All experiments on human tissues were approved by the Ethics Committee of the Zhejiang Provincial People’s Hospital, Zhejiang, China. Written informed consent was obtained from all subjects. PD-MSCs were isolated from placental villous tissues procured from healthy women during labor. Briefly, placental villous tissues were cut into approximately 1 mm${}^3$ pieces and treated with Collagenase type I (1 mg/mL, Sigma-Aldrich, St. Louis, MO, USA) for 1 h at 37 °C. After passing through a 100-µm strainer (Falcon, BD Biosciences, Franklin Lakes, NJ, USA), the cell suspension was seeded into cell culture plates with serum-free MSC culture medium (SC2013-G, TBD Science, Tianjin, China) at a density of 1 $\times$ 10${}^5$/cm${}^2$. The plates were then incubated at 37 °C in a humidified incubator with 5% CO${}_2$ and passaged after digestion with 0.25% trypsin-ethylenediaminetetraacetic acid (1:5 ratio) (15400054, Gibco, Thermo Fisher Scientific, Waltham, MA, USA).

PD-MSCs were characterized by their morphology, surface marker expression, and mesenchymal lineage differentiation, as previously described [17]. Osteogenic, adipogenic and chondrogenic inductions were performed to evaluate the multi-lineage differentiation capacity of PD-MSCs. To evaluate mineralization, the cultures were stained with Alizarin Red S according to the manufacturer’s instructions (HUXXC-90021, Cyagen Biosciences Inc, Santa Clara, CA, USA). Briefly, cells were fixed with 4% paraformaldehyde (P0099, Beyotime Institute of Biotechnology, Shanghai, China) for 30 minutes and then stained with Alizarin Red S for 5 minutes at room temperature. The lipid droplets formed following induction of adipogenesis were visualized using Oil Red O staining (HUXXC-90031, Cyagen Biosciences Inc). Successful chondrogenic differentiation was tested by Alcian blue staining (HUXXC-90041, Cyagen Biosciences Inc).

The PD-MSC phenotype after the third passage was analysed by flow cytometry (BD FACSVerse, BD Biosciences, Franklin Lakes, NJ, USA) using the following antibodies: anti-CD14 (BD Biosciences, Franklin Lakes, NJ, USA, cat. no. 555397, 1:5 dilution), anti-CD45 (BD Biosciences, cat. no. 560973, 1:5 dilution), anti-CD19 (BD Biosciences, cat. no. 563326, 1:5 dilution), anti-CD34 (BD Biosciences, cat. no. 555822, 1:5 dilution), anti-CD73 (BioLegend, San Diego, CA, USA, cat. no. 344004, 1:20 dilution), anti-CD105 (BioGems, Westlake Village, CA, USA, cat. no. 17111-60, 1:20 dilution), anti-HLA-DR (BD Biosciences, Cat. no. 555811, 1:5 dilution) and anti-CD90 (BioLegend, cat. no. 328110, 1:5 dilution). Incubation with each of these antibodies was at room temperature for 15 minutes according to the manufacturer’s instructions. A suitable isotype-matched antibody (PE; BD Biosciences, cat. no. 555749; FITC, BD Biosciences, cat. no. 555573, 1:5 dilution) incubated at room temperature for 15 minutes was utilized as a negative control. The data were analyzed using BD FACSuite software (version 1.0.5.3841, BD Biosciences, Franklin Lakes, NJ, USA).

All animal experiments were approved by the Institutional Animal Care and Use Committee of Zhejiang Provincial People’s Hospital. Pregnant Sprague Dawley rats (Shanghai SLAC Laboratory Animal Co. Ltd, Shanghai, China) at day 18 of gestation were used to isolate hippocampal neurons from the brains of late rat embryos. Isolation and culture of the hippocampal neurons was carried out using a previously described protocol [18]. Briefly, rat embryos were harvested from pregnant rats at 18 days gestation. They were then dissected under a stereomicroscope and in an ice-cold and sterile environment. After exposing the surface of the cerebral cortex, the hippocampus was gently separated, cut into 1 mm${}^3$ pieces, digested with 2 mL of 0.25% pancreatic enzyme solution (25200-072, Gibco, Thermo Fisher Scientific, Waltham, MA, USA) for 20 minutes at 37 °C, and then filtered through a 150-mesh stainless steel filter (352360, Corning, Somerville, MA, USA). A suspension containing 1 $\times$ 10${}^6$ cells was added to a 35 mm diameter sterile plastic petri dish soaked overnight in 0.01% poly-L-lysine hydrobromide (P8140, Solarbio, Beijing, China). The cells were cultured in an incubator at 37 °C under 5% CO${}_2$. All medium was replaced the next day with DMEM containing 10% FBS and a 1% penicillin/streptomycin (P/S) solution. Half the medium was replaced every three days thereafter. Cell growth was observed each day with a stereomicroscope.

