Regulation of Mouse CK2α (Csnk2a1) Promoter Expression In Vitro and in Cell Lines

Abstract

CK2α is a kinase important for essential cellular and biological processes. CK2α is ubiquitously expressed, albeit at different tissue levels, and its transcript levels are dysregulated in disease. However, there is limited knowledge on the regulation of CK2α gene expression. The best one studied, the human CSNK2A1 (CK2α) gene promoter, contains uncharacterized binding motifs for NF-κB. Our goal was to investigate the role of NF-κB in Csnk2a1 promoter regulation. We cloned the mouse Csnk2a1 promoter which had significant sequence homology with the human CSNK2A1 promoter. Using promoter deletions, we identified a minimal promoter region containing transcription factor motifs (NF-κB, Ets-1, Sp1) consistent with those published for the CSNK2A1 promoter. Electrophoretic mobility shift assays demonstrated specific NF-κB subunit binding to the minimal promoter. NF-κB subunit transfection and extracellular NF-κB stimulation in non-tumor cell lines led to increased transactivation of the mouse minimal promoter. These data, together with data on the regulation of NF-κB by CK2 kinase activity, suggest a positive-feedback loop between CK2α and NF-κB. Non-tumor cell line re-plating and increased percent confluence upregulated Csnk2a1 transcript levels which differed from tumor cell line published data. In summary, Csnk2a1 promoter is regulated by NF-κB signaling and during cellular proliferation.

1. Introduction

Protein kinase CK2 is a highly-conserved holoenzyme composed of catalytic subunits (CK2α and CK2α’) and a regulatory subunit (CK2β) [1,2,3]. There is also an intronless CK2α pseudogene (CSNK2A3) whose function is not well understood. There is significant literature showing the involvement of protein kinase CK2 proteins in various cellular and biological processes and in diseases. These include cell survival [4,5] and proliferation [6], viability of yeast [7], mammalian development [8,9], Xenopus dorsal axis formation [10] and oncogenesis [11,12,13,14]. Due to the importance of CK2 proteins, a number of studies investigated their endogenous expression. CK2 proteins are ubiquitously expressed however the protein levels differ between tissues [6,15,16,17]. There is also elevated CK2 gene expression and protein kinase CK2 activity levels in a variety of different cancer types (reviewed in Chua et al. [14]). Despite their importance, the mechanism(s) controlling expression of CK2 genes and their dysregulation in diseases are not yet understood.

Of the CK2 genes, the CK2α gene (CSNK2A1) has the most data on promoter regulation. The Pyerin laboratory has significantly contributed to the understanding of its genomic structure and expression. The human CSNK2A1 gene has features consistent with a housekeeping gene: TATA box-less, presence of GC boxes, and a CpG island around exon 1 [18]. The human minimal promoter region contains potential binding sites for the transcription factors Sp1, Ets-1, and NF-κB. These transcription factors play important roles in expression of cytokine and chemokine genes, cellular functions of lymphoid cells and angiogenesis (Ets-1), in the control of genes regulating cell growth, apoptosis, differentiation and immune responses (Sp1), and in the control of genes involved in inflammation, immunity, differentiation, cell growth, tumorigenesis, and apoptosis (NF-κB). Protein kinase CK2 also has roles in many of these cellular and biological processes.

The Sp1, Ets-1, and NF-κ B sites may be functional, as electrophoretic mobility shift assay (EMSA) analysis shows that proteins present in HeLa nuclear extracts bind specifically to CSNK2A1 gene promoter segments [19]. Importantly, mutation of these binding sites has a significant negative impact on promoter segment activity [19,20,21]. From their binding site mutation experiments, the Pyerin laboratory proposed a model where Sp1 and Ets-1 are part of a negative feed-back loop to regulate the promoter and proposed that NF-κB serves to fine tune CSNK2A1 expression. Nevertheless, the significance of NF-κB in CSNK2A1 promoter regulation has not been studied in vivo. In addition, we lack studies of CSNK2A1 promoter regulation in a species besides humans, studies utilizing non-cancer cell lines, and the effect of signaling pathways on promoter regulation.

The relationship between protein kinase CK2 and NF-κB is of particular interest to our research team as NF-κB is a well-known proto-oncogene understood to be involved in numerous cell functions, including cell growth, proliferation, anti-apoptosis, and metastasis [22]. Two distinct and interrelated signaling pathways have been proposed for NF-κB signaling activation. One is stimulated by developmental signals which leads to immune cell differentiation and maturation and secondary lymphoid organogenesis (non-canonical), and another is stimulated by inflammatory signals that leads to inflammation (canonical) [23]. Both require the inactivation of the NF-κB inhibitor IκBδ by degradation. The family of NF-κB transcriptional factors is comprised of different members: NF-κB1 (p50) and RelA (p65) which participate in both canonical and not canonical signaling, NF-κB2 (p52) and RelB which participate in non-canonical signaling, and c-Rel which participates in canonical signaling [24]. These five NF-κB monomers can yield up to 15 NF-κB family protein complexes, of which the most common is a p50/p65 heterodimer to induce or repress transcriptional activity [25,26]. In addition to its normal functions, NF-κB signaling is constitutively activated in some cancers which leads to increased proliferation, anti-apoptosis, angiogenesis, epithelial to mesenchymal transition, and metastasis [27].

The data presented here provide information regarding the regulation of mouse Csnk2a1 promoter regulation in NIH-3T3 (embryonic fibroblast cell line). We identified the gene promoter, a minimal promoter, and transcription binding motifs, and showed the binding of p50 in competitive binding assays. We determined the role of NF-κB signaling in Csnk2a1 transactivation utilizing NF-κB subunit transfections in NIH-3T3 and standard and physiological NF-κB signaling pathway agonists in mouse peripheral B lymphocytes. We propose a positive regulatory loop between CK2 and NF-κB. In contrast to studies in human cancer cell lines [20,21], Csnk2a1 transcripts increased shortly after re-plating and in conditions of increasing cell confluence in NIH-3T3 and C57MG (mouse mammary epithelia) cell lines.

2. Results and Discussion

2.1. Characterization of the Mouse Csnk2a1 Promoter

We utilized the Genome Reference Consortium Mouse Build 39 (GRCm39) to identify the Mus musculus Csnk2a1 gene. Csnk2a1 is placed at bases 152,068,468 to 152,123,772 of chromosome 2. We first compared the mouse and human sequence homology upstream of mouse and human exon 1. The Homo sapiens CSNK2A1 is placed at bases 472,498 to 543,790 (complement) on chromosome 20 in the Genome Reference Consortium Human Build 38 patch release 14 (GRCh38.p14). We used mVISTA to visualize the sequence homology of 1000 bp sequences upstream and including mouse and human exon 1 (Figure 1A). In this 1000 bp sequence alignment, there were several regions with significant sequence homology. Using SIM, the overall similarity of the two sequences was found to be 62.7%. Note that exon 1 is untranslated in both species (153 bp in mice and 119 bp in humans).

