*2.3. MARK4 Increases Lipid Droplet Accumulation in Pig Placental Trophoblast Cells*

In this study, we speculated that MARK4 could modulate lipid accumulation in porcine placental trophoblast cells. To validate our hypothesis, we initially tested whether overexpression of MARK4 influences the accumulation of fatty acid in cultured term primary pig trophoblasts exposed to 400 μM FA. The results of Bodipy 493/503 fluorescence staining and TG content assay indicated that lipid droplet accumulation was significantly increased in trophoblasts from the Myc- MARK4 group compared with the sh- MARK4 or vector control groups (*p* < 0.05; control panel in Figure 2A,B).

**Figure 2.** MARK4 promotes lipid accumulation in pig primary trophoblast cells challenged with 400 μM NEFA. (A and C) Representative images (100×) of Bodipy staining after transfection with Myc-MARK4, sh-MARK4 for 48 h in primary (trophoblast cells) isolated from pig placentas. Primary trophoblasts were then incubated with 400 μM NEFA, 2 μM GW1929 or 500 μM phloretin for 24 h (*n* = 3). (**B** and **D**) Quantification of corresponding triglyceride (TG) in (**A**) and (**C**) by ELISA analysis (*n* = 3). The values in red indicate receptor (transport proteins)-mediated fatty acid accumulation by subtracting the values in the presence of phloretin from those in the absence of phloretin. (E) LPL activity (mU/mg protein) after transfection with Myc-MARK4, sh-MARK4 for 48 h in pig primary trophoblasts. Cells were then treated with 400 μM NEFA or 2 μM GW1929 for 24 h (*n* = 3). Values are expressed as mean ± SEM. \*\* *p* < 0.01; \* *p* < 0.05 compared with the control group. Myc-MARK4 group: overexpression of MARK4 group, sh-MARK4 group: knock down of MARK4 group, Control: empty vector (EV) group.

We next examined whether MARK4 affected receptor (transport proteins)-mediated fatty acid accumulation in cultured trophoblast cells. As shown in Figure 2B, sh-MARK4 treatment increased receptor-mediated fatty acid accumulation in trophoblasts compared with Myc-MARK4 group following 24 h exposure to FA (sh-MARK4: 14.54 ± 2.41 mg/g versus Myc-MARK4: 6.09 ± 1.61 mg/g, *p* < 0.05). Previous studies have shown that PPARγ is involved in regulating fatty acid transport and accumulation in primary human placental trophoblasts [21]. We therefore hypothesized that activation of PPARγ might increase the accumulation of fatty acid in cultured pig placental trophoblast cells. To test this hypothesis, we incubated trophoblasts in the presence or absence of PPARγ-specific

agonist GW1929. As shown in Figure 2B,D, activation of PPARγ promoted receptor-mediated fatty acid accumulation in sh-MARK4 treatment following 24 h exposure to FA (sh-MARK4+GW1929: 24.37 ± 1.39 mg/g versus sh-MARK4: 14.54 ± 2.41 mg/g, *p* < 0.05), whereas non- receptor-mediated fatty acid accumulation was significantly decreased in Myc-MARK4 group following GW1929 + phloretin treatment (Myc-MARK4+GW1929: 28.75 ± 1.03 mg/g versus Myc-MARK4: 42.87 ± 1.89 mg/g, *p* < 0.05). In accord with increased receptor-mediated fatty acid accumulation in Myc-MARK4+GW1929 group (Myc-MARK4+GW1929: 12.60 ± 1.22 mg/g versus Myc-MARK4: 6.09 ± 1.61 mg/g, *p* < 0.05), the LPL activity in Myc-MARK4 + GW1929 group was markedly higher than that in Myc-MARK4 group (*p* < 0.05; Figure 2E).

