The IRE-1 kinase/RNase splices the mRNA of the XBP-1 gene, resulting in the spliced XBP-1 (XBP-1s) mRNA that encodes the functional XBP-1s transcription factor that is critically important for the growth and survival of B-cell leukemia, lymphoma, and multiple myeloma (MM). Several inhibitors targeting the expression of XBP-1s have been reported; however, the cytotoxicity exerted by each inhibitor against cancer cells is highly variable. To design better therapeutic strategies for B-cell cancer, we systematically compared the ability of these compounds to inhibit the RNase activity of IRE-1 in vitro and to suppress the expression of XBP-1s in mouse and human MM cell lines. Tricyclic chromenone-based inhibitors B-I09 and D-F07, prodrugs harboring an aldehyde-masking group, emerged as the most reliable inhibitors for potent suppression of XBP-1s expression in MM cells. The cytotoxicity of B-I09 and D-F07 against MM as well as chronic lymphocytic leukemia and mantle cell lymphoma could be further enhanced by combination with inhibitors of the PI3K/AKT pathway. Because chemical modifications of the salicylaldehyde hydroxy group could be used to tune 1,3-dioxane prodrug stability, we installed reactive oxygen species-sensitive structural cage groups onto these inhibitors to achieve stimuli-responsive activities and improve tumor-targeting efficiency.

Inositol-requiring enzyme 1 (IRE-1) is an endoplasmic reticulum (ER) stress sensing protein that contains an ER stress sensor domain in the lumen of the ER, a kinase/ribonuclease (RNase) domain in the cytoplasm, and a transmembrane segment linking the two domains (1–3). ER stress as a result of pharmacologic or pathologic insults can cause autophosphorylation, dimerization, and oligomerization of IRE-1, allowing the assembly of a functional RNase to splice the mRNA of the X-box binding protein 1 (XBP-1; refs. 4–6). The expression or increased production of the spliced XBP-1 (XBP-1s) transcription factor leads to increased expression levels of specific chaperones and lipids, bringing stressed cells back to the homeostatic state (7, 8).

Genetic deletion of IRE-1 or XBP-1 from B lymphocytes compromises plasma cell differentiation and antibody production in mice (9–11). XBP-1-deficient plasma cells not only produce significantly reduced ER content (12), but also increased levels of IRE-1 (13). Such upregulated IRE-1, in the absence of XBP-1s, is phosphorylated at serine 729 within its kinase activation loop and capable of rapidly cleaving the mRNAs of immunoglobulins via regulated IRE-1–dependent decay (RIDD), further explaining the drastically reduced production of antibodies in XBP-1–deficient plasma cells (14, 15).

The IRE-1/XBP-1 pathway has been implicated in the development of multiple myeloma (MM), a plasma cell–derived cancer. Indeed, Eμ-XBP-1s transgenic mice, in which the overexpression of mouse XBP-1s in B cells is driven by an immunoglobulin μ heavy chain promoter/enhancer, develop monoclonal gammopathy of undetermined significance or MM phenotypes (16). In addition, the expression of XBP-1s supports the growth and survival of mouse and human chronic lymphocytic leukemia (CLL; ref. 17). Eμ-TCL1 transgenic mice express human TCL1 protein in B cells, and develop a disease-resembling human CLL (18). Genetic deletion of XBP-1s from CLL cells in Eμ-TCL1 transgenic mice delayed malignant progression of CLL (19). These data support the potential use of inhibitors targeting the expression of XBP-1s for the treatment of MM and CLL.

We have developed a novel class of fluorescent tricyclic chromenone salicylaldehyde-based inhibitors that can bind to the RNase domain of IRE-1 and potently suppress its RNase activity in both splicing XBP-1 mRNA and cleaving RIDD substrates (15, 19–21). These inhibitors induce cytotoxicity in MM, CLL, mantle cell lymphoma (MCL), Burkitt lymphoma, and neuroblastoma (19–23). Sunitinib (24, 25) and staurosporine (25, 26) can bind directly to the ATP-binding site of IRE-1, resulting in a DFG-in αC helix active conformation and triggering dimerization and activation of IRE-1. Although staurosporine induces splicing of the XBP-1 mRNA in RPMI-8226 MM cells (26), sunitinib inhibits the mRNA and protein levels of XBP-1s in NCI-H929 and U266 MM cell lines (25). However, both sunitinib and staurosporine are cytotoxic to MM cells (27–30). KIRA6 (31), AMG-18 (a.k.a. KIRA8; refs. 32, 33), and GSK2850163 (26) can bind to an allosteric site on the kinase domain of IRE-1, shifting the structure of the kinase domain to the DFG-out, αC helix inactive conformation and resulting in significantly reduced RNase activity. Although KIRA6 can inhibit mast cell leukemia (34), it is also a potent inhibitor of the receptor tyrosine kinase KIT (35). Although AMG-18 has been shown to impose no cytotoxicity in more than 300 cancer cell lines, including RPMI-8226 and other MM cell lines (32), it has been recently shown to be cytotoxic to RPMI-8226 cells and clinical MM samples (36). The cytotoxicity of GSK2850163 against cancer cells has not been reported. In addition, an acridine derivative, 3,6-DMAD, disrupts oligomerization of IRE-1, inhibits its RNase activity, and is cytotoxic to human MM cells (37). Toyocamycin, an antibiotic, can also inhibit the RNase activity of IRE-1 and kill MM cells (38). Given the variable effects of these inhibitors in biochemical and cellular assays, there is a need to systematically compare their ability to inhibit the RNase activity of IRE-1 and induce cytotoxicity in MM cells.

In this article, we systematically compare the ability of these small molecules to suppress IRE-1–mediated cleavage of the fluorescently tagged XBP-1 mRNA substrate in a fluorescence resonance energy transfer (FRET) suppression assay in vitro and to inhibit the expression of XBP-1s proteins in whole cells. Based on these assays, we select inhibitors that specifically and potently suppress the expression of XBP-1s, and evaluate their cytotoxicity, as a single agent or in combination with PI3K/AKT pathway inhibitors, against MM, CLL, and MCL cells. Because tumor cells express endogenously higher levels of hydrogen peroxide (H2O2) than normal cells, we chemically install reactive oxygen species (ROS)-sensitive structural cages to selected inhibitors to further improve their tumor specificity.

General methods

UV-vis and fluorescence spectra were carried out on a BioTek Synergy NEO2 instrument.

Antibodies, reagents, and chemical compounds

Antibodies against XBP-1s (Cell Signaling Technology or CST), IRE1 (CST), phospho-AKT (S473, CST), AKT (CST), phospho-Bruton's tyrosine kinase (BTK) (Y223, CST), BTK (CST), PARP (CST), cleaved PARP (CST), phospho-tyrosine (4G10; Millipore Sigma), and p97 (Fitzgerald) were obtained commercially. LPS, catalase, and N-acetyl cysteine (NAC) were purchased from Sigma-Aldrich. Subtilase cytotoxin (SubAB) was a gift from Dr. Adrienne W. Paton and Dr. James C. Paton at the University of Adelaide, Adelaide, Australia. We purchased from reliable commercial sources 3,6-DMAD (Millipore Sigma), KIRA6 (Millipore Sigma), GSK2850163 (Millipore Sigma), sunitinib (Millipore Sigma), staurosporine (Millipore Sigma), toyocamycin (Millipore Sigma), MK2206 (Cayman), CAL-101 (Cayman), and Auranofin (AdipoGen). We developed and chemically synthesized C-B06 (19), B-I09 (19), D-F07 (21), E-D08, E-G04, E-E08, and E-F02 for our studies. Experimental procedures and NMR spectra for all newly synthesized compounds can be found in the supplementary methods and figures. We followed the published method to synthesize AMG-18 (32).

Cell culture

Mouse 5TGM1 MM cell line (provided by Dr. Lori A. Hazlehurst, West Virginia University, Morgantown, WV), human RPMI-8226 (ATCC) and NCI-H929 (ATCC) MM cell lines, human MEC2 and WaC3 CLL cell lines (both provided by Dr. Javier A. Pinilla-Ibarz, Moffitt Cancer Center, Tampa, FL), human HBL2 MCL cell line (provided by Dr. Jianguo Tao, Moffitt Cancer Center, Tampa, FL), and primary mouse naïve B and Eμ-TCL1 CLL cells were cultured at 37°C in a 5% CO2 incubator using RPMI-1640 media (Gibco) supplemented with 10% heat-inactivated FBS, 2 mmol/L L-glutamine, 100 U/mL penicillin G sodium, 1 mmol/L sodium pyruvate, 0.1 mmol/L nonessential amino acids, 100 μg/mL streptomycin sulfate, and 0.1 mmol/L β-mercaptoethanol (β-ME). Human MCL cell line Z138 (ATCC) was cultured in IMDM (Gibco) with the same supplemental nutrients. Human embryonic kidney 293T (ATCC) cells and HEPA 1-6 (ATCC) mouse hepatoma cells were cultured in Dulbecco's Modified Eagle Medium (DMEM; Gibco) also with the same supplemental nutrients. Mouse J558 (ATCC) myeloma cells were cultured in DMEM plus 10% heat-inactivated horse serum and other supplemental nutrients. All cell lines were obtained from reliable sources, and routinely authenticated and tested to ensure the absence of Mycoplasma or other contaminations.

