Ablation of DGKα facilitates α‐smooth muscle actin expression via the Smad and PKCδ signaling pathways during the acute phase of CCl4 ‐induced hepatic injury

Expression of α‐smooth muscle actin (αSMA) is constitutive in vascular smooth muscle cells, but is induced in nonmuscle cells such as hepatic stellate cells (HSCs). HSCs play important roles in both physiological homeostasis and pathological response. HSC activation is characterized by αSMA expression, which is regulated by the TGFβ‐induced Smad pathway. Recently, protein kinase C (PKC) was identified to regulate αSMA expression. Diacylglycerol kinase (DGK) metabolizes a second‐messenger DG, thereby controlling components of DG‐mediated signaling, such as PKC. In the present study we aimed to investigate the putative role of DGKα in αSMA expression. Use of a cellular model indicated that the DGK inhibitor R59949 promotes αSMA expression and PKCδ phosphorylation. It also facilitates Smad2 phosphorylation after 30 min of TGFβ stimulation. Furthermore, immunocytochemical analysis revealed that DGK inhibitor pretreatment without TGFβ stimulation engenders αSMA expression in a granular pattern, whereas DGK inhibitor pretreatment plus TGFβ stimulation significantly induces αSMA incorporation in stress fibers. Through animal model experiments, we observed that DGKα‐knockout mice exhibit increased expression of αSMA in the liver after 48 h of carbon tetrachloride injection, together with enhanced phosphorylation levels of Smad2 and PKCδ. Together, these findings suggest that DGKα negatively regulates αSMA expression by acting on the Smad and PKCδ signaling pathways, which differentially regulate stress fiber incorporation and protein expression of αSMA, respectively.

Expression of a-smooth muscle actin (aSMA) is constitutive in vascular smooth muscle cells, but is induced in nonmuscle cells such as hepatic stellate cells (HSCs).HSCs play important roles in both physiological homeostasis and pathological response.HSC activation is characterized by aSMA expression, which is regulated by the TGFb-induced Smad pathway.Recently, protein kinase C (PKC) was identified to regulate aSMA expression.Diacylglycerol kinase (DGK) metabolizes a second-messenger DG, thereby controlling components of DG-mediated signaling, such as PKC.In the present study we aimed to investigate the putative role of DGKa in aSMA expression.Use of a cellular model indicated that the DGK inhibitor R59949 promotes aSMA expression and PKCd phosphorylation.It also facilitates Smad2 phosphorylation after 30 min of TGFb stimulation.Furthermore, immunocytochemical analysis revealed that DGK inhibitor pretreatment without TGFb stimulation engenders aSMA expression in a granular pattern, whereas DGK inhibitor pretreatment plus TGFb stimulation significantly induces aSMA incorporation in stress fibers.Through animal model experiments, we observed that DGKa-knockout mice exhibit increased expression of aSMA in the liver after 48 h of carbon tetrachloride injection, together with enhanced phosphorylation levels of Smad2 and PKCd.Together, these findings suggest that DGKa negatively regulates aSMA expression by acting on the Smad and PKCd signaling pathways, which differentially regulate stress fiber incorporation and protein expression of aSMA, respectively.
The a-smooth muscle actin (aSMA) is involved in the contractile force of vascular smooth muscle cells.Recent evidence shows that aSMA also plays a critical role in nonmuscle cells such as hepatic stellate cells (HSCs) [1][2][3].HSCs are a resident nonparenchymal liver cell population that serves as a lipid-storing cell under physiological conditions [3].Under pathological conditions, however, they proliferate and undergo dramatic phenotypical activation through aSMA.Within HSCs, aSMA is incorporated into stress fibers, which provide an increased myofibroblast contractile force [4], thereby promoting tissue remodeling of the hepatic lobule.Therefore, HSCs are now appreciated as a remarkable plastic cell type that regulates hepatic growth, immunity, inflammation, energy, and nutrient homeostasis under physiological, as well as hepatic fibrosis under pathological conditions [5].Reportedly, in patients with hepatic fibrosis and experimental models, increased expression of cytokines, especially transforming growth factor b (TGFb), appears to be a key mediator in the initial step for liver fibrogenesis [6,7].Smad signaling plays a major role in the TGFb pathway, which indicates that the TGFb-Smad signaling pathway is a potential target for therapy [8].
Protein kinase C (PKC) isozymes, which belong to a family of serine/threonine kinases, are classified into conventional (a, b, and c), novel (d, g, and e) and atypical subfamilies.Numerous reports describe PKCs as involved in a wide range of biological events, including cell proliferation and differentiation, inflammation, and apoptosis [9,10].In an earlier study, an inhibitor of PKC was shown to suppress hepatic fibrosis development effectively [11].More specifically, TGFb-induced aSMA production is reduced significantly by a specific PKCd inhibitor [12].These findings suggest that PKC signaling plays another key role in the pathogenesis of hepatic injury.Because PKCd, one isozyme of novel PKCs, is activated by binding a lipid second-messenger diacylglycerol (DG), we infer that DG metabolism serves as an additional layer to regulate hepatic injury similarly to Smad signaling.
Diacylglycerol kinase (DGK) is an enzyme that converts DG into phosphatidic acids (PA), thereby regulating two signaling pathways involving these lipid messengers [13][14][15][16].The present study was conducted to ascertain how DGKs are implicated in the pathogenesis of liver injury.We specifically examined the regulation of aSMA expression by PKC and Smad signaling pathways.To this end, we used a cell culture model experiment using the DGK inhibitor R59949.Additionally, we performed an animal model experiment using carbon tetrachloride (CCl 4 ), a chemical that induces hepatocyte necrosis and aSMA expression in activated HSCs at an acute phase [17,18].
The results of an animal model of CCl 4 -induced hepatic injury show that DGKa ablation enhances aSMA expression coinciding with increased Smad2 and PKCd phosphorylation, suggesting that DGKa exerts a negative regulation on aSMA expression.We also evaluated the extent of CCl 4 -induced hepatic injury and discuss the functional implication of aSMA expression in HSCs of DGKa-KO mice after CCl 4 injury.