#### *2.4. Effect of MARK4 on Key Factors of Lipid Metabolism in Pig Placental Trophoblasts*

We first determined the overexpression of MARK4 by testing protein content of MARK4 gene following transfection and FA treatment. As shown in Figure 3A,B, MARK4 protein increased in Myc-MARK4 group, while sh-MARK4 treatment reduced MARK4 protein (*p* < 0.05). Consistent with increased lipid droplet accumulation following FA treatment, the mRNA expression of genes associated with fatty acid uptake and accumulation, including LPL and DGAT1, was significantly increased in Myc-MARK4 group, whereas the mRNA content of lipid metabolism-related genes, including PPARG (PPARγ), ADRP and ACSL1, was reduced in Myc-MARK4 group compared with the sh-MARK4 or vector control groups (*p* < 0.05; Figure 3D). GW1929, the potent and specific agonist of PPARγ (Figure 3C), was used to examine the regulatory role of PPARγ on MARK4-induced increases in lipid accumulation of trophoblasts. As shown in Figure 3D, GW1929 promoted the mRNA expression of PPARG, ADRP and ACSL1 in Myc-MARK4 group, but the mRNA content of LPL and DGAT1 was decreased in Myc-MARK4+ GW1929 treatment (*p* < 0.05). In accordance with elevated receptor-mediated fatty acid accumulation following GW1929 + sh- MARK4 treatment, GW1929 increased the mRNA content of several fatty acid transporters, including FATP1, FATP4, CD36, FABP1 and FABP4, in sh-MARK4 group (*p* < 0.05; Figure 3E).

#### *2.5. WNT Signaling Promotes Lipid Accumulation and Activation of MARK4 in Pig Trophoblasts*

Previous experiments in our laboratory and others have shown that an aberrant activation of WNT signaling contributes to significant placental lipid accumulation in obese model of rat or pig [10,11]. In order to further reveal the mechanisms responsible for the increased placental lipid accumulation induced by WNT signaling, we first performed Bodipy fluorescence staining to evaluate lipid droplet accumulation in pig trophoblasts from three groups: Flag-DKK1, sh-DKK1 and Vector control. DKK1 (dickkopf family protein1) is an inhibitor of the canonical WNT signaling pathway [22]. As shown in Figure 4A,B, Flag-DKK1 treatment reduced lipid droplet accumulation in trophoblasts following 24 h exposure to FA (*p* < 0.05), whereas activation of WNT signaling by GSK3β inhibitor LiCL, which was the downstream of DKK1 and blocked the phosphorylation of β-catenin and subsequent proteolytic degradation, significantly increased lipid accumulation in sh-DKK1 treatment (*p* < 0.01). The LPL activity was not affected by Flag-DKK1 or sh-DKK1 treatment in the presence or absence of LiCL (Figure 4C).

**Figure 3.** Effects of MARK4 on key molecules of lipid metabolism in pig primary trophoblast cells. (**A**–**B**) Representative immunoblots and densitometric quantification for MARK4 after transfection with Myc-MARK4, sh-MARK4 for 48 h in primary trophoblast cells isolated from pig placentas. Cells were then incubated with 400 μM NEFA or 2 μM GW1929 for 24 h (*n* = 3). (**C**) Primary trophoblasts were cultured and incubated for 0 h, 12 h, 24 h, 36 h and 48 h in the presence of 2 μM GW1929. Relative mRNA expression of PPARγ was detected (*n* = 3). (**D**–**E**) Relative mRNA expression of lipid metabolism-related genes (**D**) and fatty acid (FA) transporters (**E**) after transfection with Myc-MARK4, sh-MARK4 for 48 h in primary (trophoblast cells). Cells were then treated with 400 μM NEFA or 2 μM GW1929 for 24 h (*n* = 3). Values are expressed as mean ± SEM. \*\* *p* < 0.01; \* *p* < 0.05 compared with the control group. Myc-MARK4 group: over expression of MARK4 group, sh-MARK4 group: knock down of MARK4 group, Control: empty vector (EV) group.

**Figure 4.** Activation of the Wnt/β-catenin pathway promotes lipid accumulation in pig primary trophoblast cells challenged with 400 μM NEFA. (**A**) Representative images (100×) of Bodipy staining after transfection with Flag-DKK1, sh-DKK1 for 48 h in primary trophoblast cells isolated from pig placentas. Cells were then incubated with 400 μM NEFA or 20 μM Li CL for 24 h (*n* = 3). (**B**) Quantification of corresponding triglyceride (TG) in (**A**) by ELISA analysis (*n* = 3). (**C**) LPL activity (mU/mg protein) after transfection with Flag-DKK1, sh-DKK1for 48 h in pig primary trophoblasts. Cells were then treated with400 μM NEFA or 20 μM Li CL for 24 h (*n* = 3). Values are expressed as mean ± SEM. \*\* *p* < 0.01; \* *p* < 0.05 compared with the control group. Flag-DKK1 group: over expression of DKK1 group, sh-DKK1 group: knock down of DKK1 group, Control: empty vector (EV) group.

