The miR-200b-3p/ERG/PTHrP Axis Mediates the Inhibitory Effect of Ethanol on the Differentiation of Rat Fetal Chondrocytes into Articular Cartilage


 Background

This study aims to further explore cartilage development in prenatal ethanol exposure (PEE) offspring at different times to explore the specific time points and mechanism of ethanol-induced fetal cartilage dysplasia.
Methods

On gestational day (GD)14, GD17, and GD20, PEE fetal cartilage was evaluated by morphological analysis. RT-qPCR, immunohistochemistry, and immunofluorescence were used to detect the expression of cartilage marker genes and their regulatory factors. Bone marrow mesenchymal stem cells (BMSCs) were used to explore the effect of ethanol on the differentiation of chondrocytes. Additionally, we used inhibitors, overexpression plasmids and a luciferase reporter assay on GD17 chondrocytes to verify the mechanism.
Findings:

PEE significantly reduced cartilage matrix content and the expression of marker genes on GD17 and GD20 but had no effect on GD14. The inhibition of chondrogenic differentiation by PEE mainly occurred on GD14-17. Furthermore, the expression of miR-200b-3p was increased, while that of ERG and PTHrP was markedly reduced in PEE fetal cartilage. In vitro, ethanol (30–120 mM) inhibited the differentiation of BMSCs into chondrocytes in a concentration-dependent manner, accompanied by strong expression of miR-200b-3p and low expression of ERG and PTHrP. Moreover, PTHLH and ERG overexpressed, as well as a miR-200b-3p inhibitor reversed the inhibitory effect of ethanol on the differentiation of fetal chondrocytes. Furthermore, miR-200b-3p could target and negatively regulate ERG.
Interpretation:

PEE can significantly inhibit the development of articular cartilage, especially during articular cartilage formation. The mechanism is related to the decreased differentiation of fetal cartilage into articular cartilage mediated by the miR-200b-3p/ERG/PTHrP axis.


Introduction
Osteoarthritis (OA) is an age-related disease characterized by the progressive loss of articular cartilage, concomitant sclerotic changes in the subchondral bone, and the formation of osteophytes (Lawrence et al. 2008). The etiology of OA is complex, and involves age-related, genetic, developmental, and biomechanical factors (Shen et al. 2019). Epidemiological data suggest that people with low birth weight are susceptible to OA in adulthood, and a fetal origin of OA has been speculated(Aigner and Richter 2012; Jordan et al. 2005;Sayer et al. 2003). In a series of studies our team proposed for the rst time and con rmed that OA has fetal origin (Chen et  One study has also shown that the partial regulatory effect of ERG on chondrocytes is mediated by PTHrP (Okabe et al. 2011). This nding shows that ERG can regulate the differentiation and developmental direction of embryonic cartilage progenitor cells through PTHrP.
MicroRNAs (miRNAs) are small single-stranded noncoding RNAs that play pivotal roles by regulating functional genes in the process of growth and development (Martinez and

Animals and treatment
Speci c pathogen-free (SPF) Wistar rats (females weighing 210-250 g, and males weighing 270-320 g) were purchased from the Experimental Center of Hubei Medical Scienti c Academy (No. 2015-2018, certi cation number: 42000600032526, Wuhan, China), and animal handling and experimental procedures were carried out with approval from the Institute of Health Sciences Institutional Animal Care and Use Committee. All rats were fed adaptively for 1 week, and 2 females were mated with 1 male every night. The day when sperm was found on a vaginal smear was considered GD 0. Then, the pregnant rats were randomly separated into two groups: the control group and the PEE group (n = 24 in each group).
The PEE group was administered 4 g/kg·d ethanol by gavage from GD9 to GD20, while the control group was administered the same volume of distilled water. Eight randomly selected pregnant rats from each group were anesthetized with iso urane and decapitated on GD14, GD17, and GD20. At these three time points, one fetal rat was randomly selected from each litter and xed with 95% alcohol for subsequent skeleton analysis. The knee joints of the remaining rats in the two groups were removed under an anatomical microscope, the right limbs were xed in a 4% paraformaldehyde (PFA) solution, and the left limbs were stored in a -80 ℃ freezer for gene detection. The scheme of animal experimental procedures was as follows.

