Anti-tumor activities of selective HSP90α/β inhibitor, TAS-116, in combination with bortezomib in multiple myeloma.

Letter to the Editor

Heat shock protein 90 (HSP90) is a highly conserved molecular chaperone that interacts with various client proteins in eukaryotic cells[1]: Akt (PI3K/Akt pathway), IL-6R (JAK/STAT pathway), Bcr-Abl (RAS/ERK pathway), CDK4, 6, 9 (cell cycling), and IκB kinases (NF-κB pathway).[2] The expression of HSP90 is upregulated (2- to 10-fold) in tumor cells compared with normal cells, reflecting multiple oncogenic pathways and maintenance of homeostasis within tumor cells.[1, 2] Because HSP90 inhibition triggers downregulation/degradation of client proteins and triggers apoptosis, it is considered a promising target for novel targeted therapies. Indeed HSP90 inhibitors (e.g., geldanamycin analog 17-allylamino-17-demethoxy-geldanamycin (17-AAG), resorcinol derivatives, purine analogues) have shown early promising results in vitro and in vivo in solid tumors and some hematological malignancies, including multiple myeloma (MM).[3, 4] However, some clinical studies have been discontinued due to adverse effects including ocular toxicity.[3, 5] Therefore, development of a next-generation less-toxic HSP90 inhibitor remains an important therapeutic goal.

In the present study, we demonstrate in vitro and in vivo preclinical anti-MM activity of TAS-116, an oral selective HSP90α/β inhibitor, alone and in combination with BTZ. TAS-116 shows favorable oral bioavailability in rodent and non-rodent species, as well as good metabolic stability.[6] Importantly, TAS-116 demonstrates less ocular toxicity and greater anti-tumor activity in multiple xenograft models, compared to other HSP90 inhibitors at their MTD in rats.[6, 7] Our data therefore provide the preclinical framework for clinical evaluation of TAS-116, alone and with BTZ, to improve patient outcome in MM.

First we examined the growth inhibitory effect of TAS-116, a novel oral selective HSP90α/β inhibitor (Supplementary Figure S1A), in MM cell lines (Supplementary Figure S1B). TAS-116 significantly inhibited growth of these MM cell lines and patient MM cells (Supplementary Figure S1C), without affecting normal donor PBMNCs (Supplementary Figure S1D). Interestingly, we confirmed that TAS-116 was also active in N-Ras mutated cell lines (the proliferation/viability of NALM-6 is affected only at higher concentrations of 17-AAG) (Supplementary Figure S2A and S2B). We next examined the effect of TAS-116 on HSP90 client protein degradation. Significant degradation of HSP90 client proteins was triggered by TAS-116 in a dose-dependent manner in MM.1S cells (Supplementary Figure S1E).

We and others have shown that N-Ras mutation and HSP27 confers significant resistance to chemotherapies.[8, 9] Moreover, treatment with other HSP90 inhibitors induces resistance mechanisms due to the upregulation of other HSP proteins such as HSP27.[10] We therefore next examined whether TAS-116 can overcome 17-AAG-resistance associated with N-Ras mutation and upregulation of HSP27. Importantly, more significant degradation of phosho-C-Raf and phospho-MEK1/2, HSP90 client proteins and key RAS/RAF/MEK pathway regulators, was triggered by TAS-116 than 17-AAG in INA6 and NCI-H929 MM cells (Supplementary Figure S2D, 2E). In addition, HSP27 upregulation induced by TAS-116 was lower than by 17-AAG at equipotent doses (Supplementary Figure S2F). Taken together, these results indicate that TAS-116 induces cytotoxicity selectively and potently in MM cell lines and patient MM cells, even in NALM-6 cells, without toxicity in normal PBMNCs; potently targets HSP90 client proteins including C-Raf and MEK1/2; as well as inhibits upregulation of HSP27 and overcomes 17-AAG resistance mechanisms in MM cells.

We further confirmed that TAS-116 induces apoptosis in MM cells (Supplementary Figure S3A–F and Supplementary Information); inhibits Akt and ERK pathway, and overcomes the growth stimulatory effects triggered by cytokines and the bone marrow microenvironment (Supplementary Figure S4A–C, S5A–E, and Supplementary Information); and induces synergistic cytotoxicity with BTZ in vitro (Supplementary Figure S6A–D, Supplementary Table S1,2, and Supplementary Information).