To construct an in vitro hypoxic-reoxygenated cell model, hippocampal neurons were grown for six days in normal culture (37 °C, 5% CO${}_2$, 95% atmosphere) and then placed in a hypoxic tank (37 °C, 5% CO${}_2$, 95% N${}_2$) for 2.5 hours. Third-generation PD-MSCs were obtained using the isolation and identification method established earlier by our research group. Three groups were set up for further experiments: (a) H/R group: hippocampal neurons with hypoxic-reoxygenated injury (H/R cells) were cultured in Transwell double-layer plates (28419067, CoStar, Suzhou, Jiangsu, China). (b) Co-culture group: H/R cells were cultured in the lower Transwell chamber, and PD-MSCs were cultured in the upper chamber without direct contact to H/R cells. (c) Control group: normal hippocampal neurons were cultured continually until the eleventh day. All three groups were incubated in a 37 °C incubator with 5% CO${}_2$ for 5 days, with the cell culture medium changed every second day.

Cell cultures were fixed with 4% paraformaldehyde and permeabilized with 0.5% Triton X-100 (Sigma-Aldrich, St. Louis, MO, USA) for 10 minutes each, incubated with 5% bovine serum albumin (BSA) for 1 hour, and then stained overnight at 4 °C with mouse anti-neuron-specific enolase (anti-NSE; ab218388, Abcam, Waltham, MA, USA) or rabbit anti-growth-associated protein 43 (anti-GAP-43; AF0150, Beyotime Institute of Biotechnology). Subsequently, the cells were washed with PBS and incubated for 30 minutes at room temperature with Alexa Fluor 488/594-conjugated secondary antibodies (Jackson ImmunoResearch, West Grove, PA, USA). After further washes with PBS and counterstaining with 4${}^{\prime }$,6-diamidino-2-phenylindole (DAPI) solution, the cells were placed onto microslides and mounted with Dako Fluorescent Mounting Medium (Dako, Glostrup, Denmark). Cells were visualized and microphotographed using an FV1000 confocal laser scanning microscope (Olympus, Tokyo, Japan).

The HIE model was established based on previous studies by clipping the bilateral uterine arteries for 5 minutes and delaying caesarean delivery [15, 19]. Pregnant Sprague-Dawley rats (21 days gestation) underwent general anaesthesia using 3% isoflurane in a 30% O${}_2$/70% N${}_2$ atmosphere. Once unconscious, each rat was placed in the supine position on the operating table with monitoring of vital signs. An incision was made along the midline of the abdomen, and the superficialis was isolated to expose the uterus. Uterine ischaemia was induced by blocking four uterine arteries with 30 g aneurysm clips for 5 minutes. A Caesarean section was then performed after removing the clips and recovering the blood supply. The newborn pups were carefully rescued and raised manually.

The newborn pups were randomly assigned to six groups (9 pups per group) as follows: control group, HIE model group, HIE + normal saline (NS) group, and three HIE + PD-MSC treatment groups (transplantation at days 7, 14 and 28 postpartum, giving the P7, P14 and P28 groups, respectively). Five microlitres of PD-MSC cell suspension (5 $\times$ 10${}^5$ cells/µL) was transplanted through the jugular vein (injection speed of 1 µL/min) in the P7, P14 and P28 groups. Pups from the control and HIE + NS groups received the same dose of saline injection on postpartum day 14.

Rats were evaluated for neurological function at 56–60 days postpartum.

After neurological evaluation, five rats from each group were chosen randomly and the brains collected under anaesthesia. Brain tissues were dehydrated in ethanol and embedded in paraffin blocks. Five-micrometer-thick slices were dewaxed in xylene, rehydrated in progressively lower concentrations of ethanol, and then fully stained with haematoxylin and eosin (H&E) (C0105M, Beyotime Institute of Biotechnology) or cresyl violet (G1430, Solarbio).

