Targeting protein kinase CK2 and CDK4/6 pathways with a multi-kinase inhibitor ON108110 suppresses pro-survival signaling and growth in mantle cell lymphoma and T-acute lymphoblastic leukemia

Overexpression and constitutive activation of CYCLIN D1 and Casein Kinase 2 are common features of many hematologic malignancies, including mantle cell lymphoma (MCL) and leukemias such as T-cell acute lymphoblastic leukemia (T-ALL). Although both CK2 and CDK4 inhibitors have shown promising results against these tumor types, none of these agents have achieved objective responses in the clinic as monotherapies. Because both proteins play key roles in these and other hematological malignancies, we have analyzed the therapeutic potential of ON108110, a novel dual specificity ATP-competitive inhibitor of protein kinase CK2 as well as CDK4/6 in MCL and T-ALL. We show that in cell growth inhibition assays, MCL and T-ALL cell lines exhibited increased sensitivity to ON108110 when compared to other tumor types. Treatment with ON108110 reduced the level of phosphorylated RB-family proteins. In addition, ON108110 treatment resulted in concentration dependent inhibition of PTEN phosphorylation and a concomitant decrease in PI3K-AKT signaling mediated by CK2. Accordingly, cells treated with ON108110 rapidly accumulated in the G0/G1 stage of the cell cycle as a function of increasing concentration followed by rapid onset of apoptosis. Together, these results indicate that dual inhibition of CK2 and CDK4/6 may be an efficient treatment of MCL and T-ALLs displaying upregulation of CK2/PI3K and CDK4 signaling pathways.


INTRODUCTION
Mantle cell lymphoma (MCL) is a well-defined and aggressive B-cell non-Hodgkin's lymphoma (B-NHL) that is genetically characterized by a t(11;14)(q13;q32) chromosomal translocation which results in constitutive over-expression of CYCLIN D1 [1][2][3]. CYCLIN D1 overexpression is considered to be the hallmark of MCL, and is closely correlated with the proliferative rate of these malignant cells [4,5]. During cell cycle progression, CYCLIN D1 complexes with its catalytic partners, cyclindependent kinases (CDKs) 4 and 6, driving retinoblastoma www.oncotarget.com Oncotarget, 2018, Vol. 9, (No. 102), pp: 37753-37765 Research Paper www.oncotarget.com (RB) phosphorylation and progression through the G 1 phase reviewed in [6]. Although CYCLIN D1 may have CDK-independent effects, nearly all CYCLIN D1 is found in complex with CDK4 in MCL cell lines and primary samples and highlights the importance of CDK4/6 in MCL transformation [7]. In addition to these genetic alterations, MCLs also exhibit dysregulation of other signaling pathways involved in cell proliferation and survival. To this end, loss of expression or inactivation of PTEN, a negative regulator of PI3 Kinase, have been reported in some MCLs, resulting in constitutive activation of AKT [8]. Importantly, casein kinase 2 (CK2), which also regulates the PTEN/PI3K/AKT pathway, is overexpressed in lymphoproliferative disorders, including MCL [9].
Similar to MCL, T-cell Acute Lymphoblastic Leukemia (T-ALL) is an aggressive hematologic malignancy resulting from the malignant transformation of T-cell progenitors. T-ALL accounts for 10-15% and 25% of pediatric and adult ALL cases, respectively reviewed in [10,11]. PTEN inactivation and subsequent PI3K/AKT activation are common in this tumor type and are associated with CK2 mediated phosphorylation of the PTEN C-terminal tail [12]. CYCLIN D overexpression is also common in T-ALL, with the expression of particular CYCLIN D isoforms associated with distinct T-ALL subsets [13]. Importantly, the CYCLIN D3:CDK4/6 complex has been shown to have unique functions in the expansion of normally developing T cell progenitors and the induction of T cell leukemia [14,15]. Furthermore, inhibition of CYCLIN D:CDK4/6 activity has been shown to abrogate proliferation in T-ALL cell lines and primary human cells as well as progression of disease in animal models of T-ALL [14,15].
Owing to the near complete dysfunction of the RB pathway resulting from CYCLIN D over-expression, as well as CK2-mediated activation of the PI3K/AKT prosurvival pathway in the majority of MCLs and T-ALLs, we sought to determine whether dual targeting of these pathways might be an effective way by which to treat these malignancies. Towards this goal, we recently developed a dual inhibitor of CK2 and CDK4/6, ON108110, and examined its effects on the growth and apoptosis of MCL and T-ALL cell lines. X-ray crystallography studies combined with molecular modeling reveal that ON108110 binds to the active site of CK2 with high affinity and that it potently inhibits CK2 and CDK4/6 kinase activities. Furthermore, treatment of MCL and T-ALL cell lines with this compound effectively inhibits key signaling pathways mediated by these proteins, resulting in their apoptosis.

