Telomere-induced cell senescence is a component of replicative senescence. As the telomeres of normal cells shorten, the cells eventually undergo a stagnation of proliferation or division and reach senescence. By contrast, tumor cells often exhibit unlimited proliferative capacity, because the length of telomeres in tumor cells is stable (Fig. 1) (44,50,51).

Telomerase is a reverse transcriptase, the main function of which is to add the repeat base sequence TTAGGG to the end of chromosomes to increase the length of telomeres and the number of cell divisions (9,52). One study demonstrated that the telomeres of human primary fibroblasts shorten when the cells lose their proliferative capacity (53). However, the cell senescence caused by telomere shortening can be prevented via the activation of telomerase (51). Telomerase serves an important role in the process by which cells escape senescence, as its activation can stabilize the length of telomeres and even prolong the life of tumor cells. Among the 100 immortalized cell lines tested by Kim et al (54), telomerase was highly expressed in 94 tumor-derived cell lines.

Due to the complexity of telomerase, various strategies for its inhibition have been developed as potential treatments for cancer. These include the use of antisense oligonucleotides to target the RNA component of telomerase, chemical telomerase inhibitors, oligonucleotides and nucleosides, small-molecule drugs that target human telomerase reverse transcriptase (TERT), gene therapies, and molecules that target telomeres and telomerase-related proteins (17,55–59). Among these, a non-competitive inhibitor of TERT, BIBR1532, has been demonstrated to shorten telomere length, inhibit cell proliferation and induce senescence in human cancer cells (60). The aforementioned studies have shown that telomerase dysfunction can induce cell senescence, damage organ functions and shorten the human life span. However, the process of senescence may be reversed by inhibiting telomerase activation (61).

The levels of active oxygen free radicals within living organisms increases over time (62). When free radicals are present in excessive quantities or the antioxidant capacity is inadequate, oxygen free radicals oxidize unsaturated fatty acids in cells, causing lipid peroxidation and biofilm damage, thus damaging the structure and function of organelles (63). The effects of ROS on cells are positively associated with their concentration. ROS attack mitochondrial DNA and accelerate the process of tumor cell senescence. The premature senescence of human fibroblasts occurs in a high-oxygen environment (40-50% O₂), while the cell cycle is extended in hypoxic conditions (2-3% O₂) (64,65). ROS intervene in tumor senescence through processes such as lipid peroxidation, DNA damage and protein destruction (66). Busulfan has been demonstrated to create DNA-DNA and DNA-protein cross-links, and induce the senescence of human fibroblasts in a ROS-dependent manner (67).

When tumor cells are exposed to abnormal external conditions but are unable to activate their auto-repair mechanisms, they may undergo senescence or apoptosis (68). Numerous types of antitumor treatment, including cytotoxic chemotherapy drugs, ionizing radiation and topoisomerase inhibitors, are DNA-damaging agents that can induce the senescence of tumor cells and as well as healthy cells (69–72). It has been shown that chemotherapeutic compounds, such as doxorubicin, etoposide and cisplatin, which cause double or single-strand breakages of DNA, can cause healthy human fibroblasts to become senescent prematurely (73,74).

The activation of oncogenes in mammalian cells results in proliferative stress and the induction of senescence, which limits tumor growth. Genetic mutations can occur at any stage of development, usually during DNA replication or the interphase of cell division, and may affect DNA replication, DNA damage repair, carcinogenesis and senescence (75). Severe deficiencies in proteins that contribute to the sensing of DNA damage and its repair have the potential to accelerate senescence, while milder mutations in these same pathways may predispose individuals to develop cancer (76,77). In young human fibroblasts, mutation of the Ras gene has been shown to induce cell cycle arrest in the G₁ phase, a senescence-like phenotype and SA-β-Gal expression (78,79). Thus, senescence is a physiological mechanism of tumor suppression that inhibits the progression from benign tumor lesions to malignant tumors. In addition to the aforementioned factors, certain cellular pathways are also able to induce cell senescence.

