In Vitro Models of Cell Senescence: A Systematic Review on Musculoskeletal Tissues and Cells

Ageing is an irreversible and inevitable biological process and a significant risk factor for the development of various diseases, also affecting the musculoskeletal system, resulting from the accumulation of cell senescence. The aim of this systematic review was to collect the in vitro studies conducted over the past decade in which cell senescence was induced through various methods, with the purpose of evaluating the molecular and cellular mechanisms underlying senescence and to identify treatments capable of delaying senescence. Through three electronic databases, 22 in vitro studies were identified and included in this systematic review. Disc, cartilage, or muscle cells or tissues and mesenchymal stem cells were employed to set-up in vitro models of senescence. The most common technique used to induce cell senescence was the addition to the culture medium of tumor necrosis factor (TNF)α and/or interleukin (IL)1β, followed by irradiation, compression, hydrogen peroxide (H2O2), microgravity, in vitro expansion up to passage 10, and cells harvested from damaged areas of explants. Few studies evaluated possible treatments to anti-senescence effects. The included studies used in vitro models of senescence in musculoskeletal tissues, providing powerful tools to evaluate age-related changes and pathologies, also contributing to the development of new therapeutic approaches.


Introduction
Human ageing is an irreversible and inevitable biological process characterized by the physiological deterioration of physical functions and an increasing of pathological conditions starting from the third decade of age [1,2].In the Western population, the number of people over 60 years old is increasing, and it is forecasted to triple by 2050 [3].
Ageing is a significant risk factor for the development of various diseases, also affecting the musculoskeletal system, with factors such as intervertebral disc degeneration (IDD), osteoarthritis (OA), osteoporosis (OP), and sarcopenia.These pathologies result from the accumulation of cell senescence and degenerative stimuli, both from inside the cell and from the microenvironment, contributing to progressive tissue degeneration [4].Cell senescence is defined as a permanent state of cell cycle arrest in which cells become resistant to mitogenic stimuli.Senescence can act as a physiological process occurring throughout the lifespan, which prevents the replication of cells with damaged DNA, playing an important role during development and wound healing and providing an antitumorigenic function [4].
During ageing, an imbalance arises between the production of reactive oxygen species (ROS) and the ability to repair tissue damage through endogenous antioxidant defenses [5], causing an increase in the oxidative stress.This oxidative stress, in turn, adversely affects cell function (causing damage to proteins, nucleic acids, and lipids), reduces the tissue repair capability, and induces cellular senescence.Additionally, senescent cells prompt neighboring cells to undergo senescence through the paracrine signaling pathway [6].
Two modalities of induction of cellular senescence are known: the replicative senescence and the stress-induced premature senescence (SIPS).Cell senescence is caused, in the first case, by telomere attrition after repeated proliferation while, in the second case, by subtoxic doses of extrinsic or intrinsic stimuli such as irradiation, oxidative stress, drugs, oncogenic stress, and metabolic or epigenetic changes [7].
Recent hypotheses suggest that eliminating senescent cells, possibly using chemical agents like senolytic compounds, can improve tissue homeostasis and delay age-associated pathologies [12].
Over the past decade, in vitro models that simulate specific pathologies or scenarios have gained prominence.These models align with the 3R (reduction, refinement, and replacement) principles for animal use for scientific purposes because they are performed using advanced in vitro systems rather than in vivo ones [13].Actually, several in vitro culture systems reproduce the characteristics of many pathologies, including complex conditions like cancer [14].
In the literature, some in vitro models are usually employed to induce cell or tissue senescence: (1) primary cells cultured in monolayers under senescence-inducing conditions; (2) three-dimensional (3D) tissue models (such as organoids, spheroids, and tissueengineered constructs), which better replicate the native tissue microenvironment compared to cell cultures [15]; (3) cultures in bioreactors simulating dynamic conditions such as mechanical loading or fluid shear [16]; (4) co-culture models involving multiple cell types to study their mutual influence [17]; and (5) genetically modified cells or tissues [18].
The aim of this systematic review was to collect the in vitro studies conducted over the past decade in which cell senescence was induced through various methods.The objective was to evaluate the molecular and cellular mechanisms underlying senescence and to identify treatments capable of delaying senescence.
This review investigates different techniques to induce senescence in musculoskeletal cells and tissue cultures, as well as the in vitro methodologies used to assess cell senescence.The perspective of our review is to encourage further research exploring the characteristics of senescent cells and potential therapeutic treatments in this context.

