|
|
||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online November 15, 2004 as doi:10.1096/fj.04-2305fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Urology, Louisiana State University Health Sciences Center, and Stanley S. Scott Cancer Center, New Orleans, Louisiana, USA
1Correspondence: Louisiana State University Health Sciences Center, Stanley S. Scott Cancer Center, 533 Bolivar St., CSRB Rm 441, New Orleans 70112, LA, USA. E-mail: wrayfo{at}lsuhsc.edu
SPECIFIC AIMS
The first aim of this study was to characterize the pattern of events evoked by exogenous p27Kip1 in PC-3 prostate cancer cells in the presence or absence of the reagents known to affect apoptotic pathways, pan-caspase inhibitor ZVAD-fmk, and cycloheximide (CHX). The second aim was to evaluate whether death receptors TNF receptor 1 and TNF receptor 2 are involved in p27Kip1-induced signaling to apoptosis.
PRINCIPAL FINDINGS
1. Infection of PC-3 prostate cancer cells with the recombinant adenovirus expressing wild-type p27Kip1 under cytomegalovirus (CMV) promoter (Adp27) resulted in a stable protein expression throughout the 5 day experiment in 92% of the cells (200pfu/cell)
Western blot analysis showed that increasing amounts of p27Kip1 protein corresponded to decreasing levels of the hyperphosphorylated form of retinoblastoma protein (Rb-pp), in agreement with the well-known function of p27Kip1 to prevent the phosphorylation of Rb protein and inhibit G1/S cell cycle progression. Flow cytometry analysis of the cell cycle showed that overexpression of p27Kip1 led to the dose-dependent clearance of cells from S phase and cycle arrest at the G1 phase by the third day of the experiment. We demonstrated that the p27Kip1 gene was efficiently delivered, stably expressed throughout the experiment in 92% of cells, and functional.
2. Microscopy analysis of Adp27-treated PC-3 cells showed that the majority of cells had enlarged nuclei with condensed and marginalized chromatin
Flow cytometry confirmed increased granularity of the nuclei in the samples infected with Adp27. Eighty-three percent of Adp27-treated cells showed DNA fragmentation in TUNEL assay combined with propidium iodide (PI) staining; cells treated with an empty virus (AdNull) contained 10% and untreated cells <5%. Flow cytometry analysis of sub-G1 fraction showed that <16% of Adp27-treated population had DNA cleaved into low molecular weight fragments within the 5 day experiment. Analysis of the cell cycle distribution in Adp27-treated sample suggested that the cells in G1 phase were protected from undergoing typical apoptosis. Phosphatidylserine moieties were exposed in only 38% of Adp27-infected cells and 22% had the cellular membrane compromised. The proapoptotic effect of exogenous p27Kip1 may depend on cell cycle distribution. We suggest that only the cells in S or G2/M phases would develop a full apoptotic program, whereas cells in G1 phase would undergo initial nuclear enlargement, chromatin condensation and margination, and DNA fragmentation to high molecular weight fragments.
3. Pan-caspase inhibitor, ZVAD-fmk, was used to evaluate caspase-dependent steps of p27Kip1-induced cell death
Treatment of Adp27-infected PC-3 cells with ZVAD-fmk resulted in a decrease of sub-G1 cells from 13% to 2%, comparable to control levels. However, TUNEL analysis of Adp27-treated cells in the presence of ZVAD-fmk showed no inhibitory effect on DNA fragmentation. Therefore, DNA fragmentation into large fragments that occurred in the majority of cells was a caspase-independent stage. We propose that p27Kip1 induces caspase-dependent stages of apoptosis only in the fraction of cells that were in S or G2/M phases of the cycle.
4. The protein synthesis inhibitor CHX was used to evaluate whether caspase-dependent and caspase-independent stages of p27Kip1-induced cell death require de novo synthesis of proteins
Treatment of Adp27-infected PC-3 cells with 2 µg/mL CHX for 18 h resulted in a significantly elevated percentage of sub-G1 cells (12.5% vs. 36% for Adp27, 200 pfu/cell) CHX treatment of Adp27-infected cells decreased the percentage of TUNEL-positive cells (64% Adp27 vs. 54% Adp27+CHX). We suggest that low molecular DNA fragmentation is caspase dependent and CHX sensitive, whereas caspase-independent stage of p27Kip1-induced cell death may require de novo synthesis of proteins. ELISA assay for caspase 8 activity in Adp27-treated cells showed that ZVAD-fmk reduced, and CHX stimulated, caspase 8 activity compared with Adp27 alone, suggesting that CHX may block an inhibitor of caspases thus promoting internucleosomal DNA fragmentation.
