Paclitaxel Induces Inactivation of P70 S6 Kinase and Phosphorylation of Thr421 and Ser424 Via Multiple Signaling Pathways in Mitosis1

Paclitaxel induces inactivation of p70 S6 kinase and phosphorylation of Thr421 and Ser424 via multiple signaling pathways in mitosis1

Xiao-Feng Le1, Walter N Hittelman1, Jiaxin Liu2, Amanda McWatters1, Chun Li3, Gordon B Mills4 and Robert C Bast Jr*,1

1Department of Experimental Therapeutics, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA; 2Molecular Pathology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA; 3Diagnostic Radiology, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA; 4Molecular Therapeutics, The University of Texas MD Anderson Cancer Center, Houston, TX 77030, USA

The 70 kDa ribosomal S6 kinase (p70S6K) is important Keywords: paclitaxel; p70S6K; phosphorylation; activa- for cell growth and survival. Activation of p70S6K tion; mitosis requires sequential phosphorylation of multiple serine and threonine sites often triggered by growth factors and hormones. Here, we report that paclitaxel, a microtubule- damaging agent, induces phosphorylation of p70S6K at Introduction threonine 421 and serine 424 (T421/S424) in a concentration- and time-dependent manner in multiple breast and The 70 kDa ribosomal S6 kinase (p70S6K) is a mitogen- ovarian cancer cell lines demonstrated by a T421/S424 activated serine/threonine kinase that plays a critical phospho-p70S6K antibody. Phosphoamino-acid analysis role in cell growth and survival (Grammer et al., 1996; and Western blot analysis by serine-/threonine-specific Avruch et al., 2001; Harada et al., 2001; Templeton, antibodies further confirms that both serine and threonine 2001; Volarevic and Thomas, 2001). It appears that residues are phosphorylated in p70S6K following treat- there are at least three mechanisms by which p70S6K ment with paclitaxel. Paclitaxel-induced p70S6KT421/S424 regulates cell growth. First, p70S6K is a regulator of phosphorylation requires both de novo RNA and protein protein synthesis. p70S6K phosphorylates the 40S synthesis via multiple signaling pathways including ribosomal protein S6 and leads to an increase in the ERK1/2 MAP kinase, JNK, PKC, Ca++, PI3K, and rate of translation of the class of 50TOP mRNA mammalian target of rapamycin (mTOR). Despite transcripts, which encode critical components of the phosphorylation of p70S6KT421/S424, paclitaxel inactivates cellular translational apparatus, thus facilitating an this kinase in a concentration- and time-dependent manner increase in the overall rate of protein translation as illustrated by in vitro kinase assay. Inhibitors of (Jefferies et al., 1997; Thomas, 2000). Second, p70S6K mTOR, PI3K, and Ca++ impair p70S6K activity, is a regulator of the cell cycle. Impairment of p70S6K by whereas inhibitors of JNK and PKC stimulate p70S6K injection of neutralizing antibodies into fibroblasts activity. Inhibition of PKC and JNK prevents paclitaxel- arrests cells in G1 (Lane et al., 1993). Inhibition of induced p70S6K inactivation. Moreover, the paclitaxel- p70S6K with rapamycin, which depresses p70S6K induced phosphorylation and low activity of p70S6K activation by inhibiting the mammalian target of mainly occurs during mitosis. In summary, paclitaxel is rapamycin (mTOR) kinase, blocks cell cycle progression able to induce p70S6KT421/S424 phosphorylation and de- through G1 (Chung et al., 1992; Price et al., 1992). Cell crease its activity in mitotic cells via multiple signaling cycle progression involves not only an increase in pathways. Our data suggest that paclitaxel-induced protein synthesis but also the coordinated activation of p70S6KT421/S424 phosphorylation and kinase inactivation cyclins, cyclin-dependent kinase (CDK), and CDK are differentially regulated. Our data also indicate that inhibitors. Indeed, p70S6K has been shown to regulate paclitaxel may exert its antitumor effect, at least in part, cyclin D1 (Hashemolhosseini et al., 1998) and p27Kip1 via inhibition of p70S6K. (Luo et al., 1996). Finally, p70S6K is a regulator of cell Oncogene (2003) 22, 484–497. doi:10.1038/sj.onc.1206175 size. p70S6K-deficient Drosophila and p70S6K knock- out mice show a significant reduction in body size (Shima et al., 1998; Montagne et al., 1999). Control of *Correspondence: RC Bast Jr, The University of Texas MD Anderson Cancer Center, Box 355, 1515 Holcombe Blvd., Houston, TX 77030- cell size appears to be consistent with a function of 4009, USA; p70S6K in protein synthesis, but other mechanisms may E-mail: [email protected] contribute to cell size control. Our research was supported in part by a Grant CA-39930 (to RCB) p70S6K has recently been reported to regulate from the National Institutes of Health and by an Institutional survival signals through phosphorylation and inactiva- Research Grant IRG-3721206 (to XFL) from the University of Texas MD Anderson Cancer Center tion of a proapoptotic molecule, BAD (Harada et al., Received 29 April 2002; revised 22 October 2002; accepted 22 2001). Growth factors such as insulin-like growth factor October 2002 1 do not induce phosphorylation at Ser-136 of BAD in Toxol phosphorylatea and inactivates p70 S6k in mitosis X-F Le et al 485 p70S6K-deficient embryonic stem cells (Harada et al., microtubules and inhibiting formation of mitotic 2001). Therefore, inhibition of apoptosis and promotion spindles and arresting cell growth at the G2/M phase of passage through the G1–S phase of the cell cycle of cell cycle (Schiff & Horwitz, 1980). In agreement with suggest that p70S6K is an important and powerful previous reports (Schiff & Horwitz, 1980; Horwitz, regulator of cell proliferation and survival. 1992), paclitaxel arrested growth of SKBr3 breast cancer Regulation of p70S6K activity is complex. In general, cells in G2/M and induced apoptosis (Figure 1a). While phosphorylation of p70S6K has been associated with investigating signaling pathways induced by treatment activation of kinase activity. As described in several of SKBr3 cells with paclitaxel, we found that paclitaxel review articles (Grammer et al., 1996; Avruch et al., inhibited the serine/threonine phosphorylation of ERK 2001; Templeton, 2001; Volarevic and Thomas, 2001), MAP kinase, p38 MAP kinase, and AKT kinase activation of p70S6K is tightly regulated by the (Figure 1b). By contrast, paclitaxel promoted phosphor- coordinated phosphorylation of at least seven different ylation of JNK and p70S6KT421/S424 as illustrated by serine and threonine residues. Four proline-directed sites phospho-specific antibodies (Figure 1b). Paclitaxel have been identified in the autoinhibitory domain at the induced a 12.5-fold increase in phosphorylation of C-terminus of the protein (Ser411, Ser418, Thr421, and p70S6K at T421 and S424 (Figure 1c). As shown in Ser424). These residues are rapidly phosphorylated in Figure 2a and b, paclitaxel was capable of inducing response to mitogenic stimuli. Additionally, there are p70S6KT421/S424 phosphorylation in a time- and dosage- three more critical residues: Thr229 in the catalytic loop dependent manner in SKBr3 cells. The second genera- and Ser371, Thr389 in the linker region (Grammer et al., tion of paclitaxel, docetaxel (DTX), and a water-soluble 1996; Avruch et al., 2001; Volarevic and Thomas, 2001). form of paclitaxel, poly(l-glutamic acid)–paclitaxel According to one model, the first step for p70S6K to be conjugate (PG–TAX), were also able to induce robust activated is the phosphorylation of four serine and p70S6KT421/S424 phosphorylation (Figure 2c). Paclitaxel threonine residues lying within the pseudosubstrate also induced p70S6KT421/S424 phosphorylation in another domain by proline-directed kinases that remain to be breast cancer cell line, BT474, and two ovarian cancer identified. This action releases the catalytic domain from cell lines, SKOv3 and 2008 (Figure 2d), indicating that the C-terminal autoinhibitory domain and permits paclitaxel-induced p70S6KT421/S424 phosphorylation is subsequent Thr389 phosphorylation by the 3-phosphoi- not cell-type-specific event. nositide-dependent protein kinase (PDK1), mTOR, and To confirm the ability of paclitaxel to induce NIMA (never in mitosis, gene A)-related kinase-6/7 phosphorylation of p70S6KT421/S424 and to determine (NEK6/7) kinases, further facilitating Thr229 access to which amino-acid residues were phosphorylated by PDK1 (Avruch et al., 2001; Templeton, 2001; Volarevic paclitaxel, phosphoamino-acid analysis was performed. and Thomas, 2001). Phosphorylation of the four serine As shown in Figure 3a, paclitaxel treatment induced and threonine residues lying within the pseudosubstrate p70S6K phosphorylation at serine and theronine domain contributes to, but is not sufficient for, p70S6K residues. Antiphosphoserine and antiphosphothreonine activation. There is no report, however, to indicate antibodies further confirmed these results. Both threo- whether phosphorylation of any of these four serine and nine and serine were phosphorylated after paclitaxel threonine residues lying within the pseudosubstrate treatment, whereas there was no detectable tyrosine domain might link to p70S6K inactivation. phosphorylation (Figure 3b). Paclitaxel is a common chemotherapeutic agent specifically targeting microtubules (Schiff & Horwitz, Paclitaxel does not induce phosphorylation of p70S6K at 1980; Horwitz, 1992). The effect of paclitaxel on T389 and S411 residues p70S6K has not been previously studied in depth. Here we have observed for the first time that paclitaxel, a non- Some eight phosphorylation sites within p70S6K are mitogenic stimulus, induced intense phosphorylation of currently believed to mediate kinase activation in a p70S6K at threonine 421 (T421) and serine 424 (S424) hierarchical fashion (Avruch et al., 2001; Volarevic and (p70S6KT421/S424). Phosphorylation of p70S6KT421/S424 Thomas, 2001). Phosphorylation of T421 and S424 induced by paclitaxel was, however, concurrent with situated within a pseudo-substrate autoinhibitory do- inactivation of the enzyme in mitotic cells. Phosphor- main in the carboxyterminal noncatalytic tail is believed ylation of p70S6KT421/S424 and inactivation induced by to be one of the initial steps in p70S6K activation paclitaxel were found to be regulated by multiple and (Avruch et al., 2001; Volarevic and Thomas, 2001). different signaling pathways. Based on the availability of phospho-p70S6K antibodies, we have also assessed the phosphorylation of p70S6K at T389 and S411 residues after paclitaxel Results treatment in multiple breast and ovarian cancer cell lines. As shown in Figure 3c, no phosphorylation of T389 Paclitaxel induces phosphorylation of p70S6K at T421 p70S6K was found in each of the four cell lines after and S424 in a time- and concentration-dependent manner paclitaxel treatment. Only a slight increase in phosphorylation of p70S6KS411 was observed in SKOv3, SKBr3, Paclitaxel is a chemotherapeutic drug frequently used in and BT474 cells, whereas significant phosphorylation of the treatment of patients with breast and ovarian p70S6KT421/S424 was observed after paclitaxel treatment cancers. Paclitaxel inhibits tumor growth by stabilizing (Figure 3c).