At 60 days after HIE injury, brain tissue homogenates were prepared for Western blot analysis according to the manufacturer’s instructions. Briefly, brain tissue was added to a cold lysis buffer (P0013B, Beyotime Institute of Biotechnology), and the supernatants were collected and quantified using a BCA protein assay kit (P0011, Beyotime Institute of Biotechnology). Samples (20 µg) were separated on 10% polyacrylamide gels and transferred to Hybond-P membranes. Incubation of the membranes with a BSA solution (5%; ST025, Beyotime Institute of Biotechnology) at room temperature for 45 minutes was used to block non-specific binding. Primary antibodies against the following proteins were then added and incubated at room temperature for 45 minutes: B-cell lymphoma-2 (BCL-2; 12789-1-AP, Proteintech Group, Inc. Rosemont, IL, USA), BCL-2-associated X protein (BAX; AF0057, Beyotime Institute of Biotechnology), BCL-2-associated agonist of cell death (BAD; AF1009, Beyotime Institute of Biotechnology), neurofilament H/M (NF-H/M; 835403, BioLegend), microtubule-associated protein-2 (MAP-2; 822501, BioLegend), Caspase3 (AB3623, Sigma-Adlrich, St. Louis, MO, USA), and Caspase7 (AB256474, Abcam). GAPDH (AF1186, Beyotime Institute of Biotechnology) served as the loading control. The blots were then incubated with an HRP-conjugated goat anti-rabbit secondary antibody (A0208, Beyotime Institute of Biotechnology) for 45 minutes at room temperature, and subsequently analyzed using an enhanced chemiluminescence (ECL) kit (Beyotime Institute of Biotechnology). ImageJ (National Institutes of Health, Bethesda, MD, USA; http://rsb.info.nih.gov/ij/, v1.8.0) was employed to analyze the greyscale values.

ELISA kits were used to determine the concentrations of tumor necrosis factor $\alpha$ (TNF-$\alpha$), interleukin (IL)-1$\beta$, IL-10 and transforming growth factor (TGF)-$\beta$1 in brain tissue, as well as IL-10 and TGF-$\beta$1 in conditioned culture medium, according to the manufacturer’s instructions (Supplementary Table 1, MLBio, Shanghai, China).

Data are presented as the mean and standard deviation (mean $\pm$ standard deviation (SD)). One-way analysis of variance (ANOVA) was used for comparisons between groups. Statistical analysis was performed using SPSS software version 22 (SPSS Inc., Chicago, IL, USA). p 0.05 was considered statistically significant.

PD-MSC were derived from placental villous tissues procured from healthy women during labor. The cells exhibited a bipolar spindle-like shape and displayed a high capacity to adhere to plastic when maintained in standard culture conditions using tissue culture flasks (Fig. 1A). PD-MSCs showed the capacity to differentiate into adipocytes, chondrocytes, and osteoblasts after being cultured in the appropriate induction conditions (Fig. 1B). Flow cytometry showed that PD-MSCs had positive expression of CD73 (99.8%), CD90 (99.4%) and CD105 (95.2%), and negative expression of the hematopoietic lineage markers CD14 (0.32%), CD19 (0.04%), CD34 (0.03%), CD45 (1.72%) and HLA-DR (0.02%). These differentiation abilities and the surface marker expression profile were in accordance with typical MSC standards (Fig. 1C).

Fig. 1.

Characteristics of human placenta-derived mesenchymal stem cells. (A) Morphology of MSCs. (B) Differentiation of MSCs was observed after induction (original magnification $\times$ 200). Alizarin Red S staining of extracellular calcium deposits confirmed osteogenic differentiation; Oil-red-O-positive lipid droplets in the cytoplasm indicated adipocytic differentiation; Alcian Blue staining of aggrecan confirmed chondrogenic differentiation. (C) Flow cytometry results showing the immunophenotype of PD-MSCs. Scale bars = 50 µm. MSCs, mesenchymal stem cells; PD-MSCs, placenta-derived mesenchymal stem cells.

We developed a new type of intrauterine HI rat model in order to study the therapeutic efficacy of PD-MSCs and the underlying mechanism of action. Rats in the HIE group had lower body weight than those in the control group. After treatment with PD-MSCs, the mean weight was higher in the P7 (266.62 $\pm$ 10.91 g), P14 (263.13 $\pm$ 10.94 g) and P28 (244.00 $\pm$ 10.61 g) groups compared to the HIE (241.89 $\pm$ 12.68 g) and HIE + normal saline (NS) groups (243.91 $\pm$ 12.06 g, Fig. 2A).

Fig. 2.

Body weight and neurological evaluation in the HIE model. (A) Weight gain by the rats. At 56 days after PD-MSC injection, the mean body weight increased until it was almost equal to the control group. (B) Rotarod test. The significantly longer time spent on the rotarod indicated that PD-MSC treatment improved coordination after HIE injury. (C) Traction test. Significant improvement in muscle strength was observed after PD-MSC injection. (D) Escape latencies for each group from the second to the fifth day of the test. The escape latencies to locate the hidden platform gradually decreased in all three PD-MSC treatment groups. (E) The escape latencies for each group on the fifth day. The P7 group showed the best results compared to the P14 and P28 PD-MSC injection times. (F) Passing times of probe trains. The significant reduction in time spent locating the platform and the increased number of passages confirmed that PD-MSC injection improved learning memory function after HIE injury. Data are mean $\pm$ SD, n = 9. * p 0.05; ** p 0.01; *** p 0.001; **** p 0.0001. HIE, Hypoxic-ischaemic encephalopathy. SD, standard deviation.

The neurological behaviors of motor coordination and balance in the HIE model were assessed using the rotarod and traction tests. The rotarod test revealed a prolonged mean latency time in the P7 (27.19 $\pm$ 12.50 s) and P14 (15.56 $\pm$ 8.06 s) groups after PD-MSC transplantation compared with the HIE (8.29 $\pm$ 5.27 s) and HIE + NS (9.24 $\pm$ 5.17 s) groups (Fig. 2B). A significant difference was observed between the P7 and the HIE groups (p 0.01). The mean scores for the traction tests were higher in the three PD-MSC treatment groups than in the HIE and HIE + NS groups (Fig. 2C). The P7 group (2.81 $\pm$ 0.18) had a higher score than the P14 (2.56 $\pm$ 0.41) and P28 (2.11 $\pm$ 0.41) groups.

Morris water maze tests were performed to evaluate the capacity to recover learning memory function in the HIE model. No significant differences in escape latencies between each group were observed on the first day of the water maze test, suggesting that neither HIE injury nor PD-MSC injection impaired rat motility and vision. The test was then performed with the hidden platform from the second to fifth days, with the escape time becoming progressively shorter with increasing training time (Fig. 2D). Compared to the HIE group (42.90 $\pm$ 34.42 s), rats in the P7, P14 and P28 groups took significantly less time to find the platform on the fifth day (12.40 $\pm$ 16.64, 16.31 $\pm$ 16.07, 22.79 $\pm$ 23.86, respectively; p 0.01, Fig. 2E). Probe trains performed on the sixth day revealed that the frequency of crossing the platform was significantly greater in the P7 (4.00 $\pm$ 1.85) and P14 (3.78 $\pm$1.99) groups (Fig. 2F), suggesting that PD-MSC injection improved HIE pup motor function and learning ability. Of the three PD-MSC treatment groups, the P7 group displayed the best recovery from HIE injury.

Histological examination and Nissl staining were carried out to evaluate the degree of brain tissue damage in the HIE model. The H&E results revealed that PD-MSC transplantation reduced histopathological damage to the brain tissue caused by HIE (Fig. 3A). Furthermore, Nissl staining showed frequent neuronal degeneration and loss in the cerebral cortex after HIE injury (Fig. 3B). PD-MSC transplantation alleviated brain injury, and cell counts showed an increased number of neurons in the PD-MSC treatment groups (Fig. 3C). Western blot analysis showed a low level of MAP-2 expression in the HIE group. However, PD-MSC injection on the 7th day after HIE injury (P7 group) increased the level to close to the normal value (Fig. 3D). The P7 group had more neurons and greater MAP-2 expression than both the P14 and P28 groups, suggesting the optimal treatment time was approximately the 7th day after HIE injury.

Fig. 3.

Histopathological results and neuron counts. (A) H&E examination of histopathological damage in brain tissues. (B) Nissl staining was used to assess neurons 60 days after HIE damage. The number of Nissl bodies was reduced in the HIE group, with this trend reversed in the PD-MSC transplantation groups. Scale bars = 100 µm. (C) A significant increase in the number of normal neurons was found after PD-MSC transplantation, with the P7 group showing the best treatment effect. Data are mean $\pm$ SD, n = 5. (D) The MAP-2 expression level indicated that PD-MSC transplantation after HIE injury reduced brain damage compared with the HIE group. The P7 group had the highest expression of MAP-2. Results shown are mean $\pm$ SD, n = 5. * p 0.05; ** p 0.01. H&E, haematoxylin and eosin; MAP-2, microtubule-associated protein-2.

