Protein Kinase CK2 Inhibition Down Modulates the NF-κB and STAT3 Survival Pathways, Enhances the Cellular Proteotoxic Stress and Synergistically Boosts the Cytotoxic Effect of Bortezomib on Multiple Myeloma and Mantle Cell Lymphoma Cells

CK2 is a pivotal pro-survival protein kinase in multiple myeloma that may likely impinge on bortezomib-regulated cellular pathways. In the present study, we investigated CK2 expression in multiple myeloma and mantle cell lymphoma, two bortezomib-responsive B cell tumors, as well as its involvement in bortezomib-induced cytotoxicity and signaling cascades potentially mediating bortezomib resistance. In both tumors, CK2 expression correlated with that of its activated targets NF-κB and STAT3 transcription factors. Bortezomib-induced proliferation arrest and apoptosis were significantly amplified by the simultaneous inhibition of CK2 with two inhibitors (CX-4945 and K27) in multiple myeloma and mantle cell lymphoma cell lines, in a model of multiple myeloma bone marrow microenvironment and in cells isolated from patients. CK2 inhibition empowered bortezomib-triggered mitochondrial-dependent cell death. Phosphorylation of NF-κB p65 on Ser529 (a CK2 target site) and rise of the levels of the endoplasmic reticulum stress kinase/endoribonuclease Ire1α were markedly reduced upon CK2 inhibition, as were STAT3 phospho Ser727 levels. On the contrary, CK2 inhibition increased phospho Ser51 eIF2α levels and enhanced the bortezomib-dependent accumulation of poly-ubiquitylated proteins and of the proteotoxic stress-associated chaperone Hsp70. Our data suggest that CK2 over expression in multiple myeloma and mantle cell lymphoma cells might sustain survival signaling cascades and can antagonize bortezomib-induced apoptosis at different levels. CK2 inhibitors could be useful in bortezomib-based combination therapies.


Introduction
Bortezomib, a boronic acid compound targeting the chymotrypsin-like activity of the 26S subunit of the proteasome, is a firstin class proteasome inhibitor (PI) [1], which has demonstrated remarkable activity against multiple myeloma (MM) and mantle cell lymphoma (MCL), two yet incurable hematologic malignancies [2,3,4]. At present, bortezomib-based combination therapies, incorporating both traditional chemotherapeutic drugs and novel agents, represent the standard care in MM and in MCL non Hodgkin Lymphomas [5,6,7,8].
The mechanisms of bortezomib-induced apoptosis are only partially known. Initial findings described that it can affect the activation of the canonical NF-kB pathway because of the induced stabilization of IkBa, the physiological NF-kB inhibitor [9]. However, recent studies have demonstrated that bortezomib can also trigger NF-kB activity in MM cells [10]. However, bortezomib may also induce many other effects. For instance, it stabilizes the tumor suppressor p53 and the pro-apoptotic protein Bax and up regulates the proteins Noxa and Puma [11], while it induces cleavage and inactivation of the anti-apoptotic molecule Mcl1 [12,13], thereby causing the activation of the mitochondriadependent apoptosis. Bortezomib can also induce endoplasmic reticulum (ER) stress, which is a mechanism of critical importance for MM plasma cell survival due to chronic ER loading with a burden of perpetually synthesized immunoglobulins [14,15]. A terminal pro-apoptotic unfolded protein response (UPR) is elicited [16] as a result of bortezomib treatment.
CK2 is a multifaceted Serine/Threonine kinase involved in several cellular processes and over-expressed and over-active in many solid and blood tumors [17,18]. A number of studies have shown that CK2 over-expression may force the cell to acquire a pro-survival program through the direct or indirect regulation of critical molecules or signaling cascades [17,19]. Interestingly, CK2 plays a central role in the activation of many cellular protein kinases by direct regulation of the activity of the chaperone complex formed by the molecules Cdc37 and Hsp90 [20,21]. CK2 also regulates signaling cascades and molecules that are targeted by bortezomib. For instance, CK2 modulates IkBa protein turnover [22,23], p53 function [24,25], AKT activation [26] and the ER stress/UPR [27,28,29,30]. We previously described that CK2 supports MM cell survival and its inhibition enhances the cytotoxic effect of both conventional chemotherapeutic agents such as melphalan [31], as well as of novel agents targeting Hsp90, both in vitro and in vivo [29]. Interestingly, CK2 was found to impinge on the proper activation of NF-kB and STAT3 in MM cells [31]. Most importantly, a phase I clinical trial with the oral ATP-competitive CK2 inhibitor CX-4945 (Cylene Pharmaceuticals, USA) is currently ongoing in the USA on relapsed/refractory MM patients [32,33].
With the above as background, we hypothesized that CK2 could modulate the sensitivity of malignant cells to proteasome inhibitors. In the present work, we studied the effect of CK2 inhibition on bortezomib-induced cytotoxicity and evaluated the signaling pathways potentially counteracting bortezomib action in MM and MCL patients. We demonstrated that CK2 inhibitors cooperate with bortezomib in causing MM and MCL cell apoptosis by down modulating the signalling cascades of NF-kB and STAT3 and by potentiating the proteotoxic effects due to proteasome blockage. The data offer the rationale for the use of CK2 inhibitors in bortezomib-based combination therapies in these malignancies.