Oncogene Toxol phosphorylatea and inactivates p70 S6k in mitosis X-F Le et al 486 a Untreated b - + Paclitaxel % ⇐ Sub-G1: 1.8 P-ERK1/2 G1: 54.8 ← P-p38 S: 30.2 G2/M: 14.9 ← P-AKTS473 Paclitaxel % ← P-JNK Sub-G1: 20.4 ← T421/S424 G1: 66.1 P-p70S6K S: 118 ← Total p70S6K G2/M: 75.9

c 14

4 p70S6K (Fold) 2

0 Phosphorylation of T421/S424 Untreated Paclitaxel

Figure 1 Paclitaxel induces G2/M arrest and phosphorylation of p70S6KT421/S424 in human SKBr3 breast cancer cells. SKBr3 cells cultured in complete medium were treated with paclitaxel at 100 nm for 24 h. Cells were harvested, subjected to cell cycle analysis, and whole cell lysates were made as described in Experimental procedures. (a) Cell cycle distribution was analysed by ﬂow cytometry. Sub- G1 fractions represent apoptotic cells. A histogram is shown that is representative of several replicate analyses. (b) Cell signaling induced by paclitaxel. Western blotting was carried out using the indicated antiphospho antibodies: P-ERK1/2, phospho-ERK1 and ERK2 antibody; P-p38, phospho-p38 antibody; P-AKTS473, phospho-AKT at S473 antibody; P-JNK, phospho-JNK antibody; P- p70S6KT421/S424, phospho-p70S6K T421/S 424 antibody. The same blot was stripped and reprobed with a pan-p70S6K antibody. These Western blotting were representative of six independent experiments. (c) quantitation of P-p70S6KT421/S424 as shown in (b) from six experiments. Result was normalized to an untreated control

a b Time (hrs): 0 4 8 16 24 32 Dose (nM): 0 0.2 2 5 10 20 200 2000

P-p70S6KT421/S424→ P-p70S6KT421/S424→

β → -actin β-actin→

c DTX - + - d Paclitaxel - + - + - + PG-TAX - - + P-p70S6KT421/S424→ P-p70S6KT421/S424 → β-actin→ Total p70S6K → SKOv3 2008 BT474

Figure 2 Time- and dose-dependent phosphorylation of p70S6KT421/S424 by paclitaxel. SKBr3, BT474, 2008, and SKOv3 cells cultured in complete medium were treated with paclitaxel or docetaxel (DTX) or a water-soluble poly(l-glutamic acid)-paclitaxel conjugate (PG–TAX) at 20 nm or at the indicated concentrations (b) for 24 h or for indicated intervals (a). Cells were then harvested and subjected to Western blot analysis as described in Experimental procedures. (a) Time-dependent phosphorylation of p70S6KT421/S424 by paclitaxel in SKBr3 cells, which were treated at 20 nm.(b) Dose-dependent phosphorylation of p70S6KT421/S424 by paclitaxel in SKBr3 cells, which were treated for 24 h. (c) Phosphorylation of p70S6KT421/S424 induced by two paclitaxel derivatives, DTX and PG–TAX. SKBr3 cells were treated with DTX and PG–TAX at 20 nm for 24 h. (d) Further conﬁrmation of paclitaxel-induced phosphorylation of p70S6KT421/ S424 in human breast cancer cell line BT474, human ovarian cancer cell lines 2008, and SKOv3. These results were representative of three independent experiments

Oncogene Toxol phosphorylatea and inactivates p70 S6k in mitosis X-F Le et al 487 a b ← rp-Se ← p-Thr - + Paclitaxel

← p-Tyr ← Phospho-Serine

← Phospho-Threonine

← Original ← Phospho-Tyrosine

ed ← Total p70S6K

Standard Paclitaxel Untreat

c SKBr3 BT474 SKOv3 2008 - + - + - + - + Paclitaxel ← P-p70S6KT421/S424 ← P-p70S6KT389

← P-p70S6KS411

← Total p70S6K

Figure 3 Conﬁrmation of phosphorylation of p70S6K at T421/S424 and detection of phosphorylation of p70S6K at T389 (P- p70S6KT389) and at S411 (P-p70S6KS411) induced by paclitaxel. (a) Paclitaxel induces serine and threonine phosphorylation of p70S6K. SKBr3 cells treated with paclitaxel were labeled with [32P]orthophosphate in vivo. Total protein was made and immunoprecipitated with a pan-p70S6K antibody. Complete hydrolysis of the 32P-labeled p70S6K protein with 6 n hydrochloric acid was performed for 1.5 h at 1101C, and the radioactive phosphoamino-acid composition was determined by TLC plates. Encircled areas indicate the locations of phosphoserine (p-Ser), phosphothreonine (p-Thr), and phosphotyrosine (p-Tyr), visualized by ninhydrin staining. The results were representative of two independent experiments. (b) Western blot analysis with antiserine, threonine, and tyrosine antibodies. SKBr3 cells were treated with paclitaxel as described above. Total lysates were isolated and immunoprecipitated with antitotal p70S6K antibody. The immunoprecipitates were resolved on 8% SDS–PAGE. Western blotting was performed with antiserine antibody. Then the same blot was stripped and reprobed with antithreonine, tyrosine, and total p70S6K antibodies. The above results were representative of two independent experiments. (c) No signiﬁcant P-p70S6KT389 and P-p70S6KS411 after paclitaxel treatment. SKBr3, BT474, 2008, and SKOv3 cells cultured in complete medium were treated with paclitaxel at 20 nm for 24 h. Cells were then harvested, protein isolated, and Western blot analysis performed as described in Experimental procedures. The same blot was stripped and reprobed with three antiphospho-p70S6K antibodies and a pan-anti-p70S6K antibody. These results were representative of two independent experiments

Paclitaxel inactivates p70S6K despite phosphorylation of bottomed 16 h after paclitaxel treatment (Figure 4b), p70S6KT421/S424 when 6% of cells were undergoing apoptosis and 55% of cells were at the G2/M phase. Furthermore, we have It is well known that at least seven regulatory sites extended this study to another breast cancer cell line, specific for serine/threonine phosphorylation within BT474 (Figure 4c), and two ovarian cancer cell lines, p70S6K control its activity (Grammer et al., 1996; 2008 (Figure 4d) and SKOV3 (Figure 4e). Similar results Avruch et al., 2001; Templeton, 2001; Volarevic and were observed in each cell line after paclitaxel treatment. Thomas, 2001). The first step in p70S6K activation is the The reduction in p70S6K activity in paclitaxel-treated phosphorylation of the serine/threonine–proline sites cells was not because of a decrease in total p70S6K including T421 and S424 in the autoinhibitory domain abundance evidenced by Western blot analysis as (Avruch et al., 2001; Volarevic and Thomas, 2001). shown in Figure 4. Thus, paclitaxel inactivates p70S6K Therefore, we investigated the effect of paclitaxel on despite phosphorylation of p70S6KT421/S424. These data total p70S6K activity and found that paclitaxel sup- further confirm that phosphorylation of p70S6KT421/S424 pressed p70S6K activity in a dose-dependent (Figure 4a) is not necessarily concordant with activation of and time-dependent (Figure 9b) manner. The activity p70S6K.