Cell culture
Mouse fibroblast cell line NIH3T3 cells were obtained from RIKEN BRC (Tsukuba, Japan) and were cultured in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum and Penicillin G / Streptomycin under 5% CO 2 at 37 °C.Cells were incubated with transforming growth factor b (TGFb) (Millipore Sigma, Burlington, MA, USA) at 10 ngÁmL À1 or dimethyl sulfoxide (DMSO) as a vehicle.In some experiments, cells were pretreated with DGK inhibitor R59949 (Millipore Sigma) at 10 lM.Cells were harvested at 0, 30 min, and 6 h of incubation after TGFb treatment for immunoblot analysis.

Animal models
All animal experiments were carried out in accordance with the guidelines and permission of the Yamagata University Animal Ethics Committee (approval number: R5012).Mice were housed under standard conditions and maintained on a 12-h light/dark cycle.They had free access to water and were fed standard mouse laboratory chow (F-2; Oriental Yeast Co., Tokyo, Japan).An acute mouse carbon tetrachloride (CCl 4 ) model using male 6-week-old C57BL6J wildtype (WT) and DGKa-knockout (KO) mice [19] were established by intraperitoneal injection of CCl 4 at a dose of 0.6 mLÁkg À1 .Corn oil was injected at an equal amount as a control.After 48 h of injection, mice were sacrificed and the livers were removed for histological and immunoblot analyses under deep anesthesia.

Histological analysis
Livers removed from CCl 4 -induced models and controls were fixed in 4% paraformaldehyde / 0.1 M phosphate buffer at 4 °C overnight.Paraffin and frozen sections were cut into 6-and 20-lm thickness, respectively.Frozen sections were used for aSMA-immunohistochemistry and paraffin sections were used for hematoxylin and eosin (H&E) and DGKa-immunofluorescence.

aSMA immunohistochemistry
Frozen sections were soaked in 0.3% Triton X-100/PBS at room temperature (RT) for 30 min.Endogenous peroxidase was inactivated by 0.3% H 2 O 2 .After blocking with 5% NGS/PBS, sections were incubated with anti-aSMA antibody (1:400; Millipore Sigma) in a moist chamber at RT overnight.Immunoreaction was detected with diaminobenzidine tetrachloride.

