We previously reported cloning the rLot1 gene, and its human homolog (hLOT1), through analysis of differential gene expression in normal and malignant rat ovarian surface epithelial cells. Both human and rat ovarian carcinoma cell lines exhibited lost or decreased expression of this gene. Interestingly, the LOT1 gene localized at band q25 of human chromosome 6 which is a frequent site for LOH in many solid tumors including ovarian cancer. In this report we have further characterized the potential role of LOT1 in malignant transformation and developed evidence that the gene is a novel target of growth factor signaling pathway. Assays using transient transfections showed that LOT1 is a nuclear protein and may act as a transcription factor. In vitro and in vivo studies involving ovarian cancer cell lines revealed that expression of LOT1 is directly associated with inhibition of cellular proliferation and induction of morphological transformations. Additionally, we show that in normal rat ovarian surface epithelial cells Lot1 gene expression is responsive to growth factor stimulation. Its mRNA is strongly down-regulated by epidermal growth factor receptor (EGFR) ligands, namely EGF and TGF-α. Blocking the ligand-activated EGFR signal transduction pathway by the specific EGF receptor inhibitor, tyrphostin AG1478, and the MEK inhibitor, PD098059, restores the normal level of Lot1 gene expression. It appears that the regulation of Lot1 gene is unique to these ligands, as well as the growth promoting agent TPA, since other factors either did not affect Lot1 expression, or the effect was modest and transient. Altogether, the results suggest that Lot1 expression is primarily mediated via EGF receptor or a related pathway and it may regulate the growth promoting signals as a zinc-finger motif containing nuclear transcription factor.
Mammalian cells normally utilize highly coordinated signaling mechanisms involving growth factors, protooncogenes, and tumor suppressor genes to correctly mediate the repair and maintenance of tissue architecture (Abdollahi et al., 1991; Arends and Wyllie, 1991; Lord et al., 1991; Weinberg, 1991). However, increased cell division due to a variety of environmental cues can increase the probability of selecting cells with a higher growth capacity and malignant phenotype (Ames and Gold, 1990; Henderson et al., 1991; Preston-Martin et al., 1990). In many instances, the cancer cells arise as a result of genetic aberrations in growth factor signaling pathways, including unregulated action of receptors, ligands, or components of their pathway (Aaronson, 1991). These pathways can cross talk with signals sensed by other receptors such as extracellular matrix or cell surface molecules on adjacent cells. In fact, loss of the requirement for growth factor, extracellular matrix, and contact inhibition are hallmarks of malignant transformation (Batt and Roberts, 1998).
Among the growth factor receptors, alteration in epidermal growth factor receptor (EGFR) family signaling has been one of the most frequently implicated effectors of human oncogenesis, since the initial discovery that the v-erbB product is a truncated form of the EGF receptor (Aaronson, 1991; Downward et al., 1984). The EGFR protein is a 170-kDa transmembrane glycoprotein which upon binding to its ligands, EGF (Carpenter, 1981) or TGF-α (Derynck, 1988; Massague, 1983), stimulates a network of mitogenic signaling pathways that includes altering activities of a variety of kinases, phosphatases, G-proteins, adaptors, and transcription factors (Ullrich and Schlessinger, 1990). These growth factors are widely expressed by tumor cells including the ovarian cancers (Bauknecht et al., 1986; Kommoss et al., 1990) and could act in an autocrine fashion to promote tumor progression (Hunter, 1984). The EGFR gene is often amplified or overexpressed, or both, in diverse types of tumors including glioblastoma (Liebermann et al., 1985; Moscatello et al., 1995; Yamamoto et al., 1986) and carcinomas of the ovary (Berchuck et al., 1990; Ilekis et al., 1997; Morishige et al., 1991; Rodriguez et al., 1991; Scambia et al., 1992), breast (Sainsbury et al., 1995), and bladder (Mellon et al., 1995). In addition, expression of truncated EGFR has been observed in many of these tumors and its presence can adversely affect prognosis (Eskrand et al., 1994; Gotoh et al., 1994; Ilekis et al., 1997; Petch et al., 1990; Xu et al., 1984).
Ovarian cancer most frequently develops from the epithelial cells on the surface of the ovary and is believed to involve multiple genetic changes (Godwin et al., 1992; Tenaka et al., 1989). After each ovulation the surface epithelial cells become stimulated to replicate and repair the surface wound caused by ovum release. Uninterrupted cycles of such cell replacement or wound repair mechanism has been suggested as the driving force in ovarian oncogenesis (Casagrande et al., 1979; Fathalla, 1971). We recently developed experimental support for this concept using a rat model and are searching for the genetic alterations leading to malignancy in this model (Abdollahi et al., 1997a,b; 1999; Godwin et al., 1992; Testa et al., 1994). As a result, we identified the Lot1 gene and based on its pattern of lost or decreased expression, we suggested that it might function as a tumor suppressor gene (Abdollahi et al., 1997a,b). The Lot1 gene and its human homolog (LOT1) encode a ∼66 and 55 kDa protein, respectively, with seven C2H2-type zinc finger motifs tandemly located in the amino-terminal half of the protein. Other features of interest in the LOT1 protein include regions rich in glutamic acid, proline, and glutamine which are often characteristic and of functional importance in transcription factors (Courney and Tijan, 1988; Mermod et al., 1988; Williamson, 1994) and the presence of -P-X-S-P- and -H-S-P-Q-K- sequences as a potential phosphorylation substrate for MAP kinases (Gonzalez et al., 1991) and cdc2 kinase (Moreno and Nurse, 1990), respectively. Additionally, the gene is ubiquitously expressed in different tissues and cells tested, including ovary, brain, breast, liver, kidney, pancreas, uterus, testes, spleen, thymus, small intestine, leukocyte, and colon (Abdollahi et al., 1997a,b). It was also interesting to observe that the LOT1 gene localizes to the human chromosome 6 band q25, which is a frequent site for LOH in many tumors (Abdollahi et al., 1997b; Colitti et al., 1998; Fujii et al., 1996; Thrash-Bingham et al., 1995). Based on these data, we set out to further characterize the gene's function and its potential role in the biology of ovarian cancer.