Eligibility Criteria
For the selection and analysis of the relevant papers, a PICO question (Population of interest (P), Intervention (I), Comparators, and Outcomes (CO)) was formulated.
The Population considered was in vitro studies in which musculoskeletal system cells or tissues were employed.The Intervention was all in vitro cultures in which cell or tissue senescence was induced.The Comparator was cells cultured in normal conditions, without the induction of senescence.The considered primary Outcome was cell senescence evaluation through cellular, molecular, biochemical, and histological techniques.In addition, a secondary outcome was the effect of some treatments used to reduce cell senescence when present.

Search Strategy
The search was performed on 15 July 2023 and included research published from 15 July 2013 to 15 July 2023 according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.
Three electronic databases (PubMed, Scopus and Web of Science) were used to identify relevant papers using the following MESH with Boolean operators: (("In Vitro Models"[Mesh]) AND ("Aging"[Mesh] OR "Cellular Senescence"[Mesh])) AND ("Musculoskeletal System"[Mesh]).The limit was the language (English) for all the databases.
After the removal of duplicates (Mendeley 1.14, www.mendeley.com,accessed on 17 July 2023), relevant articles were screened using the title and abstract by two reviewers (FV and DC), and articles that did not meet the inclusion criteria, previously reported, were excluded.The included full-text articles were retrieved and reviewed by the two authors, and any disagreement was resolved through discussion until consensus was reached or with the involvement of a third author (VB).The researchers involved in the process of reviewing the papers used an Excel spreadsheet to independently perform the screening and data extraction.The following information was extracted from each paper to summarize the evidence reported in each study: (a) cells/tissues, (b) in vitro model of senescence, (c) evaluations, (d) main results, and (e) references (Ref.)(Table 1).

Results
The initial literature search retrieved 151 studies from PubMed, 38 from Scopus, and 23 from Web of Science, for a total of 212 articles.There were 186 identified papers after the removal of duplicates (26 records).Reading the titles and abstracts, 18 papers were excluded, and 168 full texts were evaluated for inclusion.Reviews or book chapters (n = 10), in vivo studies (n = 10), and non-inherent studies because they employed cells from other tissues different from musculoskeletal ones, or because the authors did not induce cell senescence (n = 132), were excluded, for a total of 152 studies.The remaining 16 articles were considered eligible.A further search was performed by reading the reference lists of these eligible articles, and a further six papers were added.Therefore, a total of 22 articles were included in the present systematic review (Figure 1).

