MTI-101 treatment inducing activation of Stim1 and TRPC1 expression is a determinant of response in multiple myeloma

The emergence of drug resistance continues to be a major hurdle towards improving patient outcomes for the treatment of Multiple Myeloma. MTI-101 is a first-in-class peptidomimetic that binds a CD44/ITGA4 containing complex and triggers necrotic cell death in multiple myeloma cell lines. In this report, we show that acquisition of resistance to MTI-101 correlates with changes in expression of genes predicted to attenuate Ca2+ flux. Consistent with the acquired resistant genotype, MTI-101 treatment induces a rapid and robust increase in intracellular Ca2+ levels in the parental cells; a finding that was attenuated in the acquired drug resistant cell line. Mechanistically, we show that pharmacological inhibition of store operated channels or reduction in the expression of a component of the store operated Ca2+ channel, TRPC1 blocks MTI-101 induced cell death. Importantly, MTI-101 is more potent in specimens obtained from relapsed myeloma patients, suggesting that relapse may occur at a cost for increased sensitivity to Ca2+ overload mediated cell death. Finally, we demonstrate that MTI-101 is synergistic when combined with bortezomib, using both myeloma cell lines and primary myeloma patient specimens. Together, these data continue to support the development of this novel class of compounds for the treatment of relapsed myeloma.


Treatment with MTI-101 or HYD1 Increases Intracellular Ca 2+ Levels in MM Cells.
To determine the mechanism by which HYD1 and its cyclic analogue MTI-101 induces cell death in NCI-H929 cells, we developed the HYD1-resistant isogenic cell line H929-60 15,16 . As shown in Fig. 1A the IC50 value for H929 is 1.2 +/− 1.15 uM while for, H929-60 cells the IC50 value was 9.3 +/− 1.08 uM towards MTI-101 induced growth arrest as measured by MTT assays (n = 3 independent experiments p < 0.05, t-test). Gene expression profiling of the parental and drug resistant variant was used to determine pathways that were altered as cells gained resistance to MTI-101. Specifically, the expression of PLC-β, the IP3 isoform 3 receptor, TRPC1, and TRPM7 were all reduced in H929-60 cells (Table 1). Moreover ATP2A3 the SERCA channel was also downregulated in the drug resistant line (Table 1). Interestingly, recent studies indicate that TRPC1 and Orai1 expression regulate STIM1dependent Ca 2+ entry 22,23 . Moreover, TRPM7 is a substrate of MLKL (mixed-lineage kinase domain-like protein), which is phosphorylated by Ripk3 following treatment with TNFα, linking Ca 2+ entry with the necroptotic cell death pathway 24 . As shown in Fig. 1B a subset of the genes identified by microarray analysis were validated by real-time RT-PCR (n = 3 independent experiments performed in triplicates p < 0.05, Student's t-test) and western blot analysis, respectively (n = 3 independent experiments). Interestingly, TRPM6 and TRPM8 were increased in the resistant cell line and may contribute to maintaining Ca 2+ homeostasis in the resistant cell line despite reduced expression of the store Ca 2+ operated channel pathway.