Protein isolation and immunoblotting

Cells were lysed using RIPA buffer (10 mmol/L Tris-HCl, pH 7.4, 150 mmol/L NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, and 1 mmol/L EDTA) supplemented with protease inhibitors (Roche). Protein concentrations were determined by BCA assays (Pierce). Samples were boiled in the SDS-PAGE sample buffer (62.5 mmol/L Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, and 0.1% bromophenol blue) with β-ME, and analyzed by SDS-PAGE. Proteins were transferred to nitrocellulose membranes, blocked in 5% nonfat milk (wt/vol in PBS), and immunoblotted with primary antibodies and appropriate horseradish peroxidase (HRP)–conjugated secondary antibodies (SouthernBiotech). Immunoblots were developed using Western Lightning Chemiluminescence Reagent (PerkinElmer).

XTT assays

Cells were suspended in phenol red-free media, seeded in 96-well plates, and treated with inhibitors. At indicated time points, cell viability was assessed by XTT assays (Roche). Briefly, 50 μL XTT labeling reagent, 1 μL electron-coupling reagent, and 100 μL phenol red-free RPMI media were combined and applied to each well of the 96-well plates. Cells were incubated for 4 hours in a 5% CO2 incubator to allow for the yellow tetrazolium salt XTT to be converted in metabolically active cells to the orange formazan dye, which can be detected at 492 nm. Percentages of cell viability were determined by comparing inhibitor-treated groups with controls. Data from four identical experimental groups were plotted as means ± SD. Results are representative of at least three independent experiments.

H2O2 detection

Levels of H2O2 were determined using the Amplex Red reagent in the presence of HRP. Briefly, 100,000 normal or cancerous B cells in 1 mL phenol red-free RPMI media were incubated with or without catalase (500 U/mL) in a 5% CO2 incubator for 3 hours, and seeded in 96-well cell culture plates (150 μL each well). Fifty microliters of Amplex Red reagent (containing 4 U/mL of HRP and 200 μmol/L Amplex Red) was added into each well. After 1-hour incubation in a CO2 incubator, cells were spun down, and the supernatant was transferred into 96-well black plates to measure the fluorescence intensity (Ex/Em: 571/578 nm). Each fluorescent reading value was adjusted by subtracting the background measured from phenol red-free RPMI media. For 293T and HEPA 1-6 adherent cells, 0.5 × 105 cells in 1 mL DMEM were cultured in a 12-well plate 1 day before the experiment. Cells were washed with PBS and recultured in phenol red-free DMEM containing Amplex Red reagent (1 U/mL of HRP and 50 μmol/L Amplex Red). After 1-hour incubation in a CO2 incubator, the supernatant from each well was similarly transferred and measured. The rate of H2O2 production was calculated by fitting individual data points in a linear standard equation, y = bx + c. The H2O2 standard curves were created by combining 50 μL Amplex Red reagent with 150 μL phenol red-free RPMI or DMEM containing serially diluted H2O2.

The fluorescence response of E-F02 to endogenous H2O2

Primary B cells and MEC2 CLL cells (10 × 106 cells) were washed with cold PBS and spun down. 293T and HEPA 1-6 adherent cells (0.5 × 106 cells) were washed with cold PBS, trypsinized, and spun down. Cell pellets were lysed in 50 μL of 0.5% NP-40 in cold PBS, and centrifuged for 15 minutes at 15,000 rpm at 4°C. Cell lysates were transferred to a clean tube, and proteins were precipitated from lysates using cold trichloroacetic acid (TCA). Briefly, 8 μL ice-cold 100% (w/v) TCA was combined with 50 μL lysates, incubated on ice for 10 minutes, and centrifuged at 12,000 × g for 5 minutes at 4°C. The supernatant was transferred to a clean tube, neutralized to pH 6, and centrifuged again at 13,000 × g for 15 minutes at 4°C. The supernatant was transferred to a 96-well black plate, incubated with E-F02 at the final concentration of 20 μmol/L, and measured for the fluorescence intensity (Ex/Em: 350/454 nm). Each fluorescent reading value was adjusted by subtracting the background measured from the supernatant alone.

In vitro FRET-suppression assay

The endoribonuclease activity of recombinant hIRE-1 was assayed by incubation of 10 μL of 20 nmol/L hIRE-1 and 10 μL of various concentrations (0.01–1 μmol/L) of fluorescently tagged XBP-1 RNA stem loop (5′-Cy5-CAGUCCGCAGCACUG-BHQ-3′, obtained from Sigma-Aldrich) in assay buffer (20 mmol/L HEPES, pH 7.5, 50 mmol/L KOAc, 0.5 mmol/L MgCl2, 3 mmol/L DTT, 0.4% PEG, and 5% DMSO) for up to 2 hours at room temperature (RT) in a 384 shallow well plate. Fluorescence was read with excitation and emission at 620 nm and 680 nm, respectively. The Km of purified recombinant hIRE-1 was determined using the Michaelis-Menten kinetic model. Inhibition of RNA cleavage by small molecules was determined by preincubation of 5 μL of 40 nmol/L hIRE-1 with various concentrations of compounds (10 μL) in assay buffer for 90 minutes at RT. A 200 nmol/L solution of fluorescent XBP-1 RNA (5 μL) was then added to each well and the reaction allowed to proceed for 2 hours before fluorescence reading. Final concentrations of hIRE-1 and XBP-1 RNA were 10 nmol/L and 50 nmol/L, respectively. All fluorescence readings were corrected using background values from wells containing only 20 μL of 50 nmol/L XBP-1 RNA. Dose-response experiments were carried out a minimum of 3 times. In vitro IC50 values were calculated from the mean inhibition value at each concentration.

Mice

IRE-1flox/flox, CD19Cre/IRE-1flox/flox, and Eμ-TCL1 mice were maintained at our animal facility following guidelines approved by the Wistar Institute Committee on Animal Care.

Suppression of the B-cell receptor (BCR) signal transduction by AMG-18

Mouse naïve B cells were purified from spleens of wild-type (IRE-1flox/flox) and B cell–specific IRE-1 kinase domain knockout (CD19Cre/IRE-1flox/flox) mice using CD43 Microbeads (Miltenyi), and exposed to DMSO, 10 μmol/L AMG-18, or 10 μmol/L B-I09 for 12 hours. Treated B cells were subsequently exposed to F(ab')2 fragments of the goat anti-mouse IgM antibody (SothernBiotech) at the final concentration of 20 μg/mL for 2 minutes and lysed immediately for analyses by immunoblots.

C-B06, E-D08, B-I09, D-F07, AMG-18, and 3,6-DMAD exerted low IC50 values in suppressing the RNase activity of IRE-1 in vitro

To determine the in vitro IC50 values of a select group of small molecules (Fig. 1A), which were reported to inhibit the RNase activity of IRE-1, we first purified a fresh batch of puritin-His-tagged IRE-1 kinase/RNase (aa 547–977) fusion protein from SF21 insect cells, and confirmed its activity by incubation with a synthetic XBP-1 mRNA mini-stem loop, which incorporates a Cy5 fluorophore on its 5′ end and the black hole quencher (BHQ) on its 3′ end and emits fluorescence upon site-specific cleavage by IRE-1 (Fig. 1B; ref. 19). The IRE-1 fusion protein exhibited RNase activity with a Km value of 232 nmol/L (Fig. 1C). Tricyclic chromenone compounds with the aldehyde group exposed or protected had the lowest in vitro IC50 values ranging from 21 nmol/L to 1.33 μmol/L in inhibiting XBP-1 mRNA substrate cleavage (Fig. 1D and G). Of the three allosteric kinase inhibitors, AMG-18 (IC50 = 2.33 μmol/L) was significantly more potent than KIRA6 (IC50 = 19.7 μmol/L) and GSK2850163 (IC50 = 17.1 μmol/L) in inhibiting the RNase activity of IRE-1 (Fig. 1E and G). Of the two ATP-competitive inhibitors of IRE-1, sunitinib exhibited an in vitro IC50 value of 17 μmol/L in inhibiting the RNase activity of IRE-1 while staurosporine did not have an effect (Fig. 1F and G). The acridine derivative, 3,6-DMAD, could also inhibit the RNase activity of IRE-1 with an in vitro IC50 value of 2.53 μmol/L while toyocamycin did not suppress XBP-1 mRNA substrate cleavage (Fig. 1F and G).

Figure 1.

In vitro suppression of IRE-1 RNase activity by inhibitors. A, Chemical structures of a select group of IRE-1 RNase, kinase, or oligomerization inhibitors. B, The fluorescence increase from a FRET assay was plotted as a function of hIRE-1 concentrations to determine the hIRE-1 RNase activity in cleaving fluorescently tagged XBP-1 RNA stem-loop substrates (50 nmol/L). C, The Michaelis–Menten curve for hIRE-1 showing the catalytic RNase activity in a FRET assay. Initial reaction rates were plotted as a function of different XBP1 RNA stem-loop substrate concentrations in the presence of 10 nmol/L hIRE-1. D–F, Each compound was evaluated by its activity in inhibiting the hIRE-1 RNase from cleaving fluorescently tagged XBP-1 RNA stem-loop substrates in a FRET-suppression assay. Data from at least three repeated dose-response experiments for each compound were plotted as means ± SEM. G, The in vitro IC50 value for each compound was determined from experiments shown in D–F.