Results
The initial literature search retrieved 151 studies from PubMed, 38 from Scopus, and 23 from Web of Science, for a total of 212 articles.There were 186 identified papers after the removal of duplicates (26 records).Reading the titles and abstracts, 18 papers were excluded, and 168 full texts were evaluated for inclusion.Reviews or book chapters (n = 10), in vivo studies (n = 10), and non-inherent studies because they employed cells from other tissues different from musculoskeletal ones, or because the authors did not induce cell senescence (n = 132), were excluded, for a total of 152 studies.The remaining 16 articles were considered eligible.A further search was performed by reading the reference lists of these eligible articles, and a further six papers were added.Therefore, a total of 22 articles were included in the present systematic review (Figure 1).In the studies included in this review, nucleus pulposus (NP), annulus fibrosus (AF) cells or discs [19][20][21][22][23][24][25][26][27][28][29][30], chondrocytes [31][32][33][34][35][36], myoblasts [2,37,38], and mesenchymal stem cells from bone marrow (BMSCs) [36] or from NP (NPMSCs) [39] were employed to setup in vitro models of senescence.
More precisely, the addition of both TNFα (20 ng/mL) and IL1β (20 ng/mL) to the culture medium increased the ROS, SA-β-Gal, p16, p53, and degrading enzymes' gene expressions and reduced cell proliferation, TE activity, and matrix proteins.In addition, this combination also altered the distribution of cells across cell cycle phases; the percentage of cells in G0/G1 phase was higher than that in S phase [19].
Another study focused solely on TNFα, using concentrations of 10 ng/mL for cells and 200 ng/mL for disc cultures.The authors observed reduced cell viability, TE activity, and matrix gene expressions and increased SA-β-Gal activities, ROS levels, and p53 and p16 gene expressions [20].
A study tested different TNFα concentrations (10, 20, and 40 ng/mL) in cell culture, observing dose-dependent decreases in cell proliferation, TE activities, and matrix proteins, as well as dose-dependent increases in p53, p16, and AKT gene expressions.Similar results were observed in groups treated with TNFα for 3 days with 3 days of recovery or with six consecutive days of treatment [21].
Finally, a 3D co-culture model was set-up with NP cells, cultured in presence of TNFα (20 ng/mL), and calcium alginate gel balls, empty or containing normal BMSCs or NP cells.
The addition of TNFα increased SA-β-Gal activities in NP cells, whereas the addition of balls with BMSCs reversed the senescence phenotype.This intervention reduced MMP9 protein expression and p16, p21, and p53 gene expressions and increased zinc metallopeptidase STE24 (ZMPSTE24) protein levels [22].
Discs were loaded in bioreactors with intermittent compression at two different intensities (0.1 and 1.3 MP), showing higher SA-β-gal activities, elevated expressions of p16, p21, p53, catabolic enzymes, and p38 proteins, and reduced matrix proteins and tissue inhibitors of metalloproteinase TIMP3 gene expression at 1.3 MP than at 0.1 MP [23].
One study compared static and dynamic compression (0.4 MP, 1.0 Hz for 4 h/day for 7 days) in disc culture, observing that static compression led to greater SA-β-gal, p16, and p53 protein expressions and reduced TE activity and matrix proteins compared to dynamic compression.Notably, dynamic compression seemed to revert the senescent phenotype [25].
Moreover, two different magnitudes (5% and 20% elongation) of mechanical tension (1 Hz for 8 h/day for 12 days) were applied to AF cells.Cell proliferation, TE activity, S phase, matrix, and BCL1 and LC3 proteins were significantly reduced, whereas G0/G1 phase and p16 and p53 proteins were significantly increased, with 20% more elongation than with the 5% one [26].

Hydrogen Peroxide
In two studies, hydrogen peroxide (H 2 O 2 ) added to NP cells at different concentrations (from 0.05 to 2 mM) increased ROS levels, SA-β-Gal activities, p53, p21, and p16 protein expressions, and pro-inflammatory cytokines' gene expressions, accompanied by reduced cell viability and proliferation in a dose-dependent manner [27,28], with significantly detrimental effects at concentration of 0.5 mM [28].

Irradiation
Finally, two studies irradiated human and rat NP or AF cells with γ-rays ( 60 Co gamma at a rate of 2 Gy/min or 2.5 Gy/min) [29,30].In the first study, irradiation reduced cell proliferation and increased SA-β-Gal and GL13 protein levels.This effect was more conspicuous in cells harvested from old rats or patients compared to those from young rats or healthy subjects [29].
In the second study, irradiations were given to cells cultured in two different media: the first composed by 10% fetal bovine serum (FBS), 4.5 mg/mL glucose, 300 mOsm/Kg H 2 O, and 20% O 2 , whereas the second consisted of serum-free 0.9 mg/mL glucose, 400 mOsm/Kg H 2 O, and 2% O 2 .Both media increased p16, p21, and MMP gene expressions and reduced collagen II (COLL II) gene expression.Additionally, the first medium also increased SA-β-Gal activity and reduced cell proliferation [30].
Furthermore, a study compared the effects of monoenergetic C-ions (75-95 MeV/mm) with X-rays (225 kV) on chondrocytes cultured in a monolayer or in a collagen scaffold [31].The authors observed that C-ions significantly reduced chondrocyte proliferation more than X-rays in monolayer cultures.In both monolayer and in collagen scaffold cultures, a higher increase in SA-β-gal expression was observed with C-ions than with X-rays [31].

Isolation from Damaged Cartilage
Finally, micromasses of chondrocytes harvested from severely damaged cartilage areas displayed lower cartilage matrix depositions, telomere lengths, matrix gene expressions, and higher MMP13, p16, p21, and p53 protein levels than chondrocytes isolated from preserved cartilage areas [36].