Oncogene Toxol phosphorylatea and inactivates p70 S6k in mitosis X-F Le et al 488

a 200000 b

180000 40000 160000 35000

140000 30000 cpm)

120000 cpm) * 25000 * 100000 20000 80000 ** ** ** 60000 15000 ** ** 40000 10000 p70S6K activity ( activity p70S6K ** 20000 ( activity p70S6K 5000 0 0 0 10 20 100 200 2000 0 8 16 24 32 Concentration of paclitaxel (nM) Hours after paclitaxel treatment

Total p70S6K → Total p70S6K →

cde BT474 2008 SKOv3 25000 40000 500000

400000 20000 30000 *

300000 15000 ** 20000 200000 10000 ** 10000 100000 5000 p70S6K Activity (cpm) p70S6K Activity (cpm) p70S6K Activity (cpm) 0 0 0 Untreated Paclitaxel Untreated Paclitaxel Untreated Paclitaxel

Total p70S6K → Total p70S6K → Total p70S6K →

Figure 4 Paclitaxel inactivates p70S6K in dose- and time-dependent manner and in multiple cancer cell lines. SKBr3 cells or BT474, 2008, and SKOv3 cells were treated with paclitaxel at 20 nm or indicated concentration (a) for 24 h or at indicated time points (b). Cells were harvested and subjected to total protein isolation and immunoprecipitation with a pan-p70S6K antibody. A kinase assay speciﬁc for p70S6K was then performed as described in Experimental procedures. (a) Paclitaxel inactivates p70S6K in a dose-dependent manner. SKBr3 cells were treated with paclitaxel for 24 h at the indicated concentration and total protein was prepared and subjected to p70S6K kinase assay. *, Po0.05 compared with untreated group; **, Po0.01 compared with untreated group. (b) Paclitaxel inactivates p70S6K in a time-dependent manner. SKBr3 cells were treated with paclitaxel at 100 nm for indicated time points and total protein was prepared and subjected to p70S6K kinase assay. *, Po0.05 compared with untreated group; **, Po0.01 compared with untreated group. (c) Paclitaxel inactivates p70S6K in BT474 breast cancer cells. BT474 cells were treated with paclitaxel at 100 nm for 24 h and total protein was prepared and subjected to p70S6K kinase assay. **, Po0.01 compared with untreated group. (d) Paclitaxel inactivates p70S6K in 2008 ovarian cancer cells. In all, 2008 cells were treated with paclitaxel at 100 nm for 24 h and total protein was prepared and subjected to p70S6K kinase assay. **, Po0.01 compared with untreated group. (e) Paclitaxel inactivates p70S6K in SKOv3 ovarian cancer cells. SKOv3 cells were treated with paclitaxel at 100 nm for 24 h and total protein was prepared and subjected to p70S6K kinase assay. *, Po0.05 compared with untreated group. These results were conﬁrmed in three independent experiments

p70S6KT421/S424 phosphorylation requires both de novo paclitaxel-induced p70S6KT421/S424 phosphorylation even RNA and protein synthesis at a concentration of 0.1 mg/ml (lanes 7–10), demon- strating a requirement of transcription for paclitaxel- To determine whether transcription or protein synthesis induced p70S6K phosphorylation. Similarly, CHX is required for paclitaxel-induced phosphorylation of suppressed paclitaxel-upregulated p70S6K phosphoryla- p70S6KT421/S424 in SKBr3 cells, the effects of a transcription at concentrations of 10 and 5 mg/ml (Figure 5, lanes tional inhibitor actinomycin D (AD) and a protein 3–6). Thus, protein synthesis is also required for synthesis inhibitor cyclohexamide (CHX) were analysed. phosphorylation of p70S6KT421/S424 after paclitaxel treat- As shown in Figure 5, AD completely abolished ment.

Oncogene Toxol phosphorylatea and inactivates p70 S6k in mitosis X-F Le et al 489 CHX (10 µg/ml) - - + + ------CHX (5 µg/ml) - - - - + + - - - - AD (1 µg/ml) ------+ + - - AD (0.1 µg/ml) ------+ + Paclitaxel - + - + - + - + - +

← P-p70S6KT421/S424

←β-actin

1 2 3 4 5 6 7 8 9 10

Figure 5 Paclitaxel-induced phosphorylation of p70S6KT421/S424 requires both de novo RNA and protein synthesis. SKBr3 cells were ﬁrst pretreated with cycloheximide (CHX) at two concentrations (10 and 5 mg/ml) or with actinomycin D (AD) at two concentrations (1 and 0.1 mg/ml) for 2 h in complete medium. The cells were then treated with paclitaxel at 20 nm for 24 h. Cells were harvested and subjected to total protein isolation and Western blot analysis with anti-P-p70S6KT421/S424 and anti-b actin antibodies. These results were representative of three independent experiments

Effect of inhibition of ERK1/2, p38 MAP kinase, and el-induced p70S6K inactivation (Figure 7), which JNK on paclitaxel-indued p70S6K T421/S424 indicated that JNK activity was required for paclitaxel phosphorylation to inactivate p70S6K.

Although Ras/MAP kinase pathways are not the major upstream regulators of p70S6K, MAP kinases can Effect of inhibition of protein kinase C (PKC), phosphorylate p70S6K in vitro (Mukhopadhayay et al., intracellular calcium (Ca++), and phospholipid Cg 1992; Grammer et al., 1996). To investigate whether (PLCg) pathway on paclitaxel-indued p70S6K T421/S424 activity of MAP kinase components (ERK1/2, p38, and phosphorylation JNK pathways) is necessary for paclitaxel to induce p70S6KT421/S424 phosphorylation, SKBr3 cells were in- Since PKC (Romanelli et al., 1999), Ca++ (Conus et al., cubated with several selective MAP kinase inhibitors 1998), and their upstream regulator PLCg (Grammer prior to treatment with paclitaxel. PD98059 (Figure 6a, et al., 1996) have been linked to p70S6K phosphoryla- lanes 3–4) and PD184352 (Figure 6a, lanes 5–6), which tion and activity, it was of interest to investigate whether prevent the activation of MEK1 (and hence the PLCg–PKC–Ca++ signaling pathways were involved in activation of ERK1/2), decreased paclitaxel-stimulated the phosphorylation of p70S6KT421/S424 elicited by p70S6K phosphorylation. The efficacy of PD98059 and paclitaxel. Two pan-PKC inhibitors (GF109203X and PD184352 was confirmed by antiphospho-ERK anti- RO318425), two Ca++ chelators (MAPTAM and body (Figure 6b). Paclitaxel-stimulated phosphorylation BAPTA/AM), and one PLCg (U73122) were employed. of p70S6K was not affected by SB203580, a specific p38 As shown in Figure 8a, both pan-PKC inhibitors MAP kinase inhibitor, or by SB202190 that selectively diminished paclitaxel-induced p70S6KT421/S424 phosphor- inhibits p38 MAP kinase at low concentration (3 mm) ylation. Similarly, both Ca++ chelators decreased or (Figure 6c, lanes 3–6). The efficacy of SB203580 and abrogated paclitaxel-induced p70S6KT421/S424 phosphor- SB202190 was confirmed by antiphospho-p38 antibody ylation (Figure 8b). Consistent with the above observa- (Figure 6d). However, SP600125, a selective inhibitor of tions, the inhibitor of PLCg, U73122, abolished JNK activity (Bennett et al., 2001), blocked paclitaxel- paclitaxel-induced p70S6KT421/S424 phosphorylation stimulated p70S6K phosphorylation (Figure 6e). The (Figure 8c). These results suggest that PKC and Ca++ efficacy of SP600125 was confirmed by antiphospho- signaling pathways are involved in the phosphorylation JNK antibody (Figure 6f). These results indicated that of p70S6KT421/S424 induced by paclitaxel. paclitaxel-induced p70S6KT421/S424 phosphorylation in Ca++ chelator MAPTAM suppressed p70S6K activ- SKBr3 cells may involve MEK1 and JNK activity, but ity, whereas PKC inhibitor GF109203X stimulated does not require p38 MAP kinase activity. p70S6K activity (Figure 8d). These results indicated ERK1/2 inhibitors (PD98059 and PD184352) and p38 that PKCs and calcium signaling differentially regulated inhibitor SB203580 did not significantly affect p70S6K p70S6K activity, although both were implicated in basal activity, whereas JNK inhibitor SP600125 stimu- paclitaxel-induced p70S6KT421/S424 phosphorylation. lated p70S6K basal activity (Figure 7). Inhibition of Ca++ chelator MAPTAM had no dramatically syner- both ERK and p38 MAP kinases could not alter gistic effect on p70S6K inactivation with paclitaxel paclitaxel-induced p70S6K inactivation (Figure 7), (Figure 8d). PKC inhibitor GF109203X prevented which suggested that with or without p70S6KT421/S424 paclitaxel-induced p70S6K inactivation (Figure 8d), phosphorylation, paclitaxel was able to inactivate the which indicated that PKC activity was required for kinase. However, inhibition of JNK prevented paclitax- paclitaxel to inactivate p70S6K.