Genes such as p53, p21 (CDK inhibitor 1A) and Rb play important roles in cell senescence, and determine whether cells are apoptotic or senescent (71). The ATM gene is an important component of the DNA damage checkpoint pathway, which is critical in cell cycle regulation, DNA damage response and repair (80). The loss of ATM leads to telomere shortening and damage. When DNA is damaged, ATM is activated, and induces the phosphorylation of downstream proteins such as checkpoint kinase 1 (Chk1), Chk2 and cell division cycle. This regulates the cell cycle checkpoints, so that DNA damage is repaired (81).

The p53 gene is important in cell carcinogenesis, senescence, apoptosis and gene repair (82). When cells are exposed to external stimuli that induce DNA damage, such as chemotherapy drugs, ionizing radiation, gene mutation or telomerase shorting, the p53 gene is activated (83). p53 activates p21, which inhibits CDK1, causing cells to be arrested in the G₂/M phase and decreasing the phosphorylation level of pRb by inhibiting the activity of CDK2 and CDK4. This prevents cells from entering the S phase, and ultimately leads to cell senescence (Fig. 2). In one study, mice that were genetically engineered to express altered isoforms of p53 with increased activity were shown to be resistant to cancer (84). In human and mouse cells, inactivation of the p53-p21 pathway leads to cells escaping senescence (85). The knockdown of p53 or p21 reduced the drug-induced senescence of HCT116 cell line several folds, and the same phenomenon was observed when p53 expression was suppressed in HT1080 fibrosarcoma cells (86,87). Genome-wide analysis has demonstrated that the loss of p16^INK4a expression and/or p53 function is the most common genetic event in human cancers, and may enable cancers to evade senescence (88).

P16^INK4a affects the process of senescence in tumor cells and is recognized as a tumor suppressor and biomarker of senescence (89). p16^INK4a is an upstream regulator of pRb, which accelerates the degradation of phosphorylated pRb, activates the E2 factor (E2F) family of transcription factors, arrests cells at the transition from the G₁ phase into the S phase and leads to cell senescence and apoptosis (90). Rb is a major tumor suppressor protein that serves a key role in the induction of cellular senescence and is the major substrate of CDK4/6 (91). Rb inhibits E2F regulatory factors by binding to them and shielding their transcriptional activation domains (92). When the cell genome or epigenome is damaged, the DNA damage response system is activated, which leads to the activation of protein kinase C and increased levels of ROS, which ultimately increase the expression level of p16^INK4a (93). p16^INK4a activates the pRb tumor suppressor and inhibits cell proliferation (Fig. 3) (94). A study demonstrated that activation of the p16^INK4a-pRb pathway was critical for the development of senescence in most human cell lines (95). The expression of p16^INK4a is low or undetectable in young, healthy organisms, but increases exponentially in most aged tissues with senescence (96,97).

P14 (ARF) is one of the two proteins encoded by the CDKN2A, the other of which is p16^INK4a, and is an important inducer of cellular senescence (39). The p14 (ARF) protein is expressed at low levels in normal cells, but is markedly increased as the cells reach senescence. P14 (ARF) blocks cells in the G₁ and G₂/M phases mainly through the p53 pathway, leading to cell senescence or apoptosis. It has been shown that when the p16^INK4a gene is knocked out, mice are prone to developing tumors (35). It was initially considered that the deletion of p16^INK4a was responsible for disturbing of the cell senescence pathway; however, it has since been demonstrated that ARF-negative mice with normal expression levels of p16^INK4a exhibit the same phenotype as those with p16^INK4a deletion (94). Mdm2, also known as Hdm2, is a proto-oncogene in human tissues, which induces the degradation of p53 protein and inhibits its transcriptional activity. In particular, Mdm2 specifically binds to p53, thereby inhibiting its transcriptional activity and promoting its extracellular transport, resulting in the inactivation of the p53 protein (98). Transcription of the Mdm2 gene is regulated by a p53 transcriptional domain and the RAF-MEK-MAP kinase signal transduction pathway. The upregulation of Mdm2 by the RAF-MEK-MAP kinase pathway has been shown to decrease p53-mediated apoptosis (99–101). The exogenous expression of ARF stabilizes p53, promotes p53 transcription, p21^WAF1/CIP1 expression and activation of the p53-mediated apoptotic pathway, leading to apoptosis or premature senescence. In the absence of stimulation, p53 can promote the expression of Mdm2 while, conversely, Mdm2 reduces the activity of p53. p21 is downstream of p53, and its expression blocks progression of the cell cycle, promotes apoptosis and increases the sensitivity of tumor cells to chemotherapy (102). Mdm2 promotes the proteasome-mediated degradation of p21 and also reduces the stability of p21 by binding with it. When Mdm2 expression is downregulated, the expression of p21 protein increases (Fig. 4). Research has focused on investigating the involvement of the ARF-Mdm2-p53-p21 pathway in natural states as well as those involving premature senescence (103). Notably, one study showed that the loss of p14 (ARF) expression in patients with prostate cancer was positively associated with an increased risk of disease recurrence and metastatic disease (104).

