hUC-MSCs improved fetal growth by lowering maternal blood pressure and correcting kidney function in L-NAME induced-preeclampsia rats
To determine whether hUC-MSCs had any impact on preeclampsia, we first subcutaneously injected rats with 200 mg/kg of L-NAME daily, from days 13–17 of pregnancy to induce preeclampsia, followed by tail vein injection of hUC-MSCs from days 15–19 of pregnancy (Fig. 1A). Three groups, comprising 10 rats each, were thus established: control who only received saline (Ctrl), L-NAME rats who did not receive hUC-MSC (L-NAME), and L-NAME rats who received hUC-MSC (L-NAME + MSC). After 20 days of pregnancy, fetuses, placentas, as well as maternal kidney and serum, were analysed, and representative images of fetuses and placentas from all 3 groups are shown in Fig. 1B. We found that fetal weights were significantly lower among L-NAME compared to control, while L-NAME + MSC was intermediate between the 2 (Fig. 1C). However, no significant difference between the 3 groups was present in terms of fetal survival (Fig. S2A), while a significant difference was only present between Ctrl versus L-NAME and L-NAME + MSC with respect to fetal length, where it was significantly longer in Ctrl compared to the other 2 groups (Fig. S2B). As for placental weight, it was also significantly lower among L-NAME and L-NAME + MSC, compared to Ctrl (Fig. 1D). Therefore, compared to L-NAME, there was an improvement in fetal weight among L-NAME + MSC.
Additionally, maternal blood pressure and urinary protein levels were monitored among the 3 groups, in which for both systolic and diastolic blood pressures, hUC-MSC injection alleviated the L-NAME-associated pressure increases during 15–20 days of pregnancy (Fig. 1E). With respect to urinary proteins, L-NAME had significantly higher 24 hr-urinary protein levels at day 19 of pregnancy, compared to Ctrl and L-NAME + MSC (Fig. 1F), which is likely owed to glomerular dysfunction. As observed by H&E staining, L-NAME had collapsed glomerular capillary loops, disordered structures, and enlarged renal tubule lumens, with some ruptures present. All these alterations are alleviated with hUC-MSC administration (Fig. 1G). These observations are further supported by L-NAME having significantly lower glomerular densities compared to Ctrl, which significantly increased towards that of Ctrl in L-NAME + MSC, indicating that hUC-MSCs could alleviate maternal kidney dysfunction (Fig. 1H).
Furthermore, L-NAME + MSC, compared to L-NAME, had significantly lower BUN levels, similar to that of Ctrl (Fig. 1I); a similar pattern, but without statistical significance, was present for Cr (Fig. 1J). All these findings thus suggest that hUC-MSCs improved pregnancy outcomes by lowering maternal blood pressure and improving kidney function in L-NAME induced-preeclampsia.
hUC-MSCs improved blood perfusion to the fetus in preeclampsia rats by increasing placental angiogenesis
The effects of hUC-MSCs on placental tissue in preeclampsia were analyzed by H&E staining (Fig. 2A), in which the ratios of the labyrinth/junctional zone area, as well as labyrinth/total area, are significantly lower among L-NAME rats; this lowered labyrinth area indicates lowered capabilities to ensure adequate maternal and fetal blood supply exchange (Fig. 2B). However, hUC-MSC administration restored both of those ratios towards that of Ctrl (Fig. 2B). To further examine the possible basis behind the increased labyrinth area among L-NAME + MSC, immuno-histological and immune-fluorescence staining were conducted, in which compared to Ctrl, L-NAME had significantly lower CD31+, representing capillaries. On the other hand, L-NAME having higher α-SMA+ densities, which is due to the spiral arteries in the placenta failing to undergo proper conversion to highly dilated, thin-walled vessels required for proper uteroplacental blood flow. As a result, the spiral arteries exhibited thickened vessel walls and narrowed vessel lumens, thereby resulting in higher arterial densities to compensate (Fig. 2C-D). L-NAME + MSC, though, had CD31+ and α-SMA+ densities approaching that of Ctrl (Fig. 2C-D). These observations therefore indicate that hUC-MSCs could improve blood perfusion to the fetus in preeclampsia by increasing placental angiogenesis and attenuating atherogenesis.
