Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The family of transglutaminase enzymes catalyze the Ca²⁺-dependent cross-linking of proteins using protein bound glutaminic -carboxamide groups and primary amino groups of lysines in polypeptide chains or polyamines (Lorand and Conrad, 1984). Transglutaminases have been implicated in a wide variety of biological phenomena encompassing the stabilization and protection of cell and tissue integrity, including a role in the scaffold formation of apoptotic bodies and cornified envelopes, hepatic fibrogenesis, and tissue remodeling (Fesus et al., 1987; Mirza et al., 1997; Aeschlimann and Thomazy, 2000). Furthermore, the identification of tissue transglutaminase (tTG) as a GTP-binding protein raises the idea of other possible roles in signal transduction events, including the initiation phase of programmed cell death (Nakaoka et al., 1994; Melino and Piacentini, 1998).

One of the unexpected observations to emerge from early studies on the comparison of transglutaminase activity of normal and neoplastic cells was that exposure of several tumor cell lines to antineoplastic drugs was associated with a marked increase in transglutaminase (TG) activity (Piacentini et al., 1993; Furuya and Isaacs, 1994; Lokshin et al., 1995). Because overexpression of tTG in transformed cells causes cell death, it has been suggested that induction of the enzyme may contribute to the cytotoxic effect of chemotherapeutic drugs (Gentile et al., 1992; Furuya and Isaacs, 1994; Melino et al., 1994). In addition, it has been demonstrated that some cytotoxic drugs, such as bleomycin, are TG inhibitors and can interfere with TG-mediated conjugation of polyamines (Griffin et al., 1978). Russel and Womble (1982) have shown that this inhibition can also contribute to drug-induced cytotoxicity. In contrast, several studies have suggested that the induction of tTG in cells treated with antineoplastic drugs may play a role in drug resistance of these cells. Doxorubicin-resistant human breast and lung carcinoma cells showed a much higher level of tTG expression than drug-sensitive ones (Mehta, 1994; Han and Park, 1999b).

To clarify the role of tTG in the cytotoxic activity of three unrelated antineoplastic drugs frequently used in the chemotherapy of human malignancies, cisplatin (Cis), doxorubicin (Dox) and bleomycin (Bleo), we have studied the molecular and regulatory events that take place in the `window' of time that precedes the execution phase of apoptosis. Therefore, we applied response modifiers as tools to pharmacologically perturb enzyme induction and cell death and characterized transcriptional changes in the early phase of drug-induced apoptosis.

We report herein that, upon application of various anticancer treatments, tTG undergoes uniform transcriptional up-regulation followed by the induction of enzyme cross-linking activity, occurring before apoptosis. Stimulation with PPAR or RXR agonists causes abrogation of tTG induction but still allows drug-induced cell death. Treatment of cells with interleukin-6 (IL-6) before the addition of anticancer agents effectively prevents cell death and up-regulates gp130, her2/neu (erbB2), and bcl-x_L, but does not interfere with induction of tTG. These findings suggest that tTG is involved in the apoptotic phenotype but is pharmacologically dissociable from the early phase of apoptosis.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Cultures. HepG2 cells were obtained from the European Collection of Animal Cell Cultures and grown in Dulbecco's modified Eagle's medium (Sigma, St. Louis, MO) supplemented with 10% fetal calf serum (Life Technologies, Gaithersburg, MD), 2 mM glutamine, and 2 mM Na-pyruvate but no antibiotics and maintained in an incubator with 5% CO₂ at 37°C and 90% humidity. Cells were routinely passaged twice a week by trypsinization (0.1% trypsin plus 0.02% EDTA in phosphate-buffered saline) and used for experiments when reaching 30 to 40% confluence.

Induction and Modulation of Apoptosis. HepG2 cells were treated with cisplatin (EBEWE Pharmaceuticals, Unterach, Austria) at a dose range of 1 to 100 µg/ml, doxorubicin (Farmitalia Carloerba, Milan, Italy) at a dose range of 20 ng/ml to 20 µg/ml, or bleomycin (Nippon Kayaku, Tokyo, Japan) at a dose range of 60 µg/ml to 3 mg/ml to determine optimal concentrations to induce apoptotic but not necrotic cell death most abundantly at the time point of maximal effect. Puromycin (Sigma) was applied at 10 µg/ml, WY14643 (pirinixic acid, Sigma) was used at a dose of 10 µM. LG100268 was obtained from Ligand Pharmaceuticals and applied at a concentration of 1 µM. Interleukin-6 (a kind gift from Dr. András Falus) was applied at a concentration of 20 U/ml. In each case, at least two independent experiments were performed.

