector expression system, and subsequently purified from cell-lysates using Ni-affinity chromatography (Fig. S2A). Following size-exclusion chromatography on a Superdex 200, however, demonstrated that DGKα-CD formed soluble aggregates eluting in the void volume of the column (Fig. S2B).

In our early studies, a native form of full-length DGKα has been purified from porcine thymus and this full-length form was found to be catalytically competent (Sakane, Yamada & Kanoh, 1989; Sakane et al., 1991). We therefore set to produce full-length DGKα (aa 1–735) using the same baculovirus expression system used for DGKα-CD. As expected, the vast majority of DGKα remained in soluble form after cell lysis, as shown by immunoblot analysis (Fig. 1B). Ni-affinity chromatography was conducted to purify DGKα from the cell lysis supernatant, and relatively pure DGKα was eluted in fractions containing 50 and 100 mM imidazole (Fig. 1C). To further purify DGKα, we next performed size-exclusion chromatography on a Superdex 200 column. Because DGKα bears calcium-binding EF-hand motifs and a magnesium ion was predicted to bind to the CD (Abe et al., 2003), we added 3 mM CaCl₂ and 3 mM MgCl₂ in the equilibration buffer. DGKα eluted as a single peak at 73.5 mL retention volume (Fig. 1D), which corresponds to the molecular mass of 77 kDa, based on a calibration curve obtained with molecular mass standard proteins. This result indicates that DGKα exists as a monomer in solution. DGKα was purified to near homogeneity (Fig. 1E) and the yield was approximately 1.3 mg per one L of Sf9 cell culture.

Kinase activity assay and inhibitory assay for the purified DGKα

To test whether the purified DGKα is catalytically active, we conducted the octyl-β-d-glucoside mixed micelle assay combined with a luminescence-based assay that measures ADP produced in a kinase reaction (Sato et al., 2013; Liu et al., 2016). DGKα purified from the size-exclusion chromatography was found to exhibit kinase activity with peak fractions having the maximum activity (Fig. 2A). We have previously demonstrated that DGKα activity, which has been purified from porcine thymus or expressed in COS-7 cells, is enhanced by Ca²⁺ binding to its two N-terminal EF-hand motifs (Sakane et al., 1990, 1991; Yamada et al., 1997). As predicted, the purified DGKα exhibited significantly reduced activity when the bound calcium ions were chelated with 3 mM EGTA (Fig. 2B). Furthermore, no significant changes of the activity were observed after storage of the purified DGKα at 4 °C for at least 3 months.

We also determined the kinetic parameters of DGKα for ATP and DG to assess the catalytic properties of the purified DGKα. ATP-dependent increase of the kinase activity was observed (Fig. 3A) and the K_m value was determined to be 0.24 ± 0.03 mM (Table 1), comparable with those obtained with DGKα from porcine thymus (0.1 mM) (Sakane et al., 1991) or DGKα expressed in COS-7 cells (0.1–0.25 mM) (Sato et al., 2013; Liu et al., 2016). The activity was also increased in a DG-concentration dependent manner (Fig. 3B) and the K_m value of 1.1 ± 0.21 mol% (Table 1) was consistent with those from our previous studies (3.3 mol% with DGKα purified from porcine thymus (Sakane et al., 1991), 1.9–3.4 mol% with DGKα expressed in COS-7 cells (Sato et al., 2013; Liu et al., 2016)). For both cases, compared to our previous study using crude lysates of mammalian cells (Sato et al., 2013; Liu et al., 2016), the relative activity increased nearly 50-fold when the purified DGKα was used. Furthermore, the kinase activities of our purified DGKα (1 to 2 nmol/min/μg) is comparable to those obtained with DGKα from porcine thymus (2.4 nmol/min/μg) (Sakane et al., 1991). These results demonstrate that the purified DGKα is in a fully functional state and stable during purification and storage.

We next measured the inhibitory activity of CU-3, a previously identified DGKα inhibitor (Liu et al., 2016). CU-3 is an ATP competitive inhibitor with an IC₅₀ value of 0.6 μM. Consistent with this, CU-3 inhibited the activity of DGKα in a concentration-dependent manner with an IC₅₀ value of 0.34 ± 0.1 μM (Fig. 4).

Structural characterization of the purified DGKα

We also found that DGKα solution could be concentrated using a centrifugal filter without any significant loss of the protein. Concentrated DGKα remained as a monomer as demonstrated by a size-exclusion chromatography (Fig. S3). Using the concentrated DGKα (0.32 mg/mL), we characterized the secondary structure using circular dichroism spectroscopy. The circular dichroism spectrum of DGKα and following analysis indicates that DGKα is well-folded and contains certain amounts of α-helical (18.9%) and β-strand (27.4%) structures (Fig. 5), further demonstrating that the expression of a full-length DGKα, not a solo CD, in the baculovirus-infected insect cells is suitable for producing a natively folded and active form of DGKα.