Serum from Elderly Patients
Myotubes derived from C2C12 cells were treated with 10% serum from older donors, showing lower myotube diameters and decreased nuclear factor I (NFI), AKT, p70S6K1 and eukaryotic translation elongation factor 2 (eEF2) proteins than myotubes cultured in sera from younger subjects [37].Similar results were obtained when the same cells were cultured in the presence of 5% plasma from older subjects, showing higher scratch sizes and lower cell diameters and proliferation rates than cells cultured with 5% plasma from younger subjects [2].

Microgravity
Finally, human myoblasts at the first passage were cultured under normal gravity (1 G), microgravity (1 × 10 −3 G; µG), or normal gravity at the fifth passage.Microgravity decreased myosin heavy chain (MHC) and desmin proteins, cell proliferations, and myotubes' diameters, whereas it increased the myogenic factor 5 (Myf5) protein.Both microgravity exposure and the fifth in vitro passage led to increased SA-β-Gal activity [38].
In the context of senescent NP cells, treatment with resveratrol (at concentrations of 30, 60, or 100 µM) was administered.This intervention resulted in reductions in ROS, SA-β-gal activities, expressions of p16, p53, catabolic enzymes, and NF-κB proteins, as well as a decrease in cells present in the G0/G1 phase.Additionally, resveratrol increased cell proliferation, TE activity, matrix proteins, and cells in the S phase [19,24].
Finally, senescent chondrocytes were cultured in the presence of 100 mM 4-phenylbutyric acid (4-PBA).The authors observed a decrease in GRP78 protein levels, SA-β-gal activity, cell apoptosis, and an increase in cell viability after treatment [34].