Glutamic-oxaloacetic transaminase (GOT) and glutamate-pyruvate transaminase (GPT) assay
Serum GOT/GTP levels were measured according to the manufacturer's instruction using transaminase CII test Wako (Wako Pure Chemical Industries, Osaka, Japan).

TGFb assay
Serum TGFb levels were measured according to the manufacturer's instruction using the Quantikine ELISA TGFb1immunoassay kit (R&D Systems, Minneapolis, MN, USA).

Cell culture model using DGK inhibitor
We first examined the effect of DGK inhibitor R59949 on aSMA expression and signaling pathway at the cellular level.We used the untransformed mouse fibroblast cell line NIH3T3 cells.Reportedly, a broad DGK inhibitor R59949 inhibits the enzymatic activity of class I DGK isozymes including DGKa, DGKb, and DGKc [21].Since an earlier study shows that major DGK isozymes expressed in NIH3T3 cells include DGKa, DGKd, and DGKf [22], R59949 is presumed to inhibit DGKa activity mostly, if not entirely, in this cell line.Earlier studies have reported that aSMA expression is positively regulated by TGFb that is secreted by hepatocytes and resident macrophages designated as Kupffer cells [23].As shown in Fig. 1, immunoblot analysis revealed that in the absence of R59949, aSMA expression is increased at 30 min and 6 h in response to TGFb stimulation.When pretreated with R59949 for 1 h before stimulation, aSMA expression was increased significantly at timepoint 0 (without TGFb stimulation).It remained high at 30 min.However, it had retuned toward the baseline when measured 6 h after TGFb stimulation.
Reportedly, Smad2 and PKCd play roles in TGFbinduced aSMA expression.Therefore, we next investigated the regulatory mechanism of aSMA expression using a DGK inhibitor.As shown in Fig. 1, the results showed that, in the absence of R59949, Smad2 activity as assessed by phosphorylation status is increased at 30 min and 6 h after TGFb stimulation.In the presence of R59949, Smad2 activity was significantly higher at 30 min of stimulation compared with the control without R59949.Regarding PKCd, in the absence of R59949, its phosphorylation status was increased at 6 h after TGFb stimulation.However, R59949 pretreatment alone increased significantly the PKCd phosphorylation levels at timepoint 0. They were decreased thereafter.In this regard, PKCa/b phosphorylation levels were not changed during the course of TGFb stimulation with or without R59949 (data not shown).
The results of the immunoblot analysis can be summarized as follows.Under conditions of DGK inhibitor R59949 pretreatment, aSMA expression and PKCd activity are upregulated without TGFb stimulation.On the other hand, in the presence of R59949, Smad2 activity is upregulated after 30 min of TGFb stimulation.These findings suggest that class I DGK activity inhibition acts differently on two signaling pathways.It should be also noted, however, that aSMA expression declined in R59949-pretreated cells after 6 h of TGFb stimulation.The reason for this remains undetermined, although a negative feedback loop might serve to restrain aSMA expression at 6 h.
TGFb stimulates aSMA synthesis and incorporation into stress fibers [4].Stable incorporation of aSMA into stress fibers provides an increased myofibroblast contractile force that participates in tissue remodeling [4].Therefore, we next performed immunocytochemistry to examine the morphological aspects of aSMA using the same experimental protocol (Fig. 2).In the absence of R59949 (upper panels), aSMA labeling was observed faintly in a fine granular pattern without TGFb stimulation.After 30 min of TGFb stimulation, aSMA was effectively incorporated partially into stress fibers (upper right panel).
When pretreated with R59949 for 1 h before stimulation (Fig. 2, lower panels), aSMA labeling was observed as abundant in a punctate, cytoplasmic granular pattern, although no apparent stress fiber formation was recognized (lower left panel).At 30 min of TGFb stimulation, most of the aSMA labeling was obviously visible in a cytoskeletal pattern, suggesting that aSMA is incorporated efficiently into stress fibers (lower right panel).Together, the results suggest that the class I DGK inhibitor R59949 enhances aSMA protein expression irrespective of TGFb stimulation, and that it facilitates aSMA incorporation into stress fibers in response to TGFb stimulation.