RNA Preparation and Quantitative RT-PCR. Human liver RNA was purchased from CLONTECH (Palo Alto, CA). Total RNA was isolated from 10⁶ HepG2 cells after appropriate treatment using the RNeasy kit from Qiagen (Chatsworth, CA). After DNase treatment, reverse transcription was performed at 50°C for 30 min from 100 ng of total RNA using Superscript II reverse transcriptase and specific reverse primers. Quantification based on real-time monitoring of amplification was carried out using an ABI 7700. All determinations were done in triplicate with one control reaction containing no RT enzyme to test for potential DNA contamination. Values of transcripts in unknown samples were obtained by interpolating Ct (PCR cycles to threshold) values on a standard curve derived from known amounts of cognate, amplicon-specific synthetic RNAs. Absolute numbers of mRNA molecules have been normalized to cyclophilin to correct for differences in RNA concentration. Sequences of the primers and TaqMan probes used for subsequent amplification reactions are summarized in Table 1. Synthetic sRNA standards were synthesized as described previously (Ahuja et al., 2001).

Protein and Enzyme Activity Assays. Protein concentrations were determined by the Bradford method with necessary reagents obtained from Bio-Rad Laboratories (Hercules, CA). Tissue transglutaminase activity was measured by the rate of incorporation of [³H]putrescine into N,N-dimethylated casein. Reaction mixtures consisted of the following in a total volume of 100 µl: 50 µl of crude cell homogenate, 10 µl of N,N-dimethylcasein (40 mg/ml), 20 µl of [1,4(n)-³H] putrescine (30 Ci/mmol), 10 µl of 250 mM Tris·HCl, pH 7.5, containing 150 mM -MEA and 10 µl CaCl₂ (50 mM). The reaction was initiated by the addition of CaCl₂ and incubated at 37°C. After 5 and 10 min elapsed, a 25-µl sample was taken, dropped on filter paper, and precipitated in cold trichloroacetic acid; washed intensively with 10% and 5% TCA and ethanol; and radioactivity of the filter was measured in a liquid scintillation counter. Enzyme activity was calculated as picomoles of [³H]putrescine incorporated into casein in 1 min by one milligram of cellular protein.

Cytotoxicity Assay with MTT Staining of Viable Cells. The MTT assay is a colorimetric procedure based on the ability of viable cells to reduce a soluble yellow tetrazolium salt (MTT) to blue formazan crystals. Ten thousand cells were plated into wells of 96-well plates; 24 h later, cells were treated by various chemotherapeutic drugs. After 48 to 72 h, 5 mg/ml MTT was added and the plates were further incubated at 37°C for 3 h. After eluting the dye with isopropanol/formic acid, absorbance was determined using an automated plate reader at 540 nm and 620 nm. Results are the means of at least five independent measurements and expressed as the percentage of the absorbance of untreated control cells ± S.D.

Detection of Apoptosis. Assessment of apoptotic cells by flow cytometric analysis was carried out in a Coulter EPICS-XL apparatus using the System II software (Beckman Coulter, Fullerton, CA). Floating cells were collected by centrifugation at 200 g. Adherent cells were harvested by trypsinization with 1% trypsin for 2 min. All cells were washed in PBS, fixed in 70% ethanol, and stained with 50 µg/ml propidium-iodide for 30 min (Nicoletti et al., 1991). Cells showing lower DNA contents than those in the G₀ phase were considered apoptotic.

Statistical Analyses. Differences in transcript levels, enzyme activities, and cell death rates between the control group and various treatment groups were tested by one-way analysis of variance test with Tukey's post hoc analysis. When normal distribution could not be assumed despite logarithmic transformation of the data, the corresponding nonparametric test (i.e., Kluskal-Wallis analysis of variance on ranks) was performed. Student's t test was applied for pairwise comparison of treatments. All statistical analyses were carried out using SigmaStat (Jandel, San Rafael, CA) software systems. p-Values < 0.05 were considered statistically significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Induction of tTG in Relation to Apoptosis. The tTG enzyme is activated and induced in liver cells in several models of apoptosis (Fesus et al., 1987; Cummings, 1996). To investigate the molecular mechanisms of different antineoplastic drugs in relation to apoptosis, we treated nonconfluent cultures of HepG2 hepatoma cells with 10 µg/ml cisplatin, 2 µg/ml doxorubicin, or 0.6 mg/ml bleomycin. Six hours later, gene expression was quantified in total RNA extracts from these cells using real-time, quantitative RT-PCR. According to multiple determinations, tTG transcripts were present in a low abundance in HepG2 cells, approximately 0.01% of cyclophilin, a highly abundant `housekeeping' gene.

