Diacylglycerol kinase δ and sphingomyelin synthase–related protein functionally interact via their sterile α motif domains

The δ isozyme of diacylglycerol kinase (DGKδ) plays critical roles in lipid signaling by converting diacylglycerol (DG) to phosphatidic acid (PA). We previously demonstrated that DGKδ preferably phosphorylates palmitic acid (16:0)- and/or palmitoleic acid (16:1)-containing DG molecular species, but not arachidonic acid (20:4)-containing DG species, which are recognized as DGK substrates derived from phosphatidylinositol turnover, in high glucose-stimulated myoblasts. However, little is known about the origin of these DG molecular species. DGKδ and two DG-generating enzymes, sphingomyelin synthase (SMS) 1 and SMS-related protein (SMSr), contain a sterile α motif domain (SAMD). In this study, we found that SMSr–SAMD, but not SMS1–SAMD, co-immunoprecipitates with DGKδ–SAMD. Full-length DGKδ co-precipitated with full-length SMSr more strongly than with SMS1. However, SAMD-deleted variants of SMSr and DGKδ interacted only weakly with full-length DGKδ and SMSr, respectively. These results strongly suggested that DGKδ interacts with SMSr through their respective SAMDs. To determine the functional outcomes of the relationship between DGKδ and SMSr, we used LC-MS/MS to investigate whether overexpression of DGKδ and/or SMSr in COS-7 cells alters the levels of PA species. We found that SMSr overexpression significantly enhances the production of 16:0- or 16:1-containing PA species such as 14:0/16:0-, 16:0/16:0-, 16:0/18:1-, and/or 16:1/18:1-PA in DGKδ-overexpressing COS-7 cells. Moreover, SMSr enhanced DGKδ activity via their SAMDs in vitro. Taken together, these results strongly suggest that SMSr is a candidate DG-providing enzyme upstream of DGKδ and that the two enzymes represent a new pathway independent of phosphatidylinositol turnover.

Cabukusta et al. (22,23) recently pointed out that the SAMD of sphingomyelin synthase (SMS)-related protein (SMSr/ SAMD8) (Fig. 1C), a family member of SMS, is primarily structurally similar to the DGK␦-SAMD and demonstrated that SMSr-SAMD formed a homodimer, which is a crucial determinant of the subcellular localization of SMSr. SMSr is a sixtransmembrane protein in the endoplasmic reticulum (ER), which generates DG by utilizing phosphatidylethanolamine (PE) and ceramide (24).
In this study, we investigated the interaction and functional relationship between DGK␦ and SMSr because both proteins possess SAMDs, which are structurally similar to each other. Intriguingly, we found that DGK␦2 interacted and was functionally linked with SMSr via their SAMDs. Therefore, it is possible that SMSr is a DG-providing enzyme upstream of DGK␦ and that these enzymes comprise a new pathway independent of PI turnover.

Interaction and functional linkage between DGK␦ and SMSr
pressed in COS-7 cells, and the cell lysates were used for a co-immunoprecipitation analysis using anti-V5 antibody. We found that 3ϫFLAG-DGK␦2 was co-immunoprecipitated with SMSr-V5, but not with SMS1-V5, SMS2-V5, or V5 alone (Fig. 4, A and B), indicating that DGK␦2 associated with SMSr in cells.
We next examined the effect of the deletion of the SMSr-SAMD on the interaction between SMSr and DGK␦2 by co-immunoprecipitation analysis. We found that 3ϫFLAG-DGK␦2 and either V5-tagged full-length SMSr (SMSr-V5) or SAMD-deleted SMSr (SMSr-⌬SAMD-V5) ( Fig. 1) were co-expressed in COS-7 cells. The V5-tagged proteins in cell lysates were immunoprecipitated with the anti-V5 antibody. Compared with full-length SMSr, the amount of DGK␦2 co-precipitated with the SMSr-⌬SAMD was significantly reduced (by about 75%) (Fig. 4, C and D). We also examined the effect of the

