Signal transduction activated by the cancer chemopreventive isothiocyanates: cleavage of BID protein, tyrosine phosphorylation and activation of JNK

Phenethyl isothiocyanate and allyl isothiocyanate induce apoptosis of human leukaemia HL60 cells in vitro. Apoptosis was associated with cleavage of p22 BID protein to p15, p13 and p11 fragments and activation of JNK and tyrosine phosphorylation (18 kDa and 45 kDa proteins). All these effects and apoptosis were prevented by exogenous glutathione (15 mM). Protein tyrosine phosphatase activity was unchanged. The general caspase inhibitor Z-VAD-fmk prevented apoptosis but not JNK activation – excluding a role for caspases in JNK activation, whereas curcumin prevented JNK activation but only delayed apoptosis. This suggests that in isothiocyanate-induced apoptosis, the caspase pathway has an essential role, the JNK pathway a supporting role, and inhibition of protein tyrosine phosphatases is not involved. © 2001 Cancer Research Campaign http://www.bjcancer.com

Dietary isothiocyanates such as phenethyl isothiocyanate (PEITC) and allyl isothiocyanate (AITC) have anti-carcinogenic activities and are of potential use in the chemoprevention of cancer (Hecht, 1995). Chemopreventive activity is associated with inhibition of the activation of carcinogens by cytochrome P450 isozymes (Conaway et al, 1996), increased excretion of carcinogens by quinone reductase and GSH S-transferases (Bogaards et al, 1990;Zheng et al, 1992), and also induction of apoptosis in pre-clinical tumours (Nishikawa et al, 1997;Samaha et al, 1997;Sugie et al, 1999). The mechanism of induction of apoptosis is unknown.
PEITC and AITC inhibited the growth and induced apoptosis of human leukaemia (HL60) cells in vitro (Adesida et al, 1996;Xu and Thornalley, 2000). PEITC-induced apoptosis was characterized by entry of PEITC into cells and rapid formation and expulsion of the glutathione adduct S-(N-phenethylthiocarbamoyl)glutathione (PETC-SG), and protein thiocarbamoylation -exacerbated by the decrease in cellular glutathione (Xu and Thornalley, 2001). This committed the cells to apoptosis with a critical activation of caspase-8 in the initial 3 h and later activation of caspase-3 (Xu and Thornalley, 2000). The specific caspase-8 inhibitor Z-IETD-fmk, the general caspase inhibitor Z-VAD-fmk and a high concentration of glutathione (15 mM) inhibited PEITCinduced apoptosis completely (Xu and Thornalley, 2001). PEITC was found to induce a sustained activation of JNK1 in HeLa cells (Yu et al, 1996) and Jurkat cells in vitro . This was associated with activation of MEKK1 . Overexpression of Bcl-2 and Bcl-x L suppressed both PEITCinduced activation of JNK1 and apoptosis, suggesting that Bcl-2 and Bcl-x L could intervene upstream of JNK1 activation in PEITC-induced apoptosis . Curcumin is an inhibitor of JNK signalling upstream of MEKK1  and it delayed the induction of apoptosis by PEITC (Xu and Thornalley, 2001).
Recent investigations of caspase-8 activated apoptosis have suggested a critical role of cleavage of the cytosolic protein BID (Bossy-Wetzel and Green, 1999;Gross et al, 1999) and a role for caspases in the activation of MEKK1 in the JNK pathway (Cardone et al, 1997). Protein tyrosine phosphorylation and inhibition of protein tyrosine phosphatase activity has been associated with apoptosis where JNK activation was involved (Lumelsky and Schwartz, 1996;Chen et al, 1999). Potent thiol-modifying agents such as isothiocyanates may induce apoptosis by inhibiting protein tyrosine phosphatase activity (Denu and Tanner, 1998). Our recent work has indicated that peptide and protein thiol modification by isothiocyanates may play a critical role in activating apoptosis. We describe here, for the first time, experiments designed to examine these features of isothiocyanate-induced apoptosis.
After washing, blots were developed with the ECL detection system.

Assay of protein tyrosine phosphatase
Protein tyrosine phosphatase (PTP) activity was assayed by measuring the rate of dephosphorylation of tyrosine [ 33 P]phosphorylated myelin basic protein (MyBP). [ 33 P]MyBP was prepared from MyBP with γ-[ 33 P]-ATP and Abl protein tyrosine kinase according to the manufacturer's instructions (New England BioLabs), giving a solution of 2.0 µM [ 33 P]MyBP with a tyrosine phosphorylation of 0.48. The PTP activity of HL60 cell lysates was determined by incubation of 30 µl of assay buffer (50 mM Tris-HCl, pH 7.5; 1 mM Na 2 EDTA, 0.01% Brij 35, 1 mg ml -1 bovine serum albumin), 10 µl of cell lysate, 30 µl of 2.0 µM [ 33 P]MyBP for 10 min at 30˚C. The reaction was terminated by addition of 200 µl of 20% TCA. The solution was placed on ice for 5-10 min, centrifuged (12 000 g, 5 min, 4˚C) and 200 µl of supernatant removed, scintillation cocktail added and counted. The activity of PTP is given in units where one unit of PTP activity releases one nmol of phosphate from [ 33 P]MyBP per minute under assay conditions.

Measurement of BID and BID fragmentation
The procedure was similar to the measurement of protein tyrosine phosphorylation, except that the primary antibody was goat polyclonal anti-BID antibody, and secondary antibody was HRP-conjugated donkey anti-goat IgG.

