PresentationMeeting Presentation

Matrix Metalloproteinase-7 and the 20S Proteasome Contribute to Cellular Senescence

See allHide authors and affiliations

Science Signaling  25 Mar 2008:
Vol. 1, Issue 12, pp. pt1
DOI: 10.1126/stke.112pt1
A presentation from the 11th Joint Meeting of the Signal Transduction Society (STS), Signal Transduction: Receptors, Mediators and Genes, Weimar, Germany, 1 to 3 November 2007.


A detailed understanding of aging and senescence is limited by the complex interplay of the effects of extracellular and environmental stimuli on cellular metabolic, mutational, and epigenetic phenomena. For example, STASIS (stress or aberrant signaling–induced senescence) is affected by the exposure to free radicals and conditions that cause an increased cellular production of reactive oxygen species (ROS) during normal life span. In addition, progressive telomere erosion and telomeric dysfunction contribute to a cellular feature termed replicative or cellular senescence. To focus on distinct cellular pathways that contribute to these different forms of senescence, we investigated the reversible differentiation and aging process of the human U937 leukemia cell line. This was compared to cellular senescence that occurred in normal primary human mammary epithelial cells (HMECs). These two cell systems revealed an important role of the proteolytic activity of the 20S proteasome and its activation by the nuclear protein poly(ADP-ribose) polymerase–1 (PARP-1) during "retrodifferentiation" and rejuvenation of the leukemic cells. Moreover, reduced extracellular proteolytic activity of certain matrix metalloproteinases—for example, MMP-7—is associated with accelerated aging and senescence in normal HMECs.

Presentation Notes

The complex biology of aging is affected by exogenous components as well as intracellular mechanisms. To address the question of what signaling mechanisms might be involved in senescence, we investigated the aging process of two different cultured cell models (Slide 3).

In contrast to adherent primary normal human cells, such as human mammary epithelial cells (HMECs), artificially immortalized human U937 cells are a leukemic tumor cell line growing in suspension (Slide 3). Treatment of U937 cells with the phorbol ester 12-O-tetradecanoylphorbol 13-acetate (TPA) induces a process of differentiation and STASIS (stress or aberrant signaling–induced senescence) (1). In particular, TPA triggers increased generation of reactive oxygen species (ROS), whereby oxidative stress products initiate the stress-activated protein kinase (SAPK) pathway through activation of protein kinase C-β2 (2). These effects eventually stimulate the tumor cells to differentiate along the monocyte- and macrophage-like lineage (3). During this differentiation, the cells become adherent and undergo a process of aging, characterized by the inhibition of cell growth and increased senescence-associated β-galactosidase (SA-β-gal) expression, a distinctive endogenous biomarker associated with the senescent phenotype (4, 5). Surprisingly, after about 2 to 3 weeks in culture, the differentiated and aged cells spontaneously revert all acquired differentiation markers, regain proliferative capacity, and thus rejuvenate—a phenomenon termed retrodifferentiation (68). Previous studies revealed that retrodifferentiation and rejuvenation of U937 cells is not associated with the outgrowth of TPA-resistant subclones or noninduced cells (3, 9).

Several questions arise from this retrodifferentiation and rejuvenation phenomenon: (i) What are the trigger and the signaling cascade that restore the undifferentiated phenotype? (ii) What are the mechanisms that cause the spontaneous loss of all differentiation-associated marker proteins? It is known that TPA treatment can generate ROS, which contribute to DNA strand breaks and to oxidative damage of proteins and lipids. A persistent increase in damaged DNA induces cell cycle arrest and activates the DNA damage response, leading to repair. Moreover, an accumulation of dysfunctional (oxidatively damaged) proteins in the cells may impair cellular processes; thus, the cells need to degrade these detrimental proteins (1). Accordingly, we observed significantly increased proteolytic activity of the 20S proteasome at the time of retrodifferentiation and rejuvenation (Slide 4), indicating enhanced protein turnover. The 20S proteasome preferentially degrades oxidatively modified proteins and unfolded hydrophobic proteins (10). Although the activity of the 20S proteasome was significantly increased, the abundance of the proteasome subunits surprisingly appeared unaltered. A slight decrease in the protein level of the β2 subunit was paralleled by a slight increase in the abundance of the β1 subunit. Thus, changes in transcription, translation, or stability of the proteasome subunits did not appear to cause the enhanced proteasomal activity. Therefore, the question arose: What may trigger this marked increase in 20S proteasome activity?