DGK␦2 enzyme activity was increased by SMSr through their SAMDs in vitro
We next examined whether DGK␦2 activity was enhanced by complex formation with SMSr via their SAMDs. First, 6ϫHistagged SMSr, SMSr-⌬SAMD, and DGK␦2 (Fig. 1, B and D) were expressed using COS-7 cells and purified by Ni-affinity chromatography (Fig. 7A). We tested whether the activity of 6ϫHis-tagged DGK␦2 was enhanced by SMSr in vitro in the presence of DG (in the absence of PE and ceramide). Interest-ingly, when adding purified SMSr to purified DGK␦2 (stoichiometry, ϳ3:1), DGK activity was significantly increased (Fig.  7B). However, when adding SMSr-⌬SAMD to DGK␦2, DGK activity was not significantly increased (Fig. 7C), suggesting that SMSr activated DGK␦2 in a SAMD-dependent manner. To test whether SMSr-SAMD alone enhanced DGK␦2 activity, we added GST-SMSr-SAMD (Fig. 1D), which was expressed in Escherichia coli and purified by GSH-affinity chromatography, to purified DGK␦2. The activity of DGK␦2 was significantly increased by the addition of the GST-SMSr-SAMD (Fig. 7D). These results strongly suggested that, in addition to DG supply, SMSr directly enhanced DGK␦2 activity via the SMSr-SAMD.
The results of Fig. 7D implied that the disturbance of homodimerization of DGK␦2 by SMSr-SAMD enhanced its activity. To verify this, DGK␦1, DGK␦2, and their SAMD deletion mutants were expressed in COS-7 cells, and DGK activities in cell lysates were measured using a luminescence-based (ADP-glo) kinase assay ( Fig. 8) (26). The DGK activities of SAMD deletion mutants of DGK␦ (both isoform 1 and 2) were significantly increased compared with their WT enzymes (Fig.  8B). These results suggested that the SAMD suppressed the activity of DGK␦ and that the SMSr-SAMD increased DGK␦2 activity by disrupting its homo-oligomers.

Discussion
A DG-providing pathway for DGK␦, which would play an important role in type 2 diabetes pathogenesis, has been unclear. In this study, we, for the first time, demonstrate that DGK␦2 was associated with SMSr, a DG-generating enzyme, via their SAMDs (Figs. 3 and 4). Moreover, the overexpression of SMSr significantly enhanced the production of PA in COS-7 cells overexpressing DGK␦ (Fig. 5). Thus, these data suggested the possibility that SMSr acts as a DG-providing enzyme upstream of DGK␦2 (Fig. 9). Therefore, it is likely that these enzymes compose a new pathway independent of phosphatidylinositol turnover.
We demonstrated that DGK␦-SAMD directly interacted with SMSr-SAMD (Figs. 3 and 9A). Moreover, we showed that full-length DGK␦2 and SMSr associated with each other and

Interaction and functional linkage between DGK␦ and SMSr
that their SAMDs mainly contributed to the association (Fig. 4). SAMD, which is an evolutionally conserved protein-protein interaction domain, is known to occur in a wide range of proteins (more than 200) (27). SAMDs generally form homodimers between the same proteins and between closely-related family proteins. However, in some cases, a heterodimer between completely different proteins, such as between Connector enhancer of KSR (kinase suppressor of Ras)-SAMD and Hyphen-SAMD (28) and between Src homology 2 domain-containing phosphoinositide-5-phosphatase 2-SAMD and Ephrin A2 receptor-SAMD are formed (29). In this study, we identified a new combination of heterodimers between completely different proteins, DGK␦ and SMSr, through their SAMDs.

Interaction and functional linkage between DGK␦ and SMSr
DGK␦-SAMD were critical for forming a heteromeric complex with the SMSr-SAMD, we prepared single-point mutants of the DGK␦-SAMD/D1183G, V1192E, G1193D, and K1196E. Co-immunoprecipitation analysis showed that the co-sedimented level of only the DGK␦-SAMD-G1193D mutant tended to be decreased, but there was no significant difference. 3 This result suggested that Gly-1193 was one of the key residues for the DGK␦-SAMD/SMSr-SAMD heteromeric complex. However, it was possible that other residues were also important for the heterodimerization.
To determine the functional relationship between DGK␦ and SMSr, we investigated changes in the amounts of PA molecular species in COS-7 cells overexpressing DGK␦ and/or SMSr using LC-MS/MS (Fig. 5). We found that the overexpression of SMSr significantly (more than 20%) enhanced the production of palmitic acid ( Table 1). Therefore, it was possible that SMSr acted upstream of DGK␦ and supplied DG to the enzyme (Fig. 9B). Moreover, several lines of evidence supported this hypothesis.  In this study, we demonstrated that DGK␦-SAMD directly interacts with SMSr-SAMD (A). It is possible that SMSr acts upstream of DGK␦ and supplies DG to the enzyme. There is the possibility that SMSr possesses not only ceramide phosphoethanolamine (CPE) synthase activity but also PE-specific phospholipase C activity. 3 B, SMSr-SAMD may disrupt the homodimerization of DGK␦2, which attenuates its DGK activity and, consequently, enhances DGK␦2 activity (C). Alternatively, SMSr, which is a transmembrane protein, may recruit DGK␦2 to the DG-containing membrane via their SAMDs (C). Cer, ceramide; PE; phosphatidylethanolamine; SMSr, sphingomyelin synthase-related protein.