We next determined whether inhibition of WNT signaling affected lipid metabolism in pig placental trophoblasts. Not surprisingly, overexpression of DKK1 increased DKK1 protein content (*p* < 0.05; Figure 5A,B) and reduced β-catenin protein expression within the nucleus (Figure 5C). Notably, LiCL treatment prevented DKK1-induced degradation of β-catenin (*p* < 0.05; Figure 5D); this result was also confirmed by immunofluorescence assay for β-catenin (*p* < 0.05; Figure 5E,F). Consistent with elevated lipid accumulation in sh-DKK1 group following exposure to FA + LiCL, the mRNA expression of genes associated with TG synthesis, including DGAT1, LPL, LPIN3 and PPARδ, were higher in sh-DKK1 + LiCL treatment (*p* < 0.05; Figure 5G), while LiCL treatment reduced the mRNA content of fatty acid transport -related genes, including PPARγ, FATP1, FATP4, CD36 and FABP4, in Flag-DKK1 or sh-DKK1 group (*p* < 0.05; Figure 5G,H). Moreover, phos- MARK4(Thr214) was decreased in Flag-DKK1 compared with the sh-DKK1 or vector control groups, but increased activation of Mark4 was observed in Flag-DKK1+ LiCL treatment (*p* < 0.05; Figure 5A,B).

### *2.6. WNT/β-Catenin Signal is Essential for MARK4 Activated Lipogenesis in Pig Trophoblast Cells*

Having determined that WNT signaling enhanced the accumulation of fatty acids and activation of MARK4 in pig placental trophoblast cells, we next addressed whether WNT/β-catenin pathway was involved in Mark4-induced lipid accumulation in pig trophoblasts. To test this hypothesis, we incubated trophoblasts in the presence or absence of WNT signaling pathway specific inhibitor JW74. As shown in Figure 6B,D, inhibition of WNT/β-catenin signaling by JW74 reduced nonreceptor-mediated fatty acid accumulation in sh-MARK4 group following 24 h exposure to FA + phloretin (sh-MARK4+JW74: 3.56 ± 0.80 mg/g versus sh-MARK4: 16.47 ± 1.61 mg/g, *p* < 0.05), whereas receptor-mediated fatty acid accumulation was significantly increased in Myc-MARK4 group following JW74 treatment (Myc-MARK4+ JW74: 9.76 ± 0.90 mg/g versus Myc-MARK4: 4.79 ± 1.85 mg/g, *p* < 0.05). No differences were found in the LPL activity among Myc-MARK4, sh-MARK4 and Vector control in the presence or absence of JW74 (Figure 6E).

**Figure 5.** Inhibition of the Wnt/β-catenin pathway blocks key molecules of lipid metabolism and activation of MARK4 in pig primary trophoblast cells. (**A**–**D**) Representative immunoblots and densitometric quantification for p-MARK4 (T214), DKK1 and β-catenin after transfection with Flag-DKK1, sh-DKK1 for 48 h in primary trophoblast cells isolated from pig placentas. Cells were then incubated with 400 μM NEFA or 20 μM Li CL for 24 h (*n* = 3). (**E**) Representative images (100×) of β-catenin immunofluorescent staining after transfection with Flag-DKK1, sh-DKK1 for 48 h in pig primary trophoblast cells. Cells were then incubated with 400 μM NEFA or 20 μM Li CL for 24 h (*n* = 3). (**F**) Quantification of red fluorescence intensity in (**E**) relative to control group (*n* = 3). (**G**–**H**) Relative mRNA expression of lipid metabolism-related genes (**G**) and fatty acid (FA) transporters (**H**) after transfection with Flag-DKK1, sh-DKK1 for 48 h in primary trophoblast cells. Cells were then treated with 400 μM NEFA or 20 μM Li CL for 24 h (*n* = 3). Values are expressed as mean ± SEM. \* *p* < 0.05 compared with the control group. Flag-DKK1 group: overexpression of DKK1 group, sh-DKK1 group: knock down of DKK1 group, Control: empty vector (EV) group.