RESULTS
The pro-apoptotic protein BID, when processed by caspase-8 and caspase-3 to the truncated form tBID, is a major initiator of mitochondrial dysfunction in apoptosis. When HL60 cell cytosolic extracts were blotted with anti-BID IgG, full-length BID protein of molecular mass 22 kDa was detected, with non-specific blotting of protein bands of >30 kDa ( Figure 1A, lane 1). When HL60 cells were incubated with 10 µM PEITC for 6 h, cytosolic extracts indicated a loss of full-length BID and the appearance of fragments of molecular mass at 15, 13 and 11 kDa ( Figure 1A, lane 2). This fragmentation was prevented by the addition of 15 mM GSH ( Figure 1A, lanes 3 and 4).
PEITC and AITC activated JNK activity in HL60 cells. AITC was a stronger inducer of JNK activity than PEITC although AITC was slightly less cytotoxic (Xu and Thornalley, 2000) -the TC 50 values of PEITC and AITC were 4.95 µM and 11.0 µM, respectively. The induction of JNK activity was prevented by 15 mM GSH (Figure 2A). During PEITC-induced apoptosis, the activity of JNK was high after 3 h and decreased at 6 h and 9 h ( Figure 2B).
The general caspase inhibitor Z-VAD-fmk (50 µM) did not inhibit the activation of JNK; in fact, it increased slightly the intensity of the blot, suggesting that it may have further increased JNK activity. Curcumin (50 µM), however, inhibited the activation of JNK ( Figure 2C).
The tyrosine phosphorylation status of cytosolic proteins changed during PEITC-induced apoptosis was investigated. In cytosolic extracts of HL60 cells, two major protein bands of molecular mass ca. 40 kDa and 65 kDa were detected ( Figure 1B, lane 1). When HL60 cells were incubated with 10 µM PEITC for 6 h, cytosolic extracts indicated the appearance of two new phosphoprotein bands of molecular mass 18 kDa and 45 kDa ( Figure 1B, lane 2). This tyrosine phosphorylation was prevented by the addition of 15 mM GSH ( Figure 1B, lanes 3 and 4).
Their formation was prevented by addition of 15 mM GSH. The effect of PEITC on the protein tyrosine phosphatase (PTP) activity of HL60 cells was studied. After 3 h, when the maximum binding of PEITC to cell protein occurred (Xu and Thornalley, 2001), the PTP activity was: control 2.12 ± 0.26, and + 10 µM PEITC 2.10 ± 0.11 (n = 3; P > 0.05).

DISCUSSION
When HL60 cells were incubated with PEITC in vitro, caspase-8 and caspase-3 were activated (Xu and Thornalley, 2000). Caspase-8 and caspase-3 cleave BID protein to 3 fragments, p15, p13 and p11 fragments (Bossy-Wetzel and Green, 1999;Gross et al, 1999). This was found in PEITC-induced apoptosis of HL60 cells herein. p15 interacts with Bcl-X L in mitochondria, leading to cytochrome c release and loss of mitochondrial membrane potential (Li et al, 1998). A high concentration of GSH (15 mM) added to the extracellular medium prevented BID cleavage (this study) and apoptosis (Xu and Thornalley, 2001). Glutathione prevented the binding of PEITC to cells, probably by non-enzymatic formation of PETC-SG extracellularly. This keeps the PEITC concentration below cytotoxic levels. PETC-SG fragments to reform PEITC but there is a continuous slow hydrolysis of PEITC to inactive products that diminishes its pharmacological activity (Xu and Thornalley, 2000).
Both PEITC and AITC activated JNK in HL60 cells. JNK activation was mediated by MEKK1 (Yu et al, 1996;. The strongest activation of JNK by PEITC occurred after 3 h when the concentration of cellular adducts of PEITC was maximal, the cellular GSH concentration was at a minimum and commitment to apoptosis occurred Thornalley, 2000, 2001). Exogenous GSH prevented JNK but Z-VAD-fmk did not despite being an efficient inhibitor of caspase-3 and caspase-8 and, indeed, an efficient inhibitor of PEITC-induced apoptosis (Xu and , 2000). This excludes a role for caspases in the activation of MEKK1 (Cardone et al, 1997) in PEITC-induced apoptosis. The signalling upstream of MEKK1 is unknown. Curcumin suppressed the activation of JNK by PEITC and delayed but did not prevent the development of apoptosis (Xu and Thornalley, 2001). The role of JNK in apoptosis may be to potentiate cell death -JNK activation without executioner caspases did not induce apoptosis. JNK signalling increases the expression of fas ligand for increased agonism at the fas cell death receptor (Faris et al, 1998) and counters the anti-apoptotic activity of Bcl-x L in mitochondria (Kharbanda et al, 2000). A critical role for tyrosine phosphorylation in signal transduction in apoptosis has been proposed (Chen et al, 1999), including apoptosis of HL60 cells (Lumelsky and Schwartz, 1996). Inhibition of protein tyrosine phosphatases has been shown to potentiate apoptosis (Chen et al, 1999) although the fas-associated protein tyrosine phosphatase (FAP-1) that is influential in countering fas-mediated apoptosis (Li et al, 2000) was not highly expressed in HL60 cells (Komada et al, 1997). Protein tyrosine phosphorylation was an early event in PEITC-induced apoptosis of HL60 cells. Protein tyrosine phosphatases may be susceptible to inhibition by isothiocyanates by modification of their active site cysteinyl thiol (Denu and Tanner, 1998). We were unable, however, to demonstrate an effect of PEITC on protein tyrosine phosphatases. Hence, increased protein tyrosine kinase activity is implicated in the increased protein tyrosine phosphorylation in isothiocyanate-induced apoptosis.
PEITC is in phase I clinical trial for the chemoprevention of cancer. It may eventually find use in the prevention of primary and secondary tumours in vivo. The induction of tumour apoptosis contributes to these chemopreventive effects (Nishikawa et al, 1997;Samaha et al, 1997;Sugie et al, 1999).