To test if HYD1 or MTI-101 treatment affected intracellular Ca 2+ levels, NCI-H929 and H929-60 cells were loaded with fura-2 dye, and time-lapse images were taken every ten seconds for either 60 min (HYD1) or for 30 min (MTI-101) after treatment. Treatment with both HYD1 (75 µM) and MTI-101 (3 µM) triggered increases in intracellular Ca 2+ levels in NCI-H929 and H929-60 cells. Furthermore, the levels of intracellular Ca 2+ pools that were induced by MTI-101 were reduced in H929-60 vs. NCI-H929 cells (Fig. 1D), suggesting that HYD1and MTI-101-induced MM death was due to Ca 2+ overload. NCI-H929 cells were thus examined to assess if there were differences in total Ca 2+ levels or differences in temporal Ca 2+ levels after HYD1 or MTI-101 treatment. MTI-101 is more efficient in eliciting a Ca 2+ response in MM cells than HYD1, consistent with the increased potency of MTI-101 vs. its linear counterpart on inducing cell death 16 . Collectively, these findings suggest that acquisition of resistance to this class of compounds is associated with changes in genes predicted to rewire Ca 2+ homeostasis. 2+ are Caused by Release of Ca 2+ From ER Stores and Ca 2+ Entry via Store-operated Channels. Increase in intracellular Ca 2+ occur either by influx through the plasma membrane channels or exchangers (i.e., voltage-gated channels, SOC, Na + -Ca 2+ exchanger) or via release from internal stores such as the ER [25][26][27] . One means of Ca 2+ release from ER stores is through the opening of the IP3R caused by PLC activation 25 . To test if activation of PLC was involved in MTI-101-induced intracellular Ca 2+ increases, NCI-H929 or U266 cells were pre-treated with 2 µM or 1 µM of the PLC inhibitor U73122 28 , respectively, and were then treated with MTI-101. This assay was performed in the presence and absence of extracellular Ca 2+ to determine the effects of MTI-101 on Ca 2+ influx to cells. After treatment with MTI-101, the effects on Ca 2+ levels were measured using Fura-2. U73122 treatment decreased MTI-101-induced Ca 2+ levels in both NCI-H929 ( Fig. 2A) and U266 cells (data not shown); a finding that was further attenuated upon removal of extracellular Ca 2+ indicating that increased levels of Ca 2+ induced by MTI-101 occur via both release of ER stores and Ca 2+ entry via store-operated channels. Although, we could not detect a decrease in peak levels of Ca 2+ in the absence of the PLC inhibitor (Fig. 2B,C), removing extracellular Ca +2 reduced the MTI-101-induced sustained increases in Ca 2+ levels (as measured by the area under the curve) in both H929 and U266 cells (Fig. 2D,E). Together these data indicate that Ca 2+ release from the ER, via activation of the IP3R, contributes to the peak levels and Ca 2+ entry through plasma membrane channels contribute to the sustained high levels of Ca 2+ observed following treatment with MTI-101.

MTI-101-induced Increases in Intracellular Ca
These findings suggest that blocking Ca 2+ release from ER stores and/or Ca 2+ entry would inhibit MTI-101-induced cell death. Thus, we tested whether 2-APB, which blocks Ca 2+ release through the IP3R as well as SOC-mediated Ca 2+ entry 29 , affected MTI-101-induced myeloma cell death. Notably, both NCI-H929 and U266 cells pretreated with 2-APB are resistant to MTI-101-induced cell death (Fig. 3A,B). To determine whether Ca 2+ entry into mitochondria contributes to death we blocked the mitochondria uniporter with RU360. As shown in Fig. 3A,B pretreatment with RU360 blocks MTI-101 induced cell death (p < 0.05, students T-test). These data  suggest that MTI-101-induced cell death is, at least in part, due to Ca 2+ overload of mitochondria. As shown in Fig. 3C,D, cells pre-treated with 2-APB and then treated with MTI-101 inhibits MTI-101-mediated increases in intracellular Ca 2+ (Fig. 3C,D). Thus, blocking increases in intracellular Ca 2+ is sufficient to inhibit MTI-101 induced cell death.
To test whether MTI-101 induces activation of STIM1 as determined by the formation of punctae and trafficking to ER/PM junctions, we used TIRF microscopy, which images 100 nm proximity of the surface of the slide. As shown in Fig. 4A,B, U266 cells which overexpress STIM1 m-cherry have significantly increased levels of Stim1 as determined by total fluorescence intensity near the plasma membrane (p < 0.05, Two-way ANOVA). To test the specific role of TRPC1, a component of the store-operated channels we used shRNA to reduce the expression of TRPC1. As shown in Fig. 4C,D, reducing the expression of TRPC1 significantly inhibits MTI-101 induced cell death (p < 0.05, student's t-test).