By transfecting cells with LOT1 fusion proteins we show that LOT1 is a nuclear protein which may act as a transcriptional activator. Lot1 mRNA expression is found to be strongly down-regulated by the EGFR ligands, EGF and TGF-α, in the ovarian surface epithelial cells and dependent upon protein synthesis. The tumor promoting agent 12-O-tetradecanoylphorbol-13-acetate (TPA) also exerts a similar effect on Lot1 expression, while other factors or proteins apparently either have no effect on Lot1 mRNA expression or the response is weak or transient. Introduction of exogenous LOT1 cDNA spanning the complete coding region into cells decreased growth rate of the human ovarian cancer cell line, A2780, and the rat malignantly transformed ovarian surface epithelial cell line, NuTu 26, in vitro and in vivo. Taken together, these results suggest that the LOT1 gene is a novel target/part of the epidermal growth factor receptor signaling pathway or a related pathway and may play a significant role as a zinc-finger transcription factor that modulates growth suppression.
LOT1 gene encodes a nuclear protein with a potential role as a transcription factor
Analysis of the amino acid sequences from both the human and rat full-length cDNA clones suggested that the protein may have a role as a transcription factor and as such should be localized in the nucleus. To examine this possibility, we transfected Swiss 3T3 fibroblasts with a Flag-tagged pcDNA3/LOT1 expression construct. Immunofluorescence staining with the antibody directed to the Flag epitope and use of confocal microscopy demonstrated that LOT1 is a nuclear protein. The subcellular localization was confirmed by Hoechst staining of the same cells (Figure 1a). Further confirmation of the intracellular distribution of LOT1 protein was accomplished by using a GFP-LOT1 construct which was transfected into NIH3T3 fibroblasts. We found that in contrast to GFP alone, which is distributed diffusely throughout the cell, the GFP-LOT1 protein is located mainly in the nucleus (Figure 10b).
To evaluate the transcriptional regulatory function of LOT1, we fused the carboxyl terminal domain (amino acid residues 208 – 465) of the protein lacking the complete seven zinc finger region to the carboxyl terminus of a bacterial DNA-binding protein, LexA (Brent and Ptashne, 1985) (Figure 1b). Co-transfection of LexA or the fusion protein construct, LexA-LOT1, with either tk-CAT or LexA operator tk-CAT reporter constructs into NIH3T3 cells revealed that only co-transfection with LexA-LOT1, but not LexA, activates the Lex operator tk-CAT reporter (Figure 1b). These results demonstrate that the carboxyl terminus of the LOT1 protein can transform Lex protein into a sequence-specific transcription factor and that the non-zinc-finger region serves as an activator domain. This notion is further supported by a recent report published during the preparation of this manuscript (Kas et al., 1998). The report compared members of the LOT1 family of zinc-finger proteins, including PLAG1, and found strong LOT1-mediated transcriptional capacity of GAL4-luciferase reporter constructs, particularly in the epithelial cells.
LOT1 is a target/part of the epidermal growth factor receptor signaling pathway
The idea that LOT1 may be involved in a growth factor signaling pathway was based on the following considerations: (i) EGF and its homolog, TGF-α, are among the major growth factors implicated in the ovarian cancer; (ii) many transcription factors are components of growth factor receptor pathways; (iii) loss of growth factor-responiveness is one of the important hallmarks of malignant transformation; (iv) LOT1 expression is lost or decreased in the transformed ovarian surface epithelial cells. We initiated experiments to test this hypothesis using normal rat ovarian surface epithelial cells. Treatment of the normal cells with EGF at a concentration as little as 5 ng/ml significantly down-regulated the expression of Lot1 (Figure 2a). Interestingly, the result was found not to be affected by serum in the culture medium. A time course study showed that EGF-mediated down-regulation of Lot1 mRNA expression occurs at 3 h post treatment and, unlike the c-Jun and c-Fos proto-oncogenes which demonstrate a maximum change in expression at about 1 h (Figure 2b), Lot1 remains down-regulated at 24 h (Figure 2b) or even longer (48 h; data not shown) in the presence of EGF. Treatment of the cells with TGF-α also significantly decreased Lot1 expression, while the effect of TGF-β was partial and transient (Figure 3a,b and unpublished data). Interestingly, TPA which is a tumor promoting agent also significantly lowered the expression of Lot1 mRNA (Figure 3b). Lot1 expression was not affected by other factors tested (Figure 3a,b and unpublished data). These include PDGF, NGF, VEGF, TNF-α, IL-6, and the presence of insulin in the culture medium. These data suggest that LOT1 is involved in a specific pathway originating from EGFR.
Protein synthesis inhibition blocks the EGF-mediated down-regulation of Lot1 mRNA
The time course study described above (Figure 2b) suggested that early events such as protein synthesis may be necessary for the regulation of Lot1 gene expression by EGF. To test this hypothesis, we pretreated the normal rat ovarian surface epithelial cells with cycloheximide for 15 min before stimulation with EGF and found that cycloheximide blocks the EGF-mediated down-regulation of Lot1 mRNA (Figure 4). Northern analysis of Lot1 mRNA at 6 h post treatment with EGF showed that addition of cycloheximide at different time points, up to 3 h, could reverse the down-regulation response (data not shown). It appears that the cycloheximide effect is not due to superinduction mechanism since the protein synthesis inhibitor alone for 4 h compared with the control (no CHX; no EGF) did notably increase the c-Jun, but not Lot1, mRNA expression (Figure 4). These data indicate that an early mechanism involving protein synthesis (Pontecorvi et al., 1988) is essential for the regulatiaon of Lot1 gene expression. It also appears that the Lot1 RNA is highly stable as is evidenced by treating the cells with actinomycin D to block freshly synthesizing transcripts (Figure 4b).
