Constitutive Phosphorylation of STAT3 by the CK2–BLNK–CD5 Complex Free

Chronic lymphocytic leukemia (CLL), the most common leukemia in the Western Hemisphere, is characterized by a gradual accumulation of mature-appearing lymphocytes coexpressing typical B-cell surface antigens (1) and CD5. CD5 is usually expressed on T cells (2), although only on a rare B-cell subtype, but not on most B cells (3). CLL cells are also characterized by constitutive activation of the STAT3 (4, 5).

Typically, STAT3 is activated by extracellular molecules such as cytokines and growth factors that bind to their corresponding receptors and induce the phosphorylation of STAT3 on tyrosine 705 residues. This phosphorylated (p) STAT3 forms dimers, translocates to the nucleus, binds to DNA, and activates STAT3-regulated genes (6). STAT3 plays a key role in cell growth, suppression of apoptosis, cell motility (7), tumorigenesis, and malignant transformation (8).

Unlike in normal B cells, in CLL cells, STAT3 is constitutively phosphorylated on serine 727 residues rather than tyrosine residues (4, 5). Serine pSTAT3 has biologic activities similar to those of tyrosine pSTAT3: Serine pSTAT3 is shuttled to the nucleus, binds to DNA, activates genes known to be activated by tyrosine pSTAT3, and provides CLL cells with a survival advantage (5, 9–12) and proliferation capacity (12, 13).

STAT3 is ubiquitously expressed in various cell types (14), and its phosphorylation and biologic activation are usually induced by tyrosine kinases, such as Janus kinase 2 (15). What induces the phosphorylation of STAT3 on serine residues is currently unknown. Unexpectedly, although STAT3 is constitutively phosphorylated exclusively on serine residues in CLL cells (14), several large-scale whole-exome sequencing analysis did not identify a recurrent activating mutation in a serine kinase nor inactivating mutation in a phosphatase (16, 17). Therefore, we hypothesized that a combination of several proteins uniquely assembled in CLL cells induces the phosphorylation of STAT3 on serine 727 residues.

After Institutional Review Board approval and written informed consent were obtained, peripheral blood samples were obtained from 28 patients with CLL who were treated in the Leukemia Department at The University of Texas MD Anderson Cancer Center (Houston, TX) from 2011 to 2016. The clinical characteristics of all the patients are summarized in Supplementary Table S1.

To isolate low-density cells, peripheral blood cells were fractionated using Histopaque-1077 (Sigma-Aldrich). More than 95% of the fractionated peripheral blood lymphocytes obtained from these patients were CD19⁺/CD5⁺, as assessed by flow cytometry.

Western immunoblotting was performed as described previously (5). The following primary antibodies were used: monoclonal mouse anti-human STAT3, mouse anti-human CD5, and mouse anti-human α-catalytic subunit of casein kinase 2 (CK2; BD Biosciences) and rabbit anti-human serine pSTAT3 and rabbit anti–B-cell linker protein (BLNK; Cell Signaling Technology). Densitometry analysis was performed using an Epson Expression 1680 scanner (Epson America, Inc.). Densitometry values were normalized by dividing the numerical value of each sample signal by the numerical value of the corresponding β-actin signal, used as a loading control.

Mass spectrometry was performed as described previously (18). Briefly, the silver-stained bands from the pull-down assay were destained and subjected to in-gel digestion. The resulting peptides were analyzed by nano-liquid chromatography-coupled ion trap mass spectrophotometry. Electrospray ion trap mass spectrometry was performed on an LTQ linear ion trap mass spectrometer (Thermo Fisher Scientific). The resulting proteins were then identified by a database search for the fragment spectra using the National Center for Biotechnology Information database of nonredundant proteins. Resulting peptide matches were manually curated.

A total of 1 × 10⁶ cells/mL were placed in flasks in minimum essential medium alpha (Invitrogen) supplemented with 10% bovine calf serum (HyClone Laboratories). CD5-neutralizing antibodies (Abcam) were added at a concentration of 1 to 2,500/mL. Cells were incubated at 37°C for 16 hours and harvested for further studies.