MTI-101/Bortezomib Combination demonstrates Increased Anti-Myeloma Activity in vitro and
In Vivo. Several reports indicated that blocking L-type Ca 2+ channels or blocking Ca 2+ uptake by inhibiting the mitochondria uniporter blocks cell death induced by bortezomib. In contrast, 2-APB was not shown to block bortezomib-induced cell death. It is attractive to speculate that increased Ca 2+ via L-type, and store-operated channels converge to cause synergistic cell death via Ca 2+ overload of the mitochondria in myeloma cells 30,31 . Thus we next sought to determine the activity of MTI-101 in combination with the standard of care agent bortezomib in 5TGM1 cells (Fig. 5A). For 5TGM1 cells we utilized a bone marrow stroma co-culture model and time lapse imaging of individual cells to determine viability as function of time. Combination index was calculated at select time points using the Calcusyn software 32,33 . As shown in Fig. 5A, in 5TGM1 cells when MTI1-101 is combined with bortezomib synergy is observed, as defined by a combination index of less than 1 cells as early as 30 hours and was maintained to the endpoint of 72 hours. To test whether MTI-101 in combination with Bortezomib has anti-tumor activity in vivo C57BL/KaLwRijHsd mice (6-to 8-week old) were injected i.v. with syngeneic 5TGM1 MM cells and were then treated with MTI-101 (10 mg/kg) and/or bortezomib (0.5 mg/kg). As predicted, mice treated with MTI-101 or bortezomib had improved survival versus vehicle treated mice (Fig. 5B). Notably, the MTI-101/bortezomib combination had superior anti-myeloma activity versus single-agent treatment. Specifically, mean survival times for vehicle control, bortezomib only, MTI-101 only, and MTI-101/bortezomib combination therapy were 36, 42, 45.5, and 72.5 days, respectively (p < 0.05, Log-Rank Test) (Fig. 5B). To assess tumor burden in transplanted mice, IgG2B serum levels were determined (Fig. 5C). Interestingly we could not detect decreases in IgG2B levels with the combination at 4 weeks, which may reflect decreased fitness for cells that survive the combination treatment. We had previously reported that selection for resistance to MTI-101 resulted in a compromised CAM-DR phenotype and decreased adhesion to fibronectin and stroma cells 15 . Sensitivity to MTI-101 is augmented in Relapsed vs. Primary Multiple Myeloma. To test the sensitivity of primary vs. relapsed myeloma specimens to MTI-101, CD138-positive malignant plasma cell fractions from thirteen newly diagnosed and twelve relapsed primary MM specimens were treated with MTI-101. Consistent with our previous studies using the HYD1 linear analog of MTI-101 15 , MTI-101 was significantly more potent (p < 0.05, Student's t-test) in relapsed than newly diagnosed MM specimens (Fig. 6A).
As 2-APB pretreatment inhibits MTI-101-induced death of MM cell lines, we tested whether this Ca 2+ blocker would also impair MTI-101-induced cell death of primary MM specimens. Indeed, patient samples pretreated with 50 µM 2-APB were also more resistant to MTI-101 treatment ( Fig. 6B p < 0.05, Student's t-test). Thus, similar to MM cell lines, primary MM cells treated with MTI-101 die, at least in part, due to Ca 2+ overload. Finally, as shown in Fig. 6C, the synergy between MTI-101 and bortezomib was observed in 5 out of the 6 primary patients, with pt135 showing synergy at 30 but not 60 hrs. Together these data indicate that combining MTI-101 with bortezomib maybe a good strategy for the treatment of multiple myeloma.

Discussion
MM is a disease that initially responds to chemotherapy. However, despite the range of mechanistically distinct therapies, the disease will eventually relapse and become refractory to further treatment, underscoring the need for novel treatment strategies. MM has high basal levels of ER stress due to the load of IgG production and is thus prone to initiate an unfolded protein response 34 . This increased ER load appears to contribute to the sensitivity of MM towards treatment with proteasome inhibitors. Our data support the notion that MM cells also rewire their Ca 2+ circuitry, and that targeting Ca 2+ signaling is an attractive therapeutic strategy for drug-resistant MM. Indeed, this may be a general feature of cancer cells, as demonstrated by increased expression of SOC and STIM1 in multiple tumor types 3,4 and the required roles for elevated Ca 2+ pools in cancer cell migration, invasion and metastasis 5,7 . Our findings suggest that remodeling the Ca 2+ circuitry occurs at a cost to fitness, rendering cells vulnerable to death induced by Ca 2+ overload.