Whole-mount skeletal staining
Whole-mount skeletal staining using Alcian blue and Alizarin red was performed as described previously (Singh et al. 2018). Rat embryos of GD17 and GD20 were collected, rinsed in PBS and xed in 95% ethanol for three days, followed by overnight xation in 100% acetone. Then, the tissues were stained for three days in a 1:1:1:17 volume mixture of glacial acetic acid:0.3% Alcian blue 8GX (Sigma-Aldrich) in 95% ethanol:0.1% Alizarin red in 70% ethanol:70% ethanol. After staining, the tissues were washed using 1% KOH and imaged under a Canon EOS 70D camera (Canon, Tokyo, Japan).

Histological measurement
Rat embryonic limbs on GD14, GD17, and GD20 were dissected, xed in 4% PFA at 4°C for three days, and embedded in para n, and 4 µm sections were cut along the parasagittal plane using a microtome. For histological analysis, Alcian blue and safranin O were applied to detect the glycosaminoglycan (GAG) content. Histological images were captured using a Nikon NIS Elements BR light microscope (Nikon, Tokyo, Japan).
Immunohistochemical staining and tissue immuno uorescence staining were performed following the manufacturer's protocol. Brie y, after dewaxing, EDTA containing antigen retrieval buffer (pH 8.0) was used for antigen retrieval. BSA was used to block the previously added primary antibody, and the primary antibody dilution ratios were as follows: anti-Col2a1 (1:200 dilution), anti-Col10a1 (1:200 dilution), anti-PTHrP (1:250 dilution), and anti-ERG (1:250 dilution). Immunohistochemistry was conducted using a DAB staining kit (GeneTech Company, Ltd., Shanghai, China). For tissue immuno uorescence, the primary antibody was detected with a uorescent Cy3-conjugated goat anti-rabbit IgG (H + L) (1:50 dilution) secondary antibody, after which the tissue sections were stained with 4′, 6-diamidino-2-phenylindole (DAPI) and sealed with an anti-uorescence quenching agent. All images were captured and then analyzed with a Nikon NIS Elements BR light microscope (Nikon, Tokyo, Japan). The staining intensity was calculated by measuring the mean integrated optical density (MOD) in 10 different elds for each sample.

Chondrogenic differentiation of BMSCs in alginate bead culture
The extraction of BMSCs and their culture in alginate beads were conducted as described in a previous study (De Ceuninck et al. 2004;Deng et al. 2012). Monolayer cultures were trypsinized, washed, and centrifuged. The isolated BMSCs were suspended at a concentration of 1×10 7 cells/ml in 1.25% alginate (Sigma-Aldrich, USA) in 0.15 M NaCl. The cell suspension was drawn into a syringe and slowly added to a 10 2 mM calcium chloride solution dropwise with a needle. Beads with approximately 5×10 5 cells/bead were cultured in chondrogenic medium, containing high-glucose DMEM supplemented with 1% insulin, transferrin, and selenous acid (ITS) (Sigma-Aldrich, USA), 100 nM dexamethasone (Sigma-Aldrich, USA), 50 µg/ml ascorbic acid-2-phosphate (Sigma-Aldrich, USA), 40 µg/ml l-proline (Sigma-Aldrich, USA), and 10 ng/ml transforming growth factor-β1 (TGF-β1) (PeproTech, USA). After chondrogenic differentiation for 3 weeks, alginate bead sections were stained with safranin-O to assess the number of cells and to locate GAG deposits. During the period of chondrogenic differentiation, the culture medium with or without ethanol at concentrations of 0, 30, 60, and 120 mM was replaced every other day.

Primary chondrocyte culture
Chondrocytes were isolated from GD17 rats and plated at a density of 2 × 10 5 cells per well in 6-well plates in medium (DMEM/F12 medium supplemented with 10% fetal bovine serum, 100 mg/ml streptomycin, and 100 U/ml penicillin). Primary chondrocytes at 80% con uence were used for further experiments. During the culture period, the cells were incubated at 37°C in a humidi ed atmosphere of 5% CO 2 and 95% air.
Subcon uent GD17 rat primary chondrocytes in 6-well plates were transfected in triplicate with 2.5 µg of ERG and PTHLH overexpression plasmids using Lipofectamine™ 3000 and P3000™ transfection reagent (Invitrogen, USA) according to the manufacturer's protocol. For miRNA inhibitor and mimic transfection, cells were transfected with 100 nM miR-200b-3p inhibitor or 50 nM miR-200b-3p mimic using Lipofectamine™ 3000 transfection reagent (Invitrogen, USA) according to the manufacturer's protocol. After 8 hours, the transfection e ciency was determined using a uorescence microscope. After 24 and 48 hours, the cells were collected for further analysis of gene and protein levels.