Derivation of a small molecule inhibitor of MCL and T-ALL cell growth
As a part of this study, we developed a library of a novel ATP mimetic chemotype, the benzo thiazolidinnes, that contain a core thiazole structure and have synthesized approximately 300 compounds containing the basic backbone shown in Figure 1A. A primary screen aimed at determining the growth inhibitory (GI 50 ) value against a panel of solid tumor and leukemic cell lines revealed that one of these compounds, ON108110 ( Figure 1B), exhibited a 10-20-fold higher cytotoxicity against MCLs and T-ALLs compared to other tumor types (Table 1). Importantly and in contrast to its potent anti-cancer activity, ON108110 was found to exhibit little or no cytotoxicity towards normal human cells (Table 1). Subsequent testing of this compound for inhibitory activity against a panel of 285 functional kinases revealed that ON108110 is a multi-kinase inhibitor, with highest inhibitory activity against CDK4/6 and CK2 (Table 2), two kinases intimately associated with growth, survival and anti-apoptotic phenotypes of human tumor cells.

Inhibition of CDK4/6 and CK2 kinase activity by ON108110
To confirm the that CDK4/6 and CK2 catalytic subunits were indeed the targets of ON108110, we independently tested its inhibitory activity in in vitro kinase assays using recombinant CDK4, CDK6, CK2ɑ1 and CK2ɑ2 ( Figure 1). Our results showed that ON108110 is a potent inhibitor of both CDK4 and CDK6, with IC 50 values of 6 and 10nM, respectively ( Figure 1C). Palbociclib, an FDA-approved CDK4 inhibitor which showed similar inhibition at IC 50 value of 6nM (data not shown) was used as a positive control. In addition to CDK4/6, ON108110 potently inhibited kinase activity of both CK2 catalytic subunits CK2ɑ1 and CK2ɑ2 with IC 50 values of 3 and 2nM, respectively ( Figure 1D). Silmitasertib (CX-4945), a commercially available CK2 inhibitor, inhibited these subunits with similar IC 50 values (data not shown).

ON108110 binds to protein kinase CK2
To determine the affinity and structural basis of ON108110-mediated inhibition and interaction with CK2, we performed differential scanning flourimetry (DSF) using recombinant CK2ɑ1 kinase domain. While the CK2ɑ1 apo protein showed a melting temperature (T m ) of 58°C, incubation of this protein with ON108110 induced a 6.8 o C shift in the T m of CK2ɑ1 to 64.8°C (Figure 2A), which is reflective of binding. We next determined the binding affinity constant of ON108110 to CK2ɑ1 using microscale thermophoresis (MST) [16]. A highly purified preparation of recombinant CK2ɑ1 was fluorescently labeled according to the instructions of the manufacturer and mixed with increasing concentrations of ON108110 (0.1nM-1000nM) and subjected to MST. The K d values obtained from this analysis showed that ON108110 binds to CK2 with a K d of 2.5nM ( Figure 2B), demonstrating high affinity binding. www.oncotarget.com X-ray Crystallographic Structure of the Nuclear Magnetic Resonance Analysis of the CK2ɑ1-

ON108110 interaction
To further understand the structural basis of CK2 inhibition, we performed crystallographic studies with CK2ɑ1 in complex with ON108110. The co-crystal structure, which resolved at 1.75Å, revealed that ON108110 tightly binds in the CK2 active site pocket and that the compound established hydrogen bonding with the critical amino acids in the protein ( Figure  2C). Interestingly, the O3 and N4 of the drug take the positions of the O6 and N1 groups of GMP-PNP, respectively, and maintain the same interactions with the backbone atoms of His115 and Val116. Moreover, the NO2 group of the drug is near the position that is occupied by the alpha phosphate of ATP or GTP and interacts by establishing a hydrogen bond with the side chain of Lys68 ( Figure 2D).

