Human Decidua Basalis mesenchymal stem/stromal cells reverse the damaging effects of high level of glucose on endothelial cells in vitro

Abstract Recently, we reported the therapeutic potential of mesenchymal stem/stromal cells (MSCs) from the maternal decidua basalis tissue of human term placenta (DBMSCs) to treat inflammatory diseases, such as atherosclerosis and cancer. DMSCs protect endothelial cell functions from the negative effects of oxidative stress mediators including hydrogen peroxide (H2O2) and monocytes. In addition, DBMSCs induce the generation of anti‐cancer immune cells known as M1 macrophages. Diabetes is another inflammatory disease where endothelial cells are injured by H2O2 produced by high level of glucose (hyperglycaemia), which is associated with development of thrombosis. Here, we investigated the ability of DBMSCs to reverse the damaging effects of high levels of glucose on endothelial cells. DBMSCs and endothelial cells were isolated from human placental and umbilical cord tissues, respectively. Endothelial cells were incubated with glucose in presence of DBMSCs, and their functions were evaluated. The effect of DBMSCs on glucose‐ treated endothelial cell expression of genes was also determined. DBMSCs reversed the effects of glucose on endothelial cell functions including proliferation, migration, angiogenesis and permeability. In addition, DBMSCs modified the expression of several genes mediating essential endothelial cell functions including survival, apoptosis, permeability and angiogenesis. We report the first evidence that DBMSCs protect the functions of endothelial cells from the damaging effects of glucose. Based on these results, we establish that DBMSCs are promising therapeutic agents to repair glucose‐induced endothelial cell injury in diabetes. However, these finding must be investigated further to determine the pathways underlying the protective role of DBMSCs on glucose‐stimulated endothelial cell Injury.


| INTRODUC TI ON
Diabetes is an inflammatory disease where hyperglycaemia (high level of glucose) stimulates the production of reactive oxygen species products, such as hydrogen peroxide (H 2 O 2 ) in endothelial cells that causes injury in the endothelium associated with vascular damage leading to the development of thrombosis. [1][2][3][4][5][6][7][8] Therefore, repairing the damage induced by high levels of glucose in endothelial cells is one of the therapeutic options that aims to restore endothelial cell functional activities and preventing complications (ie atherosclerosis and thrombosis) associated with diabetes.
Recently, we reported the therapeutic potential of DBMSCs [mesenchymal stem/stromal cells (MSCs) from the maternal decidua basalis tissue of human term placenta] to treat inflammatory diseases, such as atherosclerosis and cancer. DBMSCs can protect the functions of endothelial cells from the damaging effects of H 2 O 2 and monocytes. 9,10 In addition, DBMSCs induce the generation of anti-cancer immune cells known as M1 macrophages. 11 In this study, we extended our research on the suitability of
The viability of DBMSCs and HUVEC was determined using Trypan blue. DBMSCs (passage 3) and HUVEC (passages 3-5) of twenty placentae and umbilical cords, respectively, were used in this study.

| HUVEC proliferation by MTS assay
Four cell treatment groups [ (CMDBMSC) was produced as previously described. 10 Before adding DBMSCs (whole cells) to HUVEC culture, they were treated with Mitomycin C to inhibit their proliferation. 10 Blank was cells incubated alone with MTS in complete endothelial cell growth medium.
The concentration of glucose (100 mmol/L) and culture time (72 hours) were chosen based on our previous findings at which concentration and incubation time, the proliferation of HUVEC was significantly reduced without affecting their viability (>95%). 13 The viability of DBMSCs at 100 mmol/L glucose was also >95%.
The concentration of CMDBMSC and ratio of DBMSCs were also chosen because these parameters significantly increased the proliferation of HUVEC. 10 Results were presented as means (± standard errors). Each experiment was conducted in triplicate and repeated with five independent HUVEC (passage 3-5) and DBMSCs (passage 3) preparations.