Oncogene Toxol phosphorylatea and inactivates p70 S6k in mitosis X-F Le et al 490 a b PD98059 - - + + - - PD184352 - - - - + + PD98059 - - + Paclitaxel- + - + - + PD184352 - + - P-p70S6KT421/S424 → Phospho-ERK1/2 ⇒

β-actin → TotalERK1/2 ⇒ 1 2 3 4 5 6

c SB203580 - - + + - - d Sorbitol - + - + - + SB202190 - - - - + + SB203580 -- + + -- Paclitaxel - + - + - + SB202190 - - - - + + P-p70S6KT421/S424 → Phospho-p38 →

β-actin → Total p38 → 1 2 3 4 5 6

e Paclitaxel - + - + f Sorbitol - + - + SP600125 - - + + SP600125 - - + + P-p70S6KT421/S424 → Phospho-JNK1 →

Total p70S6K→ Total JNK1 →

Figure 6 Paclitaxel-induced phosphorylation of p70S6KT421/S424 requires ERK1/2 MAP kinase and JNK, but not p38 MAP kinase. SKBr3 cells were first pretreated with individual inhibitors for 1 h in complete medium. The cells were then treated with paclitaxel at 20 nm for 24 h. Cells were harvested, total protein was isolated, and resolved on SDS–PAGE for Western blot analysis with anti-P- p70S6KT421/S424 and anti-b actin antibodies. (a) ERK1/2 MAP kinase is required. Cells were pretreated with MEK1-specific inhibitors PD98059 at 50 mm or PD184352 at 50 nm. These results were representative of three independent experiments. (b) Confirmation of the efficacy of PD98059 and PD184352. SKBr3 cells have relatively high basal level of phospho-ERK expression. (c) p38 MAP kinase is not required. Cells were pretreated with p38 MAP kinase inhibitors SB203580 at 2 mm or SB202190 at 3 mm. This experiment was repeated three times with similar results. (d) Confirmation of the efficacy of SB203580 and SB202190. SKBr3 cells were pretreated with inhibitors SB203580 at 2 mm or SB202190 at 3 mm for 1 h and then treated with 0.5 m sorbitol for 15min. (e) JNK is required. Cells were pretreated with SP600125 at 10 mm for 1 h. This experiment was repeated three times with similar results. (f) Confirmation of the efficacy of SP600125. SKBr3 cells were pretreated with SP600125 at 10 mm for 1 h and then treated with 0.5 m sorbitol for 15 min. Western blotting was performed with anti-phospho-JNK antibody

Effect of inhibition of PI3K and mTOR/FRAP/RAFT activity was observed in the combination with paclitaxel pathways on paclitaxel-indued p70S6KT421/S424 (Figure 9c). phosphorylation Phosphorylation of p70S6KT421/S424 and inactivation of PI3K and mTOR/FRAP/RAFT are believed to be the p70S6K are associated with the mitotic cell fraction major upstream effectors of p70S6K (Grammer et al., among paclitaxel-treated cells 1996; Avruch et al., 2001; Templeton, 2001; Volarevic and Thomas, 2001). Therefore, their involvement in the As mentioned above, paclitaxel induced significant G2/M induction of p70S6KT421/S424 phosphorylation was exam- arrest and apoptosis in SKBr3 cells, which resulted in ined using selective PI3K inhibitors (LY294002 and two cell fractions: suspended and attached. We have wortmannin) and the mTOR kinase inhibitor, rapamy- used a shake-off technique as described in Experimental cin. Cotreatment of the cells with paclitaxel and procedures to examine which cell fraction is responsible LY294002 or wortmannin resulted in decreased or for the phosphorylation of p70S6KT421/S424 and inactiva- absent p70S6KT421/S424 phosphorylation (Figure 9a). As tion of p70S6K after paclitaxel treatment. SKBr3 cells in expected, rapamycin also decreased paclitaxel-induced 175-cm2 culture flasks were treated with 100 nm pacli- p70S6KT421/S424 phosphorylation (Figure 9b). The inhi- taxel for the indicated intervals and then gently shaken, bitor studies indicated that PI3K and mTOR were permitting us to separate a detached (suspended) required for the induction of phosphorylation of fraction and an attached fraction of cells. The cells p70S6KT421/S424 by paclitaxel. As expected, both rapamy- from the two fractions at each time point were subjected cin and LY294002 significantly suppressed p70S6K to Giemsa staining and cell cycle analysis. Giemsa activity (Figure 9c). No further decrease in p70S6K staining clearly showed that there are few mitoses

Oncogene Toxol phosphorylatea and inactivates p70 S6k in mitosis X-F Le et al 491 60000 * 50000

40000

30000

20000 Kinase activity (in CPM) 10000

0 PD98059 PD59+Tax PD184352 PD52+Tax SB203580 SB+Tax SP600125 SP+Tax Untreated Paclitaxel

Total p70S6K →

Figure 7 Inhibition of JNK prevented paclitaxel-induced S6K inactivation, whereas inhibition of ERK and p38 MAP kinases could not. SKBr3 cells were ﬁrst pretreated with individual inhibitors for 1 h in complete medium. The cells were then treated with paclitaxel at 20 nm for 24 h. A kinase assay speciﬁc for p70S6K was then performed as described in Experimental procedures. *, Po0.05 compared with untreated group. This experiment was repeated twice with similar results a GF109203X - - + + - - c U73122 - - + + RO318425 - - - - + + Paclitaxel - + + - Paclitaxel - + - + - + P-p70S6KT421/S424 → P-p70S6KT421/S424 → β-actin → β-actin→

d 60000 *

50000

40000 b BAPTA/AM - - - - + + MAPTAM - - + + - - 30000 Paclitaxel - + - + - + 20000 * Kinase activity (in CPM) 10000 P-p70S6KT421/S424 →

0 U Pac U73122 U7 M M G GF+Tax n F A A 3 t 109203X li

β → r PT +Ta +Ta

-actin e taxe a t AM e x x d l

Total p70S6K →

Figure 8 Paclitaxel-induced phosphorylation of p70S6KT421/S424 requires PLCg–PKC–Ca++ signaling pathways. SKBr3 cells were first pretreated with individual inhibitors for 1 h in complete medium. The cells were then treated with paclitaxel at 20 nm for 24 h. Cells were harvested and subjected to total protein isolation and Western blot analysis with anti-P-p70S6KT421/S424 and anti-b actin antibodies. (a) PKC pathway is required. Two highly selective pan-PKC inhibitors GF109203X at 5 mm and RO 318425 at 2.5 nm were used for the pretreatment. (b) Intracellular Ca++ signaling is required. Ca++ chelators MAPTAM at 20 mm and BAPTA/AM at 25 mm were used for the pretreatment. (c) PLCg signaling is required. PLCg inhibitor U73122 at 4 mm was used for pretreatment. (d) p70S6K kinase assay. SKBr3 cells were first pretreated with individual inhibitors for 1 h in complete medium. The cells were then treated with paclitaxel at 20 nm for 24 h. A kinase assay specific for p70S6K was then performed. *, Po0.05 compared with untreated group. This experiment was repeated twice with similar results

