Human umbilical cord mesenchymal stem cells alleviate hypoxic-ischemia-induced white-matter injury in neonatal rats by regulating polarization of microglia

Abstract

The protective impact and unique mechanisms of human umbilical cord mesenchymal stem cells (HUC-MSCs) transplantation following hypoxic-ischemic (HI)-induced brain white-matter injury (WMI) were explored. To establish a WMI model, Sprague-Dawley rats with three days after birth underwent unilateral carotid artery ligation, followed by hypoxic exposure (8% oxygen and 92% nitrogen). Subsequently, HUC-MSC transplantation was performed into the lateral ventricle. Molecular and behavioral experiments were conducted to assess how it would influence NLRP3 inflammasome activation, M1/M2 microglial polarization, and spatial cognitive abilities. HUC-MSCs promoted myelin regeneration and improved spatial cognitive function by blocking NLRP3 inflammasome activation. Furthermore, HUC-MSCs modified microglial polarization away from the M1 phenotype by downregulating the expression of CD86 and iNOS proteins and attenuating the release of proinflammatory cytokines such as TNF-α and IL-1β. They promoted anti-inflammatory cytokine production, such as TGF-β and IL-10, and the upregulation of CD206 and Arg-1 protein expression, thereby helping microglia transition to the M2 phenotype. HUC-MSCs inhibited NLRP3 inflammasome activation by antagonizing TLR4 receptors, induced microglial polarization towards the M2 phenotype in neonatal rats with WMI. HUC-MSCs seem to be a promising therapeutic option for treating WMI in premature infants.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-42445-8.

Introduction

The global incidence rates of premature infants are exceedingly high, with 1 million (6.7%) out of 15 million premature infants each year estimated to experience death¹. Babies born extremely prematurely account for over 2% of all live babies. With advances in obstetrics and neonatal nursing technology, the survival rate remains above 85% for this population². However, nearly half of the surviving extremely premature infants are accompanied by a series of adverse outcomes such as cognitive impairment, audio-visual impairment, and cerebral palsy³. White-matter injury (WMI) is considered the main cause of adverse outcomes in premature infants and its pathogenesis exhibits significant heterogeneity^4,5, mainly involving oxidative stress, glutamate excitotoxicity, free radical damage, and inflammatory response. Among these mechanisms, epidemiological analysis has found a high correlation between neuroinflammation and the development of WMI in premature infants⁶. As hypoxic ischemia (HI) induces pro-inflammatory cytokine mediated inflammatory responses in susceptible infants⁷, ameliorating HI-induced neuroinflammation is considered a promising strategy for treating WMI.

As resident phagocytic cells in the brain, activated microglia mediate neuroinflammation, which is crucial in the pathogenesis of WMI^8,9. A key component of microglia-mediated inflammation involves the engagement of Nod-like receptor nucleotide-binding domain leucine-rich repeat containing protein 3 (NLRP3), a Nod-like receptor family member¹⁰. By activating specific signaling pathways, microglia can differentiate into two specific phenotypes: the classical activation phenotype (M1: proinflammatory) or the alternative activation phenotype (M2: anti-inflammatory). The M1 state releases pro-inflammatory cytokines and increases the expression of inducible nitric oxide synthase (iNOS) and cluster of differentiation markers 86 and 16/32 (CD86, CD16/32). The M2 state is further subdivided into M2a, M2b, and M2c, which express anti-inflammatory cytokines as well as arginase-1 (Arg1), transforming growth factor β1 (TGFβ1), CD206, and chitinase-3-like-3 (Ym1)¹¹. M1 microglia amplify the inflammatory response and aggravate brain damage by releasing proinflammatory cytokines, whereas M2 microglia provide immune protection by secreting anti-inflammatory cytokines and trophic factors. Both M1 and M2 phenotypic microglia are highly plastic and can dynamically switch in response to environmental brain variables¹². The study found that modulation of microglia M1/M2 polarization ameliorates GMH-induced brain. MCC950 modulates the imbalance of microglia M1/M2 polarization thereby exerting a protective role in inflammatory damage induced by HI-induced cerebral WMI, which may be related to the inhibition of NLRP3 inflammasome¹³. In addition, previous reports have found that inhibiting the release of pro-inflammatory mediators and regulating the differentiation of microglia towards M2 direction after neonatal HI is crucial for the recovery of neurological function. Therefore, focusing on regulating the phenotype transition of microglia can be meaningful for seeking new therapeutic targets for WMI in the brain^13,14.

Cell transplantation presents distinctive benefits for addressing ischemic damage and has emerged as a central focus in interdisciplinary studies¹⁵. Human umbilical cord mesenchymal stem cells (HUC-MSCs) have robust regenerative capacity, reduced immunogenic profile, diminished risk of harboring viruses, and lower likelihood of age-induced mutations, which collectively make them the favored option for transplantation purposes^15,16. Accumulating evidence indicates that HUC-MSCs exhibit multifaceted therapeutic effects, particularly in modulating inflammation, reducing oxidative stress, regulating immune responses, and promoting vascularization^17–20. Notably, HUC-MSC administration has been reported to enhance long-term functional recovery in neonatal rats by reversing the pathological changes in glial cells after WMI²¹. Moreover, existing evidence suggests that MSCs attenuate microglial activation and microglia-derived neuroinflammation by the TLR4-dependent signaling mechanisms²², and have also confirmed that the TLR4 receptor plays a dominant role in microglia-mediated neuroinflammation¹⁰. For example, Zhang et al. demonstrated that the TLR4-specific agonist CRX-527 induced microglial activation and exacerbated cerebral ischemia-reperfusion injury²³. Experimental evidence indicates that suppressing the TLR4/NLRP3 signaling axis attenuates neuroinflammation and preserves blood-brain barrier integrity in a rat model of ischemic stroke²⁴. Although HUC-MSCs have been found exert neuroprotective effects, particularly regarding the brain, the exact mechanism by which HUC-MSCs provide neuroprotection against HI-induced WMI remains unclear, and whether HUC-MSCs antagonize TLR4 and regulate microglial polarization also requires further investigation.

The current research aimed to elucidate how HUC-MSCs transplantation exerts neuroprotective effects against HI-induced WMI, investigate the specific mechanisms underlying this protection, and determine whether these effects involve the antagonism of TLR4 receptors. We hypothesized that HUC-MSCs provide neuroprotection against HI-induced WMI in neonatal rats by modulating microglial phenotypic polarization through the TLR4/NLRP3 signaling pathway.