| HUVEC adhesion and proliferation using xCELLigence system
Six treatment groups of cells were used [Table S1(ii) and illustrated in Figure  The adhesion and proliferation of HUVEC were evaluated using the xCELLigence system (RTCA-DP; Roche Diagnostics) as previously described. 9,10,13 Briefly, 2 × 10 4 HUVEC [Table S1(ii) and illustrated in Figure 1A-F] were cultured in complete endothelial cell growth medium in quadruplicate wells of 16-well culture E-Plates (Catalogue number 05469813001, Roche Diagnostics), and the proliferation index (mean ± standard errors) was measured as previously described. 9,10,13 For cell adhesion, data were measured at 2 hours, whereas the rate of cell proliferation was determined by calculating the normalized cell index at 24, 48 and 72 hours after the adhesion time-point (2 hours). Each experiment was performed and repeated as described above.

| HUVEC migration using xCELLigence system
Eight groups of HUVEC treatments were used in the migration experiment [Table S1(iii) and illustrated in Figure 2A-H]. Migration of HUVEC was evaluated using CIM migration plates (Catalogue number 05665825001, Roche Diagnostics) in the xCELLigence system as previously described. 9,10,13 In all migration groups, the upper chamber contained serum free medium whereas the lower chamber contained HUVEC medium supplemented with 30% FBS. To initiate the experiment, 2 × 10 4 HUVEC [Table S1(iii) and illustrated in Figure 2A-H] were seeded in quadruplicate wells in the upper chamber, and the migration index was then measured at 24 hours as previously described. 9,10,13 Positive (30% FBS) and negative (without 30% FBS) controls were used. Each experiment was performed and repeated as described above.

| DBMSC effect on monocyte invasion through a monolayer of endothelial cells
The permeability of HUVEC was evaluated by determining the ability of monocytes (THP-1) to invade through a monolayer of HUVEC using the xCELLigence system as previously described. 13 Six treatment groups were used in the invasion experiments [Table S1(iv)].
The invasion experiments were initiated by seeding HUVEC (2 × 10 4 ) in 16-well culture E-Plate, and THP-1 (10 4 ) were added to HUVEC monolayer after HUVEC reached growth plateau as previously described. 13 After 10 hours, the cell invasion index (mean ± standard errors) was measured by calculating the normalized cell index at pausing time (15-20 hours) of HUVEC growth. 13 Five experiments were performed in triplicate using HUVEC and DBMSCs as described above.

| Capillary network formation Assays
Four cell groups were used in the tube formation experiments [  Corning), and capillary network formation of HUVEC was then observed for 14 hours as previously described. 13 Images were taken from each well using a bright field microscope at a magnification of 4×. Nodes of capillary network were counted, and the results were presented mean ± standard deviation. Experiments were conducted in triplicate and repeated as described above.

| Real-time polymerase chain reaction
Expression of an array of genes in endothelial cells [  13 Data were analysed and expressed as fold change by calculating the ΔΔ −2 values. The relative expression of an internal housekeeping genes was used as control as previously described. 13 Experiments were conducted in triplicate and repeated three times using HUVEC and DBMSCs as indicated above.

| Statistical analysis
GraphPad Prism 5 software was used to analyse the data by the t test (unpaired t test, two-tailed). Results were considered to be statistically significant if P < .05. HUVEC migration in response to 100 mmol/L glucose alone [100 (out)], added to the lower chamber of the migration plate, as compared to glucose-untreated HUVEC after 24 hours (J). Migration of HUVEC pre-treated with glucose (Pre-Glu), or with 100 mmol/L glucose and CMDBMSC (Pre-CM) compared to glucose-untreated HUVEC (K). Each experiment was performed in triplicate using HUVEC (passage 3-5) and DBMSCs (passage 3) from five independent umbilical cord tissues, and placentae, respectively. Bars represent standard errors

| HUVEC proliferation in response to glucose and DBMSCs
DBMSCs as previously described 12 were used to study their effects on proliferation of glucose-treated HUVECs.
The MTS results show that the treatment with 100 mmol/L glucose [100] for 72 hours significantly reduced the proliferation of HUVEC (P < .05) as compared to untreated HUVEC ( Figure 3).

| Effects of DBMSCs and glucose on the reversibility of HUVEC proliferation
The results of xCELLigence system showed that at 24 and 48 hours, proliferation of HUVECs pre-treated with glucose [Pre-Glu] was not significantly changed (P > .05) ( Figure 4A,B), but was significantly (P < .05) reduced at 72 hours as compared to untreated control ( Figure 4C). These results show that inhibitory effect of glucose on HUVEC proliferation is irreversible. In contrast, the effects of CMDBMSC on glucose-treated HUVEC proliferation are reversible.
As shown in Figure

| HUVEC adhesion in response to glucose and DBMSCs
The results of the xCELLigence system showed that at 2 hours,

| HUVEC migration in response to glucose and DBMSCs
The  Figure 2I.

