Advanced maternal age causes premature placental senescence and malformation via dysregulated α‐Klotho expression in trophoblasts

Abstract Advanced maternal age (AMA) pregnancy is associated with higher risks of adverse perinatal outcomes, which may result from premature senescence of the placenta. α‐Klotho is a well‐known antiaging protein; however, its expression and effect on the placenta in AMA pregnancies have not yet been fully elucidated. The expression patterns of α‐Klotho in mouse and human placentas from AMA pregnancies were determined by Western blotting and immunohistochemistry (IHC) staining. α‐Klotho expression in JAR cells was manipulated to investigate its role in trophoblastic senescence, and transwell assays were performed to assess trophoblast invasion. The downstream genes regulated by α‐Klotho in JAR cells were first screened by mRNA sequencing in α‐Klotho‐knockdown and control JAR cells and then validated. α‐Klotho‐deficient mice were generated by injecting klotho‐interfering adenovirus (Ad‐Klotho) via the tail vein on GD8.5. Ablation of α‐Klotho resulted in not only a senescent phenotype and loss of invasiveness in JAR cells but also a reduction in the transcription of cell adhesion molecule (CAM) genes. Overexpression of α‐Klotho significantly improved invasion but did not alter the expression of senescence biomarkers. α‐Klotho‐deficient mice exhibited placental malformation and, consequently, lower placental and fetal weights. In conclusion, AMA results in reduced α‐Klotho expression in placental trophoblasts, therefore leading to premature senescence and loss of invasion (possibly through the downregulation of CAMs), both of which ultimately result in placental malformation and adverse perinatal outcomes.

AMA pregnancies have higher risks of adverse perinatal outcomes, such as fetal loss, preterm birth, low birthweight, and preeclampsia (Abel et al., 2002;Jolly et al., 2000;Kortekaas et al., 2020;Sultana et al., 2018), all of which are believed to be due to premature placental senescence and consequent dysfunction. Senescence of the placenta is characterized by reduced telomerase activity (Biron-Shental et al., 2010), increased DNA damage and DNA oxidation, and increased expression of senescent biomarkers [including tumor suppressor p53 and cyclin-dependent kinase (CDK) inhibitors p16 and p21 (Londero et al., 2016)] and senescence-associated secretory phenotype biomarkers IL-6 and IL-8 (Lu et al., 2017). Although numerous studies have reported that α-Klotho is widely expressed in human placental tissue, the involvement of α-Klotho in AMA-related placental senescence has not yet been reported. In this study, we aimed to explore the role of α-Klotho in regulating trophoblastic senescence in the context of AMA.

| AMA is associated with placental senescence
Human term placentas from AMA pregnancies showed significantly upregulated expression of multiple well-known biomarkers of senescence, including p53, p21, and p16, compared with that in control placentas ( Figure 1a). Interestingly, such elevations in p53, p21, and p16 expression in AMA placentas can be detected as early as the first trimester ( Figure 1b). Similarly, the expression levels of p53, p21, and p16 in placentas from aged mice were also significantly higher than those in placentas from young mice ( Figure 1c). In accordance with these findings, more positive senescence-associated β-galactosidase (SAβ-Gal) staining was observed in both term human placental tissues and first trimester villi of AMA pregnancies than in corresponding controls from young pregnancies (Figure 1d), while a similar manifestation was also observed in mouse placentas ( Figure 1e). Taken together, these facts strongly indicate that AMA caused more severe senescence in the placenta, which started from the early stage of placentation.

K E Y W O R D S
advanced maternal age, placenta, senescence, trophoblast, α-klotho F I G U R E 1 AMA placentas showed a senescence phenotype. (a) Western blotting of p21, p53, and p16 protein expression in human term placentas, n = 30 in the control and n = 37 in the AMA groups; (b) Western blotting of p21, p53 and p16 protein expression in human first trimester villi, n = 6 in each group; (c) Western blotting of p21, p53, and p16 protein expression in mouse placentas collected on GD18.5, n = 6; (d) representative images of SAβ-Gal staining of human term placenta sections. Quantification of the area of positive signal per sample (n = 3 patients per group, 3 random fields per patient). Scale bars, 100 μm; (e) representative images of SAβ-Gal staining of sections of mouse placentas collected on GD18.5. Quantification of the area of positive signal per mouse (n = 3 mice per group, 3 random fields per mouse). Scale bars, 100 μm. All data are presented as the mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001. Mann-Whitney U test. NS, nonsignificant; AU, arbitrary unit. All experiments were performed in triplicate.

