*2.4. Plagl1 Is Associated with Blood Vessel Development and Insulin Response*

*2.4. Plagl1 Is Associated with Blood Vessel Development and Insulin Response* The PLAGL1 motif was identified in 233 e9.5-specific enhancers that were predicted to associate with genes involved in fetal growth, placental labyrinth morphology, and insulin response (Figure 4a). These enhancers were also predicted to associate with genes involved in obesity and The PLAGL1 motif was identified in 233 e9.5-specific enhancers that were predicted to associate with genes involved in fetal growth, placental labyrinth morphology, and insulin response (Figure 4a). These enhancers were also predicted to associate with genes involved in obesity and overnutrition

overnutrition (Figure 4a). Interestingly, maternal gestational diabetes mellitus (GDM) has been

(Figure 4a). Interestingly, maternal gestational diabetes mellitus (GDM) has been associated with defective insulin signaling within the placenta [43] and increased vascularization [44], leading us to hypothesize that *Plagl1* could be misexpressed in GDM placentas. *Int. J. Mol. Sci.* **2020**, *21*, x FOR PEER REVIEW 6 of 21

**Figure 4.** PLAGL1 is associated with placental morphology and *Plagl1* shows sex-specific differences in GDM mouse placentas. (**a**) The 233 e9.5-specific enhancers containing a PLAGL1 motif are associated with fetal growth, vasculature development, and other processes important in the placenta. (**b**) *Plagl1* is significantly overexpressed in the male placenta from mothers modeling GDM, but not in female placentas (*p*-value ≤ 0.05(\*)). **Figure 4.** PLAGL1 is associated with placental morphology and *Plagl1* shows sex-specific differences in GDM mouse placentas. (**a**) The 233 e9.5-specific enhancers containing a PLAGL1 motif are associated with fetal growth, vasculature development, and other processes important in the placenta. (**b**) *Plagl1* is significantly overexpressed in the male placenta from mothers modeling GDM, but not in female placentas (*p*-value ≤ 0.05 (\*)).

To determine if *Plagl1* expression differs between control placentas and those from GDM mothers, we measured its gene expression in placentas from a mouse model for GDM [45]. Mice were fed a high-fat, high-sucrose diet a week prior to mating and throughout pregnancy resulting in glucose intolerance during pregnancy [45]. When comparing placentas from control mice and placentas from GDM mice at e17.5, we observed no significant difference in *Plagl1* expression (*p*value = 0.07). Since the placental environment has been shown to affect males and females differently [46,47], and *PLAGL1* has also been shown to have sex-specific differences in fetuses [15], we analyzed the sexes separately. First, we compared *Plagl1* gene expression in control female placentas and control male placentas and found no difference (*p*-value = 0.98). Then, we compared *Plagl1* expression in GDM placentas to control placentas for each sex, and found significant upregulation of *Plagl1* in the GDM placentas in males only (*p*-value = 0.031) (Figure 4b; Supplemental Table S3). These findings indicate that GDM affects *Plagl1* in the placenta in a sex-specific manner in a murine model of GDM. To determine if *Plagl1* expression differs between control placentas and those from GDM mothers, we measured its gene expression in placentas from a mouse model for GDM [45]. Mice were fed a high-fat, high-sucrose diet a week prior to mating and throughout pregnancy resulting in glucose intolerance during pregnancy [45]. When comparing placentas from control mice and placentas from GDM mice at e17.5, we observed no significant difference in *Plagl1* expression (*p*-value = 0.07). Since the placental environment has been shown to affect males and females differently [46,47], and *PLAGL1* has also been shown to have sex-specific differences in fetuses [15], we analyzed the sexes separately. First, we compared *Plagl1* gene expression in control female placentas and control male placentas and found no difference (*p*-value = 0.98). Then, we compared *Plagl1* expression in GDM placentas to control placentas for each sex, and found significant upregulation of *Plagl1* in the GDM placentas in males only (*p*-value = 0.031) (Figure 4b; Supplemental Table S3). These findings indicate that GDM affects *Plagl1* in the placenta in a sex-specific manner in a murine model of GDM.