Our findings indicate that the acquisition of resistance to HYD1 and MTI-101 is associated with remodeling the circuitry that controls Ca 2+ homeostasis. Specifically, drug resistance correlated with the suppression of IP3R, PLC, TRPC1, and TRPM7 expression and increased expression of TRPM6 and TRPM8 (Table 1). This switch in expression might reflect a compensatory response allowing for similar basal levels of Ca 2+ to support growth in the face of MTI-101 treatment. Moreover, in this study, we showed that MTI-101 treatment as a single agent induces increased intracellular Ca 2+ levels; a finding that contributes to MTI-101 induced cell death in MM cells. Evidence to support the role of Ca 2+ release from the ER included data showing that the PLC inhibitor partially blocks MTI-101 induced Ca 2+ levels. In addition, we showed that the SOC channel and IP3 receptor inhibitor 2-APB attenuates MTI-101 mediated cell death in MM cell lines as well as primary patient derived CD138 plasma cells. Finally, we showed that MTI-101 induces STIM1 punctae near the plasma membrane, an event required for activation of store operated channels. Selection for resistance towards MTI-101 correlated with reduced expression of TRPC1 a component of the store operated channel. This finding guided our focus on the role of TRPC1. Indeed, we showed that a reduction TRPC1 partially inhibits MTI-101 induced MM cell death. Together, our results suggest that MTI-101 mediated increase in Ca 2 is caused by release of Ca 2+ from ER stores and Ca 2+ entry via store-operated channel and that TRPC1 expression is a determinant of response in MM.
Treatment of cancer is typically most effective when combination strategies are implemented. Thus, it is important to place novel treatment strategies in context with standard of care agents. In this study, we showed that treatment with MTI-101 in combination with the standard of care agent Bortezomib demonstrates increased anti-myeloma activity in vitro and in vivo. Importantly, the synergy was maintained using CD138 cells derived from patients and cultured in an organotypic model which contained bone marrow stroma cell and collagen. These data indicate that MTI-101 in combination with the proteasome inhibitors is a tractable strategy for the development of this class of inhibitors. Mechanism of action of bortezomib includes inhibition of NF-κB activity and the induction of the unfolded protein response (UPR) 34,35 . Accordingly, alterations in NF-κB pathways, the UPR and the induction of autophagy have been implicated in mediating resistance to bortezomib 36,37 . Further, acquired resistance to bortezomib in cell line cultures has been linked to the identification of missense mutations (Ala49Thr) in the β5-subunit (PSMB5) of the proteasome and to increased expression of PSMB5 38,39 . However, sequencing efforts have failed to identify mutations in PSMB5 in clinical resistance to bortezomib 40 . Several Ft/F0 is the pixel intensity at the indicated time as denoted on the X axis divided by the average pixel density prior to drug treatment. Mean intensity data for STIM1 mCherry at each time point were extracted for comparisons. A representative cell from the control group and treatment group is shown before treatment (0 min) and after treatment (5 and 10 min). Shown is the mean and standard error of 7 cells for a representative experiment (p < 0.05, Two Way ANOVA). The experiment was repeated 3 independent times and similar data was obtained. (B and C). TRPC1 expression was reduced by using retroviral TRPC1 shRNA construct and expression was determined by Western Blot in U266 and MM1.S cells respectively. (B and C See Supplemntal Data S1). MM1.S and U266 cells were treated with different concentrations (5, 10 and 20 μM) of MTI-101 for 48 hr and cell death was measured using PI staining. Cells with reduced TRPC1 levels cells showed a significant reduction in MTI-101 induced cell death (p < 0.05, Student's t-test).
reports indicated that blocking L-type Ca 2+ channels or blocking Ca 2+ uptake by inhibiting the mitochondria uniporter blocks cell death induced by bortezomib 30,31 . It is attractive to speculate that synergistic cell death between MTI-101 and Bortezomib is driven via Ca 2+ overload of the mitochondria in myeloma cells. However, recent studies indicate that inhibition of the proteasome can lead to necroptosis in a RIPK3 dependent pathway 41 . Thus, it is also feasible that the observed synergy between MTI-101 and bortezomib is due to augmentation of activation of RIPK3 leading to increased phosphorylation of MLKL and necroptosis. We currently favor the hypothesis that that as primary samples become resistant to bortezomib, myeloma cells may rewire Ca 2+ homeostasis to favor the induction of Ca 2+ flux via store-operated channels rather than L-type channels, which underlies increased sensitivity to MTI-101 induced cell death in specimens obtained from patients relapsing on therapy. More studies are required to fully understand the mechanistic underpinning of synergy when MTI-101 is combined with bortezomib and the increased activity of MTI-101 in relapsed patient specimens. In summary, MTI-101 remains an attractive novel class of compounds to test as a front line combination strategy with proteasome inhibitors or in the setting of proteasome inhibitor refractory disease.