Hepatic injury model
Reportedly, DGKa is intimately involved in liver function under pathophysiological conditions [24] and positively regulates proliferation and invasion of human hepatocellular carcinoma cells [25].Immunohistochemical analysis showed that DGKa immunoreactivity is  diffusely detected throughout the hepatic lobule.Importantly, it was clearly visible in cells with a slender shape and long cytoplasmic extensions, suggesting that DGKa is expressed in HSCs, if not solely (arrows in upper left panel of Fig. 5A).Therefore, we focused on DGKa in an animal model and strove to gain further insight into the functional implications of DGKa in aSMA expression in HSCs.We undertook the CCl 4 intoxication model, which engenders aSMA induction.In addition to DGKa-KO, we used DGKe-KO and DGKf-KO mice as controls.
Activation of HSCs is a well-known event, occurring at the beginning of CCl 4 -induced hepatic injury.Reportedly, a single dose of CCl 4 (0.6 mLÁkg À1 body weight) is sufficient to elicit liver damage as quickly as 24 h after injection [26].In fact, aSMA, a widely characterized cytoskeletal protein, represents the hallmark of myofibroblast activation and differentiation in liver injury.Therefore, we next examined the levels of aSMA as a marker for HSC activation in an acute injury model.
In immunoblot analysis, it should be noted first that in WT liver after 48 h of CCl 4 injection, DGKa protein expression levels are decreased, whereas aSMA levels are significantly increased (Fig. 3).In this regard, the aSMA levels were robustly increased in DGKa-deficient liver.Similar results were obtained in the expression levels of vimentin, another HSC activation marker.These results suggest that decreased DGKa levels lead to upregulated HSC activation upon CCl 4 intoxication.We next examined the activation status of upstream pathways regulating aSMA expression.Upon CCl 4 injection, the phosphorylation levels of Smad2 and PKCd were significantly increased in WT liver that exhibited decreased DGKa expression levels.Consistent with this, phosphorylation levels of those proteins were vigorously enhanced in DGKadeficient liver.
In DGKe-and DGKf-deficient livers, DGKa protein levels were decreased, and aSMA expression levels were significantly increased upon CCl 4 injection, showing an inverse relationship between levels of DGKa expression and aSMA induction.This is principally consistent with the results obtained in WT and DGKa-KO mice.However, the details seemed somewhat different: Levels of PKCd phosphorylation were not significantly increased in DGKe-and DGKf-KO mice.The absence of DGKe or DGKf might exert some yet-unknown effects on the response to CCl 4 intoxication.
We next examined whether TGFb secretion levels are altered in DGKa-KO mice 48 h after CCl 4 Fig. 3. Immunoblot analysis in acute hepatic injury model.Wildtype (WT) and DGKa-knockout (DGKa-KO) mice were injected with a single dose of CCl 4 (0.6 mLÁkg À1 body weight) or corn oil (control).After 48 h of injection, livers were removed and examined for immunoblot analyses.As a control, corn oil was injected with the same amount.DGKe-KO and DGKf-KO mice were used as controls.Liver homogenates were immunoblotted for aSMA, phospho-PKCd (Thr505), total PKCd, vimentin, phospho-Smad2, and total Smad2/3.GAPDH was used as a control.Immunoblot signals were quantified by densitometry and the values were normalized to the GAPDH level.A representative result of three repeated experiments is shown.injection.To this end, we performed an enzyme-linked immunosorvent assay (ELISA) for serum TGFb levels.As presented in Fig. 4, serum TGFb levels were increased slightly in both WT and DGKa-KO mice to the same extent 48 h after CCl 4 injection.The results suggest that increased aSMA expression is not attributed to an increased TGFb secretion, but rather to the activated signaling pathway downstream of the TGFb receptor in HSCs of DGKa-KO mice.
Having shown that upon CCl 4 intoxication DGKadeficient livers exhibit activated signaling pathways of Smad2 and PKCd that lead to enhanced aSMA expression, we next assessed how these conditions affect the progress or degree of hepatic damage after 48 h of CCl 4 injection in DGKa-KO mice.The acute liver injury model of this experimental scheme is shown to induce centrilobular necrosis, which is characterized by hepatocyte necrosis and lymphocyte infiltration around a central vein [17].As shown in Fig. 5A, confocal microscopic observation demonstrated that the number of DGKa immunoreactive cells was decreased in CCl 4injected WT liver compared with that of oil-injected WT liver.These data were consistent with the immunoblot analysis.The results also revealed that after 48 h of CCl 4 injection, aSMA-immunoreactive cells was observed in the centrilobular portion of the liver in both WT and DGKa-KO mice (Fig. 5B).Consistent with the immunoblot analysis, aSMA-immunoreactive cells were significantly more abundant in DGKa-deficient livers than  in WT ones.We next performed H&E staining to examine the degree of hepatic damage after 48 h of CCl 4 injection.On liver sections of both WT and DGKa-KO mice, hepatocytes exhibited a paler cytoplasm together with pyknotic nuclei in the centrilobular area, showing centrilobular hepatic damage at an acute phase (Fig. 5C).It should be noted, however, that the necrotic area seems slightly smaller in DGKa-deficient liver than in WT liver.
Serum glutamic-oxaloacetic transaminase (GOT) / glutamate-pyruvate transaminase (GPT) assays are useful biochemical markers to assess hepatocyte damage.To be noted, after 48 h of CCl 4 injection GOT and GPT levels tended to be lower in DGKa-KO mice than those in WT mice, although statistically not significant (Fig. 6).Collectively, the results suggest that DGKa-deficient liver is less vulnerable to CCl 4induced injury than WT liver, judging from the smaller centrilobular necrotic area and lower serum biomarkers, despite the upregulated aSMA expression.