Ethics statement
Samples and biopsies from normal bone marrows, 26 MCL, 5 monoclonal gammopathy of undetermined significance (MGUS)   3 and CK2 low if the sum was , 3; MIPI = mantle cell international prognostic index according to (44)

Patients and cell cultures
Malignant CD138 + plasma cells were purified using the Rosette Sep Kit according to the Manufacturer's protocol (Stem cell Technologies, USA). Normal peripheral blood mononuclear cells (PBMC), MM cell lines RPMI 8226, U-266 and INA-6, the human stromal cell line HS-5 were isolated and cultured as described in [29,34]. PBMC from MCL patients were obtained from peripheral blood or bone marrow as per standard Ficoll PaqueH protocol. MM International Staging System (ISS) [35] and Mantle Cell Lymphoma International Prognostic Index (MIPI) [36] were applied to the patients population analyzed. MCL cell lines Granta-519, Jeko-1 and Rec-1 purchased from Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ, Germany) were cultured in DMEM supplemented with 10% fetal bovine serum (FBS), RPMI 1640 supplemented with 20% FBS, and RPMI supplemented with 10% FBS, respectively. Testing for Mycoplasma infection was carried at a monthly basis.

Assessment of drug concentration-effect and calculation of the combination index (CI)
The experiments were performed as described in [29].

Statistical analysis
Data obtained were evaluated for their statistical significance with the two-tail paired Student's t test or analysis of variance (ANOVA) with post-hoc corrections. Values were considered statistically significant at p values below 0.05.

CK2 is highly expressed in MM and MCL, two bortezomib sensitive blood tumors and is essential for MCL cell survival
To investigate the expression of CK2 in bortezomib-sensitive B cell tumors, the expression of CK2a (catalytic subunit) and CK2b (regulatory subunit) proteins were examined by immunohistochemistry in lymph node biopsies from MCL patients (n = 21), in normal lymphoid tissues (tonsils or reactive lymph nodes)(n = 3), in bone marrow biopsies obtained from monoclonal gammopathy of undetermined significance (MGUS) (n = 5) and MM patients (n = 17). As shown in Figure 1, panels B and C, CK2a and CK2b were found mostly expressed in the germinal centers (GC) of normal reactive lymph nodes (hematoxylin and eosin (H&E) staining in Figure 1A), while a much lower expression of both CK2 subunits was detected in the mantle zone. Conversely, the paracortical zone, in which mostly T-cells reside, showed strong and diffuse nuclear positivity for CK2. Remarkably, in lymph nodes of 62% of MCL patients in our series (13/21 cases) both CK2a and CK2b were strongly expressed in tumor cells as compared to normal residual areas and their expression overlapped the expression pattern of Cyclin D1 (Figure 1, panels D-E-F). In the BM of MGUS patients CK2a and CK2b expression was scattered in hematopoietic cells of myeloid and erythroid lineage as well as in megakaryocytes but plasma cells did not display a significant over expression of the kinase as compared to normal hematopoietic cells ( Figure 1H-I; figure 1G shows H&E staining). On the opposite, in 88% of MM patients analyzed (15/ 17 cases) CK2 was found highly expressed by malignant plasma cells in the BM (Figure 1, panels K-L; Figure 1J shows H&E staining). In the case of MGUS and MM samples, double immunostaining for CD138 and either CK2a or CK2b is also shown in supplemental Figure S1 to emphasize the expression of the kinase in non malignant (Figure S1 A,  Furthermore, by western blot analysis we found that CK2a was expressed at relatively low levels in normal B lymphocytes, while in all the MCL and MM cell lines tested it was found highly expressed ( Figure 1M). To note, the expression of CK2a was found to be higher in the Rec-1 MCL cell line, which is fairly less sensitive to bortezomib [40]. The features of the MCL and MM cell lines used are listed in supplemental Table 4 and 5.