#### *2.5. PLAGL1 Knockdown in the HTR-8/SVneo Cell Line Predicts A Role in Blood Vessel Remodeling 2.5. PLAGL1 Knockdown in the HTR-8*/*SVneo Cell Line Predicts a Role in Blood Vessel Remodeling*

We next investigated a potential role for PLAGL1 in human placenta. Using data from the human protein atlas [48] and the TissueEnrich tool [49], we found that *PLAGL1* has placenta-enriched gene expression (Figure 5a). The human protein atlas further showed that the placenta sections used for analysis were comprised primarily of trophoblast cells (Figure 5b). Therefore, we sought to investigate the role of *PLAGL1* in human trophoblast cells. First, we evaluated RNA-seq data from several human cell lines [50] for *PLAGL1* expression, including: choriocarcinomas representing villous trophoblast (BeWo [51] and JEG3 [52]); syncytiotrophoblast (PHTd\_Syncytio [53]) differentiated from term placenta cytotrophoblast (PHTu\_Cyto [53]); BAP treated human embryonic stem cells (ESCd [53]); and a cell line derived from chorionic villi explants of first trimester placenta We next investigated a potential role for PLAGL1 in human placenta. Using data from the human protein atlas [48] and the TissueEnrich tool [49], we found that *PLAGL1* has placenta-enriched gene expression (Figure 5a). The human protein atlas further showed that the placenta sections used for analysis were comprised primarily of trophoblast cells (Figure 5b). Therefore, we sought to investigate the role of *PLAGL1* in human trophoblast cells. First, we evaluated RNA-seq data from several human cell lines [50] for *PLAGL1* expression, including: choriocarcinomas representing villous trophoblast (BeWo [51] and JEG3 [52]); syncytiotrophoblast (PHTd\_Syncytio [53]) differentiated from term placenta cytotrophoblast (PHTu\_Cyto [53]); BAP treated human embryonic stem cells (ESCd [53]); and a cell line derived from chorionic villi explants of first trimester placenta (HTR-8/SVneo [54]). We proceeded with

commonly used to model endovascular differentiation of extravillous trophoblast [55–59].

(HTR-8/SVneo [54]). We proceeded with our experiments in HTR-8/SVneo cells, since they had the

our experiments in HTR-8/SVneo cells, since they had the highest expression of *PLAGL1* (Figure 5c). When grown on Matrigel, HTR-8/SVneo cells express genes associated with blood vessel development, can form endothelial tube-like structures, and are commonly used to model endovascular differentiation of extravillous trophoblast [55–59]. *Int. J. Mol. Sci.* **2020**, *21*, x FOR PEER REVIEW 7 of 21

**Figure 5.** PLAGL1 regulates genes involved in blood vessel development. (**a**) Bar graph of PLAGL1 expression in human tissues, generated using TissueEnrich. Expression reported using transcripts per million (TPM). (**b**) Percentage of trophoblast and endothelial cells in human placental samples. Data is from the Human Protein Atlas. (**c**) Bar graph of PLAGL1 expression in multiple human cell lines. (**d**) *PLAGL1* expression is significantly reduced by two siRNAs (*p*-value ≤ 0.01 (\*\*)) compared to a negative control. Values are normalized to the negative control siRNA. (**e**) 4003 genes are differentially expressed after *PLAGL1* knockdown in HTR-8/SVneo cells. Genes which increase (1964) in expression are indicated as orange dots on the volcano plot and those that decrease (2039) are purple. (**f**) Cell-adhesion terms are associated with genes that are upregulated when *PLAGL1* is knocked down. (**g**) Vasculature development terms are associated with genes that are downregulated when *PLAGL1* is knocked down. **Figure 5.** PLAGL1 regulates genes involved in blood vessel development. (**a**) Bar graph of PLAGL1 expression in human tissues, generated using TissueEnrich. Expression reported using transcripts per million (TPM). (**b**) Percentage of trophoblast and endothelial cells in human placental samples. Data is from the Human Protein Atlas. (**c**) Bar graph of PLAGL1 expression in multiple human cell lines. (**d**) *PLAGL1* expression is significantly reduced by two siRNAs (*p*-value ≤ 0.01 (\*\*)) compared to a negative control. Values are normalized to the negative control siRNA. (**e**) 4003 genes are differentially expressed after *PLAGL1* knockdown in HTR-8/SVneo cells. Genes which increase (1964) in expression are indicated as orange dots on the volcano plot and those that decrease (2039) are purple. (**f**) Cell-adhesion terms are associated with genes that are upregulated when *PLAGL1* is knocked down. (**g**) Vasculature development terms are associated with genes that are downregulated when *PLAGL1* is knocked down.