Interaction and functional linkage between DGK␦ and SMSr
First, the fatty acid compositions of PA species produced by DGK␦ ( Fig. 5 and Table 1) were similar to those produced under physiological conditions. We previously revealed that the suppression of DGK␦2 expression decreased high-glucoseinduced production of 16 Table 2).
Third, both SMSr-SAMD and DGK␦2 (DGK␦2-SAMD) exist in the cytoplasm and are able to associate with each other as shown in Fig. 9. Topological analysis using the N-glycosylation gel-shift assay showed that the N-terminal SAMD is cytosolic (30). DGK␦2 is also cytosolic (21,31). Therefore, it was possible that the DGK␦2-SAMD interacted with SMSr via the SMSr-SAMD in the cytoplasm. In addition, DG is known to quickly diffuse across the lipid bilayer by the flip-flop mechanism (32, 33). Therefore, it is likely that the DG produced by SMSr immediately transverses the ER membrane from the lumen side to the cytosol leaflet and, consequently, is provided to DGK␦, as illustrated in Fig. 9B.
Fourth, SMSr and DGK␦2 were distributed in the same cells and tissues. DGK␦ is highly expressed in skeletal muscle (31), and SMSr is also reported to be abundantly expressed in skeletal muscle (34). It was revealed that the mRNA level of SMSr was markedly higher than DGK␦ in skeletal muscle-derived C2C12 myoblast cells. 3 Taken together, it is possible that endogenous DGK␦2 is able to interact with SMSr in skeletal muscles and myoblast cells. In addition to skeletal muscle, DGK␦ was reported to be broadly expressed in multiple mouse tissues, in particular, and to be highly expressed in the brain and testis (35). Moreover, it was reported that SMSr is also widely expressed, with the highest expression in testis, brain, kidney, and pancreas (34). Thus, the expression pattern of DGK␦2 is similar to that of SMSr, suggesting that DGK␦2 and SMSr act in the same cells and tissues. Overall, these results allow us to speculate that DGK␦2 functionally links with SMSr, which acts upstream of DGK␦2 as a DG-supply enzyme. However, further studies are needed to clarify the linkage in more detail.
SMSr is known to show only weak ceramide phosphoethanolamine synthase activity (24). Interestingly, we unexpectedly found that purified SMSr generated DG even in the absence of ceramide when glycerophospholipid, such as PE or phosphatidylcholine (PC) alone, was added as a substrate. 3 Therefore, it is possible that, in addition to ceramide phosphoethanolamine synthase, SMSr is able to act as a PE-and PC-phospholipase C (PLC) to produce DG. It is interesting to further verify this possibility. We previously reported the possibility that D609sensitive enzymes, SMS and PC-PLC, are candidates for the DG-supply enzyme to DGK␦ (15). Thus, it is noteworthy that D609 partly (ϳ50%) inhibited the PLC activity of SMSr. 3 The results of Fig. 7 suggested that SMSr activated DGK␦2 in a SAMD-dependent manner. We previously reported that DGK2-SAMDs formed homodimers and suppressed the catalytic activity of DGK2 (19). We confirmed that the DGK␦-SAMD also inhibited the catalytic activity of DGK␦2 by homodimerization (Fig. 8). Therefore, it was possible that SMSr-SAMD disrupted the homodimerization of DGK␦2, which attenuated its DGK activity and, consequently, enhanced the DGK␦2 activity (Fig. 9C). Alternatively, SMSr, which is a transmembrane protein, may recruit DGK␦2 to DG-containing micelles via their SAMDs (Fig. 9C).
In summary, this study, for the first time, showed that DGK␦2 and SMSr formed a heteromeric complex via their SAMDs. Moreover, the SAMD of SMSr activated DGK␦2. These results allow us to propose the alternative DG metabolic pathway "PE 3 SMSr 3 DG 3 DGK␦2 3 PA" (Fig. 9). This pathway metabolized palmitic acid (16:0) and/or palmitoleic acid (16:1)-containing glycerolipids, but it did not utilize arachidonic acid (20:4)-containing glycerolipids derived from PI turnover. Therefore, it is likely that this new pathway is independent of PI turnover, although the substrate of DGK is generally thought to be derived from the PI-dependent pathway. The decrease of DGK␦ protein and DG accumulation is known to regulate the pathogenesis of type 2 diabetes (14). Future studies exploring the mechanism by which SMSr-dependent DG supply and the linkage between SMSr and DGK␦ are regulated will provide further insights into how type 2 diabetes is exacerbated. Antibodies-Rabbit polyclonal anti-His-tag antibody (PM032) and anti-DDDDK (FLAG)-tag antibody (PM020) were obtained from Medical and Biological Laboratories (Nagoya, Japan). Mouse monoclonal anti-V5 antibody (clone E10/ V4RR, catalog no. MA5-15253) and Alexa Fluor 594conjugated goat anti-mouse IgG (A-11005) were obtained from Thermo Fisher Scientific (Waltham, MA). Mouse monoclonal anti-FLAG-tag antibody (F1804) was purchased from Sigma. Mouse monoclonal anti-GFP antibody (sc-9996) and anti-GST antibody (sc-138) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). A peroxidase-conjugated goat anti-mouse IgG antibody was purchased from Bethyl Laboratories (Montgomery, TX). A peroxidase-conjugated goat anti-rabbit IgG antibody (11-036-045) was obtained from Jackson ImmunoResearch (West Grove, PA). We also used rabbit polyclonal anti-DGK␦ antibody (WB-1), which was prepared previously (31).