CK2 inhibitors leads to MCL cell apoptosis
Next, in order to see whether -alike to MM cells -CK2 could control MCL survival, we assessed the effects of the clinically tested ATP-competitive CK2 inhibitor CX-4945 on MCL cells isolated from five MCL patients. As shown in Figure 1N, CX-4945 caused an increase of apoptotic annexin V-expressing CD19 + MCL B cells.
Altogether, these data indicate that protein kinase CK2 is over expressed in a substantial fraction of MCL and MM, both in tissue biopsies and in cell lines, and that MCL is another B-cell malignancy, whose growth could be regulated by CK2.

Inhibition of CK2 in MM and MCL cells empowers bortezomib-induced cell proliferation in a synergistic mode
Then, we investigated whether CK2 inhibition could affect the cytotoxic effects of bortezomib on MM and MCL cell lines. Employing CX-4945 and the previously described tTBB-derivative K27 [41], a first set of experiments tested whether the simultaneous inhibition of CK2 and the proteasome could cause either an additive or a synergic effect in terms of cell growth arrest. To this aim, we performed 3 H-thymidine incorporation assays evaluating the rate of cell proliferation at increasing concentration of bortezomib (range: 1-30 nM), K27 (range: 1-15 mM) and CX-4945 (range: 1-40 mM) and the combination of bortezomib either Table 3. Summary of CK2a, CK2b and STAT3 positivity scores in CD138+ plasma cells of MGUS and Multiple Myeloma cases.

MGUS (n = 5)
Multiple Myeloma (n = 17) S c o r e * 0 1 with K27 or CX-4945. The results were analyzed to obtain the IC 50 for the three agents and the constant ratio drug combination assay was performed, giving the combination indexes (CI) according to the method described in [29]. The results showed that treatment of MM cells ( Figure 3D) also demonstrated a cooperation between CK2 inhibitors and bortezomib in inducing cell apoptosis (p,0.05, n = 3-6). The effectiveness of HS-5 bone marrow stromal cells to induce INA-6 cell resistance to chemotherapy was confirmed in experiments employing doxorubicin as a control (Figure S2 A). Moreover, similar results on the enhanced bortezomib cytotoxicity upon CK2 inhibition were obtained with the other CK2 inhibitor K27 (Figure S2 B). To note, the enhanced apoptotic effect of the combination of bortezomib and CK2 inhibitors was seen on all the MCL cell lines used, irrespective of their different sensitivity to bortezomib [40]. Significantly, the analysis of the effects of this treatment on normal B lymphocytes obtained from healthy donors showed that the exposure to bortezomib, K27, CX-4945 or the combination of the drugs induced a much lower rate of apoptosis, which was never cooperative, indicating little toxicity on normal cells (p.0.05, n = 5-7, Figure S2 C). Lastly, to further validate these findings, INA-6 and Rec-1 cells were subjected to the intracellular ATP generation assay, which measures cell proliferation and viability. As shown in Figure S2 D, both cell lines were greatly impaired in their capability to produce ATP upon treatment with bortezomib, K27, CX-4945. Most importantly, the combination of the two inhibitors with bortezomib caused a stronger reduction of ATP production (p,0.05, n = 3 for both cell lines).
Altogether, these data clearly indicate that CK2 inhibitors and bortezomib cooperate in inducing MCL and MM cell apoptosis.