Results

Surface markers and differentiation identification of HUC-MSCs

According to the International Society for Cell Therapy (ISCT) standards, MSCs express specific surface markers and demonstrate multidirectional differentiation potential, enabling differentiation into osteoblasts, adipocytes, and chondrocytes. Through our experiment, we verified that HUC-MSCs fulfilled these essential characteristic. HUC-MSCs were characterized using flow cytometry to analyze the expression of surface markers in cultured cells. The results revealed a protein expression profile consistent with MSCs. As depicted in Fig. 1A, most cells exhibited high levels of CD90, CD105, CD73, and CD166, while expressing CD11b, CD14, HLA-DR, and CD184 at low levels. Furthermore, we successfully induced the differentiation of HUC-MSCs into osteoblasts (Fig. 1B-D), adipocytes (Fig. 1E-G), and chondrocytes (Fig. 1H-J), as confirmed by lineage-specific gene expression and staining. These results indicate that HUC-MSCs meet the MSC criteria defined by the ISCT.

Fig. 1.

Identification of HUC-MSCs. (A) HUC-MSC markers were analyzed using flow cytometry. (B) Osteogenic differentiation of HUC-MSCs assessed by Alizarin Red S staining. Magnification: 200×. (C) Osteogenic differentiation of HUC-MSCs assessed by detecting the expression of the characteristic gene Runx2 using Western blot. Uncropped gel/spot images are shown in supplementary information file Fig. 1C. (D) Quantitative analysis of Runx2 expression in HUC-MSC group and differentiation group. (E) Adipogenic differentiation of HUC-MSCs assessed by Oil Red O staining. Magnification: 200×. (F) Adipogenic differentiation of HUC-MSCs assessed by detecting the characteristic gene PPAR-γ in HUC-MSC and differentiation group. Uncropped gel/spot images are shown in supplementary information file Fig. 1F. (G) Quantitative analysis of PPAR-γ expression in HUC-MSC and differentiation group. (H) Chondrogenic differentiation of HUC-MSCs assessed by alcian blue staining. Magnification: 200×. (I) Chondrogenic differentiation of HUC-MSCs assessed by detecting the characteristic gene collagen II in HUC-MSC and differentiation group. Uncropped gel/spot images are shown in supplementary information file Fig. 1I. (J) Quantitative analysis of collagen II expression in HUC-MSCs and in the differentiation group. ***P < 0.01, compared with the HUC-MSC group. All values were the mean ± SEM, N = 6.

HUC-MSCs treatment alleviated WMI in neonatal rats

To evalute the interventional effects of HUC-MSCs on WMI in neonatal rats, a HI model was established. HUC-MSCs were locally injected in 8 h post-modeling, followed by continuous monitoring (Fig. 2A). The TTC staining technique was utilized to analyze the impact of HUC-MSCs on HI-induced brain injury. The brain injury area was significantly greater in HI versus sham rats at 48 h post-HI injury (P < 0.05), while HUC-MSCs treatment substantially reduced this pathology (P < 0.05; Fig. 2B, C). HE staining results at 7-, 14-, and 21-days post-treatment further corroborated above findings (Fig. 2D). In the sham group, the white matter structure remained intact, with normal cellular morphology and orderly arranged nerve fibers. In the HI group, significant white matter disruption was observed, characterized by increased vacuolization around cells, sparse cribriform nerve fibers, and glial scar formation at different time points. Contrastingly, the pathological damage of the HUC-MSCs group was alleviated, the cell morphology was well preserved, and the nerve fibers were relatively orderly, indicating the protective effect of HUC-MSCs against WMI in neonatal rats.

Fig. 2.

HUC-MSCs reduce the cerebral infarction area after HI in neonatal rats with WMI. (A) Construction of a WMI model in neonatal rats was induced by HI injury, with specific sampling times and locations. (B) TTC staining profiles of the sham group, HI group, and HUC-MSC group. (C) Measurement and analysis of the cerebral infarction area, achieved by comparing the infarct area with the total area. (D) HE staining results of each group at different time points after HI. Magnification: 200×. **P < 0.05, compared with the sham group, P < 0.05, compared with the HI group. All values were the mean ± SEM, N = 6.

Oligodendrocyte precursor cells encounter obstacles during their differentiation into mature oligodendrocytes, impairing effective myelin repair in WMI^3–5. Here, we assessed the cytoprotective actions of HUC-MSCs on oligodendrocytes by evaluating the expression of myelin basic protein (MBP) and proteolipid protein (PLP). IHC results showed that MBP and PLP were predominantly expressed in the corpus callosum region (Fig. 3A, B). The expression of PLP and MBP decreased significantly at 7, 14, and 21 days post-HI compared to that in the sham group (both P < 0.01), but was partially restored after HUC-MSC treatment (both P < 0.01; Fig. 3C, D). Transmission electron microscopy revealed a decrease in the number and thickness of myelinated axons after HI, indicating hindered myelination, which significantly improved at 7, 14 and 21 days post-HUC-MSC treatment (Fig. 3E). Western blot experiments indicated lower expression levels of PLP and MBP after HI, compared to those in the sham group (both P < 0.05), with their expression being upregulated from 7, 14, and 21 days post-HUC-MSC treatment (P < 0.05; Fig. 3F–H). These data suggest that HUC-MSCs may promote oligodendrocyte maturation and thereby contribute to WMI repair.

Fig. 3.

HUC-MSCs treatment alleviated WMI in neonatal rats. IHC staining of PLP (A) and MBP (B). Magnification: 400×. Analysis of PLP (C) and MBP (D) mean optical density. (E) Electron microscopy images in the sham group, HI group, and HUC-MSC group. Magnification: 8000×. (F) Western blot detection of PLP and MBP. Uncropped gel/spot images are shown in supplementary information file Fig. 3F. Analyses of relative PLP (G) and MBP (H), with β-actin used for normalization. *P > 0.05, compared with the sham group, P > 0.05, compared with the HI group; **P < 0.05, compared with the sham group, P < 0.05, compared with the HI group; ***P < 0.01, compared with the sham group, P < 0.01, compared with the HI group. All values are presented as mean ± SEM, N = 6.

To assess axonal integrity, we performed NF-L IHC staining at 14 days post-HI. The results showed that axonal density was significantly reduced compared to that in the sham group (P < 0.01). However, HUC-MSC treatment significantly attenuated this axonal loss compared to the HI group (P < 0.01) (Supplementary Fig. S1A-B). These data suggest that HUC-MSCs may exert both myelination-promoting and axonal protective effects on the WMI.

