Deacetylation of cortactin by SIRT1 promotes cell migration.

Cortactin binds F-actin and promotes cell migration. We showed earlier that cortactin is acetylated. Here, we identify SIRT1 (a class III histone deacetylase) as a cortactin deacetylase and p300 as a cortactin acetylase. We show that SIRT1 deacetylates cortactin in vivo and in vitro and that the SIRT1 inhibitor EX-527 increases amounts of acetylated cortactin in ovarian cancer cells. We also show that p300 acetylates cortactin in vivo and that cells lacking or depleted of p300 express less-acetylated cortactin than do control cells. Deletion analysis mapped the SIRT1-binding domain of cortactin to its repeat region, which also binds F-actin. Mouse embryo fibroblasts (MEFs) lacking sir2alpha (the mouse homolog of SIRT1) migrated more slowly than did wild-type cells. The expression of SIRT1 in sir2alpha-null cells restored migratory capacity, as did expression of a deacetylation-mimetic mutant of cortactin. SIRT1 and cortactin were more abundant in breast tumor tissue than in their normal counterparts, whereas SIRT1 expression inversely correlates with the ratio of acetylation cortactin versus total cortactin. These data suggest that deacetylation of cortactin is associated with high levels of SIRT1 and tumorigenesis. Finally, breast and ovarian cancer cell lines expressing an acetylation mimetic mutant of cortactin are less motile than that of control cells, whereas cells expressing the deacetylation mimetic mutant of cortactin migrate faster than that of control cells in Transwell migration assays. In summary, our results suggest that cortactin is a novel substrate for SIRT1 and p300 and, for the first time, a possible role for SIRT1 in cell motility through deacetylation of cortactin.

Ultraviolet irradiation- and dimethyl sulfoxide-induced telomerase activity in ovarian epithelial cell lines.

Information about telomerase regulation is incomplete, especially since various studies suggest complexity in telomerase regulation. Given the important association between telomerase and cancer, it is imperative to design and develop a model system in which telomerase activity can be regulated and studied. We employed ultraviolet (UV) radiation or dimethyl sulfoxide (DMSO) to transiently induce telomerase activity in a telomerase-positive cell line and, most importantly, in a telomerase-negative cell line. UV- or DMSO-induced telomerase activity was associated with increased hTRT, but not hTR, mRNA transcription in the telomerase-negative cells. However, no changes in hTRT or hTR mRNA transcription were noted with UV- or DMSO-induced telomerase activity in the telomerase-positive cells. Inhibition of protein synthesis or the phosphotidyl inositol 3-kinase (PI3K) pathway suppressed telomerase induction and/or activity in all cell lines examined, suggesting telomerase activity was dependent on protein synthesis and PI3K-mediated phosphorylation. Furthermore, enhanced telomerase activity was limited to UV and DMSO, since a variety of chemotherapeutic agents failed to induce telomerase activity. Therefore, our data provide a useful culture model system to study telomerase regulation in telomerase-negative and -positive cell lines and from which to obtain information about telomerase as a target for cancer intervention.

Telomeric instability and reduced proliferative potential in ovarian surface epithelial cells from women with a family history of ovarian cancer.

OBJECTIVE: Increased telomeric instability in normal ovarian surface epithelium may contribute to ovarian carcinogenesis in women from families with a high frequency of breast/ovarian cancer. To test this hypothesis, we compared proliferative potential, mean telomeric length, and telomerase activity in SV-40 large T-antigen transfected cell lines derived from normal ovarian surface epithelium of women with and without a familial history of breast/ovarian cancer. METHODS: Telomeric instability was examined in SV-40 large T-antigen transfected cell lines of normal ovarian surface epithelium from patients with (FHIOSE, N = 5) and without (NFHIOSE, N = 11) a history of familial breast/ovarian cancer. The duration and total attainable number of population doublings, mean telomeric length, rate of telomeric loss, and telomerase activity were determined by cell counts, Southern blot analysis, and PCR ELISA. RESULTS: FHIOSE cells attained fewer population doublings than NFHIOSE cells and doubled at approximately half the rate of NFHIOSE cells, indicating a reduced proliferative capacity in FHIOSE cells. While telomerase activity was not detected in FHIOSE or NFHIOSE cell lines, mean telomeric lengths in FHIOSE were generally 1 kb shorter than in NFHIOSE cells and the rate of telomeric loss as a function of population doublings was up to threefold greater in FHIOSE cells. CONCLUSIONS: Increased telomeric instability and reduced growth potential suggest greater proximity to replicative senescence in ovarian surface epithelium from women with a familial history of breast/ovarian cancer. Consequently, an accumulation of genetic aberrations due to accelerated cellular aging may contribute to the enhanced susceptibility for malignant transformation and earlier onset in heritable ovarian cancer.

Telomeric length in individuals and cell lines with altered p53 status.

Telomeres play an important role in maintaining chromosomal stability and are often shortened in transformed cells. p53 is the most commonly mutated gene in cancers and its status is thought to reflect the level of genomic stability. We measured telomeric length by Southern blot analysis in cells from cancer-prone syndromes and in selected cancer cells with altered p53 status. Mean telomeric lengths in the cancer-prone syndromes Li-Fraumeni syndrome, Fanconi's anemia, and ataxia telangiectasia, were shorter in the affected individuals than in their unaffected parents. We also found that altered p53 expression in selected cancer cell model systems may be associated with shortened telomeric length, but did not appear to be associated with significant alterations in telomerase activity.

Telomerase activity is elevated in early S phase in hamster cells.

Telomerase is a ribonucleoprotein that elongates telomeric repeats de novo. We examined the possibility that telomerase activity is cell cycle regulated by examining telomerase activity in cell cycle synchronized Chinese hamster ovary (CHO) B11 cells. Overall telomerase activity was similar in growing and quiescent cells. Further, cells synchronized in G1, S, or G2/M showed similar levels of telomerase activity. However, a detailed analysis of cells within S phase showed that there was a higher level of telomerase activity in early S phase when compared with other points in the cell cycle. These results suggest a relationship between telomerase activity and cell cycle regulation.

Telomere reduction and telomerase inactivation during neuronal cell differentiation.

Telomerase adds (TTAGGG)n hexanucleotide repeats to the ends of mammalian telomeres. This compensates for telomeric loss with successive rounds of cellular replication. Telomerase activity is detected in many neoplastic cells, but not in most normal somatic cells. To determine whether telomeric length and telomerase activity are associated with cellular differentiation, we measured telomeric lengths and telomerase activity in embryonic NT2 precursor cells prior to and following differentiation into post mitotic hNT neurons. This system allows for studies in a direct neuronal cell lineage and, thus, provides a unique model for studying the role of neuronal telomerase activity. Our results show that telomerase activity was present in precursor cells, but not in neuronal cells. Telomeres were consistently longer in NT2 cells than in hNT cells. These results suggest that changes in telomeric length and loss of telomerase activity play a role in neuronal cellular differentiation.

DNA damage and repair in telomeres: relation to aging.

We have established a method for the detection of DNA damage and its repair in human telomeres, the natural ends of chromosomes which are necessary for replication and critical for chromosomal stability. We find that ultraviolet light-induced pyrimidine dimers in telomeric DNA are repaired less efficiently than endogenous genes but more efficiently than inactive, noncoding regions. We have also measured telomeric length, telomeric DNA damage, and its repair in relation to the progression of aging. Telomeres are shorter in fibroblasts from an old donor compared to fibroblasts from a young donor, shortest in cells from a patient with the progeroid disorder Werner syndrome, and relatively long in fibroblasts from a patient with Alzheimer disease. Telomeric DNA repair efficiency is lower in cells from an old donor than in cells from a young donor, normal in Alzheimer cells, and slightly lower in Werner cells. It is possible that this decline in telomeric repair with aging is of functional significance to an age-related decline in genomic stability.

Reciprocal interactions between human ovarian surface epithelial cells and adjacent extracellular matrix.

The human ovarian surface epithelium (OSE), or ovarian mesothelium, is functionally complex as seen by its capacity to proliferate, migrate, and contribute to ovulation and ovulatory repair in response to cyclic hormonal and environmental changes. We wished to determine whether this phenotypic versatility is reflected in cell-extracellular matrix interactions in primary and low-passage culture. Comparisons of cultures maintained on different substrata revealed that these cells form cohesive monolayers on plastic, while fibrin clots enhance cell dispersion, and thus may provide a migratory cue. The cells invaded Matrigel as multicellular aggregates, while collagen gels mediated a morphologic epithelial-mesenchymal conversion. On plastic, the cells produced extracellular matrix components characteristic of epithelial basement membrane (laminin and collagen type IV), as well as stroma (collagen types I and III). In addition, ovarian surface epithelial cells secreted serine proteases and matrix metalloproteinases. The levels of chymotrypsin- and elastase-like proteases were dictated by the substratum: low levels were secreted by cells grown on plastic, intermediate levels on collagen gels and fibrin clots, and most protease was produced on Matrigel. The rate of cell proliferation varied with the substrata and was inversely related to protease secretion. Integrin expression was greatest on plastic and least on collagen gels where integrins were downregulated with time. alpha 6/beta 4 was absent from all cells while varying levels of alpha 2, alpha 3, alpha 5, beta 1, and vitronectin receptor were detected depending on the culture substratum employed. In low-passage cultures of human ovarian surface epithelial cells, then, cell shape, growth, protease production, and integrin expression are modulated by the extracellular matrix. The cells, in turn, alter extracellular matrix by synthesis, lysis, and physical remodeling, and express both stromal and epithelial characteristics. The broad repertoire of these functions may be related to their mesodermal origin, and may reflect the expression of a dual, epithelio-mesenchymal phenotype by relatively immature, uncommitted cells. The results demonstrate the great complexity and versatility of these interactions which render OSE cells capable of participating in numerous physiological and pathological processes.

Characterization of cultured human ovarian surface epithelial cells: phenotypic plasticity and premalignant changes.

BACKGROUND: The ovarian surface epithelium (OSE) is a modified mesothelium that gives rise to most human ovarian carcinomas. In culture, OSE cells tend to assume atypical morphologies that make it difficult to accurately identify normal OSE cells and to recognize pathologic changes. The present study was undertaken to improve the accuracy of OSE identification and to distinguish phenotypic variations of normal OSE cells from early (pre)neoplastic changes. EXPERIMENTAL DESIGN: The expression of epithelial and stromal markers was compared between OSE cultures in low passage, three simian virus 40-immortalized OSE lines (IOSE lines) and two ovarian carcinoma lines, using immunofluorescence microscopy, immunocytochemistry, and Western blots, with fibroblasts and vascular endothelial cells as controls. RESULTS: Whereas keratin remained a convenient and specific epithelial marker for normal OSE, it was not expressed by all cells, and it diminished with passages in culture. E-cadherin and desmoplakins were absent in cultured OSE, mucin was detected in few cells, and microvilli diminished within one to two passages. Laminin and collagen IV were uniformly expressed and stable with time but were also found in endothelial cells. In contrast to endothelial cells, OSE lacked Factor VIII and did not bind Ulex Europaeus Lectin. The three IOSE lines were more stable than OSE morphologically, and keratin was expressed consistently in 100%, 90%, and 0% of the cells, respectively. All IOSE cells produced laminin and collagen IV but lacked E-cadherin. Microvilli persisted in 50% of the cells in one IOSE line and were lacking in the others. The antibody to breast/ovarian carcinoma, 2G3, reacted with few OSE cells but with significantly more IOSE cells. All fibroblast markers tested (vimentin, collagen types I and III, and prolyl-4-hydroxylase) were expressed in OSE and IOSE cultures, concurrently with the epithelial markers. There was no consistent relationship between any of the markers and cell morphology. CONCLUSIONS: Cultured OSE is more accurately identified if the demonstration of keratin is supplemented by 2G3, laminin, or the lack of endothelial markers. The modulation to a fibroblast-like morphology by OSE cells may reflect the expression of their dual epithelio-mesenchymal phenotype rather than epithelio-mesenchymal conversion. Possible indicators of early neoplastic change in immortalized OSE cells include reduced morphologic plasticity and increased 2G3 binding.

A line of rat ovarian surface epithelium provides a continuous source of complex extracellular matrix.

A spontaneously immortalized, yet non-tumorigenic rat ovarian surface epithelial (ROSE 199) cell line, deposits large amounts of extracellular matrix (ECM) in response to crowding. The characteristics and components of ROSE 199-derived cell-free ECM were compared after three different preparative techniques: treatment with 20 mM ammonium hydroxide, with 1% sodium deoxycholate, or by repeated freeze-thaws. The ECMs were analyzed by histochemistry, immunofluorescence, electron microscopy, and Western immunoblotting. Components of ROSE 199 ECM included laminin, fibronectin, and collagen types I and III. Even though ROSE 199 is an epithelial cell line, striated collagen fibers formed a major part of its matrix. Thus, ROSE 199 matrix consists of both basement membrane and stromal matrix components. This matrix supported the adhesion, spreading, and growth of several cell types without altering their morphology or growth pattern, and enhanced the attachment of some cell types that spread on plastic only with difficulty. Immunofluorescence, electron microscopy, and dry weight determinations indicated that a greater proportion of matrix was retained in preparations obtained by ammonium hydroxide or freeze thaw techniques than after sodium deoxycholate treatment. Ammonium hydroxide and freeze-thaw treated matrices were also superior to sodium deoxycholate preparations as evidenced by enhanced initial cellular adhesion and spreading compared to cells plated on plastic. Residual nuclear material did not seem to affect the biological activity of this matrix. ROSE 199 extracellular matrix provides a novel, complex substratum for cell culture and for studies of matrix functions and synthesis.

Human ovarian surface epithelial cells are capable of physically restructuring extracellular matrix.

OBJECTIVE: After ovulation the human ovarian surface epithelium proliferates at the wound edges, migrates over the ovulatory defect, and contributes to its repair primarily by the action of proteolytic enzymes and by the deposition of new matrix material. We examined the potential for human ovarian surface epithelial cells to physically remodel extracellular matrix in culture, similar to collagen gel lattice contraction by fibroblasts, a well-known culture model for wound repair, as an additional role of human ovarian surface epithelium in wound repair. STUDY DESIGN: Human ovarian surface epithelium cells from ovarian biopsies of 11 patients were grown in culture and plated onto a combination of collagen gel and rat ovarian surface epithelial-derived extracellular matrix. The degree of matrix contraction was measured as the percentage of the original culture diameter. RESULTS: Human ovarian surface epithelial cells surrounded and contracted the combination of matrices into a dense matrix organoid. The degree of organoid contraction was related to the number of human ovarian surface epithelial cells plated per organoid and to the inclusion of fibroblasts within the collagen gel but was not affected either by adding epidermal growth factor and hydrocortisone to the culture medium or by reducing the serum component of the medium. CONCLUSION: Human ovarian surface epithelial organoids may be useful for the study of normal and abnormal ovarian events such as ovulatory wound repair and cyst formation.

Simian virus 40-transformed human ovarian surface epithelial cells escape normal growth controls but retain morphogenetic responses to extracellular matrix.

OBJECTIVE: We aimed to improve the availability of experimental models for the study of human ovarian surface epithelium. STUDY DESIGN: Low-passage cultures of human ovarian surface epithelium were transfected with SV40 large- T antigen and the resulting lines were characterized. RESULTS: Three immortalized lines were obtained. They formed epithelial monolayers resembling ovarian surface epithelium in serum-free medium, expressed large-T antigen and overexpressed p53, produced laminin, and were CA 125 negative. Two lines expressed keratin. On plastic surfaces, the growth rate and growth potential of immortalized ovarian surface epithelium were increased over the growth of ovarian surface epithelium, but on extracellular matrices their growth and morphologic features resembled ovarian surface epithelium. The lines were not tumorigenic in Nu/Nu mice. CONCLUSION: The immortalized ovarian surface epithelium lines resemble cells early in neoplastic progression and should be useful to study ovarian carcinogenesis.

Ovarian surface epithelium: autonomous production of connective tissue-type extracellular matrix.

The ovarian surface epithelium (OSE) is known to contribute to postovulatory repair of the ovarian cortex by proliferation and migration over the site of follicular rupture, and by deposition of a basement membrane. We examined the production of other extracellular matrix components in culture by OSE cells of the rat (ROSE), using immunofluorescence microscopy, electron microscopy, and proline incorporation. We compared recently explanted cells in low passage, the immortal line ROSE 239, whose growth pattern resembles low passage cultures, and the immortal line ROSE 199, which forms ridges and papillae. The epithelial nature of all three cell types was confirmed by the presence of keratin and laminin. All three cell types secreted collagen types I and III and at least one (ROSE 199) produced highly polymerized banded fibrils, which are characteristic for stromal or interstitial extracellular matrix. Simultaneous production of collagen types I and III, keratin, and laminin by cloned subpopulations ruled out an origin of the lines in mixed epithelial/fibroblast populations. The results demonstrate that OSE has the capacity to synthesize major components of connective tissue stroma. They suggest that this epithelium, in addition to its postulated proteolytic role, may also express synthetic activity in the remodelling of the ovarian cortical stroma. A capacity of OSE cells to produce stromal components autonomously might be an important factor in the formation of ovarian surface papillae and in neoplastic progression of OSE-derived ovarian carcinomas.

A simplified method to culture human ovarian surface epithelium.

The ovarian surface epithelium (OSE) is thought to give rise to over 85% of human ovarian carcinomas. In spite of its clinical importance, no animal models for the in vivo investigation of this tissue exist, and available culture methods have yielded limited success. In this study, OSE cells from 55 normal ovarian biopsy specimens were used to improve and simplify the methodology for OSE culture and to define the influence of clinical parameters on cultured OSE cells. An improved explanation method was developed which takes advantage of the tenuous attachment of OSE to underlying tissues: the surface epithelium was scraped off the ovarian surface with a rubber scraper, generating epithelial fragments which produced monolayers in culture, with little contamination by other cell types. The scrape method is superior to the explant method previously described (Siemens CH, Auersperg N: J Cell Physiol 134:347, 1988) in terms of speed, simplicity, higher purity of cultures, and increased cell yield. An improved nutrient medium (199/MCDB105/15%FBS) resulted in OSE lines that maintained the original epithelial phenotype for up to 12 population doublings. OSE, detached from the ovary, remained viable if frozen in liquid nitrogen either before culture or in primary culture on strips of plastic, providing OSE independently of the availability of surgical specimens. Growth was not influenced by diagnosis (nonmalignant gynecologic disorders), patient age (mean range: 40.5, 20 to 62 years), or the presence of inclusion cysts or large follicles in the biopsy specimen. This culture system provides conditions for in depth studies of OSE physiology and pathology.

Histochemical classification based upon reaction types and its application to carbohydrate histochemistry.

A method of classification is presented, which divides histochemical visualization reactions into categories based on general reaction types. This scheme is dependent upon the reaction between two elements, the substrate and the probe. The substrate represents a tissue component(s) with one or more reactive groups that can combine directly with the probe. Alternatively, the substrate reactive groups are chemically modified (activation) with a suitable reagent before reaction with the probe. Probes are of three types: those that yield a coloured product, those that result in a colourless product, and those that produce a coloured product only after a further reaction. Methods used in carbohydrate histochemistry are divided into one, two and three probe reactions. Two probe reactions are further subdivided into sequences involving one or two coloured products (one and two dye sequences); three probe reactions into sequences involving one, two or three coloured products (one, two and three dye sequences). This classification permits the rationalization and organization of methods, and provides a framework for the examination of existing methods and the development of new ones.

Heterogeneous expression of keratin, involucrin, and extracellular matrix among subpopulations of a poorly differentiated human cervical carcinoma: possible relationships to patterns of invasion.

Undifferentiated cervical carcinomas vary considerably in their intercellular organization and patterns of invasion. In spite of its clinical significance, the basis for such variation is poorly understood. We investigated the cellular properties that may be responsible for this diversity, using as a model two human cervical carcinoma cell lines that were derived from the same tumor specimen and the same clone. It was shown previously that, in spite of their common origin, each line forms a histologically distinct type of undifferentiated carcinoma when heterotransplanted in vivo: cells of line C-4I grow as compact expanding masses with central necrosis, while tumors of line C-4II infiltrate host tissues as small, well-vascularized, dispersed cell groups. The characteristic behavior of each line was retained in culture, where C-4I cells formed highly multilayered cohesive colonies, while C-4II cells formed diffuse, monolayered colonies and shed into the culture medium. These observations as well as ultrastructural data suggested that each line may be arrested at a different stage of stratified squamous differentiation. In the present study, this hypothesis was tested by examining specific differentiation markers. An analysis of the cultures by immunofluorescence microscopy and immunoblotting revealed that keratin was more abundant in the compact C-4I line than in the dispersed C-4II line. C-4I cells expressed keratins 5, 6, 8, 16, 18, and 19, while C-4II expressed only keratins 8, 16, 18, and 19. In the multilayered C-4I colonies, involucrin-positive cells occurred in the apical cell layers only. In C-4II, involucrin-positive cells occurred in monolayers and domes, and they were most consistently located apically in crowded cultures. Laminin was secreted by both lines, but only C-4II cells deposited a fibronectin matrix. The results suggest that C-4I cells resemble normal cervical cells at the spinous stage of stratified squamous differentiation, while C-4II cells resemble basal/suprabasal cells. The different growth patterns of the tumors, formed by the lines in vivo, therefore likely reflect functional and behavioral differences that normally exist between spinous and basal cervical epithelial cells. The results suggest that differentiation-related functional properties may lead to histological diversity among cervical carcinomas that are categorized as undifferentiated by histopathological criteria.