hUC-MSCs promoted placental angiogenic and anti-inflammatory factors, along with suppressing oxidative stress/apoptotic protein expression in preeclampsia rats
To further verify the association between hUC-MSCs and the promotion of placental angiogenesis, we examined the levels of angiogenesis-related factors in maternal blood. We found that L-NAME had significantly lower levels of pro-angiogenic factors VEGF and PIGF, which were restored towards that of Ctrl in L-NAME + MSC (Fig. 3A). On the other hand, significantly higher levels of anti-angiogenic factors sFLT-1 and sENG were present in L-NAME, which were reduced towards that of Ctrl in L-NAME + MSC (Fig. 3A).
Other processes involved in preeclampsia pathogenesis are excessive inflammation and oxidative stress. Indeed, we found that the levels of pro-inflammatory factors IL-6, IFN-γ, and TNF-α significantly increased among L-NAME; these increases, however, were reversed towards that of Ctrl among L-NAME + MSC (Fig. 3B). Conversely, the anti-inflammatory factor IL-10 is significantly lower among L-NAME versus Ctrl, while hUC-MSC administration restored IL-10 to similar levels as that of Ctrl (Fig. 3B). Similar patterns were present for oxidative stress indicators, in which the oxidative stress indicator, MDA, significantly increased among L-NAME versus Ctrl. This increase, though, was reversed towards that to Ctrl among L-NAME + MSC (Fig. 3C). On the other hand, anti-oxidative markers eNOS, NO, and SOD were significantly lower in L-NAME than Ctrl; these levels were restored towards that of Ctrl in L-NAME + MSC (Fig. 3C). All these findings regarding increased angiogenesis-related factors, as well as lowered inflammation and oxidative stress associated with hUC-MSC administration, are further supported by L-NAME placentas having increased pro-apoptotic Bax, and lowered anti-apoptotic Bcl-2 expression, compared to Ctrl (Fig. 3D-E). hUC-MSC administration, however, reversed those levels towards that of Ctrl (Fig. 3D-E). In summary, hUC-MSC was able to improve placental angiogenesis by increasing pro-angiogenic factor production, along with lowering inflammation, oxidative stress, and apoptosis.
hUC-MSCs counteracted LPS-induced detrimental effects on HTR8/HUVEC cell proliferation, migration and apoptosis in vitro
To further examine the effects of hUC-MSC on placental cells in preeclampsia, an in vitro preeclampsia model was established, where in a Transwell, HTR-8/SVneo trophoblasts or HUVECs were seeded in the lower chamber, followed by LPS administration to induce cell states similar to that observed in preeclampsia. hUC-MSCs were then seeded into the upper chamber to rescue those cells (Fig. 4A). Under the clonogenic assay, we found that LPS administration significantly lowered HTR-8 and HUVEC cloning efficiencies, reflecting lowered cell proliferation (Fig. 4B-C). These reductions, though, were restored back towards that of control after hUC-MSCs were added in the LPS + MSC group (Fig. 4B-C). This lowered cell proliferation in LPS was further correlated with flow cytometry analysis of cell cycle progression, in which LPS, compared to Ctrl, had significantly higher proportions of arrested cells, at the G1 phase, and lower proliferative S and G2 phase cells, for both HTR-8 and HUVECs. However, these proportions were more similar to Ctrl in LPS + MSC (Fig. 4D-E).
To verify whether hUC-MSC administration was also able to counteract against apoptosis in vitro, flow cytometry of Annexin-V/7AAD stained cells was carried out, in which among both HTR8 and HUVECs, LPS had significantly increased apoptotic cell counts. hUC-MSCs, though, were able to reverse those levels towards that of Ctrl in LPS + MSC (Fig. 4F-G).