Combined treatment of CK2 inhibitors and bortezomib causes mitochondrial apoptosis of MM and MCL cells
To investigate the mechanisms leading MM and MCL cells to apoptosis upon exposure to CK2 inhibitors and bortezomib, we analyzed mitochondrial membrane potential by flow cytometry analysis of JC-1 dye [34] as well as Bcl2 family members expression by WB. As shown in Figure 4A and B, MM and MCL cell lines displayed an accumulation of JC-1 green upon the different treatments. The association of bortezomib with CK2 inhibitors caused a stronger mitochondrial membrane potential depolarization as compared to the single treatments (p,0.05, n = 3-6). Again, for MM cells this effect was present when cells were grown both in suspension and on bone marrow stromal cells HS-5 (which could confer drug resistance). For MCL cells, this effect was present both in Jeko-1 and in Rec-1 cells. Bcl2 family members expression analysis revealed that treatment with bortezomib plus CK2 inhibitors caused a stronger reduction of the anti-apoptotic Bcl2 and Mcl1 proteins and a higher accumulation of the pro-apoptotic Bak and Bax proteins in MM ( Figure 4C) and MCL cells ( Figure 4D). However, in MCL cells we observed a different pattern of changes since, at variance with MM cells and according to previous studies [40], bortezomib caused a rise of Mcl1 levels that was partly antagonized by the combination with CK2 inhibitors. In Granta-519 the most evident changes were for Bcl2 and Mcl1, while Bax and Bak seemed not to vary significantly. To note, in the case of Rec-1, a less bortezomibsensitive cell line, apoptosis was achieved even by combining subapoptotic concentrations of bortezomib (15 nM) with CK2 inhibitors (as judged by JC-1 and PARP cleavage experiments and H3-thymidine incorporation, Figure 4B and 4D).
In sum, these results on changes in Bcl2 family proteins (even if somewhat variable among the various cell types) -taken together with all the other findings described above (H3-thymidine incorporation, PAPR cleavage analysis, annexin V staining, ATP generation assay, JC-1 staining) -clearly indicate that CK2 inhibitors and bortezomib strongly cooperate to destroy the mitochondrial homeostasis and the balance of Bcl2 family members, causing a shift towards the pro-apoptotic cascade.