View larger version (34K): [in this window] [in a new window]   Fig. 1.   Effect of the administration of 10 µg/ml cisplatin (Cis), 2 µg/ml doxorubicin (Dox) or 0.6 mg/ml bleomycin (Bleo) on tTG mRNA levels and apoptosis of HepG2 cells compared with untreated control cells (Ctr, *p < 0.05). A, comparison of tTG transcript levels in HepG2 cells treated with antineoplastic drugs for 6 h (normalized to cyclophilin and given as mean ± SD). B, illustration of the progression of viable cell counts (bars) after 24 and 48 h of treatment with chemotherapeutic drugs and the accumulation of trypan blue positive cells (lines). C, comparison of MTT positivity as a measure of viability in cell cultures treated with Cis, Dox, or Bleo for 48 h, relative to untreated ones (100%). D, flow cytometric analysis of propidium iodide-stained HepG2 cells after exposure to Cis, Dox, or Bleo; ratios of the gated apoptotic populations are shown.

Modulation of tTG Expression and Cell Death by Response Modifiers. Having determined the induction of tTG at the transcriptional level, we then measured cross-linking activity in whole-cell lysates to confirm that the induction of RNA resulted in the production of functional enzyme upon drug treatment. As expected, the induction of enzyme activity followed a trend similar to that of the change at the transcript level (Fig. 2A). The most pronounced induction was measured after bleomycin treatment. The observed response, to a more modest extent, could be elicited in HeLa cervical carcinoma and K562 erythroleukemia cells, human cell lines derived from different sources of tissue (Fig. 2B).

View larger version (30K): [in this window] [in a new window]   Fig. 2.   Tissue transglutaminase activity in human cell lines following chemotherapeutic treatments. A, comparison of tTG cross-linking activity in lysates of HepG2 cells treated with antineoplastic drugs for 48 h (values shown as mean ± SD, *p < 0.05). B, comparison of tTG activity in human cell lines HepG2, K562, and HeLa 48 h after the addition of 0.6 mg/ml Bleo.

View larger version (41K): [in this window] [in a new window]   Fig. 3.   Effect of the pretreatment of HepG2 cells with PPAR and RXR ligands and IL-6 on tTG expression and cell viability following antineoplastic treatment. A, comparison of the effects of a 12-h pretreatment with vehicle only (), 10 µM WY14643 (), or 1 µM LG100268 () on tTG mRNA levels subsequently induced by antineoplastic drugs at 6 h. Normalized values are expressed as fraction of untreated controls (*p < 0.05). B, comparison of MTT reduction capacity in cells without () or with () preadministration of 20U/ml IL-6 before chemotherapeutic treatment. Measurements were made 48 h after the addition of drugs. Values are shown as mean ± SD (*p < 0.05). C, comparison of the effects of 24-h pretreatment with vehicle only () or 20U/ml IL-6 () on tTG mRNA levels induced by antineoplastic drugs. Normalized values are expressed as fraction of untreated controls.

Expression Pattern of Other Apoptosis-Related Genes. To evaluate whether the induction of tTG was coincident with changes of other genes in HepG2 cells, we quantified the transcripts of genes commonly regarded as key effectors in apoptotic cells. These included fasL, caspase3, and bax  along with genes with antiapoptotic potential, bcl-x_L, the signal transducer gp130, and her2/neu (Table 2). Our results show that both cisplatin and bleomycin selectively induce tTG and fasL. Expression levels of caspase 3 were unchanged and bax  was induced only by cisplatin. As opposed to tTG, the expression of the receptor tyrosine kinases gp130 and her2/neu were down-regulated by all three drugs. Also, bcl-x_L transcript levels decreased upon treatment with cisplatin and doxorubicin but not bleomycin. Induction of both tTG and fasL mRNAs were transitory, because transcript levels already decreased 12 h later, whereas changes in bcl-x_L and her2 expression showed a continuous trend over 18 h. Interestingly, IL-6 pretreatment resulted in marked and sustained induction of gp130 when combined with either anticancer drug (Fig. 4A), paralleled by a smaller induction of her2/neu mRNA (Fig. 4B). IL-6 pretreatment also up-regulated bcl-x_L expression, maintaining its induced state even after drug treatments (Fig. 4C). Administration of WY14643 or LG100268 did not have a similar effect (data not shown).