HUC-MSCs improve spatial cognition in neonatal rats with WMI

The MWM paradigm (Fig. 4A) was utilized to examine how HUC-MSCs influence spatial learning and memory in WMI neonatal rats. During the training phase, the escape latency decreased progressively across the rat groups. On day 1 of training, no statistical difference in escape latency was observed among the groups (all P > 0.05). However, from days 2 to 5, the HI group showed significantly longer escape latency compared to the sham group (all P < 0.01). In contrast, the results demonstrated a marked reduction in escape latency in the HUC-MSC group compared to the HI group (all P < 0.01; Fig. 4B). The swimming trajectories during the probe trial phase, compared to those on training day 1, are shown in Fig. 4C. The HI group showed significantly fewer platform crossings and shorter dwell time in the target quadrant compared to the sham group (both P < 0.01; Fig. 4D, E). Notably, the HUC-MSC treatment significantly increased platform traversal ability and prolonged dwell time in HI-injured rats. No statistically significant differences in locomotor distance were observed across the three groups (all P > 0.05, Fig. 4F), suggesting that the results were not influenced by variations in physical activity. In summary, HUC-MSCs improved the spatial cognitive abilities of neonatal rats with WMI.

Fig. 4.

HUC-MSCs improve spatial cognition in neonatal rats with WMI. (A) Schematic diagram of the Morris water maze installation. Abbreviations: (n) north; (s) south; (w) west; (e) east. (B) Line chart of escape latency among the sham group, HI group, and HUC-MSC group at days 28–32 post-treatment. (C) Pseudo-color images of the walking trajectory of newborn rats were obtained on the first day of training and during the detection experiment. (D) Number (times) of platform crossing during the probe experiment. (E) Time (s) spent in the target quadrant during the probe experiment. (F) Moving distance (cm) during the probe experiment. *P > 0.05, compared with the sham group, P > 0.05, compared with the HI group; **P < 0.05, compared with the sham group; ***P < 0.01, compared with the sham group, P < 0.01, compared with the HI group. All values are presented as mean ± SEM, N = 6.

HUC-MSCs suppress NLRP3 inflammasome activation in neonatal rats with WMI

NLRP3 inflammasome triggering constitutes a pivotal mechanism underlying both the initiation and progression of neural pathologies. Therefore, we assessed the localization and expression of NLRP3 inflammasomes and investigated how HUC-MSCs influence them in the brains of neonatal rats with WMI. Immunofluorescence analysis revealed that NLRP3 expression was significantly upregulated at 7, 14, and 21 days following HI versus sham rats (P < 0.01). However, HUC-MSC administration markedly reduced NLRP3 expression at the 14 days post-HI (P < 0.01), while no significant differences were observed at 7 or 21 days post-HI (P > 0.05, Fig. 5A, B). mIHC dual-localization analysis revealed that at 14 days post-HI injury, the HI group exhibited a markedly higher count of Iba1/NLRP3 co-positive cells compared to the sham group (P < 0.01). Following HUC-MSC treatment, this co-localization was significantly reduced relative to the HI group (P < 0.01, Fig. 5C, D), indicating that microglial activation is associated with NLRP3 inflammasome upregulation.

Fig. 5.

HUC-MSCs inhibit the expression of NLRP3 inflammasomes in neonatal rats with WMI. (A) Fluorescently localized expression of NLRP3 in the sham group, HI group, and HUC-MSC group at different time points. Magnification: 400×. (B) Analysis of the mean fluorescence intensity of NLRP3. (C) Iba1 (red) and NLRP3 (green) fluorescence co-localization staining in each group 14 days after HI. Magnification: 400×. (D) Analysis of co-localization of Iba1 with NLRP3. (E) Western blot bands of NLRP3, pro-caspase-1, cleavage caspase-1, and IL-1β. Uncropped gel/spot images are shown in supplementary information file Fig. 5E. Analyses of relative NLRP3 (F), pro-caspase-1 (G), cleavage caspase-1 (H), and IL-1β (I) expression, with β-actin used for normalization. *P > 0.05, compared with the sham group, P > 0.05, compared with the HI group; **P < 0.05, compared with the sham group, P < 0.05, compared with the HI group; ***P < 0.01, compared with the sham group, P < 0.01, compared with the HI group. All values are presented as mean ± SEM, N = 6.

Western blot analysis showed significantly higher expression levels of NLRP3-associated proteins at various time points following HI compared to the sham group (P < 0.01 or P < 0.05). In contrast, HUC-MSC treatment significantly reduced the expression of NLRP3 and IL-1β (P < 0.01 and P < 0.05, respectively), as well as pro-caspase-1 and caspase-1 (P < 0.05; Fig. 5E-I), relative to the HI group. These findings demonstrate that NLRP3 inflammasome activation was strongly induced in microglia after HI, and that HUC-MSCs effectively suppressed this activation in neonatal rats with WMI.

HUC-MSCs inhibit microglial activation and shift microglia polarization towards the M2 phenotype in the brains of neonatal rats with WMI

To comprehensively elucidate the regulatory effects of HUC-MSCs on microglial protein expression and cellular morphology, we employed both IHC staining and Western blot analysis. IHC staining showed that the microglial marker Iba1 was mainly present in the corpus callosum and subventricular zone (Fig. 6A). Following HI, the number of Iba1 positive-cells was significantly elevated at 7-, 14-, and 21-days relative to the sham group, but this increase was attenuated by HUC-MSC administration (Fig. 6B, C). Western blot analysis supported the above results (Fig. 6D, E). ImageJ and Fiji software were used to assess the morphology of microglia in the subventricular zone (Fig. 6F, left) and to conduct cytoskeletal analyses (Fig. 6F, right). The sham group exhibited resting microglia with small cell bodies and slender branched processes. Following HI, microglia were activated and underwent significant morphological remodeling, as evidenced by significant increases in both endpoint counts and branch lengths (all P < 0.01). Compared to the HI group, the microglial endpoint counts and branch lengths in the HUC-MSC group showed no significant change (all P > 0.05; Fig. 6G, H). These results indicate that while HUC-MSCs significantly inhibit microglia proliferation, they do not reverse the activated morphological changes induced by HI.

Fig. 6.

HUC-MSCs inhibit microglial activation in neonatal rats with WMI. (A) IHC staining of Iba1 in the corpus callosum region and subventricular zone. Magnification: 400×. Relative positive area analysis of Iba1 in the corpus callosum region (B) and subventricular zone (C). (D) Western blot bands of Iba1. Cropped gel/spot images are shown in supplementary information file Fig. 6D. (E) Analysis of relative Iba1 expression, with β-actin used for normalization. (F) Morphological reconstruction of microglia in the subventricular zone (left) and microglial skeletal diagram reconstruction process (right). (G) Endpoint analysis of the number of microglia. (H) Analysis of microglia process length. *P > 0.05, compared with the sham group, P > 0.05, compared with the HI group; **P < 0.05, compared with the sham group, P < 0.05, compared with the HI group; ***P < 0.01, compared with the sham group, P < 0.01, compared with the HI group. All values are presented as mean ± SEM, N = 6.

The transition between pro-inflammatory M1 and anti-inflammatory M2 microglial phenotypes contributes to the development of central nervous system disorders. We employed mIHC staining to assess the co-localization of Iba1 with CD86/iNOS (M1 markers) and CD206/Arg1 (M2 markers) in the subventricular zone at 14 days post-treatment. Following HI injury, the proportion of Iba1 and the M1 markers CD86 and iNOS co-localization was significantly increased compared to the sham group (all P < 0.01). Similarly, a marked increase was also observed in the co-localization of Iba1 with the M2 markers CD206 and Arg1 (all P < 0.01). Notably, HUC-MSCs treatment attenuated the HI-induced increase in Iba1/CD86 and Iba1/iNOS microglia (all P < 0.01; Fig. 7A-C), while further enhancing the co-localization of Iba1 with CD206 and Arg1 (all P < 0.01; Fig. 7D-F).

Fig. 7.

HUC-MSCs modulate the phenotypic polarization of M1/M2 microglia in the brains of neonatal rats with WMI. (A) mIHC staining showed co-localization of Iba1 (green) with CD86 (red) or iNOS (red) in the subventricular zone. Magnification: 400×. Analysis of co-localization and expression of Iba1 with CD86 (B) and iNOS (C) in the sham group, HI group, and HUC-MSC group, respectively. (D) mIHC staining showed co-localization of Iba1 (red) with CD206 (green) or of Iba1 (green) with Arg1 (red) in the subventricular zone. Magnification: 400×. Analysis of co-localization and expression of Iba1 with CD206 (E), and Arg1 (F) in the sham group, HI group, and HUC-MSC group, respectively. (G) Western blot bands for CD86, iNOS, CD206, and Arg1. Uncropped gel/spot images are shown in supplementary information file Fig. 7G. Analyses of relative CD86 (H), iNOS (I), CD206 (J), and Arg1 (K) expression, with β-actin used for normalization. *P > 0.05, compared with the sham group, P > 0.05, compared with the HI group; **P < 0.05, compared with the sham group, P < 0.05, compared with the HI group; ***P < 0.01, compared with the sham group, P < 0.01, compared with the HI group. All values are presented as mean ± SEM, N = 6.

To investigate the temporal dynamics of HUC-MSC-mediated modulation of microglial polarization, we performed Western blot analysis at multiple time points following HI injury. Quantitative analysis demonstrated a marked temporal progression in polarization marker expression. Compared to the sham group, M1 phenotype markers (CD86 and iNOS) showed progressive upregulation at 7-, 14-, and 21-days post-HI (all P < 0.01). Interestingly, M2 markers (CD206 and Arg1) were also significantly elevated during this period (P < 0.05). Notably, HUC-MSCs treatment induced a significant phenotypic shift, with substantial downregulation of M1 markers (P < 0.05; Fig. 7G, H–I) accompanied by concurrent upregulation of M2 markers (all P < 0.05; Fig. 7G, J-K). Taken together, these results suggest that HUC-MSCs may exert a dual regulatory effect, effectively suppressing M1 microglial polarization while promoting M2 phenotype conversion.

HUC-MSCs inhibit NLRP3 Inflammasome activation and regulate the polarization of microglia towards the M1 and M2 phenotypes by antagonizing TLR4 receptors

To investigate whether TLR4 is a critical target through which HUC-MSCs exert their effects, we administered the TLR4-specific agonist LPS to HI-injured rats that concurrently received HUC-MSC transplantation (denoted as the HUC-MSC + LPS group). mIHC analysis revealed a marked increase in the number of Iba1/TLR4 co-labeled cells at 14 days post-HI compared to the sham group (P < 0.01). HUC-MSC treatment significantly attenuated this HI-induced co-localization of Iba1 and TLR4 (P < 0.01). Furthermore, the suppressive effect of HUC-MSCs was more pronounced than that observed in the HUC-MSC + LPS group (all P < 0.01, Fig. 8A–B). Western Blot results showed that, compared with the Sham group, TLR4 expression was increased in the HI group (P < 0.01). HUC-MSC treatment significantly inhibited TLR4 expression in the HI group (P < 0.01), and this inhibitory effect was reversed in the HUC-MSC + LPS group (P < 0.01; Fig. 8C-D). Notably, the HUC-MSC + LPS group exhibited significantly elevated CD86 expression compared to the HUC-MSC group (P < 0.01; Fig. 8C, E), whereas CD206 levels were markedly reduced (P < 0.01; Fig. 8C, F). Importantly, the HUC-MSC + LPS group abolished the suppressive effects of HUC-MSCs on the NLRP3 inflammasome and its associated proteins, such as cleaved caspase-1 and IL-1β (both P < 0.01; Fig. 8G–J). Based on the above data, HUC-MSCs appeared to inhibit NLRP3 inflammasome activation and modulate M1/M2 microglial polarization by antagonizing TLR4 receptors.

Fig. 8.

HUC-MSCs inhibit NLRP3 inflammasome activation in neonatal rats with WMI and regulate M1/M2 microglial polarization by antagonizing TLR4 receptors. (A) Fluorescence co-localization staining of Iba1 and TLR4 at 14 days after HI. Magnification: 400×. (B) Co-localization analysis of Iba1 and TLR4. (C) Western blot bands for TLR4, CD86, and CD206. Uncropped gel/spot images are shown in supplementary information file Fig. 8C. Analyses of relative TLR4 (D), CD86 (E), and CD206 (F) expression, with β-actin used for normalization. (G) Western blot bands for NLRP3, cleavage caspase 1, and IL-1β. Uncropped gel/spot images are shown in supplementary information file Fig. 8G. Analyses of relative NLRP3 (H), cleavage caspase-1 (I), and IL-1β (J) expression, with β-actin used for normalization. Levels of M1-related cytokines, IL-1β (K), TNF-α (L), M2-related cytokines, IL-10 (M), and TGF-β (N). *P > 0.05, compared with the sham group, P > 0.05, compared the with HI group; **P < 0.05, compared with the sham group, P < 0.05, compared with the HI group, ^##P < 0.05, compared with the HUC-MSC group; ***P < 0.01, compared with the sham group, P < 0.01, compared with the HI group, ^###P < 0.01, compared with the HUC-MSC group. All values are presented as mean ± SEM, N = 6.