Bortezomib triggers CK2-dependent Cdc37 and NF-kB p65 phosphorylation
Since CK2 is a stress-responsive kinase [29], which protects from apoptosis, we asked whether it could be downstream from bortezomib-induced cellular stress. Thus, to check whether bortezomib affects CK2 activity, we treated MM and MCL cells with increasing concentrations of the proteasome inhibitor and assessed by western blot analysis the CK2-dependent phosphorylation of two known CK2 targets, i.e. the Hsp90 co-chaperone Cdc37 on Ser13 [20] and the NF-kB member p65 on Ser529 [42,43]. As shown in Figure S3, bortezomib treatment for 8 hours at increasing concentrations (1, 5 or 10 nM) did not affect the high levels of Cdc37 Ser13 phosphorylation already present in MM cells; however, it caused a dose-dependent increase of NF-kB p65 Ser529 phosphorylation in all the three MM cell lines tested. On the opposite, in MCL cell lines Granta-519 and Jeko-1 we observed lower basal levels of phospho Ser13 Cdc37 but a dose-dependent increase upon bortezomib treatment; differently, in the MCL cell line Rec-1, the levels of phospho Ser13 Cdc37 werelike in MM cells -higher and not inducible. At variance with MM cell lines, phospho Ser529 p65 was inducible by bortezomib only in Granta-519, less in Jeko-1 but not at all in Rec-1 cells. Therefore, it is conceivable that this kinase is variably constitutive and active also on Cdc37 and NF-kB p65 in MM and MCL cells and can be increased by bortezomib.
CK2 inhibition diminishes the constitutive STAT3 Ser727 phosphorylation and NF-kB activation in MM and MCL cells CK2 activates survival pathways, which could be pivotal for the growth of both MM and MCL cells, namely the NF-kB and STAT3 cascades [44,45,46,47,48]. To check the STAT3 activation status in our patients' series, we analyzed the expression of phospho Ser727 STAT3 by immunohistochemistry in the same samples as in Figure 1. As shown in Figure 5A-D, accordingly to previous work cited above, we found a strong expression of nuclear phospho Ser727 STAT3 in MCL and MM as compared to normal lymphoid tissue or MGUS (see Tables 1 and 2 Table 3. Interestingly, 60% of the CD138+ MGUS plasma cells displayed a low staining score (0-1) for nuclear phospho Ser 727 STAT3, while 76% of CD138+ MM plasma cells displayed higher scores (2-3). Thus, in MCL and MM, similarly to phospho Ser529 NF-kB p65 ( Figure S3), another CK2 target, i.e. phospho Ser727 STAT3, is highly expressed. Hence, we tested the effects of CK2 inhibitors, bortezomib and the combination of the two drugs, on STAT3 and NF-kB phosphorylation in MM and MCL cells. In MM cells, CK2 inhibition caused a strong down regulation of phospho Ser727 STAT3, both in freshly isolated malignant plasma cells from MM patients ( Figure 5E) and in U-266 cells or INA-6 cells grown alone or on HS-5 stromal cells ( Figure 5F, G). Also, in U-266 cells the bortezomib-induced up regulation of phospho Ser529 p65 was markedly down regulated by the combination with CK2 inhibitors (Figure 5F, G). In MCL cells, these experiments revealed comparable, though more variable, effects of CK2 inhibitors on these pro-survival signaling pathways, since CX-4945 and K27 caused a marked reduction of phospho Ser727 STAT3 and phospho Ser529 p65 in all the three MCL cell lines, which appeared milder in Granta-519 cells ( Figure 5H). Altogether, these results clearly suggest that CK2 controls the extent of NF-kB and STAT3 activation in MM and MCL cells.

CK2 down regulation by RNA interference causes higher rate of bortezomib induced MM cell apoptosis
To strengthen the findings obtained with the chemical inhibitors, we also performed RNA interference experiments against CK2a and CK2b in INA-6 cells. Cells were transfected with siGLO Green scrambled non-specific siRNAs or siGLO Green plus CK2a and CK2b-directed siRNAs and were left untreated or were treated with bortezomib (5 nM) for 18 hours. Cellular apoptosis evaluated by annexin V staining and FACS analysis demonstrated that, while both bortezomib and CK2 silencing caused a moderate amount of apoptosis, the combined treatment was followed by almost 1.5 to two-fold increase of the apoptotic rate compared to the single treatments ( Figure 6A). Immunoblot analysis also demonstrated that CK2 silencing caused a reduction of Bcl2 and Mcl1 protein levels and of STAT3 phosphorylation; the levels of CK2a and CK2b were significantly reduced ( Figure 6B). Thus, CK2 silencing causes INA-6 MM cell apoptosis and augments bortezomib-dependent cytotoxicity.