Immunoprecipitation studies were done as described previously (19). Briefly, CLL cell lysates were incubated with anti-serine pSTAT3, anti-CK2, anti-CD5, or anti-BLNK antibodies for 16 hours at 4°C. Protein A agarose beads (EMD Millipore) were added for 2 hours at 4°C. For negative controls, the cytoplasmic lysates were incubated either with rabbit serum plus protein A agarose beads or with protein A agarose beads alone. After three washes with RIPA, the beads were suspended in SDS sample buffer, boiled for 5 minutes, and removed by centrifugation, and the supernatant proteins were separated by SDS–PAGE. Human embryonic kidney 293 (HEK293) cells, HeLa cells, Jurkat T lymphoblastic leukemia cells, and cells from the myeloid-derived lines K-562 and RAMOS were used as controls in the immunoprecipitation studies.

Nondenatured nuclear and cytoplasmic extracts of CLL cells were prepared using an NE-PER Extraction Kit (Thermo Fisher Scientific) and confirmed by Western immunoblotting–based detection of the nuclear protein lamin B and the cytoplasmic ribosomal protein S6 (5).

Recombinant STAT3 from Novus Biologicals and active CK2 from New England Biolabs were used according to the manufacturers' instructions. Briefly, 360 ng STAT3 was incubated with 100 ng CK2 in adenosine 5′-triphosphate–supplemented (200 μmol/L) kinase buffer [20 mmol/L Tris-HCl (pH 7.5), 10 mmol/L MgCl₂, 5 mmol/L dithiothreitol] for 30 minutes at 37°C. Subsequently, the reaction was stopped with 2× Laemmli buffer, and phosphorylation of STAT3 was determined using anti-serine pSTAT3 antibodies in Western immunoblotting.

After thawing in warm water, cells were washed twice with RPMI 1640 Medium (Gibco), and TRIzol (Thermo Fisher Scientific) was added. The RNA was purified using an RNeasy purification procedure (QIAGEN). RNA quality and concentration were analyzed with a spectrophotometer (ND-1000; NanoDrop Technologies).

We used 500 ng of total RNA in one-step qRT-PCR (Applied Biosystems) with an ABI PRISM 7700 sequence detection system (Applied Biosystems) using a TaqMan gene expression assay for CD5 and BLNK according to the manufacturer's instructions. Samples were run in triplicate, and relative quantification was performed by using the comparative C_t method.

Five microliters of siPORT NeoFX agent and 50 pmol of CD5 siRNA, BLNK-siRNA, or FAM-labeled siRNA-targeting human GAPDH or scrambled control (Applied Biosystems) were each diluted in 50 μL of OPTI-MEM I and then mixed together and incubated at room temperature for 10 minutes. A total of 5 × 10⁶ cells suspended in 0.2 mL of OPTI-MEM I medium containing the siRNA and transfection agents were incubated at room temperature. After 1 hour of incubation, transfections were performed by electroporation (Bio-Rad Laboratories), and then the cells were cultured in complete RPMI1640 medium. Transfection efficiency was calculated on the basis of the GFP-conjugated siRNA measured by flow cytometry (Becton, Dickinson and Company). The apoptosis rate of the evaluated transfected cells was ≤30% as assessed by Annexin V/PI using flow cytometry analysis.

CLL cells were incubated in microtubes with PBS supplemented with 2% bovine serum (HyClone) and with anti-S6 ribosomal protein antibodies (Cell Signaling Technology) and anti-CD5 antibodies (Becton, Dickinson and Company) for 1 hour. After being washed three times with PBS, the cells were suspended in 5 mg/mL solution of 4′,6-diamidino-2-phenylindole (DAPI) dye (Invitrogen) for 5 minutes and then washed in PBS to remove the unbound dye. The cells were then placed into μ-slide VI^0.4 chamber slides (ibidi, LLC) for microscopic analysis. The slides were viewed using an Olympus FluoView 500 confocal laser scanning microscope (Olympus America), and images were analyzed using the FluoView software (Olympus America).