Figure 1.

In vitro suppression of IRE-1 RNase activity by inhibitors. A, Chemical structures of a select group of IRE-1 RNase, kinase, or oligomerization inhibitors. B, The fluorescence increase from a FRET assay was plotted as a function of hIRE-1 concentrations to determine the hIRE-1 RNase activity in cleaving fluorescently tagged XBP-1 RNA stem-loop substrates (50 nmol/L). C, The Michaelis–Menten curve for hIRE-1 showing the catalytic RNase activity in a FRET assay. Initial reaction rates were plotted as a function of different XBP1 RNA stem-loop substrate concentrations in the presence of 10 nmol/L hIRE-1. D–F, Each compound was evaluated by its activity in inhibiting the hIRE-1 RNase from cleaving fluorescently tagged XBP-1 RNA stem-loop substrates in a FRET-suppression assay. Data from at least three repeated dose-response experiments for each compound were plotted as means ± SEM. G, The in vitro IC50 value for each compound was determined from experiments shown in D–F.

Close modal

B-I09, D-F07, and AMG-18 suppressed the expression of XBP-1s in both human and mouse myeloma cells in a dose-dependent manner

We next evaluated the ability of selected inhibitors to suppress XBP-1s expression in mouse 5TGM1 and human RPMI-8226 MM cell lines using immunoblots. Levels of XBP-1s protein in response to 24-hour treatment with each inhibitor at increasing concentrations were quantified by densitometry. Consistent with our previous results, tricyclic chromenone inhibitors with the functional aldehyde group protected as a 1,3-dioxane acetal as in B-I09 and D-F07 potently suppressed the expression of XBP-1s in both 5TGM1 and RPMI-8226 cells with in-cell IC50 values ranging from 0.33 to 1.1 μmol/L (Fig. 2A–C and E–G). As a control, treatment of 5TGM1 cells with B-I09 or D-F07 indeed led to the accumulation of the unspliced XBP-1 (XBP-1u) proteins (Supplementary Fig. S1A and S1B). Tricyclic chromenone inhibitors with the exposed aldehyde group as in C-B06 and E-D08 were less potent than their respective protected counterparts in suppressing the expression of XBP-1s in both 5TGM1 and RPMI-8226 cells (Fig. 2AC and E–G).

Figure 2.

Suppression of XBP-1s expression in MM cells by IRE-1 inhibitors. A, 5TGM1 cells were dose-dependently treated with the indicated inhibitors for 24 hours, lysed, and analyzed for XBP-1s and p97 (loading control) by immunoblots. B and C, Dose-response curves (B) and in-cell IC50 values (C) for inhibition of XBP-1s expression in 5TGM1 cells by indicated inhibitors, as determined by immunoblots and densitometry (N = 3). D, 5TGM1 cells were treated with DMSO (control) or indicated inhibitors at 20 μmol/L for 24 hours and subjected to XTT assays. E, RPMI-8226 cells were dose-dependently treated with inhibitors for 24 hours, lysed, and analyzed by immunoblots. F and G, Dose-response curves (F) and in-cell IC50 values (G) for inhibition of XBP-1s expression in RPMI-8226 cells by indicated inhibitors were similarly determined (N = 3). H, RPMI-8226 cells were treated with DMSO or inhibitors at 20 μmol/L for 24 hours and subjected to XTT assays.

Figure 2.

Suppression of XBP-1s expression in MM cells by IRE-1 inhibitors. A, 5TGM1 cells were dose-dependently treated with the indicated inhibitors for 24 hours, lysed, and analyzed for XBP-1s and p97 (loading control) by immunoblots. B and C, Dose-response curves (B) and in-cell IC50 values (C) for inhibition of XBP-1s expression in 5TGM1 cells by indicated inhibitors, as determined by immunoblots and densitometry (N = 3). D, 5TGM1 cells were treated with DMSO (control) or indicated inhibitors at 20 μmol/L for 24 hours and subjected to XTT assays. E, RPMI-8226 cells were dose-dependently treated with inhibitors for 24 hours, lysed, and analyzed by immunoblots. F and G, Dose-response curves (F) and in-cell IC50 values (G) for inhibition of XBP-1s expression in RPMI-8226 cells by indicated inhibitors were similarly determined (N = 3). H, RPMI-8226 cells were treated with DMSO or inhibitors at 20 μmol/L for 24 hours and subjected to XTT assays.

Close modal

Of the three allosteric kinase inhibitors, AMG-18 dose dependently suppressed the expression of XBP-1s with an in-cell IC50 value of 1.2 μmol/L determined in 5TGM1 cells, and 0.26 μmol/L in RPMI-8226 cells (Fig. 2A–C and E–G). KIRA6 exhibited an in-cell IC50 value of 6.1 μmol/L in suppressing XBP-1s in 5TGM1 cells, but did not suppress XBP-1s in a dose-dependent manner in RPMI-8226 cells. GSK2850163 had little effect in suppressing the expression of XBP-1s in both 5TGM1 and RPMI-8226 cells even when used at the concentration of 20 μmol/L.

Of the two ATP-binding site competitors of IRE-1, sunitinib did not suppress the expression of XBP-1s in a dose-dependent fashion in either 5TGM1 or RPMI-8226 cells (Fig. 2A–C and E–G), and the sudden disappearance of XBP-1s in cells treated with 20 μmol/L sunitinib is the outcome of cell death (Fig. 2D and H). Staurosporine exhibited an in-cell IC50 value of 0.48 μmol/L in suppressing the expression of XBP-1s in RPMI-8226 cells. However, when 5TGM1 cells were treated with staurosporine at nanomolar concentrations, it did not suppress the expression of XBP-1s in a dose-dependent manner, and the use of staurosporine at > 1 μmol/L induced rapid apoptosis of 5TGM1 cells (Fig. 2D).

The acridine derivative, 3,6-DMAD, also did not inhibit the expression of XBP-1s in a dose-dependent manner. When 5TGM1 cells were treated with 10 or 20 μmol/L 3,6-DMAD, they produced significantly increased levels of XBP-1s (Fig. 2A). Conversely, treatment with 3,6-DMAD caused RPMI-8226 cells to undergo rapid apoptosis (Fig. 2H) and so reduced levels of XBP-1s (Fig. 2E). Toyocamycin also rapidly induced apoptosis in 5TGM1 and RPMI-8226 cells, resulting in no detectable XBP-1s (Fig. 2A, D, E, and H).

Although B-I09 and D-F07 could continuously suppress the expression of XBP-1s, they were not as potent as AMG-18 in inducing apoptosis in 5TGM1 and RPMI-8226 cells

Of all compounds examined, B-I09, D-F07, and AMG-18 could suppress the expression of XBP-1s dose-dependently in both 5TGM1 and RPMI-8226 cells (Fig. 2A–C and E–G). The dose-dependent suppression of XBP-1s by these three inhibitors was also confirmed in NCI-H929 human myeloma, J558 mouse myeloma, and LPS-stimulated primary mouse Eμ-TCL1 CLL cells (Supplementary Fig. S1C–S1E). We next treated 5TGM1 and RPMI-8226 cells with 20 μmol/L B-I09, D-F07, and AMG-18 for a course of 3 days (Fig. 3A and B). We chose to treat these MM cells with inhibitors at 20 μmol/L because only at this dose could AMG-18 completely suppress the expression of XBP-1s in 5TGM1 cells (Fig. 2A). AMG-18 is much more cytotoxic than B-I09 and D-F07, with almost no 5TGM1 and RPMI-8226 cells surviving for more than 48 hours in cultured media containing 20 μmol/L AMG-18. When compared with B-I09 or D-F07, AMG-18 was indeed more potent in inducing apoptosis, as evidenced by the cleavage of PARP in 5TGM1 and RPMI-8226 cells treated with AMG-18 for 12 hours (Fig. 3C and D). To establish the correlation between the suppressed levels of XBP-1s and apoptosis, we exposed 5TGM1 and RPMI-8226 cells to a reduced dose of 10 μmol/L AMG-18, B-I09, or D-F07 for 3 days so that we could obtain live AMG-18–treated 5TGM1 and RPMI-8226 cells for the analysis of XBP-1s by immunoblots (Fig. 3E and F). Unexpectedly, while 10 μmol/L B-I09 or D-F07 could continuously suppress the expression of XBP-1s in both 5TGM1 and RPMI-8226 cells for 3 consecutive days, 10 μmol/L AMG-18 began to significantly lose its inhibitory effect after 24 hours in 5TGM1 cells (Fig. 3E) and after 48 hours in RPMI-8226 cells (Fig. 3F). We further treated 5TGM1 cells with 5 μmol/L AMG-18, B-I09 or D-F07 (Supplementary Fig. S1F). Congruous with data shown in Fig. 3E, B-I09 or D-F07 dosed at 5 μmol/L could still significantly suppress the expression of XBP-1s in 5TGM1 cells on day 3 while AMG-18 dosed at 5 μmol/L completely lost its inhibitory effect (Supplementary Fig. S1F).