CK2 inhibitors suppress NF-kB/STAT3 target gene expression in MM cells
We next assessed the expression of some pro-survival, proapoptotic NF-kB/STAT3 regulated genes in MM. To reproduce the MM milieu, we performed these experiments on INA-6 cells grown in presence of HS-5 stromal cells. Cells were treated with bortezomib, CX-4945 or with the combination of the two drugs. Quantitative real time RT-PCR was performed to analyze the mRNA levels of NF-kB (IAP-2, TNFa, COX-2, IL-6, BCL2, NOS-2) and STAT3 (Cyclin D1, IL-6) target gene expression. As shown in Figure 7, bortezomib caused an up regulation of IAP-2, Cyclin D1 and TNFa mRNAs that was markedly opposed by the addition of CX-4945. Furthermore, bortezomib induced a reduction of the mRNA level of the NF-kB targets COX-2, BCL2 and NOS-2. CX-4945 strongly down regulated the expression of COX-2, IL-6, BCL2 and NOS-2. These data indicate that the addition of CK2 inhibitors to bortezomib causes a block of bortezomib-induced upregulation of some and keeps markedly down regulated other NF-kB and STAT3 targets. Combined treatment with bortezomib and CK2 inhibitors enhances the proteotoxic stress in MM and MCL cells.
Since CK2 promotes a protective ER stress/UPR by maintaining the cellular levels of Ire1a [29] and facilitates the clearance of poly-ubiquitylated (poly-Ub) proteins through the autophagic cascade by mean of its phosphorylation of the docking protein p62 [49], we hypothesized that CK2 could influence the cell survival outcome upon bortezomib-induced proteotoxic stress [50]. To check if bortezomib and CK2 inhibitors could interact in the ER stress pathways and proteotoxicity, we performed WB analysis of Ire1a, total and phospho Ser51 eIF2a and poly-Ub proteins in MM cells upon exposure to bortezomib, CK2 inhibitors or the combination of the two drugs. As shown in Figure 8A,B bortezomib determined a rise of Ire1a in U-266 and Granta-519 cells and a reduction of phospho Ser51 eIF2a in Granta-519. Remarkably, the exposure to CX-4945 led to a down modulation of these effects causing a reduction of Ire1a and an increase of phospho Ser51 eIF2aprotein levels. Moreover, bortezomib led to an accumulation of poly-Ub proteins in INA-6, Granta-519 and Rec-1 cells (Figure 8 C, D, E), as shown by the appearance of a smeared signal upon probing with an anti-Ub antibody and caused an up regulation of the proteotoxic stress marker Hsp70. Interestingly, CK2 inhibition caused a mild increase of poly-Ub proteins as well, however, most remarkably, the accumulation of poly-Ub proteins was much higher in the samples treated with the combination of bortezomib and CK2 inhibitors. The detection of a parallel rise of Hsp70 protein levelswhich reflects the extent of cellular proteotoxic stress -also confirmed the stronger action caused by the association of bortezomib plus CK2 inhibitors.
Thus, CK2 inhibitors act by opposing bortezomib-induced UPR changes, which could represent homeostatic compensatory mechanisms, and greatly enhance the intracellular proteotoxicity.