Klotho expression in the placenta
We then assessed the expression pattern of α-Klotho in placental tissue. In humans, α-Klotho is ubiquitously expressed in various types of placental trophoblast cells, including cytotrophoblasts (CTBs), cell column trophoblasts (CCT), syncytiotrophoblasts (STB), and interstitial extravillous trophoblast (iEVT) cells (Figure 2a, b & Figure S1,S2). Importantly, IHC staining also suggested that α-Klotho expression was compromised in CTBs, STB in floating villi (FV), and iEVTs in BPs of term placentas and first-trimester decidua collected from AMA pregnancies, which was confirmed by Western blotting (Figure 2c,d). Similarly, α-Klotho was expressed in both the labyrinth zone (Lz) and junctional zone (Jz) areas in the mouse placenta ( Figure   S3). The reduction in placental α-Klotho protein expression in human AMA pregnancies was confirmed in placentas collected from aged mice at GD18.5 compared with those of young controls ( Figure 2e).
Placental α-Klotho expression remained stable throughout gestation in both humans and mice ( Figure S4). Our findings suggest that AMA is associated with deficiency of α-Klotho expression in placentas during placentation, which may be responsible for the senescence of the placenta.

| α-Klotho deficiency induces senescence in JAR cells and compromises its invasiveness
To investigate whether α-Klotho is a determinant of senescence in placental trophoblasts, α-Klotho-knockdown (Sh-KL) and overexpression (OE-KL) JAR cells were generated, whereby α-Klotho expression was repressed and elevated by nearly 50%, respectively Furthermore, DNA synthesis among groups was not significantly different ( Figure S5).

| α-Klotho-deficient mice demonstrated a placental senescence phenotype and compromised placentation
To verify our findings from the in vitro studies, α-Klotho-deficient mice were generated by injecting GFP-tagged klotho interfering adenovirus (Ad-ADPr-klotho-GFP, Ad-Klotho) via the tail vein into pregnant females on GD8.5. The experimental design is depicted in Figure 4a. A significant reduction in placental α-Klotho expression was observed on GD14.5 and GD18.5 ( Figure 4b). As a result, the expression levels of p53, p21, and p16 were significantly
F I G U R E 5 α-Klotho regulates the genes of the CAM pathway in JAR cells. (a) Volcano plot of the significant differences in gene expression levels between Sh-KL and Sh-NC JAR cells; genes showing the greatest differences were analyzed by (b) KEGG and (c) GO analysis; (d-e) mRNA levels and protein levels of CDH4, CLDN3, and ITGAM were examined separately by RT-qPCR and Western blotting in various JAR cells, n = 3 per group; (f-g) mouse placentas collected at different gestational ages: GD14.5 (Ad-Ctrl, n = 6; Ad-Klotho, n = 6) and GD18.5 (Ad-Ctrl, n = 5; Ad-Klotho, n = 5); (h-i) placentas from young and aged mice on GD18.5, n = 6; (j-k) human term placentas, protein levels (n = 6 per group), gene level, n = 7/8; and (l-m) human first trimester villi, protein levels (n = 6 per group), gene level, n = 7/8. All data are presented as the mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001. Mann-Whitney U test or one-way ANOVA. NS, nonsignificant. All experiments were performed in triplicate.