Prediction of ON108110 binding to the CDK6 kinase domain
To gain an understanding of the possible mode by which ON108110 binds to the ATP binding pocket of CDK4/6, molecular docking and energy minimization studies were conducted using the X-ray co-crystal structure of CDK6-VCYCLIN-PD0332991 as reference [17]. Binding of ON108110 to the CDK6 ATP binding site, like palbociclib, appears to be stabilized by various hydrogen bonds and multiple van der Waals interactions ( Figure 3A). Specifically, two hydrogen bonds are predicted to form between the benzothiazinone N4 and O3 with the alpha chain of hinge residue Val101 ( Figure 3B). Other hydrogen bonds are also predicted to form between the nitro and hydroxyl groups at positions 4 and 3 of the benzylidene moiety and Asn150 and of Asp163 ( Figure 3B). Some of these hydrogen bonds are equivalent to those observed for palbociclib in the co-crystal structure. Overall, ON108110 is predicted to be binding tightly in the ATP binding site of CDK4/6, stabilized by numerous interactions, which explains the high affinity and inhibitory activity observed for this molecule towards CDK4/6.

CK2 and CDK4/RB are overexpressed in MCL and T-ALL
Based on the targets of ON108110 and the increased sensitivity of MCL and T-ALL cell lines to the antiproliferative effects of this compound, we next examined the mRNA and protein levels of CK2 subunits and the phosphorylation status of target pRB in a panel of MCL and T-ALL cell lines. Figures 4A & 4B show that both MCL and T-ALL cell lines expressed significantly higher levels of CKSN2A1 (CK2α1) and CKSN2A2 (CK2α2) transcripts as well as protein subunits compared to normal peripheral blood mononuclear cells (PBMCs). In addition, western blot studies showed constitutive phosphorylation of pRB which is not seen in PBMCs. pRB and CK2 levels correlated with the sensitivity of these cell lines to the effects of ON108110 treatment as evidenced by their IC 50 values (>100-fold more sensitive) when compared to that of PMBCs in vitro ( Figure 4C).

Effect of ON108110 on cell cycle progression and induction of apoptosis of MCL and T-ALL cells
We next examined the effect of ON108110 treatment on cell cycle progression in MCL and T-ALL cell lines that are most sensitive to this compound, Granta-519 and CEM, respectively. For these studies, exponentially growing cells were treated with increasing concentrations of ON108110 for 24 hours. The cells were then harvested and subjected to propidium iodide staining and flow cytometric analysis. While there was no significant change in the profile of cells treated with DMSO over the course of the experiment, a rapid accumulation of the cells in the G 1 phase of the cell cycle was evident following treatment with ON108110 ( Figure 5A). However, at higher concentrations of the compound, a large proportion of the cells progressed through the S and G2/M phases and eventually began accumulating in the sub-G 1 phase, suggesting induction of apoptosis at these concentrations ( Figure. 5A). To determine whether the sub-G 1 phase cells correlated with the onset of apoptosis in cells treated with the ON108110, we analyzed the frequency of apoptotic cells using Annexin V staining. The results presented in ( Figure  5B & 5C) show that while there is no evidence of activation of apoptosis DMSO treated cells, we observed a concentration-dependent increase in the frequency of apoptotic cells 24 hours post-treatment. Caspase activity and PARP cleavage also increased as a function of cell death in ON108110-treated cells ( Figures 5D and 6). Together, these results suggest that dual CK2/CDK inhibition by ON108110 induced a potent cell cycle arrest in MCL and T-ALL cell lines that is accompanied by the induction apoptosis.

ON108110 inhibits CK2 and CDK4/6 mediated oncogenic signaling in MCL and T-ALL cell lines
It has now been well established that the RB tumor suppressor is a primary target of CDK4 and CDK6 kinases. To determine the effects of ON108110 on the CDK/pRB signaling in MCL and T-ALL cell lines, Granta-519 and CEM cells were treated with increasing concentrations of ON108110 for 24 hours. Whole cell lysates were subjected to western blot analysis using antibodies against phosphorylated RB (Ser780 and Ser 807/811). The results of this study showed that while DMSO treated cells had high levels of phosphorylated RB, cells treated with ON108110 showed CK2 has been shown to play an important role in hematopoietic malignancies, especially MCL and T-ALL [9,18]. CK2 mediated phosphorylation and inactivation of tumor suppressor PTEN activates the PI3K/AKT signaling pathway, which is important in the survival of these and other tumor types. Because ON108110 inhibits CK2 kinase activity, we next examined phosphorylation status of the PTEN and AKT proteins in cells treated with this compound. As seen in Figure 6, ON108110-treated cells expressed lower levels of phospho-PTEN Ser380 , a CK2 specific phosphorylation site that inactivates the PTEN phosphatase activity. Accordingly, we also detected a reduction in the level of AKT Ser129 , which is phosphorylated by CK2 and promotes its activity [19]. These results provide a molecular basis for the observed induction of cell death in cells treated with ON108110.