Figure 3.

AMG-18 was more cytotoxic than B-I09 and D-F07 due to its off-target effects. A and B, 5TGM1 (A) and RPMI-8226 (B) cells were treated with DMSO, B-I09, D-F07, or AMG-18 at 20 μmol/L for 3 days and subjected to XTT assays. C and D, 5TGM1 (C) and RPMI-8226 (D) cells were treated with indicated inhibitors at 20 μmol/L for 12 hours, lysed, and analyzed by immunoblots. E and F, 5TGM1 (E) and RPMI-8226 (F) cells were treated with indicated inhibitors at 10 μmol/L for 3 days, lysed, and analyzed by immunoblots. G and H, Naïve B cells from wild-type (G) and B cell–specific IRE-1KO (H) mice were treated with DMSO, AMG-18 (10 μmol/L), or B-I09 (10 μmol/L) for 12 hours, subsequently stimulated with F(ab′)2 goat anti-mouse IgM for 2 minutes, and immediately lysed for immunoblot analyses.

Figure 3.

AMG-18 was more cytotoxic than B-I09 and D-F07 due to its off-target effects. A and B, 5TGM1 (A) and RPMI-8226 (B) cells were treated with DMSO, B-I09, D-F07, or AMG-18 at 20 μmol/L for 3 days and subjected to XTT assays. C and D, 5TGM1 (C) and RPMI-8226 (D) cells were treated with indicated inhibitors at 20 μmol/L for 12 hours, lysed, and analyzed by immunoblots. E and F, 5TGM1 (E) and RPMI-8226 (F) cells were treated with indicated inhibitors at 10 μmol/L for 3 days, lysed, and analyzed by immunoblots. G and H, Naïve B cells from wild-type (G) and B cell–specific IRE-1KO (H) mice were treated with DMSO, AMG-18 (10 μmol/L), or B-I09 (10 μmol/L) for 12 hours, subsequently stimulated with F(ab′)2 goat anti-mouse IgM for 2 minutes, and immediately lysed for immunoblot analyses.

Close modal

Because AMG-18 is based on a kinase inhibitor scaffold, we tested whether it could also suppress phosphorylation of certain tyrosine kinases associated with BCR activation. We purified naïve B cells from both wild-type (IRE-1flox/flox) and B cell–specific IRE-1KO (CD19Cre/IRE-1flox/flox) mice (15), exposed these cells to DMSO or 10 μmol/L AMG-18 or B-I09 for 12 hours, subsequently stimulated these cells with F(ab')2 to activate the BCR, and examined tyrosine phosphorylation of BCR activation-associated kinases and proteins by immunoblots using an anti-phospho-tyrosine antibody, 4G10 (Fig. 3G and H). Genetic deletion or pharmacologic inhibition of XBP-1s in LPS-stimulated mouse B cells or CLL cells had been shown to compromise BCR signaling (13, 19). However, because naïve B cells from both wild-type and B cell–specific IRE-1KO mice did not express IRE-1 and XBP-1s, inhibition of the expression of XBP-1s using B-I09 did not suppress BCR signaling in naïve B cells (Fig. 3G and H; ref. 13). In contrast to B-I09, AMG-18 potently suppressed phosphorylation of many BCR activation–associated kinases and proteins, and one of these was BTK, a critical BCR signaling molecule supporting the growth and survival of CLL (Fig. 3G and H). Such results suggested that AMG-18 has off-target effects that contribute to its cytotoxicity.

Cytotoxicities of B-I09 and D-F07 were enhanced by PI3K/AKT pathway inhibitors in MM, CLL, and MCL cells

The cytotoxicity induced by AMG-18 in 5TGM1 and RPMI-8226 cells could not be completely attributed to the suppression of XBP-1s, because AMG-18–treated cells still expressed XBP-1s (Fig. 3E and F; Supplementary Fig. S1F). On the other hand, B-I09 and D-F07 almost completely suppressed the expression of XBP-1s in 5TGM1 and RPMI-8226 cells, but they were not as effective as AMG-18 in killing these cells. We thus decided to explore whether we could enhance the cytotoxicity of B-I09 and D-F07 against MM cells by rational combinations with other FDA-approved tumoricidal inhibitors. We showed previously that B-I09 could synergize with ibrutinib (a BTK inhibitor) to induce apoptosis in CLL, MM, and MCL cells potentially due to strong suppression of phosphorylated AKT (19). We thus tested whether combining IRE-1 RNase inhibitors with inhibitors of the PI3K/AKT pathway could enhance cytotoxicity against MM, CLL, and MCL cells.

We first exposed 5TGM1 or RPMI-8226 cells to 10 μmol/L MK2206 (a potent inhibitor for AKT1, AKT2, and AKT3) or 10 μmol/L CAL-101 (a potent inhibitor for PI3Kδ) for 4 days, and showed that MK2206, but not CAL-101, could suppress the growth of both MM cell lines (Fig. 4AD). The treatment with MK2206 alone was potent in suppressing the growth of RPMI-8226 cells (Fig. 4C). When B-I09 (20 μmol/L) or D-F07 (20 μmol/L) was combined with MK2206 (10 μmol/L) or CAL-101 (10 μmol/L), we found that MK2206 could also significantly enhance the cytotoxicity of B-I09 and D-F07 in 5TGM1 cells (Fig. 4A). Although treatment with CAL-101 alone did not affect the growth of 5TGM1 and RPMI-8226 cells, CAL-101 clearly enhanced the cytotoxicity of B-I09 and D-F07 in both MM cell lines (Fig. 4B and D). We further confirmed the effect of combination treatments on the expression of XBP-1s, IRE-1, and AKT in 5TGM1 and RPMI-8226 cell lines (Supplementary Fig. S2A–S2D). Although we observed that B-I09 and D-F07 could effectively suppress the expression of XBP-1s, we could not detect in 5TGM1 and RPMI-8226 cells the phosphorylation of AKT at Ser473, which is readily phosphorylated in human MEC2 CLL cells (Supplementary Fig. S2A–S2D). We hypothesize that MM cells may be different from CLL and MCL cells in that they constitutively activate different phosphorylation sites of AKT.

Figure 4.

B-I09- or D-F07–induced cytotoxicity in human MM and CLL cell lines was enhanced by PI3K/AKT pathway inhibitors. A, 5TGM1 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (10 μmol/L), or the combinations for 4 days, and subjected to XTT assays. B, 5TGM1 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), CAL-101 (10 μmol/L), or the combinations for 4 days and subjected to XTT assays. C, RPMI-8226 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (10 μmol/L), or the combinations for 4 days and subjected to XTT assays. D, RPMI-8226 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), CAL-101 (10 μmol/L), or the combinations for 4 days and subjected to XTT assays. E, MEC2 cells were dose-dependently treated with B-I09 or D-F07 for 24 hours, lysed, and analyzed by immunoblots. F, Dose-response curves together with in-cell IC50 values for inhibition of XBP-1s expression in MEC2 cells by B-I09 and D-F07 were determined by immunoblots and densitometry (N = 3). G, MEC2 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (1 μmol/L), or the combinations for 4 days and subjected to XTT assays. H, MEC2 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), CAL-101 (1 μmol/L), or the combinations for 4 days and subjected to XTT assays. I, MEC2 cells were treated with DMSO, MK2206 (1–20 μmol/L), B-I09 (20 μmol/L), or the combination for 24 hours, and lysed for immunoblot analyses. J, MEC2 cells were treated with DMSO, CAL-101 (1–50 μmol/L), B-I09 (20 μmol/L), or the combination for 24 hours, and lysed for immunoblot analyses.

Figure 4.

B-I09- or D-F07–induced cytotoxicity in human MM and CLL cell lines was enhanced by PI3K/AKT pathway inhibitors. A, 5TGM1 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (10 μmol/L), or the combinations for 4 days, and subjected to XTT assays. B, 5TGM1 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), CAL-101 (10 μmol/L), or the combinations for 4 days and subjected to XTT assays. C, RPMI-8226 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (10 μmol/L), or the combinations for 4 days and subjected to XTT assays. D, RPMI-8226 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), CAL-101 (10 μmol/L), or the combinations for 4 days and subjected to XTT assays. E, MEC2 cells were dose-dependently treated with B-I09 or D-F07 for 24 hours, lysed, and analyzed by immunoblots. F, Dose-response curves together with in-cell IC50 values for inhibition of XBP-1s expression in MEC2 cells by B-I09 and D-F07 were determined by immunoblots and densitometry (N = 3). G, MEC2 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (1 μmol/L), or the combinations for 4 days and subjected to XTT assays. H, MEC2 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), CAL-101 (1 μmol/L), or the combinations for 4 days and subjected to XTT assays. I, MEC2 cells were treated with DMSO, MK2206 (1–20 μmol/L), B-I09 (20 μmol/L), or the combination for 24 hours, and lysed for immunoblot analyses. J, MEC2 cells were treated with DMSO, CAL-101 (1–50 μmol/L), B-I09 (20 μmol/L), or the combination for 24 hours, and lysed for immunoblot analyses.