Interaction and functional linkage between DGK␦ and SMSr
Mice C57BL/6N mice were obtained from SLC Japan, Inc. (Shizuoka, Japan). Tissues were removed immediately after decapitation. All procedures using C57BL/6N mice were conducted in compliance with the National Institutes of Health: Guide for the Care and Use of Laboratory Animals and the Institutional Animal Care and Use Committee of the University of Chiba approved the protocol (permission number: .

Reverse transcriptase-PCR
Total RNA was isolated from mouse testis using QIAzol Lysis Reagent (Qiagen, Venlo, Netherlands) and Direct-zol TM RNA Miniprep (Zymo Research, Irvine, CA) according to the protocol from the manufacturer. The cDNA was generated using transcriptor reverse transcriptase (Roche Diagnostics, Mannheim, Germany), as described previously (36).

Expression and purification of 6؋His-TF-fusion proteins
BL21 cells that harbored expression plasmids encoding 6ϫHis-TF alone or 6ϫHis-TF-DGK␦2-SAMD were grown at 37°C in LB medium supplemented with 100 g/ml ampicillin until the cell density reached an OD 600 of 0.45. The cells were incubated at 16°C for 24 h in the presence of 0.1 mM IPTG and then were harvested by centrifugation. The pellets were lysed by sonication on ice with a lysis buffer (50 mM sodium phosphate, pH 8.0, containing 500 mM NaCl, 1 mM PMSF, 1 mM DTT, 20 mg/ml aprotinin, 20 mg/ml leupeptin, 20 mg/ml pepstatin, and 1 mM soybean trypsin inhibitor) followed by centrifugation (15,000 ϫ g for 40 min at 4°C). Ammonium sulfate was added gradually to the supernatant to 40% saturation and incubated at 4°C for 1 h with shaking. After centrifugation (15,000 ϫ g for 45 min at 4°C), the precipitates were removed, and the ammonium sulfate concentration in the remaining supernatants was further increased to 80% saturation. After 2 h of incubation at 4°C with continuous shaking, precipitated proteins were collected by centrifugation (15,000 ϫ g for 45 min at 4°C), resuspended in 1 ml of 50 mM sodium phosphate, pH 8.0, and dialyzed in buffer (25 mM Tris-HCl, pH 8.0, 500 mM NaCl). The dialyzed proteins were purified by affinity chromatography on a Ni-Sepharose 6 Fast Flow column (GE Healthcare). The beads were washed with 30 ml of wash buffer (50 mM imidazole, 50 mM sodium phosphate, pH 8.0, 300 mM NaCl). Subsequently, the bound proteins were eluted with elution buffer (500 mM imidazole, 50 mM sodium phosphate, pH 8.0, 300 mM NaCl). The purified proteins were dialyzed in a buffer (20 mM sodium phosphate, pH 8.0, 150 mM NaCl).