We analyzed the promoter sequences to identify similarities between both species. Exon 1, the transcriptional start site1 (TSS1), is untranslated in both species (153 bp in mice and 119 bp in humans). Sequence alignment showed that the last 119 bp of the mouse Csnk2a1 exon 1 has an 89.9% similarity with the 119 bp human CSNK2A1 exon 1 (Figure 1B). The sequence of the first 34 bp of mouse exon 1 has a 91.2% similarity to the sequence upstream of the TSS1 in humans (Figure 1B).

Review of mouse Csnk2a1 transcript variants listed in NCBI found that the transcriptional start site (TSS) was variable. Eight of the 13 transcript variants contained the 153 bp exon 1, which we utilized in this study. Two more variants contained the same exon 1 start, however with extended exon 1 sequences. The other three variants did not contain any portion of our selected exon 1, and two of these were significantly truncated transcript variants.

In mice, a putative TATA box with the sequence 5′-TATAA-3′ is found at position −90, which is likely non-functional, as most TATA-box motifs are located between positions −20 to −35 [28]. The human CSNK2A1 promoter was reported not to include a TATA box or CAAT box [18,29]. Using the last human genome compilation, we confirmed that there were no TATA box sequences upstream of human CSNK2A1 exon 1. Similarly, there is no CAAT box upstream of the mouse promoter. There is a CAAT box in the human promoter found at −117, most likely not functional, as CAAT box motifs are between −75 to −80 [28].

Taken together, the CSNK2A1 mouse and human promoters have similar characteristics indicating conservation of regulation cross-species.

2.2. Identification of the Mouse Csnk2a1 Minimal Promoter

To understand the regulation of the mouse Csnk2a1 gene, we next identified its minimal promoter region using similar methods to Krehan et al. [19]. To determine the mouse minimal promoter, we generated plasmids pLuc1 (−3955 to +153), pLuc2 (−735 to +153), pLuc3 (−453 to +153), pLuc4 (−273 to +153), pLuc5 (−273 to +31), pLuc6 (−183 to +153), pLuc7 (−135 to +153), pLuc8 (−95 to +153), and pLuc9 (−273 to −80) (Figure 2A). We transfected these plasmids into subconfluent NIH-3T3 cells along with Renilla to control transfection efficiency. The luciferase activity of the Csnk2a1 promoter deletion variants was then normalized to the activity of the full-length Csnk2a1 promoter, pLuc1 (Figure 2B). pLuc2-4 and pLuc6-8 maintain complete activity, suggesting that the region upstream from −735 is not essential for promoter activity. Welch’s t-test was performed on all pLuc constructs against pLuc1 and it was found that luciferase activity for pLuc4 (p = 0.002) and pLuc7 (p = 0.011) were significantly different from that of pLuc1. pLuc5 (−273 to +31) construct retains full activity, therefore the +31 to +153 (the majority of exon 1) region is not required for Csnk2a1 promoter activity. pLuc9 (−273 to −81), containing a deletion of the entire region downstream of the putative TATA sequence, exhibits a total loss of activity. With Welch’s t-test it was found that the activity of pLuc5 (p = 0.015) and pLuc9 (p = < 0.001) was significantly different from that of pLuc1. This suggests that the minimal functional mouse Csnk2a1 promoter must span or be contained within −95 to +31. Further studies could use additional deletion constructs to investigate whether smaller sequences upstream of exon 1 retain promoter function. The results also suggest that the region upstream of the minimal promoter (−3955 to −735) may contain a transcriptional silencer(s), as pLuc1 has less activity than the 5′ deletions, although this needs to be investigated. The region upstream of the minimal promoter (−735 to −95) may also contain a transcriptional silencer(s) given that pLuc1 through pLuc7 have decreased luciferase activity relative to pLuc8. However, it should be noted that this observation is not significant when we compared the activity of all pLuc constructs to that of pLuc8 using Welch’s t-test.

Next, we investigated the homology between the mouse and human minimal promoters. We performed sequence alignment using our identified mouse Csnk2a1 minimal promoter and the Krehan et al. human CSNK2A1 minimal promoter (Figure 3). Krehan et al. identified the human CSNK2A1 minimal promoter in a region comprising positions −89 to 15 using the current position for exon 1, with a region of maximal promoter activity located from positions −59 to −4 (Note: Krehan et al. identified two transcription start sites (TSS1 and TSS2); in this work, we use the current start of human CSNK2A1 exon 1 which is TSS2 by Krehan et al.) [19]. The 126 bp mouse Csnk2a1 minimal promoter and 104 bp human CSNK2A1 minimal promoter, when aligned, have a region of overlap of 85 bp with 79.2% similarity as determined by SIM. Contained in this 85 bp region of overlap is 55 bp of the 56 bp region of maximal activity identified by Krehan et al [19]. Looking just at this region of maximal activity, there is 80.7% similarity between the mouse and human promoter sequences.

2.3. Transcription Factor Binding Analysis of the Minimal Csnk2a1 Promoter

The mouse Csnk2a1 and human CSNK2A1 minimal promoter sequences were analyzed to predict transcription factor binding via the TRANSFAC R3.4 database utilizing TFBIND. The analysis generated the similarity to the consensus binding sequence, and we used 0.8 as the similarity cut-off to generate the list of potential transcription binding sites. This analysis identified potential binding sites for SP1 (91.4% similarity in mice, 95.1% similarity in humans), NF-κB p50 (83.9%, 91.4%), NF-κB p65 (85.4%, 96.2%), C-Rel (92.1%, 98.1%), Elk1 (93.0%, 85.3%), Ets-1 (96.4%, 96.4%), and Nrf2 (93.7%, 93.7%) (Figure 3). These findings largely agree with the transcription factors identified by Pyerin et al. (NF-κB, Ets-1, Sp1) as having a potential role in the human CSNK2A1 promoter [19]. Supporting the role of Ets-1, a study in hepatocellular carcinoma cells showed that the histone deacetylase inhibitors AR42 and MS-275 increased CK2α transcription and enhanced the recruitment of Ets-1 to the CK2α promoter. This occurred without altering the expression of Ets-1 [30]. Future studies could investigate whether Sp1 and NF-κB are also recruited to the Csnk2a1 promoter under these and other conditions.