To substantiate these findings, cytokine levels were quantitatively analyzed by ELISA. HUC-MSC treatment significantly attenuated M1-associated pro-inflammatory cytokines (IL-1β and TNF-α) compared to the HI group (P < 0.05 or P < 0.01, respectively; Fig. 8K, L), while concurrently enhancing M2-associated anti-inflammatory cytokines (IL-10 and TGF-β) (P < 0.05 or P < 0.01; Fig. 8M, N). Moreover, the therapeutic effect of HUC-MSCs was offset by the HUC-MSC + LPS group (P < 0.05 or P < 0.01, respectively). Collectively, these data suggest that HUC-MSCs regulate microglial polarization by targeting TLR4.

Discussion

Premature infants often experience the consequences of perinatal brain damage, characterized by impaired oligodendrocytes maturation and myelination. This condition is linked to an increased risk of neurodevelopmental disorders, and there are currently no effective treatments available. Stem cells have been widely studied in various fields because of its paracrine effects and multipotent differentiation potential. Among them, HUC-MSCs have been widely utilized in the clinical trial stage across various disease models^25–27. Previous studies have found that HUC-MSC transplantation can exert neuroprotective effects and has the potential to reverse functional brain damage^25,28,29. Specifically, it was able to achieve neurorestorative effects on WMI in neonatal rats by promoting oligodendrocyte maturation and reducing neuronal death³⁰. Extracellular vesicles of HUC-MSC can significantly reduce cerebral infarction, microglia activation and pyroptosis³¹. This study aimed to elucidate the protective effects and underlying mechanism of action of HUC-MSCs against WMI in neonatal rats. The results showed that HUC-MSCs promoted the maturation of oligodendrocytes and facilitated myelination after WMI in neonatal rats and improved spatial cognitive ability. HUC-MSCs inhibited the activation of microglia after WMI in neonatal rats and regulated the polarization imbalance of M1/M2 microglia. Further, HUC-MSCs inhibited NLRP3 inflammasome activation and regulated M1/M2 microglial polarization by antagonizing TLR4 receptors.

Stem cell therapy is considered a promising treatment method because of its regenerative and immunoregulatory abilities. We aimed to determine whether HUC-MSCs offer therapeutic benefits for WMI in neonatal rats. The selection of the optimal transplant dose is the primary task in the early stage of this study. Based on prior studies^32–34 and previous research by our group, we transplanted HUC-MSCs at a dose of 2 × 10⁵ directly into the lateral ventricles. MBP and PLP are two major myelin proteins in the central nervous system³⁵. HUC-MSC transplantation has been found to significantly reduce the loss of MBP and induce myelin regeneration after HI^16,21 Our findings align with the results of prior reports, demonstrating that HUC-MSCs remarkably reduce the area of cerebral infarction, increase the level of mature oligodendrocyte markers PLP and MBP, and facilitate myelin formation. In addition to these effects, our data indicate that HUC-MSCs also exert a direct protective effect an axonal integrity, preventing their loss after injury. However, this study has certain limitations, as we cannot determine whether the protective effect of HUC-MSCs against WMI is due to their multipotent differentiation potential or their paracrine function. A large amount of research^36–38 has focused on the efficacy of extracellular vesicles secreted by HUC-MSCs; however, research on the stem cells themselves has been scarce, which may be attributable to the difficulty in controlling the amount of stem cell transplantation.

In our study, transplanting a certain amount of HUC-MSCs into the lateral ventricle after brain injury not only improved the pathological changes of brain tissue, but also had a significant impact on its behavior. Several studies have shown that transplantation of MSCs restored the use of lateral limbs in HI-induced rats, shortened the escape latency of rats seeking the submerged platform, and increased their number of entries into the platform quadrant and their dwell time during the probe test phase, suggesting an improvement in behavioral skills in brain-injured rats^16,39,40. Similarly, our MWM experiment revealed that HUC-MSCs could enhance spatial cognitive ability in neonatal rats following WMI. This was evidenced by a significantly reduced escape latency, an increased number of platform crossings, and more time spent in the target quadrant, which are consistent with the results reported in the literature. HUC-MSCs demonstrated notable efficacy in the MWM test, which is likely related to oligodendrocyte maturation and myelin regeneration. Studies^16,39,40 have demonstrated that HUC-MSCs offer neuroprotective effects against WMI in premature infants. While the specific mechanism of action remains unclear, it is likely caused by anti-inflammatory or immune regulation.

Microglia are resident phagocytic cells in the brain that are crucial for maintaining normal brain function. Their morphology and function are closely related⁴¹. The activation status of microglia is closely linked to the degree of WMI, with morphological changes in these cells reflecting this relationship⁴². Classical studies have firmly established that in many acute neuroinflammatory and degenerative contexts, activated microglia adopt an amoeboid morphology with retracted processes, aligning with a phagocytic, pro-inflammatory phenotype^43,44. Our findings, however, alongside other focused investigations in neonatal HI injury models⁴⁵, reveal a more complex and context-dependent picture. In our neonatal rat WMI model, HI activation drove microglia toward a “hyper-ramified” or reactive morphology, characterized by somatic hypertrophy concomitant with increased process branching and length rather than wholesale retraction. This discrepancy is not contradictory but rather highlights the remarkable plasticity and disease-specific adaptation of microglia. It is noteworthy that while HUC-MSC transplantation successfully prevented the development of microglial proliferation when administered early after HI, it did not reverse the established morphology.