5. Considering the characteristic features of p27Kip1-induced cell death such as late onset, the enhancing effect of CHX, and DNA degradation to high molecular weight fragments, we hypothesized that TNF receptors 1 and 2 (TNFR1 and TNFR2) may play a role in signaling from overexpressed p27Kip1 to cell death
Overexpression of exogenous p27Kip1 resulted in increased expression of TNFR1 and TNFR2. RT PCR analysis of mRNA expression for TNF-
demonstrated that PC-3 cells expressed TNF-
in all samples tested (Adp27, AdNull, untreated). Treatment of Adp27-infected PC-3 cells with anti-TNFR1 neutralizing antibody resulted in a dose-dependent inhibition of DNA fragmentation, as detected by TUNEL (Fig. 1
), but anti-TNFR2 neutralizing antibody did not affect DNA fragmentation in PC-3 cells overexpressing p27Kip1 (Fig. 1)
. Adp27-infected cells treated with TNF
(10 ng/mL, 6 h) in the presence of CHX (10 µg/mL, 6 h) showed increased sub-G1 fraction vs. AdNull-infected cells (Adp27+TNF+CHX vs. AdNull+TNF+CHX). Such sensitization of PC-3 cells by Adp27 to undergo TNF-mediated apoptosis and the inhibitory effect of neutralizing anti-TNFR1 strongly suggest that TNF signaling, particularly from TNFR1, is involved in the proapoptotic effect of p27Kip1 in prostate cancer cell line PC-3.
|
CONCLUSIONS AND SIGNIFICANCE
There is much controversy about the role of p27Kip1 in apoptosis. Proapoptotic effects of exogenous p27Kip1 were observed in prostate, cervical, renal, lung, melanoma, and colorectal cancer cell lines infected with adenovirus. Accumulation of endogenous p27Kip1 by inhibition of proteosome function was shown to induce apoptosis in oral squamous cell carcinoma and lung cancer cells. In contrast, antiapoptotic actions of p27Kip1 were reported in human leukemic and lymphoid cells. The final outcome of p27Kip1 overexpression may depend on cell type or cell type-specific signaling pathways in malignant cells.
We used a variety of techniques to characterize the extent and pattern of cellular events during cell death induced by p27Kip1 in PC-3 cells. We did not observe typical cellular and nuclear shrinkage in Adp27-infected cells. The majority of Adp27-infected cells showed nuclear enlargement, chromatin condensation, and caspase-independent DNA cleavage into large fragments. Our observations are supported by reports showing that exogenous p27Kip1 caused nuclear enlargement in HeLa cells and lung carcinoma cell line A549. Nuclear enlargement and chromatin condensation were also observed upon accumulation of endogenous p27Kip1 when using antisense S-phase kinase-associated protein 2 (Skp2). Our results show that p27Kip1 may induce a transient arrest at the G2/M checkpoint. Such enrichment of cells in G2-phase after Adp27 infection was also reported in HeLa and A549 cells.
Our results suggest that cells in G1 phase are protected from executing full apoptosis. Only cells in S- and G2/M phases were clearing after p27Kip1 overexpression, suggesting this would be the fraction able to develop the typical caspase-dependent features of apoptosis. We speculate that distribution of cells in the cell cycle can determine the fate of the Adp27-treated cells. We propose that exogenous p27Kip1 leads primarily to chromatin condensation and DNA breaks introduced in a caspase-independent manner. Only cells committed to S phase and/or approaching G2/M checkpoint would be able to activate caspase-cascade and run the full program of apoptosis while the cells in G1 phase would stay protected. This interpretation would explain the antiapoptotic effects of p27Kip1 reported for cell lines arrested in G1 phase.
We evaluated the role of TNF-
signaling in PC-3 cells because we observed that p27Kip1-induced cell death 1)had late onset; 2)involved DNA-cleavage into high molecular weight fragments; and 3)was enhanced by CHX. The combination of TNF-
and CHX is a common way to induce apoptosis in PC-3 cells. TNF-
exerts its cellular function through TNF receptors 1 and 2 and may lead to either survival through TRAF2 and NF-
B or apoptosis through activation of caspase 8. This is the first report showing that overexpression of p27Kip1 causes higher expression of TNFR1 and TNFR2, and that neutralization of TNFR1 down-regulates proapoptotic signaling from exogenous p27Kip1in prostate cancer cells.
TNFR1-induced apoptosis generally occurs when antiapoptotic signaling through activated NF-
B is down-regulated. It was shown that TNFR1 needs to be recruited to lipid rafts in order to mediate NF-
B activation, and any interference with the lipid raft organization may switch TNF signaling from NF-
B activation to apoptosis in the human fibrosarcoma cells. Membrane rafts in normal lymphocytes were reported to contain p27Kip1, and it was suggested that p27Kip1 may play a role in modulating membrane signaling, since it was excluded from the rafts upon lymphocytes activation. Taking these two together, we speculate that in Adp27-infected PC-3 cells, p27Kip1 remains in lipid rafts, preventing activation of NF-
B and promoting the signaling to apoptosis.
|
FOOTNOTES
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.04-2305fje;
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |