An emerging role for the nuclear localization of maspin in the suppression of tumor progression and metastasis.

Maspin, a member of the serpin family of serine protease inhibitors, was originally identified as a tumor suppressor that is expressed in normal mammary epithelial cells but is reduced or absent in breast carcinomas. Early enthusiasm for maspin as a biomarker for disease progression has been tempered by clinical data that associates maspin with favourable outcomes in some studies and poor prognosis in others. Here, we review all of the published clinical studies for maspin in breast and ovarian cancers and propose that the apparent discordance between clinical reports is a consequence of differential cellular distribution of maspin. Indeed, it was thought that an extracellular pool of maspin possessed tumor suppressor activity, acting by inhibiting migration and increasing cell adhesion. Recent evidence from our group and others indicates, however, that the nuclear localization of maspin in cancer cells is necessary for its tumor suppressor activity. We provide additional data here to demonstrate that nuclear-localized maspin binds to chromatin and is required to effectively prevent cells from metastasizing. Our knowledge of other serpins that localize to the nucleus should help to inform future studies of nuclear maspin. Elucidation of the molecular mechanisms regulating the localization and activities of maspin should pave the way for the development of improved diagnostics and therapies for cancer.

Nuclear localization of maspin is essential for its inhibition of tumor growth and metastasis.

Maspin (mammary serine protease inhibitor or SerpinB5) acts as a tumor suppressor when overexpressed in aggressive cancer cell lines. However, its role in human cancer is controversial. Maspin expression has been associated with a poor prognosis in some studies, whereas in others, with favorable outcome. The clinical data suggest, however, that nuclear-localized maspin is associated with improved survival. We hypothesized that the tumor suppressor activity of maspin may require nuclear localization, and that the discordance between clinical and experimental reports is a consequence of the variable subcellular distribution of maspin. Furthermore, we surmized that nuclear maspin could function as a tumor suppressor through the regulation of genes involved in tumor growth and invasion. Maspin or maspin fused to a nuclear export signal were expressed in metastatic human breast and epidermoid carcinoma cell lines. We found that pan-cellular localized maspin inhibited in vivo tumor growth and metastasis when assessed in xenograft chicken embryo and murine mammary fat pad injection models. However, when maspin was excluded from the nucleus via a nuclear exclusion signal, it no longer functioned as a metastasis suppressor. Using chromatin immunoprecipitation, we show that nuclear maspin was enriched at the promoter of colony-stimulating factor-1 (CSF-1) and associated with diminished levels of CSF-1 mRNA. Our findings demonstrate that the nuclear localization of maspin is required for its tumor and metastasis suppressor functions in vivo, and suggest that its mechanism of action involves, in part, direct association of maspin with target genes.

Proteolytic processing of cut homeobox 1 by neutrophil elastase in the MV4;11 myeloid leukemia cell line.

Proteolytic processing by cathepsin L generates p110 Cut homeobox 1 (CUX1) at the end of the G(1) phase, whereas an alternative transcript encodes p75 CUX1. These short CUX1 isoforms were reported to be overexpressed in cancer cells, and transgenic mice overexpressing the p75 isoform were found to develop myeloproliferative disease-like myeloid leukemias. In the present study, we report that the neutrophil elastase can also generate a short CUX1 isoform in the MV4;11 acute myeloid leukemia cell line. Proteolytic processing was so efficient that the full-length CUX1 protein was detected only when cells were maintained in the presence of the specific elastase inhibitor III. In agreement with these findings, higher levels of the processed cyclin E isoforms were also detected in MV4;11 cells. Reappearance of full-length cyclin E and CUX1 could be induced upon the treatment of MV4;11 cells with the differentiation inducer phorbol 12-myristate 13-acetate or, unexpectedly, following overexpression of a short recombinant CUX1 protein. In both cases, the mechanism involved transcriptional repression of the neutrophil elastase gene. This result revealed a negative feedback loop whereby CUX1 shuts down the expression of the protease that cleaves it. Overall, the findings in MV4;11 and other cancer cells suggest that various mechanisms are used in cancer to favor the expression of short CUX1 isoforms.

Increased expression and activity of nuclear cathepsin L in cancer cells suggests a novel mechanism of cell transformation.

It is generally accepted that the role of cathepsin L in cancer involves its activities outside the cells once it has been secreted. However, cathepsin L isoforms that are devoid of a signal peptide were recently shown to be present in the nucleus where they proteolytically process the CCAAT-displacement protein/cut homeobox (CDP/Cux) transcription factor. A role for nuclear cathepsin L in cell proliferation could be inferred from the observation that the CDP/Cux processed isoform can accelerate entry into S phase. Here, we report that in many transformed cells the proteolytic processing of CDP/Cux is augmented and correlates with increased cysteine protease expression and activity in the nucleus. Taking advantage of an antibody that recognizes the prodomain of human cathepsin L, we showed that human cells express short cathepsin L species that do not contain a signal peptide, do not transit through the endoplasmic reticulum, are not glycosylated, and localize to the nucleus. We also showed that transformation by the ras oncogene causes rapid increases both in the production of short nuclear cathepsin L isoforms and in the processing of CDP/Cux. Using a cell-based assay, we showed that a cell-permeable inhibitor of cysteine proteases is able to delay the progression into S phase and the proliferation in soft agar of ras-transformed cells, whereas the non-cell-permeable inhibitor had no effect. Taken together, these results suggest that the role of cathepsin L in cancer might not be limited to its extracellular activities but may also involve its processing function in the nucleus.