Under inflammatory conditions, microglia not only display morphological differences as described above, but also demonstrate significant variations in cell phenotypes at the microscopic level. It is well-seen that microglia exhibit remarkable functional adaptability, allowing them to transition between two distinct states: M1 and M2⁴⁶. Modulation of microglia phenotype and promotion of cellular transition into M2 direction may alleviate WMI and improve neurological prognosis^47–49. Increasing evidence suggests that MSC and their exosomes can shift microglia phenotype from a pro-inflammatory M1-like state to an anti-inflammatory M2-like phenotype, with underlying mechanisms involving the regulation of receptors, gene expression and so on^50–52. Our research results confirm that HUC-MSCs promote the shift of microglia phenotype. This is achieved by downregulating the expression of the M1 microglial markers CD86 and iNOS, reducing the levels of pro-inflammatory cytokines IL-1β and TNF-α, and simultaneously upregulating the expression of the M2 microglial markers CD206 and Arg1, while increasing the levels of anti-inflammatory cytokines IL-10 and TGF-β. Previous studies have identified pro-inflammatory M1 microglia as key contributors to WMI in the brain, primarily through the production of a range of multiple inflammatory mediators. However, anti-inflammatory M2 microglia modulate WMI and participate in myelin regeneration by producing growth factors⁵³. Therefore, interventions that promote microglia polarization toward M2 may offer significant therapeutic potential for treating WMI. Therefore, our study provides experimental evidence that the therapeutic potential of HUC-MSCs against WMI is related to their ability to regulate microglial polarization.

Pyroptosis plays a pivotal role in brain damage, as it results in the release of various inflammatory factors, triggering an inflammatory cascade that exacerbates ischemic damage⁵⁴. In this study, NLRP3 was localized and expressed in the microglia after HI, and its related protein expression was significantly reduced after HUC-MSC treatment. Interestingly, HUC-MSCs showed a remarkable therapeutic effect at 14 days after HI. The results of protein and fluorescence experiments are not completely consistent, which may be due to the fact that NLRP3 is not only activated in one cell type, but HUC-MSC has a certain time effect on the inhibition of NLRP3. A growing number of evidence indicate that NLRP3 activation is a key factor in microglial pyroptosis and the progression of ischemic brain damage^54,55. Of note, NLRP3 activation in astrocytes is also involved in the development of neuroinflammation⁵⁶. Previous studies have found that pramipexole, a dopamine receptor agonist, exerts anti-neuroinflammatory activity by antagonizing NLRP3 activation in astrocytes in a mouse model of Parkinson’s disease⁵⁷. In addition, targeting NLRP3 inflammasome in astrocytes can alleviate blood-brain barrier breakdown in a mouse model of ischemic stroke⁵⁸. Therefore, we speculated that NLRP3 activation may be involved in microglial phenotype polarization after WMI in neonatal rats.

Microglia express TLRs, which are crucial for the development of neuroinflammation. Our study showed that TLR4 receptors were highly expressed after HI and that HUC-MSCs downregulated its expression level, suggesting that the role of HUC-MSCs is likely related to TLR4 receptors. Previous studies have also shown that microglia highly expressed TLR4 and its co-receptor CD14⁵⁹ and that TLR4 deficiency might lead to the promotion of the M2 microglial phenotype by inhibiting autophagy, upregulating CD206 and Arg-1, and reducing white-matter demyelination during ischemic injury⁶⁰. Research on WMI had found that TLR4 receptor deficiency caused a shift of microglia towards an anti-inflammatory phenotype, providing protection against WMI^59,60. We speculated that the regulation of M1/M2 microglial polarization by HUC-MSCs might be highly correlated with the inhibition of TLR4. The experimental findings of this study demonstrate that co-administration of the TLR4 agonist LPS with HUC-MSCs reverses their therapeutic benefits—restoring NLRP3 inflammasome activation, M1 microglial polarization, and the release of pro-inflammatory cytokines—which provides direct functional evidence that TLR4 is a critical downstream target through which HUC-MSCs exert their neuroprotective effects. These findings are in line with research in other areas^61–63 that have also confirmed the inhibitory effect of HUC-MSCs on TLR4. These results suggest that HUC-MSCs have a regulatory effect on both the activation and polarization of microglia following WMI in neonatal rats.

This study has several limitations. First, only male neonatal rats were included, which limits the generalizability of the findings across sexes. Second, although the TLR4 agonist experiment provided critical functional evidence, more direct evidence—such as TLR4 gene knockout—is required in future studies to further validate the mechanism. Finally, our analyses focused on specific time windows after injury and did not cover the full course of the pathology. These limitations provide important directions for future research.

In summary, our study confirms that HUC-MSCs inhibit NLRP3 inflammasome activation and regulate microglial polarization by antagonizing TLR4 receptors, thereby exerting a protective effect on WMI. These findings, as well as research on adult neurological disorders, highlight the potential of exogenous HUC-MSCs transplantation as a promising method for treating neonatal brain injury.

Methods

Animals and umbilical cord acquisition

Prior to the experiment, ethical approval was obtained from the Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University (No. IACUC-20200318-65; Urumqi, China) and strictly adhered to the 3R principles of animal welfare. At the same time, the collection of umbilical cord samples was approved by the Ethics Review Committee and conducted after obtaining informed consent from the family members, it was confirmed that all experiments were performed in accordance with the relevant guidelines and regulations. 80 three-day-old male Sprague-Dawley (SD) (purchased from the Animal Experiment Center of the First Affiliated Hospital of Xinjiang Medical University) rats with an average body weight of 9.56 ± 0.69 g were chosen for this study and fed in a particular pathogen-free environment within a controlled barrier system. The housing conditions, including temperature (21 ± 2℃), humidity (55 ± 5%), and lighting (12 h of alternating light and dark), were carefully regulated to ensure adaptation to their living environment. Based on the support of the national fund project, the Ethics Review Board thoroughly reviewed and approved the procedure for obtaining umbilical cords. With informed consent from the families, umbilical cords (length: approximately 4 cm) were collected from the department of obstetrics, the first affiliated hospital of xinjiang medical university and transported in sterile refrigerated containers maintained at 4 °C.

Extraction and identification of HUC-MSCs

The sterile sample collected at low temperature underwent washing, vascular removal, separation of Wharton’s jelly, and tissue homogenization to complete the pretreatment process and prepare the stem cell suspension. HUC-MSCs were subsequently cultured in Minimum Essential Medium-α supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin. Half-medium exchange was performed on days 5–6 of primary culture, followed by full-medium exchange. When 80–90% confluency was reached, the cells were passaged. HUC-MSCs at passage 3 were used in subsequent experiments unless otherwise specified.