| DISCUSS ION
Physiological aging of the placenta is an inevitable biological process (Sultana et al., 2018). However, aberrant or premature placental senescence disturbs the normal physiological function of the placenta, particularly the nutrient-oxygen exchange function, which could lead to adverse pregnancy outcomes (Cox & Redman, 2017;Sultana et al., 2018). α-Klotho is a well-known antiaging protein, and its expression declines with increasing age. However, the understanding of its role in placental senescence, especially in the context of AMA, is very limited. The present study revealed that α-Klotho is expressed in various types of trophoblasts in the human placenta, which is consistent with previous results (Fan et al., 2016;Loichinger et al., 2016), implying that α-Klotho expression possesses critical biological and pathological relevance in the placenta. Indeed, it has been reported that α-Klotho expression levels in placentas from women diagnosed with preeclampsia or who delivered offspring that were small for gestational age (SGA) are significantly lower than those in placentas from women with healthy pregnancies (Fan et al., 2016;Iñiguez et al., 2018). α-Klotho plays a role in the activation of the IGF-I signaling pathway, which is vital for fetal growth (Iñiguez et al., 2018;Ohata et al., 2011) and may be associated with maintaining the balance of the fetomaternal calcium gradient (Ohata et al., 2011). Therefore, we speculated that these mechanisms may underlie the unfavorable birth outcomes associated with AMA.
The average maternal age of AMA subjects in this study was 40.86 years old, which is consistent with the reported age associated with a decline in α-Klotho expression (Yamazaki et al., 2010).
Studies have shown that α-Klotho expression wanes in the skeletal muscle of aged male mice (Sahu et al., 2018) and aged human skin (Behera et al., 2017). These results indicated that the expression level of α-Klotho may be associated with increased age. Previous findings from mouse muscle cells suggested that methylation of the Klotho gene promoter may be involved in the regulation of Klotho expression (Sahu et al., 2018); therefore, epigenetic modification could be the underlying mechanism of the downregulation of α-Klotho expression in AMA trophoblasts.
Interestingly, our findings also indicated that gestational age had no effect on α-Klotho expression, as placental α-Klotho expression remained stable from the beginning to the end of gestation in both humans and mice. Moreover, in our animal experiments, paternal influence was excluded since all the female mice were bred with the same young male mouse. The results strongly suggested that AMA is an independent risk factor for the loss of placental α-Klotho expression beginning in the early phase of pregnancy, implying that placental α-Klotho expression is barely influenced by the life cycle of the placenta but is rather determined by maternal factors of the embryo. However, a previous study in another group showed that α-Klotho expression in human placentas was downregulated as gestational age advanced (Iñiguez et al., 2018). The inconsistency between these findings may be due to the discrepancy in specimen collection. In our study, all human term placentas were collected from cesarean sections before parturition, whereas two-thirds of the samples in Iñiguez's study were collected after natural delivery. Therefore, the effects of parturition and mechanistic stress from vaginal birth on placental α-Klotho levels require further exploration.
To further investigate the effects of α-Klotho deficiency on trophoblasts, we modulated α-Klotho expression in JAR cells, which highly express α-Klotho ( Figure S11), and found that α-Klotho deficiency promotes premature senescence, as reflected by the increased expression of p53 and p21. Interestingly, the expression of p16 in Sh-KL JAR cells did not exhibit the upregulation observed in AMA placentas, possibly because p21 is downstream of p53 and is therefore expressed accordantly to mediate irreversible DNA damage and telomere shortening and additionally contribute to the stability and maintenance of cellular senescence. Our findings are consistent with a report that the downregulation of Klotho induces premature senescence of human primary fibroblast cells via a p53/ p21-dependent pathway (de Oliveira, 2006) without an increase in p16 expression. p16 mediates senescence through the retinoblastoma (Rb) pathway, inhibiting the action of cyclin-dependent kinases and leading to G1 cell cycle arrest (Rayess et al., 2012). p16 is also a tumor suppressor (Romagosa et al., 2011) and can be overex- In accordance with our results, α-Klotho deficiency accelerated senescence in primary human fibroblast cells (MRC-5) (de Oliveira, 2006). In contrast, supplementation with α-Klotho expression relieves cellular senescence in human umbilical vein endothelial cells (HUVECs) and MRC-5 cells, as shown by the decline in p53 and p21 expression (Ikushima et al., 2006). In this study, overexpression of α-Klotho did not result in a significant reduction in either SAβ-gal staining or the expression of p53, p21, and p16, probably because of the high expression levels of α-Klotho in JAR cells, which therefore maintains low expression levels of senescent biomarkers under basal conditions. Nevertheless, the relationship between senescence and α-Klotho expression in trophoblasts remains unclear. Therefore, we manipulated the senescence status of JAR cells and found that senescence was the result rather than the cause of α-Klotho deficiency in trophoblasts ( Figure S12). Moreover, the secretome of senescent trophoblast cells did not disturb α-Klotho expression in adjacent cells ( Figure S13).