Carboxyl-terminal proteolytic processing of CUX1 by a caspase enables transcriptional activation in proliferating cells.

Proteolytic processing at the end of the G(1) phase generates a CUX1 isoform, p110, which functions either as a transcriptional activator or repressor and can accelerate entry into S phase. Here we describe a second proteolytic event that generates an isoform lacking two active repression domains in the COOH terminus. This processing event was inhibited by treatment of cells with synthetic and natural caspase inhibitors. In vitro, several caspases generated a processed isoform that co-migrated with the in vivo generated product. In cells, recombinant CUX1 proteins in which the region of cleavage was deleted or in which Asp residues were mutated to Ala, were not proteolytically processed. Importantly, this processing event was not associated with apoptosis, as assessed by terminal dUTP nick end labeling assay, cytochrome c localization, poly(ADP-ribose) polymerase cleavage, and fluorescence-activated cell sorting. Moreover, processing was observed in S phase but not in early G(1), suggesting that it is regulated through the cell cycle. The functional importance of this processing event was revealed in reporter and cell cycle assays. A recombinant, processed, CUX1 protein was a more potent transcriptional activator of several cell cycle-related genes and was able to accelerate entry into S phase, whereas mutants that could not be processed were inactive in either assay. Conversely, cells treated with the quinoline-Val Asp-2,6-difluorophenoxymethylketone caspase inhibitor proliferated more slowly and exhibited delayed S phase entry following exit from quiescence. Together, our results identify a substrate of caspases in proliferating cells and suggest a mechanism by which caspases can accelerate cell cycle progression.

A novel proteolytically processed CDP/Cux isoform of 90 kDa is generated by cathepsin L.

The Cut-like genes code for multiple isoforms of the CDP/Cux transcription factor. The full-length protein contains four DNA-binding domains: Cut repeats 1, 2 and 3 (CR1, CR2 and CR3) and the Cut homeodomain (HD). The p75 isoform is expressed from an mRNA that is initiated within intron 20 and contains only CR3 and HD. The p110 isoform is generated by proteolytic processing by cathepsin L and contains CR2, CR3 and HD. In the present study, we show that an additional isoform of 90 kDa is expressed in many cell lines of epithelial origin. Mapping experiments with deletion mutants indicated that the N-terminus of p90 is located upstream of CR2, between amino acids 918 and 938. Indeed, p90 and p110 displayed similar DNA-binding and transcriptional activities. The p90 isoform, like p110, was found to be generated by proteolytic processing. The steady-state level of both p90 and p110 correlated with the level of cathepsin L activity. Importantly, co-expression with a cathepsin L mutant that is initiated at downstream AUG sites also stimulated the generation of p90 and p110. These results strongly suggest that p90, like p110, is generated by cathepsin L isoforms that are devoid of a signal peptide.

The p110 isoform of the CDP/Cux transcription factor accelerates entry into S phase.

The CDP/Cux transcription factor was previously found to acquire distinct DNA binding and transcriptional properties following a proteolytic processing event that takes place at the G1/S transition of the cell cycle. In the present study, we have investigated the role of the CDP/Cux processed isoform, p110, in cell cycle progression. Populations of cells stably expressing p110 CDP/Cux displayed a faster division rate and reached higher saturation density than control cells carrying the empty vector. p110 CDP/Cux cells reached the next S phase faster than control cells under various experimental conditions: following cell synchronization in G0 by growth factor deprivation, synchronization in S phase by double thymidine block treatment, or enrichment in G2 by centrifugal elutriation. In each case, duration of the G1 phase was shortened by 2 to 4 h. Gene inactivation confirmed the role of CDP/Cux as an accelerator of cell cycle progression, since mouse embryo fibroblasts obtained from Cutl1z/z mutant mice displayed a longer G1 phase and proliferated more slowly than their wild-type counterparts. The delay to enter S phase persisted following immortalization by the 3T3 protocol and transformation with H-RasV12. Moreover, CDP/Cux inactivation hindered both the formation of foci on a monolayer and tumor growth in mice. At the molecular level, expression of both cyclin E2 and A2 was increased in the presence of p110 CDP/Cux and decreased in its absence. Overall, these results establish that p110 CDP/Cux functions as a cell cycle regulator that accelerates entry into S phase.

Complete and limited proteolysis in cell cycle progression.