Close modal

Because B-I09 and D-F07 are potent in suppressing XBP-1s in mouse Eμ-TCL1 CLL cells (Supplementary Fig. S1E), we similarly determined that B-I09 (in-cell IC50 = 0.9 μmol/L) and D-F07 (in-cell IC50 = 0.15 μmol/L) were also potent inhibitors of XBP-1s expression in human MEC2 CLL cells (Fig. 4E and F). After MEC2 cells were treated with MK2206 or CAL-101 at 1 μmol/L for 4 days, we found MK2206 and CAL-101 suppressed approximately 60% and 50% cell viability, respectively (Fig. 4G and H). Although B-I09 and D-F07 were not as effective as MK2206 or CAL-101 in killing MEC2 cells, their combinations with MK2206 or CAL-101 significantly enhanced cytotoxicity against MEC2 cells (Fig. 4G and H). B-I09 at 20 μmol/L did not alter the levels of phosphorylated AKT in MEC2 cells (Fig. 4I and J). As a response to treatment with MK2206 or CAL-101, MEC2 cells dose-dependently upregulated their expression of XBP-1s without the involvement of phosphorylation of S729 in the kinase activation loop of IRE-1 (Supplementary Fig. S2E), and such upregulated levels of XBP-1s could be suppressed by B-I09 (Fig. 4I and J). We next treated primary CLL cells freshly purified from Eμ-TCL1 mice with MK2206 or CAL-101 at 1 μmol/L, or with B-I09 or D-F07 at 10 μmol/L for 3 days. We found that CAL-101 was more potent than MK2206, B-I09, and D-F07 in inducing cytotoxicity in mouse CLL cells (Fig. 5A and B). Combinations of B-I09 or D-F07 with MK2206 or CAL-101 significantly enhanced cytotoxicity against mouse CLL cells (Fig. 5A and B). B-I09 or D-F07 at 20 μmol/L did not significantly alter the levels of phosphorylated AKT in mouse CLL cells but could potently suppress XBP-1s even when XBP-1s was upregulated in response to treatment with MK2206 or CAL-101 (Fig. 5C).

Figure 5.

B-I09- or D-F07–induced cytotoxicity in mouse Eμ-TCL1 CLL cells and human MCL cell lines was enhanced by PI3K/AKT pathway inhibitors. A, Mouse CLL cells were treated with DMSO, B-I09 (10 μmol/L), D-F07 (10 μmol/L), MK2206 (1 μmol/L), or the combinations for 3 days and subjected to XTT assays. B, Mouse CLL cells were treated with DMSO, B-I09 (10 μmol/L), D-F07 (10 μmol/L), CAL-101 (1 μmol/L), or the combinations for 3 days and subjected to XTT assays. C, 48-hours LPS-stimulated mouse CLL cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (20 μmol/L), CAL-101 (20 μmol/L), or the combinations for 24 hours, and lysed for immunoblot analyses. D, Z138 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (1 μmol/L), or the combinations for 4 days and subjected to XTT assays. E, Z138 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), CAL-101 (1 μmol/L), or the combinations for 4 days and subjected to XTT assays. F, Z138 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (20 μmol/L), CAL-101 (20 μmol/L), or the combinations for 24 hours, and lysed for immunoblot analyses. G, HBL2 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (10 μmol/L), or the combinations for 4 days and subjected to XTT assays. H, HBL2 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), CAL-101 (10 μmol/L), or the combinations for 4 days and subjected to XTT assays. I, HBL2 cells were treated with DMSO, B-I09 (20 μmol/L), MK2206 (20 μmol/L), CAL-101 (20 μmol/L), or the combinations for 24 hours, and lysed for immunoblot analyses.

Figure 5.

B-I09- or D-F07–induced cytotoxicity in mouse Eμ-TCL1 CLL cells and human MCL cell lines was enhanced by PI3K/AKT pathway inhibitors. A, Mouse CLL cells were treated with DMSO, B-I09 (10 μmol/L), D-F07 (10 μmol/L), MK2206 (1 μmol/L), or the combinations for 3 days and subjected to XTT assays. B, Mouse CLL cells were treated with DMSO, B-I09 (10 μmol/L), D-F07 (10 μmol/L), CAL-101 (1 μmol/L), or the combinations for 3 days and subjected to XTT assays. C, 48-hours LPS-stimulated mouse CLL cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (20 μmol/L), CAL-101 (20 μmol/L), or the combinations for 24 hours, and lysed for immunoblot analyses. D, Z138 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (1 μmol/L), or the combinations for 4 days and subjected to XTT assays. E, Z138 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), CAL-101 (1 μmol/L), or the combinations for 4 days and subjected to XTT assays. F, Z138 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (20 μmol/L), CAL-101 (20 μmol/L), or the combinations for 24 hours, and lysed for immunoblot analyses. G, HBL2 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), MK2206 (10 μmol/L), or the combinations for 4 days and subjected to XTT assays. H, HBL2 cells were treated with DMSO, B-I09 (20 μmol/L), D-F07 (20 μmol/L), CAL-101 (10 μmol/L), or the combinations for 4 days and subjected to XTT assays. I, HBL2 cells were treated with DMSO, B-I09 (20 μmol/L), MK2206 (20 μmol/L), CAL-101 (20 μmol/L), or the combinations for 24 hours, and lysed for immunoblot analyses.

Close modal

Before conducting combination studies in MCL, we first demonstrated that B-I09 and D-F07 could dose-dependently suppress XBP-1s in two chosen MCL cell lines, Z138 and HBL2 (Supplementary Fig. S3). When these two cell lines were treated with B-I09 or D-F07 in combination with MK2206 or CAL-101, we again observed that the combined treatment was superior to that with a single agent in inducing cytotoxicity (Fig. 5D and E, G and H). Like the results from human and mouse CLL cells, B-I09 or D-F07 did not alter the levels of phosphorylated AKT or AKT but could potently suppress XBP-1s in Z138 and HBL2 cells regardless of treatment by MK2206 or CAL-101 (Fig. 5F and I).

Installation of an ROS-sensitive boronate cage on the hydroxy group of D-F07 stabilized the 1,3-dioxane acetal

We previously reported that both B-I09 and D-F07 could emit strong fluorescence and that installation of a photolabile cage onto the hydroxyl group of D-F07 could stabilize the 1,3-dioxane acetal protecting group and allow for specific stimulus-mediated control of XBP-1s inhibition in cells (21). To develop an inhibitor that could target the expression of XBP-1s in tumor cells, we chose to install ROS-sensitive boronate cages onto the C8 hydroxy of D-F07 (as in E-G04 and E-E08) and B-I09 (as in E-F02; Fig. 6A). We envisioned that replacement of the C8 hydroxy with a boronate could reduce the fluorescence of D-F07 or B-I09 due to the electron withdrawing nature of boron and that cleavage of the boronate could be induced by H2O2 or other ROS to allow for liberation of D-F07 or B-I09. We anticipated that intracellular H2O2 could also affect this cleavage (39, 40) because its concentration is often significantly elevated in tumor cells (41–44). We synthesized ROS-sensitive boron-caged E-G04, E-E08, and E-F02 (Fig. 6A), incubated these compounds with increasing concentrations of H2O2, and measured increased fluorescence intensity resulting from liberated D-F07 (Fig. 6B and C) or B-I09 (Fig. 6D). When E-G04 was incubated with increasing concentrations of H2O2, we observed dose-dependent fluorescence enhancement; however, incubation with high concentrations of H2O2 at 37°C was required to oxidize the 1,3-dioxaborolane moiety from E-G04 to generate D-F07 (Fig. 6B; Supplementary Fig. S4A), making E-G04 unsuitable for use in cells. In contrast to E-G04, E-E08 and E-F02 (harboring the aryl boronate) were readily oxidized after incubation with lower concentrations of H2O2 for 1 hour at RT (Fig. 6C and D; Supplementary Fig. S4A). During the synthesis of the borylated analogues, we noted that the final phenol O-methylation to give E-E08 resulted in significant overalkylation at the tertiary amine. As a result, the yield of E-E08 was much lower than that of E-F02 (∼36% vs. ∼94% calculated from the starting materials). We thus synthesized E-F02 in larger quantities and took advantage of fluorescence decay to monitor the hydrolysis rate of the 1,3-dioxane acetal group from B-I09 and E-F02 in aqueous solution (Fig. 6E). Installation of the boron cage similarly stabilized the 1,3-dioxane acetal protecting group on E-F02 (Fig. 6E).

Figure 6.