Interaction and functional linkage between DGK␦ and SMSr
and then 300 l of PBS and 20 g of 6ϫHis-TF-fused protein was added. Beads and proteins were incubated at 4°C for 2 h. Then the beads were washed with lysis buffer four times, and the proteins were eluted in 50 l of 2ϫ SDS sample buffer by incubation at 95°C for 10 min.

Cell culture and transfection
COS-7 cells were maintained on 150-mm dishes (Thermo Fisher Scientific) in Dulbecco's modified Eagle's medium (DMEM) (Wako Pure Chemicals, Tokyo, Japan) containing 10% fetal bovine serum (FBS) (Thermo Fisher Scientific), 100 units/ml penicillin G (Wako Pure Chemicals), and 100 g/ml streptomycin (Wako Pure Chemicals) at 37°C in an atmosphere containing 5% CO 2 . Cells (5 ϫ 10 5 ) were plated on poly-L-lysine-coated 60-mm dishes for immunoprecipitation analysis. For quantitation of PA and DG levels, 1 ϫ 10 6 cells were plated on 100-mm dishes. For confocal microscopy, 1.5 ϫ 10 4 cells were plated on poly-L-lysine (Sigma)-coated glass coverslips (15 mm diameter). After 24 h, plasmid cDNAs were transfected using PolyFect (Qiagen) according to the manufacturer's instruction manual. After transfection, the cells were cultured for an additional 24 h and were used for the experiments described below.

Co-immunoprecipitation analysis
COS-7 cells co-expressing FLAG-tagged and V5-tagged proteins were washed two times with PBS and lysed in a buffer containing 50 mM HEPES, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% Nonidet P-40, 1 mM PMSF, and Complete EDTA-free protease inhibitor mixture (Roche Diagnostics, Mannheim, Germany). After sonication, insoluble materials were removed by centrifugation (10,000 ϫ g for 5 min at 4°C). Anti-V5 (1.5 g) or anti-FLAG (5 l) antibody was added to the cell lysates (400 l). After a 2-h incubation at 4°C, 20 l of protein A/G PLUSagarose beads (Santa Cruz Biotechnology, catalog no. sc-2003) was added and further incubated at 4°C for 2 h. The beads were then washed five times with a wash buffer (50 mM HEPES, pH 7.4, 100 mM NaCl, 5 mM MgCl 2 , 0.1% Triton X-100, 1 mM PMSF, 1 mM EDTA, and Complete EDTA-free protease inhibitor mixture). Co-immunoprecipitated proteins were eluted in 50 l of 2ϫ SDS sample buffer by incubation at 95°C for 10 min.

Lipid extraction
Total lipids were extracted from the samples according to the method of Bligh and Dyer (43) as described previously (36,44). Briefly, 2 ml of methanol and 1 ml of chloroform were added to 700 l of sample. Internal standards (100 ng of the 14:0/14:0-PA (Avanti Polar Lipids) and 50 ng of the 14:0/14:0-DG (Cayman Chemical)) were added. To improve recovery ratio of acidic phospholipids, 100 l of 3 M HCl was added to the sample. After addition of HCl, the sample was vortexed for 30 s. After incubation for 30 min at room temperature, 1 ml of chloroform was added and vortexed for 30 s, followed by the addition of 1 ml of water and vortexing for 30 s. The sample was centrifuged at 1000 ϫ g for 10 min to separate the phases. The lower phase, containing the extracted lipids, was transferred to a new vial. The solvent-containing lipids was dried under N 2 gas, and the extracted lipids were reconstituted in 100 l of chloroform/ methanol (2:1, v/v).

Liquid chromatography
The extracted lipids (10 l) were separated on a liquid chromatography (LC) system (EXION LC, AB SCIEX, Framingham, MA). This LC system was controlled by the Analys software (AB SCIEX).