Oncogene Toxol phosphorylatea and inactivates p70 S6k in mitosis X-F Le et al 492 a - - - - + + Wortmannin c LY294002 - - + + - - 30000 Paclitaxel - + - + - + 25000 P-p70S6KT421/S424 → 20000 β-actin → 15000 10000 5000 Kinase activity (in CPM 0 b Rapamycin - - + + Untreated Paclitaxel Rapamyc Rapa+Tax LY294002 LY+Tax Paclitaxel - + - + P-p70S6KT421/S424 → in

β-actin → Total p70S6 →

Figure 9 Paclitaxel-induced phosphorylation of p70S6KT421/S424 requires PI3K and mTOR/FRAP signaling pathways. SKBr3 cells were first pretreated with individual inhibitors for 1 h in complete medium. The cells were then treated with 20 nm paclitaxel for 24 h. Cells were harvested, total protein was isolated and resolved for Western blot analysis with anti-P-p70S6KT421/S424 and anti-b actin antibodies. (a) PI3K signaling is required. Two highly selective PI3K inhibitors LY294002 at 10 mm and wortmannin at 2 mm were used for pretreatment. (b) mTOR/FRAP signaling is required. MTOR-specific inhibitor rapamycin at 55 nm were used for pretreatment. (c) p70S6K kinase assay. SKBr3 cells were first pretreated with rapamycin or LY294002 for 1 h in complete medium. The cells were then treated with paclitaxel at 20 nm for 24 h. A kinase assay specific for p70S6K was then performed. *, Po0.05 compared with untreated group. This experiment was repeated twice with similar results

among untreated SKBr3 cells and that the majority of fraction of mitotic cells in detached cells. Taken cells from the attached fraction at 24 h time point were together, these data suggest that paclitaxel-induced not mitosis (data not shown). By contrast, the majority phosphorylation of p70S6KT421/S424 and inactivation of of detached cells were arrested in mitosis. Detached cells p70S6K is associated with cells in mitosis. had a significantly higher mitotic index than did the cells from the attached fraction (data not shown). Corre- spondingly, cell cycle analysis showed that detached Discussion cells had a significantly higher G2/M component than did the attached fraction (data not shown). Interest- Phosphorylation of p70S6K in mammalian cells is often ingly, the attached cells that exhibited an increased G2/ triggered by mitogenic stimuli such as growth factors, M peak did not have higher mitotic indices. Thus, cells hormones, and amino acids (Avruch et al., 2001; in the attached fractions were in early stage of G2 phase, Volarevic and Thomas, 2001). Multiple sites of serine but not in M phase. By contrast, the detached fractions and threonine phosphorylation on p70S6K are believed that had an increased G2/M peak also had higher to be essential to its activation. One current model for mitotic indices (data not shown), indicating that p70S6K activation suggests that upon mitogen stimula- detached cells were in the M phase. tion, the four proline-directed sites at the C-terminus As shown in Figure 10a, no significant phosphoryla- (Ser424, Thr421, Ser418, and Ser411) are phosphorylated. tion of p70S6KT421/S424 was observed in the attached Subsequently, Ser404, Ser371, and Thr389 are phosphory- fractions. The p70S6K activity in the attached fractions lated. The final step in p70S6K activation involves decreased slightly and slowly within 24 h after paclitaxel phosphorylation of the Thr229 on activation loop. treatment, but recovered completely (Figure 10b). In (Grammer et al., 1996; Avruch et al., 2001; Templeton, sharp contrast, significant phosphorylation of 2001; Volarevic and Thomas, 2001). Here, we have p70S6KT421/S424 was observed in four detached fractions demonstrated for the first time that paclitaxel as a (Figure 10c, lanes 2–5), and correlated with the mitotic microtubule-damaging agent is able to induce phos- index. After 32 h, the majority of cells in detached phorylation of p70S6KT421/S424. Paclitaxel-induced phos- fractions were apoptotic cells that did not contain phorylation of p70S6KT421/S424 is concurrent with phosphorylated p70S6KT421/S424 (Figure 10c, lanes 6–8). decreased activity of p70S6K. Our data show that The p70S6K activity in the detached fractions decreased inhibitors of mTOR, PI3K, and Ca++ impair p70S6K dramatically within the first 16 h after paclitaxel treat- activity, whereas inhibitors of JNK and PKC stimulate ment and reached a minimum at 16 h (Figure 10d). p70S6K activity. Inhibition of PKC and JNK prevents p70S6K activity recovered slightly at later intervals, but paclitaxel-induced p70S6K inactivation. These results did not return to control levels (Figure 10d). A high level not only confirm that phosphorylation of p70S6KT421/S424 of cyclin B1 expression could be detectable in the is not always associated with activation of p70S6K, but detached fraction (Figure 10c), consistent with the high also demonstrate that phosphorylation of p70S6KT421/S424

Oncogene Toxol phosphorylatea and inactivates p70 S6k in mitosis X-F Le et al 493

a Hours: 0 4 8 16 24 32 48 56 72 c Hours: 0 8 16 24 32 48 56 72 p-p70S6KT421/S424 → p-p70S6KT421/S424 → Cyclin B1 → Cyclin B1 → Total p70S6K→ Total p70S6K→ 1 2 3 4 5 6 7 8

d b Attached cells Shaken-off cells 150000 200000

(cpm) 150000 100000 100000

Activity (cpm) 50000 50000 0 0 0 8 16 24 48 72 0 8 16 24 48 72

p70S6K p70S6K Activity Hours after Paclitaxel treatment Hours after Paclitaxel treatment

Figure 10 Paclitaxel induces phosphorylation of p70S6KT421/S424 and inactivates p70S6K in detached cells with high mitotic index. SKBr3 cells in exponential growth were treated with paclitaxel at 20 nm for indicated intervals. Detached cells were obtained by shaking the treated cell population firmly for 1 min and removing the medium as described in Experimental procedures. Attached cells were obtained by trypsinization. Cells were then subjected to total protein preparation, Western blot analysis, and p70S6K kinase assay. (a) No obvious phosphorylation of p70S6KT421/S424 and cyclin B1 expression were detected in the attached fractions by Western blotting. (b) The attached cells largely maintain the p70S6K activity after paclitaxel treatment. (c) Significant phosphorylations of p70S6KT421/S424 and cyclin B1 expression were detected in the detached cells by Western blotting. (d) The detached cells display significantly weak p70S6K activity after paclitaxel treatment. This experiment was repeated three times with similar results and kinase inactivation induced by paclitaxel are paclitaxel-induced inactivation of p70S6K. This study regulated by different mechanisms. has shown that inhibition of ERK1/2 MAP kinase, The significance of paclitaxel-induced Thr421 and JNK, PKC, Ca++, PI3K, or mTOR diminishes or Ser424 phosphorylation of p70S6K is not known yet. abolishes phosphorylation of p70S6KT421/S424. Inhibitors Based on our data and that in the literature, Thr421 and of mTOR, PI3K, and Ca++ impair p70S6K activity, Ser424 phosphorylation of p70S6K may or may not be whereas inhibitors of JNK and PKC stimulate activity. important for inhibition of the enzyme. A mitosis- These results suggest that phosphorylation of related inhibitory protein, human prolyl isomerase Pin1, p70S6KT421/S424 is not related to regulation of kinase has been shown to inhibit G2/M progression by activity. Our unpublished data indicated that conversion interacting with phospho-proteins, such as NIMA, of residues Thr421 and Ser424 to Ala fails to decrease the Cdc25, and Wee1 (Zhou et al., 1999; Nakamura et al., basal level of p70S6K activity and to retard paclitaxel- 2001). Interestingly, Pin1 is reported to interact with induced inactivation of p70S6K. There are at least two p70S6K (Yaffe et al., 1997). Data in Figure 10 show that reports, which indicate that mutation of four carboxyl- paclitaxel-induced phosphorylation and inactivation of terminal phosphorylation sites has little or no effect on p70S6K occur exclusively in mitotic cells. One possibi- the activity of p70S6K (Edelmann et al., 1996; lity is that paclitaxel-induced phosphorylation of Thr421 Mahalingam & Templeton, 1996). Thus, the significance and Ser424 of p70S6K could create docking sites for Pin1- of paclitaxel-induced Thr421 and Ser424 phosphorylation like inhibitory factors, which can then bind and of p70S6K remains to be elucidated. Probably, it may isomerize the phosphorylated Ser/Thr–Pro peptide play a role in subsequent activation or inactivation bond. As a result of the binding of Pin-like inhibitory steps. factor(s), the structure of p70S6K becomes changed and Which kinase phosphorylates Thr421 and Ser424 of its activity suppressed. Indeed, Ser411 of p70S6K, a close p70S6K in paclitaxel-treated cells is not clear. Auto- residue of Thr421 and Ser424, has been shown to be inhibitory domain kinases have been proposed as phosphorylated by Cdc2-cyclin B complex in vivo, which candidates, including Cdc2 and the mitogen-activated may cause the inactivation of p70S6K during mitosis kinase ERK1/2 MAP kinase (Mukhopadhayay et al., (Papst et al., 1998). Our unpublished data also indicate 1992). ERK1/2 MAP kinase is unlikely the kinase to that SKBr3 cells in exponential growth display higher phosphorylate these two sites, since paclitaxel treatment phosphorylation of Thr421 and Ser424 and concurrent inactivates ERK1/2 (Figure 1b). Cdc2 (Cdk1) has been lower p70S6K activity at cell cycle G2/M phase, where shown to phosphorylate p70S6K at Ser411 in mitosis protein synthesis is suppressed. These data suggest that (Papst et al., 1998). However, there is no significant phosphorylation of Thr421 and Ser424 may play a role in phosphorylation of Ser411 in paclitaxel-treated cells the inactivation of p70S6K. (Figure 3b). Therefore, Cdc2 will not be a strong However, our data and others suggest that phosphory- candidate. Other members of the proline-directed kinase lation of p70S6KT421/S424 may not be important for the family may phosphorylate these sites. p38 MAP kinase