DISCUSSION
Aberrant cell cycle regulation is a common feature in most forms of cancer reviewed in [20]. In hematologic malignancies, including MCL and ALL, deregulated expression of D-type CYCLINS is frequently observed, which closely correlates with the high proliferative rate of these cells [4,5]. Although there have been many efforts to develop selective inhibitors of CDK4/CDK6, initial studies with palbociclib, an FDA-approved selective small molecule inhibitor of CDK4 and CDK6 [21,22], revealed that the drug was cytostatic when used as a monotherapy for MCL, resulting in disease stabilization with an 18% overall response rate [23]. In addition to CYCLIN D1, CK2 overexpression is a hallmark of T-ALL [18] and regulates multiple hematopoiesis-associated signaling cascades, including those involving PI3K and AKT, which are activated in MCL and T-ALL reviewed in [8]. In light of the key roles played by CYCLIN D1 and CK2 in hematological malignancies, it is not surprising that clinical trials aimed at testing small molecule inhibitors of these and other proteins in combination has gained increasing momentum. In this communication, we examined the effects of ON108110, a dual inhibitor of CDK4/6 and protein kinase CK2 on the growth and viability of MCL and T-ALL. Our results show that ON108110 potently inhibits the kinase activities of these proteins in vitro. We also provide evidence demonstrating that targeting CK2 and CDK4/6-mediated pathways in MCL and T-ALL leads to potent cell cycle arrest and the induction of apoptosis. The G 1 arrest seen with ON108110 treatment is clearly due to inhibition of phosphorylation of members of the RB family, a direct result of CDK4 and CDK6 inhibition. Furthermore, ON108110 acts as a potent inhibitor of CK2 mediated PI3K/AKT activation as measured by inhibition of CK2-specific phosphorylation sites of its substrates, PTEN Ser380 and AKT Ser129 . Given the significance of the PTEN/PI3K/AKT signaling axis in T-ALL cell survival [18], our results provide a molecular basis for the observed induction of apoptosis in ON108110 treated cells and a strong rationale for targeting CK2 and CDK4/6 in MCL and T-ALL.
To gain insight into the structural basis of selectivity of ON108110 for CK2 and CDK6 we determined high-resolution crystal structures of the CK2-ON108110 complex. Structural studies showed that ON108110 binds tightly in the active site pocket of CK2α1. Furthermore. the co-crystal structure shows that ON108110 binds in the cleft between two domains that normally accommodate ATP. Importantly, ON108110 interacts with conserved amino acids and establishes hydrogen bonding with the surrounding amino acids and water molecule. Collectively, this extensive combination of direct and water-mediated hydrogen bonds and van der Waals contacts observed between ON108110 and CK2ɑ establishes the structural basis for high affinity binding. Additionally, molecular docking and energy minimization studies with CDK6-ON108110 predict that ON108110 establishes various hydrogen bonds and multiple van der Waals interactions, which were similar to those observed with palbociclib. Collectively, ON108110 binds tightly in the ATP binding site of CDK6 and is stabilized by numerous interactions, which explains the potent inhibitory effect on CDK4/6.
In conclusion, our results indicate, that combined inhibition of CYCLIN D1/CDK4/RB and CK2/PI3K/AKT via a small molecule inhibitor of the CDK4/6 and CK2 kinases is effective in inducing cell death of MCL and T-ALL, particularly those that have aberrant upregulation of both signaling pathways.