Cellular immuno uorescence staining
After treatment, the cells cultured on coated plates were washed three times with PBS, xed in 4% formaldehyde for 15 min, blocked with 3% BSA for 1 hour, and permeabilizes with 0.1% Triton X-100/PBS for 10 min. The cells were then incubated overnight at 4°C with primary antibodies in 0.5% BSA, including rabbit anti-Col2a1 (1:200 dilution), rabbit anti-Col10a1 (1: 50 dilution), mouse anti-ERG (1: 100 dilution), and mouse anti-PTHrP (1: 100 dilution). The cells were washed with PBS and incubated with 1:200 diluted Cy3-conjugated goat anti-rabbit and Fluor 594-conjugated goat anti-mouse secondary antibodies for 1 hour at room temperature. The nuclei were stained with DAPI at a 1:500 dilution for 5 min. The slides were washed twice with PBS, and uorescence images were captured using a confocal microscope (Smartproof 5, Carl Zeiss, Oberkochen, Germany). The staining intensity was determined by measuring the IOD in 10 different elds for each sample.

Western blotting
Brie y, resuspended cells were rinsed with ice-cold PBS and then lysed for 30 min at 4°C in RIPA lysis buffer containing phosphatase inhibitor cocktail, followed by analysis with the BCA Assay Kit for protein quanti cation. A total of 20 µg of protein was loaded into each lane, separated by SDS-PAGE, and blotted onto PVDF membranes (Millipore, MA, USA). The membranes were blocked in 5% nonfat milk for 1 hour and incubated overnight at 4°C with the primary antibody. The dilution concentrations of the primary antibodies were as follows: Col2a1 (1:1000), Col10a1 (1:800), PTHrP (1:3000), and ERG (1: 3000). Then, the blots were incubated with horseradish peroxidase (HRP)-conjugated secondary antibody (goat antirabbit IgG, 1:5000 and goat anti-mouse IgG, 1:5000) for 1.5 hours and visualized using ECL HRP substrate (PerkinElmer Inc., Boston, MA, USA). The antibody binding signals were detected using a ChemiDoc Image Analyzer (Bio-Rad, Hercules, CA, USA). The relative protein level was standardized to the GAPDH protein level. The protein band intensities were analyzed by ImageJ (National Institutes of Health, Bethesda, MD, USA) from 3 independent bands.

Total RNA extraction and RT-PCR
Total RNA was isolated from cartilage tissues and primary chondrocytes using TRIzol reagent following the manufacturer's protocol. The RT-PCR procedure was described by Li et al.(Qing-Xian et al. 2020). RNA was assayed for GAPDH, Col2a1, aggrecan, tenascin-C, Col10a1, PTHLH, and ERG espression. All primers were designed with Primer Premier 6.0 (Premier Biosoft International, Palo Alto, CA, USA). The primer sequences for the rat genes are shown in Table 1. GAPDH served as the reference gene to normalize the expression of other genes. The relative expression levels of target genes were calculated by the 2 −ΔΔCt method.

Luciferase reporter assay
Lipofectamine™ 3000 reagent was used to transiently transfect reporter plasmids into chondrocytes. For the dual luciferase assay, 6.0×10 4 chondrocytes in a 24-well plate were transfected with 100 nM SV40-ERG-WT or SV40-ERG-MUT. The chondrocytes were then cotransfected with 50 nM miR-200b-3p mimic or negative control (NC). After incubation for 48 h, the transfected cells were lysed and luciferase activity was detected using the Dual-Luciferase® Reporter Assay System (Promega, Madison, WI, USA). Fire y luciferase activity was normalized to Renilla luciferase activity. Three biological replicates were used for each experiment, and each experiment was repeated ve times.

Statistical analysis
Prism Graphics (GraphPad Software, La Jolla, CA, USA, version 8.0) was applied for all data analyses. Quantitative data are expressed as the mean ± S.D. The data of in vitro experiments with different ethanol concentrations were analyzed using one-way analysis of variance (ANOVA) with a post hoc test for multiple comparisons. For the data from the in vivo study, unpaired, two-tailed Student's t-tests were conducted for comparisons between the control and PEE groups. Correlation analysis of gene expression with and cartilage phenotype was performed using the Pearson correlation coe cient. Statistical signi cance was considered at P < 0.05 for all the tests.