Oncogene Toxol phosphorylatea and inactivates p70 S6k in mitosis X-F Le et al 494 has been indicated to phosphorylate the C-terminus of cells lack IFG-1-stimulated phosphorylation of both p70S6K in vitro (Grammer et al., 1996). However, Thr229 and Thr389 of p70S6K (Williams et al., 2000). It inactivation of p38 kinase by paclitaxel treatment was reported that the rate of decline in p70S6K activity (Figure 1b) has made it unlikely. JNK may phosphor- after addition of wortmannin correlated more closely ylate Thr421 and Ser424 (Grammer et al., 1996). Given the with the dephosphorylation of Thr389 than of Thr229 upregulation of JNK phosphorylation in paclitaxel- (Weng et al., 1998), pointing to a crucial role for Thr389 treated cells (Figure 1b), our data are consistent with phosphorylation in p70S6K activation. Our data here this hypothesis. Additionally, our data show that clearly showed that inhibition of the PI3K pathway paclitaxel-induced phosphorylation of p70S6K occurs decreased p70S6KT421/S424 phosphorylation, indicating exclusively in mitotic cells, suggesting that cells at cell that except Thr229 and Thr389, PI3K pathway may cycle M phase contain protein kinase(s) that phosphor- phosphorylate p70S6KT421/S424. ylate Thr421 and Ser424. Another possibility is that mTOR/FRAP is another important upstream regu- paclitaxel may activate microtubule-related kinase(s) lator of p70S6K through two distinct mechanisms. that phosphorylate Thr421 and Ser424. Further studies are mTOR/FRAP has been shown in vitro to phosphorylate needed to address these questions. Thr389 directly (Burnett et al., 1998; Isotani et al., 1999) In this study, we have identified several upstream and possibly to phosphorylate other p70S6K residues signaling pathways required for paclitaxel-induced (Dennis et al., 1996). Thr389 phosphorylation is blocked phosphorylation of p70S6K, which could be direct or by rapamycin (Pearson et al., 1995). mTOR/FRAP may indirect effects. Inhibition of JNK or ERK1/2 MAP also control p70S6K phosphorylation and activity kinase decreased paclitaxel-induced p70S6KT421/S424 through protein phosphatase 2A (PP2A). PP2A has phosphorylation (Figure 5), consistent with previous been shown to associate directly with and to depho- reports that ERK1/2 MAP kinases can contribute to sphorylate p70S6K (Peterson et al., 1999). The mTOR/ phosphorylation of serine/threonine sites at the C- FRAP kinase inhibits PP2A-mediated p70S6K depho- terminal of p70S6K (Mukhopadhayay et al., 1992; sphorylation (Peterson et al., 1999). Our data presented Grammer et al., 1996). JNK was stimulated by in Figure 7b support the requirement of the mTOR/ paclitaxel treatment as shown in Figure 1, raising the FRAP signaling pathway for p70S6K phosphorylation possibility that JNK may also contribute to phosphor- at Thr421 and Ser424 residues. ylation of p70S6K. Surprisingly, inhibition of p38 MAP This study has demonstrated that paclitaxel is able to kinase did not affect paclitaxel-induced phosphoryla- decrease substantially p70S6K activity. As p70S6K is an tion. important factor for cell growth and survival, paclitaxel Atypical PKCs, specifically PKCl and PKCz,have may exert its antitumor effect in part via inhibition of been shown to regulate p70S6K phosphorylation and p70S6K. Several factors could contribute to inactivation activity (Akimoto et al., 1998; Romanelli et al., 1999). of p70S6K in paclitaxel-treated cells. One of the major Depletion of intracellular stores of Ca2+ has been shown upstream regulators of p70S6K, AKT, is inhibited by to abolish p70S6K activation, but has no effect on AKT paclitaxel treatment (Figure 1b). Avruch’s lab has activation (Conus et al., 1998). In agreement with these recently identified a never-in-mitosis/Aspergillus reports, our data in Figure 6 demonstrate that PKC and (NIMA) kinase, termed NIMA-related kinase (NEK) Ca++ signaling pathways were involved in the paclitax- 6, as the kinase that physiologically phosphorylates el-induced phosphorylation of p70S6KT421/S424. Inhibi- Thr389 (Belham et al., 2001). The NIMA kinases tion of PLCg, the upstream regulator of PKC and Ca++, associate with mitotic spindles, and phosphorylate the also diminished phosphorylation of p70S6KT421/S424 mitotic histone H3 (Ed Souza et al., 2000), whereas (Figure 6c), further supporting the role of PKC and paclitaxel targets microtubule directly (Schiff & Ca++ signaling pathways in the phosphorylation of Horwitz, 1980; Horwitz, 1992). Thus, paclitaxel could p70S6KT421/S424. directly affect the NIMA kinases including NEK6, p70S6K has been demonstrated to be a downstream preventing phosphorylation of Thr389. Indeed, our data effector of the PI3K signaling pathway (Chung et al., in Figure 3b indicate that paclitaxel fails to phosphor- 1994; Ming et al., 1994; Reif et al., 1997). PI3K ylate Thr389 in spite of substantial phosphorylation of inhibitors, wortmannin and LY294002, block the Thr421 and Ser424. activation of p70S6K and constitutively active forms In summary, microtubule-damaging agent paclitaxel of PI3K activate p70S6K (Chung et al., 1994; Reif et al., is able to induce p70S6KT421/S424 phosphorylation and 1997). Phosphorylation of Thr229 and Thr389 can be decreases its activity simultaneously. Paclitaxel-induced inhibited by wortmannin, an inhibitor of PI3K activity p70S6KT421/S424 phosphorylation requires both de novo (Ming et al., 1994). Whereas PI3K does not appear to RNA and protein synthesis via multiple signaling phosphorylate p70S6K directly, the generation of 3- pathways including ERK1/2 MAP kinase, JNK, PKC, phosphoinositide (3-PtdIns) products by PI3K and Ca++, PI3K, and mTOR. However, these different PDK1 are required for both Thr229 and Thr389 phos- signaling pathways regulate p70S6K activity differently. phorylation (Han et al., 1995; Balendran et al., 1999). Inhibitors of mTOR, PI3K, and Ca++ impair p70S6K PDK1 has been shown to be the physiologic p70 Thr229 activity and cannot further augment paclitaxel-induced kinase (Alessi et al., 1998; Pullen et al., 1998). PDK1 is p70S6K inactivation. Inhibitors of JNK and PKC later confirmed to phosphorylate Thr389 directly as well stimulate activity and prevent paclitaxel-induced (Balendran et al., 1999). PDK1 null embryonic stem p70S6K inactivation. Inhibitors of ERK and p38 (that

