Critical Role of Cathepsin L/V in Regulating Endothelial Cell Senescence

Simple Summary Endothelial cell senescence has been considered as an initiation in the progress of vascular aging leading to the advancement of cardiovascular diseases, while the mechanism of endothelial cell senescence remains elusive. This study aimed to investigate the critical role of cathepsinL/V in endothelial cell senescence. We found that cathepsinL/V was decreased in senescent endothelial cells, which enhanced aldehyde dehydrogenase 1 family member A2 (ALDH1A2) expression and activated AKT/ERK1/2-P21 pathway, and therefore promoted cellular senescence, which may play an important role in vascular aging. These findings suggest that cathepsinL/V may be a potential therapeutic target in endothelial cell senescence. Abstract The senescence of vascular endothelial cells (ECs) is characterized as a hallmark of vascular aging, which leads to the initiation, progress, and advancement of cardiovascular diseases. However, the mechanism of the ECs senescence remains elusive. In this study, thoracic aortas were separated from young (8-week-old) and aged (18-month-old) mice. Decreased Ctsl expression and increased vascular remodeling were observed in senescent aorta. H2O2 was used to induce human umbilical vein endothelial cells (HUVECs) senescence, as shown by increased SA-β-gal positive cells and upregulated p21 level. CTSV significantly decreased after H2O2 treatment, while over-expression of CTSV by adenovirus reduced cellular senescence. RNA sequencing analysis was conducted subsequently, and ALDH1A2 was observed to significantly increased in H2O2 group and decreased after over-expression of CTSV. This result was further confirmed by RT-PCR and WB. Moreover, over-expression of CTSV reduced the increase of ERK1/2 and AKT phosphorylation induced by H2O2. Additionally, retinoic acid (RA), the major production of ALDH1A2, was added to CTSV over-expressed senescent HUVECs. Administration of RA activated AKT and ERK1/2, induced the expression of p21, and enhanced SA-β-gal positive cells, while not affecting the expression of CTSV and ALDH1A2. These results were further confirmed in doxorubicin (DOX)-induced senescent ECs. In conclude, we have identified that Ctsl/CTSV plays a key role in ECs senescence by regulating ALDH1A2 to activate AKT/ ERK1/2-P21 pathway. Therefore, targeting Ctsl/CTSV may be a potential therapeutic strategy in EC senescence.

function [20]. Recent findings suggested that CTSV decreased in the skin tissue of aged patients [21], while the mechanism of Ctsl/CTSV in aging is elusive. Since ECs senescence is the crucial step of the aging process, we speculated that the Ctsl/CTSV may regulate the ECs senescence to lead to aging.
In the present study, we found Ctsl expression decreased in the artery of aging mice. We used adenovirus to overexpress CTSV in ECs and further confirmed CTSV deficiency induced ECs senescence. Transcriptome analysis showed that endothelial CTSV participate in ECs senescence by regulating ALDH1A2 via activating AKT/ERK1/2-P21 pathway. Therefore, our results reveal that Ctsl/CTSV may emerge as a promising potential target of EC senescence.

Mice
All animal study experiments were carried out conforming to the National Institutes of Health guidelines and were approved by the Animal Care and Utilization Committee of Xiangya Hospital, China. The 18-month-old and 8-week-old C57 BL/6 J mice were purchased from Vita River laboratory. The mice were sacrificed with 1% pentobarbital sodium. Thoracic aortas were isolated and cut into two sections. One fixed in 4% paraformaldehyde for histology and immunohistochemistry (IHC), while the other stored in liquid nitrogen for western blot (WB) analysis.

Senescence-Associated β-Galactosidase (SA-β-gal) Staining
SA-β-gal staining was conducted according to the manufacturer's instructions (CST9860S, Cell Signaling Technology, Danvers, MA, USA) [25]. First, 1 × 10 6 HUVECs were seeded in 6-well plates, cells were rinsed once in PBS after treatments, and fixed in fixative solution for 15 min. After rinsing twice in PBS, fresh SA-β-gal staining solution was added and the plate was placed at 37 • C for 16 h. For each sample, total cells and SA-β gal positive cells were counted in three random visual fields using Image J. The percentage of SA-β gal positive cells was calculated as the ratio of SA-β-gal positive cells to total number of HUVECs.