Cells and Reagents. NCI-H929 and U266 multiple myeloma (MM) cells were obtained from American
Type Culture Collection (Manassas, VA) and maintained in RPMI-1640 medium supplemented with 10% FBS and 1% penicillin/streptomycin/glutamine. NCI-H929 cells were also supplemented with 0.05 mM beta-mercaptoethanol. The HYD1 drug-resistant NCI-H929 cell line H929-60 was developed as previously described 15 . The parental and drug resistant line was validated by short tandem repeat (STR) analysis. 5TGM1 myeloma cells were derived from murine myeloma 5T33 and kept in similar media as U266 cells. All cell lines were tested for mycoplasma every six months. HYD1 (kikmviswkg) and MTI-101 were synthesized by Bachem (San Diego, CA). Bortezomib was purchased from Selleck Chem. The compounds U73122 (Sigma), 2-APB (Sigma) and Ru360 (Merck Millipore) were all used to inhibit Ca 2+ signaling. Gene Expression Profiling. Gene expression profiling was performed as previously described 42 . Briefly, 100 ng of total RNA was isolated using a RNeasy mini kit (Qiagen). The oligonucleotide probe arrays used were the Human Genome U133 plus 2.0 Arrays. Data were processed and normalized using IRON. The mean expression values of the three experiments were compared between NCI-H929 and H929-60 MM and fold change were compared between mean expression in the two groups.
Real-Time PCR Analysis. 1 μg total RNA was used as a template to synthesize cDNA with the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA; Cat. no. 4368814) with a reaction volume of 50 μL. Real-time PCR was performed on the ABI 7900HT Fast Real-Time PCR System (Applied Biosystems) using assays specific for each gene of interest. Each reaction well contained 5μL of Power SYBR Green PCR Master Mix (cDNA equivalent to 20 ng of total RNA and 400 nM each of forward and reverse amplification primers in a reaction volume of 10 μL. Cycling conditions were as follows: 95 °C for 10 minutes for polymerase activation, followed by 40 cycles of 95 °C for 15 seconds and 60 °C for 1 minute. Data analysis was performed using Sequence Detection System software from Applied Biosystems, version 2.4. The experimental Ct (cycle threshold) was calibrated against the endogenous control product GAPDH. Real Time RTPCR was performed by ARQ genetics LLC Bastrop, TX. The following primer sets were used for amplification.
ATP2A3 For ward primer CCGGAACCACATGCACGAAG; Reverse primer GGGGATG GCCATTCTGACCTC Figure 6. Relapsed myelomas are more sensitive to MTI-101. Specimens were separated into two groups depending on clinical diagnosis; either newly diagnosed or relapsed patients. (A) CD138+ cells were treated with 10 µM of MTI-101 for 24 hr. After 24 hr cell death was measured by Annexin V staining and FACS analysis. CD138 cells derived from patients which have relapsed on therapy were significantly more sensitive to MTI-101 induced cell death compared to CD138 cells obtained from newly diagnosed patients (p < 0.05, Student's t-test) (B) CD138+ cells were pretreated with 50 µM 2APB or VC (vehicle control) for 30 min and then were treated with 10 µM MTI-101 for 24 hrs followed by FACS analysis. (C) Dose response surfaces for primary MM cells from patients treated ex vivo with bortezomib (50-0.6 nM, 3 fold serial dilution), MTI-101 (20-0.25 uM, 3 fold serial dilultions)) and combination. Exposure time was 96 hrs except for patient pt129 (60 hrs). The actual combination of MTI-101 and bortezomib (MTI-101 + BTZ) is more effective (higher kill) than theoretical additive effect (ADDITIVE) indicating synergy between two drugs. Combination indexes (CI) were calculated using CalcuSyn software at 30 and 60 hrs. Samples were performed in duplicate and independent combination indexes were calculated for each patient specimen. PLCB1 (span Exon 4-5) Forward primer; GTGTCCGACAGCCTCAAGAA; Reverse primer ATCCCTGAGG GTCAGTCCTC TRPC1 Forward primer CGGTTGTCAGAAACTAATGGAACG; Reverse primer CCTGAATTCCAC CTCCACAAGA TRPM7 Forward primer CTGCTTTTGATCTCCTGTCCTGT; Reverse primer CAGACAGCC CATATTGCCCT ITPR3 Forward primer GTTCCTGACGTGTGACGAGT Reverse primer GCCACCAGGCAGTACTTGAT HGAPDH-forward primer TGCCCTCAACGACCACTTTG; reverse primer CTCTTCCTCTTGT GCTCTTGCTG TIRF imaging of STIM1 mCherry. Imaging of STIM1 mCherry overexpressing U266 cells was performed using a total internal reflection fluorescence (TIRF) imaging system with objective lens 60x/1.49 Apo TIRF DIC in oil immersion using a Nikon Eclipse TE2000-E microscope (Melville, NY). Data were collected with a CoolSNAP HQ monochrome CCD camera (Tucson, AZ) and analyzed by NIS-Elements Software (Nikon Instruments, Melville, NY). mCherry-tagged STIM1 was excited with a 561-nm TIRF laser respectively, and emissions were collected using a 300-ms exposure time. TIRF images were collected every 15 seconds over 30 minutes with the experimental group receiving a single 2.  Stromal cells. Primary MM cells were co-cultured with bone marrow derived stromal cells obtained from patients' BM aspirates, as previously described 44 . Since this process takes weeks, primary MM cells from fresh biopsies were co-cultured with established stroma from prior patient samples.