We and others have previously shown that HSP90 inhibitors such as 17-AAG inhibit NF-κB signaling and induce terminal unfolded protein response (UPR).[11, 12] Whereas, BTZ induces both terminal UPR and canonical NF-κB pathway activation.[13, 14] We therefore hypothesized that TAS-116 could enhance the terminal UPR and inhibit canonical NF-κB pathway induced by BTZ, thereby augmenting BTZ-induced cytotoxicity. Although BTZ triggers activation of IκB kinase (IKKβ) and Akt, TAS-116 significantly downregulated IKKα/β in a time-dependent manner (Supplementary Figure S7A). Importantly, we observed that enhanced phosphorylation of Akt and key canonical NF-κB pathway regulators (p65, IκBα, and IKKα/β) triggered by BTZ in MM cell lines were significantly inhibited by TAS-116. Since Akt associates with IKK to induce IKK activation leading to NF-κB activation, these results indicate that TAS-116 significantly inhibits bortezomib-induced canonical NF-κB pathway.

We next evaluated the effect of this combination on UPR. TAS-116 markedly upregulated p-IRE1α, p-eIF2α, and CHOP, a transcription factor leading to apoptosis due to endoplasmic reticulum (ER) stress, at early time points (within 4 hours) (Supplementary Figure S7B). Importantly, TAS-116 in combination with BTZ enhanced phosphorylation of IRE1α and eIF2α in MM cell lines, indicating that BTZ-induced UPR was enhanced by TAS-116 (Supplementary Figure S6H, S7C). Moreover, cleavage of PARP was significantly enhanced by TAS-116 in combination with BTZ in both MM cell lines, associated with increased CHOP (Supplementary Figure S6I, S7D). These results suggest that TAS-116 augments BTZ-induced ER stress, followed by the terminal UPR and apoptosis.

We next examined the in vivo efficacy of TAS-116 in combination with BTZ using a murine xenograft model of human MM. Mice treated with TAS-116 (10 mg/kg and 15 mg/kg), BTZ, or TAS-116 plus BTZ showed significantly enhanced growth inhibition versus the vehicle control group (P = .004 and P = .0012, P = .003, and P < .0001, respectively; Figure 1A). Representative images of tumor growth inhibition by TAS-116 (10 mg/kg) are demonstrated in Figure 1B. The delay in tumor growth was greater in the combination-treated group compared with either monotherapy cohort (P = .0014 in TAS-116 vs the combination and P = .0001 in BTZ vs the combination; Figure 1A). Median overall survival of treated animals (TAS-116 10 mg/kg = 33 days, 15 mg/kg = 37 days, BTZ = 36 days, and the combination = 56.5 days) was significantly longer than vehicle control (29 days; P = .0064, P < .0001, P = .0009, and P < .0001, respectively; Figure 1C). The OS was significantly prolonged in the combination group compared with either monotherapy cohort (P = .004 in TAS-116 vs the combination, and P = .0004 in BTZ vs the combination; Figure 1C). These treatments were well tolerated, and no significant body weight loss was observed (Figure 1D). Importantly, immunohistochemical analysis of harvested human MM confirmed a significant increase in cleaved caspase-3- and TUNEL-positive cells in TAS-116 15 mg/kg-treated mice (Figure 1E). These results indicate that TAS-116 triggers enhanced in vivo anti-MM activities, both alone and in combination with BTZ, with a favorable safety profile.

Figure 1

TAS-116 inhibits human MM cell growth and enhances bortezomib-induced cytotoxicity in vivo

(A–G) SCID mice were injected subcutaneously with 5 × 106 MM.1S cells and treated with 10 mg/kg oral TAS-116 5 days a week (n = 10; green line); 15 mg/kg oral TAS-116 5 days a week (n = 10; blue line); 0.5 mg/kg subcutaneous BTZ twice a week (n = 8; purple line); or 0.5 mg/kg subcutaneous BTZ twice a week and 10 mg/kg oral TAS-116 5 days a week (n = 10; red line) for 28 days. A vehicle control group received oral vehicle only and subcutaneous saline (n = 9; black line).

(A) Tumor volume was calculated from caliper measurements every other day, and data represent mean ± SD.

(B) Representative whole-body images from a mouse treated for 29 days with control vehicle (bottom panel) or for 31 days with TAS-116 (10 mg/kg; top panel).