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Effect of the pretreatment of HepG2 cells with IL-6 on drug-induced changes in transcript levels of antiapoptotic genes. Normalized values are expressed as fraction of untreated controls (* p < 0.05). A, comparison of the effects of 24-h pretreatment with vehicle only () or 20 U/ml IL-6 () on gp130 mRNA levels after antineoplastic drug treatment. B, comparison of the effects of 24-h pretreatment with vehicle only () or 20 U/ml IL-6 () on Her2/neu mRNA levels after antineoplastic drug treatment. C, comparison of the effects of 24-h pretreatment with vehicle only () or 20U/ml IL-6 () on bcl-xL mRNA levels after antineoplastic drug treatment.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In this study, we report that in HepG2 cells, DNA-damaging chemotherapeutics elicit uniform induction of a marker of apoptosis, tTG, well before manifest cell death takes place. The finding seems to represent a general effect of antineoplastic agents, because drugs with different modes of action elicited induction of tTG in different cell types.

The expression and activity of proapoptotic genes (tTG, fasL, caspase 3, bax) are regulated at several different levels (Balajthy et al., 1997; Nicholson and Thornberry, 1997; Villunger et al., 1997; Wood and Newcomb, 2000). Therefore, it comes as no surprise that chemotherapeutic treatment fails to alter the mRNA pool of many apoptotic genes (Table 2). Antiapoptotic transcripts, however, undergo correlated changes; surprisingly, these appear in a coordinated fashion with tTG. The induction of gp130 and her2/neu along with bcl-x_L by IL-6 may be manifestations of an orchestrated process to protect the cell from apoptotic triggers. Her2/neu belongs to the same family of EGF receptors and is often up-regulated or amplified in breast cancers. It has also been implicated in chemoresistance (Tsai et al., 1993) and shown to associate with the Fas receptor (Shen and Novak, 1997), suggesting a possible connection to Fas-dependent cell death. Recent findings imply that Her2/neu may be involved in the IL-6 signaling pathway via gp130, its newly identified heterodimerization partner, to generate a receptor that has potent antiapoptotic activity in a variety of cell types (Qiu et al., 1998; Chien, 1999). Our data imply that expression of her2/neu and gp130 in HepG2 cells treated with chemotherapeutic drugs is directed by the same factor and provide further evidence of an autocrine regulatory loop existing in HepG2 cells. The identification of these mechanisms and their functional relationship to the participation of tTG in drug-induced apoptosis remain the subject of further studies.

Based on the information accumulated over the years about the possible connections of transglutaminase enzymes and drug-induced cytotoxic effects (Griffin et al., 1978; Russell and Womble, 1982; Piacentini et al., 1993; Han and Park, 1999a), important knowledge can be provided by studies clarifying the role of tTG on the molecular mechanisms of chemotherapeutic agents. Our results suggest, that although the induction of tTG is a frequent component of the apoptotic phenotype, it is pharmacologically dissociable from the early phase of apoptosis. Using agonists of a nuclear receptor heterodimer as response modifiers, the effects of cytotoxic drugs on tTG could be manipulated independently from concomitant cell death. The multiplicity of simultaneous processes occurring in the course of apoptosis can allow for cell death without early tTG up-regulation, whereas enhanced tTG induction can still permit rescue from cell death by alternative mechanisms. This result does not support the notion that tTG may be an essential player in the early, decision-making phase of apoptosis. On the other hand, the observation may be important in cancer therapy, because the ability to block the induction of tTG during drug-induced cell death may alleviate some of the limiting side effects of anticancer agents linked to transglutaminase activity, including fibrosis and neuropathies (Toida et al., 1991; Mirza et al., 1997; Igarashi et al., 1998).

The common factor to regulate tTG induction upon antineoplastic treatment of cancer cells remains to be elucidated. In macrophages, NFB signaling was shown to be involved in tTG induction by tumor necrosis factor- (Kuncio et al., 1998). NFB signaling can be blocked by the activation of PPARs (Ricote et al., 1998). Because WY14643 and LG100268, PPAR and RXR agonists, respectively, interfere with the induction of tTG in HepG2 cells, we can hypothesize that NFB participates in drug-dependent tTG up-regulation. Transcriptional regulation of fasL involves activator protein-1 and SV-40 promoter binding protein-1 sites and a functional role of NFB (Holtz-Heppelmann et al., 1998; Kasibhatla et al., 1998), all of which are present in the tTG promoter. Whether NFB acts as cis-regulatory element to retinoid or IL-6 response elements remains to be investigated. However, the differential effect of doxorubicin suggests that despite shared regulatory elements chemotherapeutic drugs may lack one common mechanism to trigger the molecular response in HepG2 cells.