Cells were collected from third-generation cultures, digested with trypsin, and divided into flow cytometry tubes. HLA-DR-PerCP-Cy5.5, CD90-FITC, CD105-PE, CD73-PE, CD166-PE, CD14-PerCP-Cy5.5, CD11b-FITC, and CD184-PerCP-Cy5.5 antibodies (Thermo Fisher Scientific, MA, USA) were added at 1 µL to every 100 µL of cell suspension and stained for 1 h. Thereafter, the samples were resuspended in 250 µL of cell fixative, detected by flow cytometry (CytoFLEX, Beckman Counlter) and analyzed by Flowjo V10 software.

Cells were collected from the third passage of subcultures according to the guidelines provided for the HUC-MSC differentiation medium. Kits from Pricella (Wuhan, China) were used for osteogenesis (PD-017), adipogenesis (PD-019), and chondrogenesis (PD-018) differentiation. Osteogenic differentiation: HUC-MSCs were cultured in osteogenic induction medium for 3 weeks before alizarin red staining and Runx2 protein detection was performed to confirm osteogenic differentiation. Adipogenic differentiation: HUC-MSCs were cultured in adipogenic induction medium ADP1 for 3 days and then replaced with dipogenic maintenance medium ADP2 for further culture. ADP1 and ADP2 media were alternated every 1–3 days for 3–5 cycles, followed by 2 weeks of continuous incubation in ADP2 maintenance medium. Oil Red O staining and PPARγ protein detection was used to confirm adipogenic differentiation. Chondrogenic differentiation: HUC-MSCs were cultured in complete chondrogenic induction medium, and the resulting cartilage particles were fixed with paraformaldehyde and embedded in paraffin. Toluidine blue staining and collagenⅡ protein detection was performed to confirm chondrogenic differentiation.

WMI rat model

The hypoxic-ischemic (HI)-induced neonatal white matter injury (WMI) model was prepared according to the Vannucci model^64,65. Briefly, SD rats were anesthetized by aerosol inhalation of 3–5% isoflurane for induction and maintained with 2% isoflurane and then fixed on anatomical plates. With an anatomical microscope, the left common carotid artery was exposed, and both the upper and lower ends were ligated. Subsequently, the vessel was transected between the ligatures. After suturing, the rats were placed next to their dam for 2 h to allow for recovery. The hypoxia device was placed in a water bath to control the temperature of the chamber at 21 ± 2℃, and the rats were placed in the chamber (8% oxygen and 92% nitrogen) for 2.5 h. The sham group only accepted arterial exposure and did not receive any other treatment. Rats was randomly divided into the sham (n = 24), HI (n = 24), HUC-MSC (n = 24), and HUC-MSC + LPS (n = 6) groups before modeling, and corresponding treatments were administered post-modeling. Following induction of anesthesia with 3% isoflurane inhalation, rats were maintained under a higher concentration of isoflurane (5%) for a minimum of 5 min until complete cessation of respiratory activity and cardiac arrest was confirmed. Rats were sampled (7-, 14-, and 21-days, post-modeling) and the behavioral changes were observed at specific time point (28-days post-modeling) with six rats per group at each time point.

HUC-MSCs transplantation and LPS administration

At the end of hypoxia, the pups were placed next to their mothers for 2 h and then, with the assistance of brain stereotaxic instrument, the HUC-MSC group was injected with a certain dose (2 × 10⁵ cells/2 µL) of HUC-MSCs into the lateral ventricle. The transplantation coordinates were determined by positioning 1.5 mm left and 1.5 mm posterior to the bregma. A microsyringe was carefully inserted until a depth of approximately 3 mm was reached, indicated by a breakthrough sensation. The stem cells were injected at a rate of 1 µL/min. Following injection, the syringe was kept in place intracranially for 3 min to prevent cell leakage. After transplantation, the pups were returned to their dams for recovery. In addition to the stem cell intervention, the HUC-MSC + LPS group received an injection of LPS (Sigma-Aldrich, Beijing, China). LPS was prepared at a dose of 1 mg/kg for SD rats. Each rat received intraperitoneal injection of 0.2 ml using a 0.2 ml insulin syringe. Injections were performed 2 h after hypoxia in a single dose. After administration, the pups were returned to their dams for continued care.

Triphenyltetrazolium chloride (TTC) staining

Rats at 48 h after modeling (sham, HI and HUC-MSC groups) were selected for TTC staining analysis. The brain tissues of rats in each group were first processed into tissue sections of uniform thickness and placed in 0.4% TTC staining solution for 15 min of incubation at 37 °C protected from light. After cleaning, they were fixed in 4% paraformaldehyde. The samples were placed on the dark plate, and the images were taken under the same visual angle (Magnification: 10x) by a camera (SonyA7M4,Japan) and the infarct region (red areas are normal brain tissue, while white areas are infarct areas) was calculated using ImageJ (NIH, Bethesda, MD). Infarct size was defined by calculating the sum of infarct areas in each coronal section of brain tissue, while the sum of the non-ischemic side was defined as the total area. The percentage of infarct region was calculated using the following formula: (infarct region / total region) × 100%.

Electron microscopy

At different time points (P7, P14, and P21) after HI, rats were anesthetized with isoflurane and fixed in the anatomical plates. After the heart was exposed, a perfusion needle was inserted into the left ventricle, and systemic circulation perfusion was performed with 0.9% physiological saline solution and 4% glutaraldehyde in turn. The left corpus callosum was carefully removed, processed into 1 mm³ tissue blocks, and placed in 4% glutaraldehyde solution. The tissues were fixed with 1% osmium acid in the dark for 1 h and rinsed three times with phosphate buffer solution. The tissues were dehydrated and then incubated overnight in a 1:1 mixture of acetone and embedding medium. Subsequently, the tissues were embedded, polymerized, and processed into 60-nm sections using an ultramicrotome (Leica, Germany). Finally, the tissues were sequentially stained with 2% uranium acetate and 2.6% lead citrate solutions, dried overnight, and imaged using a transmission electron microscope (HITACHI, Japan).

Behavioral test

The morris water maze (MWM) test was performed on days 28–33 after model establishment to assess spatial learning and memory ability. The overall appearance of the water maze device is a black cylindrical water tank (diameter 160 cm, height 30 cm), which is divided into four quadrants by a cross and defined as the northeast, northwest, southeast and southwest four regions. A small platform (diameter 12 cm, height 28.5 cm) was placed on the southwest side. The water tank is filled with water at a certain height above the platform (about 1.5 cm). A striking red five-pointed star marker was placed on one side of the water tank to facilitate the orientation of rats. A capture camera was placed directly above the dark box and connected to the computer behavioral system software to capture the parade trajectory of the rats. These tests included both an initial spatial training and a probe test. During the training phase, rats were placed in the water facing the middle of the wall in one of the four quadrants and allowed to swim and locate the hidden platform for a maximum of 60 s, and their escape latency was recorded. The next round of experiments was conducted at 30 min intervals, and each rat was subjected to 4 experiments per day, and the training was conducted for 5 consecutive days. After training, the platform was removed for detection experiments. Each rat was then placed in the location opposite to where the platform had been, and their movements and related parameters were recorded for 120 s. All data from the trial were automatically recorded using camera (Shanghai, China). The MWM test recording parameters included escape latency (s), number of platform traversals (times), target quadrant dwell time (s), and path length (cm).