Discussion
aSMA, a cytoskeletal protein in vascular smooth muscle cells, is induced in HSCs under pathological conditions.This represents an initial step for HSC activation [2].In the present study we examined the functional implication of DGK in the regulatory mechanism of aSMA expression.
First, we reveal that in a cellular model experiment, DGK inhibitor R59949 pretreatment enhances aSMA expression in NIH3T3 cells.Previous studies showed that DGKa, DGKd, and DGKf are major DGK isozymes expressed in NIH3T3 cells [22] and that DGK inhibitor R59949 mainly acts on class I DGKs (DGKa, DGKb, and DGKc) [21].Therefore, it is reasonable to consider DGKa as a major target of R59949 in NIH3T3 cells.Taken these findings together, we assumed that enhanced aSMA induction is attributed to inhibition of DGKa enzymatic activity in NIH3T3 cells, if not entirely.
This assumption is confirmed by our second experiment, i.e. an animal model of acute hepatic injury.In the liver of WT mice, CCl 4 intoxication downregulates DGKa expression levels but upregulates aSMA induction.Consistent with this, aSMA induction is robustly upregulated in the liver of DGKa-KO mice upon CCl 4 intoxication.Collectively, the results obtained from studies using cellular and animal models suggest that DGKa negatively regulates aSMA expression in an activity-dependent manner.
In CCl 4 -induced liver injury, TGFb appears to be a key mediator [6,7].The TGFb-activated Smad signaling pathway stimulates experimental hepatic fibrosis and is a potential target for therapy [8].In this regard, we found no significant difference in serum TGFb levels between wildtype and DGKa-KO mice in an acute hepatic injury model.
From these findings, we infer that increased aSMA expression is assigned to a facilitated response in HSCs themselves.As described above, the TGFb-activated Smad signaling pathway plays a pivotal role in HSC activation.Our cellular model study suggests that inhibition of DGKa enzymatic activity enhances Smad2 phosphorylation together with aSMA expression at 30 min of TGFb stimulation.Additionally, it is noteworthy that DGKa activity inhibition enhances PKCd phosphorylation together with aSMA expression in the absence of TGFb stimulation.When considering the time course of activation of Smad2 and PKCd together with that of subcellular aSMA labeling, we propose the following working hypothesis: DGKa activity inhibition enhances aSMA protein synthesis in HSCs independently of TGFb stimulation, but it does not facilitate stress fiber formation.In response to TGFb stimulation, DGKa activity inhibition enhances Smad2 activation, which engenders the effective incorporation of aSMA into stress fibers.It remains unclear why aSMA expression declines in R59949-pretreated cells after 6 h of TGFb stimulation.Since cytoskeletal proteins like a-SMA should be tightly regulated within the physiological range for cell survival, a yet undetermined negative feedback loop might serve to restrain aSMA expression at 6 h.
An important question remains unanswered, which involves the phenotype of DGKa-deficient liver.In the animal model of the present study, we found that DGKa-deficient liver exhibits a smaller necrotic area together with lower GOT/GPT levels than WT liver.This suggests that DGKa-deficiency renders cells less vulnerable to CCl 4 -induced injury compared with WT liver, despite the upregulated aSMA expression.
Does DGKa ablation induce HSC activation, thereby ameliorating liver injury?As noted in the Introduction, recent studies suggest that HSCs represent a remarkable plastic cell type that regulates both liver homeostasis and tissue repair [5,27].Whether HSC activation initiates repair processes or exacerbates liver damage might depend on the cellular and environmental context and the extent of injury.Further studies are needed to elucidate this point.
In summary, the present studies using cellular and animal models show that DGKa activity inhibition and DGKa ablation leads to activation of PKCd and TGFb-triggered Smad2 pathways, thereby cooperatively inducing aSMA expression for HSC activation.These findings suggest that DGKa negatively regulates aSMA expression, which may be exerted in an activity-dependent manner.In addition, we also suggest that the PKCd pathway regulates aSMA expression levels, whereas the TGFb-triggered Smad pathway facilitates aSMA incorporation into stress fibers.Interestingly, it is also suggested that DGKa ablation does not result in exacerbation of liver damage in CCl 4 injection, despite the upregulated aSMA expression.This raises a possibility that HSC activation is involved in the recovery process in the DGKamediated pathway.