**Figure 6.** Activation of the WNT/β-catenin pathway by MARK4 promotes lipid accumulation in pig primary trophoblast cells challenged with 400 μM NEFA. (**A** and **C**) Representative images (100×) of Bodipy staining after transfection with Myc-MARK4, sh-MARK4 for 48 h in primary (trophoblast cells) isolated from pig placentas. Cells were then incubated with 400 μM NEFA, 10 μM JW74 or 500 μM phloretin for 24 h (*n* = 3). (**B** and **D**) Quantification of corresponding triglyceride (TG) in (**A**) and (**C**) by ELISA analysis (*n* = 3). The values in red indicate receptor (transport proteins)-mediated fatty acid accumulation by subtracting the values in the presence of phloretin from those in the absence of phloretin. (**E**) LPL activity (mU/mg protein) after transfection with Myc-MARK4, sh-MARK4 for 48 h in pig primary trophoblasts. Cells were then treated with 400 μM NEFA or 10 μM JW74 for 24 h (*n* = 3). Values are expressed as mean ± SEM. \*\* *p* < 0.01; \* *p* < 0.05 compared with the control group. Myc-MARK4 group: overexpression of MARK4 group, sh-MARK4 group: knock down of MARK4 group, Control: empty vector (EV) group.

We further confirmed the role of WNT signaling on MARK4 activated lipogenesis in trophoblasts by Western blot analysis. Specifically, overexpression of MARK4 increased the protein contents of Mark4 and β-catenin (*p* < 0.05; Figure 7A,C), while no changes were noted for DKK1 expression in Myc-MARK4 or sh-MARK4 treatment in the presence or absence of JW74 (Figure 7B). Despite with JW74 treatment, MARK4 still increased the content of β-catenin within the nucleus (*p* < 0.05; Figure 7D). In accordance with increased receptor-mediated fatty acid accumulation in Myc-MARK4 + JW74 treatment, the mRNA expression of genes associated with fatty acid transport, including ACSL1, ADRP, PPARγ, FATP1, FATP4, CD36, FABP1 and FABP4, were up-regulated in Myc- MARK4 group following exposure to JW74(*p* < 0.05; Figure 7E,G), whereas the mRNA content of genes associated with TG and lipid droplet synthesis, including ACACA, FASN, DGAT1, LPIN1, LPIN3, LPL, PPARδ and SREBP-1c, were decreased in sh-MARK4 group following JW74 treatment (*p* < 0.05; Figure 7E,F,H), in agreement with reduced non- receptor-mediated fatty acid accumulation in sh-MARK4 + JW74 group following exposure to FA + phloretin.

**Figure 7.** Activation of the WNT/β-catenin pathway by MARK4 promotes lipogenesis in pig primary trophoblast cells challenged with 400 μM NEFA. (**A**–**D**) Representative immunoblots and densitometric quantification for MARK4, DKK1 and β-catenin after transfection with Myc-MARK4, sh-MARK4 for 48 h in primary trophoblast cells isolated from pig placentas. Cells were then incubated with 400 μM NEFA or 10 μM JW74 for 24 h (*n* = 3). (**E**–**H**) Relative mRNA expression of lipid metabolism-related genes (**E** and **F**), fatty acid (FA) transporters (**G**) and regulators of lipid metabolism (**H**) after transfection with Myc-MARK4, sh-MARK4 for 48 h in primary trophoblast cells. Cells were then treated with 400 μM NEFA or 10 μM JW74 for 24 h (*n* = 3). Values are expressed as mean ± SEM.\* *p* < 0.05 compared with the control group. Myc-MARK4 group: over expression of MARK4 group, sh-MARK4 group: knock down of MARK4 group, Control: empty vector (EV) group.