Discussion
This systematic review collected and analyzed in vitro studies from the literature that evaluated the existing in vitro models used to induce senescence in musculoskeletal cells and tissues over the last 10 years.
Generally, an in vitro environment can provide information about cell-cell and cellextracellular matrix interactions, expressions of surface receptors, and syntheses of proteins and growth factors.By increasing the in vitro model's complexity, we are able to represent and recapitulate the characteristics of many pathologies through the use of cells or tissues from affected subjects.Moreover, complex in vitro models are functional for the gathering of preliminary data on the safety and cytocompatibility but also on the bioactivity and therapeutic efficacy of newly developed materials, medical devices, or therapies [40].
bioactivity and therapeutic efficacy of newly developed materials, medical devices, or therapies [40].
Primary human cells were collected from IVD or cartilage of healthy subjects [28,29,34], from cartilage of patients affected by OA [36] or cadavers [32], and from patients with lumbar disc herniation [39].The range of ages of these subjects was from 28 to 73 years, even if some studies did not indicate the age.
TNFα and IL1β are the two inflammatory cytokines often used to recreate an inflammatory microenvironment in some in vitro models of tissue degeneration, including OA and disc degeneration [41][42][43].In addition, according to the literature, the two cytokines induce cell senescence in some cell types, concluding that inflammation response-mediated cell senescence plays an important role in accelerating tissue degeneration [44] and can provoke an oxidative stress that drives cells to premature senescence.However, as the inflammatory factors were maintained throughout the experiment, the inflammation environment may exhibit a direct effect but not a long-term effect.In addition, the concentration of inflammatory cytokines used in in vitro senescence models is remarkably higher than that in human knee joints.Primary human cells were collected from IVD or cartilage of healthy subjects [28,29,34], from cartilage of patients affected by OA [36] or cadavers [32], and from patients with lumbar disc herniation [39].The range of ages of these subjects was from 28 to 73 years, even if some studies did not indicate the age.
Their use to induce senescence in cells is based on ionizing radiation, such as γ-irradiation, which is a principal DNA-damaging agent [30], and C-ion irradiation, which induces healthy organ deterioration [45].On the other hand, the biologic effects measured in 2D cultures might be overexaggerated compared with the reality of clinics.
Then, intermittent, dynamic, or static compressions of discs and NP cells [23][24][25] and elongations of AF cells [26] were also employed.Indeed, it was observed that the various magnitudes of mechanical load, to which IVD was subjected every day during the daily activities, affected disc cell viability and biological function [46].The non-physiological loads could accelerate disc degeneration [47], and the excessive and continuous mechani- TNFα and IL1β are the two inflammatory cytokines often used to recreate an inflammatory microenvironment in some in vitro models of tissue degeneration, including OA and disc degeneration [41][42][43].In addition, according to the literature, the two cytokines induce cell senescence in some cell types, concluding that inflammation response-mediated cell senescence plays an important role in accelerating tissue degeneration [44] and can provoke an oxidative stress that drives cells to premature senescence.However, as the inflammatory factors were maintained throughout the experiment, the inflammation envi-ronment may exhibit a direct effect but not a long-term effect.In addition, the concentration of inflammatory cytokines used in in vitro senescence models is remarkably higher than that in human knee joints.
Their use to induce senescence in cells is based on ionizing radiation, such as γirradiation, which is a principal DNA-damaging agent [30], and C-ion irradiation, which induces healthy organ deterioration [45].On the other hand, the biologic effects measured in 2D cultures might be overexaggerated compared with the reality of clinics.
Then, intermittent, dynamic, or static compressions of discs and NP cells [23][24][25] and elongations of AF cells [26] were also employed.Indeed, it was observed that the various magnitudes of mechanical load, to which IVD was subjected every day during the daily activities, affected disc cell viability and biological function [46].The non-physiological loads could accelerate disc degeneration [47], and the excessive and continuous mechanical compression may have aggravated the cellular senescence of the discs.Only one study that evaluated elongation in the induction of cell senescence used AF cells [26].Tears and fissures of AF tissues were related with aggravations of disc degeneration [48], and the mechanical tension was considered a reason for AF structural destruction.However, the signaling pathways involved in the mechanical induction of senescence are not yet well known, and this technique is limited to a specific setting through the use of the appropriate bioreactor and appropriate stimuli.
Other authors cultured NP cells with H 2 O 2 at different concentrations [27,28] and myotubes or myoblasts with sera or plasma obtained from old individuals [2,37].
Indeed, H 2 O 2 is a commonly used inducer of stress-induced premature senescence and DNA damage, given its role as a natural inducer of oxidative stress [49].However, H 2 O 2 treatment creates a short-period senescent phenotype, which leads to rapid apoptosis and cell death in days and easily triggers cell apoptosis rather than cell senescence.
The ageing process is associated with changes in circulating endocrine factors, such as growth factors hormones and pro-inflammatory cytokines, which are known to moderate muscle mass or the Klotho factor [50].Unfortunately, this type of model cannot be standardized, as it incurs the bias that the sera of the donors have different amounts of the above-mentioned circulating factors.
One study employed microgravity on myoblasts [38] because microgravity causes a remarkable reduction in muscle mass and motor skills of astronauts [51].However, this type of model is difficult to be replicated, thus providing controversial results.Furthermore, it seems that it is more a model of disuse than of senescence.
A study proposed to create a senescence phenotype in BMSCs by an extensive in vitro expansion up to passage 10 [36], the so called "replicative senescence" [36], and another one using cultures of chondrocytes isolated from damaged areas of osteoarthritic cartilage explants, as OA shares some common mechanisms to senescence, such as low-grade inflammation and telomere shortening [36].In contrast to the previous in vitro models that resulted in rapid apoptosis and cell death, the model of "replicative senescence" induced cell senescence without the addition of external agents.Also, this model showed a limit because the excessive cell division may not represent the actual reason causing cell senescence, and the "replicative senescence" is just one type of reasons leading to senescence.