Hematoxylin eosin (HE) staining

Tissue samples were fixed with 4% paraformaldehyde, dehydrated, and immersed in wax to prepare paraffin blocks. The slices were processed into 4 μm slices with a microtome, deparaffinized, and immersed in distilled water. Hematoxylin staining was used to observe the nuclear morphology and eosin staining was used to observe the cytoplasmic morphology. After staining, the slides were sealed after dehydration, and the pathological images of the target area were observed and collected under a light microscope (Nikon, Tokyo Ni-U).

Immunohistochemical (IHC) staining

After dewaxing, hydration, and antigen repair, brain tissue slices were incubated sequentially with 3% hydrogen peroxide and 10% goat serum, followed by overnight incubation with MBP antibody (1:1000; ab218011; Abcam), Neurofilament-Light Chain (NF-L; 1:800; 12998-1-AP; Proteintech), PLP antibody (1:2000; ab254363; Abcam), and Iba1 (1:10000; ab283319; Abcam) at 4 °C. Subsequently, the samples were rewarmed at 37 °C for 30 min and incubated with biotinylated goat anti-mouse/rabbit IgG antibody and streptavidin-horseradish peroxidase for an additional 30 min. The final samples were subjected to 3, 3′-diaminobenzidine chromogenic staining, followed by hematoxylin staining, dehydration, and sealing. The images of the target area were collected and observed by optical microscope (Nikon, Tokyo Ni-U), and regional statistical analysis was executed using ImageJ.

Multiplex immunohistochemistry (mIHC) staining

After deparaffinization and antigen retrieval, the tissues were sequentially incubated with endogenous peroxidase and serum and were subsequently combined with primary antibody at 4 ℃ overnight. Next, the tissues were combined with secondary antibody for 20 min and reacted with TSA-520 fluorescent dye, applied dropwise for 5 min. After repeating the above steps sequentially, the samples were labelled with TSA-570 and TSA-690 fluorescent dyes (AFIHC024; AiFang Biological, Changsha, China). The nuclei were then counterstained with DAPI and mounted. The primary antibodies used were Iba1 antibody (1:100, ab283319; Abcam), CD86 antibody (1:100, BM4121; Boster, Wuhan, China), CD206 antibody (1:200, ab64693; Abcam), iNOS antibody (1:200, 80517-1-RR; Proteintech, IL, USA), NLRP3 (1:100, WL02635; Wanleibo, China), TLR4 (1:100, WL00196; Wanleibo, China) and Arg-1 antibody (1:100, A01106; Boster, Wuhan, China). Fluorescence images were captured using a confocal fluorescence microscope (Olympus FV3000, Japan). Image analysis was performed on three randomly selected microscopic fields from each sample, ensuring consistent microscope settings and processing parameters across all images. The images were exported and analyzed with ImageJ (version 1.48) by two independent, blinded observers.

Western Blot

The periventricular tissue samples from rats in different groups and at different time points were lysed using RIPA lysis buffer (Biyuntian, China) containing 1% PMSF protease inhibitor (vazyme, China), shaken, and centrifuged at 12,000 g for 30 min at 4 °C. The protein concentration of supernatant was determined by BCA assay (Pierce, USA). Proteins were separated on 12.5% or 7.5% SDS-PAGE, transferred to PVDF membranes (Millipore, MA, USA). The membrane was sealed with a rapid blocking solution for 10 min and combined overnight for 14 h at 4 °C with the primary antibodies diluted to the appropriate concentrations. The antibody dilution ratios were as follows: 1:1000 for MBP (ab218011; Abcam) and PLP (ab9311; Abcam), 1:1000 for CD86 (A00220-4; Boster), 1:500 for CD206 (ab64693; Abcam), 1:500 for Iba1 (ab178846; Abcam), 1:1000 for iNOS (18985-1-AP; Proteintech), 1:1000 for Arg1 (16001-1-AP; Boster), 1:2000 for NLRP3 (WL02635; Wanleibio), 1:2000 for caspase-1 (WL03450; Wanleibio), 1:1000 for interleukin (IL)-1β (ab283818; Abcam), and 1:500 for TLR4 (38519 S; Cell Signaling Technology). After washing, the membranes were incubated with the corresponding horseradish peroxidase (HRP)-conjugated secondary antibodies (SA00001-2; Proteintech), at a dilution of 1:5000 for 2 h at room temperature. Following washes, samples were incubated with ECL chemiluminescent solution and exposed in a dark room. The grayscale values of the target bands were determined using imageJ software (version 1.48) and calibrated with β-actin.

Enzyme-linked immunosorbent assay (ELISA)

IL-1β, TNF-α, IL-10, and TGF-β levels in the brain tissues were measured using ELISA kits according to the manufacturer’s instruction (Boster, Wuhan, China).

Statistical analysis

Data were organized and analyzed using SPSS version 22.0, and visual graphs were created using GraphPad Prism version 8.0. Significance was determined using two-sided Student’s t-test or one-way ANOVA followed by Dunnett’s multiple comparison test. All data were presented as mean ± standard deviation, and P < 0.05 was considered to be statistically significant. N represents the number of samples used in the experiments.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

CW and QQX extracted and identified HUC-MSCs. CW performed the experiments. YPZ and SJZ participated in this study. CW and SJZ contributed to the data acquisition and analysis. All the authors have read and approved the final version of the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (Project Name: pre-OL Transformation and Mechanism of HUC-MSC Transplantation in Neonatal Rats with WMI, Based on Single-cell Sequencing Technology; project Number: 82060288).

Data availability

Data Availability Statement: The data supporting the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics declaration

All experiments were approved by the Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University (No. IACUC-20200318-65; Urumqi, China). All methods were performed in accordance with the National Standard for Laboratory Animal Care (GB 14925 − 2020) and relevant regulations. This study is reported in accordance with ARRIVE guidelines (https://arriveguidelines.org).