Fig. 1 .
Fig. 1.Immunoblot analysis in cell culture model using DGK inhibitor.Nontransformed NIH3T3 cells derived from mouse fibroblasts were incubated with TGFb (10 ngÁmL À1 ).Some cells were pretreated for 1 h with R59949, an inhibitor for class I DGKs (10 lM).Control cells were incubated with DMSO solvent alone.Cells were harvested at 0, 30 min, and 6 h after TGFb stimulation in the presence or absence of R59949.Total cell lysates were immunoblotted for aSMA, phospho-PKCd (Thr505), total PKCd, phospho-Smad2, and total Smad2/3.b-actin was used as a control.Immunoblot signals were quantified by densitometry and the values were normalized to the b-actin level.A representative result of three repeated experiments is shown.

Fig. 2 .
Fig.2.Immunocytochemical analysis in cell culture model using DGK inhibitor.NIH3T3 cells were stimulated with TGFb in the presence or absence of R59949.After 30 min, cells were fixed and stained for aSMA (green).In the presence of R59949, aSMA immunoreactivity was significantly increased before and after TGFb stimulation.Note that aSMA staining is detected in a granular pattern before TGFb stimulation whereas it is observed in a stress fiber pattern (arrows) after TGFb stimulation.Insets are the magnified images of the squares.Scale bars = 10 lm.

Fig. 4 .
Fig. 4. Measurement of the serum TGFb levels for liver injury.TGFb levels were measured using the Quantikine ELISA TGFbimmunoassay kit.Serum TGFb levels were increased slightly in both WT and DGKa-KO mice to the same extent 48 h after CCl 4 injection.Data shown are the means AE SD (n = 5).n.s., not significant (Student's t-test).

Fig. 5 .
Fig. 5. Histological analysis of the livers in the acute hepatic injury model of WT and DGKa-KO mice.(A) Immunofluorescence using the anti-DGKa-antibody.Compared with the corn oil-injected control (upper panel), smaller numbers of DGKaimmunoreactive cells (arrowheads) are demonstrated in CCl 4 -injected WT liver (lower panel).No immunoreaction for DGKa was detected in DGKa-KO liver.Scale bar = 10 lm.(B) aSMAimmunoreactivity is intensely recognized in CCl 4 -injected DGKa-KO liver compared with WT liver (lower panels).Upper panels are oil-injected controls.Scale bar = 50 lm.(C) Paraffin sections of livers were stained with routine H&E.Note the smaller necrotic area indicated by a dashed line in DGKa-KO mice compared with WT controls.Scale bar = 50 lm.

Fig. 6 .
Fig.6.Measurement of the serum markers for liver injury.Serum GOT (A) and GPT (B) levels were measured using the transaminase CII test, Wako.Both levels tend to be lower in DGKa-KO mice than WT controls, although no significant difference was discerned.Data shown are the means AE SD (n = 4-7).n.s., not significant (Student's t-test).