Accumulating evidence suggests that the antiaging mechanism of α-Klotho is possibly derived from but not limited to activating the fibroblast growth factor (FGF) 23 signaling pathway, which inhibits tissue atrophy by promoting cell proliferation and preventing cell death (Medici et al., 2008;Sugiura et al., 2005), drives the expression of the mitochondrial enzyme manganese superoxide dismutase
In our study, we found that Sh-KL JAR cells display a decline in invasion ability, and it has been established that placental trophoblast invasion is of particular importance in spiral artery remodeling, which is critical for placental function (Pijnenborg et al., 1980). Haram et al found that the adhesion molecule-mediated invasion ability of trophoblasts is involved in the pathophysiology of preeclampsia (Haram et al., 2019). Therefore, we hypothesized that a decrease in invasion ability in placentas from AMA may be involved in pathophysiological placental development and activity.
Intriguingly, despite the unchanged expression of p53, p21, p16, and SAβ-gal, JAR cell lines overexpressing α-Klotho exhibited significantly improved invasive capability. This fact clearly suggests that α-Klotho regulates JAR cells through a mechanism different from that of general biomarkers of senescence. Nonetheless, whether this regulatory mechanism of JAR cells is derived from their trophoblastic origin or tumor origin requires further investigation. Then, unbiased mRNA sequencing was performed, and the data demonstrated that CAMs, CDH4, CLDN3, and ITGAM are potential downstream target genes regulated by α-Klotho in JAR cells. We then confirmed that CAM expression is required for the invasion of JAR cells, which is in accordance with previous reports noting that CDH4, CLDN3, and ITGAM are involved in the regulation of cell invasion. For example, Schumann et al reported that CLND3 was highly expressed in murine trophoblast giant cells and in human EVT cells and that it plays a role in the regulation of trophoblast invasion (Schumann et al., 2015;Zhao et al., 2020). Similarly, CDH4 deficiency impacts the invasion of human osteosarcoma cells , and epithelial ovarian cancer (EOC) invasion is mediated by ITGAM (Lyu et al., 2020). Moreover, the CAM pathway has been widely reported to be involved in regulating cell invasion (von Lersner et al., 2019;Thiery et al., 1988). Collectively, impaired α-Klotho expression and consequent abnormal expression of adhesion molecules may be involved in the aberrant invasive activity of trophoblasts.
Although α-Klotho−/− mice show a premature senescence phenotype and are widely used for the study of aging, their short lifespan (60.7 days) largely impedes their application in the study of AMA; therefore, the effects of α-Klotho on the reproductive system, especially in the placenta, remain unclear. In this work, we generated an α-Klotho-deficient mouse model by injecting klotho-interfering adenovirus (Ad-Klotho) via the tail vein on GD8.5; this approach allows for the exclusion of the effects of α-Klotho deficiency before placentation. In accordance with the results of cell experiments, the placentas of α-Klotho-deficient mice demonstrated a placental senescence phenotype. The mature placenta of the mouse consists of three parts: the labyrinth, spongiotrophoblast, and maternal decidua. The trophoblast and its associated fetal blood vessels form abundant branching to form a densely packed structure called the labyrinth (Watson & Cross, 2005). When the labyrinth develops, it is supported structurally by the spongiotrophoblast, which comprises a tight layer of nonsyncytial cells sandwiched between the labyrinth and the outer giant cells (Rossant & Cross, 2001). These structures are beneficial to the exchange of materials between the maternal and fetal environments. In the present study, α-Klotho-deficient mice exhibited an abnormal Jz/Lz area ratio indicative of placental malformation, which may be related to diminished transplacental nutrient transport and the consequent poor development of the fetus.
To our knowledge, this study is the first to report that AMA diminishes α-Klotho in placental trophoblasts, therefore leading to premature senescence and loss of function, ultimately resulting in malformation of the placenta, which may contribute to placental deficiency and adverse perinatal outcomes. These findings demonstrated the importance of α-Klotho in modulating placental senescence. To surmount senescence in the AMA placenta by targeting α-Klotho, specific delivery of recombinant α-Klotho into the placenta may be achievable via nanoparticles coated with placentahoming peptides (Tobita et al., 2017;Zhang et al., 2018).