#### **3. Discussion**

At present, the MARK4 gene has been widely explored in mammal species [13]. However, such information is still quite limited in *Sus scrofa* (Pig). In this study, the full-length cDNA of MARK4 was characterized from a lean breed swine (Landrace), including an ORF of 2259 bp nucleotides in length, encoding 752 amino acids (AA) residues, in agreement with the previous study [20]. Sequence alignments and phylogenetic analysis showed that MARK4 is highly conserved between *Sus scrofa* (Pig) and other mammals. In addition, several functional sites were also observed, including a protein kinase ATP-binding region, a serine/threonine protein kinase active-site and a protein kinase domain, which represent the typical characters of the protein kinase superfamily [23]. Meanwhile, the catalytic kinase domain (KD), the ubiquitin-associated domain (UBA), the kinase associated domain1 (KA1) and three conserved functional sites (lysine 88 ATP binding site, aspartic 181 active site and threonine 214 phosphorylation site) were also identified through the multiple alignment analysis, which are regarded as the typical structures of microtubule affinity regulatory kinases family [12,13]. Several studies on mammals indicated that the activation of MARK4 is mediated by the major active site (Asp 181) that is activated by phosphorylation of Thr 214 located in the activation loop (T-loop) on protein kinase domain, whereas phosphorylation of Ser 218 in T-loop inactivates MARK4 [24,25]. Accordingly, compared with other mammals, we found the AA sequences of MARK4 protein in *Sus scrofa* has a conserved T-loop sequence, LDTFCGSPP, including the regulatory phosphorylation sites of Thr 214 and Ser 218. Furthermore, the predicted tertiary protein structure of MARK4 in *Sus scrofa* showed high similarity (AA sequence identity is 99%) with that of human (*Homo sapiens*). This was further confirmed by the observation that the key structural residues (Lys 88, Asp 181 and Thr 214) of human MARK4 protein are all well conserved in that of porcine (See Figure 1).

MARK4, the fourth member of microtubule affinity regulatory kinases (MARKs) family, is implicated in the regulation of dynamic biological functions, including glucose homeostasis and energy metabolism [16]. Recently, MARK4 has been found to promote adipogenesis and trigger apoptosis in 3T3-L1 adipocytes [17], suggesting MARK4 may play an important role in regulating lipid metabolism in adipose tissue. In addition, hyperlipidemia associated with obesity has been suggested to contribute to the ectopic lipid accumulation (lipotoxicity) often seen in highly metabolic tissues, including liver, skeletal muscle and placenta [10,11,26], a process that has been implicated as an important mediator of cellular stress and altered tissue function. Regarding the impact of MARK4 on adipogenesis, we hypothesized that Mark4 may potentially stimulate lipid accumulation in other cell types besides placental trophoblast cells from porcine, and this study was designed to investigate the role of MARK4 in modulating lipid metabolic signaling in pig placental trophoblasts in vitro. We found that in pig trophoblast cells MARK4 significantly increased the expression of lipogenic genes, including FASN, ACACA, DGAT1, LPIN1, LPIN3, LPL and SREBP-1c, suggesting increased TG and lipid droplet synthesis by MARK4 expression, as evidenced by dramatically increased lipid droplet accumulation in trophoblast cells. Thus, our data indicated that Mark4 is involved in regulating lipogenesis of pig placental trophoblasts upon the status of lipotoxic insult.

Studies have suggested that activation of PPARγ stimulates fatty acid uptake and fatty acid accumulation in cultured human trophoblast cells [21,27]. However, our data suggests that the stimulating effect of Mark4 on lipid accumulation of trophoblasts is not mediated by increased activation of PPARγ. Furthermore, the MARK4 effect on fatty acid accumulation is unlikely to be due to an activation of LPL activity since MARK4 did not regulate trophoblast LPL activity in vitro. PPARγ is known to be required for placental development and placental uptake of fatty acids [21,28]. Activation of PPARγ regulates gene expression of several proteins involved in lipid transport, including FA transport proteins (FATPs/SLC27As), intracellular FA binding proteins (FABPs), FA translocase (FAT/CD36), adipose differentiation-related protein (ADRP) and Acyl CoA synthase (ACS) [21,27,29]. Our finding showed that MARK4 inhibited the mRNA expression of PPARγ, ADRP, ACSL1, FATP1, FATP4, CD36, FABP1 and FABP4 in cultured trophoblast cells, suggesting impaired FA uptake by trophoblasts in vitro, as evidenced by significantly decreased

receptor (transport proteins)-mediated fatty acid accumulation by MARK4. Recent studies determined that PPARγ and MARK4 play an opposing role in adipose inflammation response and oxidative stress [18]. Consistently, we preliminarily determined that activation of PPARγ by PPARγ-specific agonist GW1929 prevented MARK4 from stimulating lipogenesis and non-receptor-mediated lipid accumulation in cultured pig trophoblasts, suggesting that MARK4 promotes lipid synthesis in pig trophoblast cells by inhibiting the PPARγ pathways. However, the precise mechanism for inhibition of PPARγ by MARK4 in regulating lipogenesis of trophoblasts needs to be further studied.