Ex vivo assay. The ex vivo assay used to quantify chemo sensitivity of primary MM cells was described in detail previously 45 . Briefly, MM cells (CD138+) were seeded in multi-well plates with previously established human-derived stroma and collagen-I to a total volume of 8 µL containing approximately 4,000 MM cells and 1,000 stromal cells. Drugs were added using a robotic plate handler, so that every drug was tested at five concentrations (1:3 serial dilution) and two replicates. Negative controls (supplemented growth media with and without vehicle control, DMSO) were included, as well as positive controls for each drug (cell line MM1.S at highest drug concentration). Plates were placed in a motorized stage microscope (EVOS Auto FL, Life Technologies) equipped with an incubator and maintained at 5% CO 2 and 37 °C. Each well was imaged every 30 minutes for a total duration of four days.
Digital image analysis algorithm. We have developed a digital image analysis algorithm previously described 45,46 to determine changes in viability of each well longitudinally across the 96 h interval.

Drug combination and synergy analysis.
To study interactions between MTI-101 and bortezomib in patient samples, we have added two drugs diluted serially, so that their concentration ratio remained the same across all wells tested. Thus the concentrations tested were (BTZ/MTI-101): 50 nM/20 µM, 16.7 nM/6.67 µM, 5.6 nM/2.22 µM, 1.9 nM/0.74 µM and 0.6 nM/0.25 µM. To determine if there is synergy between the two drugs, we have plotted the dose-response surface of each drug individually, their actual combination, and the theoretical additive curve (Fig. 6). Here we use the classic definition of synergy, which determines that two drugs are synergistic if their combined effect (percent cells killed) is higher than the combination of their independent effects 47 . For instance, consider that drugs A and B, at a specific concentration and during a certain period of time, kill the fractions a and b of cells, where a < 1 and b < 1. Should the combination of both drugs kill a fraction equal to [1 − (1 − a) * (1 − b)], then A and B are additive. If the fraction killed is higher, the drugs are synergistic; if lower, they are antagonistic.
Murine 5TGM1 Myeloma Model. Animal studies were conducted using 6-to 8-week-old female C57BL/ KaLwRijHsd mice (Harlan) in accordance with the NIH Guide for the Care and Use of Laboratory Animals and protocols were approved by the University of South Florida IACUC committee. 5TGM1 MM cells were injected i.v. via tail vein. Establishment of MM tumors in inoculated mice was followed by assaying immunoglobulin G2b (IgG2b) monoclonal paraprotein in sera prepared from whole blood obtained by a submandibular bleed. Mouse IgG2b levels were assayed by an ELISA (Bethyl Laboratories, Montgomery, TX) on days 7, 14, 21, and 28 per manufacturer's instructions. On day 10 after inoculation, mice began treatment with 10 mg/kg MTI-101 or 0.5 mg/kg bortezomib 3 times/week (Monday, Wednesday, Friday) for 21 days, for nine treatments. All drugs were administered by IP injections. Mice were then examined until euthanization, which occurred when mice displayed hind leg paralysis or tumors grew in excess of 2 cm in diameter. On day 100, all remaining mice were euthanized.