| Patient and sample collection
Term placental tissue from women experiencing young pregnancies (20-25 years old, 37.14-41.14 weeks, n = 30) or AMA pregnancies (35-45 years old, 36.28-40.71 weeks, n = 37) who were admitted to the First Affiliated Hospital of Chongqing Medical University for cesarean sections was collected as previously described (J. E. Wagner et al., 1992). Patients with pregnancy complications, such as gestational diabetes mellitus, fetal growth restriction, spontaneous abortion, renal disease, or preeclampsia, were excluded. Young firsttrimester villi (20-25 years old, n = 20) and AMA first-trimester villi (35-40 years old, n = 12) were collected from subjects who legally and voluntarily terminated their pregnancy between 6 and 10 weeks of gestational age. Patients with a history of spontaneous abortion or ectopic pregnancy were excluded.

| Immunohistochemistry
The villous and placental tissues were washed with PBS, fixed overnight with 4% paraformaldehyde at room temperature (RT), and sectioned into 4μm-thick sections after they were dehydrated and were quantified. The detailed quantitative procedures used in our study were derived from the published literature (Jensen, 2013).
Furthermore, the area of blood sinuses in the labyrinth was quantified using ImageJ 1.50i software as previously reported (Varghese et al., 2014). Briefly, the positively stained area was divided by the total area. Three random-view fields of each sample in 3 independent experiments were quantified.

| RNA extraction and RT-qPCR
Total RNA was extracted from JAR cells and placental tissues using TRIzol reagent (Invitrogen, USA). The RNA concentration was measured by ultraviolet spectroscopy (NanoDrop2000, Thermo, MA, USA). One microgram of total RNA was reverse transcribed to cDNA with a Prime Script RT reagent kit (Roche Life Science, Germany).
The design and synthesis of primers were performed by TaKaRa

| Matrigel invasion assay
The invasion assay was performed as previously reported by our laboratory (Yang et al., 2020

| mRNA sequencing
Negative control and α-Klotho-knockdown JAR cells were collected for total RNA extraction (Invitrogen, USA). Total RNA was assessed by agarose gel electrophoresis for quality checks. Purity and integrity tests of RNA were conducted using a NanoPhotometer ® spectrophotometer (IMPLEN, USA) and an Agilent 2100 bioanalyzer (Agilent Technologies, US), respectively. A total of 1 μg of RNA from each sample served as input material for the sample preparations.
Sequencing libraries were established using the NEBNext®UltraTM RNA Library Prep Kit for Illumina® (NEB, USA) according to the manufacturer's protocol. Differential expression analysis was performed using the DESeq2 R package (1.16.1). An adjusted P-value of 0.05 and absolute fold-change value of 2 served as the thresholds for significant differential expression. Differentially expressed genes were further analyzed by GO and KEGG database analyses.

| Statistical analyses
The data are presented as the mean ± SEM. Statistical data were analyzed by the Mann-Whitney U test or one-way ANOVA. A value of p < 0.05 was considered significant. The statistical analyses were performed using Prism7 software (GraphPad Software, La Jolla, CA, USA).

CO N FLI C T S O F I NTE R E S T
The authors have declared no conflicts of interest.

AUTH O R CO NTR I B UTI O N S
CT and HQ conceived and designed the study; ZC, LX, HJ, JY, XL, HF, and LW performed the experiments and analyzed the data; MK, RS and PB interpreted the results; CT, LW and HQ provided funding resources; ZC wrote the draft; CT, MK, and PB edited the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data and materials described in the manuscript will be available upon reasonable request made to the corresponding authors, and delivery charges and agreement of usage may apply. The RNAseq data reported in this paper have been deposited in a public data depository under accession number HRA000693 and are publicly accessible at http://bigd.big.ac.cn/gsa-human