Skeletal and limb development of PEE fetal rats at different times during pregnancy
To determine the effect of PEE on cartilage development, we examined the skeletal conditions of the fetal rats at different times (GD14, GD17, and GD20) (Fig. 1a). We found that the body weight, body length, femur length, and tibia length of PEE fetal rats were signi cantly lower than those of control rats on GD17 and GD20, but there was no signi cant difference between the two groups on GD14 (Fig. 1b).
Additionally, the growth ratios of body weight, body length, femur length, and tibia length between the PEE group and the control group were 1.52, 1.3, 1.41, and 1.3, respectively, during GD14-17, while the growth ratios of the two groups during GD17-20 were 1.04, 0.99, 1.09, and 1.09, respectively. These results suggest that PEE can reduce the body weight and shorten the skeleton length of fetal rats in utero, especially during GD14-17.
On GD14, Alcian blue staining showed that the interzone formed in both groups was similar, and the articular cavity had not yet formed in either groups (Fig. 1c), indicating that there was no difference in cartilage development between the two groups during joint formation. On GD17, the articular cavity had formed in the control group, while it was not fully formed in the PEE group (Fig. 1c), indicating that the formation of articular cartilage was blocked in the PEE group. In addition, compared with that in the control group, the range of the proliferative zone (PZ) in the PEE group was signi cantly reduced, while the range of the hypertrophic zone (HZ) was signi cantly increased (Fig. 1d). On GD20, the range of the PZ and the primary ossi cation center (POC) in the PEE group was noticeably smaller than that in the control group, while the HZ in the PEE group was remarkably larger than that in the control group (Fig.  1e). These results suggest that PEE can cause intrauterine limb dysplasia, mainly during the stage of articular cartilage formation (GD14-17), including narrowing of the cartilage proliferation area, delay in POC formation, and widening of the cartilage hypertrophy area.
3.2 Morphology of local cartilage and expression of marker genes in fetal rats with PEE at different times during pregnancy We observed the morphology of the local cartilage at different times during pregnancy (GD14, GD17, and GD20). On GD14, condensation initially displayed no overt morphological signs of the knee or elbow joint, as observed by Safranin O staining. Moreover, there was no signi cant difference in the optical density of GAG or the number of chondrocytes per unit area in the knee joint between the two groups ( Fig. 2a and   2b). Additionally, by observing the cartilage development of the ankle joint and toe of GD14 fetal rats, we found that the cartilage staining of the two groups was consistent, and there was no signi cant difference in the number of chondrocytes per unit area ( Supplementary Fig. 1). On GD17, compared with that of the control group, the GAG content in the knee joint cartilage of the PEE group was decreased, and the number of chondrocytes per unit area in the knee joint area was signi cantly reduced ( Fig. 2a and   2b). Similarly, we found consistent changes in the elbow joints of fetal rats ( Supplementary Fig. 2). On GD20, the development of articular cartilage in the PEE group was signi cantly slower than that in the control group ( Fig. 2a and 2b). In summary, PEE can slow the development of intrauterine cartilage.
To more thoroughly examine the effect of PEE on cartilage, RT-qPCR was used to quantify the expression of cartilage function genes. The results showed that the mRNA levels of tenascin-C, aggrecan, and Col2a1 on GD17 and GD20 in the PEE group were signi cantly lower than those in the control group, while there was no signi cant difference between the two groups on GD14 (Fig. 2c). Additionally, during GD14-17, the change ratios of tenascin-C, aggrecan, Col2a1 and Col10a1 mRNA expression between the control group and PEE group were 2.74, 5.3, 1.97, and 0.48, respectively, while during GD17-20, the change ratios were 1.1, 1.11, 1.13, and 0.95, respectively. These ndings suggest that the toxic effect of PEE on cartilage development mainly occurs on GD14-17. The results of tissue immuno uorescence of Col2a1 and Col10a1 were consistent with those of RT-qPCR (Fig. 2d-2f). Overall, PEE reduced the expression of genes associated with fetal articular cartilage development and increased the expression of hypertrophic genes, which mainly occured in the stage of articular cartilage formation.