Installation of ROS-sensitive boronate cages on the hydroxy group of D-F07 or B-I09 stabilized the 1,3-dioxane acetal protecting group and allowed for cleavage by H2O2 to generate D-F07 or B-I09. A, A structural scheme showing that installation of boronate cages on the hydroxy group of D-F07 (as in E-G04 and E-E08) or B-I09 (as in E-F02) could stabilize the 1,3-dioxane acetal protecting moiety. B, Fluorescence of E-G04 (2.5 μmol/L in the DMSO/PBS solution (v/v = 1:99), Ex = 360 nm) increased upon incubation with 0 to 2,000 μmol/L H2O2 at 37°C for 1 hour. C, Fluorescence of E-E08 (2.5 μmol/L in the DMSO/PBS solution (v/v = 1:99), Ex = 360 nm) increased upon incubation with 0 to 200 μmol/L H2O2 at RT for 1 hour. D, Fluorescence of E-F02 (2.5 μmol/L in the DMSO/PBS solution (v/v = 1:99), Ex = 360 nm) increased upon incubation with 0 to 200 μmol/L H2O2 at RT for 1 hour.E, Fluorescence stability of B-I09 and E-F02 at 37°C was monitored from 0 to 48 hours (10 μmol/L, DMSO/PBS mixture (v/v = 1:99), Ex = 360 nm). I0 was the initial fluorescence intensity of B-I09 or E-F02 at 0 hour. Based on fluorescence decay, the decomposition rates of B-I09 and E-F02 were plotted as a function of time. F, The levels of H2O2 produced by primary mouse B, 5TGM1, 72-h LPS-stimulated primary mouse CLL, MEC2, and WaC3 cells were measured by Amplex Red with or without the addition of extracellular catalase (500 U/mL). G, Fluorescence response of E-F02 to endogenous H2O2 produced by primary B cells or MEC2 CLL cells was investigated by incubating equal amounts of cytosol from both cell types with 20 μmol/L E-F02, and monitored at RT for 0 to 90 minutes. I0 was the initial fluorescence intensity of E-F02 at 0 minute. H, The levels of H2O2 produced by 293T and HEPA 1-6 cells were measured by Amplex Red. I, Fluorescence response of E-F02 to endogenous H2O2 produced by 293T or HEPA 1-6 cells was investigated by incubating equal amounts of cytosol from both cell types with 20 μmol/L E-F02, and similarly monitored as in G. J, MEC2 cells were treated with E-F02 at indicated concentrations for 24 hours and lysed for immunoblot analyses. K, MEC2 cells were treated with E-F02 at 20 μmol/L for 16 hours, and subsequently incubated with H2O2 at indicated concentrations for additional 3 hours. Cells were lysed and analyzed by immunoblots. L, MEC2 cells were treated for 24 hours with E-F02 (20 μmol/L), Auranofin (1 μmol/L) or the combination and lysed for immunoblot analyses. M, RPMI-8226 cells were treated for 24 hours with E-F02 (20 μmol/L), Auranofin (0.1 μmol/L), or the combination and lysed for immunoblot analyses. N, MEC2 cells were treated with DMSO (control), E-F02 (20 μmol/L), Auranofin (0.1 μmol/L), or the combination for 3 days and subjected to XTT assays. O, RPMI-8226 cells were treated with DMSO, E-F02 (20 μmol/L), Auranofin (0.1 μmol/L), or the combination for 3 days and subjected to XTT assays. P, Z138 cells were treated with DMSO, E-F02 (20 μmol/L), Auranofin (0.25 μmol/L), or the combination for 3 days and subjected to XTT assays.

Figure 6.

Installation of ROS-sensitive boronate cages on the hydroxy group of D-F07 or B-I09 stabilized the 1,3-dioxane acetal protecting group and allowed for cleavage by H2O2 to generate D-F07 or B-I09. A, A structural scheme showing that installation of boronate cages on the hydroxy group of D-F07 (as in E-G04 and E-E08) or B-I09 (as in E-F02) could stabilize the 1,3-dioxane acetal protecting moiety. B, Fluorescence of E-G04 (2.5 μmol/L in the DMSO/PBS solution (v/v = 1:99), Ex = 360 nm) increased upon incubation with 0 to 2,000 μmol/L H2O2 at 37°C for 1 hour. C, Fluorescence of E-E08 (2.5 μmol/L in the DMSO/PBS solution (v/v = 1:99), Ex = 360 nm) increased upon incubation with 0 to 200 μmol/L H2O2 at RT for 1 hour. D, Fluorescence of E-F02 (2.5 μmol/L in the DMSO/PBS solution (v/v = 1:99), Ex = 360 nm) increased upon incubation with 0 to 200 μmol/L H2O2 at RT for 1 hour.E, Fluorescence stability of B-I09 and E-F02 at 37°C was monitored from 0 to 48 hours (10 μmol/L, DMSO/PBS mixture (v/v = 1:99), Ex = 360 nm). I0 was the initial fluorescence intensity of B-I09 or E-F02 at 0 hour. Based on fluorescence decay, the decomposition rates of B-I09 and E-F02 were plotted as a function of time. F, The levels of H2O2 produced by primary mouse B, 5TGM1, 72-h LPS-stimulated primary mouse CLL, MEC2, and WaC3 cells were measured by Amplex Red with or without the addition of extracellular catalase (500 U/mL). G, Fluorescence response of E-F02 to endogenous H2O2 produced by primary B cells or MEC2 CLL cells was investigated by incubating equal amounts of cytosol from both cell types with 20 μmol/L E-F02, and monitored at RT for 0 to 90 minutes. I0 was the initial fluorescence intensity of E-F02 at 0 minute. H, The levels of H2O2 produced by 293T and HEPA 1-6 cells were measured by Amplex Red. I, Fluorescence response of E-F02 to endogenous H2O2 produced by 293T or HEPA 1-6 cells was investigated by incubating equal amounts of cytosol from both cell types with 20 μmol/L E-F02, and similarly monitored as in G. J, MEC2 cells were treated with E-F02 at indicated concentrations for 24 hours and lysed for immunoblot analyses. K, MEC2 cells were treated with E-F02 at 20 μmol/L for 16 hours, and subsequently incubated with H2O2 at indicated concentrations for additional 3 hours. Cells were lysed and analyzed by immunoblots. L, MEC2 cells were treated for 24 hours with E-F02 (20 μmol/L), Auranofin (1 μmol/L) or the combination and lysed for immunoblot analyses. M, RPMI-8226 cells were treated for 24 hours with E-F02 (20 μmol/L), Auranofin (0.1 μmol/L), or the combination and lysed for immunoblot analyses. N, MEC2 cells were treated with DMSO (control), E-F02 (20 μmol/L), Auranofin (0.1 μmol/L), or the combination for 3 days and subjected to XTT assays. O, RPMI-8226 cells were treated with DMSO, E-F02 (20 μmol/L), Auranofin (0.1 μmol/L), or the combination for 3 days and subjected to XTT assays. P, Z138 cells were treated with DMSO, E-F02 (20 μmol/L), Auranofin (0.25 μmol/L), or the combination for 3 days and subjected to XTT assays.

Close modal

E-F02 could be cleaved by endogenous H2O2 produced by cancerous B cells, and its cytotoxicity could be enhanced by combination with a H2O2-inducing agent

We first surveyed the levels of H2O2 in different types of cancerous B cells (Fig. 6F; Supplementary Fig. S4B and S4C). When compared with normal mouse B cells, 5TGM1 mouse MM, LPS-stimulated primary mouse CLL, human MEC2 CLL and human WaC3 CLL cells all produced significantly higher levels of H2O2 (Fig. 6F). As an assay control, treatment of these cells with 500 U/mL extracellular catalase resulted in reduced levels of H2O2 (Fig. 6F; Supplementary Fig. S4C). To demonstrate that tumor cells could produce higher levels of ROS to cleave E-F02 and liberate fluorescent B-I09, we incubated equal amounts of cytosol from normal B cells and MEC2 cells with 20 μmol/L E-F02 for 0, 15, 30, 60, and 90 minutes, and observed time-dependent increase of fluorescence in the cytosol of MEC2 cells but not in that of normal B cells (Fig. 6G). We next compared the levels of H2O2 in SV40 large T antigen–expressing 293T cells and mouse hepatoma HEPA 1-6 cells, and showed that HEPA 1-6 cells expressed higher levels of H2O2 than 293T cells (Fig. 6H; Supplementary Fig. S4D and S4E). We similarly incubated equal amounts of cytosol from 293T and HEPA 1-6 cells with 20 μmol/L E-F02, and observed significantly higher levels of fluorescence in the cytosol of HEPA 1-6 cells than in that of 293T cells (Fig. 6I), suggesting more efficient cleavage of E-F02 into fluorescent B-I09 in HEPA 1-6 cells. We also treated MEC2 cells with increasing concentrations of E-F02 for 24 hours and observed the dose-dependent suppression of XBP-1s (Fig. 6J), suggesting that the endogenous H2O2 produced by MEC2 cells could oxidize the aryl boronate of E-F02 to generate B-I09. Enhanced suppression of XBP-1s could also be observed when E-F02–treated MEC2 cells were further exposed to exogenous H2O2 for additional 3 hours (Fig. 6K). As controls, we showed that treatment with H2O2 or NAC (an ROS scavenger) alone for 3 hours did not significantly alter the expression levels of XBP-1s in MEC2 cells (Supplementary Fig. S4F). To highlight the translatability of E-F02, we treated MEC2 CLL and RPMI-8226 MM cells with E-F02 in combination with Auranofin, an FDA-approved drug for cancer therapy and an inducer of H2O2 production. Auranofin triggered the expression of XBP-1s in MEC2 cells but not in RPMI-8226 cells (Fig. 6L and M), and it was not cytotoxic to MEC2 or RPMI-8226 cells (Fig. 6N and O). Combination of E-F02 with Auranofin further suppressed the expression of XBP-1s and induced synergistic cytotoxicity in MEC2 and RPMI-8226 cells (Fig. 6LO; Supplementary Table S1). Similar synergistic cytotoxicity of E-F02 and Auranofin was confirmed in Z138 cells (Fig. 6P; Supplementary Table S1).