Previous work described the nuclear enzyme poly(ADP-ribose) polymerase 1 (PARP-1) as a potential regulator of 20S proteasome activity (1113). Indeed, Western blot analysis revealed a significant increase in the abundance of PARP-1 at the time of retrodifferentiation that correlated with maximum 20S proteasome activity (Slide 4). To further investigate the role of PARP-1 during retrodifferentiation and rejuvenation, we targeted PARP-1 by stable transfection of U937 cells with an antisense-PARP-1 vector. Whereas differentiated U937 cells underwent retrodifferentiation and reentered the cell cycle in a manner similar to that of control vector transfectants (pTracer U937), antisense-PARP-1–transfected U937 (asPARP-1 U937) remained in a growth-arrested state (Slide 5). The characteristic SA-β-gal staining for senescent cells disappeared after 17 days in differentiated U937 cells, indicating retrodifferentiation (Slide 5). In contrast, SA-β-gal staining in asPARP-1 cells persisted after day 17 and the cells were kept in a differentiated and senescent state, which underscores the importance of PARP-1 for the retrodifferentiation and rejuvenation process (Slide 5). Initiation of the retrodifferentiation and rejuvenation program seems to be the consequence of a threshold effect of oxidative stress in the cells (2). In particular, increased ROS generation by TPA treatment may lead to an accumulation of oxidatively damaged proteins and DNA and, thus, to the induction of PARP-1 and repair processes. Subsequently, PARP-1 activates the 20S proteasome, which preferably degrades oxidatively damaged proteins. Together, both PARP-1 and the 20S proteasome appear to be crucial factors allowing retrodifferentiation and rejuvenation of U937 cells (2, 7, 13, 14).

In addition to altered intracellular signaling, there also might be extracellular processes contributing to differentiation and retrodifferentiation. Again, the U937 cells served as a model because they grow in suspension when undifferentiated and become adherent during differentiation. Another human leukemic cell line, termed TUR (TPA-U937 resistant) (American Type Culture Collection No. CRL-2367), does not differentiate along the monocyte and macrophage lineage, maintains proliferative capacity, and remains in suspension upon TPA exposure (15). Accordingly, we performed DNA microarray analysis to compare TPA-resistant TUR cells and asPARP-1–transfected TUR cells, which had recovered their susceptibility to TPA upon PARP-1 down-modulation (13, 14). This revealed a marked increase in the expression of genes encoding matrix metalloproteinases (MMPs) in the asPARP-1–transfected TUR cell line upon TPA treatment. In contrast, there was little if any change in the control TUR cells. The microarray data were further confirmed at the protein level by Western blot analysis, demonstrating a significantly elevated abundance of MMP-1, MMP-7, and MMP-9 in differentiated asPARP-1 TUR cells after 72 hours, in contrast to the undifferentiated wild-type TUR (data not shown) and control vector-transfected TUR cells (Slide 6). The increased expression and abundance of matrix metalloproteinases suggest that enhanced proteolysis of certain extracellular matrix (ECM) components and, thus, distinct changes in cell-to-cell and ECM interactions, may occur for adherent differentiated cells compared to the suspension cells. These interactions likely indicate that the microenvironment surrounding the cells participates in the determination of differentiation and senescence.

In comparison to the artificially immortalized U937 cell line, it was of interest to investigate the factors that accompany the aging process in normal primary epithelial cells (Slide 7). These primary HMECs have a limited life span and develop a senescent phenotype during the time in culture (5). Typically, senescence of HMECs is associated with several morphological and functional changes. The initially "young" HMECs grow as a small adherent population up to passage 12 (P12) with continued cell divisions [see the cell cycle data (Slide 8)]. During later passages (P13 through P16), the cells increase in size, form long cytoplasmic processes, and eventually become giant multinucleated cells, which cease to divide (Slide 8). Cell cycle analysis demonstrated a progressively decreasing proportion of cells in G0/G1 and S phase with increasing culture passages. This was paralleled by an accumulation of cells in the G2/M phase. Moreover, the appearance of an 8N DNA peak in populations at P14 could be distinguished in addition to 2N and 4N peaks, indicating aberrant mitosis. This 8N DNA population continuously increased until P16 (Slide 8). Likewise, more than 80% of the P16 cell population stained positively in the SA-β-gal assay, confirming the senescent state of these cells (Slide 9).