In addition, the TFBIND analysis showed the potential for additional regulation of the Csnk2a1 promoter by Elk1 and Nrf2. Elk1, a subclass of the Ets family, is involved in cell proliferation, differentiation and survival, hematopoiesis, angiogenesis, cognition, neurodegeneration, and cancer. Elk1 is ubiquitously expressed and highly expressed in cortex and lymphoblasts [31]. Nrf2 is also ubiquitously expressed with highest expression in blood cells, thyroid, smooth muscle, and bronchial epithelial cells. Activation of Nrf2 transactivates cytoprotective proteins [32].

In summary, the observation that the consensus sites for these transcription factors (NF-κB, Ets-1, Sp1) are preserved in the minimal promoters of both mouse and human further strengthens the importance of these transcription factors as regulators for Csnk2a1.

2.4. NF-κB Subunits Bind the Csnk2a1 Promoter

Here we focus on NF-κB interaction with the Csnk2a1 promoter as it is understudied. The NF-κB (c-Rel) site is preserved in both mouse and human minimal promoters; however, the mouse sequence has a one base difference within the consensus (Figure 3). This is noteworthy as the consensus binding site for NF-κB (c-Rel) is loose (SGGRNWTTCC, where S = C or G, W = A or T, R = A or G, Y = C or T, K = G or T, M = A or C, N = any base pair) [33].

To test a direct interaction between NF-κB and the mouse Csnk2a1 minimal promoter, EMSA was performed with a double-stranded oligonucleotide probe overlapping the putative NF-κB in the Csnk2a1 minimal promoter. Nuclear extracts were prepared from NIH-3T3 untransfected cells and cells transfected with p50 and c-Rel subunits. Potential specific bands, denoted by arrowheads, were observed (Figure 4). To assess the specificity of these bands, a competitor NF-κB-specific oligonucleotide encoding the c-myc upstream response element (URE) was introduced at concentrations of 5× and 50×. An excess of the URE should outcompete specific binding of the Csnk2a1 promoter oligonucleotide to genuine NF-κB transcription factors resulting in eliminating bands in the EMSA. Furthermore, as NF-κB signaling often involves a p50 homodimer or heterodimer, a supershift analysis was conducted using a p50 antibody. Any shift up to a slower mobility complex or alterations in the density of the shifted bands would indicate the presence of p50 in the transcriptional complex.

The band designated by the black arrowhead intensifies in the presence of the p50 supershifting antibody and disappears when challenged with a 50× competing URE oligonucleotide, indicating its specificity for NF-κB. The band designated by the white arrowhead disappears with the p50 antibody and is eliminated with both 5× and 50× competitor URE oligonucleotide. This suggests that the bands marked by the white and black arrowheads could potentially be p50 homodimers or heterodimers. The band designated by the gray arrowhead exhibits a shift with the p50 antibody, confirming the presence of p50. However, this band darkens with the competitor URE oligonucleotide and is not competed away. This observation could be attributed to p50 binding in a complex with another transcription factor and/or possibly acting on a non-canonical NF-κB site [34]. Here, we demonstrate for the first time that p50 interacted with the Csnk2a1 minimal promoter, potentially forming different NF-κB complexes. Based on the mouse and human minimal promoter homology we hypothesize that p50 is likely to interact with the human CSNK2A1 minimal promoter as well. This fits well with data from the Pyerin laboratory on the CSNK2A1 minimal promoter [19]. Oligonucleotides containing the CSNK2A1 gene promoter segment that contains the NF-κB binding sequence can bind p65 (an NF-κB subunit)) found in HeLa nuclear extracts. Electrophoretic mobility shift assay (EMSA) analysis show that proteins present in HeLa nuclear extracts bind specifically (reduced protein binding in the presence of oligonucleotides with the consensus motif mutated) to the CSNK2A1 gene promoter segment that contains the NF-κB binding sequence [19,20,21].

2.5. Csnk2a1 Transcription Is Upregulated by NF-κB Subunit Expression

The Pyerin studies, which proposed NF-κB to have a role in transcriptional regulation of the human CSNK2A1 minimal promoter, also demonstrated that point mutations (3 bases at a time) of the NF-κB consensus sequence reduced promoter activity. Our laboratory and others have previously studied biological relationships between protein CK2 and NF-κB in cancer, which together indicate that protein kinase CK2 and NF-κB interact in several signaling pathways [13]; therefore, we investigated the role of NF-κB in regulating the Csnk2a1 promoter.

We first tested whether NF-κB regulates the transcription of the Csnk2a1 promoter by transfecting NIH-3T3 with NF-κB subunits and analyzed the changes in luciferase activity of Csnk2a1 promoter constructs. For this, plasmid pLuc8, containing only the mouse minimal promoter and exon 1, was co-transfected with different combinations of NF-κB subunits (p50, p65, RelB) into NIH-3T3 cells at 30% confluence and subsequently harvested 48 h later. Bcl3 was added to some of these transfections as it has been shown to decrease repressive p50 and p52 homodimers and allow transcriptional activation from other NF-κB family complexes [35,36]. For these experiments, Renilla proved to be an inadequate control as we observed that NF-κB subunit transfection enhanced pRL-SV40 Renilla expression, potentially through NF-κB binding sites in its promoter. Therefore, luciferase activity was normalized to the respective protein concentration in each lysate. Cells were co-transfected with GFP to confirm equal transfection efficiency.

For the first series of co-transfections, 200 ng of pLuc8, was co-transfected in duplicate with 4 μg of each the NF-κB subunits (p50 + RelB, p50 + p65, p50 + Bcl3), and pCDNA3 was used as a negative control. When normalized to the activity of the pCDNA3 control, the luciferase activity of each co-transfection was as follows: p50 + RelB, 2.97-fold; p50 + p65, 2.20-fold; p50 + Bcl3, 2.40-fold.

A second series of co-transfections was performed by co-transfecting 200 ng pLuc8 in triplicate with 3 μg of the following NF-κB subunit combinations: p50 alone, p50 + cRel, p50 + RelB, p50 + p65, p50 + Bcl3, Bcl3 + p52 (Figure 5). Relative to the activity of pCDNA3, the luciferase activity of all co-transfection increased. It is interesting that the p50+p50 transfections led to increased activity of the Csnk2a1 promoter construct, as the p50 homodimer is thought to repress transcription. However, we did not confirm that the p50+p50 transfections actually led to formation of p50 homodimers. It is possible that these p50 subunits instead formed heterodimers with other members of the NF-κB family which induce transcription. Indeed, it was demonstrated in the past that p50 preferentially forms heterodimers [37,38]. In summary, diverse combinations of NF-κB subunits increased transactivation of the Csnk2a1 promoter construct. This is an important new finding, suggesting that NF-κB has a positive effect on Csnk2a1 transcriptional regulation.