Blocking the EGFR signaling pathway restores the normal level of Lot1 gene expression
Binding of the ligand to the EGFR results in activation of the receptor's kinase activity and leads to autophosphorylation of several tyrosines in the receptor protein (Schlessinger and Ullrich, 1992). These initial events trigger a phosphorylation cascade utilizing mitogen-activated protein kinases (MAPKs) (Aaronson, 1991; Ahn et al., 1990). These MAPKs (also called extracellular signal-regulated kinases, or ERKs) are themselves activated by phosphorylation by the MAPK-activating enzymes (MAPK/ERK kinases, or MEKs) (Alessandrini et al., 1992; Zheng and Guan, 1993). These signals traverse the cytoplasm and converge in the nucleus to alter specific gene expression. Tyrphostin AG1478 specifically inhibits EGFR tyrosine kinase activity (Fry et al., 1994) and PD098059 prevents activation of MAPK by selectively inhibiting MEK (Dudley et al., 1995). Therefore, we used these compounds to block the EGFR signaling pathway and then monitor effects on Lot1 gene expression. The results are shown in Figure 5a. Both these inhibitors effectively reversed the EGFR ligand-mediated Lot1 down-regulation, providing additional evidence for involvement of this growth factor signaling pathway in Lot1 gene regulation. It is also noteworthy that the expression effects coincided inversely with the pattern of cell growth characteristics exerted by each treatment, suggesting a relationship between the growth stimulation and down-regulation of Lot1 expression (data not shown). Treatment of the cells with anisomycin (20 ng/ml) apparently did not mimic the EGF-mediated response of Lot1 expression (Figure 5b), suggesting that anisomycin-activated kinases (JNK/SAPK and p38/RK) are not sufficiently involved in this process (Cano et al., 1994; Hazzalin et al., 1998).
The protein kinase C (PKC) inhibitors, H7 and calphostin C, did not block the EGF stimulation of the Lot1 down-regulation, suggesting that the pathway from EGF to Lot1 expression is fed downstream of PKC and is independent of the PKC activity (Figure 5b). This possibility is currently being investigated. Genistein (a protein tyrosine kinase inhibitor) and PP1 (a Src protein tyrosine kinase inhibitor) also did not affect the EGF response, indicating that these tyrosine kinases may not be involved. However, okadaic acid which is a potent phosphatase 1 and phosphotase 2A inhibitor reversed the action of EGF (Figure 5b), suggesting a potential role for these or similar okadaic acid-sensitive phosphatase(s) in the pathway leading to Lot1 down-regulation. Forskolin (20 μM) which increases cAMP levels by directly activating adenylate cyclase and wortmannin, an inhibitor of phosphoinositide 3-kinase (P13K; 200 nM), did not abolish EGF-mediated down-regulation of Lot1 expression (data not shown).
LOT1 is involved in negative growth control and may function as a tumor suppressor gene
The above results and the findings described previously (Abdollahi et al., 1997a,b) were suggestive of a possible growth-inhibitory role for LOT1. To directly test the negative effect on cell growth, we transfected the human ovarian cancer cell line, A2780, and the rat transformed surface epithelial cell line, NuTu 26, with LOT1 cDNA spanning the complete coding region and then selected for G418 resistant clones. We previously observed that endogenous expression of LOT1 in A2780 cells is dependent on the growth phase with the lowest expression at the exponential phase (data not shown). Therefore, it was of interest to determine the effect of constitutive LOT1 expression in this cell line. The positive cells were identified by Northern blot analysis and then confirmed by two consecutive PCR amplifications of the genomic DNA using two sets of primers, one flanking the whole cDNA insert and the other as nested primers (see Materials and methods). The first PCR showed an expected size fragment (1.5 kbp) in the LOT1 transfected cells but not in vector transfected cells (data not shown). The next amplification reaction also produced the expected size (0.5 kbp) fragment in the transfected cell line versus the control cells, confirming LOT1 cDNA insertion into the genome. These PCR fragments would not be produced in non-transfected cells due to the presence of a large (∼5.5 kbp) intron embedded between the forward and reverse primers used for the amplification reactions (Roberts et al., 1999). A representation of such analysis and PCR amplifications is given in Figure 6a.
Two of the positive cell lines derived from A2780, designated as AL-25 and AL-33, were cultured in the presence of G418 and growth was monitored and compared to the non-LOT1 transfected (A2780-V) cell line. The results showed that overexpression of exogenous LOT1 significantly suppressed the growth of A2780 ovarian cancer cells (Figure 6b). In addition, a growth kinetic assay for AL-33 cells in relation to A2780-V cells revealed a marked delay of growth rate characteristics for the transfected cell line (Figure 6b). In other words, the mathematically derived doubling time was approximately 28 h for the AL-33 transfectant and 19 h for the A2780-V control cell line.
In another set of experiments we used NuTu 26 transformed cell line which has significantly down-regulated expression of the Lot1 gene (Abdollahi et al., 1997a) and its normal progenitor cells in which the endogenous expression of Lot1 gene is abundant and has remained intact. We introduced the Lot1 cDNA into the transformed cell line and obtained cells which had expression of the gene at levels comparable to that observed in its progenitor normal cells (Figure 7). The results show that constitutive expression of the Lot1 gene in the transfectants (EN-11, NL-10, and NL-14) was capable of significantly diminishing growth rate of the cells. The fact that the NuTu 26 cell line selected for transfection was derived from related and normal progenitor cell line in which the endogenous Lot1 is intact indicates that the growth suppression is a consequence of restoring the Lot1 expression. To determine whether LOT1 overexpression can suppress or delay in vivo tumor formation by A2780 or NuTu 26 cells, nude mice (Scid mice) were injected subcutaneously with the cells on opposite flanks and tumor growth was monitored and recorded for 4 weeks. The results indicated that LOT1 overexpression is associated with marked inhibition of the ability of these cancer cells to form tumors (Figure 8). Phase-contrast and fluorescence microscopic examinations also showed that the LOT1-transfected cell lines exhibit an altered morphological characteristic (e.g., cell rounding or membrane blebbing) as compared to their corresponding control or vector transfected cells (Figures 9 and 10a,b). The alterations in morphology and pattern of growth inhibition were also evidenced in cell cultures (plus G418) of the tumors or nodules examined about 4 weeks after the injection (data not shown).
This report presents evidence supporting the potential function of LOT1 as a nuclear transcription factor involved in modulation of cell growth. We have shown both in vitro and in vivo that LOT1 expression is significantly associated with the suppression of proliferation. The evidence provided here suggests that the growth-regulatory potential of LOT1 is likely due to its involvement in mediating a growth factor signal transduction pathway since the growth promoting agents such as EGF, TGF-α, or TPA down-regulate Lot1 gene expression in normal ovarian surface epithelial cells. Inhibitors of the EGFR pathway, tyrphostin AG1478 and PD098059, blocked the growth factor-stimulated down-regulation of Lot1 expression, further substantiating a role of the mitogenic pathway in Lot1 gene regulation. Treatment of the cells with the inhibitors of protein and mRNA synthesis suggest that the Lot1 message is relatively stable but its EGF-mediated down-regulation may involve labile mRNase(s) which is (are) activated by the growth factor and disappear or become inactivated rapidly following inhibition of protein synthesis (Pontecorvi et al., 1988).
Growth factors cause quiescent cells to advance into the first gap phase (G1) of the cell division, traverse the G1 phase and then become committed to DNA synthesis or S phase (Aaronson, 1991; Pardee, 1989). The cells require sustained exposure to EGF for at least 6 – 8 h before they are committed to DNA synthesis and transition through the G1 phase (Aharonov et al., 1978; Carpenter and Cohen, 1976; Schechter et al., 1978; Weinberg, 1991). In this context, we hypothesize that LOT1, like other growth suppressor genes, may serve as a transducer of anti-proliferative signals stopping progression through the cell cycle (Weinberg, 1991). Loss of LOT1 expression caused by malignant transformation or persistent EGF stimulation (more than 6 h) may release cells from the resting or quiescent state. Disruption of EGF-generated signals by the inhibition of protein synthesis or removal of the growth factor would reverse cells to the G0 state. Interestingly, unlike the immediate early response genes such as c-Jun, c-Fos, and c-Myc which show a peak induction at about 1 h post EGF treatment, Lot1 transcript remains undetectable or low for up to 48 h when the mitogen remains present. This observation further supports the above hypothesis.
It appears that regulation of Lot1 expression may involve the MAPK pathway since the signals from EGF, TGF-α, and TPA are transduced through activation of the MAPK pathway (Howe et al., 1992), predominantly Erk1 and Erk2 (Kyriakis et al., 1994), and the specific inhibitor of MAPK activation, i.e. PD098059, effectively blocks the EGFR ligand-mediated Lot1 down-regulation (this report). Anisomycin treatment at subinhibitory concentration (20 ng/ml) did not mimic the EGF response of Lot1, suggesting anisomycin-activated kinases such as JNK/SAPK and p38/RK are not involved in this process (Cano et al., 1994; Hazzalin et al., 1998). This notion is further supported by the ineffectiveness of TNF-α, which is able to stimulate stress-activated protein kinases (SAPKs) (Kyriakis et al., 1994), in down-regulating Lot1 expression (data not shown).
Treatment of the cells with genistein (a protein tyrosine kinase inhibitor) and PP1 (a Src tyrosine kinase inhibitor) did not reverse the action of EGF on Lot1 gene. Whereas, okadaic acid (an inhibitor of phosphatase 1 and 2A) (Cohen et al., 1990) was able to inhibit the EGF response, suggesting a potential role for these or similar okadaic acid-sensitive phosphatase(s) in the pathway leading to Lot1 down-regulation. Although okadaic acid was first discovered as a tumor promoter (Suganuma et al., 1988), the above hypothesis may be reasonable to believe because okadaic acid can act differently (tumor versus anti-tumor activity) in different cell types and with varied dose (Hunter, 1995). Therefore, it has been proposed that the substrates for okadaic acid may include phosphoproteins that act downstream of MAPK but in a fashion opposite to its proliferative potential (Casillas et al., 1993; Cheng et al., 1998).
The protein kinase C (PKC) inhibitors, calphostin C and H7, were ineffective in blocking the EGF-mediated down-regulation of Lot1 in the ovarian surface epithelial cells, suggesting that the pathway from EGF to Lot1 is fed downstream of PKC. Similarly, Coughlin et al. (1985) reported that EGF is able to induce c-Myc expression in fibroblasts without activating PKC and proposed the existence of an alternate pathway from the growth factor receptor to c-Myc induction that is independent of PKC. Therefore, both EGF and TPA may activate a common pathway, likely downstream of PKC, leading to the Lot1 gene regulation. In other words, TPA may initiate a mitogenic response that mimics an early response known to be elicited by EGF (Moon et al., 1984). PKC-independent activation of the receptor tyrosine phosphorylation and other proteins has been shown previously (Emkey and Kahn, 1997; Moon et al., 1984). On the other hand, isozyme-specific transcriptional activation of c-Fos has been observed in mouse mammary epithelial cells (Kampfer et al., 1998). Work addressing alternate pathways to Lot1 regulation is in progress.
Correct cellular localization of the LOT1 protein, like other transcriptional factors, may be very critical to its normal function. These nuclear proteins may either carry a nuclear localization sequence (NLS) or interact with another protein that contains a NLS (Dingwall et al., 1982; Jans and Hassan, 1998; Li et al., 1998; Zhou and Padmanabhan, 1988). Presently, we can not exclude any of these possibilities since we have not yet examined the presence of a potential NLS in LOT1 protein. The presence of proline-rich sequence in LOT1 protein (Abdollahi et al., 1997a,b) probably suggests an interaction between LOT1 and other protein (Williamson, 1994). Also, we do not yet know the mechanism underlying the transcriptional activity elicited by LOT1 protein in the nucleus. It could be argued, therefore, that our observations provide a basis to believe that LOT1 may signal downstream genes to yield growth arrest associated with lack of EGFR signaling activation. LOT1 may activate transcription of another growth suppressor gene such as a cell cycle inhibitor protein, a protein involved in the cell's cytoskeletal organization or cell-cell interaction, or it may serve a role in suppressing an oncogene by indirectly activating a repressor protein. In any case, elucidation of the mechanism(s) leading to the transport of LOT1 into the nucleus and its role as a transcription factor may yield useful functional information.
Studies by other groups have identified a LOT1 family of zinc-finger motif containing proteins (Kas et al., 1997; Nomura et al., 1994; Spengler et al., 1997). Spengler et al. (1997) identified Zac1, which is most likely the mouse homolog of LOT1 based on their high sequence homology, in the pituitary gland and brain. They suggested a potential role of the gene product in apoptosis and involvement in cell cycle arrest. Based on our observations and the findings by Spengler et al. (1997), we hypothesize that in normal cells LOT1 may act primarily as a growth arrest gene which is capable of switching to an apoptotic gene if its expression is accompanied by the change in activity of another gene due to a gene alteration or modification of growth conditions. Kas et al. (1997) reported identification of PLAG1 as a developmentally regulated zinc finger gene located at the human chromosome 8 band q12, as a putative pleiomorphic adenoma gene (PLAG) in salivary gland epithelial cells. All these genes are highly homologous in the zinc-finger (amino terminal) domain. It is possible that they may either function in a tissue specific manner or they may compete and/or interact with each other in binding a specific DNA sequence or region to regulate gene transcription. Alternatively, each gene may be involved in independent or specific process of differentiation or development. However, the LOT1 gene was found to be ubiquitously expressed, suggesting a likely important role in other tissues.
In summary, we conclude that LOT1 may serve as a tumor suppressor nuclear transcription factor down-regulated by some specific tumor promoting agents. Although the exact mechanism responsible for the growth factor-mediated Lot1 mRNA down-regulation is unknown and is presently being investigated, it is rational to believe that any upstream genetic or functional alteration(s) which constitutively activates the EGFR mitogenic signaling pathway(s) can lead to a decreased or lost expression of the Lot1 gene. This may include mutations in the LOT1 promoter and/or silencing by gene methylation and/or histone deacetylation. Interestingly, unlike many proto-oncogenes such as c-Jun, c-Fos, and c-Myc which are responsive to serum or a large variety of growth factors, Lot1 down-regulation is apparently unique and is responsive to only a limited number of growth factors. It is possible that different growth factors use different signaling pathways such as the one transduced through LOT1 gene to mediate cell growth (Rani et al., 1997). Such divergence of growth factor signaling pathways suggests a potential therapeutic target which could by-pass many upstream signals common to diverse growth factors, cytokines, or metabolic networks.
Materials and methods
Cells and cell cultures
Normal rat ovarian surface epithelial cells were obtained from the ovaries of adult Fisher rats by selective trypsinization (Godwin et al., 1992; Testa et al., 1994). The cells were maintained in Dulbecco's modified Eagle's medium (Gibco-BRL, Gaithersburg, MD, USA) supplemented with 4% fetal bovine serum, glutamine (2 mM), insulin (10 μg/ml), penicillin (100 units/ml) and streptomycin (100 μg/ml) in a humidified 5% CO2 atmosphere at 37°C. For growth factor treatment, the cells were used 5 – 7 days later when they were almost confluent and quiescent. The human ovarian cancer cell line A2780 was grown in RPMI (10% fetal calf serum) as previously described (Abdollahi et al., 1997b). For transfections, the cells were seeded 1 day earlier to give a confluency of about 60%.
EGF, TGF-α, PDGF-AA, NGF, TNF-α, and IL-6 were from Gibco-BRL. TGF-β, Endothelin-1, TPA, VEGF, cycloheximide, actinomycin D, forskolin, and wortmannin were obtained from Sigma (St. Louis, MO, USA). Tyrphostin AG1478 was from Calbiochem (LaJolla, CA, USA) and PD098059 was obtained from Research Biochemicals International (Natick, MA, USA). The compounds PP1, genistein, anisomycin, okadaic acid, H7, and calphostin C were from Biomol (Plymouth Meeting, PA, USA). All the primers used were synthesized at Fox Chase Cancer Center.
Transfection, chloramphenicol acetyltransferase (CAT) assay, immunofluorescence and image analysis
All transfections were performed by using the TransIT-LT1 (PanVera Co., Madison, WI, USA) or Lipofectamine (Gibco-BRL) reagents and following the protocols recommended by the suppliers. Transient transfections were performed in 6-well plates with 2×105 cells/well. The cells were washed after 48 h and resuspended in 1×Reporter Lysis buffer (Promega, Madison, WI, USA). The lysates were centrifuged for 2 min and the supernatant was analysed for CAT activity by thin-layer chromatography. Cell lysates were normalized for transfection efficiency by β-galactosidase assay. Four transfections were performed per each assay (plasmid combination) and were repeated twice. To establish stably transfected A2780 and NuTu 26 cell lines, the cells were selected in the presence of G418 at a concentration of 500 μg/ml after transfection with the pUHD/LOT1 or the control plasmid.
For immunofluorescence staining and subcellular localization of the Flag-tagged LOT1 fusion protein, the transfected Swiss 3T3 cells were cultured on glass coverslips to ∼50% confluence. The cells were washed in PBS, fixed in 4% paraformaldehyde for 10 min, and then washed again with PBS. The cells were permeabilized by treatment with 0.2% Triton X-100 for 10 min and then exposed to 10 μg/ml anti-Flag M2 monoclonal antibody (Kodak, New Haven, CT, USA) in the same buffer for 1 h. Following two washes (10 min each) with 0.2% Triton X-100, the cells were treated for 30 min with fluorescein isothiocyanate (FITC)-conjugated anti-mouse IgG antibody (Sigma) in 3% BSA in PBS. The cells were washed twice with 0.1% Triton X-100 in PBS and once with PBS. The FITC was visualized at ∼525 nm using a Carl Zeiss fluorescence microscope. Nuclei were stained with Hoechst (Sigma) staining procedure. For detection of GFP, the cells were viewed using Nikon Eclips TE300 microscope equipped with a GFP Chroma excitation filter. All the fluorescence images were captured using Quintix cooled charge-coupled (CCD) camera (Photometrics, Tucson, AZ, USA).
Plasmid construction and cloning
LOT1 cDNA (A1H2) for subcellular localization and transfections was prepared by PCR using FB4-2 cDNA clone (Abdollahi et al., 1997b) as the template and primers A1 (5′-ATGGCCACGTTCCCCTG-3′) and H2 (5′-TCAATTATCTGAATGCATG-3′) with EcoRI and XhoI linkers added, respectively, for amplification of the coding region. The resulting fragment was ligated directly into pcDNA3 vector to produce pcDNA3/LOT1 plasmid. Alternatively, a pcDNA3-Flag plasmid was constructed first by inserting the Flag epitope (Kodak) into pcDNA3 vector at BstXI-EcoRI site, followed by ligation of the LOT1 cDNA fragment (A1H2), in frame with the 3′ Flag epitope, into the EcoRI-XhoI site to create pcDNA3/Flag-LOT1 plasmid. For construction of GFP-LOT1 mammalian expression vector, the GFP coding region from pFRED143 (KH1035) plasmid (a gift of Jun-Hsiang Lin) was amplified using GFP forward (5′-CGCCATGGCTAGCAAGGGCCAG-3′) and reverse (5′-CTTGTACAGCTCGTCCATGCC-3′) primers. The PCR product was ligated with XbaI linker to the LOT1 coding region and the resulting fragment was introduced downstream of CMV promoter into XhoI site of pcDNA3 plasmid.
To generate pcDNA3/Lex-LOT1 plasmid, we first synthesized I1H2 fragment (amino acid residues 208-465 in the LOT1 protein) by PCR using FB4-2 cDNA clone as the template and primers I1 and H2 (I1: 5′-GAGCTGATGAAAGAGAGCT-3′; H2: same as above) for amplification. Using these primers we obtained truncated LOT1 cDNA fragment which excluded all seven zinc finger motifs at the amino-terminal end of the protein. The I1H2 fragment was cloned into EcoRI – XhoI restriction enzyme sites of pEG202 plasmid (a gift of Erica Golemis) downstream of LexA DNA binding domain. The resulting construct was cut with HindIII and XhoI enzymes and the fragment containing LexA DNA binding domain plus I1H2 was inserted into pcDNA3 to produce pcDNA3/Lex-LOT1 plasmid. The LexA expression vector was constructed with a LexA fragment (0.8 kbp) flanked by HindIII and XhoI restriction enzyme sites of pEG202 plasmid. The empty tkCAT vector or containing a single (Lex-1-tkCAT) or double (Lex-2-tkCAT) LexA operator were a kind gift of H Leighton Grimes (Grimes et al., 1996). To construct pUHD/LOT1 plasmid, EcoRI – XbaI fragment of the LOT1 cDNA PCR product spanning the whole coding region was inserted into EcoRI – XbaI restriction sites of the vector pUHD10.3 (a gift of M Gossen and H Bujard) downstream of the CMV promoter sequence. All the plasmids were prepared using Qiagen kit (Valencia, CA, USA) and sequenced by the Fox Chase Cancer Center sequencing facility.
Northern analysis, probes, genomic DNA, and PCR amplification
Total RNA was isolated from the cells by the guanidinium isothiocyanate extraction method (Chomczynski and Sacchi, 1987) and was separated on 1% agarose gels containing 2.2 M formaldehyde. The RNA was transferred to Nylon membranes (Micron Separations, Inc.) by capillary action and hybridized using a procedure described before (Abdollahi et al., 1997a,b). The probes for visualization of rat Lot1 and human LOT1 transcripts were the same as described previously (Abdollahi et al., 1997a,b). The probe for β-actin was from Clontech, Inc. (Palo Alto, CA, USA) and the probes for c-Jun and c-Fos were as described previously (Lord et al., 1993).
High molecular weight DNA was isolated from the transfected cells collected in the lysis buffer containing 10 mM tris-HCl, 150 mM NaCl, 10 mM EDTA, 0.1% SDS, 100 μg/ml proteinase K (Maniatis et al., 1989). The lysate was phenol/chloroform extracted and precipitated with sodium acetate/ethanol. The purified total genomic DNA was used for PCR amplification and agarose gel electrophoresis. Two consecutive amplification reactions, using two different sets of primers, were carried out on the genomic DNAs to visualize the incorporation of transfected LOT1 cDNA. First set of primers (Forward, 5′-GGAGACGCCATCCACGCTGT-3′ and Reverse, 5′-TCACTGCATTCTAGTTGTGG-3′) was selected on the vector flanking the LOT1 cDNA. The first PCR product was then diluted (1 : 10) and reamplified using the second set of nested primers (P61: 5′-TCCTCACCCTGCAGAAGTTCACG-3′ and p65: 5′-GTCGCACATCCTTCCGGGTGTAG-3′) within a region of LOT1 cDNA. PCR amplifications were carried out in a final volume of 40 μl and contained 100 ng of template, 60 μM of dNTPs, 1×PCR buffer, 500 nM primer, 5% DMSO, and 2.5 units of Amplitaq DNA Polymerase (Perkin-Elmer Corp.). The cycling conditions consisted of 95°C for 5 min, 35 cycles of 94°C for 30 s, 55°C for 30 s, 72°C for 1 min followed by extending at 72°C for 5 min.
To test for in vivo tumorigenicity, AL-25, AL-33, or EN-11 cell lines overexpressing the exogenous LOT1 cDNA and the parental cells (A2780 or NuTu 26) transfected with the empty vector (control) were injected into the opposite flanks of Scid mice (6 – 8 weeks of age); Female C.B. 17/Icr Scid mice were bred and maintained at Fox Chase Cancer Center. 5×106 cells were resuspended in 0.2 ml PBS and injected subcutaneously. Tumor growth was monitored daily and recorded using an electronic caliper.
Aaronson S. . 1991 Science 254: 1146–1153.
Abdollahi A, Getts L, Sonoda G, Miller P, Taguchi T, Godwin A, Testa and Hamilton T. . 1999 J. Soc. Gynecol. Invest. 6: 32–40.
Abdollahi A, Godwin A, Miller P, Getts L, Schultz D, Taguchi T, Testa J and Hamilton T. . 1997a Cancer Res. 57: 2029–2034.
Abdollahi A, Lord K, Hoffman-Liebermann B and Liebermann D. . 1991 Oncogene 6: 165–167.
Abdollahi A, Roberts D, Godwin A, Schultz D, Sonoda G, Testa J and Hamilton T. . 1997b Oncogene 14: 1973–1979.
Aharonov A, Pruss R and Herschman H. . 1978 J. Biol. Chem. 253: 3970–3977.
Ahn N, Weiel J, Chan C and Krebs E. . 1990 J. Biol. Chem. 265: 11487–11494.
Alessandrini A, Crews C and Erikson R. . 1992 Proc. Natl. Acad. Sci. USA 89: 8200–8204.
Ames G and Gold L. . 1990 Proc. Natl. Acad. Sci. USA 87: 7772–7776.
Arends M and Wyllie A. . 1991 Int. Rev. Exp. Pathol. 32: 223–254.
Batt D and Roberts T. . 1998 J. Biol. Chem. 273: 3408–3414.
Bauknecht T, Kiechle M, Bauer G and Siebers J. . 1986 Cancer Res. 46: 2614–2618.
Berchuck A, Olt G, Everitt L, Soisson A, Bast R and Boyer C. . 1990 Obstet. Gynecol. 75: 255–262.
Brent R and Ptashne M. . (1985) Cell 43: 729–736.
Cano E, Hazzalin C and Mahadevan L. . 1994 Mol. Cell. Biol. 14: 7352–7362.
Carpenter G. . 1981 Hdbk. Exp. Pharmacol. 57: 90–126.
Carpenter G and Cohen S. . 1976 J. Cell Physiol. 88: 227–238.
Casagrande J, Pike C, Ross R, Louie E, Roy S and Henderson B. . 1979 Lancet ii: 170–173.
Casillas A, Amaral K, Chegini-Fraahani S and Nel A. . 1993 Biochem. J. 290: 545–550.
Cheng A, Balczon R, Zuo Z, Koons J, Walsh A and Honkanen R. . 1998 Cancer Res. 58: 3611–3619.
Chomczynski P and Sacchi N. . 1987 Analyt. Biochem. 162: 156–159.
Cohen P, Holmes C and Tsukitami Y. . 1990 Trends Biochem. Sci. 15: 98–102.
Colitti C, Rodabaugh K, Welch W, Berkowitz R and Mok S. . 1998 Oncogene 16: 555–559.
Coughlin S, Lee W, Williams P, Giels G and Williams L. . 1985 Cell 43: 243–251.
Courney A and Tijan R. . 1988 Cell 55: 741–753.
Derynck R. . 1998 Cell 54: 593–595.
Dingwall C, Sharnick S and Laskey R. . 1982 Cell 30: 449–458.
Downward J, Yarden Y, Mayes E, Scrace G, Totty N, Stockwell P, Ullrich A, Schlessinger J and Waterfield M. . 1984 Nature 307: 521–527.
Dudley D, Pang L, Decker S, Bridges A and Saltiel A. . 1995 Proc. Natl. Acad. Sci. USA 92: 7686–7689.
Emkey R and Kahn C. . 1997 J. Biol. Chem. 49: 31172–31181.
Eskrand A, Longo N, Hamid M, Olson J, Liu L, Collins V and James C. . 1994 Oncogene 9: 2313–2320.
Fathalla M. . 1971 Lancet ii: 163.
Fry D, Kraker A, McMichael A, Ambroso L, Nelson J, Leopold W, Connors R and Bridges A. . 1994 Science 265: 1093–1095.
Fujii H, Zhou W and Gabrielson E. . 1996 Genes Chromosomes Cancer 16: 35–39.
Godwin A, Testa J, Handel L, Liu Z, Vandeveer L, Tracey P and Hamilton T. . 1992 J. Natl. Cancer Inst. 84: 592–601.
Gonzalez F, Raden D and Davis R. . 1991 J. Biol. Chem. 266: 22159–22163.
Gotoh N, Tojo A, Muroya K, Hashimoto Y, Hattori S, Nakamura S, Takenaw T, Yazaki Y and Shibuya M. . 1994 Proc. Natl. Acad. Sci. USA 91: 167–171.
Grimes H, Chan T, Zweidler-McKay P, Tong B and Tsichlis P. . 1996 Mol. Cell. Biol. 16: 6263–6272.
Hazzalin C, Panse R, Cano E and Mahadevan L. . 1998 Mol. Cell. Biol. 18: 1844–1854.
Henderson B, Ross R and Pike M. . 1991 Science 254: 1131–1138.
Howe L, Leevers S, Gomez N, Nakielny S, Cohen P and Marshal C. . 1992 Cell 71: 335–342.
Hunter T. . 1984 Nature 311: 414–416.
Hunter T. . 1995 Cell 80: 225–236.
Ilekis J, Gariti J, Niederberger C and Scoccia B. . 1997 Gynecol. Oncol. 65: 36–41.
Jans D and Hassan G. . 1998 BioEssays 20: 400–411.
Kampfer S, Hellbert K, Villunger A, Doppler W, Baier G, Grunicke H and Uberall F. . 1998 EMBO J. 17: 4046–4055.
Kas K, Voz M, Hensen K, Meyen E and Van de ven W. . 1998 J. Biol. Chem. 273: 23026–23032.
Kas K, Voz M, Roijer E, Astrom A-K, Meyen E, Stenman G and Van de Ven W. . 1997 Nature Genetics 15: 170–174.
Kommoss F, Wintzer H, Von Kleist S, Kohler M, Walker R, Langton B, Tran K, Pfleiderer A and Bauknecht T. . 1990 J. Pathol. 162: 223–230.
Kyriakis J, Banerjee P, Nikolakaki E, Dai T, Rubie E, Ahmad M, Abruch J and Woodgett J. . 1994 Nature 369: 156–160.
Li S, Ku C-Y, Farmer A, Cong Y-S, Chen C-F and Lee W-H. . 1998 J. Biol. Chem. 273: 6183–6189.