Discussion
We showed here that protein kinase CK2 is highly expressed in MM and MCL and its inhibition enhances MM and MCL cells sensitivity to the proteasome inhibitor bortezomib. This effect is coupled with the down modulation of signaling pathways that may intersect with the mechanism of action of bortezomib. Our data provide the rationale for the use of CK2 inhibitors in the treatment of these malignancies.
Previously, we demonstrated a role for this kinase in MM cell survival [29,31]. In the present work, we showed that CK2 is over expressed also in MCL and that MCL cells are induced to apoptosis when CK2 is inhibited with small ATP-competitive compounds. Our immunohistochemical analysis is, to our knowledge, the first ever reported of CK2a and CK2b expression in normal lymphoid tissue and in a subset of human non-Hodgkin lymphomas. Intriguingly, this kinase was found strongly expressed in centroblasts and centrocytes, suggesting that it could play a role in the germinal center (GC) reaction. To note, the finding of low expression levels of CK2 in the normal mantle zone, as opposed to high expression in malignant MCL cells, strongly supports a role for CK2 in the biology of MCL. We are now extending these results in MCL to GC-derived non-Hodgkin lymphomas.
We showed that CK2 dependent phosphorylation of Cdc37 and p65 NF-kB (a mirror for CK2 activity) was inducible by bortezomib and was already high in cells, which are less sensitive to bortezomib (e.g. Rec-1, Figure S3). Therefore from a clinical standpoint, it can be speculated that CK2 over expression could represent an obstacle to a fully effective cytotoxic action of this drug.
To note, CK2 modulates MCL and MM cell survival upon proteasome inhibition: MM and MCL cells undergoing treatment with CK2 inhibitors plus bortezomib were induced to a proliferation arrest that was significantly higher than what seen in cells treated with the single agents. This effect was due to a strong synergy between the two agents, as shown by the determination of the combination indexes ( Figure 2). These findings confirm the efficacy of CK2 inhibitors as enhancers of other drugs' cytotoxicity, similarly to what we recently demonstrated using the CK2 inhibitor tTBB in association with the Hsp90 inhibitor 17-AAG. This last approach potently inhibited MM cell growth in a mouse xenograft model and provided the first in vivo evidence of an anti-myeloma effect of CK2 targeting [29]. The cooperation between CK2 inhibitors and bortezomib was confirmed in different experimental settings, including a MM microenvironment model using the stromal cells HS-5 and freshly isolated cells from MM patients (Figure 3 and S2). Most importantly, RNA interference and silencing of the kinase with siRNAs directed against CK2a and CK2b reproduced the effects obtained with the chemical inhibitors, confirming a cooperation with bortezomib in inducing cell apoptosis of INA-6 MM cells ( Figure 6A). Significantly, normal B lymphocytes, were fairly less sensitive to CK2 and proteasome inhibition ( Figure S2C). These findings were substantiated by numerous cell viability assays and evaluation of the homeostatic mitochondrial mechanism (Figure 4 and S2D). Notably, the cooperation between bortezomib and CK2 inhibitors was effective also using doses of bortezomib that are sub-apoptotic when added alone, both in U-266 MM cells and also in MCL Rec-1 cells, which are less sensitive to bortezomib [40]. Overall, these data clearly indicate that CK2 activity could somewhat lie downstream from bortezomib and that this protein kinase could partly counteract bortezomib-induced apoptosis. Also, they suggest that the less sensitive the cells to bortezomib, the higher the levels and activity of CK2 and its targets. Indeed, high levels of phosphorylated Ser529 NF-kB p65 and Ser727 STAT3 paralleled the strong expression of CK2 in MCL and MM was ( Figures S3, S4 and 5). Most importantly, CK2 inhibition obtained with chemicals caused a consistent drop of the levels of phospho NF-kB and phospho STAT3 in MM and MCL cells, with a consequent down-regulation of a set of target genes. Also, CK2directed RNA interference experiments confirmed, in INA-6 cells, that CK2 sustains the phosphorylation levels of STAT3 ( Figure  6B).
These findings are consistent with previous descriptions suggesting that CK2 might control NF-kB and STAT3 activation in MM cells and likely in other tumors [22,31]. Since bortezomib can also trigger the activation of the NF-kB pathway [10], our results are particularly meaningful in the perspective of developing strategies limiting NF-kB activation upon proteasome inhibition. Similarly, STAT3 downmodulation upon CK2 inhibition could represent an exploitable therapeutic strategy. CK2 can regulate STATs signaling by impinging on JAKs activation and by direct phosphorylation of STATs, as demonstrated for STAT1 and STAT3 [51,52,53,54]. It will be important to elucidate the exact mechanism of CK2 action on these two cascades in MM and MCL cells.
We also provided evidence for another level of regulation of the bortezomib-induced cell response by CK2, namely the ER stress/ UPR and the proteotoxic stress. Ire1a, the ER-localized sensor, is responsible for a compensatory response upon ER stress. We recently demonstrated that the levels and activity of Ire1a are influenced by CK2 [29]. Our findings emphasize the importance of this kinase in sustaining a pro-survival response not only upon ER stress or Hsp90 inhibition but also upon bortezomib treatment. Interestingly, phospho Ser51 eIF2a levels were markedly increased upon CK2 inhibition in MM and MCL cells, irrespective of the effects of bortezomib on this branch of the UPR. Since it has been proposed that maintaining high the levels of phospo Ser51 eIF2a could potentiate the cytotoxic mechanism of bortezomib [55], our result suggest the use of CK2 inhibitors to achieve this goal. As a consequence of an increased uncompensated UPR, we could demonstrate that CK2 inhibition enhanced the accumulation of poly-Ub proteins and of Hsp70 upon proteasome block ( Figure 8C, D, E). Therefore, CK2 could control the proteotoxic stress response in MM cells.
In conclusion, in this work we provide evidence that CK2 inhibition could represent a powerful way to boost bortezomibmediated cell death of MM and MCL. CK2 inhibitors may simultaneously down modulate critical pro-survival pathways, producing a multisided "wounding effect" on bortezomib-treated malignant B cells. It is conceivable that the use of CK2 inhibitors might minimize the likelihood of the emergence of bortezomibresistant clones and could improve the overall response to this and other second-generation proteasome inhibitors.