The levels of Csnk2a1 promoter activation by NF-κB shown here (1.5 to 1.8×) are larger than the levels of CSNK2A1 promoter activation by transfection of Ets-1 in HeLa cells (1.25×) [20]. This was unexpected, as Pyerin demonstrated that mutations of Sp1 and Ets-1 binding sites reduced CSNK2A1 transactivation to a much larger extent (70–75%) than mutation of NF-κB binding sites (40%), and from this posited that NF-κB may serve to fine-tune CSNK2A1 expression [19]. Nonetheless, our data here indicated that NF-κB could be a significant driver of Csnk2A1. Conservation of NF-κB motifs between the mouse and human CSNK2A1 minimal promoter further suggests its importance in regulation. The difference between the degree of upregulation between our NF-κB data and Pyerin’s Ets-1 data could be explained by differences in the cells used (non-cancer vs. cancer cells) [19]. Human cancer cells may have almost maximal expression of CSNK2A1, diminishing the effect of Ets-1 on promoter upregulation, whereas NIH-3T3, a non-cancer cell line, may have low basal Csnk2a1 expression, allowing NF-κB subunit transfection to induce stronger Csnk2a1 promoter activation. The difference may be also due to differences in the length of the promoter sequences utilized (minimal versus or not) or in transfection efficiency. Future experiments could compare the activation of the different promoters in parallel using diverse cell lines to address the contribution of the different transcription factors in promoter regulation.

2.6. Regulation of Endogenous Mouse Csnk2a1 Promoter by Extracellular NF-κB Agonists

Our experiments so far showed that NF-κB subunits bound and increased transcription of the mouse Csnk2a1 minimal promoter. We next determined whether activation of NF-κB signaling by extracellular agonists could upregulate endogenous Csnk2a1 promoter. The expectation is that if NF-κB upregulates Csnk2a1 promoter transcription, then activation of NF-κB signaling should lead to increased levels of endogenous Csnk2a1 mRNA transcripts. NF-κB signaling pathway stimulants (i.e., they signal via NF-κB family protein complexes) include phorbol myristic acid + ionomycin, IgM, CD40 and CD8a, and lipopolysaccharide which we use in this study [39,40,41,42].

For this study, mouse peripheral B lymphocytes were chosen for their ability to activate the NF-κB signaling pathway in a stimulable manner [39]. To stimulate the NF-κB signaling pathway, the following chemical and physiological growth stimulants were chosen: PMA (phorbol myristic acid) + I (ionomycin) at two different concentrations (Snow protocol and Roth protocols described in the Methods Section) [39,43], lipopolysaccharide [42,44,45], anti-mouse IgM [40,41,46], and CD40 ligand + CD8a [47]. Following B lymphocyte stimulation, RT-qPCR analysis was used to measure Csnk2a1 copy number which was normalized to endogenous β-actin (Actb).

For the first series of B lymphocyte stimulation experiments, PMA + I was used to stimulate rested B lymphocytes and samples processed after addition (0) and after 4, 20, 32, and 48 h. PMA + I resulted in significant early upregulation of Csnk2a1 mRNA at the four-hour time point (three-fold over time 0) (Figure 6A). From 24 h onwards, Csnk2a1 copy number was significantly downregulated, suggesting negative feedback.

Next, physiological agonists anti-IgM, LPS, and CD40 ligand + CD8a were used and compared with two different PMA + I stimulation protocols (Snow and Roth). Cells were collected immediately after stimulation or after 2.5 or 5 h (Figure 6B). One “P + I Snow” five-hour time point was found to have poor dissociation curves and unusable qPCR data and therefore is not included. All the agonists increased Csnk2a1 mRNA levels at 2.5 h which was maintained at 5 h. The increase in Csnk2a1 mRNA levels associated with each stimulus at 2.5 h ranged from a 1.5- to almost 4-fold increase, with the greatest increase found in PMA + I and anti-IgM. Lastly, we confirmed that our chosen chemical or physiological stimulants increased NF-κB signaling by evaluating their stimulation of IκBα transcription (Nfkiba) (Figure 6C). Nfkiba was chosen as its transcription is upregulated as part of a negative feedback mechanism in the NF-κB signaling pathway [48].

In a previous study, DeBenedette and Snow demonstrated that protein kinase CK2 activity exhibits a biphasic increase (peaking at 18 and 48 h) in primary B cells upon NF-κB signaling pathway stimulation with PMA + I [39]. They determined that the increased protein kinase CK2 activity levels resulted from stimulation with the polyamine ODC. The increase in protein kinase CK2 activity levels that they observed could be due to increased levels of CK2 protein and/or Csnk2a1 mRNA. Here, we show that the stimulants PMA + I, LPS, anti-IgM, and CD40 ligand + CD8a resulted in increased NF-κB signaling and increased Csnk2a1 mRNA levels. This increase in Csnk2a1 mRNA levels by NF-κB signaling pathway stimulation provides a possible explanation for the increase in protein kinase CK2 activity shown by DeBenedette and Snow. Future experiments could determine correlations and/or parallels between protein kinase CK2 protein and activity levels and increased mRNA levels upon NF-κB signaling.

2.7. Proposed NF-κB/CK2α Positive Feedback Loop

In normal B cells, basal NF-κB activity is limited by sequestration in a complex with inhibitor IκBα (Figure 7) [49]. Stimulation of the NF-κB pathway releases NF-κB from this complex to transactivate genes, such as Csnk2a1. NF-κB activation also induces the expression of IκBα, thereby resulting in feedback inhibition of the NF-κB pathway. Protein kinase CK2 has been well-established as a positive regulator of NF-κB signaling (Figure 7, blue arrows) [13]. CK2 phosphorylates p65, thereby activating it to enhance NF-κB signaling (Figure 7) [50,51]. CK2 also inactivates IκBα via phosphorylation [50,52,53]. The findings presented in this study suggest a potential positive feedback loop between NF-κB and CK2α, at least in normal B cells. In our proposed model, NF-κB stimulation of CK2α may have a buffering effect through inhibition of IκBα by preventing immediate signaling suppression, thereby sustaining NF-κB signaling (Figure 7).