Whereas differentiation and senescence of asPARP-1–transfected TUR cells was accompanied by significantly increased expression of genes encoding distinct matrix metalloproteinases (MMP-1, MMP-7, or MMP-9) as compared to undifferentiated TUR cells (Slide 5), few if any changes were detectable in the protein levels of MMP-9, MMP-1, or MMP-2 during the aging process of HMEC (Slide 10). However, for MMP-7, the abundance of the latent precursor form was slightly reduced between P15 and P16, and the reduction was even more pronounced for the active form of MMP-7. Thus, the amount of active MMP-7 protein continuously decreased after P14 and was below the detection limit in P16 (Slide 10). Thus, the discrepancy between the protein levels of MMP-7 in aging tumorigenic leukemia cells versus aging normal epithelial cells focused our attention on how this ECM protease might be associated with the aging process of HMECs. MMP-7, also known as matrilysin, is predominantly involved in proteolytic degradation of ECM, as well as the ectodomain shedding of membrane-associated proteins (16). Moreover, this matrix metalloproteinase is overexpressed in a variety of human tumors, including breast cancer, and is associated with poor prognosis in patients with cancer (Slide 11) (17).

To reveal possible aging effects in response to decreased MMP-7 abundance, young HMECs of passage 12 were transiently transfected with MMP-7 small interfering RNA (siRNA) (Slide 12). The transfection method was tested using a green fluorescent protein (GFP) plasmid. Fluorescence microscopy demonstrated successful expression of GFP in transfected P12 HMECs. Fluorescence-activated cell sorting analysis of fluorescein isothiocyanate (FITC)–labeled siRNA-transfected P12 HMECs revealed a transfection efficiency of ~80% (Slide 12). The successful reduction of latent and active MMP-7 abundance was confirmed by Western blot analysis between 48 and 72 hours after transfection and revealed a significantly reduced protein level when compared to HMECs transfected with nontargeting siRNA (neg. siRNA) (Slide 12). The MMP-7 transfectants were then analyzed for the presence of cells in various phases of the cell cycle and for aging by staining with SA-β-gal. The transfection itself had no influence on the distribution of cells in the different phases of the cell cycle when P12 HMECs transfected with negative siRNA were compared to nontransfected P12 HMECs (P12) (Slide 13). In contrast, reduction of MMP-7 by siRNA in P12 HMECs was associated with an altered distribution of cells in the cell cycle (Slide 13). Thus, 72 hours after transfection with MMP-7 siRNA, P12 HMECs demonstrated a significantly reduced proportion of cells in G0/G1 phase and an accumulation of cells in the G2/M phase, indicating reduced proliferative capacity. Compared to the cell cycle pattern of aging HMECs, this cell cycle resembles a distribution of HMECs between P13 and P14 (Slide 13), suggesting an accelerated aging process for HMECs after MMP-7 down-modulation. To further confirm this hypothesis, we performed staining for the aging marker SA-β-gal (Slide 14). Little if any effect was observed by the transfection itself (Slide 14); however, reduction of MMP-7 by siRNA in P12 HMECs significantly enhanced the proportion of cells that were positive for SA-β-gal activity. Whereas 6 ± 2.1% and 10 ± 2.2% of SA-β-gal–positive senescent cells were detectable in HMEC control cells and negative siRNA-transfected cells, respectively, MMP-7-siRNA transfectants demonstrated 25 ± 3.2% of SA-β-gal–positive senescent cells, supporting the hypothesis that inhibition of MMP-7 accelerated aging in young HMECs (Slide 14).

Taken together, these data demonstrate the involvement of intracellular proteolytic mechanisms during retrodifferentiation and rejuvenation of the immortalized human leukemia cell lines, whereas senescence of normal primary human mammary epithelial cells is also affected by extracellular proteolytic processes. In particular, retrodifferentiation and rejuvenation in leukemic cells require activation of PARP-1 and the 20S proteasome (7, 14), whereas decreased MMP-7 protease activity contributes to an accelerated aging of HMECs (Slide 15) (6).

Editor’s note: This contribution, in which primary data are shared with other readers, is not intended to be equivalent to an original research paper. Note, in particular, that the text and associated slides have not been peer reviewed.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
View Abstract

Navigate This Article