Cell lines
Cell lines were cultured in RPMI-1640 medium (Thermo Scientific) supplemented with 10% heat inactivated FBS and penicillin-streptomycin at 37°C under humidified conditions and 5% CO 2 . All cell lines tested negative for mycoplasma. To determine GI 50 values, cells were seeded at a density of 2.5x10 3 cells-0.1ml -1 per well of a 96-well plate. Compounds were added 24 hrs postplating at the indicated concentrations. Cell counts were determined from duplicate wells 96 hrs post-treatment using the Cell Titer Blue assay in conjunction with the GloMax plate reader (Promega).
Peripheral blood samples were obtained from patients at the Icahn School of Medicine at Mount Sinai with informed and written consent for tissue banking following a protocol approved by the Institutional Review Board (MSSM IRB#11-1669) and in accordance with the Declaration of Helsinki. Peripheral blood mononuclear cells were pelleted by low-speed centrifugation, resuspended in media composed of 90% fetal bovine serum (FBS) and 10% DMSO (Sigma), frozen slowly in the vapor phase of liquid nitrogen in multiple cryovials, and stored in liquid nitrogen.

Kinase assays
Initial testing of ON108110 against a panel of 285 functional kinases was performed by Reaction Biology Corp (Malvern, PA). Recombinant CDK4/CYCLIN D1 or CK2α1 were mixed with their respective synthetic peptide substrates (10ng) and incubated with DMSO or increasing concentrations of ON108110 in 20μl kinase buffer [50mM Tris-HCl (pH 7.5), 10mM MgCl 2 , 1mM EGTA, 2mM DTT and 0.01% NP-40] for 30 min. at room temperature. Kinase reactions were carried out for 20 min at 30°C in the presence of 5μl of substrate-ATP mixture (1μg RB, 20μM ATP, 20μCi γ-32 P-ATP). The reactions were incubated at 30°C for 20 minutes, terminated by addition of 0.5M O-phosphoric acid and 10μl of the reaction mixture was spotted on P-30 Filtermat membranes (Wallac #1450-523). The filtermat was air-dried, blocked with methanol for 2 minutes followed by three washes with 75mM phosphoric acid. The filtermat was then air-dried and analyzed using a phosphoimager. Densitometric values were plotted as a function of log drug concentration using Prism 4 GraphPad software and the IC 50 values determined by plotting sigmoidal non-linear regression curves with variable slope (n=3).

RNA expression in MCL and T-ALL cell lines
For quantitative RT-PCR (qRT-PCR), total RNA was extracted using the RNeasy mini kit (Qiagen) and subjected to reverse transcription using Superscript III (Life Technologies). Q-PCR of the resulting cDNA templates was performed using Power SYBR Green Master Mix (Applied Biosystems). Expression levels of individual genes were normalized to ß-actin. Amplification of CKSN2A1 (CK2α1), CKSN2A1 (CK2α2), and β-actin was performed using Applied Biosystems proprietary primer-probe sets.

Western blot analysis
Cells were grown as indicated, lysed in lysis buffer (20mM Tris, pH 8.0/ 137mM NaCl/ 10% glycerol/ 1% NP-40) supplemented with 1X protease and phosphatase inhibitor cocktail on ice and resolved by SDS-PAGE. The proteins were transferred to a nitrocellulose membrane and subjected to Western blot analysis using the indicated antibodies. Proteins were visualized using an Odyssey imaging system (LI-COR Biosciences).

Flow cytometric analysis
To determine cell cycle distribution, 70% ethanol fixed cells were washed with phosphate buffered saline (PBS), resuspended in PBS and incubated with RNase and propidium iodide at 37°C for 20 min. DNA content was analyzed on a FACS Calibur (BD Immunocytometry Systems) and the data analyzed using Flowjo software (Treestar). Annexin V staining was measured using the FITC Annexin V Apoptosis Detection Kit I (BD Biosciences) according to the instructions of the manufacturer except that DAPI was substituted at a concentration of 1mg ml -1 to measure DNA content.