Expression of miR-200b-3p, ERG, and PTHrP in PEE fetal rats at different times during pregnancy
To explore the mechanism of cartilage dysplasia induced by PEE, we detected the expression of cartilage regulatory genes, including ERG and PTHrP. The results of immunohistochemistry showed that the expression of ERG and PTHrP in the cartilage of PEE offspring was signi cantly lower than that in the control offspring on GD17 and GD20, but there was no signi cant difference between the two groups on GD14 (Fig. 3a-c). RT-qPCR showed the same results (Fig. 3d). Moreover, the change ratio of ERG and PTHrP mRNA expression levels between the control group and the PEE group was 3.7 and 2.95, respectively, in the articular cartilage formation stage (GD14-17); however, during GD17-20, the change ratio of the two groups was 1.03 and 1.11, respectively. There results suggest that PEE has a signi cant effect on the formation of articular cartilage. Furthermore, we found that the mRNA expression of ERG and PTHrP on GD17 was positively correlated with articular cartilage phenotypes, such as the content of articular cartilage matrix (GAG) and the number of chondrocytes per unit area (Fig. 3e), and negatively correlated with cartilage hypertrophy phenotypes, such as the mRNA expression of Col10a1 and HZ/PZ (Fig. 3f). We obtained the same results in GD20 fetal cartilage (Supplementary Fig. 3). These results indicate that the expression of ERG and PTHrP in fetal cartilage is downregulated by PEE, which is related to the decrease in articular cartilage differentiation and the increase in cartilage hypertrophy. Furthermore, we investigated the possible miRNAs targeting ERG by overlapping the predicted results of the miRDB, TargetScan, and microRNA.org databases (Supplementary Table 1). The common miRNAs are shown in the Venn diagram and include rno-miR-144-3p, rno-miR-139-5p, rno-miR-9a-3p, rno-miR-200b-3p and rno-miR-200c-3p (Fig. 3g). Then, we detected the expression of these miRNAs in cartilage tissue, which showed that only the expression of rno-miR-200b-3p in the PEE group was signi cantly higher than that in the control group in both GD17 and GD20 samples (Fig. 3h). These results suggest that the high expression of miR-200b-3p may mediate the development of articular cartilage by inhibiting the expression of ERG and PTHrP in fetal rats induced by PEE.

Effect of ethanol on chondroblast differentiation of BMSCs and the expression of the miR-200b-3p/ERG/PTHrP axis
To verify the effect and mechanism of ethanol on cartilage development, we used a BMSC threedimensional culture model to simulate intrauterine cartilage development (Supplementary Fig. 4) and found that different concentrations of ethanol had no cytotoxicity on the differentiation of BMSCs ( Supplementary Fig. 5). Safranin O staining showed that compared with that of the control group, the staining of groups treated with different concentrations of ethanol was signi cantly lighter, and the number of chondrocytes per unit area was reduced in a concentration-dependent manner (Fig. 4a-c). RT-qPCR analysis showed that the mRNA expression of tenascin-C, aggrecan, and Col2a1 in groups treated with different concentrations of ethanol was signi cantly decreased, and the mRNA expression of Col10a1 was signi cantly increased in a concentration-dependent manner (Fig. 4d-g). Moreover, compared with that in the control group, the expression of miR-200b-3p was signi cantly increased and the mRNA expression of ERG and PTHrP was signi cantly decreased in the ethanol group (Fig. 4h-j). The protein levels of Col2a1, Col10a1, ERG, and PTHrP were consistent with the mRNA expression levels, as observed by western blotting (Fig. 4k-l). These results indicate that ethanol inhibited BMSCs from differentiating into articular cartilage in a concentration-dependent manner, promoted cartilage hypertrophy, increased the expression of miR-200b-3p, and decreased the expression of ERG and PTHrP.

Regulatory effect of ERG/PTHrP on the differentiation of articular chondrocytes
To assess the involvement of ERG/PTHrP in the regulation of ethanol-induced differentiation of articular chondrocytes, we transfected GD17 primary chondrocytes with PTHLH and ERG overexpression plasmids. Immuno uorescence at 8 hours after transfection con rmed that the PTHLH and ERG plasmids were successfully transfected ( Supplementary Fig. 6). As shown in Fig. 5a, ethanol decreased the expression of tenascin-C, aggrecan, and Col2a1 and promoted the espression of Col10a1, while the overexpression of PTHLH reversed these changes, as observed by RT-qPCR. Western blotting and immuno uorescence also con rmed that overexpression of PTHLH could reverse the effect of ethanol on the protein levels of Col2a1 and Col10a1 (Fig. 5b). Similarly, overexpression of ERG in GD17 primary chondrocytes could reverse ethanol-induced changes in both the mRNA level (Fig. 5d) and protein level of marker genes (Fig. 5e-f). Additionally, overexpression of ERG could reverse the low expression of PTHrP induced by ethanol, but overexpression of PTHLH could not reverse the effect of ethanol on PTHrP expression (Fig. 5g-h). These results suggest that ERG regulates the differentiation of articular chondrocytes by acting through PTHrP.