We compared tricyclic chromenone salicylaldehyde-based compounds with seven inhibitors which were reported to potently suppress the expression of XBP-1s and induce cancer cell death. Our data showed that KIRA6, GSK2850163, sunitinib, staurosporine, and toyocamycin had weak or no activity in inhibiting IRE-1 from cleaving XBP-1 mRNA substrates in vitro (Fig. 1G). These five inhibitors also did not inhibit the expression of XBP-1s in a dose-dependent manner in at least one of two MM cell lines (5TGM1 or RPMI-8226), suggesting that they are not specific inhibitors of the IRE-1/XBP-1 pathway (Fig. 2). Although 3,6-DMAD showed an in vitro IC50 value of 2.53 μmol/L in inhibiting the RNase activity of IRE-1 (Fig. 1G), it also did not suppress the expression of XBP-1s in a dose-dependent manner in both MM cell lines (Fig. 2). Unexpectedly, 5TGM1 cells treated with 3,6-DMAD at 10 or 20 μmol/L expressed dramatically increased levels of XBP-1s (Fig. 2A), suggesting that 3,6-DMAD is not a specific IRE-1/XBP-1 pathway inhibitor. AMG-18 exhibited an in vitro IC50 value of 2.33 μmol/L in inhibiting the RNase activity of IRE-1 (Fig. 1G), and it suppressed the expression of XBP-1s in a dose-dependent fashion in both MM cell lines (Fig. 2). However, AMG-18 harbored off-target effects, as evidenced by its suppression of BCR activation-mediated tyrosine phosphorylation of kinases and proteins in naïve wild-type and IRE-1KO B lymphocytes that could not express functional IRE-1 and XBP-1s proteins (Fig. 3G and H). In terms of growth inhibition, AMG-18 and the other 6 compounds, when similarly dosed at 20 μmol/L for 24 hours, exerted much higher cytotoxicity than the 4 tricyclic chromenone-based compounds in at least one of the two MM cell lines (Fig. 2D and H), similarly suggesting that AMG-18 and the other 6 compounds may have effects other than suppressing XBP-1s. Different from these 7 compounds, tricyclic chromenone-based inhibitors when dosed at 20 μmol/L for 24 hours almost completely inhibited the expression of XBP-1s in both MM cell lines (Fig. 2A and E); however, the lack of XBP-1s resulting from such treatment did not significantly impose cytotoxicity onto both MM cell lines (Fig. 2D and H).

C-B06 and E-D08 were identified as potent tricyclic chromenone salicylaldehyde-based compounds in inhibiting the RNase activity of IRE-1 in vitro (Fig. 1G). The reactive aldehyde group of C-B06 and E-D08 was further protected with the 1,3-dioxane acetal to generate B-I09 and D-F07, respectively, to increase their potency of inhibiting XBP-1s in whole cells (Figs. 1A and 2). Installation of the 1,3-dioxane acetal allowed for the restoration of blue fluorescence from the coumarin chromophore in B-I09 and D-F07, and the gradual release of the active compounds, C-B06 or E-D08, via slow decomposition of the 1,3-dioxane acetal facilitated the sustained inhibition of XBP-1s in whole cells (19, 21). We further took advantage of a prodrug strategy to install boronate cages to B-I09 or D-F07 to achieve tumor-specific release of active compounds at efficacious levels. The higher levels of endogenous H2O2 produced by tumor cells could efficiently cleave the boronate cage of E-F02 to allow for specific release of B-I09 to suppress the expression of XBP-1s (Fig. 6FJ). The endogenous levels of H2O2 in tumor cells could be further boosted by the use of Auranofin to speed up the cleavage of the boronate cage on E-F02 to release B-I09, inhibit XBP-1s, and kill tumor cells (Fig. 6LP).

J.R. Del Valle reports grants from the NIH during the conduct of the study. No potential conflicts of interest were disclosed by the other authors.

A. Shao: Conceptualization, data curation, formal analysis, investigation, writing–original draft, writing–review, and editing. Q. Xu: Data curation, formal analysis, investigation, and writing–original draft. W.T. Spalek: Data curation, formal analysis, and investigation. C.F. Cain: Data curation, formal analysis, investigation, and writing–original draft. C.W. Kang: Data curation, formal analysis, investigation, and writing–original draft. C.-H.A. Tang: Conceptualization, data curation, formal analysis, supervision, investigation, writing–original draft, project administration, writing–review, and editing. J.R. Del Valle: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, investigation, writing–original draft, project administration, writing–review, and editing. C.-C.A. Hu: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, investigation, writing–original draft, project administration, writing–review, and editing.