We then examined the invasion capabilities of HTR8 and HUVECs, in which the bottom of a Transwell was coated with Matrigel, followed by seeding of HTR-8 or HUVECs in the upper chamber, which were treated with LPS. hUC-MSCs were seeded in the lower chamber, and the extent of cell invasion was examined after 24 hr of co-culture. We observed that compared to Ctrl, LPS-treated HTR8 and HUVECs had significantly lower 24 hr invasion rates, which increased back towards that of Ctrl in LPS + MSC (Fig. 4H-I). All these findings thus indicate that in line with in vivo findings, hUC-MSCs counteracted LPS-induced detrimental effects on HTR8/HUVEC cell proliferation, migration and apoptosis in vitro.
hUC-MSCs corrected pre-eclampsia-associated mitochondrial dysfunction by decreasing pathological autophagy
Increased apoptosis and ROS production has been associated with pathological alterations in autophagic activities. To verify whether this was present in preeclampsia, we first examined ROS levels in our in vitro model among Ctrl, LPS, and LPS + MSC groups. We found that in line with our maternal serum findings, ROS levels were significantly higher among LPS-treated HTR8 and HUVECs, which were lowered towards that of Ctrl upon hUC-MSC administration in LPS + MSC (Fig. 5A-B). We then examined the extent of autophagy among the 3 groups, in which LPS also had significantly higher MDC levels, indicating increased autophagy, among both HTR8 and HUVECs, compared to Ctrl; these levels, though, were lowered towards that of Ctrl in LPS + MSC (Fig. 5C-D). To further verify these observations in vivo, TEM analyses of placental mitochondrial ultra-structures were conducted, in which compared to Ctrl, L-NAME-treated pre-eclamptic rats had mitochondrial swelling and reduced cristae numbers. These pathological alterations, though, are partially reversed upon hUC-MSC administration (Fig. 5E). Furthermore, the density of normal-appearing mitochondria was significantly lower, and autophagosome density higher, in L-NAME compared to Ctrl, while L-NAME + MSC levels for both densities were reverted towards that of Ctrl (Fig. 5F-G). All these findings indicate that pre-eclampsia increased autophagy, which is linked to elevated oxidative stress and mitochondrial dysfunction. By contrast, hUC-MSC administration alleviated mitochondrial dysfunction by decreasing ROS and autophagy.
hUC-MSCs lowered pathological pre-eclamptic placental autophagy by activating Akt/mTOR and inhibiting AMPK/mTOR pathways
To further investigate the ability of hUC-MSC administration to lower pathological placental autophagy, relative mRNA and protein expression levels were measured for various autophagy-related mediators. We found that mRNA expression of pro-autophagic LC3 and Beclin1 significantly increased (Fig. 6A-B), and anti-autophagic P62 decreased (Fig. 6C), in L-NAME versus Ctrl. These changes were reverted towards that of Ctrl upon hUC-MSC administration (Fig. 6A-C). With respect to protein, Western blot analyses similarly found that the elevated levels of LC3 II/I ratio and Beclin1 in L-NAME was decreased upon hUC-MSC treatment (Fig. 6D-E). On the other hand, P62 protein level was significantly higher in L-NAME + MSC compared to L-NAME (Fig. 6D-E).
The mTOR-regulated signaling pathway is crucial for cell invasion and vascular remodeling, along with being involved in placental ischemia and hypoxia. To investigate if this pathway is responsible for the increased pathological autophagy in pre-eclampsia, RT-qPCR was first used to measure mRNA expression levels for mTOR pathway components mTOR, AMPK, and Akt. The results showed that mTOR and Akt mRNA levels were significantly lower in L-NAME but were restored around Ctrl levels in L-NAME + MSC (Fig. 6F-G). Conversely, AMPK expression was significantly higher in L-NAME than the other 2 groups (Fig. 6H). To determine whether these alterations were also present at the protein level, Western blot was conducted, in which the p-mTOR/mTOR and p-Akt/Akt ratios were significantly lower in L-NAME but increased to levels similar to Ctrl in L-NAME + MSC (Fig. 6I-J). The opposite, though, was the case for p-AMPK/AMPK, where expression levels were significantly higher in L-NAME than Ctrl and L-NAME + MSC (Fig. 6I-J). All these analyses thus suggested that hUC-MSCs were able to lower pathological pre-eclamptic placental autophagy, via activating Akt/mTOR and inhibiting AMPK/mTOR pathways.