Oncogene Toxol phosphorylatea and inactivates p70 S6k in mitosis X-F Le et al 495 has no effect on p70S6KT421/S424 phosphorylation) MAP p70S6KT389 were obtained from New England BioLabs, Inc. kinases have no effect on p70S6K activity and cannot (Beverely, MA, USA). Antibodies to phospho-ERK1/2, prevent paclitaxel-induced p70S6K inactivation. More- phospho-JNK, and total ERK1/2 were obtained from over, paclitaxel-induced phosphorylation and low activ- Promega Corporation (Madison, WI, USA). Monoclonal ity of p70S6K occur only in mitotic cells. Our data antibody to b-actin was obtained from Sigma (St Louis, suggest that paclitaxel-induced p70S6KT421/S424 phos- MO, USA). Rabbit antibodies to phosphoserine and phosphothreonine were obtained from Zymed (San Francisco, CA, phorylation and inactivation are differentially regulated. USA). Monoclonal antibody to phosphotyrosine was obtained Paclitaxel may exert its antitumor effect in part via from Upstate Biotechnology (Lake Placid, NY, USA). inhibition of p70S6K. Preparation of total protein and western immunoblot analysis The procedures for the preparation of total protein and Materials and methods Western immunoblot analysis were performed as described previously (Le et al., 2001). Cell culture The human breast cancer cell lines SKBr3 and BT474, and Cell cycle analysis human ovarian cancer cell line SKOv3 were obtained from the Cell cycle distribution was analysed by flow cytometry. The American Type Culture Collection (ATCC, Rockville, MD, treated cells were trypsinized, washed once with PBS, and fixed USA). The human ovarian cancer cell line 2008 was obtained overnight in 70% ethanol. Fixed cells were centrifuged at 300 g from Dr Zahid Siddik at the Department of Experimental for 10 min and washed with PBS. Cell pellets were resuspended Therapeutics, MD Anderson Cancer Center. SKBr3 was in PBS containing 50 mg/ml of RNase A and 50 mg/ml grown in complete medium containing RPMI 1640 (GIBCO, propidium iodide and incubated for 20 min at 371C with Grand Island, NY, USA) supplemented with 10% fetal bovine gentle shaking. Stained cells were filtered through nylon mesh serum (FBS) (Sigma, St Louis, MO, USA), 2 mml-glutamine, (41 mm) and analysed on a Coulter flow cytometer XL-MCL 100 U/ml penicillin, and 100 mg/ml streptomycin in humidified (Coulter Corporation, Miami, FL, USA) for relative DNA air with 5% CO at 37oC. BT474 and 2008 cells were grown in 2 content based on red fluorescence levels. Doublets and cell complete medium containing DMEM (GIBCO, Grand Island, debris were excluded from the DNA histograms. The NY, USA), supplemented with 10% FBS, 2 mml-glutamine, percentages of sub-G1 cell population were determined based 100 U/ml penicillin, and 100 mg/ml streptomycin. SKOv3 cells on relative DNA content. The percentages of cells in different were maintained in McCoy 5A medium containing 10% FBS, cell cycle compartments were determined using the MULTI- 2mml-glutamine, 100 U/ml penicillin, and 100 mg/ml strepto- CYCLE software program (Phoenix Flow Systems, San mycin. For all experiments, cells were detached with 0.25% Diego, CA, USA). trypsin-0.02% EDTA. For cell culture, 5 Â 105 exponentially growing cells were plated into 100-mm tissue culture dishes or 75–175-cm2 culture flasks. After culture for 24 h in complete Phosphoamino-acid analysis medium, cells were treated (if applicable) with or without Immunoprecipitated p70S6K from 32P-labeled SKBr3 cells was paclitaxel for the desired period in the presence or absence of separated by SDS–PAGE and transferred to a PVDF either the PLCg-specific inhibitor U73122 (4 mm), the Ca++ membrane (Immobilon-P, Millipore). The portion containing chelators MAPTAM (20 mm) and BAPTA/AM (25 mm), the phosphorylated p70S6K was excised after autoradiography, pan-PKC inhibitors Ro318425 (2.5 nm) and GF109203 (5 mm), and the membrane was washed 5 times with deionized water. the PI3K inhibitors LY294002 (10 mm) or wortmannin (2 mm), The analysis was performed as described in our previous the ERK1/2 MAP kinase inhibitor PD98059 (50 mm) and report (Le et al., 1999), with some modification. For PD184352 (50 nm), the p38 MAP kinase inhibitor SB203580 phosphoamino-acid analysis, the p70S6K on Immobilon-P (2 mm) and SB202190 (3 mm), JNK inhibitor SP600125 (10 mM), (Millipore) was incubated in 200 mlof6n HCl (Pierce) at or the FRAP/mTOR inhibitor rapamycin (55 nm) in complete 1101C for 90 min. After evaporation of HCl, the samples were medium. mixed with phosphoamino-acid standard (1 mg each of cold phosphoserine, phosphothreonine, and phosphotyrosine (Sig- Reagents ma)) in pH 1.9 buffer (H2O/88% formic acid/glacial acetic acid, 897 : 25 : 78), and the mixtures were applied to a thin- Paclitaxel and Giemsa were purchased from Sigma (St Louis, layer chromatography (TLC) plate (JT Baker) and separated MO, USA). DTX was obtained from Aventis Pharmaceuticals with an HTLE-7000 electrophoresis system (CBS, Del Mar, (Collegeville, PA, USA). A water-soluble PG–TAX was made CA, USA). Electrophoresis was carried out in pH 1.9 buffer at according to previous report (Li et al. (1998), also available 1500 V for 45 min for the one-dimension separation. Positions from Cell Therapeutics, Inc., Seattle as Xyotax). Chemical of phosphoamino acids were determined by ninhydrin (0.25% inhibitors U73122, LY294002, GF109203, the Ca++ chelators in acetone) staining. MAPTAM and BAPTA/AM, and PD98059 were purchased from BIOMOL Research Laboratories, Inc. (Plymouth Meet- Metabolic labeling and immunoprecipitation ing, PA, USA). MEK1-specific inhibitor PD184352 (2-[2- chloro-4-iodo-phenylamino]-N-cyclopropylmethoxy-3, 4-di- SKBr3 cells (1 Â 107) were incubated in phosphate-free RPMI fluoro-benzamide), wortmannin, SB203580, Ro318425, 1640 medium containing 10% dialyzed-FBS and treated with SP600125, and rapamycin were obtained from Calbiochem- 20 nm paclitaxel for 24 h, followed by the addition of Novabiochem Corp. (La Jolla, CA, USA). Antibodies to [32P]orthophosphate at 2 mCi/ml. After 4 h of labeling, cells phospho-JNK, total JNK, and pan-p70S6K were purchased were washed with phosphate-free medium twice and lysed in from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). radioimmunoprecipitation assay (RIPA) buffer (20 mm Tris Antibodies to phospho-p38, phospho-AKT, phospho- (pH 7.5), 150 mm NaCl, 1% NP-40, 0.5% deoxycholic acid, p70S6KT421/Ser424, phospho-p70S6KSer411, and phospho- 0.1% SDS, 2 mm EDTA, 1 mm EGTA) with proteinase

Oncogene Toxol phosphorylatea and inactivates p70 S6k in mitosis X-F Le et al 496 inhibitors (1 mm PMSF, 2 mg of aprotinin per ml, and 2 mgof p70S6K kinase activity (in cpm) from each sample was leupeptin per ml) and phosphatase inhibitors (50 mm NaF and corrected for background. Each sample was tested in 1mm sodium orthovanadate). The lysate was spun at 14 000 g, duplicate. and the supernatant was cleared with protein A beads for 30 min, followed by incubation with anti-p70S6K antibody at Giemsa staining 41C for 1.5 h. The immune complex was captured with protein A beads for 1 h, washed with RIPA buffer four times, SKBr3 cells untreated or treated with paclitaxel or obtained resuspended in loading buffer, and separated by SDS-PAGE. from the shake-off experiments were cytospun on slides. After The separated samples were transferred to a membrane and air drying, slides were fixed in cold methanol for 10 min and visualized by autoradiography. then subjected to Giemsa staining according to our previous report (Panditax et al., 1994). p70S6K activity assay Mechanical shake-off experiment p70S6K activity was measured by an immune complex kinase assay using an S6 peptide, RRRLSSLRA, as a substrate Asynchronous SKBr3 cells growing in 175-cm2 culture flasks (Santa Cruz). Briefly, total cell lysates were prepared from were treated with 20 nm paclitaxel for 24 h in 50 ml of culture SKBr3, BT474, 2008, or SKOv3 cells treated with diluent or medium. Fractions (detached or shake-off) containing mitotic paclitaxel 100 nm. Cells were harvested at the times indicated cells were obtained by shaking the flask firmly for 1 min. Cell and lysed in 300 ml of buffer A containing 20 mm Tris (pH 7.5), culture media were then pipetted as described previously 150 mm NaCl, 1 mm EDTA, 1 mm EGTA, 1% (v/v) Triton X- (Boonstra et al., 1981). Cells that remained adherent to culture 100, 2.5 mm sodium pyrophosphate, 1 mm glycerol phosphate, flasks were designated as attached fractions. 1mm Na3VO4,1mg/ml leupeptin, 10 mg/ml aprotinin, and 1mm phenylmethylsulfonyl fluoride. The cell lysates were Transient transfection clarified by centrifugation at 10 000 g for 10 min at 41C. The supernatant fractions containing equal amounts of proteins SKBr3 cells grown in 60 mm dishes were transfected with were incubated with a pan-p70S6K antibody at 41C for 1.5 h p70S6K wild-type and mutants constructs according to a and then for an additional 1.5 h with protein A-Sepharose previously described procedures (Le et al., 2001). beads (Santa Cruz). After washing four times with PBS, the immunoprecipitates were incubated at 301C for 30 min in a Statistical analysis m mixture of the following: 20 ml of assay dilution buffer (20 m The two-tailed Student’s t-test was used to compare two MOPS, pH 7.2, 25 mm glycerol phosphate, 5 mm EGTA, 1 mm m different groups. Values with Po0.05 were considered Na3VO4, and 1 m DTT), 10 ml of substrate mixture (S6 significant. peptide in assay dilution buffer), 10 ml of PNPP solution, and 10 mlof[g-32P]ATP (1 mCi/ml, Amersham). After a brief spin, each sample was spotted onto numbered P81 paper circles (Whatman) and washed three times (10 min each) with 2% Acknowledgments phosphoric acid and once (10 min) with acetone. Each sample We sincerely thank Drs J Avruch and DJ Templeton for paper was transferred into a scintillation vial containing 5 ml providing mutant plasmids of p70S6K and Dr Z Siddik for of scintillation fluid and then counted in a liquid scintillation providing 2008 cells. We also thank Karen Ramirez at Flow counter. At the same time, immunoprecipitates isolated by Cytometry Core Laboratory (Smith Research Building) of nonimmune IgGserum (Santa Cruz) instead of the p70S6K MD Anderson Cancer Center for her expert assistance with antibody were used as background controls. The actual flow cytometry analysis.