This study was partially supported by grants (R01CA163910 and R01CA190860) from the NIH/NCI. The authors thank Avery C. Lee for reading the manuscript and making useful suggestions.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Cox
JS
,
Shamu
CE
,
Walter
P
. 
Transcriptional induction of genes encoding endoplasmic reticulum resident proteins requires a transmembrane protein kinase
.
Cell
1993
;
73
:
1197
206
.
2.
Shamu
CE
,
Walter
P
. 
Oligomerization and phosphorylation of the Ire1p kinase during intracellular signaling from the endoplasmic reticulum to the nucleus
.
EMBO J
1996
;
15
:
3028
39
.
3.
Niwa
M
,
Sidrauski
C
,
Kaufman
RJ
,
Walter
P
. 
A role for presenilin-1 in nuclear accumulation of Ire1 fragments and induction of the mammalian unfolded protein response
.
Cell
1999
;
99
:
691
702
.
4.
Yoshida
H
,
Matsui
T
,
Yamamoto
A
,
Okada
T
,
Mori
K
. 
XBP1 mRNA is induced by ATF6 and spliced by IRE1 in response to ER stress to produce a highly active transcription factor
.
Cell
2001
;
107
:
881
91
.
5.
Shen
X
,
Ellis
RE
,
Lee
K
,
Liu
CY
,
Yang
K
,
Solomon
A
, et al
Complementary signaling pathways regulate the unfolded protein response and are required for C. elegans development
.
Cell
2001
;
107
:
893
903
.
6.
Calfon
M
,
Zeng
H
,
Urano
F
,
Till
JH
,
Hubbard
SR
,
Harding
HP
, et al
IRE1 couples endoplasmic reticulum load to secretory capacity by processing the XBP-1 mRNA
.
Nature
2002
;
415
:
92
6
.
7.
Rutkowski
DT
,
Hegde
RS
. 
Regulation of basal cellular physiology by the homeostatic unfolded protein response
.
J Cell Biol
2010
;
189
:
783
94
.
8.
Walter
P
,
Ron
D
. 
The unfolded protein response: from stress pathway to homeostatic regulation
.
Science
2011
;
334
:
1081
6
.
9.
Reimold
AM
,
Iwakoshi
NN
,
Manis
J
,
Vallabhajosyula
P
,
Szomolanyi-Tsuda
E
,
Gravallese
EM
, et al
Plasma cell differentiation requires the transcription factor XBP-1
.
Nature
2001
;
412
:
300
7
.
10.
Iwakoshi
NN
,
Lee
AH
,
Vallabhajosyula
P
,
Otipoby
KL
,
Rajewsky
K
,
Glimcher
LH
. 
Plasma cell differentiation and the unfolded protein response intersect at the transcription factor XBP-1
.
Nat Immunol
2003
;
4
:
321
9
.
11.
Zhang
K
,
Wong
HN
,
Song
B
,
Miller
CN
,
Scheuner
D
,
Kaufman
RJ
. 
The unfolded protein response sensor IRE1alpha is required at 2 distinct steps in B cell lymphopoiesis
.
J Clin Invest
2005
;
115
:
268
81
.
12.
McGehee
AM
,
Dougan
SK
,
Klemm
EJ
,
Shui
G
,
Park
B
,
Kim
YM
, et al
XBP-1-deficient plasmablasts show normal protein folding but altered glycosylation and lipid synthesis
.
J Immunol
2009
;
183
:
3690
9
.
13.
Hu
CC
,
Dougan
SK
,
McGehee
AM
,
Love
JC
,
Ploegh
HL
. 
XBP-1 regulates signal transduction, transcription factors and bone marrow colonization in B cells
.
Embo J
2009
;
28
:
1624
36
.
14.
Benhamron
S
,
Hadar
R
,
Iwawaky
T
,
So
JS
,
Lee
AH
,
Tirosh
B
. 
Regulated IRE1-dependent decay participates in curtailing immunoglobulin secretion from plasma cells
.
Eur J Immunol
2014
;
44
:
867
76
.
15.
Tang
CH
,
Chang
S
,
Paton
AW
,
Paton
JC
,
Gabrilovich
DI
,
Ploegh
HL
, et al
Phosphorylation of IRE1 at S729 regulates RIDD in B cells and antibody production after immunization
.
J Cell Biol
2018
;
217
:
1739
55
.
16.
Carrasco
DR
,
Sukhdeo
K
,
Protopopova
M
,
Sinha
R
,
Enos
M
,
Carrasco
DE
, et al
The differentiation and stress response factor XBP-1 drives multiple myeloma pathogenesis
.
Cancer Cell
2007
;
11
:
349
60
.
17.
Kriss
CL
,
Pinilla-Ibarz
JA
,
Mailloux
AW
,
Powers
JJ
,
Tang
CH
,
Kang
CW
, et al
Overexpression of TCL1 activates the endoplasmic reticulum stress response: a novel mechanism of leukemic progression in mice
.
Blood
2012
;
120
:
1027
38
.
18.
Bichi
R
,
Shinton
SA
,
Martin
ES
,
Koval
A
,
Calin
GA
,
Cesari
R
, et al
Human chronic lymphocytic leukemia modeled in mouse by targeted TCL1 expression
.
Proc Natl Acad Sci U S A
2002
;
99
:
6955
60
.
19.
Tang
CH
,
Ranatunga
S
,
Kriss
CL
,
Cubitt
CL
,
Tao
J
,
Pinilla-Ibarz
JA
, et al
Inhibition of ER stress-associated IRE-1/XBP-1 pathway reduces leukemic cell survival
.
J Clin Invest
2014
;
124
:
2585
98
.
20.
Ranatunga
S
,
Tang
CH
,
Kang
CW
,
Kriss
CL
,
Kloppenburg
BJ
,
Hu
CC
, et al
Synthesis of novel tricyclic chromenone-based inhibitors of IRE-1 RNase activity
.
J Med Chem
2014
;
57
:
4289
301
.
21.
Shao
A
,
Kang
CW
,
Tang
CH
,
Cain
CF
,
Xu
Q
,
Phoumyvong
CM
, et al
Structural tailoring of a novel fluorescent IRE-1 RNase inhibitor to precisely control its activity
.
J Med Chem
2019
;
62
:
5404
13
.
22.
Xie
H
,
Tang
CH
,
Song
JH
,
Mancuso
A
,
Del Valle
JR
,
Cao
J
, et al
IRE1alpha RNase-dependent lipid homeostasis promotes survival in Myc-transformed cancers
.
J Clin Invest
2018
;
128
:
1300
16
.
23.
Tang
CHA
,
Chang
S
,
Hashimoto
A
,
Chen
YJ
,
Kang
CW
,
Mato
AR
, et al
Secretory IgM exacerbates tumor progression by inducing accumulations of MDSCs in mice
.
Cancer Immunol Res
2018
;
6
:
696
710
24.
Korennykh
AV
,
Egea
PF
,
Korostelev
AA
,
Finer-Moore
J
,
Zhang
C
,
Shokat
KM
, et al
The unfolded protein response signals through high-order assembly of Ire1
.
Nature
2009
;
457
:
687
93
.
25.
Ali
MM
,
Bagratuni
T
,
Davenport
EL
,
Nowak
PR
,
Silva-Santisteban
MC
,
Hardcastle
A
, et al
Structure of the Ire1 autophosphorylation complex and implications for the unfolded protein response
.
EMBO J
2011
;
30
:
894
905
.
26.
Concha
NO
,
Smallwood
A
,
Bonnette
W
,
Totoritis
R
,
Zhang
G
,
Federowicz
K
, et al
Long-range inhibitor-induced conformational regulation of human IRE1alpha endoribonuclease activity
.
Mol Pharmacol
2015
;
88
:
1011
23
.
27.
de Brito
LR
,
Batey
MA
,
Zhao
Y
,
Squires
MS
,
Maitland
H
,
Leung
HY
, et al
Comparative pre-clinical evaluation of receptor tyrosine kinase inhibitors for the treatment of multiple myeloma
.
Leuk Res
2011
;
35
:
1233
40
.
28.
Dai
Y
,
Landowski
TH
,
Rosen
ST
,
Dent
P
,
Grant
S
. 
Combined treatment with the checkpoint abrogator UCN-01 and MEK1/2 inhibitors potently induces apoptosis in drug-sensitive and -resistant myeloma cells through an IL-6-independent mechanism
.
Blood
2002
;
100
:
3333
43
.
29.
Dai
Y
,
Pei
XY
,
Rahmani
M
,
Conrad
DH
,
Dent
P
,
Grant
S
. 
Interruption of the NF-kappa B pathway by Bay 11-7082 promotes UCN-01-mediated mitochondrial dysfunction and apoptosis in human multiple myeloma cells
.
Blood
2004
;
103
:
2761
70
.
30.
Pei
XY
,
Dai
Y
,
Rahmani
M
,
Li
WQ
,
Dent
P
,
Grant
S
. 
The farnesyltransferase inhibitor L744832 potentiates UCN-01-induced apoptosis in human multiple myeloma cells
.
Clin Cancer Res
2005
;
11
:
4589
600
.
31.
Ghosh
R
,
Wang
L
,
Wang
ES
,
Perera
BG
,
Igbaria
A
,
Morita
S
, et al
Allosteric inhibition of the IRE1alpha RNase preserves cell viability and function during endoplasmic reticulum stress
.
Cell
2014
;
158
:
534
48
.
32.
Harrington
PE
,
Biswas
K
,
Malwitz
D
,
Tasker
AS
,
Mohr
C
,
Andrews
KL
, et al
Unfolded protein response in cancer: IRE1alpha inhibition by selective kinase ligands does not impair tumor cell viability
.
ACS Med Chem Lett
2015
;
6
:
68
72
.
33.
Morita
S
,
Villalta
SA
,
Feldman
HC
,
Register
AC
,
Rosenthal
W
,
Hoffmann-Petersen
IT
, et al
Targeting ABL-IRE1 alpha Signaling spares ER-stressed pancreatic beta cells to reverse autoimmune diabetes
.
Cell Metab
2017
;
25
:
883
.
34.
Wilhelm
T
,
Bick
F
,
Peters
K
,
Mohta
V
,
Tirosh
B
,
Patterson
JB
, et al
Infliction of proteotoxic stresses by impairment of the unfolded protein response or proteasomal inhibition as a therapeutic strategy for mast cell leukemia
.
Oncotarget
2018
;
9
:
2984
3000
.
35.
Mahameed
M
,
Wilhelm
T
,
Darawshi
O
,
Obiedat
A
,
Tommy
WS
,
Chintha
C
, et al
The unfolded protein response modulators GSK2606414 and KIRA6 are potent KIT inhibitors
.
Cell Death Dis
2019
;
10
:
300
.
36.
Harnoss
JM
,
Le Thomas
A
,
Shemorry
A
,
Marsters
SA
,
Lawrence
DA
,
Lu
M
, et al
Disruption of IRE1alpha through its kinase domain attenuates multiple myeloma
.
Proc Natl Acad Sci U S A
2019
;
116
:
16420
9
.
37.
Jiang
D
,
Tam
AB
,
Alagappan
M
,
Hay
MP
,
Gupta
A
,
Kozak
MM
, et al
Acridine derivatives as inhibitors of the IRE1alpha-XBP1 pathway are cytotoxic to human multiple myeloma
.
Mol Cancer Ther
2016
;
15
:
2055
65
.
38.
Ri
M
,
Tashiro
E
,
Oikawa
D
,
Shinjo
S
,
Tokuda
M
,
Yokouchi
Y
, et al
Identification of toyocamycin, an agent cytotoxic for multiple myeloma cells, as a potent inhibitor of ER stress-induced XBP1 mRNA splicing
.
Blood Cancer J
2012
;
2
:
e79
.
39.
Kim
EJ
,
Bhuniya
S
,
Lee
H
,
Kim
HM
,
Cheong
C
,
Maiti
S
, et al
An activatable prodrug for the treatment of metastatic tumors
.
J Am Chem Soc
2014
;
136
:
13888
94
.
40.
Van de Bittner
GC
,
Dubikovskaya
EA
,
Bertozzi
CR
,
Chang
CJ
. 
In vivo imaging of hydrogen peroxide production in a murine tumor model with a chemoselective bioluminescent reporter
.
Proc Natl Acad Sci U S A
2010
;
107
:
21316
21
.
41.
Liu
TZ
,
Hu
CCA
,
Chen
YH
,
Stern
A
,
Cheng
JT
. 
Differentiation status modulates transcription factor NF-kappa B activity in unstimulated human hepatocellular carcinoma cell lines
.
Cancer Lett
2000
;
151
:
49
56
.
42.
Chen
WB
,
Balakrishnan
K
,
Kuang
YY
,
Han
YY
,
Fu
M
,
Gandhi
V
, et al
Reactive oxygen species (ROS) inducible DNA cross-linking agents and their effect on cancer cells and normal lymphocytes
.
J Med Chem
2014
;
57
:
4498
510
.
43.
Sen
S
,
Kawahara
B
,
Chaudhuri
G
. 
Maintenance of higher H2O2 levels, and its mechanism of action to induce growth in breast cancer cells: important roles of bioactive catalase and PP2A
.
Free Radic Biol Med
2012
;
53
:
1541
51
.
44.
Liao
Y
,
Xu
LP
,
Ou
SY
,
Edwards
H
,
Luedtke
D
,
Ge
YB
, et al
H2O2/peroxynitrite-activated hydroxamic acid HDAC inhibitor prodrugs show antileukemic activities against AML cells
.
ACS Med Chem Lett
2018
;
9
:
635
40
.