Skp2 is a member of the F-box family, which mediates cell cycle regulation and cell proliferation by degrading cell cycle regulatory proteins, including p27, p21, p53, cyclin A, cyclin E and cyclin D. The Skp2 gene is an important regulator of the cell cycle, which interacts with the S-phase kinase cyclin A-CDK2 (cyclin dependent kinase 2) complex. The expression levels of Skp2 are very low in the G₀/G₁ phase, but increase markedly in the S phase. An important substrate of Skp2 in the cell division cycle is p27 (105). p27 is an important member of the CDK family; p27^kip1 is a protein encoded by the CDKN1B gene. As a negative regulator of the cell cycle, p27^kip1 inhibits the activity of CDK complexes to coordinate the cell cycle, DNA replication and DNA repair (106,107). In a study in which 68 cases of non-small cell lung cancer tissues were compared with normal bronchial epithelial cells using tissue microarrays and immunohistochemistry, the results revealed that Skp2 was only expressed in the lung cancer tissues while p27^kip1 was expressed in both normal bronchial epithelial cells and lung cancer cells (108). In addition, the expression of p27^kip1 was found to be significantly downregulated in Skp2-positive cells and negatively correlated with Skp2 expression. In another study, Shapira et al (109) analyzed the expression of Skp2 and p27^kip1 in the tissue sections of 80 patients with colorectal tumors using immunohistochemistry, and found that Skp2 overexpression was significantly associated with the loss of p27^kip1 expression and cell differentiation.

Skp2 may contribute to a malignant phenotype, and the overexpression of Skp2 protein results in the accelerated hydrolysis of p27 and deterioration of tumor cells (110). Since Skp2 can degrade p27^kip1 and promote tumor development, interference with the Skp2-p27^kip1 pathway may induce tumor senescence (111).

The LBR is an integral membrane protein of the interphase nuclear envelope. Its N-terminus protrudes into the nucleoplasm where it binds to lamin B and heterochromatin; these interactions are disrupted during mitosis (112). A number of studies have indicated that chromatin and chromatin proteins are involved in cellular senescence (113,114). For example, the altered expression of lamins A/C and B, which form the nuclear lamina, as well as altered heterochromatin structures have been observed in senescent cells (115,116). In addition, the expression of lamin B1 (LB1) in WI-38 cells was found to decrease during cellular senescence, and the silencing of LB1 slowed the proliferation of these cells and induced premature senescence. These effects were accompanied by a reduction in p53-dependent ROS, which was repaired by growth under hypoxic conditions (15).

In addition to the aforementioned senescent pathways, studies have demonstrated that inhibition of the ERK and AKT pathways can reduce the number of senescent cells, with ERK and AKT potentially acting through the ETS variant transcription factor 6 and forkhead box O1 genes (117–120). As such, drugs that target and disrupt downstream effectors of ERK and AKT have been proposed as new therapeutic methods. It has also been demonstrated that oxytocin can alleviate cell senescence through ERK/nuclear factor erythroid 2-related factor 2 (NrF2) signaling; this implies that interference with the ERK/NrF2 pathway may induce cell senescence (121). Furthermore, the inhibition of mTORC1 has been shown to prolong the lifespan of yeast, worms, fruit flies and some mice, suggesting that mTORC1 may also be a novel target for inducing cell senescence and treating various types of cancer (122).