Mass spectrometry
The LC system was coupled on line to Triple Quad TM 4500 (AB SCIEX), a triple-quadrupole tandem mass spectrometer equipped with a turbo spray ionization source. The experimental conditions for detection of PA molecular species were as follows: ion spray voltage Ϫ4500 V, curtain gas 30 p.s.i., collision gas 6 p.s.i., temperature 300°C, declustering potential Ϫ160 V, entrance potential Ϫ10 V, collision energy Ϫ42 V, collision cell exit potential Ϫ11 V, ion source gas I 70 p.s.i., and ion source gas II 30 p.s.i.. The experimental conditions for detection of DG molecular species were as follows: ion spray voltage 5500 V, curtain gas 30 p.s.i., collision gas 7 p.s.i., temperature 300°C, declustering potential 60 V, entrance potential 10 V, collision energy 30 V, collision cell exit potential 6.0 V, ion source gas I 70 p.s.i., and ion source gas II 50 p.s.i.
PA and DG molecular species were detected in a multiple reaction monitoring mode. Ionized PA species ([M Ϫ H] Ϫ ) or DG species ([M ϩ NH 4 ] ϩ ) were isolated at the first quadrupole (Q1). Thereafter, a product ion of PA species (m/z 153 in negative ion mode) (48,49) or DG species (see Tables S1 and S2) was reselected at Q3 after fragmentation at Q2 by collisioninduced dissociation.
The cells (10 dishes 150 mm in diameter) were harvested and lysed by sonication on ice with lysis buffer (50 mM sodium phosphate, pH 8.0, 1% Triton X-100, 300 mM NaCl, 10 mM imidazole, 1 mM PMSF, 20 mg/ml aprotinin, 20 mg/ml leupeptin, 20 mg/ml pepstatin, and 1 mM soybean trypsin inhibitor) followed by centrifugation at 15,000 ϫ g at 4°C for 40 min. The lysate was filtered through a 0.45-m filter (Millipore). The proteins were purified by affinity chromatography on a Ni-Sepharose 6 Fast Flow column (GE Healthcare). The beads were washed once with lysis buffer, followed by washing with wash buffer (50 mM sodium phosphate, pH 8.0, 50 mM imidazole, 300 mM NaCl). 6ϫHis-tagged proteins were eluted with elution buffer (50 mM sodium phosphate, pH 8.0, 300 mM imidazole, 300 mM NaCl). The purified proteins were used for in vitro DGK assays described below.

In vitro DGK assay
DGK␦2 activity was measured using an octyl-␤-D-glucosidemixed micellar assay (26), followed by quantitation of PA levels using LC-MS/MS as described above. In brief, 100 l of purified proteins, prepared as described above, were used for the assay. The enzyme reaction was started by adding 150 l of reaction solution containing the final concentration of 50 mM MOPS, pH 7.4, 20 mM NaF, 10 mM MgCl 2 , 50 mmol/liter n-octyl-␤-Dglucoside, 1 mM DTT, 1 mM (1.96 mol %) 16:0/18:1-DG (Avanti Polar Lipids), and 0.2 mM ATP. The solution was incubated at 37°C for 2 h. After the reaction, lipids were extracted as described above. The 34:1-PA in the extracted lipids was detected by LC-MS/MS as described above. The DGK assay was also performed using the ADP-Glo kinase assay kit as described previously (26).

Immunoblot analysis
Proteins eluted in an SDS sample buffer were separated by SDS-PAGE. The separated proteins were transferred to a polyvinylidene difluoride membrane (Wako) and blocked with 5% skim milk in TBS-T (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.05% (v/v) Tween 20) for 1 h at room temperature. After washing with TBS-T, the membrane was incubated with an antibody in 5% BSA in TBS-T for 16 h at 4°C. The immunoreactive bands were then visualized using a peroxidase-conjugated IgG antibody and the enhanced chemiluminescence Western blotting detection system (GE Healthcare).

Statistical analysis
Data are represented as the means Ϯ S.D. and were analyzed using the Student's t test for the comparison of two groups or one-way analysis of variance followed by Tukey's post hoc test for multiple comparisons using GraphPad Prism 8 (GraphPad) to determine any significant differences. p Ͻ 0.05 was considered significant.