Regulation of ERG and articular chondrocyte differentiation by miR-200b-3p
To further explore whether miR-200b-3p is involved in the regulation of ERG by ethanol, we designed a luciferase reporter assay (Fig. 6a). Our data showed that the miR-200b-3p mimic signi cantly inhibited luciferase activity in cells transfected with SV40-ERG-WT but not in those transfected with SV40-ERG-MUT (Fig. 6b). Therefore, miR-200b-3p can target and negatively regulate ERG. RT-PCR, western blotting, and immuno uorescence further con rmed that 60 mM ethanol signi cantly decreased the mRNA and protein levels of ERG, while the miR-200b-3p inhibitor could partially reverse the inhibitory effect of ethanol on the expression of ERG; in the NC + miR-200b-3p inhibitor group, the mRNA and protein expression of ERG was signi cantly increased compared with that in the control group (Fig. 6c-e).
In addition, the mRNA expression of tenascin-C, aggrecan, and Col2a1 was markedly reduced when the cells were exposed to 60 mM ethanol, and the mRNA expression of Col10a1 was dramatically upregulated; however, the miR-200b-3p inhibitor reversed the above changes (Fig. 6f). Moreover, western blotting and immuno uorescence results showed that the miR-200b-3p inhibitor could reverse the effect of ethanol on the protein levels of Col2a1 and Col10a1 (Fig. 6g-h). Overall, these lines of evidence suggest that miR-200b-3p negatively regulates the expression of ERG and partially reverses the decrease in ERG expression and articular cartilage dysplasia induced by ethanol.

Discussion
Page 12/25 4.1 Articular cartilage dysplasia in PEE offspring originated from the second trimester In the past 60 years, people have been aware of the harm of alcohol exposure during pregnancy, but alcohol intake during pregnancy is still an important public health problem (Vall et al. 2015). It has been reported that 5.8% of women in Canada and approximately 12.5% of women in the USA drink alcohol during pregnancy (Delano et al. 2019). In China 26.9% of adult women drink alcohol, and 0.2% of them are affected by alcohol dependence(Centers for Disease and Prevention 2009). Previous studies have found that offspring with PEE have characteristic craniofacial deformities, low birth weight and limb shortening, that the effect of alcohol on fetal bone development is di cult to correct after birth, and that high-dose alcohol exposure during pregnancy leads to life-long limb shortening in offspring (Day et al. 2002;Keiver and Weinberg 2004). In this study, we found that, consistent with the studies of Snow (Snow and Keiver 2007) and Pan (Pan et al. 2016), PEE could induce hypertrophic transition cartilage, increase the length of the hypertrophic cartilage area and increase the expression of the hypertrophy marker gene Col10a1; however, limb length and POC were short, which shows the inhibition of osteogenic differentiation. These differences were most prominent at GD17 and GD20, which showed that the weight, body length, femur length and tibia length of PEE fetal rats were signi cantly lower than those of control rats. This is related to the inhibition of the terminal differentiation of transitional cartilage by PEE.
Interestingly, in this study, we also found that PEE can cause dysplasia of articular cartilage, which shares a common tissue source with transitional cartilage. The theory that adult OA has intrauterine origin was In this study, we dynamically observed the effects of ethanol on articular cartilage development in fetuses at three time points: arthrogenesis (GD14), articular chondrogenesis (GD17) and just before birth (GD20). We found that on GD14, ethanol did not signi cantly affect the development of articular cartilage in offspring, while on GD17 and GD20, the development of articular cartilage in the ethanol group was signi cantly delayed compared with that in the control group. The growth rate analysis showed that the growth rate of body length and limb length from GD14 to GD17 in the control group was signi cantly higher than that in the PEE group, but there was no signi cant difference between the two groups from GD17 to GD20. Moreover, compared with those in the control group, the change rates of expression levels of the cartilage marker genes COL2A1 and Col10a1, as well as the cartilage development regulation genes ERG and PTHrP, in PEE group were signi cantly different from GD14 to GD17, while there was no signi cant difference between the two groups from GD17 to GD20. These results indicate that the toxicity of PEE to fetal cartilage mainly occurs in the period of articular cartilage formation, i.e., GD14-17, rather than in GD17-20, indicating that there is a sensitive period (articular cartilage formation stage) for the toxic effect of PEE on cartilage development, rather than a cumulative toxic effect of PEE. Previous studies also suggested that the cartilage progenitor cells in this period had strong plasticity, and external factors could induce them to differentiate into articular cartilage or hypertrophic cartilage (Kahn et al. 2009;Ray et al. 2015). We used the chondrogenic differentiation model of BMSCs to simulate the process of intrauterine articular cartilage differentiation, and obtained consistent results, that is, ethanol can inhibit the chondrogenic differentiation of BMSCs and promote their hypertrophy.  Singh et al. 2018). In this study, it was found that at GD17 and GD20, the mRNA and protein expression levels of ERG and PTHrP in the PEE group were signi cantly lower than those in the control group, and the expression levels of ERG and PTHrP were positively correlated with the articular cartilage phenotype and negatively correlated with the cartilage hyperplasia phenotype. At GD20, POC was formed in both groups, but the length of POC in the PEE group was signi cantly shorter than that in the control group.