To identify a serine kinase that induces phosphorylation of STAT3 on serine residues in CLL cells, we immunoprecipitated protein extract of CLL cells obtained from 3 patients with anti-serine pSTAT3 antibodies and analyzed the immunoprecipitated protein using mass spectrometry. One of the 635 proteins that were pulled down with anti-serine pSTAT3 antibodies was the β-regulatory subunit of the serine/threonine kinase CK2. Therefore, we sought to determine whether CK2 is the kinase that induces constitutive phosphorylation of STAT3 on serine 727 residues in CLL cells.

Although CK2 is ubiquitously expressed in mammalian cells (20), we first sought to determine whether the catalytically active subunit of CK2 is present in CLL cells. We obtained low-density peripheral blood cells from 7 patients with CLL and, using Western immunoblotting, found that the α-catalytic subunit of CK2 as well as serine pSTAT3 were readily detected in all samples (Fig. 1A).

Then, to validate the mass spectrometry results, we immunoprecipitated CLL cell extracts obtained from 2 CLL patients using anti-STAT3 antibodies. As expected, we found the α-catalytic subunit of CK2 along with phospho-serine STAT3 coimmunoprecipitated with STAT3 (Fig. 1B, top). To confirm these findings, we immunoprecipitated CLL cell extracts from 3 patients with anti-CK2 antibodies. We found that the α-catalytic subunit of CK2 coimmunoprecipitated with STAT3 and phospho-serine STAT3 (Fig. 1B, bottom), suggesting that CK2 binds to STAT3 and phospho-serine STAT3.

Because we found that the serine/threonine kinase CK2 binds to STAT3 in CLL cells, we sought to determine whether CK2 indeed phosphorylates STAT3 on serine residues. To test this hypothesis, we incubated active CK2 with recombinant human (rh) STAT3 for 30 minutes and, as shown in Fig. 1C, found that rhSTAT3 became phosphorylated on serine 727 residues when rhSTAT3 was incubated in the presence, but not in the absence, of active CK2. Then, to confirm these observations, we transfected CLL cells with a CK2-siRNA construct and found that when the levels of CK2 transcripts were significantly downregulated, protein levels of CK2 and serine pSTAT3 were markedly reduced (Fig. 1D), suggesting that CK2 induces the induction of STAT3 phosphorylation of serine 727 residues.

CK2 and STAT3 are ubiquitously expressed in mammalian cells (14, 20). However, in most cells, unphosphorylated STAT3, but not serine phosphorylated STAT3, is commonly detected, likely because CK2 has to be activated to exert its enzymatic activity. CD5 is a transmembrane glycoprotein present on the surface of CLL cells. Only a rare population of nonmalignant B cells expresses CD5 (3). Because CD5 is thought to be biologically active upon its phosphorylation, we obtained CLL cells from 8 CLL patients and found that CD5 is constitutively phosphorylated in CLL cells (Fig. 2A). Because the interaction between CK2 and CD5 induces activation of CK2 (21–24), we sought to determine whether CD5 binds CK2 in CLL cells. Therefore, we immunoprecipitated CLL cell lysates with anti-CK2 antibodies and, using Western immunoblotting, found that CD5, as well as serine pSTAT3, coimmunoprecipitated with the α-catalytic subunit of CK2 (Fig. 2B).

Anti-CD5 antibodies have been found to block the homophilic interactions of B cells and to inhibit B-cell activation, suggesting that CD5 binds to and induces CD5 (25). To determine whether CK2 requires CD5 to induce STAT3 phosphorylation in CLL cells, we first incubated CLL cells with CD5-neutralizing antibodies, and we found a significant reduction in levels of both CD5 and serine pSTAT3 (Fig. 2C). Then, to confirm this observation, we transfected CLL cells with a CD5-siRNA construct and found that when the levels of CD5 transcripts were significantly downregulated, protein levels of CD5 and serine pSTAT3 were markedly reduced (Fig. 2D), suggesting that CD5 is indeed required for the induction of STAT3 phosphorylation of serine 727 residues.