F I G U R E 3
The proliferation of HUVECs in response to 100 mmol/L glucose and different treatments of DBMSCs. MTS assay showed that proliferation of HUVEC in response to 100 mmol/L glucose significantly reduced as compared to glucoseuntreated HUVEC controls after 72 hours, whereas treatment with 25% CMDBMSC (100 + CM) and 100 mmol/L glucose for 72 hours significantly increased HUVEC proliferation as compared to glucose-untreated and treated HUVECs. HUVEC proliferation in response to 100 mmol/L glucose and DBMSCs (100 + DBMSC) significantly reduced as compared to glucose-untreated HUVEC and did not significantly change as compared to glucose-treated HUVEC after 72 hours. Each experiment was performed in triplicate using DBMSCs (passage 3) and HUVEC (passage 3-5) from five independent placentae and umbilical cord tissues. *P < .05. Bars represent standard errors

| DBMSCs reduce the effect of glucose on HUVEC permeability
The permeability of endothelial cells was assessed by examining the ability of monocytes (THP-1) to invade a monolayer of HUVECs using the xCELLigence system as previously reported. 13 Increased invasion of monocytes through HUVEC monolayer is defined as a reduction in the cell index due to the detachment of HUVECs as a result of infiltration by monocytes, whereas increased cell index reflects the reduction in monocyte invasion. 13 As compared to un-

| DBMSCs protect HUVEC capillary network formation from the damaging effect of glucose in vitro
The results of capillary network formation show that after 14 hours, untreated HUVECs formed capillary networks ( Figure 7A) whereas their incubation with 100 mmol/L glucose disturbed this ability of HUVECs to form capillary networks ( Figure 7B). Addition of CMDBMSC to HUVEC culture in presence of 100 mmol/L glucose protected capillary network formation of HUVECs ( Figure 7C), whereas the co-culture of HUVECs with DBMSCs in presence of 100 mmol/L glucose did not reverse the inhibitory effect of glucose on HUVEC capillary network formation (data not shown). We quantified the number of capillary nodes formed and found that there was significant decrease in the number of HUVEC nodes formed after their treatment with 100 mmol/L glucose. However, in the presence of CMDBMSC coupled with 100 mmol/L glucose, node formation was significantly restored (P-value <.05) ( Figure 7D).

| DBMSCs modify glucose effect on endothelial cell expression of various genes
The results of RT-PCR showed that DBMSCs modified a number of genes in HUVECs after treatment with glucose. These genes have important functional roles in the HUVEC biology. (Tables 1-6).