identified 4003 genes that were differentially expressed (fold ≥ 1.5, adjusted *p*-value ≤ 0.05) between the *PLAGL1* and negative control knockdown samples using DESeq2 [60] (Figure 5e; Supplemental

To understand the global impacts PLAGL1 has on gene expression, we performed an siRNA

To understand the global impacts PLAGL1 has on gene expression, we performed an siRNA knockdown of *PLAGL1* followed by RNA-seq in the HTR-8/SVneo cells. We first tested two siRNAs, and found that both siRNAs knocked down *PLAGL1* by 73–75% on average (Figure 5d). Since both siRNAs showed similar knockdown efficiencies, we proceeded with RNA-seq using siRNA 1. We identified 4003 genes that were differentially expressed (fold ≥ 1.5, adjusted *p*-value ≤ 0.05) between the *PLAGL1* and negative control knockdown samples using DESeq2 [60] (Figure 5e; Supplemental Table S4). The 1964 genes that were increased upon *PLAGL1* knockdown included several protocadherins such as *PCDH1*, *PCDH10*, and *PCDH7*. The only terms enriched in this group were related to cell-cell adhesion (Figure 5f). On the other hand, the 2039 genes that decreased as a result of *PLAGL1* knockdown were enriched for terms related to blood vessel development, cell migration, and both type 1 (insulin dependent) and type 2 (non-insulin dependent) diabetes (Figure 5g; Supplemental Figure S3). *Int. J. Mol. Sci.* **2020**, *21*, x FOR PEER REVIEW 8 of 21 Table S4). The 1964 genes that were increased upon *PLAGL1* knockdown included several protocadherins such as *PCDH1*, *PCDH10*, and *PCDH7*. The only terms enriched in this group were related to cell-cell adhesion (Figure 5f). On the other hand, the 2039 genes that decreased as a result of *PLAGL1* knockdown were enriched for terms related to blood vessel development, cell migration, and both type 1 (insulin dependent) and type 2 (non-insulin dependent) diabetes (Figure 5g; Supplemental Figure S3).

To determine if *PLAGL1* could play a role in the ability of HTR-8/SVneo cells to mimic endothelial cell tube-like structure formation, we performed a tube formation assay. Ten hours after plating the cells transfected with the *PLAGL1* siRNA, we see a significant decrease in cord formation, as determined by the total branching length (*p*-value = 0.0093) and number of enclosed regions (meshes; *p*-value = 0.0024) compared to cells transfected with a negative control (Figure 6). To determine if *PLAGL1* could play a role in the ability of HTR-8/SVneo cells to mimic endothelial cell tube-like structure formation, we performed a tube formation assay. Ten hours after plating the cells transfected with the *PLAGL1* siRNA, we see a significant decrease in cord formation, as determined by the total branching length (*p*-value = 0.0093) and number of enclosed regions (meshes; *p*-value = 0.0024) compared to cells transfected with a negative control (Figure 6).

**Figure 6.** *PLAGL1* knockdown decreases cord formation ability in HTR-8/SVneo cells. (**a**) Representative images showing cord formation is reduced after PLAGL1 is knocked down (left) compared to a control (right). (**b**) Branch length was significantly reduced after PLAGL1 knockdown (*p*-value ≤ 0.01(\*\*)). (**c**) Enclosed regions, or meshes, were significantly reduced after PLAGL1 knockdown (*p*-value ≤ 0.01(\*\*)). **Figure 6.** *PLAGL1* knockdown decreases cord formation ability in HTR-8/SVneo cells. (**a**) Representative images showing cord formation is reduced after PLAGL1 is knocked down (left) compared to a control (right). (**b**) Branch length was significantly reduced after PLAGL1 knockdown (*p*-value ≤ 0.01 (\*\*)). (**c**) Enclosed regions, or meshes, were significantly reduced after PLAGL1 knockdown (*p*-value ≤ 0.01 (\*\*)).

similar to the terms associated with the genes we initially predicted to contain PLAGL1-binding motifs in their enhancer regions in the mouse genome (Figures 5g and 4a). Therefore, we tested whether the specific genes associated with terms enriched for predicted PLAGL1 enhancers

*2.6. Relationship between Mouse and Human Plagl1 Results*
