Dear Editor,

Of the several signaling molecules important for cell survival, one that has garnered much attention is the serine/threonine kinase Akt.1 Downstream of phosphoinositide 3′-kinase (PI3K), Akt is activated by phosphorylation and regulates proapoptotic substrates, including Bad, caspase-9 and the forkhead transcription factor, which are all inactivated by phosphorylation.2,3,4 We have previously demonstrated that the p85 regulatory subunit of PI3K interacts with, and is negatively modulated by, the cytosolic protein tyrosine phosphatase SHP-1.5 SHP-1-mediated events are increasingly associated with genotoxic stress and, at least in our examination of somatostatin-induced apoptosis, appear to be dependent on caspase-8 recruitment, and increased p53 and Bax expression.6,7 Activation of caspases during apoptosis is a process tightly regulated by various Bcl-2 family members.

The histone deacetylase (HDAC) inhibitors butyrate and trichostatin A (TSA) are gaining interest as potential anti-cancer drugs. Experimentally, TSA induces caspase activity and apoptosis in the MCF-7 breast cancer cell line via a cytochrome c-dependent pathway,8 thereby implicating caspase-9-dependent events. Interestingly, both butyrate and TSA can induce a shift in cellular tyrosine phosphorylation; one mechanism involves repression of mRNA and protein expression of the c-Src tyrosine kinase by inhibiting activity of each of its promoters,9 whereas the other mechanism involves induction of SHP-1 expression by increasing the activity of the tissue-specific P1 promoter in MCF-7 cells.10

We now demonstrate that the sensitivity of MCF-7 cells (HTB-22: ATCC) to TSA-induced apoptosis was heightened by stable overexpression of SHP-1, but not by overexpression of the catalytically inactive mutant of SHP-1, that is, SHP-1-C455S. Expression of p85, the regulatory subunit of PI3K, in the MCF-7 parental cell line was transiently increased by TSA treatment indicating activation of cell survival mechanism(s). In contrast, this transient increase was absent in SHP-1-overexpressing cells and p85 immunodetection was, in fact, completely lost by TSA-treatment for 48 h. Stable overexpression of SHP-1 during TSA treatment may induce an actual downregulation of total cellular p85 protein rather than a diminished phosphorylation of p85, which was proposed to account for the regulation of p85 activity by transient SHP-1 expression.5,11 The reduction in phosphorylation of Akt observed in TSA-treated SHP-1 transfectants was not observed in SHP-1-C455S transfectants, concurring with the observation that SHP-1, but not the catalytic inactive mutant, can negatively regulate p56Lck-induced phospho-PI3K activity and phosphorylation of Akt (Figure 1).11

Figure 1
figure 1

Effect of overexpressed SHP-1 on TSA-induced apoptosis in MCF-7 cell lines. (a) Neomycin-resistant pools of stable transfectants expressing wildtype SHP-1 [SHP-1] and the catalytic mutant SHP-1-C455S [C455S] were generated (the respective immunoblots are shown below the corresponding labels) and treated with TSA (300 ng/ml). Apoptotic events were assessed by fluorescence-activated cell sorting (using annexin-V-FLUOS; 10 000 gated events/sample). MCF-7-SHP-1 transfectants were extremely sensitive to TSA treatment. The results (mean±S.D., n≥3) are presented as the percentage of apoptotic cells in control cultures (black bars) vs TSA-treated (48 h) cultures (gray bars). *: P<0.05 and **: P<0.01 (by ANOVA) vs the respective controls or between the indicated sampling groups. (b) TSA treatment induced a transient increase in expression of p85 in MCF-7 and a return to basal levels by 48 h. Expression of p85 in similarly treated MCF-7-SHP-1 cells was almost completely abolished by 48 h of TSA treatment. Phosphorylation of Akt ([p]-Akt) was selectively decreased by TSA treatment in MCF-7-SHP-1 cells [SHP-1]. TSA also induced proapoptotic changes in the phosphorylation and/or expression of Bcl-2 family members, for example, Bad, Bcl-2 and Bax, almost exclusively in the MCF-7 SHP-1 overexpressing cell line. (c) Cleavage of procaspase-9 to the active p35 subunit was detected only in MCF-7-SHP-1 [SHP-1] cells at 48 h of TSA treatment. The MCF-7 cell line is caspase-3 null, but does express the executioner caspase-7, which, because of substrate recognition, is able to cleave the same substrates as caspase-3. A strong 21 kDa band corresponding to activated caspase-7 cleaved from the 34 kDa precursor was only detectable in MCF-7-SHP-1 transfectants treated with TSA for 48 h. Cleavage fragment sizes are indicated on the right. Representative blots probed for β-actin are included in (b) and (c) to confirm equivalent protein loading.

Full size image

Treatment with the tyrosine phosphatase inhibitor vanadate inhibits apoptosis by activation of PI3K/Akt, and phosphorylation of Bad.12,13 The present study revealed that basal phosphorylation of Bad-Ser136 was decreased by SHP-1 overexpression and further affected by TSA treatment. The basal and response levels of phospho-Akt were similarly affected in these same cells and, thus, support the notion that phosphorylation of Bad-Ser136 is dependent on Akt kinase activation.14 Basal phosphorylation of Bad-Ser112 was increased by SHP-1 overexpression, perhaps indicating a heightened cell survival response mechanism(s) to stress, and was diminished by TSA treatment of these cells. In contrast, increased phosphorylation of Bad-Ser112 was evident in TSA-treated MCF-7 parental cells, even though the phosphorylation of ERK1/2 in these same cells was greatly inhibited (data not shown). These combined data suggest either that phosphorylation of Bad-Ser112 is not as specific to ERK1/2 activation in MCF-7 cells as it has been demonstrated for other cell lines14 or simply that a redundancy in cell survival signaling pathways is revealed when ERK1/2 activation is compromised. What is germane to the present hypothesis is that SHP-1 overexpression leads to a loss of Bad phosphorylation (at Ser112 and Ser136) by 48 h of TSA treatment. It is well known that unphosphorylated Bad sequesters antiapoptotic molecules such as Bcl-2 away from the proapoptotic Bax and, once liberated, Bax translocates to the mitochondrial membrane where it modulates the formation of the permeability transition pore. This event allows intramembranal cytochrome c to escape and to activate downstream caspases. The present observation of a complete ablation of Bcl-2 protein by SHP-1 overexpression concurrently with an increased expression of Bax protein would shift the cells’ response from one promoting survival to one of a greater predisposition to an apoptotic outcome. This corroborates the angiotensin II-mediated (apoptotic) ERK inactivation, and the subsequent Bcl-2 inactivation and Bax upregulation observed in PC12W and R3T3 cells.15 Our observation of proapoptotic changes in Bcl-2-related proteins coincides with increased sensitivity of MCF-7 cells to apoptogenic insult and the cleavage of caspase-9 and of the executioner caspase-7. Processing of these caspases corresponded to a 170% increase in the activity of caspase-9 [F(3,15)=3.775, P=0.0406] and a 160% increase in the activity of caspase-7 [F(3,15)=6.592, P=0.0070] (data not shown).

Perhaps what dictates sensitivity to apoptogenic stimuli is the level of expression of molecules such as Akt family members. It is known that AKT1/RACα is overexpressed in less aggressive human breast cancer cell lines such as MCF-7,16 whereas AKT3/RACγ is overexpressed in more aggressive forms of breast tumors.17 Our present data suggest that targeted manipulation of the Akt-dependent antiapoptotic signaling pathway in combination with a relative shift in cellular protein tyrosine phosphatase activity may provide a novel approach to enhancing the efficacy of adjuvant therapies in the clinic.