| D ISCUSS I ON
Recently, we reported the therapeutic potential of DBMSCs to treat inflammatory diseases, such as atherosclerosis and cancer. [9][10][11] DMSCs protect endothelial cell functions from the negative effects  Table 1.
Previously, we reported that DBMSCs express IL-6, IL-10 and VEGF. 12 These molecules play important roles in cell survival, cell proliferation and cell migration activities. [20][21][22] Therefore, these molecules may mediate the protective functions of DBMSCs on glucose-treated endothelial cells. However, this possibility will be addressed in future by performing functional studies to reveal the molecular pathways involved in this property.
CMDBMSC but not ICDBMS reversed the inhibitory function of glucose on endothelial cell migration as we and others have reported already 13,23 (Figure 2A). The stimulatory effect of CMDBMSC on endothelial cell migration is irreversible ( Figure 2L) and is possibly mediated by a mechanism that involves a number of pathways involving ADAM17, 24 ICAM1 19 and VEGFA 25 as summarized in Table 2. These data further confirm the protective roles that DBMSCs employ on glucose-treated endothelial cells.
Angiogenesis is an important biological functions of endothelial cells 10 that are also disturbed by glucose as recently reported by us 13 and others. 17 This inhibitory effect of glucose on the angiogenic activity of endothelial cells was also confirmed in this study ( Figure 7B).
We found that anti-angiogenic effect of glucose on endothelial cells can be reversed by CMDBMSC ( Figure 5C,D), but not by DBMSCs, as previously reported for pMSC (MSCs isolated from the chorionic villi F I G U R E 7 HUVEC tubule formation in presence of glucose and DBMSCs. After 14 hours, HUVECs were able to form tube networks (A). But, treatment with 100 mmol/L glucose reduced this property is HUVECs. (B) However, HUVECs cultured with 100 mmol/L glucose and subsequent treatment with 25% CMDBMSC restored their tube formation potential. (C) Semi-quantitative analysis for the number of capillary nodes formed was done by counting the number of nodes formed in each experimental group(d)A significant decrease in the number of nodes formed was observed in HUVECs treated with 100 mmol/L glucose. However, in the presence of CMDBMSC with 100 mmol/L glucose node formation was significantly restored (D). Each experiment was performed in triplicate using HUVEC (passage [3][4][5] and DBMSCs (passage 3) from five independent umbilical cord tissues, and placentae, respectivel of human term placenta). 13 We also found that CMDBMSC increased and reduced the expression of pro-angiogenic genes (ICAM1, 19 CCL2 26 and VEGFA 25 ) and anti-angiogenic gene (CASP1 27 ), respectively, in glucose-treated endothelial cells, (Table 2). In addition, ICDBMSC increased the expression of a number of anti-angiogenic genes including TIMP-1, 28 PF4 29 and AGT 30 (Table 5). Expression data of these genes suggest that they may mediate the effects of DBMSCs (CMDBMSC and ICDBMSC) on glucose-treated endothelial cell angiogenic activity.
However, ICDBMSC also reduced and increased the expression of anti-angiogenic gene (CASP1 27 ) and pro-angiogenic genes (TYMP, 31 KIT 32 and CCL2 26 ), respectively, in glucose-treated endothelial cells (Table 5). These data suggest that DBMSCs may have TA B L E 1 CMDBMSCs modulate the expression of genes involved in endothelial cell (EC) proliferation, survival, injury and inflammation  13,35 Sequentially, this will increase monocyte infiltration through endothelial cells.. 13,35 CMDBMSCs, but not ICDBMSC, reduced glucose-inducing monocyte infiltration through endothelial cells ( Figure 4) via a mechanism that may involve a reduction in endothelial expression of PLG, which is involved in leucocyte infiltration, 36 (Tables 1 and 4). However, DBMSCs also increased glucose-treated endothelial cell expression of genes that induce their injury (CX3CL1, 50 THBD 51 and TNF 52 ), apoptosis (CAV1 53 and CCL2 54 ) and inflammation (THBS1 55 ) (Tables 1   and 4).

| CON CLUS IONS
This is the first report to demonstrate the protection of endothelial cell from harmful effects of glucose by DBMSCs. The mechanisms involve multiple genes involved in endothelial cell function.
This study indicates that DBMSCs are promising therapeutic agents for the treatment of complicated diseases such as thrombosis and atherosclerosis that are associated with endothelial cell injury in diabetes. The therapeutic significance of DBMSCs must be validated in pre-clinical animal studies.

ACK N OWLED G EM ENTS
We also acknowledge King Abdul Aziz Medical City for providing us with placentae.
Dedication: This paper is dedicated to the late 'Professor Mohamed AbuMaree', who was the main architect of this study. We will continue to be inspired by his dedication and profound passion for the stem cell research at KAIMRC. May you rest in peace.

CO N FLI C T O F I NTE R E S T
No competing financial interests exist. The authors declare that there is no conflict of interests regarding the publication of this paper. Abomaray, contributed to data analysis and interpretation of results.

AUTH O R CO NTR I B UTI
All authors reviewed the manuscript.

E TH I C S A PPROVA L A N D CO N S E NT TO PA RTI CI PATE
The institutional review board (IRB) at King Abdulla International Medical Research Centre (KAIMRC), Saudi Arabia, approved this study. Samples (ie placentae and umbilical cords) were obtained from uncomplicated human pregnancies (38-40 gestational weeks) following informed patient consent.
Consent for publication: 'Not applicable'. All authors agree to publish this manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated during this study are included in this published article.