In this study we determined fatty acid accumulation in trophoblast cells which is dependent upon uptake as well as cellular metabolism. As previously documented, WNT signaling pathway is involved in increased placental lipid accretion in obesity-prone rats or obese women [5,10]. In support of the role of WNT signaling in regulating lipid synthesis, our data showed that inhibition of WNT signaling by DKK1 remarkably reduced the mRNA expression of genes associated with TG and lipid droplet synthesis in pig trophoblasts, including DGAT1, LPL, LPIN3 and PPARδ, which is confirmed by decreased lipid droplet accumulation by DKK1. On the contrary, we found that activation of WNT signaling by GSK3β inhibitor LiCL significantly decreased the expression of PPARγ and several FA transporters, including FATP1, CD36, FABP4 and FATP4, in cultured trophoblast cells. Previous studies have shown that β-catenin (a key target of WNT signaling) and PPARγ functionally interact to negatively regulate each other's activity, and activation of WNT signaling prevents induction of C/EBPα and PPARγ during preadipocyte differentiation [30,31]. Hence, accumulation of fatty acids in pig trophoblast cells in response to lipotoxic insult may be attributed to altered intracellular metabolism of fatty acids rather than changes in cellular uptake.

This study pointed out a significant correlation between MARK4 and WNT signaling. The WNT signal is a cytosolic sensor which activates and promotes β-catenin nuclear translocation and DNA binding [22]. Sun et al. reported that Par-1, the mammalian ortholog of MARKs, is a positive regulator of Wnt/β-catenin pathway in mammalian cells and Drosophila embryos [32]. Consistently, we demonstrated that MARK4 was potent to activate WNT signaling through promoting translocation of β-catenin into the nucleus in cultured pig trophoblasts. It is noticed that activation of WNT signaling pathway by LiCL prevented DKK1 from inhibiting phosphorylation of endogenous Mark4 (Thr214) in trophoblast cells, in agreement with previous studies that WNT signaling stimulates endogenous Par-1 kinase activity [32]. Activation of WNT pathway leads to the phosphorylation of Dishevelled (Dvl) protein, which then inhibits the activity of GSK3β [33]. GSK3β has been shown to inhibit MARK4 protein by phosphorylating the serine residue (Ser218), near the threonine activation site (Thr214) in the activation loop of MARK4 [24]. Therefore, inhibition of GSK3β could be a possible mechanism involved in the activation of MARK4 by WNT signaling. In addition, our experiments employing the WNT specific inhibitor JW74 further confirmed that the WNT pathway is involved in the promotion of lipogenesis via MARK4, suggesting WNT signal is central to MARK4 performing lipid synthesis function in pig trophoblast cells in response to lipotoxic insult.

#### **4. Materials and Methods**

#### *4.1. Experimental Animals and Reagents*

For the analysis of full-length cDNA cloning of MARK4 gene and isolation of porcine placental trophoblast cells, samples of placenta from *Sus scrofa* (Landrace) were collected at Research Farm of Nan Jing Agricultural University. The collection of porcine full-term placental tissue was specifically approved by the Laboratory Animal Care and Use Committee of Nan Jing Agricultural University. (SYXK2015-0072, 6 September 2015)

For the isolation of porcine placental trophobalst cells, the following reagents were purchased, including Ham's F12/Dulbecco's Modified Eagle Medium(DMEM/F12) (HyClone, Logan, UT, USA), fetal bovine serum (FBS) (HyClone, Logan, UT, USA), Trypsin (Gibco, Grand Island, NY, USA), Phosphate-buffered saline (PBS) (Life Technologies, Grand Island, NY, USA), Bovine serum albumin

(BSA) (Amresco, Solon, OH, USA); Percoll (Pharmacia, London, UK), 100× Penicillin-Streptomycin (10,000 U/mL) (Invitrogen, Carlsbad, CA, USA), 100× Insulin–Transferrin–Selenium (ITS; Sigma, Saint Louis, MO, USA) and epidermal growth factor (EGF; Invitrogen, Carlsbad, CA, USA).