CK2α1 protein expression and purification
Amino acid residues 3-335 of CK2α1 were cloned using BamH1 and HindIII into pET28bsmt3-SUMO fusion vector, which encodes a N-Terminal His6 Tag followed by a SUMO (small ubiquitin-related modifier) protein and SUMO protease cleavage site. The nucleotide sequence was confirmed by DNA sequencing. CK2α1 was expressed in Escherichia coli BL21-DE3 (RIL), cells and lysed by sonication. Cell lysate was clarified by centrifugation at 18,000 rpm for 1 hour at 4°C. The supernatant was filtered and loaded onto a Ni-NTA column equilibrated with 25mM Tris pH-8.5, 0.5M NaCl, 20mM imidazole pH 8.0, 86 10% Glycerol and 2mM β-mercaptoethanol. The protein was eluted using elution buffer containing 25mM Tris pH8.5, 0.3M NaCl, 250mM imidazole pH 8.0 and 2mM β-mercaptoethanol. Fractions containing CK2α1 were pooled and dialyzed overnight against 25mM Tris, pH8.5, 0.3M NaCl and 2mM β-mercaptoethanol. The SUMO tag was cleaved by adding ULP1 protease (5mg ml-1) to the dialysis buffer. Cleaved CK2α1 was then passed over a second Ni2+-NTA column to collect CK2α1 and remove the His-Tag SUMO and His Tag-ULP1 protease. The flowthrough was collected, concentrated, passed through a Heparin Column and further purified using size exclusion chromatography using a Superdex 75 gel filtration column. Fractions with >95% pure CK2α1 (as monitored by SDS-PAGE electrophoresis) were collected and concentrated using an Amicon centrifugal concentrator. The purified protein was quantified aliquots were snap frozen and stored at −80°C.

CK2α1-ON108600 co-crystallization
2μl of 10 mg ml-1 purified CK2α1 protein solution was mixed with an equal volume of reservoir solution (21% PEG-5000, 0.2M ammonium sulfate, 0.1M Tris HCl and 10 mM AMP-PNP). Thin, needle-like crystals were obtained by the vapor diffusion method at 18°C. Crystals obtained from microseeding were soaked with ON108110 dissolved in DMSO (1mM). Co-crystallization was also performed by the hanging drop vapor diffusion method at 18°C. Crystals appeared over a period of 7 days. For data collection, the crystals were cryoprotected by stepwise soaking for 5 minutes in mother liquor solutions containing 5-30% (v/v) glycerol and subsequently flash frozen in liquid nitrogen. X-ray diffraction data were collected on beamline X25 at Brookhaven National Laboratory. The crystals diffracted to a 1.75-Å resolution with synchrotron radiation at the platinum LIII absorption edge (λ = 1.0722 Å). The crystals belonged to the space group P21.

Molecular docking of ON108110-CDK6
ON108110 small molecule was generated using SKETCHER (CCP4 suite) and superimposed in COOT to the ligand of the X-ray co-crystal structure of CDK6-Vcyclin-PD0332991 (2EUF). Binding positions of ON108110 were analyzed using restrained docking in FireDock and the predicted lowest energy binding was visualized and graphed using PyMOL.

Differential scanning fluorimetry
1μM of in-house purified recombinant CK2α1 was diluted in protein buffer and added to a final concentration of 1X of Sypro Orange Dye in the presence of DMSO or increasing concentrations of ON108110. Following a 20 min incubation at room temperature, 20μl samples were pipetted into a clear MicroAmp Fast 96-well Reaction PCR plates and sealed with MicroAmp Optical Adhesive Film. The assay was performed using a StepOne Plus RT-PCR instrument (Applied Biosystems) using an increasing temperature gradient of 20 to 90° C in 0.2° C increments using a 200-millisecond stabilization delay. Florescence readings were acquired with excitation and emission wavelengths of 580±10 nm and 624±14 nm respectively. A customized analysis program (Step One Plus Protein Thermal Shift Software) was used to calculate the T m from each fluorescence profile (Boltzmann method) and the T m of the fluorescence (derivative method). Derivative T m values were obtained from the top of the peak in the derivative plot, while the Boltzmann T m values calculated from the inflection point of the fluorescence plot.

Author contributions
AP designed and performed experiments, analyzed data and wrote the manuscript, OR and SCC designed and performed experiments and analyzed data, RV analyzed data, VP and VSDRC performed experiments, SJB analyzed data and wrote the manuscript, AA and MVRR designed experiments and analyzed data, EPR designed experiments, analyzed data and wrote the manuscript.

CONFLICTS OF INTEREST
EPR is an equity holder, board member and a paid consultant of Onconova Therapeutics, Inc. EPR and MVRR are named inventors on pending and issued patents filed by Temple University. MVRR is a stockholder and paid consultant of Onconova Therapeutics Inc. SJB and SCC are paid consultants of Onconova Therapeutics Inc. All other authors declare no conflicts of interest.