Additional evidence of a positive self-regulation loop implicating protein kinase CK2 activity in its own expression comes from the study from Lupp et al. demonstrating that expression of CK2α or CK2α’, but not CK2β, are sufficient to transactivate the CK2α and CK2β promoters. In addition, Lupp et al. show that CK2 activity is necessary for CK2α and CK2β promoter activity as 90% inhibition of CK2 kinase activity resulted in a decrease of CK2α and CK2β promoter activity by 50% and 25%, respectively [54]. The mechanism involved in this positive self-regulation loop has not been determined; nonetheless, several recent lines of evidence show that CK2α does not bind to the promoter of CK2α or CK2β, including chromatin immunoprecipitation assays [54] and the yeast-based one-hybrid system [55]. Based on our data, it is plausible that Lupps’ positive self-regulation loop is due to the expressed CK2 proteins regulating NF-κB activity.

As CK2α and NF-κB are both known proto-oncogenes that contribute to cellular transformation and the development of cancer, the proposed CK2α-NF-κB positive feedback model has significant implications in cancer biology [13,53,56,57,58]. Thus, upregulation of Csnk2a1 or constitutive NF-κB signaling could initiate a perpetually activated positive feedback loop resulting in increased protein kinase CK2 activity and NF-κB signaling. To test this proposed model, future experiments could investigate whether NF-κB signaling activation correlates with increased levels of CK2α transcript, protein, and activity. Additionally, pharmacological blockade of NF-κB or non-degradable IκB super-repressor constructs could be employed to demonstrate that NF-κB signaling is indeed necessary for the early Csnk2a1 transcriptional increase observed upon NF-κB signaling activation. Lastly, pharmacological blockade of CK2 at different times during NF-κB signaling activation can be used to show that CK2 is needed to inhibit the IκB feedback-loop inhibition.

The proposed sustained elevation of CK2α and NF-κB signaling aligns closely with observations in cancerous cells, where co-overexpression of both CK2α and NF-κB has been observed. For example, mouse mammary gland cancers and human breast cancers that had constitutive NF-κB expression were also found to have elevated protein kinase CK2 levels [52,59,60]. This feedback model could also explain the ability of Csnk2a1 to induce lymphoma when driven by a µ heavy chain enhancer and promoter in transgenic mice [61]. It is important to note that the resulting lymphomas required a “second hit” to drive cancer proliferation. Similarly, CK2α overexpression in mouse mammary glands leads to late tumor formation, at a median age of almost 2 years [62]. Therefore, alterations in Csnk2a1 expression may be an important factor in carcinogenesis but may not be sufficient to drive cancer proliferation on its own as reported in other studies for CK2α’ [63].

In contrast, a negative feed-back mechanism was proposed for Ets-1 and Sp1 by Pyerin’s team. Notably, protein kinase CK2 phosphorylates Sp1, and potential CK2 phosphorylation sites have been identified for Ets-1 [19,50,51,64,65]. Protein kinase phosphorylation of Sp1 decreases its binding to DNA, providing a plausible mechanism for negative self-regulation of CK2 gene transcription [64]. Both the negative and positive feedback mechanism could contribute to precise regulation of the CK2α promoter. Other signaling pathways can also regulate the CSNK2A1 promoter. Indeed, Csnk2a1 levels are upregulated in mouse mammary gland cells transfected with Wnt-1 [66], and in human breast cancer cells exposed to estrogen [67].

2.8. Re-Plating and Percent Confluence Increased Csnk2a1 Transcript Levels

We examined changes in expression levels of Csnk2a1 transcripts in non-cancer cell lines. For this we investigated Csnk2a1 transcriptional regulation in NIH-3T3 upon different conditions, such as cell re-plating and cell confluence.

For cell re-plating, NIH-3T3 cells were first grown to 40% confluence. New media was then added, and cells were grown for 10 h. The media was removed, and cells were trypsinized, washed with PBS two times, centrifuged at 1000 rpm for 6 min, and then resuspended in the original media. Cells were collected for RNA analysis both before trypsinization and at 120, 180, 285, 405, and 510 min after re-plating. RT-qPCR was used to evaluate the copy number of Csnk2a1 mRNA which was normalized to Actb mRNA (Figure 8A). Re-plating led to a transient rise in Csnk2a1 mRNA levels after around 120 min. This transient upregulation of Csnk2a1 mRNA levels after re-plating suggests that transient upregulation of Csnk2a1 is beneficial to cells when their adhesion is limited. We can draw parallels between these results and Giusiano’s findings of elevated levels of CK2α protein and protein kinase CK2 activity associated with increased metastatic risk [68]. Future re-plating experiments could use CK2 chemical inhibitors, inducible RNAi/siRNA, and inducible CK2 expression to investigate whether the transient upregulation of CK2 is essential for cell survival when adhesion is limited. In addition, parallel studies on cells replated at varying confluences could help determine whether the alterations in Csnk2a1 mRNA are attributable to the initial adhesion or if they reflect specific confluence levels.

To study the effects of increasing cell confluence on Csnk2a1 transcription, NIH-3T3 and C57MG cells were plated at a very low cell density. These cells were harvested, and the RNA was extracted at increasing levels of confluence: 30%, 50%, 100%, after 8 h at 100% confluence, and after 24 h at 100% confluence. RT-qPCR was used to evaluate the copy number of Csnk2a1 mRNA which was normalized to Actb (Figure 8B). In both NIH-3T3 and C57MG cells, levels of Csnk2a1 mRNA increased with cell confluence. Next, we investigated whether this effect is mediated by the Csnk2a1 minimal promoter. For this, the pLuc8 plasmid was transfected into NIH-3T3 at 25% confluence on a plate and allowed to grow 24 h. The cells were then replated at 5–10% confluence. These cells were then harvested for luciferase assay and measurement of protein concentration at increasing levels of confluence: 10%, 50%, and 100% confluence. Due to its 3 h half-life in mammalian cells, the luciferase protein does not accumulate over time. As such, luciferase activity does not reflect an accumulation of protein but rather an increase in Csnk2a1 minimal promoter activity. Increasing cellular confluence was associated with an increase in Csnk2a1 minimal promoter activity and resulting luciferase activity (Figure 8C). Both the RT-qPCR and luciferase reporter assays showed that Csnk2a1 promoter activity increases with increasing cell confluence. This may have implications for cancers in which tumor cells operate at a steady state of “confluence” due to crowded tumor microenvironment. For these “confluent” tumors, Csnk2a1 transcription may be upregulated, potentially leading to further transformation.

Previous studies showed that the expression levels of CSNK2A1 transcripts do not change at different confluence percentages in HeLa and JEG-3 cancer cell cultures [21]. The difference with our data, as we discussed above, could be that cancer cells may have almost maximal expression of CSNK2A1, diminishing the effect of proliferation on promoter upregulation. Future experiments could address the mechanism underlying this difference. It is plausible that in these proliferation experiments, NF-κB is at least partially responsible for the increase in Csnk2a1 transcription. In HEK/293 cells, human embryonic kidney fibroblasts, NF-κB-signaling reporter increases with increased confluency [69]. However, once cells are confluent, NF-κB signaling reporter activation decreases in HEK/293 cells. Therefore, it is plausible that NF-κB stimulation of Csnk2a1 transcription is strongest during proliferation, and at 100% confluence, other transcription factors have greater control on Csnk2a1 expression in normal cells.