Wound Healing Assay
First, 5 × 10 5 cells were seeded in 12-well plates. After treatments, clear wounds were created using the end of 200 µL pipette tips. Images of different times were captured using a phase-contrast microscope (Eclipse, Nikon, Tokyo, Japan).
Original Western blots used in this manuscript were uploaded as "Supplemental Materials".

Quantitative Real-Time PCR (qRT-PCR)
First, 1 × 10 6 HUVECs were seeded in 6-well plates, total RNA was isolated from cells using AG RNAex Pro Reagent (AG21102, Accurate Biology, Kunming, China) with instructions of the manufacturer. Nano Drop 2100 (Thermo Fisher Scientific, Waltham, MA, USA) was used to qualify and quantify total RNA. cDNA was obtained using an Evo M-MLV Mix Kit with gDNA Clean for qPCR (AG11728, Accurate Biology, China) and the mRNA expression was quantified by qRT-PCR using SYBR Green Premix Pro Taq HS qPCR (Rox Plus) Kit (AG11718, Accurate Biology, China). Relative expression levels were obtained using the 2−∆∆Ct method and normalized by β-actin. The primers involved in this study were synthesized by Sangon Biotech (Shanghai, China) and are shown in Table S1.

RNA Sequencing and Bioinformatic Analysis
Total RNA was isolated using Trizol (Invitrogen, Waltham, MA, USA) and purified using Oligo(dT)-attached magnetic beads. A-tailing mix and RNA index adapters were added after RNA fragmentation and cDNA generation. Agilent Technologies 2100 bioanalyzer was used for quality control and a library was constructed subsequently. BGIseq500 platform (BGI-Shenzhen, Shenzhen, China) was used to sequence the library. The sequencing data were filtered with SOAPnuke (v1.5.2) [26]. Differentially expressed genes were recognized using the DESeq2 (v1.4.5) with Q value < 0.05 [27]. Heatmap was drawn by pheatmap (v1.0.8).

Intracellular Retinoic Acid (RA) Analysis
First, 5 × 10 5 cells were seeded in 12-well plates. After treatments, cells were lysed with ultrasonicator on ice and the supernatants were collected. Concentration of retinoic acid was measured using a human retinoic acid ELISA kit (CSB-E16712h, CUSABIO, Wuhan, China) according to the manufacture's description [28].

Statistical Analyses
Data were expressed as mean ± SEM. If the data were normally distributed, t-test or one-way ANOVA with Tukey Kramer test was used; otherwise, non-parametric test was applied. Data analyses were performed by ImageJ 1.51j8, SPSS version 23 or Graph Prism 6. p value < 0.05 was considered statistically significant.

Ctsl Expression Is Decreased in Senescent Aorta
To investigate the relationship between Ctsl and vascular aging, we separated the thoracic aortas of 18-month-old and 8-week-old mice. A significant increase of MT, MT/LD, MA, and collagen/MA was observed in the aortas of aged mice by histological analyses ( Figure 1A-C), indicating the presence of vascular remodeling. The expression of Ctsl in young and senescent aortas was analyzed by IHC and WB subsequently. As shown in Figure 1D,E, Ctsl significantly decreased in senescent aortas, with increased expression of p21. Therefore, we may speculate that Ctsl participates in the procedure of the vascular aging.