We found that CK2 and CD5 conjoin in phosphorylating STAT3 on serine 727 residues in CLL cells. However, unlike in CLL cells, STAT3 is not constitutively phosphorylated in T lymphocytes, although both CK2 and CD5 are commonly detected there (26). Therefore, we wondered whether an additional factor not present in T cells is required for the induction of STAT3 phosphorylation on serine residues in CLL cells. BLNK, also known as SLP-65, is an adaptor protein that is expressed in B cells (27) but not in T cells and was found to be required for lipopolysaccharide-induced STAT3 phosphorylation (28). To determine whether BLNK is present and activated in CLL cells, we performed a Western immunoblot analysis, and we detected BLNK and tyrosine pBLNK in CLL cells in 7 of 7 patients (Fig. 3A). Furthermore, when we immunoprecipitated CLL cell extracts with anti-BLNK antibodies, we found that BLNK coimmunoprecipitated CK2, CD5, STAT3, and serine pSTAT3 (Fig. 3B). We then transfected CLL cells with BLNK siRNA and found that when BLNK transcript levels were significantly downregulated, protein levels of BLNK decreased and that while the levels of STAT3 did not change, the levels of serine pSTAT3 were markedly reduced, suggesting that BLNK is required for serine phosphorylation of STAT3 in CLL cells (Fig. 3C).

Our data suggest that both CD5 and BLNK, uniquely present in CLL cells, bind CK2 and are required for the induction of STAT3 phosphorylation. Therefore, we hypothesized that CD5, CK2, BLNK, and STAT3 form a complex that enables the phosphorylation of STAT3 by CK2. To test this hypothesis, we immunoprecipitated CLL cell protein extracts with anti-STAT3 antibodies. Then, using Western immunoblotting, we found that STAT3 coimmunoprecipitated with CD5, CK2, BLNK, and serine pSTAT3 (Fig. 4A). To confirm these data, we immunoprecipitated CLL cell protein extracts with anti-CD5 antibodies and, using Western immunoblotting, found that CD5 coimmunoprecipitated with STAT3, CK2, BLNK, and serine pSTAT3 (Fig. 4B). Taken together, our data suggest that a protein complex consisting of CD5, CK2, and BLNK binds STAT3 and that the entire protein complex is required for the phosphorylation of STAT3 on serine 727 residues.

Because CD5 is a surface protein, we wondered whether the STAT3 phosphorylation complex is cell surface bound or whether a CD5 intracytoplasmic isoform participates in the phosphorylation complex. Using confocal microscopy of CLL cells, we found that CD5 remains membrane bound (Fig. 5A), suggesting that CD5 anchors the STAT3 phosphorylation protein complex to the cell membrane.

In a previous study, we demonstrated that after STAT3 phosphorylation, phospho-serine STAT3 forms dimers and translocates to the nucleus, where it binds to DNA (5). To determine whether any of the components of the STAT3 phosphorylation protein complex are coshuttled with serine pSTAT3 to the nucleus, we obtained CLL cell cytoplasmic and nuclear fractions and performed a Western immunoblotting analysis. While proteins from the entire complex, CD5, CK2, BLNK, STAT3, and serine pSTAT3, were detected in the cytosolic fractions, only serine pSTAT3 was detected in the nuclear extracts (Fig. 5B). Taken together, these data suggest that the STAT3 phosphorylation protein complex is anchored by CD5 to the cell membrane, and that upon phosphorylation, pSTAT3 detaches from the complex and is shuttled to the nucleus.

We and others have previously shown that in circulating CLL cells, STAT3 is constitutively phosphorylated on serine 727 residues (4, 5); however, what induces this posttranslational modification of STAT3 was largely unknown.

Here, we show that a protein complex consisting of CK2, CD5, and BLNK induces phosphorylation of STAT3 on serine 727 residues. Whereas CK2 is expressed in all hematopoietic cells, BLNK is expressed exclusively in B cells (27), and CD5, primarily present on the surface of T cells (29), is usually detected on the cellular membrane of CLL cells. Unlike normal B cells, CLL cells coexpress CD5 and BLNK (3), and these two proteins participate in the formation of a protein complex that is likely unique to CLL cells. Because pSTAT3 activates numerous pathways that provide CLL cells with a survival advantage (9, 10, 19, 30, 31), the cellular mechanism that induces STAT3 phosphorylation is a potential therapeutic target to inhibit in CLL.