An overall limitation in this study is that CK2α protein levels were not assessed due to the processing of cell lines to test for luciferase levels. Future experiments could build on this study by investigating levels of Csnk2a1 mRNA, luciferase, CK2α protein levels, and protein kinase CK2 activity at varying degrees of confluence for all the experiments in this work.

3. Materials and Methods

3.1. Bioinformatics Analysis

Ensembl was used to identify the mouse (ENSMUST00000099224.10) and human (ENST00000217244.9, RefSeq Match NM_177559.3) CK2α gene transcripts. To align and visualize the mouse and human promoter sequences, NCBI BLAST+2.15.0 alignment software [70], mVISTA v2.0 (https://genome.lbl.gov/vista/index.shtml), SIM (https://web.expasy.org/sim/, accessed 1 December 2023), and EMBOSS Needle 6.6.0.0 (https://www.ebi.ac.uk/jdispatcher/psa/emboss_needle) were used. To predict potential transcription factor binding sites within the minimal promoter, TFBIND (which queries the transcription factor database TRANSFAC R3.4) was used [71].

3.2. Cloning and Plasmid Constructions

For the plasmid construction described below, all the plasmids and oligonucleotide primers are presented in Supplementary Tables S1 and S2, respectively. All plasmid DNA was isolated from bacteria using the Qiagen Spin Mini Kit or Qiagen Plasmid Maxi Kit. For the isolation of plasmid fragments, as discussed below, restriction digested plasmid DNA was extracted from 1% agarose gel following the manufacturer’s instructions of the QIAquick Gel extraction kit protocol.

To generate the mouse Csnk2a1 promoter–luciferase constructs, a bacterial artificial chromosome (BAC) clone (clone # 34983798473) containing the full mouse Csnk2a1 genomic region was utilized, and fragments were cloned into pGL3-Basic, a plasmid that contains the luciferase gene Luc. We cloned a 4108 bp fragment into pGL3-Basic (construct named pLuc1), which included exon 1 (1 to 153) and 4108 bp upstream from the end of exon 1, as follows. We amplified by a 4108 bp fragment by PCR (−3955 to +153) using elongase enzyme (GibcoBRL Life Technologies, Waltham, MA, USA), a forward primer containing a MluI restriction endonuclease site and a reverse primer containing a NheI restriction endonuclease site. The amplified PCR product was extracted out of an agarose gel and digested with MluI and NheI, which were also used to digest the plasmid pGL3-Basic plasmid (Promega, Madison, WI, USA) which codes for the luciferase enzyme and ampicillin resistance. Following this, calf intestinal phosphatase was used to remove 5′ phosphate groups from the pGL3-Basic and Quick Ligase (New England Biolabs, Ipswich, MA, USA) was used to ligate the resulting linearized plasmid with the 4108 bp promoter fragment. Restriction digestion was used to screen for positives, and the resulting plasmid containing the 4108 bp mouse Csnk2a1 promoter and luciferase gene was designated pLuc1 (−3955 to +153).

Plasmid deletion variants of pLuc1 were created via additional restriction digests of the pLuc1 plasmid. To create the pLuc2 plasmid, pLuc1 was digested with MluI and EcoRI to remove the first 3221 bp of the promoter. T4 DNA Polymerase was then used to blunt the ends of this linearized plasmid, and it was re-ligated with Quick Ligase resulting in pLuc2 (−735 to +153), a plasmid containing an 888 bp Csnk2a1 promoter fragment. To create the pLuc3 plasmid, pLuc1 was digested with KpnI to remove the first 3503 bp of the promoter. Re-ligation with Quick Ligase yielded pLuc3 (−453 to +153), a plasmid containing a 606 bp Csnk2a1 promoter fragment.

To create the pLuc4 plasmid, pLuc3 was amplified via PCR using a single primer which can anneal both downstream (−273) and upstream (base 4801 of pGL3-Basic) of exon 1. The amplified DNA was then treated with methylation sensitive endonuclease DpnI to only digest parental DNA. Transformed E. coli colonies were screened by restriction digest with NheI and MluI. The resulting plasmid pLuc4 (−273 to +153) contained a 426 bp Csnk2a1 promoter fragment. To create the pLuc5 plasmid, pLuc4 was digested with PstI and NheI to remove the region downstream of +31. T4 DNA Polymerase was used to blunt the ends of this linearized plasmid, and it was re-ligated with Quick Ligase. This yielded plasmid pLuc5 (−273 to +31). pLuc5 was transformed into E. coli and the resulting colonies were screened by restriction digest with NotI and StyI.

Two-primer deletional PCR was used to create plasmids pLuc6, pLuc7, and pLuc8. To create the pLuc6 plasmid, pLuc1 underwent two-primer deletional PCR with primers. The forward primer binds at −184 and the reverse primer which binds at 4801 bp of pGL3-Basic. To create the pLuc 7 plasmid, pLuc1 underwent two-primer deletional PCR with the forward primer which binds at +47 and the reverse primer which binds at 4801 bp of the pGL3-Basic. To create the pLuc8 plasmid (−95 to +153), pLuc1 underwent two-primer deletional PCR with the forward primer which binds at −95 and the reverse primer which binds at 4801 bp of the pGL3-Basic. Following these amplifications, EcoRI was used to digest the amplified DNA, and it was then ligated with Quick Ligase. These plasmids were then transformed into E. coli and the resulting colonies were screened by restriction digest with NotI and StyI.

To create the pLuc9 plasmid (−273 to −80), pLuc4 underwent two-primer deletional PCR on pLuc4 with the forward primer which binds downstream of the primer insert in pLuc4, but upstream of the luciferase open reading frame, at base 55 of pGL3-Basic and the reverse primer which binds at −80. Following amplification, ApaI was used to digest the amplified DNA before ligation with Quick Ligase. pLuc9 was transformed into E. coli and the resulting colonies were screened by restriction digest with NotI and StyI.

For the transformations used to create the above plasmid deletion variants, either 1–5 ng of circular DNA or 50 ng of ligated DNA was introduced to competent E. coli DH5α or INVαF. The cultures were then incubated on ice for 30 min, heat shocked for 30 s at 42 °C and transferred back to ice to sit for 30 min. Five hundred μl of SOC media or LB broth was then added to the cultures before incubating on shakers at 37 °C for 1 h. Then, 100–300 μL of the transformation was incubated on agar plates supplemented with ampicillin for a period of 12–20 h. Subsequently, colonies were selected and cultured in LB broth with ampicillin.