References

Akimoto K, Nakaya M, Yamanaka T, Tanaka S, Matsuda S, Chung J, Kuo CJ, Crabtree GR and Blenis J. (1992). Cell, 69, Weng QP, Avruch J and Ohno S. (1998). Biochem. J., 335, 1227–1236. 417–424. Conus NM, Hemmings BA and Pearson RB. (1998). J. Biol. Alessi DR, Kozlowski MT, Weng QP, Morrice N and Avruch Chem., 273, 4776–4782. J. (1998). Curr. Biol., 8, 69–81. Dennis PB, Pullen N, Kozma SC and Thomas G. (1996). Mol. Avruch J, Belham C, Weng Q, Hara K and Yonezawa K. Cell. Biol., 16, 6242–6251. (2001). Prog. Mol. Subcell. Biol., 26, 115–154. Edelmann HM, Kuhne C, Petritsch C, Ballou LM. (1996). J. Balendran A, Currie R, Armstrong CG, Avruch J and Alessi Biol. Chem., 271, 963–971. DR. (1999). J. Biol. Chem., 274, 37 400–37 406. Ed Souza CP, Osmani AH, Wu LP, Spotts JL and Osmani SA. Belham C, Comb MJ and Avruch J. (2001). Curr. Biol., 11, (2000). Cell, 102, 293–302. 1155–1167. Grammer TC, Cheatham L, Chou MM and Blenis J. (1996). Bennett BL, Sasaki DT, Murray BW, O’Leary EC, Sakata ST, Cancer Surv., 27, 271–292. Xu W, Leisten JC, Motiwala A, Pierce S, Satoh Y, Bhagwat Han JW, Pearson RB, Dennis PB and Thomas G. (1995). J. SS, Manning AM and Anderson DW. (2001). Proc. Natl. Biol. Chem., 270, 21 396–21 403 Acad. Sci. USA, 98, 13 681–13 686. Harada H, Andersen JS, Mann M, Terada N and Boonstra J, Mummery CL, Tertoolen LGJ, van der Saag PT Korsmeyer SJ. (2001). Proc. Natl. Acad. Sci. USA, 98, and de Laat SW. (1981). J. Cell Physiol., 107, 75–83. 9666–9670. Burnett PE, Barrow RK, Cohen NA, Snyder SH and Sabatini Hashemolhosseini S, Nagamine Y, Morley SJ, Desrivieres S, DM. (1998). Proc. Natl. Acad. Sci. USA, 95, 1432–1437. Mercep L and Ferrari S. (1998). J. Biol. Chem., 273, Chung J, Grammer TC, Lemon KP, Kazlauskas A and Blenis 4424–4429. J. (1994). Nature, 370, 71–75. Horwitz SB. (1992). Trends Pharmacol. Sci., 13, 134–136.

Oncogene Toxol phosphorylatea and inactivates p70 S6k in mitosis X-F Le et al 497 Isotani S, Hara K, Tokunaga C, Inoue H, Avruch J and Pearson RB, Dennis PB, Han JW, Williamson NA, Kozma Yonezawa K. (1999). J. Biol. Chem., 274, 34 493–34 498. SC, Wettenhall RE and Thomas G. (1995). EMBO J., 14, Jefferies HB, Fumagalli S, Dennis PB, Reinhard C, Pearson 5279–5287. RB and Thomas G. (1997). EMBO J., 16, 3693–3704. Peterson RT, Desai BN, Hardwick JS and Schreiber SL. Lane HA, Ferna´ ndez A, Lamb NJC and Thomas G. (1993). (1999). Proc. Natl. Acad. Sci. USA, 96, 4438–4442. Nature, 363, 170–172. Price DJ, Grove JR, Calvo V, Avruch J and Bierer BE. (1992). Le XF, Groner Y, Kornblau SM, Gu Y, Hittelman WN, Science, 257, 973–977. Levanon D, Mehta K, Arlinghaus RB, Chang KS. (1999). J. Pullen N, Dennis PB, Andjelkovic M, Dufner A, Kozma SC, Biol. Chem., 274, 21 651–21 658. Hemmings BA and Thomas G. (1998). Science, 27, Le XF, Marcelli M, McWatters A, Nan B, Mills GB, O’Brian 707–710. C and Bast Jr RC. (2001). Oncogene, 20, 8258–8269. Reif K, Burgering BMT and Cantrell DA. (1997). J. Biol. Li C, Yu DF, Newman RA, Cabral F, Stephens LC, Hunter Chem., 272, 14 426–14 433. N, Milas L, Wallace S. (1998). Cancer Res., 58, 2404–2409. Romanelli A, Martin KA, Toker A and Blenis J. (1999). Mol. Luo Y, Marx SO, Kiyokawa H, Koff A, Massague J and Cell. Biol., 19, 2921–2928. Marks AR. (1996). Mol. Cell. Biol., 16, 6744–6751. Schiff PB and Horwitz SB. (1980). Proc. Natl. Acad. Sci. USA, Mahalingam M and Templeton DJ. (1996). Mol. Cell. Biol., 77, 1561–1565. 16, 405–413. Shima H, Pende M, Chen Y, Fumagalli S, Thomas Gand Ming XF, Burgering BM, Wennestrom S, Claesson-Welsh L, Kozma SC. (1998). EMBO J., 17, 6649–6659. Heldin CH, Boss JL, Kosma SC and Thomas G. (1994). Templeton DJ. (2001). Curr. Biol., 8, 69–81. Nature, 371, 426–429. Thomas G. (2000). Nat. Cell Biol., 2, E71–E72. Montagne J, Steward MJ, Stocker H, Hafen E, Kozma SC and Volarevic S and Thomas G. (2001). Prog. Nucleic Acid Res. Thomas G. (1999). Science, 285, 2126–2129. Mol. Biol., 65, 101–127. Mukhopadhayay NK, Price DJ, Kyriakis JM, Pelech SL, Weng QP, Kozlowski M, Belham C, Zhang A, Comb MJ and Sanghera J and Avruch J. (1992). J. Biol. Chem., 267, 3325– Avruch J. (1998). J. Biol. Chem., 273, 16 621–16 629. 3335 Williams MR, Arthur JS, Balendran A, van der Kaay J, Poli Nakamura M, Zhou XZ, Kishi S, Kosugi I, Tsutsui Y and Lu V, Cohen P and Alessi DR. (2000). Curr. Biol., 10, 439–448. KP. (2001). Curr. Biol., 11, 1512–1516. Yaffe MB, Schutkowski M, Shen M, Zhou XZ, Stukenberg Panditax TK, Gregoire V, Dhingra K and Hittelman WN. PT, Rahfeld JU, Xu J, Kuang J, Kirschner MW, Fischer G, (1994). Cytogenet. Cell Genet., 67, 94–101. Cantley LC, Lu KP. (1997). Science, 278, 1957–1960. Papst PJ, Sugiyama H, Nagasawa M, Lucas JJ, Maller J and Zhou XZ, Lu PJ, Wulf Gand Lu KP. (1999). Cell Mol. Life Terada N. (1998). J. Biol. Chem., 273, 15 077–15 084. Sci., 56, 788–806.

Oncogene