The threonine/serine kinase CK2 is ubiquitously expressed in eukaryotic cells (20). However, although STAT3 is one of the approximately 300 well-established CK2 substrates (32), STAT3 is rarely phosphorylated on serine residues in unstimulated B cells, suggesting that to induce STAT3 phosphorylation, CK2 has to be activated and attach to STAT3. CD5 is known to associate with and activate CK2 (21). Our data show that in CLL cells, CD5 contributes to STAT3 phosphorylation, as transfection of CLL cells with CD5-siRNA significantly downregulated CD5 mRNA and protein levels and markedly reduced the levels of serine pSTAT3. In addition, CD5-neutralizing antibodies, known to inhibit the activation of CD5, significantly reduced CD5 and serine pSTAT3 protein levels in CLL cells, likely because CD5 and STAT3 form a feed-forward loop (33).

Both CLL cells and T cells express CD5. Yet, although in T cells CD5 activates CK2 (34), STAT3 is typically not constitutively phosphorylated, suggesting that another component, not shared by CLL cells and T cells, is required for the induction of STAT3 phosphorylation on serine residues. Like their normal B-cell counterparts, CLL cells express BLNK. BLNK is a cytoplasmic B-cell–specific protein that plays a critical role in B-cell development (35) and participates in the induction of STAT3 phosphorylation in CLL cells (36). We found that in CLL cells, BLNK is a crucial component of the STAT3 phosphorylation complex and is required for maximal STAT3 serine phosphorylation. Remarkably, BLNK is also a part of a protein complex that includes phospholipase Cγ2 and Bruton tyrosine kinase phosphorylation in CLL cells (37).

CK2 is a stable tetrameric enzyme consisting of two α-catalytic and two β-regulatory subunits (38). Recent studies suggested that some of the CK2 β-subunit functions are independent of the CK2 tetramer (39). In the search for a serine kinase that phosphorylates STAT3 in CLL cells, we used a proteomics approach and found that the β-regulatory subunit of CK2 coimmunoprecipitated with STAT3. Then, to determine whether CK2 phosphorylates STAT3, we obtained recombinant human tetrameric CK2 and found that it phosphorylated rhSTAT3 on serine 727 residues. Because in all other experiments we used anti-CK2 antibodies directed against the α-catalytic subunit or CK2-α-siRNA, we were unable to determine whether the CK2 β-subunit plays a role in the induction of STAT3 phosphorylation.

Several investigators reported that CK2 is overexpressed and activated in a variety of hematologic malignancies, such as multiple myeloma (40, 41), mantle cell lymphoma (40, 41), follicular lymphoma (42), diffuse large B-cell lymphoma (42), acute B-lymphoblastic leukemia (43–45), and CLL (46–49). These observations led to the development of CK2 inhibitors. One such inhibitor CX-4945, whose activity is being investigated in clinical trials, was found to possess antineoplastic activity in CLL (50). Ex vivo studies showed that a combination of fludarabine and CX-4945 significantly reduced CLL cell viability (47) and that CX-4945 synergistically interacted with the Bruton tyrosine kinase inhibitor ibrutinib (47) in eliminating CLL cells. Because CK2 is present in all mammalian cells, the specificity of CK2 inhibitors might be a crucial limiting factor in the development of CK2 inhibitors as antineoplastic agents. In contrast, targeting the STAT3 phosphorylation complex might prove to be a specific and effective approach.

Z. Estrov is a consultant/advisory board member for Incyte. No potential conflicts of interest were disclosed by the other authors.

Conception and design: U. Rozovski, A. Ferrajoli

Development of methodology: D.M. Harris, P. Li

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): D.M. Harris, P. Jain, A. Ferrajoli, J. Burger, S. O'Brien, M.J. Keating

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): U. Rozovski, I. Veletic, A. Ferrajoli, W. Wierda, M.J. Keating, Z. Estrov

Writing, review, and/or revision of the manuscript: U. Rozovski, P. Jain, A. Ferrajoli, J. Burger, S. O'Brien, P. Thompson, N. Jain, W. Wierda, M.J. Keating, Z. Estrov

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Z. Liu, P. Jain, P. Bose

The authors thank Sarah J. Bronson for editing our manuscript.

This study was supported by a grant from the CLL Global Research Foundation and the Cancer Center Support Grant from the NIH/NCI, P30 CA016672.