Liebermann T, Nusbaum H, Razon N, Kris R, Lax I, Soreq H, Whittle N, Waterfield M, Ullrich A and Schlessinger J. . 1985 Nature 313: 144–147.
Lord K, Abdollahi A, Hoffman-Liebermann B and Liebermann D. . 1991 Cell Growth Diff. 1: 637–645.
Lord K, Abdollahi A, Hoffman-Liebermann B and Liebermann D. . 1993 Mol. Cell. Biol. 13: 841–851.
Maniatis T, Fritsch E and Sambrook J. . 1989 Molecular cloning: A laboratory manual., 2nd edn. Cold Spring Harbor Laboratory.
Massague J. . 1983 J. Biol. Chem. 258: 13614–13620.
Mellon K, Wright C, Kelly P, Horne C and Neal D. . 1995 J. Urol. 153: 919–925.
Mermod N, O'Neill E, Kelly T and Tijan R. . 1988 Cell 58: 741–753.
Moon S, Palfrey H and King A. . 1984 Proc. Natl. Acad. Sci. USA 81: 2298–2302.
Moreno S and Nurse P. . 1990 Cell 61: 549–551.
Morishige K, Kurachi H, Amemiya K, Fujita Y, Yamamoto Y, Miyake A and Tanizawa O. . 1991 Cancer Res. 51: 5322–5328.
Moscatello D, Holgadomadruga M, Godwin A, Ramirez G, Gunn G, Zoltick P, Biegel J, Hayes R and Wong A. . 1995 Cancer Res. 55: 5536–5539.
Nomura N, Nagase T, Miyajima N, Sazuka T, Tanaka A, Sato S, Seki N, Kawarabayas Y, Ishikawa K and Tabata S. . 1994 DNA Res. 1: 223–229.
Pardee A. . 1989 Science 246: 603–608.
Petch L, Harris J, Raymond V, Blasband A, Lee D and Earp H. . 1990 Mol. Cell. Biol. 10: 2973–2982.
Pontecorvi A, Tata J, Pylillaier M and Robbins J. . 1988 EMBO J. 7: 1489–1495.
Preston-Martin S, Pike M, Ross R, Jones P and Henderson B. . 1990 Cancer Res. 50: 7415–7421.
Rani C, Wang F, Fuir E, Berger A, Wu J, Sturgill T, Beitner-Johnson D, LeRoith D, Varticovski L and Spiegel S. . 1997 J. Biol. Chem. 272: 10777–10783.
Roberts D, Weinstein J, Abdollahi A and Hamilton T. . 1999 Submitted for publication.
Rodriguez G, Berchuck A, Whitaker R, Schlossman D, Clarke-Pearson D and Bast R. . 1991 Am. J. Obstet. Gynecol. 164: 745–750.
Sainsbury J, Frandon J, Sherbert G and Harris A. . 1985 Lancet I: 364–366.
Scambia G, Benedetti-Panici P, Battaglia F, Ferrandina G, Briocchi G, Greggi S, deVincenzo R and Mancuso S. . 1992 J. Clin. Oncol. 10: 529–535.
Schlessinger J and Ullrich A. . 1992 Neuron 9: 383–391.
Shechter Y, Hernaez L and Cuatrecasas P. . 1978 Proc. Natl. Acad. Sci. USA 75: 5788–5791.
Spengler D, Villalba M, Hoffman A, Pantaloni C, Houssami S, Bockaert J and Journet L. . 1997 EMBO J. 16: 2814–2825.
Suganuma M, Fujiki H, Suguri H, Yoshizawa S, Hirota M, Nakayasu M, Ojika M, Wakamatsu K, Yamada K and Sugimura T. . 1988 Proc. Natl. Acad. Sci. USA 85: 1768–1771.
Tenaka K, Boice C and Testa J. . 1989 Cancer Genet. Cytogenet. 43: 1–14.
Testa J, Getts L, Salazar H, Liu Z, Handel L, Godwin A and Hamilton T. . 1994 Cancer Res. 54: 2778–2784.
Thrash-Bingham C, Salazar H, Freed J, Greenberg R and Tartof K. . 1995 Cancer Res. 55: 6189–6195.
Ullrich A and Schlessinger J. . 1990 Cell 61: 203–212.
Weinberg R. . 1991 Science 254: 1138–1146.
Williamson M. . 1994 Biochem. J. 297: 249–260.
Xu Y-H, Ishii S, Clark A, Sullivan M, Wilson R, Ma D, Roe B, Merlino G and Pastan I. . 1984 Nature 309: 806–810.
Yamamoto T, Kamata N, Kawano H, Shimizu S, Kuroki T, Toyoshima K, Rikimaru K, Nomura N, Ishizaki R, Pastan I, Gamou S and Shimizu N. . 1986 Cancer Res. 46: 414–416.
Zheng C-F and Guan K. . 1993 J. Biol. Chem. 268: 11435–11439.
Zhou L and Padmanabhan R. . 1988 Cell 55: 1005–1015.
We are grateful to H Leighton Grimes, Andrew Godwin, Joseph Testa, Philip Tsichlis, Jun-Hsiang Lin and Yasuhiro Mitsuuchi for helpful discussions, and Daphne Bell, Binaifer Balsara, and Jonathan Boyd for confocal microscopy and image analysis, and Cathy Thompson for secretarial work. This research was supported by Public Health service grant CA56916/TC Hamilton.
About this article
Cite this article
Abdollahi, A., Bao, R. & Hamilton, T. LOT1 is a growth suppressor gene down-regulated by the epidermal growth factor receptor ligands and encodes a nuclear zinc-finger protein. Oncogene 18, 6477–6487 (1999). https://doi.org/10.1038/sj.onc.1203067
- epidermal growth factor
Reproductive Sciences (2016)
Altered expression of the PLAGL1 (ZAC1/LOT1) gene in colorectal cancer: Correlations to the clinicopathological parameters
International Journal of Oncology (2015)
PLOS ONE (2015)
Regulation of cell proliferation and migration in gallbladder cancer by zinc finger X-chromosomal protein
STEM CELLS (2013)