An important mechanism of regulation that controls progression through the cell cycle involves the timely degradation of specific regulatory proteins. In parallel to the main degradative pathways, it appears that the function of certain proteins may also be modulated by a process called limited proteolysis. We have recently shown that the CDP/Cux transcription factor is proteolytically processed at the G(1)/S transition by the cathepsin L protease. Two aspects of these findings are discussed in the context of the cell cycle. Firstly, together with the cohesin subunit Scc1 and the HCF-1 factor, CDP/Cux represents a third example whereby the process of "limited proteolysis" plays a role in the control of cell cycle progression. Secondly, our findings provides compelling evidence that the cathepsin L protease, which was believed to be obligatorily targeted through the endoplasmic reticulum to the lysosomes or the extra-cellular milieu, could also be present in the nucleus and modulate the function of transcription factors.

A cathepsin L isoform that is devoid of a signal peptide localizes to the nucleus in S phase and processes the CDP/Cux transcription factor.

The subclass of cysteine proteases termed lysosomal cathepsins has long been thought to be primarily involved in end-stage protein breakdown within lysosomal compartments. Furthermore, few specific protein substrates for these proteases have been identified. We show here that cathepsin L functions in the regulation of cell cycle progression through proteolytic processing of the CDP/Cux transcription factor. CDP/Cux processing in situ was increased following ectopic expression of cathepsin L but was reduced in Cat L(-/-) cells. Furthermore, catalytically active cathepsin L was localized to the nucleus during the G1-S transition as detected by immunofluorescence imaging and labeling using activity-based probes. Trafficking of cathepsin L to the nucleus is accomplished through a mechanism involving translation initiation at downstream AUG sites and the synthesis of proteases that are devoid of a signal peptide. Overall, these results uncover an as yet unsuspected role for cysteine proteases in the control of cell cycle progression.

CDP/Cux stimulates transcription from the DNA polymerase alpha gene promoter.

CDP/Cux (CCAAT-displacement protein/cut homeobox) contains four DNA binding domains, namely, three Cut repeats (CR1, CR2, and CR3) and a Cut homeodomain. CCAAT-displacement activity involves rapid but transient interaction with DNA. More stable DNA binding activity is up-regulated at the G(1)/S transition and was previously shown to involve an N-terminally truncated isoform, CDP/Cux p110, that is generated by proteolytic processing. CDP/Cux has been previously characterized as a transcriptional repressor. However, here we show that expression of reporter plasmids containing promoter sequences from the human DNA polymerase alpha (pol alpha), CAD, and cyclin A genes is stimulated in cotransfections with N-terminally truncated CDP/Cux proteins but not with full-length CDP/Cux. Moreover, expression of the endogenous DNA pol alpha gene was stimulated following the infection of cells with a retrovirus expressing a truncated CDP/Cux protein. Chromatin immunoprecipitation (ChIP) assays revealed that CDP/Cux was associated with the DNA pol alpha gene promoter specifically in the S phase. Using linker scanning analyses, in vitro DNA binding, and ChIP assays, we established a correlation between binding of CDP/Cux to the DNA pol alpha promoter and the stimulation of gene expression. Although we cannot exclude the possibility that stimulation of gene expression by CDP/Cux involved the repression of a repressor, our data support the notion that CDP/Cux participates in transcriptional activation. Notwithstanding its mechanism of action, these results establish CDP/Cux as an important transcriptional regulator in the S phase.

Characterization of a tissue-specific CDP/Cux isoform, p75, activated in breast tumor cells.

Two isoforms of the CCAAT-displacement protein/cut homeobox (CDP/Cux) transcription factor have been characterized thus far. The full length protein, p200, which contains four DNA binding domains, transiently binds to DNA and carries the CCAAT-displacement activity. The p110 isoform is generated by proteolytic processing at the G1-S transition and is capable of stable interaction with DNA. Here we demonstrate the existence of a shorter CDP/Cux isoform, p75, which contains only two DNA binding domains, Cut repeat 3 and the Cut homeodomain, and binds more stably to DNA. CDP/Cux p75 was able to repress a reporter carrying the promoter for the cyclin-dependent kinase inhibitor p21 gene and to activate a DNA polymerase alpha gene reporter. Expression of CDP/Cux p75 involved a novel mechanism: transcription initiation within intron 20. The intron 20-initiated mRNA (I20-mRNA) was expressed at higher level in the thymus and in CD4+/CD8+ and CD4+ T cells. I20-mRNA was expressed only weakly or not at all in normal human mammary epithelial cells and normal breast tissues but was detected in many breast tumor cells lines and breast tumors. In invasive tumors a significant association was established between higher I20-mRNA expression and a diffuse infiltrative growth pattern (n = 41, P = 0.0137). In agreement with these findings, T47D breast cancer cells stably expressing p75 could not form tubule structures in collagen but rather developed as solid undifferentiated aggregates of cells. Taken together, these results suggest that aberrant expression of the CDP/Cux p75 isoform in mammary epithelial cells may be associated with the process of tumorigenesis in breast cancer.