(C) Survival was evaluated from the first day of treatment using Kaplan-Meier curves.

(D) Change of body weight was expressed from the first day of treatment. Data represent mean ± SD.

(E) Tumors harvested from TAS-116- (15 mg/kg) and vehicle control- treated mice after 3 days of treatment were subjected to immunohistochemical analysis for cleaved caspase-3 and TUNEL staining.

Finally, we investigated the ocular toxicity profile of TAS-116, since ocular toxicity is one of the most notable toxicities limiting the clinical application of other HSP 90 inhibitors. Because others have shown that geldanamycin and its analogue 17-AAG inhibit proliferation of normal human retinal pigment epithelial ARPE-19 cells essential for the support of photoreceptors by inducing cell cycle arrest and apoptosis,[15] we first examined the growth inhibitory effect of TAS-116 and 17-AAG in ARPE-19 cell line (Figure 2A). Importantly, TAS-116 was less toxic to ARPE-19 cells than 17-AAG. In addition, we assessed the cytotoxicity of TAS-116 or HSP90 inhibitor PF-04928473 (SNX-2112) in combination with BTZ in ARPE-19 cells. Surprisingly, low-dose BTZ significantly ameliorated the cytotoxicity induced by TAS-116, compared with PF-04928473 (SNX-2112) (Figure 2B). We next investigated the ocular toxicity profile of TAS-116 using an in vivo murine xenograft model. Importantly, TAS-116 15 mg/kg–treatment (maximum tolerated dose) did not induce ocular toxicity in mice, in contrast to PF-04929113 (SNX-5422) 40 mg/kg, which was associated with increased photoreceptor cell death in all retinal layers (Figure 2C). Taken together, these results indicate that TAS-116 demonstrates a safer ocular toxicity profile than PF-04929113 (SNX-5422).

Figure 2

TAS-116 is less toxic to human retinal pigment epithelial cells than other HSP90 inhibitors, and does not trigger ocular toxicity in mice

(A) Human retinal pigment epithelial ARPE-19 cell lines and NCI-H929 MM cells were cultured with TAS-116 or 17-AAG (0–5 μM) for 48 hours. Cell viability was assessed by CellTiter-Glo® assay of triplicate cultures, expressed as percentage of untreated control. Data represent mean ± SD.

(B) ARPE-19 cells were cultured for 48 hours with BTZ (0–2 nM) in combination with TAS-116 (0 μM: gray, 0.125 μM: gold, 0.25 μM: light orange, 0.5 μM: orange) or PF-04928473 (SNX-2112) (0 μM: gray, 0.125 μM: light green, 0.25 μM: sea green, 0.5 μM: green). Cell proliferation was assessed by CellTiter-Glo® assay of triplicate cultures, expressed as percentage of untreated control. Data represent mean ± SD. (* P < .01; ** P < .001)

(C) TAS-116 (15 mg/kg; 5 days a week), PF-04929113 (SNX-5422) (40 mg/kg; 3 times per week), or vehicle were administered orally in SCID mice for two weeks. Retinal morphology and photoreceptor cell death were evaluated by TUNEL staining. ONL indicates outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer.

In summary, we show the anti-MM activities of a novel HSP90α/β-selective inhibitor TAS-116 in MM. TAS-116 triggered cytotoxicity via apoptosis, associated with downregulation of phosphorylation of client proteins including Akt and ERK. TAS-116 overcame the anti-apoptotic effect triggered by IL-6, IGF-1, or bone marrow stromal cells. TAS-116 also induced cytotoxicity even in acute B cell leukemia cell line NALM6, as well as N-Ras mutated NCI-H929 and INA6 MM cell lines, and overcame 17-AAG resistance mechanisms. TAS-116 also enhanced bortezomib (BTZ)-induced MM cytotoxicity, due to inhibition of BTZ-triggered canonical NF-κB activation and enhancing endoplasmic reticulum stress. We confirmed that TAS-116, alone and in combination with BTZ, was well tolerated; triggered significant tumor growth inhibition; and prolonged host survival in a murine xenograft model of human MM. Importantly, TAS-116 showed lower ocular toxicity, a known toxicity of HSP90 inhibitors, than PF-04929113 (SNX-5422). Taken together, our studies show that TAS-116 blocks MM cell growth both in vitro and in vivo, and is well tolerated, providing the framework for its clinical evaluation to improve patient outcome in MM.