3.3. Cell Culture and Transfections

All cell work was approved by the Boston University Medical Campus Institutional Safety Committee (protocol 14-1289). All transfections were performed utilizing the Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) standard protocol with a few modifications [72]. Adherent NIH-3T3 cells (untransformed mouse fibroblasts, #CRL-1658) were obtained from the American Type Culture Collection (www.atcc.org). C57MG (mouse mammary epithelia) cell line was obtained from Dr. Gail Sonenshein (Cat. # SCC273; Merck/Millipore, Billerica, MA, USA). Cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) (Cellgro, Manassas, VA, USA) at 37 °C and were seeded in 6-well plates without antibiotics at a concentration of 0.5–0.7 × 10⁴ cells per ml. After a 24-hour incubation period, the cells were suspended in 2 mL of fresh medium without antibiotics and replated. For each transfection sample, 4 µg of plasmid DNA was added to a 250 µL of Opti-MEM. A separate mixture of 250 µL of Opti-MEM and 10 µL of Lipofectamine 2000 was made and then mixed with the DNA mixture before incubating for 20 min to allow DNA-lipid complexes to form. This 500 µL solution was then added to each well containing the NIH-3T3 cells in 2 mL of fresh DMEM. The cells were allowed 24 h for transfection before the media was replaced. To control for transfection efficiency, simultaneous co-transfections of renilla (pRL-SV40, Promega) were performed for each sample using this same transfection protocol.

Following transfections, the Promega Dual Luciferase Reporter Assay System was utilized to quantify the activity of luciferase and renilla. First, the transfected cells were washed with phosphate-buffered saline (PBS) and then lysed with 250 µL of 1× Passive Lysis Buffer. To record luciferase activity, 10 µL of each sample was mixed with 50 µL of Luciferase Assay Reagent II in a polypropylene tube. Luminescence was recorded for 10 s using a luminometer. To record renilla activity, 50 µL of 1× Stop and Glo solution was added to the tube, and a second set of 10 s readings was performed.

3.4. Analysis of Gene Transcription

For the RNA copy number experiments, RNA extraction was performed as follows. Cultured cells were grown as described in the results (e.g., different confluences). Cells were washed with PBS and scraped, before being spun at 1000× g for 5 min at 4 °C. The pellet was resuspended in 1 mL of TRIzol Reagent (Invitrogen, cat # 15596-026) and homogenized by pipetting before incubating for 5–10 min at 15–30 °C. At this stage, cells were often frozen at −80 °C for later extraction. Following incubation, 0.2 mL of RNase-free chloroform was then added to the sample, and it was shaken vigorously for 15 s. The sample was then incubated at room temperature for 5 min before centrifugation at 12,000× g for 15 min at 4 °C. The clear RNA-containing supernatant was then pipetted to a new tube and RNase free isopropyl alcohol was added. Samples were then incubated at room temperature for 10 min before centrifugation at 12,000× g for 10 min at 4 °C to pellet the RNA. After discarding the supernatant, the RNA pellet was washed with 1 mL of RNase-free 75% ethanol and subsequently centrifuged for 5 min at 75,000× g at 4 °C. The supernatant was discarded and the RNA pellet was spun at low speed in a SpeedVac for 5 min before adding 30 µL of DEPC-treated, RNase-free water. RT was performed as described [73]

The PrimerExpress primer analysis program was utilized to select oligonucleotide primers for qPCR which would anneal at 59–61 °C (optimally 60 °C) and spanned an intron, when possible (Table S3). Ensembl mouse (GRCm39) and human (GRC38.p14) genomic sequences were used for primer selection. To verify primer sets, qPCR amplification was performed with the primers on cellular cDNA and plasmid DNA, if applicable. Only primers found to have distinct dissociation curve peaks and linear quantization of cDNA (over 10 two-fold dilutions) and plasmid DNA (over 10 two-fold dilutions) were used for experimental qPCR analysis.

For the qPCR reaction, 25 µL reactions were performed in the presence of 1× BIO-RAD SYBR Green Supermix, 400 nM of each primer, 2 µL of cDNA, and nuclease-free water. qPCR cycling was conducted with a Stratagene mx3000P real-time machine utilizing the SYBR green setting. The parameters for qPCR were: 3 min at 95 °C, followed by 40 cycles of 15 s at 95 °C and 1 minute at 60 °C, followed by 1 cycle of 30 s at 95 °C, 30 s at 60 °C, and an increase to 95 °C for dissociation curve.

Quantification of DNA and RNA was done via UV spectrophotometry at wavelengths of 260 nm and 280 nm with a Bio-Rad SmartSpec Plus spectrophotometer (Bio-Rad Laboratories, Hercules, CA, USA). Conversion ratios of 50 and 40 were applied for DNA and RNA, respectively. The 260/280 ratio served as an indicator for protein contamination (ratio < 1.8).

3.5. Protein Analysis

For protein extraction, cells were washed twice with cold PBS and then centrifuged at 3000 rpm for 4 min. The resulting pellet was washed again with PBS and then lysed in 50 µL of lysis buffer (50 mM Tris–HCl/pH 8.0, 125 mM NaCl, 1% NP-40, 10 mM NaF, 1 mM PMSF, 10 mM Na₃VO₄, l0 mM NaPPi, 5 µg/ml each of leupeptin, aprotinin, antipain and pepstatin) on ice for 10 min. After centrifugation at 12,000× g, the protein-containing supernatant separated and stored at −80 °C. To analyze protein concentration, the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA) protocol was used. 10 µL of serial dilutions of Bovine Serum Albumin (BSA) in PBS was added to 190 µL of Pierce BCA reagents (solution A to B at a ratio of 50:1). To determine sample protein concentration, the sample absorbance was measured at 570 μm via spectrophotometry (Bio-Rad SmartSpec Plus spectrophotometer) and compared to the BSA dilution standard curve.

3.6. EMSA

Electrophoretic mobility shift assays (EMSA) were performed using a ³²P-end-labeled fragment of the Csnk2a1 mouse promoter containing the c-Rel/NF-κB transcription factor binding site (covering bases 4 to 13 of mouse exon 1). To label the two oligonucleotides (5′ GCATTGGGAGGGGGCTTCCGCTTC 3′ and 5′ GCATGAAGCGGAAGCCCCCTCCCA 3′), 30 to 100 ng of oligonucleotides were boiled in 11 µL of water for 5 min and then rested on ice for 5 min. A mixture of 2 µL of hexanucleotide mix (dATP, dGTP, dTTP; Roche, Cat. No. 11 277 081 001), 3 µL of α³²P dCTP, and 1 µL of Klenow enzyme (Roche, Cat. No. 11 008 404 001) were added and the oligonucleotide probe was labeled for 30 min at room temperature. The oligonucleotides were annealed.