CTSV Expression Is Inhibited in H 2 O 2 -Induced Cellular Senescence
Since ECs senescence has been recognized as a primary process of vascular aging, we explored the connections between CTSV and endothelial cell senescence in vitro. H 2 O 2 was widely used as stimulation of inducing cell senescence; therefore, in our study, HUVECs were treated with different dose of H 2 O 2 (0 µM, 50 µM, 100 µM, 200 µM, and 400 µM) for 24 h to induce cellular senescence. With the raised concentration of H 2 O 2 increased, SA-β-gal positive cells accordingly increased, as well as the expression of P21 and P16, indicating the presence of senescence. Meanwhile, CTSV significantly decreased after treated with 400 µM H 2 O 2 ( Figure 2A,C and Figure S1A,B), thus the concentration of 400 µM H 2 O 2 was adapted for further study. However, the expression of P53 remained unchanged at both mRNA and protein level, indicating that P53 was not involved in H 2 O 2 -induced cellular senescence, which was consistent with previous studies [29]. These findings demonstrated that CTSV decreases with cellular senescence in vitro.   (E) Western Blot of Ctsl and p21 in young and senescent aortas (n = 3-4). Data were presented as mean ± SEM. T-test was used. * p < 0.05 vs. young, ** p < 0.01 vs. young, *** p < 0.001 vs. young. induced cellular senescence, which was consistent with previous studies [29]. The findings demonstrated that CTSV decreases with cellular senescence in vitro.

Over-Expression of CTSV Reduces Cellular Senescence
To further investigate the effects of CTSV on cellular senescence, adenovirus w applied to enhance the expression of CTSV in HUVECs before treating with H2O2. It shown in Figure 3D,E that CTSV significantly increased after transfection with ad-CTS Meanwhile, SA-β-gal positive cells decreased after over-expression of CTSV ( Figure 3A Moreover, EdU assay and wound healing assay were conducted to determine E proliferation and migration were influenced in this process. As shown in Figure 3B H2O2 treatment impaired cell proliferation and migration, while over-expression of CT reversed the effect of H2O2. These data indicate that CTSV reduces cellular senescence a enhances cell proliferation and migration in HUVECs. Additionally, the increase of P induced by H2O2 was reduced by ad-CTSV at both mRNA and protein level ( Figure 3D, while the expression of P53 and P16 was unaffected by ad-CTSV ( Figure S1C,D Therefore, CTSV is supposed to reduce cellular senescence via P21-dependent pathwa

Over-Expression of CTSV Reduces Cellular Senescence
To further investigate the effects of CTSV on cellular senescence, adenovirus was applied to enhance the expression of CTSV in HUVECs before treating with H 2 O 2 . It is shown in Figure 3D,E that CTSV significantly increased after transfection with ad-CTSV. Meanwhile, SA-β-gal positive cells decreased after over-expression of CTSV ( Figure 3A). Moreover, EdU assay and wound healing assay were conducted to determine ECs proliferation and migration were influenced in this process. As shown in Figure 3B,C, H 2 O 2 treatment impaired cell proliferation and migration, while over-expression of CTSV reversed the effect of H 2 O 2 . These data indicate that CTSV reduces cellular senescence and enhances cell proliferation and migration in HUVECs. Additionally, the increase of P21 induced by H 2 O 2 was reduced by ad-CTSV at both mRNA and protein level ( Figure 3D,E), while the expression of P53 and P16 was unaffected by ad-CTSV ( Figure S1C,D). Therefore, CTSV is supposed to reduce cellular senescence via P21-dependent pathways. Biology 2023, 12, x 8 of 18 Data are presented as mean ± SEM. One-way ANOVA test was used. * p < 0.05, ** p < 0.01, *** p < 0.001.

CTSV Regulates ECs Senescence via ALDH1A2
To explore the anti-senescence mechanism of CTSV, RNA sequencing analysis was conducted. Compared with the control group, 3706 upregulated and 3580 downregulated genes were recognized after H2O2 treatment. As for HUVECs treated with H2O2 and ad-CTSV, 1 upregulated and 11 downregulated genes were recognized compared with H2O2 group. The Venn diagram shows that 8 genes which were upregulated after H2O2 treatment were downregulated after ad-CTSV ( Figure 4A). The heatmap of these genes is presented in Figure 4B. . Data are presented as mean ± SEM. One-way ANOVA test was used. * p < 0.05, ** p < 0.01, *** p < 0.001.