ERG/PTHrP mediated articular chondrodysplasia in PEE
Meanwhile, the HZ region adjacent to the POC and the PZ region in the metaphysis of the PEE group were also signi cantly shorter than those in the control group. Similar to the changes in ERG expression, PTHrP mRNA expression in the PEE group was signi cantly lower than that in the control group at GD17 and GD20. Overexpression of PTHLH or ERG reversed the inhibitory effect of ethanol on the expression of articular cartilage marker genes in GD17 primary chondrocytes, indicating that low expression of ERG and PTHrP mediates ethanol-induced dysplasia of articular cartilage. In addition, the present study found that overexpression of ERG could reverse the low expression of PTHrP induced by ethanol, but overexpression of PTHLH could not reverse the low expression of ERG induced by ethanol, suggesting that ERG functions by regulating the expression of PTHrP. Previous studies also have con rmed that ERG can regulate

Summary
In conclusion, this study con rmed the existence of articular cartilage dysplasia in PEE offspring, which mainly occurred in the period of articular cartilage formation, namely, GD14-17. The mechanism is related to the decreased differentiation of chondrogenic progenitor cells into articular chondrocytes due to the upregulation of miR-200b-p expression by ethanol and subsequent downregulation of PTHrP expression by targeted inhibition of ERG. In this study, for the rst time, the development of cartilage at different stages of intrauterine development was detected, and the speci c time course and mechanism of articular chondrodysplasia induced by ethanol were clari ed, providing a more accurate and in-depth theoretical basis for the analysis of the developmental toxicity of PEE in fetal cartilage.
Ethics approval All animal experimental procedures were conducted in accordance with the Guidelines for the Care and Use of Laboratory Animals of the Chinese Animal Welfare Committee. Consent to participate: Not applicable. Consent for publication: All authors agree that this article will be published in Cell Biology and Toxicology.
Availability of data and material: The data that support the ndings of this study are available from the corresponding author upon reasonable request.    Col10a1 at different times during pregnancy. n=5. Data are reported as the mean ± SD. *P < 0.05, **P < 0.01, compared with the control group.  l) The protein expression and quanti cation of Col2a1, Col10a1, ERG, and PTHrP after the differentiation of BMSCs in the control group and the 60 mM ethanol group. n=3. Data are reported as the mean ± SD. *P < 0.05, **P < 0.01, compared with the control group.  n=3. (f) RT-qPCR showed that the miR-200b-3p inhibitor could reverse the effect of ethanol on the mRNA levels of tenascin-C, aggrecan, Col2a1, and Col10a1. (g-h) Immuno uorescence and western blotting showed that the miR-200b-3p inhibitor could reverse the effect of ethanol on the protein levels of tenascin-C, aggrecan, Col2a1, and Col10a1. n=3. Data are reported as the mean ± SD. *P < 0.05, **P < 0.01, compared with the NC group; ##P < 0.01, compared with the ethanol group.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Supplementarymaterial.docx