To summarize, the above-mentioned in vitro models of senescence can be essentially divided into two groups: stress-induced premature senescence and replicative senescence.Irradiation, mechanical stimuli, the addition of H 2 O 2 or sera, and microgravity are part of the first group, whereas the addition of inflammatory cytokines and several in vitro passages expansion belong to the second group.
TE is a protein that adds a telomere repeat sequence to the 3 end of telomeres, regions of repetitive sequences at each end of the chromosomes that protect the end of the chromosome from DNA damage.During aging, TE activity is reduced, causing a shortening of telomere [56].Its activity was evaluated through enzyme-linked immunosorbent assay (ELISA) in these studies.In regard to ROS, they induce oxidative damage to proteins, DNA, and lipids, promoting cell senescence, tissue destruction, and inflammation.They directly affect the rate of telomere shortening, further accelerating cell senescence [57].
As the phenotype of senescent cells displays variability, a single unambiguous biomarker for the identification of senescent cells has not been recognized yet.A combination of multiple biomarkers is usually accepted to distinguish between senescent and quiescent or terminally differentiated cells, reflecting the dynamicity of the senescence process.In addition, some biomarkers are not specific to senescent cells and can be present also in apoptotic cells or quiescent cells, for example.This represents a limitation in the isolation and characterization of senescent cells and a challenge for future research.
The purpose of performing in vitro studies with senescence-induced cells is to find and test agents with the ability to counteract senescence.
Resveratrol is a natural phenol produced by several plants in response to injury, such as attack of bacteria or fungi, and shows protective effects, including anti-inflammatory and anti-aging properties [58].In the analyzed studies, resveratrol reduced ROS production, SA-β-Gal activity, p16 and p53 proteins' production, and increased cell proliferation in senescent NP cells [19,24].
As NP cells contain estradiol receptor (ER) [59], it was observed that cell proliferation and TE activity increased, whereas SA-β-Gal activity, ROS levels, and p53 and p16 proteins decreased after exposure to E2 [20].
Finally, 4-PBA, which alleviates ER stress in both cell lines and animal models [60], reduced apoptosis and SA-β-Gal activity and increased cell viability in senescent chondrocytes [34].
These in vitro models provide a controlled and reproducible environment to investigate the cellular and molecular changes associated with senescence, ultimately contributing to our understanding of age-related musculoskeletal diseases and potential therapeutic interventions.However, it is important to note that in vitro models have their limitations, and findings from these models must be further validated using in vivo studies for translations to human health.
Indeed, studying senescence in musculoskeletal tissues and cells using in vitro models has still some limitations, which can impact the translatability and relevance of the findings to the complex in vivo environment.Some of these limitations include the lack of: (1) tissue complexity, which can lead to oversimplification of the biological processes involved in senescence; (2) cellular heterogeneity; (3) mechanical loading; (4) immune system interactions; and (5) standardized protocols.
In addition, a short-term culture may not capture the long-term effects and progression of senescence observed in aging musculoskeletal tissues and may not fully capture the genetic and epigenetic changes that occur during aging and senescence in vivo.
Many senescence studies use immortalized cell lines, which may not fully represent the senescence process in primary cells.Using primary cells more often can provide more physiologically relevant insights into senescence mechanisms.
Developing 3D models for senescence studies can provide more physiologically relevant data, and the combination of genomics, transcriptomics, proteomics, and metabolomics can provide a comprehensive view of senescence.
While in vitro models are valuable, they should be validated in animal models to confirm the relevance of findings in vivo.This step is critical for translating laboratory discoveries into potential clinical applications.
However, in vivo and clinical studies have their advantages and limitations, and they complement each other in aging research.The advantages in the in vivo models concern complexity, the possibility to perform longitudinal studies, and genetics, whereas the limitations regard species differences, ethical concerns, and high costs and time consumption.The advantages of human aging studies are their direct relevance, ethical considerations, and the possibility to perform clinical observations, but the main limitations are the ethical and practical constraints; the fact that long-term human aging studies can be time consuming, expensive, and subject to dropouts; and genetic diversity.
In summary, in vitro models are useful for controlled experiments and initial insights, animal models allow for more complex biological contexts and genetic manipulation, and human studies are essential for understanding human-specific aging processes and interventions.A combination of these approaches, along with advanced computational modeling, can provide a more comprehensive understanding of the aging process and improve the development of strategies to promote healthy aging.

Conclusions
Studies included in the present systematic review used in vitro models of senescence in musculoskeletal tissues, providing powerful tools to analyze the complex processes of aging in musculoskeletal tissues and to provide valuable insights into age-related changes and pathologies, ultimately contributing to the development of new therapeutic approaches to mitigate the effects of aging on musculoskeletal health.
The cells and tissues that were mostly investigated were those obtained from IVD, followed by cartilage and muscle tissues.

Figure 1 .
Figure 1.Flow-chart of the included studies according to PRISMA principles.

Figure 1 .
Figure 1.Flow-chart of the included studies according to PRISMA principles.

Figure 2 .
Figure 2. Percentage of each cell and tissue type employed in the studies of the review.

Figure 2 .
Figure 2. Percentage of each cell and tissue type employed in the studies of the review.

Figure 3 .
Figure 3. Percentages of the different methods to induce cell or tissue senescence in vitro.

Figure 3 .
Figure 3. Percentages of the different methods to induce cell or tissue senescence in vitro.

Table 1 .
Summary of the results of the included studies.

Table 2 .
Summary of anti-senescence treatments of the included studies.