CTSV Regulates ECs Senescence via ALDH1A2
To explore the anti-senescence mechanism of CTSV, RNA sequencing analysis was conducted. Compared with the control group, 3706 upregulated and 3580 downregulated genes were recognized after H 2 O 2 treatment. As for HUVECs treated with H 2 O 2 and ad-CTSV, 1 upregulated and 11 downregulated genes were recognized compared with H 2 O 2 group. The Venn diagram shows that 8 genes which were upregulated after H 2 O 2 treatment were downregulated after ad-CTSV ( Figure 4A). The heatmap of these genes is presented in Figure 4B.  . Data are presented as mean ± SEM. One-way ANOVA test was used. * p < 0.05, ** p < 0.01, *** p < 0.001.
In addition, the mRNA expression of the recognized 8 genes was confirmed by RT-PCR and 5 of which showed the same tendency as RNA sequencing, including ALDH1A2, IGFBP5, NOS3, THBD, and SEMA3F ( Figures 4C and S2). Since the role of ALDH1A2 and SEMA3F has not been widely studied in cellular senescence [30][31][32], we further analyzed the protein level of ALDH1A2 and SEMA3F by WB. As shown in Figures 4D and S3, the protein expression of SEMA3F was basically unchanged, while ALDH1A2 was significantly increased in H2O2 group and decreased after over-expression of CTSV. Therefore, ALDH1A2 may be a novel mediator of cellular senescence. It was reported that ALDH1A2 mainly acts as an enzyme that catalyzes the synthesis of retinoic acid (RA) from retinaldehyde, while RA can upregulate the expression of p21 via ERK1/2 and AKT pathways [33,34]. We analyzed the expression and phosphorylation of ERK1/2 and AKT One-way ANOVA test was used. * p < 0.05, ** p < 0.01, *** p < 0.001.
In addition, the mRNA expression of the recognized 8 genes was confirmed by RT-PCR and 5 of which showed the same tendency as RNA sequencing, including ALDH1A2, IGFBP5, NOS3, THBD, and SEMA3F ( Figure 4C and Figure S2). Since the role of ALDH1A2 and SEMA3F has not been widely studied in cellular senescence [30][31][32], we further analyzed the protein level of ALDH1A2 and SEMA3F by WB. As shown in Figure 4D and Figure S3, the protein expression of SEMA3F was basically unchanged, while ALDH1A2 was significantly increased in H 2 O 2 group and decreased after over-expression of CTSV. Therefore, ALDH1A2 may be a novel mediator of cellular senescence. It was reported that ALDH1A2 mainly acts as an enzyme that catalyzes the synthesis of retinoic acid (RA) from retinaldehyde, while RA can upregulate the expression of p21 via ERK1/2 and AKT pathways [33,34]. We analyzed the expression and phosphorylation of ERK1/2 and AKT subsequently and found that ad-CTSV reduced the increase of ERK1/2 and AKT phosphorylation induced by H 2 O 2 .

CTSV Attenuates Cellular Senescence through ALDH1A2-AKT/ERK1/2-p21 Pathway
To verify whether CTSV attenuates cellular senescence by ALDH1A2, exogenous RA was added to HUVECs. RA reversed the anti-senescence effect of CTSV in H 2 O 2 treated cells, as shown by increased SA-β-gal positive cells and impaired cell proliferation and migration ( Figure 5A-C). Additionally, RA did not affect the expression of CTSV and ALDH1A2 but reversed the effect of ad-CTSV on P21, p-ERK1/2, and p-AKT ( Figure 5D).  To further confirm the effect of CTSV in cellular senescence, a doxorubicin (DOX)induced senescence model was used. Administration of DOX induced cellular senescence and impaired cell proliferation and migration. Simultaneously, it downregulated CTSV expression and upregulated ALDH1A2, RA, p-AKT, p-ERK1/2, and P21. Ad-CTSV reversed the effect of DOX, while further administration of RA canceled the function of ad-CTSV. However, ad-CTSV or RA did not affect the expression of P16 increased by DOX. Moreover, DOX, ad-CTSV, and RA had little influence on the expression of CTSL and P53 (Figures 6,  S1E and S4A).  Furthermore, siRNAs were used to verify the causality of downregulation of CTSV in endothelial senescence. Si-CTSV had a similar effect on ECs as DOX. Additionally, inhibition of ALDH1A2 by siRNAs reversed the effect of si-CTSV and DOX (Figures 7, S1F  and S4B). These results indicate that CTSV attenuates cellular senescence via ALDH1A2-AKT/ERK1/2-P21 pathway.  . Data are presented as mean ± SEM. One-way ANOVA test was used. * p < 0.05, ** p < 0.01, *** p < 0.001.