Since the number of normal neurons was notably higher in the three PD-MSC transplantation groups, we further evaluated the expression levels of apoptosis-related proteins in brain tissue, including BAX, BAD and BCL-2. Western blot analysis showed that BCL-2 expression was upregulated after PD-MSC transplantation, whereas BAD and BAX expression levels were downregulated (Fig. 4A). The BCL-2 expression level was the highest in the P7 group, suggesting that PD-MSCs showed the best anti-apoptosis effects on the 7th day after HIE injury.

Fig. 4.

Expression of apoptosis-related proteins and inflammatory factors. (A) The expression of apoptosis-related proteins as determined by Western blot assay. PD-MSC transplantation increased the level of BCL-2, which inhibits cell apoptosis. In contrast, PD-MSC transplantation decreased the levels of BAX and BAD, which promote apoptosis. GAPDH was used as a loading control. (B–E) Protein expression levels of IL-1$\beta$, TNF-$\alpha$, IL-10 and TGF-$\beta$1 in brain tissue, as determined by ELISA. Results shown are mean $\pm$ SD, n = 5. * p 0.05; ** p 0.01. BAX, BCL-2-associated X protein; BAD, BCL-2-associated agonist of cell death; ELISA, Enzyme-linked Immunosorbent Assay.

To study the inflammation processes in the HIE model, the expression levels of anti-inflammatory cytokines were also examined. PD-MSC transplantation reduced production of the proinflammatory cytokines IL-1$\beta$ and TNF-$\alpha$ compared to the HIE group (Fig. 4B,C and Supplementary Document 1). The anti-inflammatory factors IL-10 and TGF-$\beta$1 were expressed at low levels in the HIE group, but were upregulated after PD-MSC injection (Fig. 4D,E and Supplementary Document 1). The best moderating effects were observed when the PD-MSC were transplanted 7 days after HIE injury, which was determined to be the optimal time for treatment.

We next conducted flow cytometry with annexin V/propidium iodide (PI) to investigate the effects of PD-MSCs on the apoptosis of injured neurons. As shown in Fig. 5A, approximately 8% of neurons were apoptotic (p 0.05) after co-culture with PD-MSCs, whereas the percentage apoptosis in the H/R model was approximately 20%. The reparative effect of PD-MSCs on the injured neurons was manifested after 24 hours of co-culture, and increased with prolonged culture time. Immunofluorescent staining showed shrinkage of the cell bodies and axons in the H/R model, while the co-culture group showed a dense neural fiber network and typical morphological characteristics (Fig. 5B).

Fig. 5.

Apoptosis was inhibited by PD-MSCs in vitro. (A) Annexin V-FITC/propidium iodide (PI) flow cytometry was performed to investigate neuronal apoptosis. Results are mean $\pm$ SD, n = 3. (B) Morphology of neurons. Immunofluorescence showing the cell bodies labelled with NSE (red), neurites labelled with GAP-43 (green), and nuclei stained with DAPI (blue). **** p 0.01, Scale bars = 50 µm. NSE, neuron-specific enolase; GAP, growth-associated protein; DAPI, 4${}^{\prime }$,6-diamidino-2-phenylindole; H/R, hypoxic-reoxygenated.

Protein extracted from the brain tissue of HIE model animals was used to stimulate PD-MSCs. The results showed that protein stimulation significantly upregulated the expression levels of IL-10 and TGF-$\beta$1 in PD-MSCs (p 0.05), which indirectly reflects the feedback regulatory effect of the HIE microenvironment on PD-MSCs (Fig. 6A,B and Supplementary Document 1). The expression levels of IL-10, TGF-$\beta$1, FGF2, Caspase3 and Caspase7 were also evaluated after co-culture of PD-MSCs with H/R neurons. ELISAs showed that the expression levels of IL-10, TGF-$\beta$1 and FGF2 were upregulated (Fig. 6C–E and Supplementary Document 1), while Western blotting showed decreased expression of Caspase3 and Caspase7 (Fig. 6F–H). These results suggest that the neuroprotective effects of PD-MSCs may be partly due to the secretion of anti-inflammatory and anti-apoptotic cytokines.

Fig. 6.

Expression levels of apoptosis-related proteins and inflammatory factors in vitro. (A,B) Anti-inflammatory cytokines in PD-MSCs were upregulated after stimulation with proteins from the HIE model. Results shown are mean $\pm$ SD, n = 5. (C–E) The expression levels of IL-10, TGF-$\beta$1 and FGF2 were upregulated in the co-culture model. Results shown are mean $\pm$ SD, n = 5. (F–H) The expression levels of Caspase3 and Caspase7 were downregulated in the co-culture model. Results shown are mean $\pm$ SD, n = 3. ** p 0.01.