The 50 µL annealing reaction contained both primers at a final concentration of 1 mM each in the presence of 20 mM Tris-Cl pH 7.6, l0 mM NaCl, 2 mM MgCl₂, and 0.2 mM EDTA. The reaction was performed at 85 °C for 10 min and then at room temperature for 4 h to form the probe listed in Table S3. For isolation of this radiolabeled oligonucleotide probe, STE buffer (0.438 g NaCl in 50 mL TE pH 8.0) was added to the radiolabeling reaction to reach a total volume of 50 µL. One µL of this mixture was taken for sample quantification. The remaining 49 µL was pipetted onto beads in a pre-spun (735× g for one minute) ProbeQuant G-50 Micro Column (Amersham Biosciences, Piscataway, NJ, USA). After addition of the probe to the column, it was spun at 735× g for 2 min. The flow through volume was assessed and another 1 µL was taken for sample quantification. The pre and post-spin samples were each added to 1 mL of scintillation fluid (Fisher Scientific, Hampton, NH, USA) for measurement in a beta-scintillation counter. The scintillation counts were normalized to volume and compared to assess the incorporation percentage of radiolabeled nucleotide into the oligonucleotide probe.

In the gel shift binding reactions, 1.2 µg of nuclear extract was added to 5 µL of 5× sample buffer (5 µL of Triton X-100, 25 µL of glycerol, 16 µL of 250 M DTT solution, 100 µL of 100 mM HEPES pH 7.5, 845 µL of H₂O) 2.5 µL dIdC, and DR buffer (420 mM KCl, 20 mM HEPES pH 7.9, 1.5 mM MgCl₂, 0.2 mM EDTA, 20% glycerol, 1 mM DTT, 0.5 mM PMSF, 10 µg/ml leupeptin, and 1% NP-40). In total, there were 100,000 counts of radiolabeled probe in a total of 25 µL of reaction mixture. For the competition experiments, 1 µL or 5 µL of unlabeled upstream response element (URE) probe (5′ GATCCAAGTCCGGGTT TTCCCCAAC 3′), kindly provided by Dr. Gail Sonenshein, was also added. The binding reaction proceeded for 60 min at room temperature.

A 15 cm acrylamide gel (7.5 mL 37% Acrylamide, 5 mL l0× TBE, 39.8 mL H₂0, 35 µL TEMED, and 175 µL of 20% APS) was prepared and pre-run at 100 mV for 30 min. Samples were loaded, including a lane for the 6× loading buffer (also with DR buffer, dIdC, and SB buffer) and a lane for the free probe (also with DR buffer, dIdC, and SB buffer). The loaded gel was run at 200 mV for 3 hours. After, the gels were placed on Whatman paper and wrapped in saran wrap before drying for 1 hour in a gel drier. Gels were then exposed to film overnight at −80 °C.

3.7. Mice B Cell Isolation and Stimulation

All mice were maintained and managed at the Boston University Medical Center Animal Facility in uncrowded cages and husbandry was provided by the ASC (Animal Science Center) personnel. Animal protocol (AN-15183.2015.05) was approved in 2011 by the Boston University Medical Campus IACUC. Balb/c were euthanized using an acceptable method of euthanasia: they were anesthetized with isoflurane followed by cervical dislocation and vital tissue removal. To harvest splenic B cells, spleens were crushed and then cells were treated with anti-thymocyte globulins (gift from Dr. Thomas Rothstein) on ice for 40 min. To lyse any contaminating T cells, the cells were washed and treated with Rabbit complement. The cells were washed again and then layered onto 4 mL of warm Lympholyte-M. The cells were then spun at 1000–1500× g at room temperature, and as they spun the B lymphocytes became visible as a fuzzy white layer between the Lympholyte-M layer and the medium. The B lymphocyte layer was removed and washed twice.

To stimulate Balb/c mice splenic B lymphocyte growth, cells were added to RPMI (10% FBS, 2% L-glutamine, 1% penicillin/streptomycin) medium along with various chemical and physiological growth stimulants. For the first series of B lymphocyte stimulation experiments, 20 ng/ml of PMA and 0.7 ng ionomycin/ml were used to stimulate B lymphocytes. For the second series of B lymphocyte stimulation experiments, the following were used: 20 ng/ml and ionomycin at 709 ng/ml (Snow protocol [35]), 200 ng/ml and ionomycin at 600 ng/ml (Roth protocol [43]), 15 μg/ml lipopolysaccharide, 15 μg/ml anti-mouse IgM (AffiniPure (Fab’)2 Fragment Goat Anti-Mouse IgM(μ) Chain 1.3 μg/ml, Jackson ImmunoResearch Laboratories), and a 1:10 dilution of CD40 ligand followed by a 1:40 dilution of CD8a 10 min later. The CD40 and CD8a dilutions were provided by Rothstein using their protocol [37]. Notably, it has been thought that the use of anti-IgM to stimulate B lymphocytes may induce apoptosis after 24 hours, but these anti-IgM experiments only used it for 5 hours. Furthermore, trypan blue staining was used to ensure that B lymphocytes remained viable throughout the experiment.

As B cells were stimulated, RNA was collected immediately after stimulation or at other time points. RNA was extracted, RT performed and then qPCR was utilized to generate copy numbers of Csnk2a1, Actb, and Nfkbia. Levels of Csnk2a1 mRNA and Nfkbia mRNA were normalized to Actb mRNA. IκBα mRNA levels were measured as a positive control for NF-κB signaling pathway activation.

3.8. Statistical Analysis

Statistical analysis was performed for comparison of luciferase activity and gene copy number was performed with Welch’s t-test in R (4.4.2).

4. Conclusions

In summary, we showed that the mouse Csnk2a1 promoter sequence contains a minimal promoter sequence in the region 81 bp upstream from the start of exon 1. We showed for the first time that the NF-κB subunit p50 can bind directly to the Csnk2a1 minimal promoter. We showed for the first time increased Csnk2a1 transcription upon NF-κB signaling stimulation, either via NF-κB subunit transfection or via stimulation of B lymphocytes with NF-κB signaling agonists. These findings were used to develop a model for NF-κB and Csnk2a1 positive feedback regulation, one which may have significant implications for cancer biology. Finally, Csnk2a1 transcriptional regulation was shown to depend on the state of re-plating and confluence, which also may have implications for cancer proliferation and migration.