Discussion
This study provides new insights into the critical role of Ctsl/CTSV in cellular senescence. The main findings are as follows: 1. The expression of Ctsl/CTSV decreased in senescent aorta and endothelial cells; 2. over-expression of CTSV reduced cellular senescence in vitro; 3. CTSV attenuated cellular senescence by enhancing ALDH1A2 expression and activating AKT/ERK1/2 pathway (Figure 8). We firstly revealed that Ctsl/CTSV was decreased in senescent endothelial cells, which enhanced ALDH1A2 expression and activated AKT/ERK1/2-P21 pathway, and therefore promoted cellular senescence, which may play an important role in vascular aging.

Discussion
This study provides new insights into the critical role of Ctsl/CTSV in cell senescence. The main findings are as follows: 1. The expression of Ctsl/CTSV decre in senescent aorta and endothelial cells; 2. over-expression of CTSV reduced cel senescence in vitro; 3. CTSV attenuated cellular senescence by enhancing ALDH expression and activating AKT/ERK1/2 pathway (Figure 8). We firstly revealed Ctsl/CTSV was decreased in senescent endothelial cells, which enhanced ALDH expression and activated AKT/ERK1/2-P21 pathway, and therefore promoted cel senescence, which may play an important role in vascular aging. Endothelial senescence is the crucial step of vascular aging leading to advancement of CVD [7]. To access senescence in vitro, SA-β-gal staining was used in study, as well as cellular functions including proliferation and migration. Consistent previous studies [13,24], increased SA-β-gal positive cells and decreased cell prolifera and migration abilities were observed in H2O2-induced senescent cells. The expressio P53, P16, and P21 has been widely recognized as markers of senescence, which demonstrated to regulate cellular senescence alone or jointly [35]. p-Coumaric ac reported to suppress cellular senescence via P53 and P16 pathways [36], while ERK1 involved in H2O2-induced senescence through P53-independent P21 pathway [29]. In present study, P21 and P16 were increased in senescent HUVECs, while P53 was involved in H2O2-induced cellular senescence. In addition, over-expression of C reduced the expression of P21 but not P16, suggesting that CTSV reduced cel senescence via P21-dependent pathways. Senescent cells may secrete many factors, as cytokines, chemokines, metalloproteases and extracellular vesicles, to influence around them, which is called SASP [13]. Some factors were commonly found upregul in different senescent cells, including IL-1β, IL-6, IL-8, ICAM-1, etc. [37]. This study fo Endothelial senescence is the crucial step of vascular aging leading to the advancement of CVD [7]. To access senescence in vitro, SA-β-gal staining was used in this study, as well as cellular functions including proliferation and migration. Consistent with previous studies [13,24], increased SA-β-gal positive cells and decreased cell proliferation and migration abilities were observed in H 2 O 2 -induced senescent cells. The expression of P53, P16, and P21 has been widely recognized as markers of senescence, which are demonstrated to regulate cellular senescence alone or jointly [35]. p-Coumaric acid is reported to suppress cellular senescence via P53 and P16 pathways [36], while ERK1/2 is involved in H 2 O 2 -induced senescence through P53-independent P21 pathway [29]. In the present study, P21 and P16 were increased in senescent HUVECs, while P53 was not involved in H 2 O 2 -induced cellular senescence. In addition, over-expression of CTSV reduced the expression of P21 but not P16, suggesting that CTSV reduced cellular senescence via P21-dependent pathways. Senescent cells may secrete many factors, such as cytokines, chemokines, metalloproteases and extracellular vesicles, to influence cells around them, which is called SASP [13]. Some factors were commonly found upregulated in different senescent cells, including IL-1β, IL-6, IL-8, ICAM-1, etc. [37]. This study found that IL-1β, IL-6, and ICAM-1 were upregulated in DOX-induced senescent ECs, indicating the existence of SASP. Furthermore, previous studies suggested that the physiological function of ECs was impaired during cellular senescence. Li et al. found that H 2 O 2 significantly inhibited the migration and proliferation of HUVECs [38]. Similarly, the proliferation and migration ability of HUVECs also decreased in Trimethylamine-N-oxide-induced senescent cells [16]. In this article, impaired cell proliferation and migration were also observed in both H 2 O 2 -and DOX-induced senescent ECs. The secretion of metalloproteases by senescent cells may take response to the disability of cell migration [39], which remains to be further studied.
As a pivotal member of cathepsin family, Ctsl is reported to express in endothelial progenitor cells and plays an essential role in retinal and choroidal neovascularization [20], indicating Ctsl may involve in EC functions, while its role in ECs senescence remains elusive. Ctsl increased in the renal tissue of aged rat as well as the muscle mass of aged mice [40,41], while in the rat brain tissues, the expression and activity of Ctsl decreases with age [42]. In this study, we analyzed the expression of Ctsl in aged thoracic aortas and found that Ctsl obviously decreased in senescent aorta, which was consistent with a previous study [43].
Human CTSV evolved from mouse Ctsl, and they have 80% homology [15]. Since CTSV is specifically expressed in human, the role of CTSV in senescence is still unknown. Previous studies showed that CTSV is highly expressed in endothelial cell [44], while its role in endothelial cells and vascular tissues is not clear. In addition, CTSV decreased in the skin tissue of aged patients [21], suggesting that the expression of CTSV may participate in the process of age. Here in our study, we found that CTSV decreased in senescent HUVECs, while over-expression of CTSV reduced cellular senescence, and enhanced cell proliferation and migration. These results indicated that CTSV plays a crucial role in ECs senescence.
Among the 19 members of the aldehyde dehydrogenase super family, aldehyde dehydrogenase 1 family member A2 (ALDH1A2, also known as RALDH2) is one of the most widely distributed members in human tissues [45]. It is reported that ALDH1A2 increases in the hippocampus of senescent female mice [46]. In contrast, Takano et al. demonstrate that ALDH1A2 decreases with age in mesenteric lymph node dendritic cells [47]. In our work, we firstly found that ALDH1A2 was increased in H 2 O 2 -induced senescent endothelial cells by using RNA sequencing and confirmation with RT-PCR and WB, while it decreased after over-expression of CTSV, which suggested the significance of ALDH1A2 in the regulation of cellular senescence by CTSV. As an enzyme, the main function of ALDH1A2 is to catalyze the synthesis of RA from retinaldehyde. Isotretinoin (13-cis-RA), the isoform of RA, has been widely applied in the treatment of severe, recalcitrant acne. However, Bershad et al. noticed that dyslipidemia occurs during isotretinoin therapy for acne, including increased cholesterol and low-density-lipoprotein cholesterol and decreased high-density-lipoprotein cholesterol in plasma [48]. In addition, plasma homocysteine increases significantly after isotretinoin treatment [49]. These changes would increase the risk of CVD if sustained over a long period, indicating the potential unfavorable effects of RA on cardiovascular system. It has been reported that RA decreases in ALDH1A2 knockout endothelial cells, leading to decreased p21 expression, while administrating RA reverses this effect [50]. Furthermore, RA can upregulate p21 to induce cellular senescence via ERK1/2 and AKT pathways in neuroblastoma cells [34]. In the present study, RA was added to CTSV overexpressed senescent ECs. We found that administration of RA significantly activated AKT and ERK1/2, induced the expression of P21, and enhanced ECs senescence. Our results suggested that CTSV may attenuate ECs senescence via ALDH1A2-AKT/ERK1/2-P21 pathway.
As for the limitations, studies on over-expressed Ctsl and genetic knockdown mice should be conducted to further validate our results. Additionally, the regulatory mechanisms of CTSV on ALDH1A2 should be further studied.

Conclusions
We have demonstrated that decreased Ctsl/CTSV is associated with vascular aging in mice. Moreover, our results suggest that decreased Ctsl/CTSV induces cellular senescence through ALDH1A2-AKT/ERK1/2-P21 pathway. Thus, Ctsl/CTSV may be a potential therapeutic target for EC senescence.